Investigation of the aluminium binding in Al(III)-treated neuroblastoma cells

Abstract

Different hyphenated techniques have been employed for the first time for a comparative study of control and metal-exposed cell lines, one of the important research areas in the field of metallomics. Speciation of aluminium (an element implicated in a variety of neurological disorders) was investigated in neuroblastoma cells exposed to aluminium lactate by size-exclusion high-performance liquid chromatography-inductively coupled plasma mass spectrometry (SE-HPLC-ICP-MS) and capillary electrophoresis (CE-ICP-MS). Whereas in control cells the most intense Al compound co-eluted with the Al–transferrin complex, in size-exclusion chromatography, in exposed cells a low molecular weight bioligand was synthesized. The latter bound all the Al(III) metabolised by exposed cells in the form of a negatively-charged complex. This complex turned out to be more stable than the protein complex of Al(III) in control cells. The isolated low molecular weight Al species did not produce a signal in electrospray mass spectrometry (ES-MS) and its behaviour investigated by size-exclusion, reversed-phase HPLC and CE excluded its being hydroxide, citrate, phosphate or residual lactate from the culture medium.

---

Introduction

Aluminium is a widely recognized neurotoxin for more than a century and a toxic factor in several human diseases.^1,2 Its accumulation by patients with renal failure is a well-known hazard.^3,4 The neurological symptoms are similar to Alzheimer’s disease;^4,5 others include bone pain, muscle weakness, microcytic anaemia and hypercalcaemia.^3–5

Despite many efforts to understand the origin and the mechanisms of the neurotoxicity of aluminium,^6,7 its localization(s) in the human brain and its mechanism(s) of action have remained virtually unknown. There is a growing acceptance of the fact that the understanding of the mechanisms of the Al toxicity is critically dependent on the knowledge (on the molecular level) of the in vivo Al interactions with bioligands. This implies the determination of the Al speciation, i.e., the particular chemical forms in which the element is transported and stored in the human body.

Aluminium speciation in biological systems has been exhaustively reviewed.^8–11 Citrate is considered to be the main low molecular weight ligand and transferrin the main carrier protein in human serum.¹² In biofluids with the lower concentrations of these ligands, Al(III) displaces Mg(II) from nucleoside di- and triphosphates. In their absence catecholamines become Al(III) binders.⁸ Citrate was proposed as the Al-carrier in brain extracellular fluid.¹³

Analytical approaches used to determine the Al speciation in biological fluids have been based on the on-line coupling of anion-exchange fast protein liquid chromatography with Al detection by electrothermal atomic absorption spectrometry (AAS),^14–17 inductively coupled plasma atomic emission spectrometry (ICP-AES)¹⁸ and ICP-MS.¹⁹ They allowed the discrimination between transferrin and albumin and the identification of another unidentified carrier protein.²⁰ Capillary zone electrophoresis with fluorimetric detection was proposed to study protein binding of Al in human serum.²¹ Electrospray tandem mass spectrometry (MS-MS) was successfully used to confirm the presence of citrate, phosphate and citrate–phosphate Al(III) complexes.¹⁷

The use of cell cultures instead of experimental animals is becoming a marked trend in toxicological studies of trace elements.²² Cells exposed to a metal species can produce metabolites of great toxicological significance. In particular, animal²³ and human²⁴ neuroblastoma cells, represent an interesting model for the neurotoxicology of aluminium. Neuroblastoma cells are malignant tumour cells that strongly resemble embryonic neuroblasts. During the cell differentiation a variety of neuronal properties are expressed, including the formation of neuritis and the presence of excitable membranes, as well as the production of neuron-specific enzymes and neurotransmitters. For these reasons, neuroblastoma cells are a convenient system for studying the neurotoxicity of aluminium in vitro in a stabilized neurone-like cell culture system. Different lines of neuroblastoma cells have been used to study aluminium toxicity.²⁵ Former studies by scanning electron microscopy of murine neuroblastoma cells exposed to aluminium at concentrations less than those affecting cell viability (15–30 µg ml⁻¹ as Al) revealed high toxicity of aluminium acetylacetonate and maltolate and scarce toxicity of aluminium lactate.²⁶

The demonstration of a bioinduction of new ligands in metal exposed tissues and cell lines and the identification of the biosynthesized metabolites are the main challenges to the analytical biochemistry of trace metals.^27,28 In order to correlate the toxicity symptoms with the Al concentration in the culture medium the use of Al chelates that are soluble at the physiological levels is mandatory.²⁴ The goal of this study was to investigate, on the molecular level, the response of neuroblastoma cells to aluminium (added as the Al–lactate complex to the cell growing medium). Techniques employing different separation mechanisms (size-exclusion HPLC, reversed-phase HPLC and capillary electrophoresis), coupled with sensitive Al-specific detection by ICP-MS, were developed to monitor the bioinduction of Al compounds, biosynthesized as a cell response to metal stress. This is, to our knowledge, the first time that hyphenated techniques have been used for a comparative study of control and metal-exposed cell lines, one of the important research areas in the field of metallomics.

Experimental

Instrumentation

Size-exclusion (SE) HPLC was carried out using a HP Model 1100 HPLC pump (Hewlett-Packard, Wilmington, DE, USA) as the sample delivery system. Injections were performed using a Model 7725 injection valve with a 100 µl injection loop (Rheodyne, Cotati, CA, USA). All the connections were made of polyether ether ketone (PEEK) tubing (id 0.17 mm). Analyte species were separated on 10 × 300 mm × 13 µm Superdex^™-75 column (Pharmacia Biotech, Uppsala, Sweden) with an exclusion limit of 100 kDa (an effective separation range 0.5–80 kDa).

Microbore HPLC was performed using an ABI 140C microbore syringe pump, an ABI Model 112A injection module and an ABI Model 785A absorbance detector equipped with a microbore cell (Applied Biosystems, Foster City, CA, USA). Analyte species were separated using a Vydac C₈ (150 mm × 1 mm × 5 µm) column (Hesperia, CA, USA) with the pore size of 300 Å.

Aluminium in HPLC effluents was detected using an ELAN 6000 ICP mass spectrometer (SCIEX, Concord, ON, Canada) equipped with a cross-flow nebulizer and a standard Ryton spray chamber. For microbore HPLC a microconcentric nebulizer (MicroMist) and low-volume spray chamber (Glass Expansion, Romainmotier, Switzerland) were used; the makeup flow was supplied from a peristaltic pump via a T-piece between the column and the nebulizer. The data was processed using the Turbochrom software (SCIEX Perkin-Elmer, Ontario, Canada). The ICP-MS conditions were optimized daily using the standard built-in procedure (daily check).

CE experiments were carried out with a Beckman P/ACE 2200 (Beckman Instruments, Inc., Fullerton, CA, USA). CE separations were carried out in a 1 m uncoated 75 µm id fused silica capillary (Beckman). The ICP-MS instrument used was an Agilent Model 7500 (Yokogawa, Yamanashi-Ken, Japan). The instrument was automatically tuned by the Chemstation software using the ion signal of ⁸⁹Y in the makeup flow. The CE-ICP-MS interface was based on a microconcentric nebulizer (MCN-100, CETAC, Omaha, NE, USA) in which the original nebulizer capillary was replaced by one with a narrower diameter, and a small volume (ca. 5 mL) spray chamber (CETAC). The interface was described in detail elsewhere.^29,30 In brief, the makeup liquid, grounded by a Pt electrode, was mixed with the CE buffer at the end of the CE capillary. It was transported to the nebulizer by self-aspiration at the rate of 6 µL min⁻¹. The makeup flow was necessary to compensate for the difference between the electroosmotic flow rate used in capillary electrophoresis (ca. 1 µl min⁻¹) and that required by the nebulizer (6 µl min⁻¹). Under optimized conditions the intensity of the ⁸⁹Y signal did not vary by more than 3%.

Electrospray MS experiments were performed using a PE-SCIEX API 300 Ion-spray triple-quadrupole mass spectrometer (Thornhill, ON, Canada).

Chemicals

Water purified using a Milli-Q water purification system (Millipore, Bedford, MA, USA) was used. Chemicals were from Sigma–Aldrich (St. Quentin Fallavier, France) unless stated otherwise. The Tris–HCl buffer solution was prepared by dissolving 30 mM or 12 mM of Tris in water for size-exclusion LC and capillary electrophoresis, respectively, and adjusting the pH to 7.5 with hydrochloric acid (1 ∶ 10). For reversed-phase HPLC the acetate buffer solution was prepared by dissolving 5 mmol of ammonium acetate in 1 l of water and adjusting the pH to 6.0 with acetic acid. Buffers were degassed in an ultrasonic bath before use. Reagents used for the preparation of cell cultures were supplied by Sigma–Aldrich (Milano, Italy).

Cell cultures

Murine neuroblastoma cells C1300 (clone Neuro-2A, N2A) were obtained from American Type Culture Collection (ATCC, Rockville, MD, USA), seeded in Petri dish, and grown in Dulbecco's modified Eagle's medium (DMEM) with 10% heat inactivated foetal calf serum (Seromed, Milan, Italy) at 37 °C in a humidified atmosphere of 5% CO₂–95% air.

Cells were seeded at a density of 100 000/35 mm diameter culture plastic dish²⁵ and resuspended in a medium containing 135 mM NaCl, 5 mM KCl, 4 mM Mg₂SO₄, 4 mM NaHCO₃ and 20 mM HEPES pH 7.3. Aluminium lactate (5 mM) was put into the medium before the addition of bovine calf serum. In the control cell cultures were treated with Na lactate 15 mM. Cell cultures were maintained in the presence or in the absence of Al lactate for several months.

Cells were harvested after several washings with culture medium without Al(III) and thus with Tris–HCl buffer. Cells were sonicated and supernatant was collected after centrifugation in an Eppendorf tube at 15 000 rpm (Beckman Coulter, microcentrifuge 18). Supernatant was collected and analyzed as reported below.

Size-exclusion chromatography. For analytical SEC-ICP-MS a sample of 100 µl was filtered for the removal of particles, injected onto the column and eluted with 30 mM Tris-HCl buffer (pH 7.5) with a flow rate of 0.75 ml min⁻¹. For preparative SEC a sample was filtered for the removal of particles, diluted 1 ∶ 1 with Tris–HCl 30 mM buffer at pH 7.5 and introduced onto the column. The elution was accomplished with 30 mM Tris–HCl buffer at pH 7.5 at a rate of 0.5 ml min⁻¹. The eluate was collected every 30 s, giving the equivalent of 250 µL in each fraction. Aluminium was determined in each fraction by ICP-MS, the intensity of ²⁷Al was noted and an elution ‘profile’ was reconstructed using Microsoft Excel software. Fractions in the range between 33–41 were pooled and lyophilized.

Microbore reversed-phase HPLC-ICP-MS. A sample of 100 µl was filtered for the removal of particles, injected onto the column and eluted with 5 mM ammonium acetate buffer (pH 6.0) with a flow rate of 40 µl min⁻¹.

Capillary electrophoresis. Sample was injected hydrodynamically at 0.5 psi for 3 s, which corresponded to an 8 nl injection volume. The running buffer was identical with the makeup buffer and was 12 mM Tris–HCl (pH 7.5). Separations were carried out at 25 °C using a voltage of 20 kV that corresponded to a current of ca. 20 µA.

Results and discussion

Verification of the bio-induction of aluminium-complexing metabolites in Al-treated neuroblastoma cells

From the analytical point of view the N2A neuroblastoma cells cytosol is a complex sample containing a number of ligands, including proteins, such as transferrin or albumin, which are able to bind aluminium in complexes with various stability constants. These complexes are usually of limited interest since the detoxification mechanism is based on the bio-induction of an Al-complexing ligand, usually absent in the control sample. Therefore, the first step of the analytical approach should allow one to establish whether any new Al-binding ligands occur in the exposed sample in comparison with the control sample. This can be readily achieved by comparing a chromatogram obtained for an exposed sample with a corresponding chromatogram of a control sample spiked with aluminium. Size-exclusion HPLC is a suitable chromatographic mechanism for that comparison since it allows the simultaneous screening of high and low-molecular weight compounds. In order to minimize the risk of changes of the original Al(III) speciation, the separation was investigated at the pH of 7.5 ± 0.1, which is very close to the physiological value.

Fig. 1(a) shows a size-exclusion LC-ICP-MS chromatogram of a supernatant solution obtained after centrifugation of cells exposed to Al(III), as described in the Procedure. It shows a major peak in the low molecular weight range that accounts for over 95% of the total Al present. Two minor peaks, one close to the void corresponding to a compound with a molecular weight of 60–80 kDa and one in the middle of the chromatogram, are present. No peak is present in the corresponding chromatogram of a control sample (Fig. 1(b)). The spiking of the control sample with Al(III) results in the complexation of the latter by a high molecular weight ligand present in the sample which is likely to be transferrin (Fig. 1(c)), the main protein ligand for aluminium.^14,31–34 The same peak appears in the chromatogram of the exposed sample spiked with 10-fold excess of aluminium (Fig. 1(d)). However, the peak corresponding to the bioinduced species remains unaffected. This indicates that the stability of the biosynthesized complex is higher than that of Al–protein complexes.

Fig. 1 Size-exclusion LC-ICP-MS chromatograms of neuroblastoma cell cytosols. (a) Sample exposed to Al stress; (b) control sample; (c) control sample spiked with Al(III); (d) exposed samples spiked with Al(III). See the Procedure section for the experimental conditions.

The SEC elution profile remained the same in the pH range investigated: 6.0–8.8. The reproducibility of the chromatograms (triplicate injection, peak area mode) was 7–8% for the major peaks, 11% for the bioinduced peak and ca. 15–20% for the minor peaks. The stability of the compounds was investigated heart-cutting an Al-containing fraction and re-chromatographing it in the same conditions. The Al-complexes can be considered as stable since more than 85% of the initial compound could be recovered.

Identification of the bio-induced species

Our first concern was to exclude the possibility of the supposed-to-be bio-induced compound being a simple artefact due to the residual Al–lactate adsorbed on the cells and contaminating the cytosol. In order to verify this, a series of synthetic solutions containing Al(III) and lactate at different ratios were injected to produce the chromatograms shown in Fig. 2(a). It is clearly seen that the morphology of the chromatogram is independent of the Al ∶ lactate ratio. Also, the elution volumes of the complexes present are distinctly higher than that of the bio-induced compound.

Fig. 2 Size-exclusion LC-ICP-MS chromatograms of reconstituted Al complexes. (a) Lactate; (b) phosphate; (c) citrate; (d) transferrin. The different chromatograms for a given complex correspond to the different metal-to-ligand ratios.

The identification attempts were based on the comparison of elution volumes of the bioinduced compound with those of synthetic standards of Al complexes with ligands frequently reported in the literature: citrate, phosphate and transferrin. The Al–phosphate complex elutes (Fig. 2(b)) later than the bioinduced compound. Citrate forms two different complexes with Al of which the intensity ratio is a function of the metal to ligand ratio (Fig. 2(c)). The bigger of these complexes co-elutes with the bioinduced compound. Since the sample does not show the characteristic doublet, citrate is unlikely to be the ligand. This was confirmed further on by a separation technique using a different principle (CE). The Al complex with transferrin produces a peak (Fig. 2(d)) at the elution volume corresponding to the peak obtained by spiking a control sample with Al.

In order to gain a deeper insight into the identity of the bio-induced Al compound, the latter was isolated by semi-preparative size-exclusion chromatography (Fig. 3) and analysed further by reversed-phase HPLC-ICP-MS, CE-ICP-MS and ES-MS.

Fig. 3 Isolation of the bio-induced Al complex by semi-preparative size-exclusion LC-ICP-MS. The collected fraction is shaded. The inset shows a reversed-phase HPLC-ICP- MS chromatogram of the collected fraction.

Analysis of the bio-induced compound by two-dimensional (size-exclusion reversed-phase) HPLC-ICP-MS

The inset in Fig. 3 shows a reversed-phase HPLC-ICP-MS chromatogram of the Al-rich fraction isolated by size-exclusion chromatography. It can be seen that Al elutes in the void of the reversed-phase column, which indicates the strongly polar (ionic) character of the complex formed. On this basis the possibility needs to be excluded that a metallothionein-like protein was bio-induced as indicated by the elution volume of the complex in SE-HPLC. It should be noted that because of the high polarity and interactions with the pores of the size-exclusion packing the elution volume of the bio-induced complex is likely to be smaller than it would result from its size.

Analysis of the bio-induced compound by SEC-CE-ICP-MS

The compound isolated and purified by SEC was further analyzed by CE-ICP-MS (Fig. 4). It was found to migrate like an anionic species (12.63 min, relative standard deviation of peak area 4.8%) (Fig. 4(a)). The Al(III) complex with lactate elutes significantly earlier (Fig. 4(b)). Citrate and phosphate form with Al(III) under these conditions strongly anionic species that are unable to leave the capillary (Figs. 4(c) and (d)). The sum of the CE-ICP-MS experiments confirms that a previously unreported bio-induced compound was formed as a response of neuroblastoma cells to exposure to Al(III).

Fig. 4 CE-ICP-MS electropherograms of: (a) Al containing fraction isolated by size-exclusion LC in Fig. 3; (b) Al–lactate complex; (c) Al–citrate complex; (d) Al–phosphate complex.

Analysis of the bio-induced compound by ES-MS-MS

The Al-containing fraction was also analysed by ES-MS-MS. In this case only the fraction corresponding to the apex of the peak shown in Fig. 3 was analysed in order to assure the maximum preconcentration factor and the maximum freedom from co-eluting interferents. No significant data could, however, be obtained, probably because of a too low concentration of the Al complex. The estimated concentration after preconcentration was 100 ng ml⁻¹ (as Al). Such a low concentration may remain undetected by ES-MS even if the compound of interest is readily ionized. Indeed, the only successful demonstration of ES-MS-MS for the identification of Al complexes in biological fluids was carried out on spiked serum.¹⁷ Note that in our case, the problem is the quantity of ligand and not that of Al since the additional Al(III) spiked on the sample was entirely bound to protein, probably transferrin.

Conclusions

Hyphenated techniques have been demonstrated for the first time to be an attractive analytical tool for the detection of metal-binding ligands induced in experiments involving the exposure of cell lines to metal stress. The sub-pg detection limits of ICP-MS in HPLC and capillary electrophoresis make HPLC-ICP-MS and CE-ICP-MS convenient techniques for monitoring the stability of the metallospecies detected during multidimensional chromatographic purification protocols. The use of a larger mass of cells and a more sensitive ES-MS-MS instrument would be required for the identification of the Al(III) compounds detected. However, already this first demonstration of the bioinduction of a novel ligand in neuroblastoma cells treated with Al(III) is an interesting contribution to the studies of the mechanisms of the aluminium neurotoxicity.

References

1. M. Nicolini, P. F. Zatta and B. Corain, Aluminum, in Chemistry, Biology and Medicine, Cortina International, Verona, 1991.
2. R. Yokel, Neurotoxicology, 2000, 21, 813.
3. T. B. Drueke, Nephrol. Dial. Transplant., 2002, 17(Suppl 2), 13.
4. A. J. Freemont, Semin. Dial., 2002, 15, 227.
5. D. J. Burn and D. Bates, J. Neurol. Neurosurg. Psychiatry, 1998, 65, 810.
6. M. Venturini-Soriano and G. Berthon, J. Inorg. Biochem., 1998, 71, 135.
7. R. R. Crichton, A. Florence and R. J. Ward, Coord. Chem. Rev., 2001, 228, 365.
8. R. B. Martin, Ciba Found. Symp., 1992, 169, 5.
9. E. DeVoto and R. A. Yokel, Environ. Health Perspect., 1994, 102, 940.
10. 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.
11. G. F. Van Landeghem, M. E. De Broe and P. C. D'Haese, Clin. Biochem., 1998, 31, 385.
12. L. O. Ohman and R. B. Martin, Clin. Chem., 1994, 40, 598.
13. R. A. Yokel, D. D. Allen and D. C. Ackley, J. Inorg. Biochem., 1999, 76, 127.
14. A. Sanz-Medel, A. B. Soldado Cabezuelo, R. Milacic and T. Bantan Polak, Coordination Chemistry Reviews, 2002, 228, 373.
15. A. B. Soldado Cabezuelo, E. Blanco Gonzalez and A. Sanz Medel, Analyst, 1997, 122, 573.
16. T. Bantan, R. Milacic and B. Pihlar, Talanta, 1998, 47, 929.
17. T. Bantan, R. Milacic, B. Mitrovic and B. Pihlar, J. Anal. At. Spectrom., 1999, 14, 1743.
18. T. Bantan, R. Milacic and B. Pihlar, Talanta, 1998, 46, 227.
19. A. B. Solado Cabezuelo, M. Montes Bayon, E. Blanco Gonzalez, J. I. Garcia Alonso and A. Sanz Medel, Analyst, 1998, 123, 865.
20. M. Favarato, C. A. Mizzen and D. R. McLachlan, J. Chromatogr., 1992, 576, 271.
21. E. Blanco Gonzalez, A. B. Soldado Cabezuelo and A. Sanz Medel, Biomed. Chromatogr., 1998, 12, 143.
22. T. Hartung, M. Balls, C. Bardouille, O. Blanck, S. Coecke, G. Gstraunthaler and D. Lewis, Altern. Lab. Anim., 2002, 30, 407.
23. P. F. Zatta, P. Zambenedetti and S. Masiero, Neurotoxicology, 1994, 15, 789.
24. C. B. Dobson, J. P. Day, S. J. King and R. F. Itzhaki, Toxicol. Appl. Pharmacol., 1998, 152, 145.
25. P. Zatta, M. Perazzolo, L. Facci, S. D. Skaper, B. Corain and M. Favarato, Mol. Chem. Neuropathol., 1992, 16, 11.
26. P. Zatta, P. Zambenedetti and R. Milačič, Analusis, 1998, 26, M72.
27. J. Szpunar, Analyst, 2000, 125, 963.
28. R. Lobinski, Fresenius' J. Anal. Chem., 2001, 372, 113.
29. A. Prange and D. Schaumlöffel, J. Anal. At. Spectrom., 1999, 14, 1329.
30. D. Schaumlöffel and A. Prange, Fresenius' J. Anal. Chem., 2000, 364, 452.
31. S. J. A. Fatemi, D. J. Williamson and G. R. Moore, J. Inorg. Biochem., 1992, 46, 35.
32. M. F. Van Ginkel, G. B. Van der Voet, H. G. Van Eijk and F. A. de Wolff, J. Clin. Chem. Clin. Biochem., 1990, 28, 459.
33. G. A. Trapp, Life Sci., 1983, 33, 311.
34. G. Kubal, A. B. Mason, P. J. Sadler, A. Tucker and R. C. Woodworth, Biochem. J., 1992, 285, 711.

---