Investigation of low molecular weight Al complexes in human serum by fast protein liquid chromatography (FPLC)-ETAAS and electrospray (ES)-MS-MS techniques Tjasœa Bantan,a Radmila Milacœ icœ,*a Bojan Mitrovic�b and Boris Pihlarc aDepartment of Environmental Sciences, Jozœef Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia. E-mail: radmila.milacic@ijs.si bLek Pharmaceutical and Chemical Company d.d., Celovsœka 135, 1000 Ljubljana, Slovenia cFaculty of Chemistry and Chemical Technology, Asœkercœeva 5, 1000 Ljubljana, Slovenia Received 25th May 1999, Accepted 15th September 1999 Speciation of low molecular weight (LMW) Al complexes was performed in human serum from eight healthy volunteers in order to investigate the individual variability in the percentage and composition of LMW-Al species.Spiked samples (100±120 ng cm23 Al3z) were microultraÆltered through a membrane Ælter (cut-off 30 000 Da) to separate Al bound to transferrin from LMW-Al complexes.A 0.5 cm3 volume of the Æltrate was injected onto an anion-exchange fast protein liquid chromatography (FPLC) column and aqueous 4 mol dm23 NH4NO3 linear gradient elution was applied for 10 min to separate LMW-Al complexes. Fractions of 0.2 cm3 were collected throughout the chromatographic run and Al was determined `off-line' by electrothermal atomic absorption spectrometry (ETAAS). The characterisation of LMW-Al species in spiked serum was performed not only on the basis of the retention time (ETAAS detection), but also by electrospray (ES)-MS-MS analysis.A tandem quadrupole mass spectrometer equipped with a Z spray ion source as LC-MS interface was used for the identiÆcation of LMW ligands eluted under the chromatographic peaks. It was found experimentally that the amount of LMW-Al species in spiked serum ranged from 14 to 55%. On the basis of FPLC-ETAAS and ES-MS-MS analysis, it was found that the main LMW-Al species present in serum were Al-citrate, Alphosphate and ternary Al-citrate-phosphate complexes.The distribution of these species varied among particular individuals. In some of them Al-citrate and Al-phosphate were the main LMW-Al species in serum, while in others the ternary Al-citrate-phosphate complex was also present. The serum of some other individuals did not contain Al-phosphate and the main LMW-Al species were either Al-citrate and Al-citrate-phosphate complexes or Al-citrate species alone. The limit of detection for the separated Al species on the FPLC column was 5.0 ng cm23, while the RSD was found to be 8%.Introduction Aluminium toxicity, particularly in patients with chronic renal failure, is related to many clinical disorders. Its accumulation in the brain and bone is associated with dialysis encephalopathy1 and osteomalacia.2 It has been demonstrated that the gastrointestinal absorption of Al is greatly increased by the presence of low molecular weight (LMW) ligands,3,4 most intensively by citrate, although its role in Al accumulation in the body is not yet clear.5±8 It has been demonstrated that the dominant Al species in serum is Al-transferrin.9±14 The remaining Al is bound to LMW ligands, presumably citrate, phosphate and hydroxide.There are contradictory reports on speciation of LMW-Al complexes in serum.15±20 Martin21,22 and O» hman and Martin23 theoretically predicted that citrate is the most likely LMW binder of Al in human serum.Jackson24 calculated that 50% of Al in serum is bound to transferrin and the remaining 50% as Al(HPO4)OH, [Al(citrate)OH]2 and [Al(HPO4)citrate]22. Clevette and Orvig25 presumed Al-citrate to be the dominant LMW species in serum, while in the model of Harris26 it was predicted that 81% of Al is bound to transferrin and the remainder exists primarily as Al(PO4)(OH)2 with minor amounts of Al-citrate and Alhydroxide species. Another study20 proposed Al(OH)3 and Al(PO4) as the main LMW species in serum.A qualitative proton NMR study27 indicated that Al3z added to plasma ultraÆltrate was initially bound to citrate. Kiss and co-workers investigated the interaction of Al with phosphate28 and also the Al-citrate-phosphate interaction29 by potentiometry and 31P NMR techniques. It was found that at physiological pH the negatively charged binary species Al-citrate and Al-phosphate, as well as ternary Al-citrate-phosphate complexes, are present.29 For quantitative determination of LMW-Al species, microultraÆltration was used to fractionate high molecular weight (HMW) from LMW Al complexes in serum.30,31 Reported data indicate that 9±19% of Al in spiked pooled serum samples of healthy volunteers corresponded to ultraÆltrable LMW-Al species.Speciation of Al-citrate was also investigated by employing the HPLC-ETAAS method, but only moderate recoveries of Al-citrate were achieved.32 A study of the speciation of Al in biological Øuids by size-exclusion chromatography and ETAAS detection33 indicated the presence of two LMW fractions of Al in the spiked haemoÆltrate of uremic patients.In our group, a systematic study of Alcitrate by anion-exchange fast protein liquid chromatography (FPLC)-ICP-AES34 and anion-exchange FPLC-ETAAS31 was performed. It was demonstrated that LMW-Al present in spiked serum was quantitatively eluted under the peak of Alcitrate, 31 but the existence of this species was not deÆnitely proved.Since in addition to citrate, phosphate is also considered to be an important LMWligand for binding of Al in serum,20,24,26 the aim of our work was to provide more detailed information on the presence of Al species eluted under the chromatographic peak.31 For this purpose the characterisation of LMW-Al species in spiked serum was performed not only on the basis of the retention time, but also by electrospray (ES)-MS-MS analysis of theLMWligands eluted under the chromatographic J.Anal. At. Spectrom., 1999, 14, 1743±1748 1743 This Journal is # The Royal Society of Chemistry 1999peak. The study was performed on spiked serum samples from eight healthy volunteers in order to estimate individual variability in the percentage and composition of LMW-Al species. Experimental Instrumentation A strong anion-exchange FPLC column of Mono Q (Pharmacia, Uppsala, Sweden) was employed for the separation of negatively charged Al species. The column was connected to a Merck±Hitachi (Darmstadt, Germany) 6200 gradient highpressure pump equipped with a Rheodyne (Cotati, CA, USA) Model 7161 injector (0.5 cm3 loop).Total Al in human serum as well as the concentration of Al in separated species was determined by ETAAS on a Hitachi (Hitachi, Tokyo, Japan) Z-8270 polarized Zeeman atomic absorption spectrometer equipped with an autosampler at 309.3 nm. A Micromass (Micromass UK, Manchester, UK) Quatro LC tandem quadrupole mass spectrometer equipped with a Z spray ion source as LC-MS interface, employing negative electrospray ionisation, was used for the identiÆcation of LMW ligands in separated Al species.A Heraeus (Osterode, Germany) Model 17S Sepatech biofuge was used in the microultraÆltration procedure. Reagents Merck (Darmstadt, Germany) suprapur acids and water doubly distilled in quartz were used for the preparation of samples and standard solutions. All other reagents were of analytical-reagent grade.A stock Al3z solution (100 mg cm23 Al) was prepared in a 100 cm3 calibrated Øask by dissolving 0.1388 g of Al(NO3)3?9H2O (Riedel-de Hae»n, Hannover, Germany) in water. A stock Al-citrate solution (100 mg cm23 Al) was made weekly by mixing 0.0694 g of Al(NO3)3?9H2O and 3.5 g of citric acid (Merck) in a 50 cm3 Øask (100 : 1 citric acid to Al molar ratio). Imidazole (C3H4N2, 0.2 mol dm23) (Merck) buffer solution with the addition of an appropriate amount of hydrochloric acid (0.1 mol dm23) was used to adjust the pH of synthetic samples to 7.4.The 4 mol dm23 ammonium nitrate eluent was prepared by dissolving 320.16 g of NH4NO3 in 1 dm3 of water. Chelex 100 (Naz form, 100±200 mesh) chelating ion-exchange resin (Sigma, St. Louis, MO, USA) and a silica-based LiChrosorb -18 HPLC column (15064.6 mm id) were used for cleaning of reagents.31 Centricon 30 concentrators (Amicon, Witten, Germany) with a nominal cut-off of 30 000 Da were used in the ultramicroÆltration of human serum.Sample preparation In order to study the behaviour of LMW-Al complexes on an anion-exchange column at pH 7.4, synthetic solutions of LMW-Al species were prepared daily in 50 cm3 TeØon calibrated Øasks. A synthetic solution of Al-citrate (100 ng cm23 Al) was prepared by mixing 0.1 cm3 of stock Al-citrate solution in imidazole-HCl buffer solution with a pH of 7.4 (100 : 1 citrate to Al molar ratio). A synthetic solution of Al-phosphate (100 ng cm23 Al) was made by mixing 0.1 cm3 of stock Al3z solution and 0.1 cm3 of 2 mol dm23 H3PO4 in imidazole-HCl buffer solution (pH~7.4, 1000 : 1 PO4 32 to Al molar ratio).The Al to citrate and/or phosphate molar ratios were the same as in human serum. A synthetic solution of Al- ATP (100 ng cm23 Al) was prepared by mixing 0.023 g of adenosine triphosphate (ATP) (Merck) and 0.1 cm3 of stock Al3z solution in imidazole-HCl buffer solution (pH~7.4). The Al to ATP molar ratio was 1 : 100 to ensure complex formation.Blood from healthy volunteers was collected into Al-free Becton±Dickinson vacutainers without additives. Samples were centrifuged for 10 min at 3000 rpm. Serum was transferred into a TeØon Øask with a polyethylene pipette and was analysed within 12 h. Total Al was Ærst determined. In order to study the distribution of LMW-Al species, 2 cm3 of serum were spiked with 0.05 cm3 of Al3z solution, so that the Ænal concentration of Al in the spiked serum ranged from 100 to 120 ng cm23.Spiked serum was left to equilibrate for 4±6 h, after which it was microultraÆltered (cut-off 30 000 Da) to separate HMWfrom LMW-Al species. Speciation of Al was then performed in the serum Æltrate by the recommended analytical procedure. Recommended procedures Sample preparation, chromatographic separations and determination of Al by ETAAS were carried out under clean-room conditions (Class 10 000). To avoid contamination by extraneous Al, polyethylene or TeØon ware was treated with 10% HNO3 for 24 h, rinsed well with water and dried at room temperature.In order to lower the blank in FPLC separations, the NH4NO3 eluent and the FPLC columns were puriÆed by the cleaning procedure reported previously.31 A cleaning procedure was also applied to remove trace amounts of Al from the microultracentrifugation membranes of the Centricon 30 concentrators.31 Negatively charged LMW-Al complexes were separated on a Mono Q strong anion-exchange FPLC column.A 0.5 cm3 volume of sample was injected onto the column and aqueous (0±100% 4 mol dm23 NH4NO3) linear gradient elution was applied for 10 min at a Øow rate of 1 cm3 min21. Eluate was collected in 0.2 cm3 fractions and diluted to 0.5 cm3 with water in Eppendorf polyethylene cups. The concentration of Al was determined `off-line' by ETAAS under optimum measurement conditions as described previously.31 For the identiÆcation of LMW ligands eluted under the chromatographic peak, fractions were diluted 1z9 with water and analysed by the ES-MS-MS technique employing a Z spray ion source.A 20 mm3 volume of sample was injected into the Micromass Quatro LC mass spectrometer. The mobile phase was acetonitrile±0.005 mol dm23 ammonium acetate±formic acid (600z399z1, v/v/v). The electrospray probe voltage and sample cone voltage were set at 2.5 kV and 35 V, respectively. The source temperature of the mass spectrometer was held at 80 �C, while the desolvation temperature was 350 �C.The MS analyses were performed by scanning negative ions. The Q1- scan represented the pseudo-molecular ions (M2H)2 in the mass range m/z 50±1000. MS-MS collision-induced dissociation (CID) experiments were performed by introducing argon (2.061023 mbar) into the collision cell and setting a collision energy of 18 eV. The Ærst quadrupole analyser was set to transmit only the user-selected precursor ion. Results and discussion Distribution of Al-citrate, Al-phosphate and Al-ATP on an anion-exchange FPLC column A method for quantitative determination of Al-citrate in a wide pH range has been developed and validated by our group previously.31 The procedure was applied to the determination of LMW-Al species in spiked human serum (pooled sample).It was proved experimentally that LMW-Al species in serum were quantitatively eluted under the chromatographic peak which corresponded to Al-citrate.Since there are reports in the literature on the possibility of co-existence of aluminium phosphate and citrate species in the LMW serum fraction, the chromatographic peak was investigated more carefully. For this purpose, 0.2 cm3 fractions were collected and the behaviour of Al-citrate, Al-phosphate and Al-ATP at pH 7.4 was Ærst examined in synthetic solutions containing 100 ng cm23 of total Al. Separation of these species by 1744 J. Anal. At. Spectrom., 1999, 14, 1743±1748anion-exchange FPLC with ETAAS detection is presented in Fig. 1. It is evident that Al-citrate was quantitatively eluted from 3.0 to 3.8 min with two maximum peaks at 3.2 and 3.6 min, respectively. This indicates the presence of different negatively charged Al-citrate species, which is in agreement with the theoretical calculations of Martin.21 The Al-phosphate system is difÆcult to study because of Al-phosphate precipitation. However, at ppb concentration levels, the soluble species exist and it was found that 21°4% of Al-phosphate was eluted as negatively charged species from 2.2 to 3.2 min.The remaining 79% was strongly adsorbed on the resin column and did not disturb further separations. Data from Fig. 1 further indicate that 50% of Al-ATP was eluted from 2.8 to 3.2 min. The reproducibility of measurement was tested for six consecutive separations of Al-citrate, Al-phosphate and Al- ATP (100 ng cm23 Al, pH~7.4). The relative standard deviation (RSD) was found to be 2% for Al-citrate and Al- ATP, while for Al-phosphate it was 15%.The limit of detection (LOD) for determination of separated Al species on the FPLC column was found to be 5.0 ng cm23. IdentiÆcation of LMW-Al ligands by the ES-MS-MS technique In order to identify LMW-Al ligands, the ES-MS-MS technique using a Z spray ion source was applied. The conÆguration of this ionisation source permitted the analysis of samples with a high salt content, such as samples of eluted fractions after the chromatographic separation.First, mass spectra of standard solutions of Al-citrate, Al-phosphate and Al-ATP were recorded. The MS analysis was performed by scanning negative ions. Since ES is a `soft' ionisation technique very little fragmentation was observed and the most intense ion in the mass spectra of the investigated Al species was the deprotonated ligand ion (M2H)2. This precursor (parent) ion was selected for a further CID experiment and the product (daughter) ion mass spectra were recorded under the optimum conditions.The mass spectra and the corresponding MS-MS scans of Al-citrate, Al-phosphate and Al-ATP standard solutions are presented in Fig. 2. MS analysis of the blank indicated that m/z 81, 91, 108, 127, 137 and 154 corresponded to signals from the eluent used in the ES-MS procedure. It is evident (Fig. 2A) that in the mass spectrum of Al-citrate the peak with m/z 191 is the most intense and corresponds to deprotonated citric acid.This peak was selected as a parent ion for the CID experiment. After fragmentation, masses of 111, 87 and 85 were present in the resulting daughter ion mass spectrum. In the mass spectrum of Al-phosphate (Fig. 2B), a peak at m/z 97, which was selected as a parent ion for CID analysis, appeared. In the daughter ion mass spectrum, two peaks were present (m/z 97 and 79). The Al-ATP mass spectrum (Fig. 2C) resulted in a peak at mass 506.In the daughter ion mass spectrum of m/z 506, fragment ions of m/z 159, 177, 273, 408 and 426 appeared. The same standard solutions were also injected onto the FPLC coland ESMS- MS scans recorded for the separated fractions eluted under the chromatographic peaks. It was found experimentally that in the ES-MS scans additional peaks appeared (m/z 80, 103, 125, 142, 147, 171 and 188) which all resulted from the eluent of the chromatographic run (NH4NO3). ES-MS-MS peaks at m/z 191, 97 and 506 were the same as in standard solutions before chromatographic separation.Determination of LMW-Al species in spiked human serum In order to investigate individual variability in the percentage and distribution of LMW-Al species in serum, eight healthy volunteers were involved in the study. After sampling, total Al was Ærst determined by ETAAS using the standard additions calibration method. To reduce the matrix effects of proteins,35 5 ml of 32% nitric acid were added to the graphite tube before each determination and Al was determined under the optimum measurement conditions.31 The accuracy of the determination of total Al was checked by the determination of Al in Seronom‘ Trace Elements serum certiÆed reference material obtained from Nycomed Pharma.Good agreement between determined Al (61.4°0.3 ng cm23) and the reported certiÆed value (63°4 ng cm23) was obtained. The results for total Al concentrations in the serum of eight healthy volunteers are presented in Table 1.It can be seen that natural Al concentrations in serum ranged from 5 to 11 ng cm23. Caroli et al.36 reported that the normal value of Al in serum ranged from 0.5 to 8 ng cm23. Since the results in Table 1 are higher than the reported normal Al concentrations,36 the problem of contamination cannot be completely ruled out under the adopted working conditions. However, the natural Al concentrations were too low to perform speciation analysis.Therefore, the samples were spiked with Al3z solution, so that the Ænal concentration of Al in the spiked serum ranged between 100 and 120 ng cm23 Al. The concentrations of Al in spiked serum were similar to those that could be found in the serum of some haemodialysis patients. After equilibration (4± 6 h), serum was microultraÆltered (cut-off 30 000 Da) and the Æltrate injected onto the anion-exchange FPLC column. Speciation was performed as described under Recommended procedures.Fractions were collected throughout the chromatographic run and Al was determined by ETAAS. The results are presented in Table 2; data represent the average of three successive separations. It can be seen that the concentration of microultraÆltrable Al which was separated on the column ranged from 14 to 55% of the total Al in spiked serum samples. In the literature there are some reports on the determination of microultraÆltrable Al in spiked human serum.30,31 Data indicate that the amount of microultraÆltrable Al ranged from 10 to 20%.Since these studies were performed on pooled serum samples it was not possible to follow individual variability. In the present study, data from Table 2 further indicate that three LMW-Al species are separated on the column and that the distribution of these species varies among particular individuals. The Al species that was eluted from 0.8 to 1.2 min represents 3±37% of the LMW-Al species separated on the column.On the basis of our previous investigations,31,34 the presence of Al(OH)4 2 species, which at higher pH values was eluted at the same retention time, can be presumed. However, on performing mass spectrometric analysis it was not possible to identify Al(OH)4 2 species, nor other Al binding ligands with a mass up to 1000. Al species that were eluted from 2.4 to 2.6 min have the same retention time as Al-phosphate while Al-species eluted from 3.0 to 3.4 min appeared at the retention time of Al-citrate.In order to identify the Al binding ligands, ES-MS-MS analysis was performed on each separated fraction containing Al. Typical mass spectra and the corresponding MS-MS scans for Sample No. I are presented in Fig. 3. From the data of Fig. 3 it is evident that in the mass spectrum of the fraction eluted from 2.4 to 2.6 min a Fig. 1 Typical chromatograms of Al-citrate, Al-ATP and Al-phosphate (100 ng cm23 Al) at pH7.4. Separation was performed on an anion-exchange FPLC Mono Q HR 5/5 column and separated species were detected by ETAAS.Sample volume, 0.5 cm3; aqueous NH4NO3 (4 mol dm23) linear gradient elution; Øow rate, 1 cm3 min21; fraction collection, 0.2 cm3; n~3. J. Anal. At. Spectrom., 1999, 14, 1743±1748 1745characteristic peak at m/z 97 is present. In the daughter ion spectrum two masses with m/z 97 and 79 were observed, which conÆrmed the presence of phosphate as a binding ligand. This is in agreement with the Ændings of Kiss and co-workers,28,29 Jackson,24 Harris26 and Dayde et al.,20 who, on the basis of computer-aided speciation studies, predicted that phosphate is also an Al binding ligand in serum.The amount of Alphosphate in the LMW-Al fraction of serum sample No. I was 21%. The mass spectrum of the fraction eluted from 3.0 to 3.2 min indicates the presence of a characteristic peak with m/z 191. The corresponding daughter ion spectrum with characteristic masses of 111, 87 and 85 conÆrmed that the binding ligand was citrate.The same spectra were found for the fraction eluted from 3.2 to 3.4 min. The total amount of Al-citrate in sample No. I represented 55% of the LMW-Al species present in the serum. On the basis of the retention times and ES-MS-MS analysis it was found that in sample Nos. II and III, the Al species eluted on the column corresponded to Al-phosphate and Al-citrate. The amount of Al-phosphate in the LMW-Al fraction was 61% for sample No.II and 40% for sample No. III, while for Al-citrate it was 29 and 44%, respectively. In Fig. 4 typical mass spectra and the corresponding MS-MS scans are presented for sample No. IV. It is evident that in the mass spectrum of the fraction eluted from 2.4 to 2.6 min a characteristic peak at m/z 97 is present. The daughter ion spectrum of m/z 97 (masses 97 and 79) conÆrmed the presence of phosphate as the binding ligand. The amount of Alphosphate in the LMW fraction of serum sample No. IV was 18%.The mass spectrum of the fraction eluted from 3.0 to 3.2 min indicates the presence of characteristic peaks with m/z 191 and 97. The corresponding daughter ion spectrum of m/z 191 resulted in characteristic masses of 111, 87 and 85, while in the daughter ion spectrum of m/z 97, peaks with m/z 97 and 79 are present. These data conÆrm the presence of citrate and phosphate binding ligands. The same spectra were found for the fractions eluted from 2.8 to 3.0 and from 3.2 to 3.4 min.On the basis of these data it can be presumed that the Al which is eluted from 2.8 to 3.4 min is present as Al-citrate and ternary Al-citrate-phosphate complexes. The presence of Al-phosphate in this chromatographic peak could be excluded on the basis of the retention time. The ternary Al-citrate-phosphate complex has also been predicted as one of the possible LMW-Al species in serum by computer-aided speciation.24,29 The amount of Alcitrate and ternary Al-citrate-phosphate complexes in sample No.IV represented 78% of the LMW-Al species. With the Fig. 2 ES-mass spectra and corresponding daughter ion mass spectra for synthetic solutions of Al-citrate (A), Al-phosphate (B) and Al-ATP (C) (100 ng cm23 Al) at pH 7.4. Table 1 Concentrations of total Al (ng cm23) in serum from eight healthy volunteers determined by ETAAS (n~3). Serum sample (No.) Total concentration of Al/ng cm23 I 9.0°1.0 II 9.0°1.0 III 6.0°0.5 IV 7.0°1.0 V 7.5°1.0 VI 6.0°0.5 VII 5.0°0.5 VIII 11.0°1.0 1746 J. Anal.At. Spectrom., 1999, 14, 1743±1748speciation procedure described it was not possible to distinguish quantitatively between these two Al species. On the basis of the retention times and ES-MS-MS analysis, it was found that in sample Nos. V and VI, the Al species eluted from 2.4 to 2.6 min corresponded to Al-phosphate (26 and 43% of the LMW-Al species, respectively). The LMW-Al species eluted from 2.8 to 3.6 min corresponded to Al-citrate and ternary Al-citrate-phosphate complexes.The amount of these species in the LMW-Al fraction was 62% for sample No. V and 47% for sample No. VI. In contrast to the other samples analysed, sample Nos. VII and VIII did not contain Al-phosphate. The Al peak that was eluted from 2.8 to 3.2 min (Table 2) corresponded, on the basis of ES-MS-MS analysis, to Al-citrate and ternary Al-citrate- Table 2 Separation of microultraÆltrable (LMW) Al in spiked human serum from eight healthy volunteers on an anion-exchange FPLC column and determination of Al (ng cm23) in separated fractions by ETAAS (n~3) Concentration of Al in separated fractions/ng cm23 Sample No.Time/min I II III IV V VI VII VIII 0±0.2 a a a a a a a a 0.2±0.4 a a a a a a a a 0.4±0.6 a a a a a a a a 0.6±0.8 a a a a a a a a 0.8±1.0 5.7°0.3 1.0°0.2 a a 1.5°0.2 1.0°0.2 6.9°0.3 a 1.0±1.2 13.4°0.4 1.0°0.2 1.0°0.2 1.0°0.2 1.8°0.2 1.0°0.2 a 5.9°0.3 1.2±1.4 a a a a a a a a 1.4±1.6 a a a a a a a a 1.6±1.8 a a a a a a a a 1.8±2.0 a a a a a a a a 2.0±2.2 a a a a a a a a 2.2±2.4 a a a a a a a a 2.4±2.6 12.6°0.4 12.2°0.4 5.9°0.4 6.2°0.3 7.1°0.3 9.6°0.4 a a 2.6±2.8 a 1.0°0.2 1.0°0.2 a a a a a 2.8±3.0 a 1.0°0.2 a 6.5°0.3 a 1.0°0.2 1.5°0.2 1.5°0.2 3.0±3.2 24°0.5 5.0°0.3 7.0°0.3 16.9°0.4 6.6°0.3 6.8°0.4 15.4°0.4 8.5°0.4 3.2±3.4 6.0°0.4 a 1.2°0.2 3.4°0.2 5.0°0.4 2.8°0.3 a a 3.4±3.6 a a a a 5.6°0.4 a a a 3.6±3.8 a a a a a a a a 3.8±4.0 a a a a a a a a 4.0A10.0 a a a a a a a a LMW Al species (%) 55 17.5 14.50 30 24 20 21 14.5 aBelow instrumental LOD (0.5 ng cm23) Fig. 3 ES-mass spectra and corresponding daughter ion mass spectra of m/z 191 and 97 for eluted fractions from 2.4±2.6 and 3±3.2 min, respectively, on an anion-exchange FPLC column for serum sample No. I. J. Anal. At. Spectrom., 1999, 14, 1743±1748 1747phosphate complexes (sample No. VII), and to the Al-citrate complex alone (sample No. VIII). The amounts of these LMWAl species eluting from 2.8 to 3.2 min were 71 and 63%, respectively. On the basis of this study it is evident that the percentage and distribution of LMW-Al species in spiked human serum from healthy volunteers varies between particular individuals involved in the study.Acknowledgements This work was Ænancially supported by the Ministry of Science and Technology of Slovenia. References 1 A. I. Arieff, Am. J. Kidney Dis., 1995, 6, 317. 2 D. M. Grekas, H. A. Ellis, M. K. Ward, A.M. Martin, I. Parkinson and D. N. S. Kerr, Uremia Investig., 1984, 8, 9. 3 J. W. Coburn, M. G. Mischel, W. G. Goodman and I. B. Salusky, Am. J. Kidney Dis., 1991, 16, 708. 4 J. S. Lindberg, J. B. Copley, K. G. Koenig and H. M. Cuhner, South. Med. J., 1993, 86, 1385. 5 B. Quartley, G. Esselmont, A. Taylor and M. Dobrota, Food Chem. Toxicol., 1993, 31, 543. 6 M. F. van Ginkel, G. B. van der Voet, P. C. D'Haese, M. E. De Broe and F. A. de Wolff, J. Lab. Clin. Med., 1993, 121, 453. 7 P. Slanina, Y. Falkeborn, W. Frech and A. Cedergren, Food Chem. Toxicol., 1984, 22, 391. 8 D. C. Ackley and R. A. Yokel, Toxicology, 1997, 120, 89. 9 J. I. Garcý�a Alonso, A. Lo�pez Garcý�a, J. Pe�rez Parajo�n, E. Blanco Gonza� lez, A. Sanz-Medel and J. B. Cannata, Clin. Chim. Acta, 1990, 189, 69. 10 P. C. D'Haese, G. F. van Landeghem, L. V. Lamberts and M. E. de Broe, Mikrochim. Acta, 1995, 120, 83. 11 K. Wro� bel, E. B. Gonza� lez, K. Wro� bel and A. Sanz-Medel, Analyst, 1995, 120, 809. 12 A. B. Soldado Cabezuelo, E. Blanco Gonza� lez and A. Sanz-Medel, Analyst, 1997, 122, 573. 13 A. B. Soldado Cabezuelo, M. Montes Bayon, E. Blanco Gonzalez, J. I. Garcia Alonso and A. Sanz-Medel, Analyst, 1998, 123, 865. 14 H. B. Ro» llin and C. M. C. A. Nogueira, Eur. J. Clin. Chem. Clin. Biochem., 1997, 35, 215. 15 W. R. Harris, G. Berthon, J. P. Day, C. Exley, T. P. Flaten, W. F. Forbes, T. Kiss, C. Orvig and P. F. Zatta, J. Toxicol. Environ. Health, 1996, 48, 543. 16 C.Orvig and G. Berthon, in Handbook of Metal±Ligand Interactions in Biological Fluids, ed. G. Berthon, Marcel Dekker, New York, 1995, pt. 5, p. 1266. 17 J. E. Gregor and H. K. J. Powell, Aust. J. Chem., 1986, 39, 1851. 18 T. Kiss, I. Sovago and R. B. Martin, Inorg. Chem., 1991, 30, 2130. 19 M. Venturini and G. Berthon, J. Inorg. Biochem., 1989, 37, 69. 20 S. Dayde, M. Filella and G. Berthon, J. Inorg. Biochem., 1990, 38, 241. 21 R. B. Martin, Clin. Chem., 1986, 32, 1797. 22 R. B. Martin, J. Inorg. Biochem., 1986, 28, 181. 23 L. O. O» hman and R. B. Martin, Clin. Chem., 1994, 40, 598. 24 G. E. Jackson, Polyhedron, 1990, 9, 163. 25 D. J. Clevette and C. Orvig, Polyhedron, 1990, 9, 151. 26 W. R. Harris, Clin. Chem., 1992, 38, 1809. 27 J. D. Bell, G. Kubal, S. Radulovic, P. J. Sadler and A. Tucker, Analyst, 1993, 118, 241. 28 K. Atkari, T. Kiss, R. Bertani and R. B. Martin, Inorg. Chem., 1996, 35, 7089. 29 A. Lakatos, T. Kiss and I. Banyai, COST D8 and ESF Workshop on Biological and Medicinal Aspects of Metal Ion Speciation, Szeged, Hungary, 1998, p. L26. 30 J. Pe�rez Parajo�n, E. Blanco Gonza� lez, J. B. Cannata and A. Sanz-Medel, Trace Elem. Med., 1989, 6, 41. 31 T. Bantan, R. Milacœicœ and B. Pihlar, Talanta, 1998, 47, 929. 32 A. K. Datta, P. J. Wedlund and R. A. Yokel, J. Trace Elem. Electrolytes Health Dis., 1990, 4, 107. 33 H. Keirsse, J. Smeyers-Verbeke, D. Verbeelen and D. L. Massart, Anal. Chim. Acta, 1987, 196, 103. 34 T. Bantan, R. Milacœicœ and B. Pihlar, Talanta, 1998, 46, 227. 35 J. Sœ cœ ancœar, R. Milacœicœ, M. Benedik and P. Bukovec, Clin. Chim. Acta, 1999, 283, 139. 36 S. Caroli, A. Alimonti, E. Coni, F. Petrucci, O. Senofonte and N. Violante, Crit. Rev. Anal. Chem., 1994, 24, 363. Paper 9/904213J Fig. 4 ES-mass spectra and corresponding daughter ion mass spectra of m/z 191 and 97 for eluted fractions from 2.4±2.6 and 3±3.2 min, respectively, on an anion-exchange FPLC column for serum sample No. IV. 1748 J. Anal. At. Spectrom.,