摘要:

Testing of chelating agents and vitamins against lead toxicity using mammalian cell cultures† Anna B. Fischer*, Cristine Hess, Tilo Neubauer and Thomas Eikmann Institute of Hygiene and Environmental Medicine, Justus-Liebig-University, Friedrichstrasse 16, 35392 Giessen, Germany Mammalian cell cultures were used to determine the capacity of antidotes to modify (a) lead uptake, (b) lead toxicity and (c) lead release from cells. The following chelating agents were tested: Na, Ca-ethylenediaminetetraacetic acid (EDTA), diethylenetriaminepentaacetic acid (DTPA), nitriloacetic acid, ethylene glycol-bis(aminoethyl)tetraacetic acid (EGTA), D,L-mercaptosuccinic acid (MSA), meso-2,3-dimercaptopropanesuccinic acid (DMSA), D,L-2,3-dimercaptopropane-1-sulfonic acid (DMPS), penicillamine (PA), N-acetylpenicillamine (NAPA), and diethylcarbodithioate (DDTC).The following vitamins were tested: thiamine (B1), riboflavine (B2), pyridoxine (B6), cobalamin (B12) and ascorbic acid (C).Inhibition of lead uptake was produced by EDTA, EGTA, DMSA, DMPS, MSA, PA, NAPA and vitamins B1, B6 and C, vitamins B2 and B12 being ineffective. The same compounds reduced lead cytotoxicity. Interestingly DDTC and DTPA increased lead uptake, but did not exacerbate lead toxicity. Significant release of lead from preloaded cells was caused by DTPA, NAPA, DMPS and PA, while the other chelators were ineffective. Keywords: Lead; chelating agents; B vitamins; thiamine; riboflavine; pyridoxine, cobalamin; cellular metal uptake; cellular metal release; cellular metal toxicity Mammalian cell cultures are established tools in many medical and biological disciplines. Several working groups have realised the usefulness of this experimental system for toxicological studies into the effects of chelating agents.1–4 On the one hand cell cultures can be used for the screening of antidotes to identify possible candidates against toxic metals before they are tested in experimental animals; on the other hand events at the cellular level can be elucidated. Thus the use of cell cultures may also help to spare or reduce animal experiments, a goal that is desirable for financial as well as ethical reasons. In previous studies an experimental approach for the assessment of the antidotal efficacy of chelating agents directed against cadmium-inducd cytotoxicity was presented and the salient methodological points for the determination of the toxicity of the antidotes and their modifying effects on metal uptake and cytotoxicity as well as metal mobilization were discussed.5,6 In the present work, tests with chelating agents of potential and proven therapeutic efficacy against lead are communicated.About a decade ago the B vitamin complex were found to prevent lead poisoning in rats.7 We were intrigued to find out if these antidotes were also effective in our experimental system.8 These studies, which identified thiamine and pyridoxine as inhibitors of lead intoxication, have been extended and are reported here.Experimental Chemicals The chelating agents, vitamins and their sources were as follows: Na, Ca-ethylenediaminetetraacetic acid (EDTA) from Serva, Heidelberg, Germany; diethylenetriaminepentaacetic acid (DTPA), d,l-mercaptosuccinic acid (MSA), meso-2,3-dimercaptopropanesuccinic acid (DMSA), d,l-2,3-dimercaptopropane- 1-sulfonic acid (DMPS), penicillamine (PA), Nacetylpenicillamine (NAPA) from EGA-Chemie, Steinheim, Germany; ethylene glycol-bis(aminoethyl)tetraacetic acid (EGTA), nitriloacetic acid (NTA) and diethylcarbodithioate (DDTC) from Merck, Darmstadt; the vitamins thiamine (B1), riboflavine (B2), pyridoxine (B6), cobalamine (B12) and ascorbic acid (C) from Merck, Darmstadt, Germany and cell culture media and reagents from Seromed, Berlin, Germany.Cell cultures Chinese hamster peritoneal cells, line B14F28, were serially cultivated in minimal essential medium (MEM) supplemented with 5% new-born calf serum (NCS), non-essential amino acids, glutamine and penicillin/streptomycin. The pH of the medium was stabilised to 7.0 ± 0.1 with 20 mmol l21 HEPES buffer.9 Procedure Cytotoxicity experiments Nearly confluent Chinese hamster cell cultures (inoculum of 30 000 cells per 25 cm2 flask, 3 d old; four replicates per experimental group) were exposed to 0.6 mmol l21 PbCl2 and different concentrations of the chelating agents or vitamins in serum-free medium.Cell numbers were determined 24 h later by detaching the cells from the flasks with trypsin (0.5% in PBS) and counting aliquots in an electronic counter (Coulter Counter).10 The results are expressed as the percentage of the attained cell increase of the cultures treated only with lead; the proliferation of these lead-treated cultures was usually reduced by 20–30% compared with untreated controls. The standard deviation of the replicate cell counts within experimental groups was generally < 4%.Lead uptake experiments Semi-confluent cultures (four replicates per experimental group) were exposed to 0.6 mmol l21 PbCl2 and different concentrations of the chelating agents or vitamins in serum-free medium. After 24 h incubation at 37 °C the cells were washed four times with PBS, detached by trypsinization and aliquots counted. Cells were processed and their lead contents analysed by atomic absorption spectrometry (AAS). The lead content of the samples was related to cell numbers.Lead release experiments Semi-confluent cultures (5 000 000 cells per 75 cm2 flask, 3 d old; four replicates per experimental group) were exposed to † Presented at The Sixth Nordic Symposium on Trace Elements in Human Health and Disease, Roskilde, Denmark, June 29–July 3, 1997. Analyst, January 1998, Vol. 123 (55–58) 5575 mmol l21 PbCl2 for 24–27 h at 37 °C in medium containing 5% serum. After washing the cells four times with PBS they were incubated for 24 h in serum-free medium without addition of test substances followed by 24 h incubation with two concentrations of the chelating agents in serum-free medium.This protocol was chosen because ca. 30–40% of the incorporated lead was released from the cells even in the absence of chelating agents during the first 24 h, while spontaneous release during the next 24 h amounted to ca. 20%. Cell numbers were determined, the lead contents of the cells were analysed and related to cell numbers.Lead analyses The preparation of the samples and lead analysis were performed, with modifications due to the fact that cell cultures are an unusual matrix, according to methods used for biological material in our laboratory.8,11 Cells, including untreated controls, were processed by wet ashing in HNO3 (Baker-Instra- Analized), samples were slowly heated to 180 °C and the residues dissolved in HNO3. Standard lead solutions were used for calibration.In addition to the original samples standard dosages were added to aliquots of the samples, which were analysed in parallel. The lead contents were determined by AAS (PU 92100, TJA-Unicam, VG Elemental, Offenbach, Germany) with an electrothermal atomiser. The retrieval rate obtained in round robins was > 97%. Statistics Student’s t-test was used for the evaluation of cytotoxicity and lead uptake. For the lead release experiments one-way analysis of variance was combined with Scheffe’s test.Results Chelating agents In previous experiments it was shown that the majority of the employed chelating agents were well tolerated by the cells; a 50% reduction of cell proliferation (IC50, 2 d exposure) occurred at > 1 mmol l21. The exceptions were MSA, with a slightly elevated IC50 of 0.1–0.2 mmol l21, and DTPA and DDTC with a substantially higher ID50 of 10–20 mmol l21.11 Therefore DTPA and DDTC were applied at one tenth the strength of the other chelating agents in the experiments reported here.The results of the simultaneous exposure of B14F28 cells to 0.6 mmol l21 lead and the chelating agents on cellular lead uptake is shown in Fig. 1. The substances DMSA, PA, DMPS, EDTA, NTA, MSA, EGTA and NAPA caused a dose-dependent decrease of lead incorporation and correspondingly the cytotoxicity of lead was also diminished by these chelating agents (Fig. 2). A different result was observed in the case of DTPA and DDTC, which caused a clear dosedependent increase of cell-bound lead.However, contrary to an expected increase in cytotoxicity, the proliferation of the cells exposed to lead and these two agents was unchanged compared with the cells treated with lead alone (Figs. 1 and 2). The following chelating agents were able to mobilise lead from preloaded cells: PA, NAPA, DTPA, DMPS and MSA, while DDTC, EDTA, NTA, DMSA and EGTA were ineffective under the given experimental conditions (Fig. 3). Vitamins Studies of the effect of the vitamins on cellular lead uptake showed that thiamine and pyridoxine as well as vitamin C depressed lead incorporation markedly. A small, but statistically insignificant effect was noted in the case of riboflavine and cobalamin (Fig. 4). The substances depressing lead incorporation also had a favourable effect on lead toxicity (Fig. 5). Discussion In earlier studies we demonstrated the reduction of lead-induced cytotoxicity in the mouse fibroblast line L-A by chelating agents.The most effective substances were DMSA, EDTA, NTA, and PA, while NAPA, DMPS, and EGTA had no significant effect and MSA even increased the cytotoxic effects of lead.12 These present results with the Chinese hamster cell line are partly in agreement, partly contradictory; a greater number of chelating agents was found to be effective under the new, apparently more sensitive experimental conditions. This can be either due to specific characteristics of the employed cell cultures or to other experimental variables.Therefore, it would be desirable to use different cell cultures for the screening of metal antidotes. In the present investigations we found that the reduction of lead toxicity correlated with a reduced metal uptake. An interesting deviation from this rule was observed with DTPA and DDTC, which did not exacerbate lead toxicity in spite of markedly increased lead uptake.A similar finding was obtained when the combination of cadmium and DDTC was tested.5 We Fig. 1 Effect of chelating agents on lead uptake. Semi-confluent cultures of B14F28 cells were simultaneously exposed for 24 h to 0.6 mmol l21 PbCl2 and 60–240 mmol l21 of the chelating agents. DDTC and DTPA were applied at 6–24 mmol l21. Lead was determined by AAS. Fig. 2 Effect of chelating agents on lead toxicity. Semi-confluent cultures of B14F28 cells were simultaneously exposed to 0.6 mmol l21 PbCl2 and 60–240 mmol l21 of the chelating agents and cell numbers were determined 24 h later.DDTC and DTPA were applied at 6–24 mmol l21. 56 Analyst, January 1998, Vol. 123interpret this to mean that the cells incorporate and store the complex formed between the metal and the chelating agent, which is less toxic than the free metal ions since these are no longer available to react with cellular ligands. For a possible therapeutic use chelating agents should, in addition to being non-toxic, be able to mobilise metals from cells.This was clearly demonstrated for NAPA, DTPA, MSA, and PA. The ability of NAPA to mobilise intracellular lead was also observed in mice exposed to lead acetate; although survival was not improved after intraperitoneal treatment with NAPA, lead concentrations in liver, kidney and brain were significantly reduced.13 Penicillamine (PA) has been clinically used to treat lead poisoning and it has also been shown in experimental animals that tissue lead is reduced and renal lead elimination increased by PA.13–15 Our finding that DMPS reaches the intracellular space is also supported by investigations which found that DMPS is taken up into erythrocytes by an anion transport protein and can even be concentrated there.16,17 In rats 40% of the parenterally applied DMPS was found in the bile which indicates that a cellular passage occurs in the liver.18 The low cytotoxicity of this substance and its protective effect which were observed in our in vitro studies has been confirmed in vivo.DMPS increased the survival of lead-poisoned mice clearly19 and in rats it caused a substantial elevation of renal lead elimination as well as a reduction of the lead contents of blood, liver, kidney and bones.3,20,21 In clinical studies with leadintoxicated children lead concentrations decreased also.22 As shown above, DTPA increased lead uptake without an accompanying increase in lead toxicity.This indicates that the substance is transported into the cells, which was also postulated because of the chelating agent’s pharmacokinetics. 23,24 Although it has been demonstrated that DTPA is effective against lead in mice19,24 and that it is well-tolerated,25 its use has so far been restricted to the treatment of transuranic elements in humans. In our experiments MSA also showed a good efficiency in mobilising lead and it also decreased lead uptake and cytotoxicity in B14F28 cells. The depression of proliferation caused by MSA in L-A cells indicates that the substance enters the cells and possesses a certain toxic potential, which would make it unsuitable for therapeutic use.A related substance DMSA caused no lead release from the cells under our experimental conditions. In vivo studies confirmed that distribution in the organism was restricted to the extracellular compartment.18 In experimental animals this chelating agent increased lead elimination and lowered tissue lead levels.13 The therapy of lead-intoxicated persons had similar results.Since DMSA is well tolerated and can be applied by the oral route so that hospitalisation can be avoided, it has several advantages compared with EDTA.26,27 In our experimental system no lead release could be induced by EDTA although the chelator showed clear protective effects both in B14F28 and in L-A cells. Our results are in contrast to in vitro tests with osteoclasts and hepatocytes where EDTA mobilised lead significantly.28,29 This difference is probably due to the fact that bone and liver represent typical target cells for lead compared to the fibroblasts which were used in this study. In vitro studies have shown that the substance is extracellularly distributed and this has been confirmed by experiments conducted in vivo.24 Although the mechanism of action is still unclear, it has been proved that the renal excretion of lead is stimulated and tissue levels can be decreased by EDTA and it has been regularly used to treat lead-intoxicated patients.15,27,30 The remaining chelating agents NTA and EGTA, which did not affect cellular lead release, will not be discussed further here, since their therapeutic potential appears to be small.The tests with the vitamins demonstrate that the B vitamins pyridoxine and thiamine have a protective effect against lead uptake and toxicity. This corresponds to our earlier results with the mouse fibroblasts L-A11 and to the experiments by Tandon and co-workers.These authors showed in rats that lead intoxication could be prevented by the application of vitamin B complex and different parameters of lead poisoning were aggravated by vitamin B deficiency.10,31 The authors also treated lead poisoned rats with B vitamins, alone and in Fig. 3 Effect of chelating agents on lead release. Semi-confluent cultures of B14F28 cells were exposed for 24 h to 75 mmol l21 PbCl2, washed and incubated for another 24 h in serum-free medium followed by 24 h with 50 and 100 mmol l21 of the chelating agents.Cell numbers were determined and lead was analysed by AAS and related to cell counts. Fig. 4 Effect of vitamins on lead uptake. Semi-confluent cultures of B14F28 cells were simultaneously exposed for 24 h to 0.6 mmol l21 PbCl2 and 60–240 mmol l21 of the vitamins. Lead was determined by AAS and related to cell counts.Fig. 5 Effect of vitamins on lead toxicity. Semi-confluent cultures of B14F28 cells were simultaneously exposed to 0.6 mmol l21 PbCl2 and 60–240 mmol l21 of the vitamins and cell numbers were determined 24 h later. Analyst, January 1998, Vol. 123 57combination with CaNa2-EDTA, and found that folic acid and pyridoxine might be the factors responsible for the favourable effects.32 In our experiments vitamin C was also highly effective against lead-induced cellular intoxication.This was also observed in rats.33 To our knowledge the promising approach of employing vitamins for the prevention and therapy of lead intoxication has not yet been introduced into human medicine. Our tissue culture studies confirm that there may be some scope in this strategy. Our results demonstrate the usefulness of the employed experimental system to identify substances that are capable of modifying the uptake, cytotoxicity and release of lead from mammalian cells.This system can easily be standardised and is comparatively rapid and inexpensive. However, it must be stated that effective substances can be missed (false negatives), a shortcoming that may be reduced by the use of several parallel culture systems. In the case of substances exhibiting antidotal properties in vitro, animal tests must confirm the findings, particularly for pharmacokinetic considerations. References 1 Aaseth, J., Alexander, J., and Deverill, J., Chem. Biol. Interact., 1981, 36, 287. 2 Bakka, A., Aaseth, J., and Rugstad, H. E., Acta Pharmacol. Toxicol., 1981, 40, 432. 3 Borenfreund, F., and Puerner, J. A., Toxicology, 1986, 39, 121. 4 Fischer, A. B., and Seibold, G., in International Conference on Heavy Metals in the Environment, ed. Lekkas, T. D., CEP Consultants, Edinburgh, 1985,vol. 2, pp. 110–112. 5 Fischer, A. B., Analyst, 1995, 120, 975. 6 Fischer, A. B., Falk, A., and Seibold, G., Plzen. Lek. Sb., Suppl., 1990, 62, 35. 7 Flora, S.J., Singh, S., and Tandon, S. K., Z. Gesamte Hyg., 1984, 30, 409. 8 Prucha, J., Zbl. Bakt. Hyg., 1987, 185, 273. 9 Fischer, A. B., Xenobiotica, 1985, 15, 751. 10 Fischer, A. B., Zellul�are Toxizit�at von Schwermetallen. Akute und chronische Wirkung auf S�augerzellkulturen, Wissenschaftlicher Fachverlag, Gießen, Germany, 1995. 11 Deutsche Forschungsgemeinschaft, Blei, in Analytische Methoden zur Pr�ufung gesundheitssch�adlicher Arbeitsstoffe. Analysen im biologischen Material, ed.Greim, H., Verlag Chemie, Weinheim, 1980, vol. 2. 12 Fischer, A. B., and Falk, A., in International Conference on Heavy Metals in the Environment, ed. Vernet, J. P., CEP Consultants, Edinburgh, 1989, vol. 1, pp. 410–413. 13 Xu, Z. F., and Jones, M. M., Toxicology, 1988, 53, 277. 14 Dhawan, M., Flora, S. J. S., Singh, S., and Tandon, S. K., Biochem. Int., 1989, 19, 1067. 15 Tsuchiya, K., in Handbook on the Toxicology of Metals, ed. Friberg, L., Nordberg, G. F., and Vouk, V.B., Elsevier, Amsterdam, 1986, vol. 2, 298–353. 16 Wiedemann, P., Fichtel, B., and Szinicz, L., Biopharm. Drug Dispos., 1982, 3, 267. 17 Wildenauer, D. B., Reuter, H., and Wegner, N., Chem. Biol. Interact., 1982, 42, 165. 18 Zheng, W., Maiorino, R. M., Brendel, K., and Aposhian, H. V., Fundam. Appl. Toxicol., 1990, 14, 598. 19 Lloblet, J. M., Domingo, J. L., Paternain, J. L., and Corbella, J., Arch. Environ. Contamin. Toxicol., 1990, 19, 185. 20 Twarog, T., and Cherian, G., Toxicol. Appl. Pharmacol., 1984, 72, 550. 21 Catsch, A., Naturwissenschaften, 1968, 10, 473. 22 Chisolm, J. J., and Thomas, D. J., J. Pharmacol. Exp. Ther., 1985, 235, 665. 23 Foreman, H., Lusbaugh, C. C., Magee, M., and Humanson, G., LAMS, 1960, 2445, 67. 24 Hammond, P. B., Toxicol. Appl. Pharmacol., 1971, 18, 296. 25 Kemper, F. H., Jekat, F. W., Bertram, H. P., and Eckard, R., in Basic Science in Toxicology: Proceedings of the 5th International Congress of Toxicology, England 1989, London, 1990, p. 523. 26 Graziano, J. H., Lolacono, N. J., and Meyer, P., J. Pediatr., 1988, 113, 751. 27 Haust, H. L., Inwood, M., Spence, J. D., Poon, H. C., and Peter, F., Clin. Biochem., 1989, 22, 189. 28 Pounds, J. G., Wright, R., and Kodell, R. L., Toxicol. Appl. Pharmacol., 1982, 66, 88. 29 Rosen, J. F., Kraner, H. W., and Jones, K. W., Toxicol. Appl. Pharmacol., 1982, 64, 230. 30 Chisolm, J. J., Jr., Am. J. Dis. Child, 1987, 141, 1256. 31 Tandon, S. K., Flora, S. J., and Singh, S., Ind. J. Med. Res., 1984, 80, 444. 32 Tandon, S. K., Flora, S. J., and Singh, S., Pharmacol. Toxicol., 1987, 60, 62. 33 Flora, S. J., and Tandon, S. K., Acta Pharmacol. Toxicol., 1986, 58, 274. Paper 7/05518H Received July 30, 1997 Accepted October 28, 1997 58 Analyst, January 1998

ISSN:0003-2654    

DOI:10.1039/a705518h

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Lead concentrations and isotope ratios in street dust determined by electrothermal atomic absorption spectrometry and inductively coupled plasma mass spectrometry† Sophie M. Nageottea and J. Philip Day*b a Ecole Nationale Sup�erieure de Chimie de Montpellier, 34000 Montpellier, France b Department of Chemistry, University of Manchester, Manchester, UK M13 9PL. E-mail: philip.day@man.ac.uk A major source of environmental lead, particularly in urban areas, has been from the combustion of leaded petrol.Street dust has previously been used to assess urban lead contamination, and the dust itself can also be a potential source of lead ingestion, particularly to children. The progressive reduction of lead in petrol, in recent years, would be expected to have been reflected in a reduction of lead in urban dust. We have tested this hypothesis by repeating an earlier survey of Manchester street dust and carrying out a comparable survey in Paris. Samples were collected from streets and parks, lead was extracted by digestion with concentrated nitric acid and determined by electrothermal atomic absorption spectrometry.Lead isotope ratios were measured by inductively coupled plasma mass spectrometry. Results for Manchester show that lead concentrations have fallen by about 40% (street dust averages, 941 mg g21 (ppm) in 1975 down to 569 ppm in 1997). In Paris, the lead levels in street dust are much higher and significant differences were observed between types of street (not seen in Manchester).Additionally, lead levels in parks were much lower than in Manchester. Samples collected under the Eiffel Tower had very high concentrations and lead isotope ratios showed that this was unlikely to be fallout from motor vehicles but could be due to the paint used on the tower. Isotope ratios measurements also revealed that lead additives used in France and the UK come from different sources. Keywords: Lead analysis; lead isotopes; street dust; petrol; motor vehicles; electrothermal atomic absorption spectrometry; inductively coupled plasma mass spectrometry; public health; Manchester; Paris It has been well established for many years that lead pollution can have serious effects on human health.1–6 Lead exposure in humans arises from inhalation and ingestion, and over the past 30 years many of the sources have been progressively identified and reduced.5 For example, very little lead is now used in household paint, lead waterpipes have largely been replaced by copper or plastics and lead is no longer used in the solder for food-containing cans.One major source of environmental lead and of lead exposure to humans, both through direct inhalation and from ingestion following contamination of food chains, has been from the combustion of leaded petrol.7 From the 1930s onward, tetraalkyllead additives have been used as antiknock agents in the petrol for motor vehicles, and by the 1970s lead fallout from the combustion of petrol accounted for around 90% of the atmospheric lead input in urban environments.4 It was suggested1 that street dust would act as a good marker for generalised urban lead pollution, and in 1975 a survey of street dust in Manchester found relatively uniform concentrations of lead, around 0.1%, over the whole of the urban area.8 Comparable surveys in other large cities showed similar levels7,9 and there were concerns that the adventitious inhalation and ingestion of such dust might present a previously neglected health hazard, particularly to children.3 In the subsequent 20 years, and particularly since 1985, measures have been taken to reduce the scale of lead emissions from motor vehicles.10,11 In particular, EC regulations10 introduced in 1985 resulted in the introduction of unleaded petrol and also the rapid reduction of the lead content of leaded petrol, with the consequence that in the UK lead emissions from motor vehicles are now around 20% of those in 1975.12,13 A result has been that human lead exposure, gauged from average blood-lead levels in the general population, is less than half that of 20 years ago.5,6 The primary objective of the present study is to evaluate present levels of lead contamination of city dust, in part by repeating the methodology of the 1975 Manchester survey and in part by examining levels of lead pollution in Paris, another city in western Europe with high traffic densities, and where similar legislation and reductions in lead emission from motor vehicles will have applied.Experimental The experiment was intended to assess changes in the amounts of lead released to the environment over the past 20 years by measuring the concentrations of lead in street dust samples, and comparing these to those of the previous survey. Additionally, by measuring isotope ratios, information on the pollutant sources could be obtained.Lead concentrations were measured by electrothermal atomic absorption spectrometry (ETAAS) and isotope ratios by inductively coupled plasma mass spectrometry (ICP-MS). Sample collection Fifty five samples from Manchester were collected during the period October 1996 to January 1997, and forty six samples were collected in Paris in January 1997. The samples were of street dust, dirt and soil taken by scraping the roadside gutter or the sidewalk with a spatula.Sample locations were subdivided into the same four categories used in the 1975 survey,8 namely: A, major roads and streets with moderate or heavy traffic ( > 100 cars per hour); B, minor roads and streets on which traffic is light or moderate (10–100 cars per hour); C, streets with very light traffic ( < 10 cars per hour); D, school playgrounds, children’s play areas in parks, gardens and wasteland. † Presented at The Sixth Nordic Symposium on Trace Elements in Human Health and Disease, Roskilde, Denmark, June 29–July 3, 1997.Analyst, January 1998, Vol. 123 (59–62) 59The 1975 Manchester survey showed that the geographical distribution of lead concentrations was relatively homogenous. Therefore, for practical reasons, on this occasion we collected samples only in the south-east sector of Greater Manchester, at distances from 1 to 15 km from the city centre. For the Paris survey, samples were collected from the central (intra muros) area, within the triangle formed by the Eiffel Tower, Montmartre and the Panth�eon.Sample preparation Solutions for analysis were contained in acid-leached glassware (Pb > 10 mg l21) or poly(propylene) bottles (Pb < 10 mg l21). Chemicals and reagents were of AnalaR grade (BDH-Merck, Poole, Dorset, UK) unless otherwise stated. For lead determination by atomic absorption and ICP-MS, we followed the same treatment protocol as before.8 The samples were dried (120 °C; 12 h), ground to a homogeneous fine powder and passed through a sieve (500 mm); objects, mostly small stones, which could not be ground or sieved in this way were rejected (very little lead is normally associated with this component14).Lead was extracted from the samples by boiling portions ( ~ 1 g) in 10 ml of concentrated nitric acid for 2 h (test experiments showed that this digestion time was appropriate). After the digestion, the mixtures were rinsed with distilled water, filtered through a glass fibre paper (grade GF/A; Whatman, Maidstone, Kent, UK) and made up to 100 ml.Reagent blanks were prepared with each batch of samples. The solutions were analysed by ETAAS and ICP-MS after appropriate dilutions (30–300 mg l21 and 10–100 mg l21 , respectively, both in 0.1 m nitric acid). Two large volume samples for quality control purposes were also prepared, using ~ 1 kg of earth taken from a grassed area. One portion of earth was ground and sieved when dry whilst the other was ground and sieved as a slurry with water, to prevent the loss of very fine particles. The measured concentrations were 771 ± 8 and 1216 ± 35 mg g21 (n = 10) for the two materials, respectively, showing that a significant amount of lead was contained in easily re-suspended particles, as previously noted.14 ETAAS Lead concentrations were determined by ETAAS (Zeeman- 3030; Perkin-Elmer, Beaconsfield, Buckinghire, UK), at 283.3 nm, using direct injection of nitric acid solutions into standard (non-pyrocoated) graphite tubes.A previously optimised heating programme was used;15 the sample was ashed at 750 °C and atomised at 2200 °C, with the internal gas flow maintained at 300 ml min21 but reduced to 0 ml min21 during the reading. Calibration standards were made from a solution of lead nitrate in 0.1 m nitric acid. The stability of calibration was checked at regular intervals using the previously described quality control materials, and the instrument was re-calibrated when the deviation exceeded 5% (the slope of the calibration line tends to decrease as the tube ages).The validity of the direct calibration was tested by applying the method of additions to a representative selection (n = 20) of real sample solutions. Recoveries were in the range 94–103%, demonstrating the absence of any significant matrix effect. The detection limits were ~ 1 mg l21 and ~ 1 mg g21, for lead in the 0.1m nitric acid solutions and in the dust samples, respectively.The absolute accuracy of the lead calibration was within 2% (95% confidence limit), determined by measurement against a certified reference standard (JTB5765; LGC, Teddington, Middlesex, UK). ICP-MS Mass spectrometry is a very useful tool to characterise lead pollutant sources.16 Lead deposits at a particular location may differ in isotope ratio depending on their origins, for example either from paint or from petrol additives.When the different Pb isotope ratios for the various samples are plotted one against the other, a single source of the element will form a cluster on the graph while linear arrays tend to appear if several disparate sources are mixed. Lead isotope ratios were determined by ICP-MS (Plasma- Quad II; VG Elemental/Fisons, Winsford, UK), with the following instrument parameters: ICP incident power, 1350 W; peak jumping mode; dwell time, 10.24 ms). The dilutions were adjusted to ensure lead concentrations in the range 50–100 mg l21 in 0.1 m nitric acid.Isotope ratios were calibrated against a standard reference material for lead isotopes (RM981; National Bureau of Standards, Washington, DC, USA). The results are presented as the isotope ratios: 206Pb : 207Pb, 206Pb : 208Pb and 206Pb : 204Pb, in conformity with published results. Elemental Pb concentrations were also calculated from the ICP-MS measurements and the correlation between ETAAS and ICP-MS was very good (r = 0.993; n = 28).Results and discussion The Manchester survey The measured lead concentrations are reported in Table 1. There is no significant variation between categories of location, and the overall average value for the street locations (A–C) is 569 ± 37 mg g21. This represents a statistically significant fall (P < 0.001) of about 40% from the 1975 mean, 941 ± 19 mg g21. There has been a similar change in the distribution of concentrations; the distribution maximum is now between 400–600 mg g21 (Fig. 1) and very few samples show Pb levels over 1000 mg g21, which may be compared to the 1975 distribution,8 having a maximum between 900 and 1000 mg g21 and Pb concentrations up to 5000 mg g21. Again, as observed in 1975, Pb concentrations in dust and soil in parks and playgrounds (category D) are indistinguishable from those in all categories of street. As expected, the lead concentrations in the environment have decreased significantly, although it is noteworthy that the fall has not been as large as the reduction in lead emissions from motor vehicles.The most important single lead input to the urban environment in 1975 was held to be the lead emissions from the combustion of petrol containing lead additives (around 0.45 g l21 at that time). Following much public pressure and concern, an EC Directive10 in March, 1985, required member states to reduce the lead content of leaded petrol to 0.15 g l21 as soon as possible, and to introduce a supply of unleaded petrol by Table 1 Measurements of lead in Manchester street dust; average concentrations (mg g21) and numbers of samples, with data subdivided according to the type of locality (A to D*).Comparison between the 1975 and the 1997 surveys Mean† Pb concentration/mg g21 (number of samples) Category 1975 survey 1977 survey A 1001 ± 40 (180) 577 ± 53 (17) B 888 ± 57 (68) 594 ± 55 (17) C 933 ± 186 (53) 536 ± 93 (13) D 1014 ± 206 (49) 572 ± 77 (8) A–C 941 ± 19 (301) 569 ± 37 (47) * Definition of the categories: A, major roads/street; moderate/heavy traffic ( > 100 cars per hour); B, minor roads/streets; light/moderate traffic (10–100 cars per hour); C, street with very light traffic only ( < 10 cars per hour); D, school playgrounds, children’s play areas in parks, garden, wasteland. † Means for individual categories as between 1975 and 1997 are all significantly different (P < 0.01). 60 Analyst, January 1998, Vol. 123October, 1989. In the UK, the progressive reduction of the lead content of petrol from 1985 resulted in a fall in lead emissions by over 60% within one year.12,13 Unleaded petrol first became available in the UK in 1986, and currently nearly 40% of all petrol consumed is unleaded13 (see Fig. 2). Although both the number of motor vehicles and the total fuel consumption has increased since 1986, fuel efficiency has also tended to improve and the proportion of diesel engined vehicles has increased.The overall results is that total lead emissions from motor vehicles in the UK in 1997 are about 20% of those in 1975, and one might have expected that trends in Manchester would have followed those for the UK in general. However, the fall in street dust lead, to around 60% of the 1975 value, is only half of that which would have been expected from the trends in vehicle emissions, and a more complex explanation must be sought. Clearly, one possibility is that up to half of the street dust lead in 1975 arose from non-vehicle sources.Another is that vehicle lead emissions in Manchester (or large cities in general) have not reduced as much as emissions nationally, perhaps if traffic densities in cities have risen disproportionately. We do not at present have enough information to elucidate this question further. The Paris survey The Paris survey results are presented in Table 2. The average value for the street locations (A–C) is 1273 ± 212 mg g21, more than twice the Manchester value.There are a number of interesting features about the Parisian results. Firstly, the overall range (84 to > 5000 mg g21) is much greater than in Manchester (see Fig. 1). Secondly, Pb levels in streets in category B are significantly higher (P < 0.02) than those in either categories A or C. And thirdly, Pb levels in parks and play grounds (category D) are significantly lower (P < 0.002) than any category of street, and also significantly lower than the same category samples from Manchester.The major differences between lead levels and distribution in Manchester and Paris are not easy to explain, although there are three obvious differences between the types of location sampled in Manchester and in Paris. Firstly, from qualitative observation, traffic densities in Paris are generally far higher. Secondly, streets in category B in Paris tend to be much narrower than both streets in category A (in Paris), and streets in both categories A and B in Manchester.And thirdly, street cleaning in Paris is apparently much more frequent and rigorous than in Manchester, and often involves spraying with water. Additionally, streets in the differing categories are cleaned in different ways. It was particularly noteworthy that it was very difficult to collect reasonable quantities of dust from many streets in Paris, particularly those in category A. The differences between categories A, B and C in Paris, and between Paris and Manchester generally, may result from these causes although it is difficult to advance totally plausible explanations.Additionally, and unlike Manchester, there are no previous surveys for Paris available to which we can refer. Therefore, we are unable to say whether lead concentrations in Paris have changed in the past twenty years, although it is likely that this will have occurred because France has been submitted to the same changes in European standards as the United Kingdom.Lead concentrations measured in samples taken from under and around the Eiffel Tower (category E) range from 333 to 4282 mg g21, but should be considered separately. There is no heavy traffic where the samples were collected (the area under the tower and Champ de Mars), but it might be supposed that the lead found there would have originated from the paint used for the Eiffel Tower (which is wrought ironwork, almost certainly painted until recently with red lead oxide paint).Even if the lead percentage in the paint now used is low, the quantity needed to paint the whole monument is so great that lead concentrations nearby could be very high. In order to test the hypothesis that the environmental lead from the vicinity of the Eiffel Tower originates from a source distinct from the streets of the rest of central Paris, isotope ratio measurements have been carried out to categorise the lead from the various sample locations.Lead isotope ratios Dust samples from Paris and Manchester streets and from the vicinity of the Eiffel Tower were subjected to lead isotope ratio determination by ICP-MS. There are four lead isotopes (mass numbers 204, 206, 207 and 208) which can be determined, and Fig. 1 Sample distribution by lead concentration: comparison between the Manchester and Paris (1997) surveys (22% of the Paris samples had Pb concentrations in the range 1700–5500 mg g21; these are not shown in the distribution).Fig. 2 Evolution of lead emissions and the fuel consumption in the UK over the past twenty years. Symbols: lead emissions (2); fuel consumption (filled symbols), leaded petrol (5); unleaded petrol (-); diesel fuel (!). Table 2 Lead measurements for Paris (1997); average concentrations and numbers of samples for the various types of location (A–E*) Probability (P) of the Mean Pb difference concentration/ between the mg g21 categories (number of occurring by Category samples) chance A 718 ± 82 (15) A–B: P = 0.011 B 2254 ± 501 (12) B–C: P = 0.016 C 846 ± 149 (8) A–C: P = 0.468 D 176 ± 40 (6) C–D: P = 0.002 E 1686 ± 744 (5) A–C 1273 ± 212 (35) * A, B, C, D are the same categories as for Manchester (see Table 1), E corresponds to the samples collected around and under the Eiffel Tower.Analyst, January 1998, Vol. 123 61the relative quantities in natural lead deposits vary according to the radiogenic characteristics of their location, although the ratio 206Pb : 207Pb is the most commonly used to characterise lead sources.16 Typical samples from the three groups (but all samples from the Eiffel Tower category, E) are plotted in Fig. 3.The results clearly form three clusters, each of which corresponds closely to one of the sample sets. We conclude that the lead content in each set comes from a different source: particularly, Eiffel Tower lead (206Pb : 207Pb = 1.174) can be distinguished isotopically from lead in the streets in the remainder of Paris (206Pb : 207Pb = 1.151), and Paris street lead can further be distinguished from Manchester lead (206Pb : 207Pb = 1.107).The results are consistent with the hypothesis that high lead concentrations under the Eiffel Tower are due to the lead contained in paint. The lead contained in the other two sets comes mostly from lead additives in petrol, but their differing isotope ratios reflect the fact that whilst the UK imports most of its lead from Broken Hill, Australia (typically, 206Pb : 207Pb = 1.037), the remainder of Europe mostly uses lead from Missouri, USA (where the lead is more radiogenic, typically 206Pb : 207Pb = 1.385).17 These results confirm the view that lead isotope ratios provide an effective and convenient method for tracing the origins of environmental lead contamination.We gratefully acknowledge the support (to S.M.N.) of the European Union ERASMUS/SOCRATES student exchange programme.References 1 Department of the Environment Central Unit on Environmental Pollution, Lead in the Environment and its Significance to Man, a report of an Inter-Departmental Working Group on Heavy Metals, HMSO, London, 1974. 2 Lawther, D., Lead and Health, report of a DHSS Working Party on lead in the environment, HMSO, London, 1980. 3 Bryce-Smith, D., and Stephens, R., Lead or Health, Conservation Society, UK, 1981. 4 Committee on Biologic Effects of Atmospheric Pollutants, Lead: Airborne Lead in Perspective, Division of Medical Sciences, National Research Council USA, Washington, DC, 1972. 5 Saryan, L. A., and Zenz, C., in Occupational Medicine, ed. Zenz, C., Dickerson, O. B., and Horvath, E. P., Mosby, St. Louis, MO, 3rd edn., 1994, pp. 506–541. 6 Stroemberg, U., Schutz, A., and Skerfving, S., Occup. Environ. Med., 1995, 52, 764–9. 7 Farmer, P., Lead Pollution from Motor Vehicles 1974–1986: a Selected Bibliography, Elsevier, Amsterdam, 1987. 8 Day, J. P., Hart, M., and Robinson, M. S., Nature, 1975, 253, 343. 9 Department of the Environment Central Unit on Environmental Pollution, Lead Pollution in Birmingham, report of the Joint Working Party on Lead Pollution around Gravelly Hill, HMSO, London, 1978. 10 EEC Directive, 85/210, 1985, No. L 96/25. 11 Barret, B., and Howells, R., Sci. Total Environ., 1984, 33, 1. 12 Transport Statistics Report: Road Traffic Statistics GB, HMSO, London 1996. 13 Central Statistical Office, Annual Abstract of Statistics, HMSO, London, 1985 and 1996 edns. 14 Alchalabi, A. S., Hawker, D., Sci. Total Environ., 1996, 187, 105. 15 Church, H. J., PhD Thesis, University of Manchester, 1991. 16 Chow, T. J., Snyder, C. B., and Earl, J. L., in Isotope Ratios as Pollutant Sources and Behaviour Indicators, 1975, Proceedings of a Symposium, Vienna, 1974, IAEA and FAO, pp. 95–108. 17 Hurst, R. W., Davis, T. E., and Chinn, B. D., Environ. Sci. Technol., 1996, 30, 304A. Paper 7/04940D Received July 10, 1997 Accepted September 29, 1997 Fig. 3 Lead isotope ratios for samples taken from: Manchester streets (5); Paris streets (“); and the vicinity of the Eiffel Tower (-). 62 Analyst, January 1998, Vol. 123

ISSN:0003-2654    

DOI:10.1039/a704940d

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Selective uptake of selenite by red blood cells† Kazuo T. Suzuki*a, Yamato Shiobaraa, Makiko Itoha and Masayoshi Ohmichib a Faculty of Pharmaceutical Sciences, Chiba University, Inage, Chiba 263-0022, Japan b Chiba City Institute of Health and Environment, Mihama, Chiba 261-0001, Japan Both organic and inorganic forms of selenium (Se) can be utilized in the body, and the biotransformation of selenite into an organic form of Se in the bloodstream is the first step for the utilization of inorganic Se.Selenite injected intravenously into rats was shown to be taken up rapidly and selectively by red blood cells (RBCs) through the anion-exchange carrier. The uptake of selenite by RBCs was inhibited by 4,4A-diisothiocyano-2,2A-stilbene disulfonate, a specific inhibitor of the anion-exchange carrier (band 3 protein). The uptake was also inhibited by chromate owing to the glutathione deprivation in RBCs, which was confirmed by the inhibition by azodicarboxylic acid bis(dimethylamide).The presence of hydrogencarbonate in the incubation solution slightly retarded the uptake of selenite by RBCs. Although Se effluxed into the plasma was bound selectively to albumin, plasma proteins (albumin) did not accelerate the uptake process. Based on these results, the rapid and selective uptake of selenite by RBCs was explained by the selective and efficient uptake through the anion-exchange carrier, followed by reduction by glutathione. Keywords: Selenite; anion-exchange carrier; red blood cells; band 3 protein; selenium; inductively coupled plasma mass spectrometry Selenium (Se) is an essential element, and both inorganic and organic Se can be utilized as nutrients in the body.An inorganic form of Se, selenite, is reduced by glutathione (GSH) in the body owing to its intrinsic low activation potential.1,2 The reduced forms of Se, Se0 and/or Se22, are effluxed from red blood cells (RBCs), bound to albumin3 and then transferred to the liver.4 In the liver, Se is incorporated into selenoproteins through selenocysteinyl tRNA, which is synthesized through the transfer of selenol groups from selenophosphate to seryl tRNA,5,6 or excreted after being methylated sequentially to monomethylselenol, dimethylselenide and trimethylselenonium ions.7,8 Therefore, the metabolic reactions of Se in the body can be divided into three categories, namely reduction, protein synthesis and methylation.Thus, selenite is incorporated into selenoproteins or methylated to metabolites.However, little is known about the reduction of selenite before being biotransformed to organic forms. Our recent study on the metabolic fate of selenite injected intravenously (i.v.) into rats revealed that it disappears rapidly from the plasma, is taken up by the RBCs and then reappears in the plasma after being reduced in the RBCs.9 Although the reduced form of Se has not yet been identified, it was shown to bind selectively to albumin,3,9 and the Se is then transferred to the liver.The anion-exchange carrier (band 3 protein) is the major integral membrane protein of RBCs and catalyses the passage of anions across the membrane.10,11 This carrier protein is specifically inhibited by 4,4A-diisothiocyano-2,2A-stilbene disulfonate (DIDS) through binding to the external transport site of the band 3 protein.12 The band 3 protein exchanges carbonate with chloride in RBCs under physiological conditions.As other anions can also act as substrates for the band 3 protein, the rapid and selective uptake of selenite by RBCs is considered to be mediated through this carrier. In this study, the mechanism underlying the uptake of selenite by RBCs was investigated in vivo and in vitro and the uptake of selenite was shown to take place through the anion-exchange carrier (band 3 protein). Experimental Reagents Sodium selenite enriched with 82Se was prepared by oxidation of the enriched metal (97.02% enriched; Oak Ridge National Laboratory, Oak Ridge, TN, USA).Oxidation of the metal was achieved by dissolving the metal in concentrated metal-free nitric acid and subsequent neutralization with 1 m NaOH.9,13 Sodium selenite (natural abundance) and azodicarboxylic acid bis(dimethylamide) (Diamide) were purchased from Sigma (St. Louis, MO, USA). 4,4A-Diisothiocyano-2,2A-stilbene disulfonic acid disodium salt (DIDS), sodium hydrogencarbonate, sodium chromate and trifluoroacetic acid (TCA) of analytical-reagent grade were obtained from Wako (Osaka, Japan), pentobarbital from Takeda Pharmaceutical (Osaka, Japan) and heparin from Dainippon Pharmaceutical (Osaka, Japan).Animal experiments Male Wistar rats were purchased at 8 weeks of age from a breeder (Clea Japan, Tokyo, Japan), and were fed a commercial diet (CE-2; Clea Japan) and tap water ad libitum. In in vitro experiments, RBCs were separated from heparinized blood, re-suspended in Dulbecco’s phosphate-buffered saline (PBS) and then centrifuged at 1060g for 15 min to remove residual plasma.Analytical procedures The concentrations of Se in whole blood and plasma were determined, after wet ashing with a mixed acid [HNO3–HClO4 (4 + 1 v/v)], at m/z 78 for natural abundance Se and m/z 82 for enriched Se by ICP-MS (HP 4500; Yokogawa Analytical Systems, Musashino, Japan). A 0.1 ml aliquot of plasma was applied to a size exclusion column (Asahipak GS 520, 500 3 7.6 mm id; Showa Denko, Tokyo, Japan), and then subjected to HPLC (Model 576; GL Sciences, Tokyo, Japan) with 50 mm TRIS–HCl buffer (pH 7.4) as eluent at a flow rate of 1.0 ml min21.The eluate was monitored with a UV detector and then introduced directly into the nebulizer tube of the ICP-MS system to detect Se (m/z 78 and 82). The concentrations of Se in the incubation solution were determined by ICP-MS using the flow injection method for sample introduction.The flow rate was maintained at 1.0 ml min21. A 0.1 ml aliquot of each sample solution was applied. † Presented at The Sixth Nordic Symposium on Trace Elements in Human Health and Disease, Roskilde, Denmark, June 29–July 3, 1997. Analyst, January 1998, Vol. 123 (63–67) 63Results Rapid and selective uptake of selenite by RBCs The concentrations of Se in whole blood and plasma were determined after a single i.v. injection of 82Se-enriched selenite at a dose of 125 mg Se kg21 body mass into a rat.The concentration of Se in RBCs was calculated by subtracting the concentration of Se in plasma from that in the whole blood. The concentration of 82Se-enriched selenite injected i.v. was estimated to represent a concentration of 2.5 mg ml21 of Se in whole blood on the assumption that the blood volume comprises 5% of the body mass. Therefore, 82Se was found to be present mostly in the bloodstream at 1 min after injection (Fig. 1). However, 82Se was not present in plasma but distributed in RBCs, indicating that selenite was taken up selectively by RBCs within 1 min. The concentrations of 82Se in whole blood and RBCs decreased with time after 1 min and 82Se disappeared from the bloodstream within 10 min (Fig. 1). During the disappearance process, 82Se was detected at only a low concentration in plasma, suggesting that the 82Se effluxed into the plasma was cleared promptly from the bloodstream. The chemical forms of 82Se in plasma were determined in vitro after incubating 82Se-enriched selenite in heparinized whole blood by the HPLC–ICP-MS method, as shown in Fig. 2. 82Se mixed with whole blood in the form of selenite was recovered in plasma and eluted as a single Se peak before incubation. However, after incubation for 1 min, 82Se recovered from plasma comprised only a small amount, and it was distributed in the high molecular mass protein fraction, i.e., albumin and selenite, indicating that [82Se]selenite was mostly taken up by RBCs and the 82Se recovered in the plasma was present mostly in chemical forms other than selenite.After incubation for 10 min, 82Se again appeared in the plasma, where it was mostly bound to albumin. Selenite incubated in plasma for 10 min without RBCs remained without any change (data not shown). Uptake of selenite by RBCs through anion-exchange carrier (band 3 protein) The mechanisms underlying the rapid and selective uptake of selenite by RBCs were studied by observing the effect of DIDS, a specific inhibitor of the anion-exchange carrier (band 3 protein), on the uptake of [82Se]selenite by RBCs, as shown in Fig. 3. RBCs were pre-treated with 0, 5 and 50 mm DIDS and then the effect of the pre-treatment on the uptake of selenite Fig. 1 Uptake of selenite by RBCs after a single intravenous injection of selenite into a rat. A male Wistar rat (body mass 330 g) was cannulated at an artery and vein of a lower limb under pentobarbital anesthesia, and then the rat received a single i.v.injection of 82Se-enriched selenite at a dose of 125 mg Se kg21 body mass. Heparinized blood (approximately 0.5 ml each) was obtained at 1, 3, 5, 10, 30 and 60 min after the injection, and plasma was separated by centrifugation at 8000g for 10 s. The concentrations of Se in plasma and RBC were expressed as ng ml21 of whole blood, the concentrations of Se in RBCs (2) were calculated by subtracting the concentrations in plasma (.) from the concentrations in whole blood (8).Fig. 2 Changes in the distribution of 82Se in plasma on incubation of 82Seenriched selenite with whole blood, as judged with the HPLC–ICP-MS method. Heparinized blood was incubated with 82Se-enriched selenite at a concentration of 30 ng Se ml21 blood at 37 °C for 1 and 10 min, and then centrifuged at 8000g for 10 s to give the plasma. The distribution of Se at m/z 82 in plasma was determined on a gel filtration column by the HPLC– ICP-MS method. Control profiles of Se (0 min) and sulfur (S) at m/z 34 were obtained without incubation.Fig. 3 Time- and concentration-dependent inhibition of the uptake of selenite into RBCs by 4,4A-diisothiocyano-2,2A-stilbene disulfonate (DIDS). RBCs suspended in isotonic saline at 10% v/v and were incubated at 37 °C for 1 min, and then DIDS in 10 mm phosphate buffer (pH 7.4), was added to the suspension of RBCs to a final concentration of 0 (8), 5 (2) and 50 mm (5) circles (A) or 0, 5, 10, 20, 40 and 60 mm (B), followed by incubation at 37 °C for 1 min. To the pre-incubated suspension, selenite (natural abundance) was added to final concentrations of 20 (a) and 100 mm (b) in A or 20 mm in B.RBCs were removed by centrifugation at 8000g for 10 s and the concentration of Se in the supernatant was determined by ICP-MS. The data in B are presented as means ± s (n = 3). Asterisks indicate significantly different from the control at P < 0.05. 64 Analyst, January 1998, Vol. 123(natural abundance Se) was observed as the decrease in Se from the incubation solution. Se disappeared from the incubation solution with time on incubation with RBCs after pre-treatment without or with DIDS at 5 mm [Fig. 3(A)]. However, Se remained in the incubation solution, not being taken up by RBCs that had been treated with DIDS at 50 mm [Fig. 3(A)]. The uptake of selenite by RBCs occurred within 3 min under the present in vitro conditions at both selenite concentrations examined (20 and 100 mm).Therefore, the effect of the dose of DIDS on the uptake by RBCs was examined after incubation for 3 min. RBCs were pre-treated with various concentrations of DIDS and then the concentrations of Se in the incubation solutions were determined to observe the dose-related effect, as shown in Fig. 3(B). The ED50 of DIDS for the inhibition of the uptake of selenite by RBCs was calculated to be 14.5 mm. Effect of hydrogencarbonate on the uptake of selenite by RBCs The effects of pre-treatment with hydrogencarbonate on the uptake of selenite by RBCs were examined as shown in Fig. 4. Incubation of selenite at concentrations of 20 and 100 mm with RBCs gave similar time-course changes for the interaction with hydrogencarbonate, indicating that the uptake of selenite was not dependent on the concentration of selenite at concentrations lower than 100 mm under the present conditions [Fig. 4(A)].The uptake (decrease in the concentration of Se from the incubation solution) had almost plateaued by 3 min after the incubation in the control and pre-treated groups with hydrogencarbonate at 5 mm, while it was retarded by a higher concentration of hydrogencarbonate, i.e., at 25 mm [Fig. 4(A)]. The dose-dependent effect of hydrogencarbonate on the uptake of selenite indicated an inhibition of ED50 = 21.8 mm [Fig. 4(B)]. As the dissolution of hydrogencarbonate may change the pH of the incubation solution to alkaline and hence may affect the uptake of selenite, the effect of pH on the uptake of selenite was examined, as shown in Fig. 5. An alkaline pH similar to that of 25 mm hydrogencarbonate in Fig. 4(A) (pH 8.4) inhibited the uptake only slightly. Effect of chromate on the uptake of selenite by RBCs The effect of pre-treatment with chromate on the uptake of selenite by RBCs was examined, as shown in Fig. 6. The timedependent changes of the pre-treatment shown in Fig. 6(A) indicate that although the uptake was inhibited by chromate at 20 and 100 mm, it plateaued by 3 min with incubation of selenite at concentrations of both 20 and 100 mm [Fig. 6(A) (a) and (b), respectively]. The dose-dependent effects of pre-treatment with various concentrations of chromate on the uptake of selenite by RBCs (decrease in concentration of Se in the incubation solution) were determined, as shown in Fig. 6(B). The uptake of selenite was inhibited on incubation of chromate, i.e., at ED50 = 4.5 mm.Effect of GSH deprivation on the uptake of selenite by RBCs As the underlying mechanism for the inhibition of the uptake by chromate, deprivation of GSH on oxidation by chromate may be possible.14 Therefore, the effect of GSH deprivation in RBCs was determined in the presence of azodicarboxylic acid bis(dimethylamide) (Diamide), as shown in Fig. 7. The time courses of the uptake of selenite by RBCs were plotted against the incubation time in the presence of various concentrations of Diamide, as shown in Fig. 7(A). The uptake was inhibited completely by Diamide at concentrations higher than 1 mm. The effect of the dose of Diamide on the uptake was also determined after incubation for 3 min at 37 °C, as shown in Fig. 7(B). The uptake was completely inhibited in the presence of Diamide at 0.58 mm, the ED50 being calculated to be 0.51 mm. Fig. 4 Effect of hydrogencarbonate (NaHCO3) on the uptake of selenite by RBCs.RBCs were suspended in isotonic saline at 10% v/v and were incubated at 37 °C for 1 min. In the time-dependent experiment (A), NaHCO3 in 10 mm phosphate buffer (pH 7.4) was added to the RBC suspension to final concentrations of 0 (8), 5 (2) and 25 mm (5) and then selenite was added to final concentrations of 20 (a) and 100 mm (b). The suspensions were incubated at 37 °C for up to 8 min. In the dose-dependent experiment (B), various concentrations of NaHCO3 in phosphate buffer (pH 7.4) were added to the suspension and then selenite was added to the final concentration of 20 mm.The suspensions were incubated at 37 °C for 3 min. RBCs were removed by centrifugation at 8000g for 10 s and then the concentration of Se in the supernatant was determined by ICP-MS. The data in B are presented as means ± s (n = 3). Asterisks indicate significantly different from the control at P < 0.05. Fig. 5 Effect of pH on the uptake of selenite by RBCs. RBCs were suspended in 50 mm TRIS–HCl buffer (10% v/v) containing 0.9% NaCl at pH 7.4 (8), 7.9 (2) or 8.4 (5).Each suspension was pre-incubated at 37 °C for 1 min, 20 mm selenite (final concentration) was added and then the suspensions were incubated at 37 °C for up to 8 min. RBCs were removed by centrifugation at 8000g for 10 s and then the concentrations of Se in the supernatants were determined by ICP-MS. Analyst, January 1998, Vol. 123 65Effect of plasma proteins on the uptake of selenite by RBCs Selenium taken up by RBCs and altered chemically in the RBCs was shown to be effluxed from RBCs and bound selectively to albumin, as shown in Fig. 1. Therefore, plasma proteins, especially albumin, may affect (accelerate) the uptake of selenite by RBCs. The effect of the presence of plasma proteins in the incubation solution on the uptake of selenite by RBCs was examined, as shown in Fig. 8. The incubated solution was treated with TCA to remove Se that had effluxed and bound to proteins.The uptake of selenite (decrease in concentration of Se in the TCA-treated incubation solution) was not accelerated but rather retarded, although only slightly under the present conditions. Discussion The HPLC–ICP-MS method has been applied to speciate Se in biological samples.9,13,15,16 In this study, an enriched stable isotope, 82Se, was used to trace exogenous Se in vivo and in vitro, and the chemical form of Se administered as [82Se]selenite was speciated by the HPLC–ICP-MS method. The results in Figs. 1 and 2 indicate that selenite is taken up rapidly (within 1 min) and selectively by RBCs in vivo without being distributed to other organs and also in vitro under the present conditions. The selenite taken up by RBCs was metabolized (reduced) in them (within 10 min), and then effluxed into the plasma, where Se was selectively bound to albumin. Although Se effluxed from RBCs into the plasma was cleared promptly from the plasma in vivo (Fig. 1), it was present bound to albumin in vitro (Fig. 2). The difference in the recovery of Se from plasma between in vivo and in vitro experiments further indicates that Se bound selectively to albumin was cleared promptly from the bloodstream (plasma), suggesting that the efflux from RBCs is the rate-limiting step in the disappearance of selenite from the bloodstream. The rapid and selective uptake of selenite by RBCs was assumed to occur through the anion-exchange carrier (band 3 Fig. 6 Time- and concentration-dependent inhibition of the uptake of selenite into RBCs by chromate (Na2CrO4). RBCs suspended in isotonic saline at 10% v/v were incubated at 37 °C for 1 min and then Na2CrO4 in 10 mm phosphate buffer (pH 7.4) was added to the preincubated RBCs to a final concentration of 0 (8), 5 (2) or 10 mm (5) (A). In the dose-dependent experiment (B), Na2CrO4 was added to final concentrations of 0, 0.5, 1, 2, 5, 10 and 15 mm and the mixtures were incubated at 37 °C for 1 min.To the pre-incubated suspension, selenite (natural abundance) was added to final concentrations of 20 (a) and 100 mm (b) in A or 20 mm in B. RBCs were removed by centrifugation at 8000g for 10 s and the concentrations of Se in the supernatants were determined by ICP-MS. The data in B are presented as means ± s (n = 3). Asterisks indicate significantly different from the control at P < 0.05. Fig. 7 Effect of deprivation of GSH by Diamide on the uptake of selenite by RBCs.RBCs suspended in isotonic saline at 10% v/v were incubated at 37 °C for 1 min and then Diamide in 10 mm phosphate buffer (pH 7.4) was added to the pre-incubated RBCs to a final concentration of 0 (8), 0.2 (.), 1 (2) or 2 mm (5) (A) or 0, 0.1, 0.5, 1.0, 1.5 or 2.0 mm (B) and then the mixtures were incubated at 37 °C for 1 min. To the pre-incubated suspensions, selenite (natural abundance) was added to a final concentration of 20 mm and then the solutions were incubated for up to 8 min (A) or 3 min (B).RBCs were removed by centrifugation at 8000g for 10 s and then the concentrations of Se in the supernatants were determined by ICP-MS. The data in B are presented as means ± s (n = 3). Asterisks indicate significantly different from the control at P < 0.05. Fig. 8 Effect of plasma proteins (albumin) on the uptake of selenite by RBCs. RBCs were suspended in isotonic saline (10% v/v) containing plasma at 0 (8), 30 (2) or 60% (5) and then incubated in the presence of selenite at the final concentration of 20 mm at 37 °C for up to 8 min.RBCs were removed by centrifugation at 8000g for 10 s and then the incubated solution was treated with TCA (15%) to remove Se bound to proteins. The concentrations of Se in the supernatants were determined by ICP-MS. 66 Analyst, January 1998, Vol. 123protein), and it was examined by means of inhibition experiments with DIDS, a specific inhibitor of the anion-exchange carrier through binding to the active site,12 as shown in Fig. 3. Under the present conditions, the uptake of selenite by RBCs occurred within 3 min with any dose of selenite [Fig. 3(A)]. The uptake was inhibited in a dose-dependent manner and the ED50 was calculated to be 14.5 mm from Fig. 3(B). Carbonates are the major substrates of the anion-exchange carrier under physiological conditions. We therefore examined whether or not hydrogencarbonate interferes with selenite ions in the uptake process.The presence of hydrogencarbonate ions was shown to retard the uptake of selenite by RBCs at ED50 = 21.8 mm under the present in vitro conditions [Fig. 4(B)]. Chromate is known to be reduced by glutathione (GSH) after being incorporated into cells,17 which makes the cells deficient in the reduced form of GSH.14 Selenite is reduced in RBCs by GSH owing to its intrinsic low redox potential. However, selenite taken up by RBCs is considered not to be reduced under GSH-deficient conditions and it may disturb the uptake of selenite by RBCs.In fact, the uptake of selenite was shown to be inhibited on pre-treatment of RBCs with chromate [Fig. 6(B)]. It was further suggested that the decrease in the concentration of GSH in RBCs was the primary cause of the inhibition of the uptake of selenite through the deprivation of the reduced form of GSH by Diamide (Fig. 7), which oxidizes acidic low molecular mass thiols in preference to protein thiols.18 As a result, selenite was shown to be taken up selectively (Fig. 1) by RBCs through the anion-exchange carrier (band 3 protein) (Fig. 3). Although the Se effluxed from RBCs was present bound selectively to albumin in vitro (Fig. 2), plasma proteins were shown to affect the uptake of selenite by RBCs to only a limited extent (Fig. 8). Although the uptake was also inhibited slightly in the presence of hydrogencarbonate, this was suggested to be caused partly by the shift of the pH to alkaline on the dissolution of sodium hydrogencarbonate (Fig. 5). Selenium effluxed into the plasma disappeared promptly from the bloodstream in vivo and the concentration of Se in the plasma remained low throughout (Fig. 1), whereas the effluxed Se was present bound to albumin in vitro (Fig. 2), indicating that the Se effluxed and bound to albumin is transferred to the liver,3,4 in the form of an albumin–Se complex. Furthermore, the step for the chemical modification (metabolism) of selenite, i.e., reduction to selenide by GSH or the efflux step, can be assumed to be the rate-limiting step for the disappearance of Se from the bloodstream.As GSH deprivation effects the uptake of selenite (Fig. 7), the former step may be the rate-limiting step. The uptake of selenite was not retarded in the absence of plasma proteins, as demonstrated in Fig. 8, indicating that the uptake by the efflux from RBCs are independent steps. Summarizing the present observations, selenite is rapidly and selectively taken up by RBCs through the anion-exchange carrier (band 3 protein), metabolized (reduced) in the RBCs and then effluxed into the plasma, where it binds selectively to albumin.Depletion of GSH inhibits the uptake whereas plasma proteins do not. Selenium bound to albumin is promptly cleared from the plasma. References 1 Hsieh, H. S., and Ganther, H. E., Biochemistry, 1975, 14, 1632. 2 Gasiewicz, T. A., and Smith, J. C., Chem.-Biol. Interact., 1978, 21, 299. 3 Sandholm, M., Acta Pharmacol. Toxicol., 1975, 36, 321. 4 Sandholm, M., Acta Pharmacol. Toxicol., 1973, 33, 1. 5 Burk, R. F., and Hill, K. E., Annu. Rev. Nutr., 1993, 13, 65. 6 Veres, Z., Kim, I. Y., Scholz, T. D., and Stadtman, T. C., J. Biol. Chem., 1994, 269, 10 597. 7 Ganther, H. E., J. Am. Coll. Toxicol., 1986, 5, 1. 8 Zeisel, S. H., Ellis, A. L., Sun, X. F.,Pomfret, E. A., Ting, B. T. G., and Janghorbani, M., J. Nutr., 1987, 117, 1609. 9 Suzuki, K. T., and Itoh, M., J. Chromatogr. B, 1997, 692, 15. 10 Steck, L. T., Ramos, B., and Strapazon, E., Biochemistry, 1976, 15, 1154. 11 Appell, C. K., and Low, S. P., Biochemistry, 1982, 21, 2151. 12 Ramjeesingh, M., Gaarn, A., and Rothstein, A., Biochim. Biophys. Acta, 1981, 641, 173. 13 Suzuki, K. T., Yoneda, S., Itoh, M., and Ohmichi, M., J. Chromatogr. B, 1995, 670, 63. 14 Aaseth, J., Alexander, J., and Norseth, T., Acta Pharmacol. Toxicol., 1982, 50, 310. 15 Suzuki, K. T., Itoh, M., and Ohmichi, M., J. Chromatogr. B, 1995, 666, 13. 16 Suzuki, K. T., Itoh, M., and Ohmichi, M., Toxicology, 1995, 103, 157. 17 Alexander, J., and Aaseth, J., Analyst, 1995, 120, 931. 18 Kosower, N. S., and Kosower, E. M., Methods Enzymol., 1995, 251, 123. Paper 7/06230C Received August 26, 1997 Accepted September 19, 1997 Analyst, January 1998, Vol. 123 67

ISSN:0003-2654    

DOI:10.1039/a706230c

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Nickel, cobalt, zinc and copper levels in brown trout ( Salmo trutta) from the river Otra, southern Norway† Rory M. Brotheridgea, Kenneth E. Newtonb, Mark A. Taggartc, Paul H. McCormickc and Stuart W. Evans*a a Division of Clinical Sciences, School of Medicine, University of Leeds, Leeds, UK LS2 9JT b Department of Chemical Pathology, Leeds General Infirmary, Leeds, UK LS1 3EX c School of Earth Sciences, University of Leeds, UK LS2 9JT The Flåt Nickel mine at Evje in southern Norway was mined extensively from 1914 to 1945 with little regard for any potential environmental effect.Much of the ore extracted was smelted at a site adjacent to the river Otra south of Evje. Recent studies have revealed heavy metal pollution in the land surrounding the smelter and in water draining from the mine leading to concern for the aquatic ecosystem in the river Otra. Brown trout were sampled from an uncontaminated lake 9 km upstream from the smelter, from the base of the Oddebekken (a tributary draining the mine water into the Otra), from sites immediately upstream and down stream of the smelter and from a site 4 km down stream from the smelter.Fish from sites adjacent to the smelter and the base of the Oddebekken were smaller than those from the lake and down stream site. Concentrations of the metals were highest in fish sampled where the mine water entered the Otra and gradually decreased in fish further down the river. Fish from the uncontaminated lake had the lowest level of metals.Keywords: Brown trout; nickel; cobalt; copper; zinc; liver; muscle; river; Norway The Otra river valley in the Evje region of southern Norway lies on metamorphic bedrock with diorite intrusions and associated Ni mineralisation. Ni ore was worked intensively from about 1914 to 1945. The Flåt mine situated above the town of Evje was once the biggest in Europe. Much of the extracted ore, estimated at 2 000 000 tons, was processed in a smelter south of Evje.1 Throughout the operation of both mine and smelter little attention was paid to the potential environmental effects of the mining and smelting of the metals.There is now increasing concern over the possibility of toxic metals leaching from the mine tailings and from the mines themselves. Little succession has occurred in the flora at the mine site as a result of the pollution. Further studies have found elevated levels of four metals; Ni, Co, Cu and Zn in the wells used to provide local drinking water leading to suggestions that the water table has been contaminated by drainage from the mine and slag heaps.1 A detailed knowledge of the mineralogy of this area suggests that Hg and As would not be expected to be present as significant pollutants.Neither metal was detected in a preliminary analysis of water samples.1 Concentrations of Al were elevated but did not exceed reported lowest known biological effect concentrations. Other work has established that the organic horizon of soil has been contaminated around the smelter and that this contamination extends for up to 1.5 km from the smelter.2,3 High concentrations of all four metals were found and the soil concentrations of three of the metals Ni, Zn and Cu exceeded the threshold values recommended by ICRCL.1,4,5 Each of these metals is potentially toxic to the aquatic ecosystem in the river Otra.In particular there is concern that the metals accumulate in fish.Ni accumulates in the gills, kidney, liver, brain and white muscle of fish.6 Sublethal concentrations of Ni have been reported to lead to changes in behaviour such as aggression and stress related movement.7 The resultant stress may also be responsible for muscle glycogenolysis observed in some Ni poisoned fish.8 Haematological parameters such as total and differential leucocyte count, erythrocyte count, haemoglobin concentration and hematocrit are also affected by Ni accumulation.7 Accumulation of Co in muscle can adversely affect muscle function.A number of other systems in fish such as the lateral line system sense organs have also been reported to be affected by Co.9 Zn is generally thought to be non-toxic except at very high concentrations10,11,12 and trout are believed to acclimatise to high concentrations of Zn in their environment.13 It has also been suggested that liver Zn concentrations may correlate with induction of metallothionein and thus act as a biomarker of exposure to heavy metals.14 Exposure to Cu has been observed to interfere with lipid metabolism in the gills,15 to induce apoptotic and necrotic lesions in the basal pavement cells of skin,16 and apoptosis of the olfactory epithelium.17 Fish sampled from a Cu contaminated lake were reported to have a high incidence of tumours.18 Cu bioaccumulates in the brain and liver of fish.19 Accumulation in the liver can result in liver necrosis and in the brain to neurotoxicity.20,21 The Otra is a fertile river with the diversity of habitats required to support thriving communities of trout, perch and non-migratory salmon.As all four metals polluting the Otra river valley have the potential to be toxic, there is concern for the effects of the metals on fish populations and hence the local commercial and recreational fishing community. In this study fish sampled from several sites in the river Otra have been examined for the presence of Ni, Co, Zn and Cu.Experimental Collection of fish Fish were collected from five sites along the river Otra over a period of one week (Fig. 1), using 20–24 mm gill nets positioned in the early afternoon and left overnight to be collected at 9 am in the morning. The length of each fish was recorded. Fish from site B, C, D and E were immediately placed in a labelled plastic bag and frozen. In the laboratory, fish were thawed and weighed, the liver was removed and weighed separately.The operculum was removed to determined age group. Fish from site A were dissected immediately after capture. The operculum, samples of muscle tissue and the liver were removed and frozen. Analysis of metals in fish tissues Approximately 0.5 g of muscle and liver were removed from each fish, weighed and dissolved in 5 ml of concentrated HNO3 in individual vials. The dissolved tissue was then analysed for † Presented at The Sixth Nordic Symposium on Trace Elements in Human Health and Disease, Roskilde, Denmark, June 29–July 3, 1997.Analyst, January 1998, Vol. 123 (69–72) 69Ni, Co, Zn and Cu. A certified reference material (DOLT-2 Dogfish liver, National Research Council of Canada) was used as a control. The recovery of metals from four samples of the certified reference material were determined: Ni (91%, 99%, 109% and 135%); Co (75%, 78%, 78% and 85%); Cu (101%, 102%, 102% and 105%); and Zn (94.5%, 97.7%, 98.3% and 100.9%).Calculated mean values ± range were Ni (108.5 ± 17.5%), Co (79 ± 4%), Cu (102.5 ± 1.5%) and Zn (97.75 ± 3.75%). Ni and Co concentrations were determined using a Perkin Elmer (Beaconsfield, Bucks, UK) Atomic Absorption Spectrometer 1100B and a combined Cr, Co, Cu, Mn and Ni Intensitron lamp. The analysis of the tissue samples for Cu and Zn was carried out using an Instrumentation Laboratory (Lexington, MA, USA) aa/ae Spectrophotometer 157 with individual Cu and Zn lamps and wavelength settings of 213.9 nm for Zn and 324.7 nm for Cu (Table 1).All primary standards were produced by diluting BDH laboratory standards (BDH/ Merck, Poole, Dorset, UK). Analysis of metals in water and sediments Sediment samples were air dried and weighed. Moisture was driven off overnight at 110 °C and the sample reweighed. Samples were ashed in a muffle furnace at 450 °C, the residue was ground to a fine powder, mixed with PVA glue and pressed into a pellet under 10 tons pressure. The pellet was dried at 110 °C overnight and stored desiccated until analysis by X-ray fluorescence spectroscopy (XRF).Water samples were analysed by ICP-AES. For both XRF and ICP-AES, replicate samples were used as internal controls. Analysis of five sets of replicate samples (three determinations for each sample) produced errors, reported as relative standard deviation (RSD), of Ni (0–9%), Co (0–10%), Cu (0–10%) and Zn (1–7%). Values determined for replicate samples analysed in commercial laboratories were in close agreement with those reported in this paper.Statistical analysis The two tailed Student’s t test was used to test for statistical significance. Data for metal concentrations was first transformed (log10) to a normal distribution. Results and discussion Population data In April, 1996 three species of fish were caught in the nets; trout, salmon and perch. Only trout were caught in sufficient numbers for proper statistical analysis (Table 2).The mean and median length of fish from sites A and B were larger, and on average the fish from site B weighed more, than those from sites C, D and E. Ranking of fish by length or weight gave the same sequence: weight ranking; B >> C > D > E; and length ranking, A > B >> C > D > E. The mean liver weight as a percentage of total body weight was approximately 0.7% at all sites except for site D. Livers at site D comprised approximately 0.9% of total body weight.Values could not be calculated for site A because total body weight was not determined in this group. No statistically significant differences were found between male and female fish sampled from sites B, C, D or E. In fish from site E the liver was often orange, or brown in extreme cases, as compared to the blood red colour of liver from fish sampled at site A. In large sample groups (B, C, D and E) the ratio of male to female fish approached 1:1 as would be expected.The modal average age was 2+ at sites A and C and 3+ at sites B, D and E. The growth of trout sampled from site E, and to a lesser extent sites C and D, appeared to be adversely affected by their environment. Fish from site D had enlarged livers and fish from site E may also have suffered significant liver necrosis. It was thought that one or more of the metals polluting the Otra river valley around Evje might be responsible for these effects.Thus the liver and muscle from the fish sampled were analysed for Ni, Co, Zn and Cu. Measurement of metals in fish liver and muscle The metal concentrations measured in individual trout are summarised in Fig. 2 (liver) and Fig. 3 (muscle). The mean concentration of Ni measured in trout was highest at site E; 0.33 ± 0.07 mg g21 for liver and 0.10 ± 0.03 mg g21 for muscle. The lowest concentrations of Ni were found in fish from site A; 0.05 ± 0.03 mg g21 for liver and 0.02 ± 0.01 mg g21 for muscle. Ni concentrations in fish from sites B, C and D were intermediate between those at site A and site E.Concentrations Fig. 1 Sites from which samples were collected: A, Byglandsfjord; B, Hornnes; C, South of Evje smelter; D, North of the Evje smelter; and E, Base of the Oddebekken. Water samples were collected at site F in the Oddebekken. Table 1 Assay performance for analysis of metals in fish LOD/mg l21 RSD (%) s/mg g21 wet weight Ni 0.8 8.7 ±5.9 Co 0.8 10.9 ±5.9 Zn 0.011 7.5 ±5.9 Cu 0.03 5.9 ±5.7 Table 2 Summary of population data for trout Site n Weight/g* Length/cm* Liver weight/g* Liver (% total weight)* Age/years Female fish (%) A 7 ND† 24.7 ± 1.0 0.68 ± 0.12 ND 2+ 14.3 B 20 119.3 ± 6.0 24.0 ± 0.6 0.88 ± 0.09 0.73 ± 0.06 3+ 55.0 C 24 90.7 ± 5.7 21.2 ± 0.5 0.67 ± 0.07 0.74 ± 0.06 2+ 45.8 D 25 65.5 ± 8.2 19.5 ± 0.9 0.58 ± 0.10 0.86 ± 0.06 3+ 52.0 E 29 79.0 ± 6.7 20.6 ± 0.7 0.53 ± 0.05 0.68 ± 0.05 3+ 55.2 * Values are means ± confidence limit (0.95).† ND, not determined. 70 Analyst, January 1998, Vol. 123at sites B, C, D and E were significantly greater than those at site A (P @ 0.0001). The lowest concentrations of Co were found in fish from site A, with values for liver being 0.06 ± 0.03 mg g21 and for muscle 0.02 ± 0.01 mg g21. Concentrations in fish from sites B, C, D and E were similar; mean concentrations in the liver were approximately 0.1 mg g21 and in the muscle 0.04 mg g21.Co concentrations in fish from all five sites were not statistically significantly different from each other (P > 0.0001). Zn concentrations at site A were lower than at all the other sites, with values for liver being 42.8 ± 5.2 mg g21 and for muscle 7.4 ± 1.5 mg g21. Zn concentrations in fish from sites B, C, D and E were not statistically significantly different from each other. Mean concentrations in the liver were approximately 60 mg g21 and in the muscle 20 mg g21.Concentrations at sites B, C, D and E were significantly greater than those at site A (P @ 0.0001). Cu concentrations were lowest in fish taken from site A. Values for liver were 29.6 ± 12.3 mg g21 and muscle 0.52 ± 0.28 mg g21. Highest values were found in fish taken from site E; 117.2 ± 33.5 mg g21 in liver and 11.5 ± 12.2 mg g21 in muscle. Values observed in the livers of fish taken from sites B, C and D were approximately 60 mg g21; intermediate between those from sites A and E.Cu concentrations measured in muscle of fish taken from sites B, C and D were 7.9 ± 1.7 mg g21, 5.1 ± 0.9 mg g21 and 1.5 ± 0.4 mg g21 respectively. Concentrations in the livers of fish from sites B, C, and D, and concentrations in the muscle of fish from sites C, D, and E were significantly greater than those from site A (P @ 0.0001). Ni, Co, Zn and Cu values determined in samples from fish taken from site A were tightly grouped at the low end of the concentration range.Similar values were also measured in some fish sampled from other sites. However, at sites B, C, D and E higher values were also measured. These observations suggest that no fish were contaminated at site A and that not all fish from the other sites were contaminated. These result were consistent for both liver and muscle samples. Concentrations of Zn and Cu in fish from site A were comparable to values reported elsewhere. Mean Zn concentrations for wild and hatchery-reared salmon smolts were 16.8 and 11.5 mg g21 wet weight, respectively.Mean Cu concentrations were 1.9 and 1.7 m g21 wet weight.22 We are not aware of any comparable published data on Ni and Co concentrations in fish. Fish collected at site E were found to be more highly polluted than those taken from the other sites. The effects of this pollution were also evident. Fish taken at site E were small, had relatively smaller livers and weighed less than those from other sites even though they were of comparable or greater age.No statistically significant differences were observed for fish of different ages at any site. This suggests that the difference in modal average age between fish from sites A and C (2+) and sites B, D, E (3+) (Table 2) does not affect the metal concentrations reported for each site. Source of contamination Preliminary studies have been performed to determine possible sources of contamination. Ni, Co, Zn and Cu concentrations were measured in water samples taken from the Oddebekken over a period of 12 months from July, 1994 to June, 1995 (Table 3).Concentrations of each metal changed during the 12 months. The concentration of Ni ranged from 11 mg l21 in January, 1995 to 567 mg l21 in November, 1994. Co was not detected in some samples and the highest concentration detected was 10 mg l21 in September, November and December, 1994 and February, 1995. Zn concentrations ranged from 5 mg l21 in September, 1994 to 103 mg l21 in June, 1995.The concentration of Cu measured ranged from 10 mg l21 in February and June, 1995 to 25 mg l21 in September, 1994. Although the pH of the Fig. 2 Scatter plot of metal concentrations measured in the livers of individual trout. Sites A–E refer to locations identified in Fig. 1. Fig. 3 Scatter plot of metal concentrations measured in the muscle of individual trout. Sites A–E refer to locations identified in Fig. 1. Analyst, January 1998, Vol. 123 71Oddebekken is low (3.9–4.7) the flux of the river Otra is so high that the measured pH of the Otra itself does not fall below pH 5.The pH of the river Otra falls along its course as humic streams drain into the river. Typical values for pH are 7 at Byglandsfjord (site A), 5.82 south of Evje (site B) and 5.77 north of Evje (site E).23 The pH values of 5.7–5.8 are typical of waters draining from humic areas known to support trout. It is unlikely that changes in pH are directly responsible for the physiological effects on fish reported in this study.Otra river bed sediment samples, collected from three sites (B, D and E) in April, 1996, were analysed for Ni, Co, Zn and Cu (Table 4). The concentrations of Ni, Co and Zn were similar at all three sites with average concentrations of 50.3 mg g21 (Ni), 14.7 mg g21 (Co) and 62.3 mg g21 (Zn). The Cu concentration ranged from 20 mg g21 at site B to 80 mg g21 at site D. The two main sources of exposure for fish are metals in the food materials, associated non-edible particulate material and metals dissolved in the water.Although the sediment data shows that the area at the base of the Oddebekken did contain Ni, Co, Zn and Cu, the levels found were much less than those in soils surrounding the smelter.4 The concentration of three metals (Ni, Co and Zn) were similar at sites B, D and E. Since the concentrations of metals found in fish from site B were considerably less than in fish from site E, it is probable that the sediment is not the main source of contamination for the fish.Superficial sediments collected from Swedish lakes contain between 100 and 500 mg g21 of Zn and between 20 and 50 mg g21 of Cu.24 Sediment values from the Otra river bed are comparable with those from the Swedish survey. Concentrations of several metals have been reported for an extensive survey of Swedish watercourses.25,26 The 75 percentile values for different regions ranged from 0.59–0.75 mg l21 (Ni), 6.5–16 mg l21 (Zn) and 1.1–2.7 mg l21 (Cu).In water from the Oddebekken, concentrations of Ni, Zn and Cu were greater than values reported in Swedish waters by up to three orders of magnitude for Ni and one order of magnitude for Zn and Cu.25 The concentrations of Ni, Zn and Cu measured in water samples taken from the Oddebekken exceeded lowest known biological effect concentrations, reported for aquatic systems, by up to an order of magnitude.27 Concentrations of metals in the river Otra were not determined in this study and only a small number of sediment samples have been analysed.In view of the effects of the aquatic environment on the fish reported here it would be of interest to determine the distribution of metals in this aquatic ecosystem. Conclusions The fish from site A were found to be much less polluted than those from any of the other sites. Site E was found to be the most polluted and the fish from this area were most severely affected physiologically by their environment.Sites B, C and D were less polluted and the fish in these regions were affected less. However, concentrations of all four metals were still considerably higher than those from the control site and clearly present a serious hazard to the aquatic ecosystem. Taken in isolation the concentrations of the metals measured in trout do not appear to constitute a risk to the health of humans consuming the fish.However, these results need to be placed in the context of metal concentrations in other local produce particularly drinking water. An assessment of the total metal exposure and the possible implications of exposure to multiple toxic metals in the diet of this population should be considered. Thanks to S. Uleberg, O. Neset and Professor R. Raiswell without whom the work would not have been possible. References 1 McCormick, P. H., MSc Thesis, University of Leeds, UK, 1995. 2 Ring, D. A., MSc Thesis, University of Leeds, UK, 1995. 3 Watkins, P. M., MSc Thesis, University of Leeds, UK, 1995. 4 Taggart, M. A., MSc Thesis, University of Leeds, UK, 1996. 5 ICRCL 59/83, ICRCL 59/83, EPTSSP187-5RH3, CDEP/EPTS, Interdepartmental Committee on the Redevelopment of Contaminated Land, London, 1987. 6 Sreedevi, P., Suresh, A., Sivaramakrishna, B., Prabhavathi, B., and Radhakrishnaiah, K., Chemosphere, 1992, 24, 29. 7 Alkahem, H. F., J. Univ. Kuwait, Sci., 1994, 21, 243. 8 Chaudhry, H. S., Toxicol. Lett., 1984, 20, 115. 9 Hassan, E. S., Abdellatif, H., and Biebricher, R., J. Comp. Physiol. A,, 1992, 171, 413. 10 Alam, M. K., and Maughan, O. E., J. Environ. Sci. Health, Part A, 1995, 30, 1807. 11 Radhakrishnaiah, K., Suresh, A., and Sivaramakrishna, B., Acta Biol. Hung., 1993, 44, 375. 12 Sharma, A., and Sharma, M. S., J. Environ. Biol., 1995, 16, 157. 13 Hogstrand, C., and Wood, C. M., Mar. Environ. Res., 1995, 39, 131. 14 Gagne, F., and Blaise, C., Environ.Toxicol. Water Qual., 1996, 11, 319. 15 Hansen, H. J. M., Olsen, A. G., and Rosenkilde, P., Comp. Biochem. Physio. C: Pharmacol. Toxicol. Endocrinol., 1996, 113, 23. 16 Iger, Y., Lock, R. A. C., Jenner, H. A., and Bonga, S. E. W., Aquat. Toxicol., 1994, 29, 49. 17 Julliard, A. K., Saucier, D., and Astic, L., Tissue Cell, 1996, 28, 367. 18 Black, J. J., Harshbarger, J., Zeigel, R. F., and Bock, F. G., Proc. Am. Assoc. Cancer Res., 1981, 22, 134. 19 Bunton, T. E., and Frazier, J. M., J. Fish Biol., 1994, 45, 627. 20 Bunton, T. E., J. Wildl. Dis., 1995, 31, 99. 21 Baatrup, E., Comp. Biochem. Physiol. C: Pharmacol. Toxicol. Endocrinol., 1991, 100, 253. 22 Felton, S. P., Grace, R., and Landolt, M., Dis. Aquat. Org., 1994, 18, 233. 23 Arnesen, R. T. I. E. R., Kartlegging av Forurensing fra Flatt Nikkelgruve, Evje, Notat 0-91144, Norsk Institutt for Vannforsking, 1993. 24 Johansson, K., Water Air Soil Pollut., 1989, 47, 441. 25 Metals and the Environment, ed. Natter, M., Swedish Environmental Protection Agency, Solna, Sweden, 1993, pp. 77–91. 26 Johansson, K., Bringmark, E., Lindevall, L., and Wilander, A., Water Air Soil Pollut., 1995, 85, 779. 27 Lithner, G., Sci. Total Environ., 1989, 87–8, 365. Paper 7/04883A Received July 8, 1997 Accepted September 15, 1997 Table 3 Heavy metal concentrations in the river Oddebekken (in mg l21) Date pH Ni Co Zn Cu July 1994 3.90 260 ND* 18 12 August 1994 4.18 105 7 12 11 September 1994 4.35 269 10 5 25 October 1994 4.50 299 ND 31 21 November 1994 4.35 567 10 21 21 December 1994 4.36 165 10 41 21 January 1995 4.32 11 ND 41 21 February 1995 4.50 103 10 31 10 June 1995 4.67 103 ND 103 10 * ND, not detected. Table 4 Metal concentrations in the sediment of the Otra in mg g21 Site Moisture (%) LOI (%)* Ni Co Zn Cu B 0.3 3.02 48 13 69 20 D 0.4 8.87 46 15 51 80 E 2.0 14.68 57 16 67 52 * LOI, loss on ignition. 72 Analyst, January 1998, Vol. 123

ISSN:0003-2654    

DOI:10.1039/a704883a

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Abnormal arsenic accumulation by fish living in a naturally acidified lake† Akiko Takatsu* and Akira Uchiumi National Institute of Materials and Chemical Research, 1-1 Higashi, Tsukuba, Ibaraki 305, Japan, The fish Tribolodon hakonensis live in good health in the naturally acidified Lake Usoriko, located in Aomori Prefecture, Japan, which has been acidified (pH 3.4–3.8) by sulfuric and hydrochloric acid of volcanic origin. The contents of various metals in various tissues of T. hakonensis, from Lake Usoriko were examined extensively by ICP-AES and it was discovered that large amounts of arsenic were accumulated in the eye tissues.This might be partly related to the fact that the lake water contains relatively large amount of arsenic and little phosphorus. Keywords: Arsenic; freshwater fish; inductively coupled plasma atomic emission spectrometry; eyes It is well known that many species of marine plants and animals accumulate arsenic, but little information is available on the accumulation of arsenic by freshwater fish.Lake Usoriko is a caldera lake located at the center of the Shimokita Peninsula in Aomori prefecture in Japan (Fig. 1). This lake is acidified by the inflow of strongly acidic water of volcanic origin from the northern part of the lake.1 One of the distinctive characteristics of the fauna of Lake Usoriko is the distribution of an acidtolerant fish, Tribolodon hakonensis in the strongly acidic water.2 Lake Usoriko is also subject to environmental pollution with arsenic and the lake sediment contains 1.6% m/m of arsenic, mainly as sulfides.3 Hot springs and solfataric activities are considered to be the source of the arsenic in the lake.In this work, we determined arsenic and other elements in various organs of T. hakonensis from Lake Usoriko. The function of the arsenic in the fish is discussed. Experimental Reagents Working standard solutions of elements and anions were prepared by diluting 1000 mg l21 stock standard solutions (Wako, Osaka, Japan).A standard solution of AsV was prepared by dissolving Na2HAsO4·7H2O (Wako) in water. For acid digestion, ultra-pure nitric acid and perchloric acid (Cica-Merck Ultrapur; Kanto Chemical, Tokyo, Japan) were used. For hydride generation, the reductant solution of sodium tetrahydroborate was prepared by dissolving NaBH4 powder (95% purity) in 1% aqueous sodium hydroxide. Potassium iodide solution (40%) was used for pre-reduction of AsV.Certified oyster tissue reference material (NIST SRM 1566a) was obtained from NIST (Gaithersburg, MD, USA). Instrumentation Elemental analyses were performed by inductively coupled plasma atomic emission spectroscopy (ICP-AES) (IRIS/AP; Thermo Jarrell Ash, Franklin, MA, USA). For the determination of arsenic, ICP-AES with hydride generation apparatus (HYD- 10; Nippon Jarrell Ash, Kyoto, Japan) was employed. The prepared sample solution, dilute hydrochloric acid (1 + 1) and KI solution, all of which were introduced by peristaltic pumps, were mixed at the beginning of a pre-reduction coil heated at 80 °C, then the pre-reduced solution was mixed with NaBH4 solution.The arsine generated was purged with argon gas and introduced into the ICP. The emission line of arsenic at 189.0 nm was used for quantification. Anions in water samples were determined using ion chromatography (IC7000; Yokogawa, Tokyo, Japan) with an ICS-A23 column, using 2.5 mm Na2CO3–1.0 mm NaHCO3 as the eluent at a flow rate of 1.0 ml min21.Sample collection and preparation Fish samples T. hakonensis were collected from the Maruyamasawa stream which is one of the inflow streams of Lake Usoriko with neutral pH. T. hakonensis living in the lake swim up this stream in order to lay eggs in early summer.2 The gills, eyes, bone and other parts of the fish were dissected under anesthesia. Five fish were killed and analyzed. The dissected organs were weighed and decomposed in a quartz beaker using nitric acid–perchloric acid (5 + 1) in a closed chamber as reported elsewhere.4 The decomposed samples were diluted to 5 or 10 ml with purified water in a calibrated flask.The concentrations of major and minor elements were determined by ICP-AES. For comparison, T. hakonensis from the River Tenryu, Japan, which is neither acidic nor polluted with arsenic, was also collected and analyzed. † Presented at The Sixth Nordic Symposium on Trace Elements in Human Health and Disease, Roskilde, Denmark, June 29–July 3, 1997.Fig. 1 Location of Lake Usoriko and sampling stations. Analyst, January 1998, Vol. 123 (73–75) 73Water samples Lake water samples were collected in polyethylene bottles at several locations of the shore of the lake. The sampling points are shown in Fig. 1. The pH and temperature were measured at the sites using a portable pH meter (DKK Model HPH-130) with a glass electrode, and the pH of water in the polyethylene bottles was subsequently measured in the laboratory using a pH meter (DKK Model PHL-40) with a glass electrode.The elements in the water were determined using ICP-AES. Anions were determined by ion chromatography. Results and discussion To confirm the analytical performance for arsenic, oyster tissue certified reference material (NIST SRM 1566a) was analyzed. The result for arsenic was 9.3 ± 0.8 mg g21 (n = 3), compared with the certified value of 14.0 ± 1.2 mg g21.The analytical recovery was 66%. Detection limit was about 0.6 ng ml21 as a solution. Tables 1 and 2 give the contents of major and trace elements in the various fish tissues. Five fish were analyzed in each instance and the means and standard deviations are given. Arsenic was detected in some organs of the fish from Lake Usoriko, whereas we could not detect arsenic in the fish from the River Tenryu. This indicates that the origin of the arsenic in the former fish could be environmental such as the lake water.From Table 2 it is noted that the arsenic content in the eyes of the fish from Lake Usoriko is very high. Small amounts of arsenic were detected in the bone and gills but no arsenic was detected in the muscle. Previous papers have reported that the target tissues for arsenic are the blood, spleen and liver.5–8 Therefore this work might represent the first finding of large amounts of arsenic localized in the eye tissues of fish.Table 3 gives the characteristics of water samples taken around Lake Usoriko. The lake water is strongly acidic and the pH of the hot springs which are the source of the acids in the lake was about 2.0. From Table 3, it can be seen that the water of Lake Usoriko contains large amounts of arsenic whereas its phosphorus content is very low compared with ordinary natural waters.9 Two cases have been reported previously: Saccharomyces carlsbergensis contains high concentrations of arsenic in the inositide parts10,11 and algae that had been cultivated with the culture ground loaded with H3AsO4 labeled with 74As contained 74As in the lipid section, especially in the phosphatidylethanolamine fraction.12 The major membrane phospholipid constituents of chick cultured retinal pigment epithelial cells are reported to be phosphatidylcholine and phosphatidylethanolamine. 13–17 It is suggested that arsenic might be taken into the lipid fraction of retinas.Arsenic and phosphorus have similar chemical properties and it is therefore likely that arsenic may participate in the metabolism process in place of phosphorus and behave like phosphorus when phosphorus is lost drastically. 18 In future studies we intend to investigate the chemical and physical functions of arsenic compounds and intracellular constituents. Table 2 Contents of minor and trace elements in fish samples [mg g21 wet mass; mean ± s (n = 5)] Sampling location Sample As Cu Fe Mn Zn Lake Usoriko (July 1996) Gills 0.28 ± 0.08 0.91 ± 0.17 35.4 ± 2.9 1.9 ± 0.6 21.6 ± 3.6 Eye 6.1 ± 1.2 0.66 ± 0.12 28.5 ± 11.5 0.6 ± 0.5 310 ± 55 Muscle < 0.03 1.0 ± 0.6 7.7 ± 4.1 0.18 ± 0.06 6.0 ± 2.4 Bone 0.27 ± 0.11 0.63 ± 0.24 5.5 ± 1.8 6.0 ± 1.9 55 ± 14 River Tenryu (December 1996) Gills < 0.02 0.66 ± 0.12 29.6 ± 4.5 3.9 ± 0.5 22.8 ± 2.3 Eye < 0.05 0.57 ± 0.08 11.5 ± 2.1 0.64 ± 0.16 251 ± 32 Muscle < 0.03 0.70 ± 0.37 8.7 ± 6.2 0.51 ± 0.11 9.0 ± 1.8 Bone < 0.05 0.34 ± 0.09 2.9 ± 0.6 14.2 ± 5.4 42.1 ± 3.5 Table 3 Characteristics of water in Lake Usoriko (sampling: June 7, 1997) Concentration/mg l21 Sampling point Temperature/ °C pH Na K Ca Fe Mg Mn P As Cl2 SO4 22 Lake water— St. 1 13.9 3.6 18.1 2.0 9.2 0.65 1.8 0.13 n.d.* 0.01 39 49 St. 2 14.0 3.3 23.1 2.3 9.8 0.86 1.8 0.14 n.d.* 0.03 54 57 St. 3 14.4 3.0 21.9 2.4 9.9 0.98 1.9 0.16 n.d.* 0.02 46 62 St. 4 14.9 3.2 47.4 5.2 13.8 1.57 2.9 0.53 n.d.* 0.45 120 80 St. 5 14.5 3.0 24.6 2.7 10.9 2.83 2.0 0.18 n.d.* 0.02 48 130 Spring— St. 6 60 2.0 1803 251 169 10.4 3.7 1.3 n.d.* 1.4 2800 620 St. 7 60 2.0 1714 230 161 11.6 4.1 1.5 n.d.* 0.01 2700 650 St. 8 60 2.0 1494 178 158 23.7 3.2 0.67 n.d.* 0.01 2300 1300 Stream (Maruyama-sawa)— St. 9 9.9 7.1 5.5 0.33 4.4 0.01 1.2 n.d. n.d.* n.d. 12 2.8 * Detection limit = 0.02 mg l21 Table 1 Contents of major elements in fish samples [mg g21 wet mass; mean ± s (n = 5)] Sampling location Sample Ca Mg P Lake Usoriko Gills 11.6 ± 1.7 0.42 ± 0.11 7.65 ± 0.79 (July 1996) Eye 3.98 ± 2.03 0.21 ± 0.10 3.11 ± 1.19 Muscle 0.52 ± 0.23 0.34 ± 0.09 2.28 ± 0.13 Bone 60.9 ± 2.5 1.28 ± 0.11 32.6 ± 1.2 River Tenryu Gills 12.9 ± 2.6 0.43 ± 0.03 9.26 ± 1.56 (December 1996) Eye 2.21 ± 0.27 0.10 ± 0.01 1.98 ± 0.16 Muscle 0.85 ± 0.37 0.26 ± 0.07 3.09 ± 0.33 Bone 61.1 ± 15.5 0.90 ± 0.16 31.2 ± 7.6 74 Analyst, January 1998, Vol. 123We thank Dr. K. Satake for discussions on the ecosystem of Lake Usoriko. We also thank A.Oyagi for his help in the field and S. Eyama and K. Terajima for technical assistance. References 1 Aoki, M., Rep. Geol. Surv. Jpn., 1992, No. 279, 16. 2 Satake, K., Oyagi A., and Iwao, Y., Water Air Soil Pollut., 1995, 85, 511. 3 Soma, M., Tanaka, A., Seyama, H., and Satake, K., Geochim. Cosmochim. Acta, 1994, 58, 2743. 4 Uchiumi, A., to be published. 5 Hove, E., Elvehjem, C. A., and Hart, E. B., Am. J. Physiol., 1938, 124, 205. 6 Hunter, F. T., Kip, A. F., and Irvine, J.W., Jr., J. Pharmacol. Exp. Ther., 1942, 76, 207. 7 Lanz, H., Jr., Wallace, P.C., and Hamilton, J. G., Univ. Calif. Publ., Pharmacol., 1950, 2, 263. 8 Mealey, J., Jr., Brownell, G. L., and Sweet, W. H., AMA Arch. Neurl. Psychiatry, 1959, 81, 310. 9 Riley, J. P. and Chester, R. , Introduction to Marine Chemistry, Academic Press, New York, 1971. 10 Cerbon, J., J. Bacteriol., 1970, 102, 97. 11 Cerbon, J., J. Bacteriol., 1969, 97, 658. 12 Irgolic, K. J., Woolson, E. A. , Stockton, R. A., Newman, R. D., Bottino, N. R., Zingaro, R. A., Kearney, P. C., Pyles, R. A., Maeda, S., McShane, W. J., and Cox, E. R., Environ. Health Perspect., 1977, 19, 61. 13 Bok, D., and Young, R.W., in The Retinal Pigment Epithelium, ed. Zinn, K. M., and Marmor, M. F., Harvard University Press, Boston, 1979, p. 148. 14 Tsunematsu, Y., Funahashi, M., and Nakajima, A., Dev. Growth Differ., 1981, 23, 313. 15 Michell, R. H., Biochim. Biophys. Acta, 1975, 415, 81. 16 Nozawa, Y., Maku, 1985, 10, 53. 17 Nakashima, S., Tsunematsu, Y., and Nozawa, Y., Acta Soc. Ophthalmol. Jpn., 1989, 93, 142. 18 Cooney, R. V., Mumma, R. O., and Benson, A. A., Proc. Natl. Acad. Sci., USA, 1978, 75, 4262. Paper 7/04877G Received July 8, 1997 Accepted September 23, 1997 Analyst, January 1998, Vol. 123 75

ISSN:0003-2654    

DOI:10.1039/a704877g

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Occupational arsenic exposure and glycosylated haemoglobin† Gunde E. Jensen*a and Michael L. Hansenb a The Danish Labour Inspection, National Register of Chemical Substances and Products, Lersø Parkalle 105, DK-2100 Copenhagen Ø, Denmark b The Danish Organization for the Control of Circulatory Diseases, Esplanaden 34 B, DK-1263 Copenhagen K, Denmark In a group of 40 workers occupationally exposed to arsenic (As workers) biological markers for cardiovascular diseases were studied. The median arsenic concentration in urine samples from the exposed group was 22.3 nmol of As per mmol of creatinine, while the individual maximum level was 294.5 nmol of As per mmol of creatinine.That of the reference group was 12 nmol of As per mmol of creatinine and significantly below the level of the exposed group (p < 0.001). The arsenic concentration in urine samples from colleagues of the persons working with arsenic containing products was similar to the arsenic concentration in urine samples from the As workers. The concentration of glycosylated haemoglobin (Hgb A1C) was increased in whole blood from the As workers.The level of the As workers was 5.4% (median), similar to that of colleagues (5.5%), while that of the reference group was 4.4%. The differences were significant (p < 0.001). Multiple regression analysis showed a significant connection (p = 0.034) between the concentration of Hgb A1C in whole blood and the arsenic level in urine from the As workers.The systolic blood pressure was 125 mm Hg in the As workers and 117 mm Hg in the control group. The difference was significant (p = 0.023). It is concluded that arsenic exposure has an influence on carbohydrate metabolism, increases the systolic blood pressure and finally may result in increased risk of development of cardiovascular diseases. Keywords: Arsenic; occupational exposure; glycosylated haemoglobin; blood pressure; diabetes; cardiovascular diseases Arsenic exposure may lead to development of cardiovascular diseases (CVD).1 However, further data are needed to confirm this suggestion.2 A study of biomarkers for cardiovascular diseases in a group occupationally exposed to As may support the suggestion.3 Increased glucose level is found in whole blood from CVD patients. The increase is due to a decreased glucose tolerance.4,5 The glucose is taken up passively by the erythrocytes; an increased level in serum results in an increased concentration of glucose in the erythrocytes.Due to a reaction with haemoglobin an increased level of glycosylated haemoglobin results.6 As an indicator for increased glucose level in the blood Hgb A1C is measured.7,8 Increased blood pressure is found in CVD patients.9,10 Increased diastolic as well as systolic blood pressure can be used as an indicator for increased risk of CVD. The aim of this work was: (i) to describe a group of occupationally exposed As workers, i.e., taxidermists and persons impregnating or working with As impregnated wood; (ii) to evaluate the concentration of glycosylated haemoglobin (Hgb A1C) in whole blood from As exposed workers; (iii) to evaluate the blood pressure in the As exposed group; and (iv) to discuss the results in relation to the development of cardiovascular diseases.Experimental Chemicals Inorganic arsenic in the oxidation state AsV and dimethylarsinic acid (DMAA) were obtained from Merck (Darmstadt, Germany) and Sigma (Poole, Dorset, UK), respectively. Pure samples of arsenite, arsenate, monomethylarsonic acid (MMAA), DMAA, arsenocholine and arsenobetaine were obtained from the Community Bureau of Reference (BCR), Brussels, Belgium.11,12 When not stated otherwise, all chemicals were of the highest purity obtainable from Merck.Groups of persons studied Urine samples and venous blood samples were collected from three groups of persons. Group 1: 26 persons (5 females) aged 20–60 years old without any known arsenic exposure (reference group); Group 2: 6 colleagues of workers who directly handle As containing products; and Group 3: 40 persons (4 females) working with arsenic containing products (As workers).Group 3 was divided into following occupational categories: (a) taxidermists (n = 13), stuffing animals and birds using arsenic as a preservative; (b) wood workers (n = 8, 4 females), producing garden fences; (c) wood workers (n = 6), producing weekend cottages; (d) workers impregnating wood (n = 2), using an autoclave (length, 12 m) for impregnation; and (e) workers impregnating electric pylons (n = 4), using a hand impregnation tool for the impregnation.The occupational categories as well as the As exposure are described in detail elsewhere.3,13 Table 1 shows the matching of the exposed and the reference group with respect to age (mean = 37 years) and smoking habits (40% versus 46%). The mean age of colleagues was 29 years and the percentage of smokers was 33%.Questionnaire interview In connection with blood and urine sampling a standardized personal interview based on a structured questionnaire was † Presented at the Sixth Nordic Symposium on Trace Elements in Human Health and Disease, Roskilde University, Denmark, June 29–July 3, 1997. Table 1 Subjects in the study characterized by age and smoking habits Subjects, Age, Smokers, Non-smokers, Groups n mean ±s n n As workers 40 37.2 ± 13.2 16 24 Colleagues 6 29.2 ± 9.8 2 4 References 26 37.5 ± 10.4 12 14 Analyst, January 1998, Vol. 123 (77–80) 77carried out. Information obtained from the interview included name and age, type of work, duration of work with arsenic containing products, chemicals and products used, alcohol intake, cigarette smoking, physical activities, consumption of fish and fish products, consumption of vitamins and medicine as well as personal and family history of diseases including hypertension, diabetes, cardiovascular disease and cancer.3 Collection of urine samples Two urine samples from each participant were collected seven days apart between 12:00 pm and 6:00 pm.They were collected in the middle of the week, the initial sample on the day of blood sampling. The samples were collected in acid-washed, 250 ml polyethylene containers and were divided into smaller polyethylene containers and stored at 220 °C until arsenic and creatinine concentrations could be determined.The participants were asked not to eat fish or fish products for three days and to wash hands prior to giving the urine samples. Methods When not stated otherwise the analyses were performed at the National Institute of Occupational Health, Copenhagen. Uncertainties and limits of detections were standard deviations (s) at low concentrations or blank samples and 2 3 s, respectively. Statistical control of the methods were documented by use of control cards.Determination of arsenic in urine samples The total amount of inorganic arsenic and metabolites (MMAA and DMAA) was determined by the direct hydride method.12,13 A Perkin Elmer 1100 B Atomic Absorption Spectrophotometer with flow injection and autosampler AS-90, equipped with an electrodeless discharge lamp was used for detection of arsenic hydrides (Perkin Elmer, Norwalk, CT, USA). The resonance line at 197.3 nm was used.12 Standard additions with different concentrations of As (as DMAA) was used.12 The method had a recovery of arsenite, MMAA and DMAA in urine not significantly different from 100%, whereas a reduced sensitivity to arsenate was observed. The reduced sensitivity to arsenate has previously been considered to be acceptable since this inorganic species represents only approximately 1% of the total urinary excretion of the arsenic compounds.12 The results were expressed as nmol of As per mmol creatinine (1 nmol of As per mmol creatinine = 0.662 mg of As per g of creatinine).All arsenic values are the means of the two samples from each person.13 The uncertainty was 12.9 nmol l21 and the limit of detection (3 3 s) was 38.7 nmol l21. Determination of creatinine Creatinine in urine was determined by Jaffe’s reaction in a Beckmann spectrophotometer.14 The results were expressed as mmol l21. The uncertainty was 0.5 mmol l21 and the limit of detection was 1.0 mmol l21. Collection of blood samples A 10 ml venous blood sample was collected from each participant.The blood was bled into venoject tubes containing EDTA. The samples were stored at 280 °C until the measurement of glycosylated haemoglobin. Determination of glycosylated haemoglobin (Hgb A1C) Glycosylated haemoglobin in whole blood was measured by affinity chromatography according to Little et al.6 The results were reported in % of total haemoglobin. The glycosylated haemoglobin assay was performed at the Medical Laboratory, Copenhagen.The uncertainty was 0.4% and the limit of detection was 0.8%. Blood pressure measurement Blood pressure, diastolic as well as systolic, were measured before the samplings and after the person had rested for 10 min. Initially three registrations were made; if the level was stable the registrations were used; otherwise the person rested until stable blood pressures were obtained. The blood pressures were automatically registered by use of digital blood pressure equipment, UA-751 (Takeda Medical, Japan).3 The blood pressures used in the study were the means of three registrations.Table 2 shows the samples, analyses and measurements involved in the study. Statistical methods The median, mean, and standard deviation were calculated by use of the MINITAB personal computer program.15 The differences between median values were tested by the Kruskal– Wallis non-parametric test. Multiple regression analysis was used for estimation of connection between parameters.For all tests performed, the level of statistical significance was set at 5%. Individual data points are reported in a preliminary study.3 Results In the As workers the maximum urine concentration 80 nmol of As per mmol of creatinine (median) was found in the group impregnating electric pylons, individual maximum level was 294.5 nmol of As per mmol of creatinine. The median value of all As workers was 22.3 nmol of As per mmol of creatinine. That of the reference group was 12 nmol of As per mmol of creatinine, which was significantly different from the concentrations in the group of all As workers as well as the subgroups consisting of taxidermists, garden fence makers and electric pylon impregnators, respectively (p < 0.001) (Table 3).The arsenic concentration in urine samples from colleagues of the persons working with arsenic containing products was at the level of occupational exposure. The concentration of Hgb A1C in whole blood from the As workers was 23% higher than that in the references.The level of the As workers was 5.4% (median), similar to that of colleagues (5.5%), while that of the reference group was 4.4%. The differences were significant (p < 0.001) (Table 4). The Hgb A1C level was increased in smokers compared to non-smokers, the difference was, however, only significant for the AsIII exposed group (p = 0.039) (Table 5). Multiple regression analysis showed a significant connection (p = 0.034) between the concentration of Hgb A1C in whole blood and the As level in urine from the As workers (Table 6).Table 2 Samples, analyses and measurements in the study Samples/ Samples from Samples from Samples from Analyses/ As workers, colleagues, references, Measurements n n n Urine samples, arsenic 40 5 26 Blood samples, Hgb A1c 32 6 26 Blood pressure 34 5 25 78 Analyst, January 1998, Vol. 123Table 7 shows the systolic blood pressure in the As workers to be 125.0 mm Hg (median value).That of their colleagues was 120.0 mm Hg, while the level in the reference group was 117.0 mm Hg. The Kruskal–Wallis test showed the systolic blood pressure to be significantly increased in the As workers (p = 0.023). The diastolic blood pressure was also increased in the As workers, however, the difference was not significant. The levels were 77.9 versus 74.7 mm Hg. In the As exposed group as well as in the reference group increased blood pressure levels were found in smokers, and increased blood pressure levels were found in As exposed nonsmokers as compared to non-smoker references; however, evaluation on this limited data showed the differences not to be significant.Discussion The results of this study showed an increased concentration of Hgb A1C in whole blood from As workers as well as from colleagues. The increase in the occupationally exposed As persons was about 25%. An increased Hgb A1C level is a well known marker for diabetes where the increase in Hgb A1C in whole blood has been estimated as 70% compared to a control level.8 At the time of sampling none of the As exposed persons in this study suffered from diabetes mellitus.However, other studies relate As exposure to diabetes. Thus an increased frequency of diabetes mellitus has been found in As exposed persons in Taiwan as well as in Sweden.16,17 Thus several data indicate that exposure to arsenic influences the blood glucose level.The biochemical mechanism by which As functions is probably not due to an effect on the insulin segregation from the pancreas. Thus according to an experiment performed by Kim and Na arsenic has no effect on the serum insulin level during intoxication.18 However, arsenic inhibits the activity of enzymes in the citric acid cycle, i.e., the activity of succinic acid dehydrogenase, thus inhibiting the carbohydrate metabolism.19 On the other hand the increased Hgb A1C level may also indicate an initial response against As intoxication.Thus animal experiments have shown that gluconeogenesis and interstitial glucose uptake are decreased by severe arsenic intoxication. 20,21 Treatment with glucose counteracts this toxic effect. 22 The results of the present study indicate that smoking contributes to an increased Hgb A1C level in whole blood. However, this increase may be due to an increased As exposure caused by the smoking.3,13 Smoking increases the systolic blood pressure in the reference group as well as in As workers.However, smoking habits may increase the As exposure in the As workers, which may contribute to the increase in the blood pressure in As workers who smoke. This may contribute to the finding of a significant systolic blood pressure in the As exposed workers in this study. That As exposure may increase the blood pressure is further supported by Chen et al., who found increased prevalence of hypertension in long-term As exposed persons.23 Epidemiological studies show a high prevalence of CVD in populations living in arsenic rich areas.1 Serum levels of Table 3 Sum of inorganic arsenic, MMAA and DMAA (arsenic*) in urine samples from individuals working with arsenic containing products, from their colleagues and from reference persons Occupational category n† Mean Median s‡ Range§ Taxidermists 13 30.0 21.0¶ 20.9 12.0–84.5 Garden fence makers 8 33.5 34.9¶ 11.4 14.6–48.8 Weekend cottage constructors 6 18.2 17.8 3.7 14.0–24.5 Colleagues 2 7.5 and 25.0 Wood impregnators 2 11.5 and 17.0 Electric pylon impregnators 4 127.4 80.0¶ 112.5 55.0–294.5 Colleague 1 61.5 New house constructors 7 19.1 15.0 7.3 13.0–29.0 Colleagues 2 7.0 and 20.0 All As workers 40 35.9 22.3¶ 46.2 11.5–294.5 All colleagues 5 24.2 20.0¶ 22.3 7.0–61.5 References 26 14.5 12.0 8.9 6.0–44.0 * Arsenic values are reported in nmol of As per mmol of creatinine.† n is the number of subjects in each category.‡ s is the standard deviation. § Minimum to maximum values. ¶ Significantly different from the reference group as estimated by the Kruskal–Wallis test. Table 4 Glycosylated haemoglobin (Haemoglobin A1c)* in whole blood from individuals working with arsenic containing products, from their colleagues and from reference persons Occupational category n Mean Median s Range All As workers† 32 5.7 5.4‡ 1.1 4.0–10.1 All colleagues 6 5.8 5.5‡ 1.0 4.7–7.5 References 26 4.4 4.4 0.5 3.2–5.1 * Values are reported in % of total haemoglobin.† Data from garden fence makers not available. ‡ Significantly different from the reference group as estimated by the Kruskal–Wallis test. Table 5 Glycosylated haemoglobin in whole blood from smokers and nonsmokers Exposure* n Mean Median s Range AsIII†, 2r 11 5.5 5.1 1.7 4.0–10.1 +r§ 6 6.2 6.2 0.8 5.3–7.5 AsV‡, 2r 8 5.4 5.6 0.8 4.6–6.2 +r 7 5.9 6.0 0.7 5.0–6.9 Reference, 2r 14 4.3 4.2 0.6 3.2–5.1 +r 12 4.5 4.4 0.4 4.0–5.1 * 2r = non-smokers; +r = smokers.† Workers exposed to AsIII, i.e., the taxidermists and the electric pylon impregnators. ‡ Workers exposed to AsV. § Significant difference between smokers and non-smokers. Table 6 Statistical connection between glycosylated haemoglobin and arsenic concentration in urine, years of work with arsenic containing products, and the age of the persons Regression equation* p y1 = 4.6 + 0.0130x1 0.596 + 0.0169z1 0.459 + 0.0078q1 0.034 * y1 = glycosylated haemoglobin, percentage of total haemoglobin; x1 = age, years; z1 = years of work with arsenic containing products; q1 = arsenic concentration, nmol of As per mmol of creatinine.Table 7 Systolic blood pressure in persons working with arsenic containing products, colleagues and references Group n Mean Median s Range As exposed workers* 34 127.5 125.0 14.4 100.0–158.0 Colleagues 5 122.8 120.0 7.7 115.0–132.0 Reference 25 119.9 117.0 11.9 105.0–153.0 * Significantly different from reference value.Analyst, January 1998, Vol. 123 79cholesterol, triglycerides and selenium are known biomarkers for CVD.24–27 However, the concentrations of cholesterol, triglycerides and selenium in serum from As workers have all been found to be at the reference level.3,28 These results indicate that increased levels of Hgb A1C(glucose) and blood pressure could be preliminary biological markers for As exposure and subsequently for increased risk of CVD.Finally, it can be concluded that As exposure influences the carbohydrate metabolism resulting in increased blood glucose levels, identified by Hgb A1C. Diabetes as well as increased blood pressure are known factors for development of cardiovascular diseases. As exposure by initiating increased blood pressure and diabetes subsequently may contribute to development of cardiovascular diseases. The study further supports the suggestion that As exposure can cause CVD.The financial support of the Danish Organization for the Control of Circulatory Diseases is greatfully acknowledged. Furthermore, the authors express gratitude to Dr. I. D. Beck and Dr. A. J. L. M�urer for valuable discussions and to E. Holst, for assistance with the statistical analysis, and to A. Abildtrup for technical assistance. References 1 Pershagen, G., and Vahter, M., Arbete och h�alsa, 1991, 9, 1. 2 Kristensen, T. S., Scand. J. Work Environ. Health, 1989, 15, 245. 3 Jensen, G. E., Arseneksponering ved arbejde med imprægneret træ og ved udstopning af dyr og fugle, Arbejdstilsynet, Arbejdsmiljøinstituttet, København, 1995, pp. 1–72. 4 Fuller, J. H., Shiply, M. J., Rose, G., Jarrett, R. J., and Keen, H., Lancet, 1980, i, 1973. 5 Phillips, G. B., Castelli, W. P., Abbott, R. D., and McNamara, P. M., Am. J. Med., 1983, 74, 863. 6 Little, R. R., England, J. D., Wiedmeyer, H. M., and Goldstein, D. E., Clin. Chem. (Winston-Salem, N.C.), 1983, 29, 1080. 7 D’Alessandro, A., Simon, D., Coignet, M. C., Cenee, S., and Giorgino, R., Diabetes Metabol., 1990, 16, 213. 8 Bruns, D. E., Lobo, P. I., Savory, J., and Wills, M. R., Clin. Chem. (Winston-Salem, N.C.), 1984, 30, 569. 9 Hein, H. O., Christensen, T. S., Suadicani, P., and Gyntelberg, F., in Hjertekredsløbssygdom og arbejdsmiljø, Arbejdsmiljøfondet, Denmark, 1990, volume 1, p.1. 10 Welin, L., Larsson, B., Sv�ardsudd, K., Wilhelmsen, L., and Tibblin, G., Lancet, 1983, i, 1087. 11 Larsen, E. H., J. Anal. Atom. Spectrom., 1991, 6, 375. 12 M�urer, A. J. L., Abildtrup, A., Poulsen, O. M., and Christensen, J. M., Talanta, 1992, 39, 469. 13 Jensen, G. E., and Olsen, I. L. B., J. Environ. Sci. Health, 1995, A30(4), 921. 14 Bartels, H, and Bohmer, M., Clin. Chim. Acta, 1971, 32, 81. 15 Ryan, T. A., Joiner, B. L., and Ryan, B. F., MINITAB Student Handbook, Duxbury, North Scituate, MA, 1976. 16 Lai, M. S., Hsueh, Y. M., Chen, C. J., Shyu, M. P., Chen, S. Y., Kuo, T. L., Wu, M. M., and Tai, T. Y., Am. J. Epidemiol., 1994, 139, 484. 17 Rahman, M., and Axelson, O., Occup. Environ. Med., 1995, 52, 773. 18 Kim, E., and Na, K. J., Toxicol. Appl. Pharmacol., 1991, 110, 251. 19 Tsutsumi, S., Usui, Y., and Matsumoto, Y., Folia. Pharmacol. Jpn., 1974, 70, 515. 20 Muckter, H., Islambouli, S., Doklea, E., Hopfer, C., Szinicz, L., Fichtl, B., and Forth, W., Toxicol. Appl. Pharmacol., 1993, 121, 118. 21 Hunder, G., Nguyen, P. T., Schumann, K., and Fichtl, B., Res. Commun. Chem. Pathol. Pharmacol., 1993, 80, 83. 22 Reichl, F. X., Kreppel, H., Szinicz, L., Fichtl, B., and Forth, W., Vet. Hum. Toxicol., 1991, 33, 230. 23 Chen, C. J., Hsueh, Y. M., Lai, M. S., Shyu, M. P., Chen, S. Y., Wu, M. M., Kuo, T. L., and Tai, T. Y., Hypertension, 1995, 25, 53. 24 Th`eriault, G. P., Tremblay, C. G., and Armstrong, B. G., Am. J. Ind. Med., 1988, 13, 659. 25 Vines, G., New Scientist, 1986, 11, 26. 26 Breier, C., Drexel, H., Lisch, H. J., M�uhlberger, V., Herold, M., Knapp, E., and Braunsteiner, H., Lancet, 1985, i, 1242. 27 Salonen, J. T., Alfthan, G., Pikkarainen, J., Huttunen, J. K., and Puska, P., Lancet, 1982, ii, 175. 28 Jensen, G. E., and Clausen, J., 1997, in preparation. Paper 7/05699K Received August 5, 1997 Accepted November 3, 1997 80 Analyst, January 1998

ISSN:0003-2654    

DOI:10.1039/a705699k

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Inductively coupled plasma mass spectrometric determination of molybdenum in urine from a Danish population† Bent S. Iversen*a, Charlotte Menn�ea, Mark A. Whitea, Jesper Kristiansenbc, Jytte Molin Christensenbc and Enrico Sabbionia a Commission of the European Communities, Environment Institute, Joint Research Centre, 21020, Ispra (VA), Italy b Department of Chemistry and Biochemistry, National Institute of Occupational Health, Lersø Parkall�e 105, 2100, Copenhagen, Denmark c The Glostrup Population Studies, Medical Department C, Glostrup Hospital, University of Copenhagen, Copenhagen, Denmark Molybdenum creatinine levels in urine were measured in 128 Danish inhabitants by ICP-MS in order to establish reference intervals of molybdenum in urine for the Danish population as a part of the EURO-TERVIHT project (Trace Element Reference Values in Human Tissues).The Mo concentration was determined using the isotopes 95Mo and 98Mo. The values measured based on 98Mo were about 2% lower than those calculated using 95Mo, a negligible difference in the context of reference values.The limit of detection was 0.2 mg l21, the precision was 8.6% and the recovery of added NIST 1643c certified reference material was 94%. The distribution of the data, with and without correction for creatinine concentration, was log-normal. The mean concentration measured was 42.5 mg l21, (3.89 mg Mo mmol21 creatinine) using 95Mo and 41.5 1 mg l21 (3.81 mg Mo mmol21 creatinine) using 98Mo, with the 95% parametric reference intervals 10.0–124.0 mg l21 (0.89–11.50 mg Mo mmol21 creatinine) and 9.6–122.6 mg l21 (0.84–11.47 mg Mo mmol21 creatinine), respectively.The difference between men and women reached the level of significance only after the values were corrected for the creatinine concentration. There was no influence of age on the Mo concentration. Specific effects of different food and beverage intakes could not be demonstrated, with the exception of a positive correlation between butter consumption and Mo concentration.Keywords: Molybdenum determination; urine; reference values; inductively coupled plasma mass spectrometry; sex; age; diet Molybdenum is an essential element, but also potentially toxic. Its main role as an essential element is related to its function as a prosthetic group in enzymes,1 of which sulfite oxidase, xanthine oxidase and aldehyde oxidase are active in animal tissues. Molybdenum is mainly obtained from molybdenite (MoS2) and it is used as an alloying agent in steels, as electrodes and in petroleum refining and nuclear energy applications. A risk of toxic exposure of workers in these industrial fields exist and environmental contamination near the industrial plants may be possible.It still remains unknown which tissue best reflects the nutritional status and/or eventual toxic exposure. Only a few papers have reported molybdenum concentrations in serum and urine.In serum it seems reasonable to suggest a Mo concentration lower than 1 mg l21,2,3 and levels of 100,4 30.7–33.35 and 4.8–205.6 mg l21 in urine6 have been reported, in the last instance with only two out of 19 values over 50 mg l21. Four review papers and an experimental paper covering five different countries have been published within the EURO-TERVIHT project (Trace Element Reference Values in Human Tissues), which aims to establish and compare trace element reference values in human tissues from inhabitants of the European countries.For the Danish,7 Italian8 and Czech and Slovak populations9 no values were given for Mo in urine. For the Belgian population10 a concentration range of 11.1–88.0 mg l21 was given based on only nine subjects plus values between 220 and 400 mg l21 for three additional subjects, and it was clearly pointed out that the limited number of data provide only a rough estimate of the concentration and concluded that a systematic study of Mo is desirable. Hamilton et al.11 argued that a level of < 10 mg l21 for Mo in urine could be used as a provisional concentration level for reference values for the UK population.No systematic studies on urine Mo levels in general populations or in exposed workers, to the knowledge of the authors, have been carried out. This paper presents for the first time Mo reference values for a Danish population. A direct method for the determination of Mo in undigested human urine by electrothermal atomisation atomic absorption spectrometry (ETAAS) using Triton X-100 and chemical modification has been published,6 showing that problems with low sensitivity and matrix effects can be overcome.In serum, where the concentration is significantly lower, a more sensitive inductively coupled plasma mass spectrometric (ICP-MS) method alone12 or in combination with ETAAS3 has been applied for Mo determination. The very sensitive radiochemical neutron activation analysis has been used for urine13 and serum14 analysis.For this study, the method chosen was ICPMS. Experimental Study population The 128 urine samples analysed in this experiment came from 71 male and 57 female subjects living in the Glostrup area, in the neighbourhood of the Danish capital Copenhagen. The population of the area is considered to represent closely the general population in Denmark. The subjects covered the age groups 40, 50, 60 and 70 years.Long-range atmospheric transport is typical for Mo, but existing or former Mo mining would result in local increased background levels.15 The zone where the present subjects live is without a molybdenum mining history or industrial molybdenum emitters, and it is therefore assumed that it represents the situation for the entire Danish population with respect to the atmospheric concentration of the element. † Presented at the Sixth Nordic Symposium on Trace Elements in Human Health and Disease, Roskilde, Denmark, June 29–July 3, 1997. Analyst, January 1998, Vol. 123 (81–85) 81Urine sampling The urine samples were collected in 100 ml polystyrene flasks with polypropylene lids at Glostrup Hospital. Questionnaires The subjects were asked to complete a personal questionnaire including a broad range of items such as occupation, health information and eating, smoking and drinking habits. Sample preparation and analysis A molybdenum 1.000 g l21 certified standard solution for atomic spectrometry was purchased from Spectroscan Teknolab (Drøbak, Norway).In all samples, calibration standards and blanks, indium (Merck, Darmstadt, Germany) was used as an internal standard. Throughout the procedure, Suprapure double sub-boiling distilled HNO3 (Romil Chemicals, Loughborough, UK) was used. All dilutions were made using Milli-Q ultrapure water (Millipore, Molsheim, France). Pipette tips (Eppendorf, Hamburg, Germany) and sampler cups (Perkin-Elmer, Norwalk, CT, USA) were made of polypropylene.Samples stored at 220 °C in screw-capped polypropylene tubes were allowed to defrost at room temperature and were well mixed. Samples were diluted 1 + 9 with 2% HNO3. Blanks were Milli-Q ultrapure water also diluted 1 + 9 with 2% HNO3. The calibration function was obtained by the standard addition method. The ICP-MS system used was a Perkin-Elmer SCIEX (Thornhill, ON, Canada) Elan 6000, equipped with a cross-flow nebulizer.Details of the operating conditions used throughout this work are given in Table 1. The performance of the instrument was checked daily following the manufacturer’s optimization programme. Isotopes and interference For the two isotopes (95Mo and 98Mo) used for quantification, interference is possible. For 95Mo the main risk of interference comes from the polyatomic ions 79Br16O+ and 40Ar39K16O+. The 81Br17O+ ion interferes with 98Mo. The 98Mo signal was corrected for the interference of 98Ru using the equation intensity (mass 98) 20.1095 [101Ru]. Vanhoe et al.12 preferred 98Mo for the analysis of serum, because their experiment suggested less interference from BrO+, but no significant differences between the two isotopes were found by Schramel and Wendler3 in serum analyses.Precision and accuracy The limit of detection (corrected for the dilution factor) for method applied was 0.17 mg l21 using 95Mo and 0.15 mg l21 using 98Mo (both 0.2 mg l21 to one decimal place), an order of magnitude lower than the lowest value found in this study.The precision of the method was established with 10 replicate measurements on the same urine sample, RSD = 8.6% (X = 32.6 mg l21, s = 2.9 mg l21). The accuracy could not be defined directly, owing to a lack of suitable certified reference material for Mo in urine. It was therefore assessed indirectly, using NIST 1643c certified reference material for water in a recovery experiment; 52.2 mg l21 (0.5 ml to 1 ml of urine) of this material were added to 10 different urine samples in the dilution step in the sample preparation.The mean recovery was 49.1 mg l21 (94%). In an ongoing project, a nearly identical ICP-MS method was compared with neutron activation analysis (NAA), 16 samples were analysed with both methods and the results are given in Table 2. The Wilcoxon matched pairs test was applied for statistical analysis. The results obtained from NAA were significant lower (p = 0.001), but no significant difference was observed between the means, 46.1 mg l21 (s = 17.5 mg l21) and 40.7 mg l21 (s = 15.1 mg l21).Statistical analysis The influence of different factors, such as gender, age and eating and drinking habits, were evaluated using both nonparametric and parametric statistical methods. Basic statistics and reference intervals were performed with in-house developed software and the remainder with the STATISTICA 4.5 software package.Basic statistics and reference intervals are presented with concentrations calculated using both isotopes (95Mo and 98Mo), the results from further statistical evaluations Table 1 Instrumental operating conditions ICP system— Instrument Perkin-Elmer SCIEX Elan 6000 Rf power 1050 W Frequency 40.68 MHz free running Auxillary gas flow rate 0.8 l min21 Plasma gas flow rate 15 l min21 Lens voltage 9.5 V Detector Nebulizer Cross-flow Sampler cone Nickel Skimmer cone Nickel Sample introduction system— Sample uptake flow rate 1.2 ml min21 Nebulizer gas flow rate 0.90 l min21 Delay time 30 s Wash time 3 min Mass spectrometer settings— Pressure Interface region 1 3 1023 Torr Ion lens region 1 3 1025 Torr Quadrupole region 1.3 3 1026 Torr Measurement parameters— Scanning mode Peak hopping Resolution Normal Points per peak 1 Sweeps per reading 15 Reading per replicate 2 Number of replicates 2 Dwell time 50 ms Autolens On Table 2 Comparison of results from ICP-MS and NAA (n = 16) Concentration of Mo in urine/mg l21 Sample No. ICP-MS NAA 1 55.4 49.2 2 39.5 35.4 3 54.6 48.7 4 78.6 69.0 5 56.6 50.2 6 50.3 43.7 7 20.8 20.0 8 59.7 51.5 9 35.2 33.9 10 45.9 37.8 11 29.5 20.6 12 56.6 46.5 13 60.1 53.0 14 10.4 8.9 15 55.6 49.7 16 28.4 33.0 Mean ± SD 46.1 ± 17.5 40.7 ± 15.1 82 Analyst, January 1998, Vol. 123did not differ significantly for the two isotopes, and are therefore presented only for 95Mo based concentrations. Results and discussion In nutritional, toxicological or environmental exposure studies, the assessment of the nutritional status, exposure and/or toxicological risk for humans is carried out by quantification of the element in a selected biological fluid.The concentration of the element in urine depends not only on the exposure or nutritional status, but also on drinking habits, differences in urine excretion, etc. It is therefore more reliable to relate the concentration of the element to the concentration of creatinine, which is excreted in a very stable manner.This was done for all the samples measured. The raw concentration data are also presented for completeness (Table 3). The mean concentration measured was 42.5 mg l21 (3.89 mg Mo mmol21 creatinine) using 95Mo and 41.5 1 mg l21 (3.81 mg Mo mmol21 creatinine) using 98Mo. Because the sex subgroups of the population were sampled with different rates, the weighted mean assuming 50% male and 50% females was calculated; using 95Mo it was 42.3 mg l21 (3.97 mg Mo mmol21 creatinine), and using 98Mo it was 41.3 mg l21 (3.89 mg Mo mmol21 creatinine).The distribution of the data was checked by the Anderson– Darling test.16 The distributions for the measurements with 95Mo and 98Mo, before and after correction for the concentration of creatinine, were significantly different from a normal distribution, and only after a logarithmic transformation of the data did the distributions fit well with the normal distribution (Table 3).The conclusion about the distribution of the data was further verified by the Lillifors test for normality; Fig. 1 shows the test for the entire population (95Mo). The corresponding reference intervals calculated as recommended by the IFCC (International Federation of Clinical Chemistry)17 are presented in Table 4. The values calculated on the basis of the two different isotopes agree very well. All 128 data pairs were compared by regression analysis, using a method which takes errors in both variables into account.18 The fitted straight line is represented by the equation [98Mo] = 20.333 + 0.983 [95Mo], and is shown in Fig. 2. The 95% confidence interval for the intercept includes zero, indicating no difference between the results, but the 95% confidence interval for the slope was [0.974:0.992], indicating a significant relative difference in the concentration values calculated using the two isotopes. The difference (from 0.8 to 2.6%) is so small that it is negligible in the context of reference values.Individual variations The subjects of the study population came from the same area, so individual variations due to differences in the atmospheric background level could be neglected. Sex and age factors Before the Mo concentrations in the urine were related to the concentration of creatinine, neither the non-parametric Mann– Whitney U-test for the original data (p = 0.57) or the parametric t-test for the logarithmic transformed data (p = 0.53) showed significant difference between the two sexes.However, the difference (female = 4.66, male = 3.28 mg Mo mmol21 creatinine) became significant after the Mo concentration was corrected for creatinine, with p = 0.01 for the Mann–Whitney U-test and the t-test. It is well known that men excrete more creatinine than women, approximately 1.7 against 1.1 g d21.19 The mean value for males in this study was 14.4, and for females 10.2 mmol l21, a highly significant difference (p < 0.001); the ratio of 1.4 is nearly identical with the ratio between Table 3 Basic statistics for Mo concentration in urine for 128 Danish subjects, total population and sex subgroups (male = 71, female = 57).Concentrations calculated using 95Mo and 98M. Values in bold are in mg l21 and values in italics in mg Mo mmol21 creatinine Arithmetic Geometric Distribution test (p)* Isotope used Range Mean SD Mean SD Median Normal Log-normal 95Mo, all 3.8–182.7 42.5 27.0 35.3 1.9 37.6 < 0.01 0.057 95Mo, female 5.4–162.3 41.0 26.1 33.9 1.9 38.3 < 0.01 > 0.15 95Mo, male 3.8–182.7 43.5 27.6 36.4 1.9 36.9 < 0.01 > 0.15 95Mo 2creatinine 0.45–14.46 3.89 2.55 3.19 1.92 3.13 < 0.01 > 0.15 95Mo 2creatinine, female 0.45–14.46 4.66 3.06 3.77 1.99 4.05 < 0.01 > 0.15 95Mo 2creatinine, male 0.59–10.10 3.28 1.87 2.80 1.82 2.93 < 0.01 > 0.15 98Mo, all 3.6–178.6 41.5 26.5 34.2 1.9 35.8 < 0.01 0.061 98Mo, female 5.3–157.9 40.1 25.9 32.9 1.9 35.8 < 0.01 > 0.15 98Mo, male 3.6–178.6 42.6 27.1 35.4 1.9 35.8 < 0.01 > 0.15 98Mo 2 creatinine 0.44–14.43 3.81 2.55 3.10 1.95 3.04 < 0.01 > 0.15 98Mo 2 creatinine, female 0.44–14.43 4.57 3.06 3.66 2.03 3.73 < 0.01 > 0.15 98Mo 2 creatinine, male 0.52–10.00 3.21 1.85 2.72 1.84 2.85 < 0.01 > 0.15 * Anderson–Darling test for normality.Fig. 1 Lilliefors test for normality Analyst, January 1998, Vol. 123 83the corrected means for Mo, and verifies that the higher mg Mo mmol21 creatinine values found for women are due to the lower concentration of creatinine in their urine. Tsongas et al.20 found that for the population of the United States, there was a significant difference in the intake between men and women, with a higher intake for men, possibly due to the different amounts of food consumed. The mean age of the subjects was 54 years; the four age groups 40, 50, 60 and 70 years were covered in a fairly uniform way with 31, 42, 25 and 30 subjects, respectively.Using a univariate statistical approach (GLM), no significant effect of age on the concentration of Mo in urine was found before (p = 0.11) or after (p = 0.91) correction for the concentration of creatinine (p = 0.12 and 0.98, respectively, after logarithmic transformation of the data). The scatterplots in Fig. 3 show these findings. The Spearman Rank Order correlation coefficient with sex ranked 1–2 revealed no significant correlation (p = 0.60) between sex and age, supporting the univariate approach used for these factors.Beverages and food Even though the questionnaire from which information was drawn was not constructed specifically for the purpose of this work, it was found worthwhile to see if any interesting relationships between the eating/drinking habits and the concentration of Mo in urine could be detected. The body burden of Mo is dependent on the intake of the essential element.In a comprehensive Italian study,21 the concentrations of a wide range of elements, including Mo, in different beverages were determined. Molybdenum was found in all samples of five different beverages, wine, mineral water, beer, tea infusion and instant coffee. The mean concentration ± s of Mo in 32 wine samples was 1.3 ± 2.2 mg l21, nearly equal to the value found for tea infusion (1.7 ± 2.4 mg l21); the concentration was about five times higher in mineral water, 6.6 ± 16.9 mg l21, and beer, 6.5 ± 3.7 mg l21, with the concentration in instant coffee being 3.45 ± 1.38 mg l21.The authors estimated the contribution of beverages to the weekly total dietary intake of Mo to be 4.3%, indicating that food intake is the far most important factor for Mo intake in humans. The Mo coming from beverages was estimated assuming a weekly consumption of 5 l of mineral water, 2 l of wine, one can of beer, two cups of tea and ten cups of instant coffee.The estimate was based on a reported weekly total dietary intake of 900 mg of Mo. For the average person in United States the average daily intake of Mo via the diet was estimated to be 180 mg,20 and a British study22 estimated a daily intake of 128 ± 34 mg. If no accumulation of Mo in the body is taking place, assuming an equilibrium between intake and excretion, and unrealistically assuming urine to be the only route of excretion and 10 l excreted in a week, a concentration of 90 mg l21, about double the value found in our study, is calculated.The consumption of beverages for the Danish population is markedly different, with a mean weekly consumption of 4.5 cans of beer, range [0–60], 3.7 glasses of wine [0–25], 36.5 cups of coffee [0–128] and 7.9 cups of tea [0–70]. No significant correlation was found between the concentration of Mo and the intake of the four beverages as demonstrated in Table 5.The correlation is based on a rank order test, and is therefore identical with the results obtained for the logarithmic transformed data. The findings were further supported by multivariate analysis [GLM and principal component analysis (PCA)] (not shown). The results are in good agreement with the suggestions from the Italian study21 that beverages play only a minor role in the dietary intake of Mo. Information about the consumption of 11 different foods was selected from the questionnaires for the statistical analysis: butter, cheese, rye bread, potato, vegetables (boiled), fruit, rice, Table 4 Reference intervals of Mo in a Danish population.Values in bold are in mg l21 and values in italics in mg Mo mmol21 creatinine. Isotope Non-parametric 0.95 reference interval Parametric 0.95 reference interval (Log-transformed data) 95Mo, all 10.5–107.1 10.0–124.0 95Mo, female 10.7–125.8* 9.5–121.1* 95Mo, male 9.53–126.0* 10.4–127.0* 95Mo 2 creatinine 0.82–11.63 0.89–11.50 95Mo 2 creatinine, female 1.06–14.22a 0.98–14.55* 95Mo 2 creatinine, male 0.63–8.37* 0.87–8.99* 95Mo, all 9.9–105.0 9.6–122.6 95Mo, female 10.2–123.6* 9.0–120.5* 95Mo, male 9.1–123.8* 10.0–125.0* 95Mo 2 creatinine 0.80–11.62 0.84–11.47 95Mo 2 creatinine, female 0.91–14.10* 0.92–14.60* 95Mo 2 creatinine, male 0.61–8.25* 0.83–8.93* * Tentative reference intervals based on less than 120 observations.Fig. 2 Fitted regression line for the Mo concentration in 128 urine samples calculated using 95Mo and 98Mo.Fig. 3 Concentration of molybdenum in urine as a function of age, with and without correction for the concentration of creatinine. Logarithmic transformed values. 84 Analyst, January 1998, Vol. 123pasta, meat, egg and fish. The consumption for each factor was divided into eight categories, never, @1 time per month, 2 times per month, 1 time per week, 2–3 times per week, 1 time per day, 2–3 times per day, !4 times per day.The first step in the statistical analysis was to see if a significant correlation (Spearman) between food factors and Mo concentration exists. The only significant correlation found was between butter consumption and Mo before correction for creatinine (0.1929, p = 0.048). Because the food factors were highly intercorrelated, linear regression would at best be misleading. The food factors were therefore transformed into 11 new non-correlated components by PCA.To simplify the interpretation of these new components (factors) they were further rotated using the Varimax strategy. After the rotation the pattern of the data was very clear, with only one high loading for one variable in each factor. These factors were then used in multiple regression analysis, and only the factor dominated by butter (loading = 0.969) was significant for the regression model with Mo as the dependent variable, both before (Table 6), and after logarithmic transformation. The conclusion of the analysis should be treated with caution, first because the regression model is bad, describing only 13% (r2 = 0.13) of the total variability for the data, and second because information about food intake is an extremely complex subject where each factor may be confounded with other factors, unknown or not involved in the analysis.The significant effect of butter on the Mo concentration in urine could not be supported by the findings of Tsongas et al.,20 who estimated the concentration of Mo in fats, oils, butter and margarine to be as low as < 0.01 mg g21 wet mass.Relating Mo to diet is more interesting from a nutritional than from a toxicological point of view, and Ward23 pointed out ‘that it would be unlikely for humans to get a toxic dose of Mo from milk or meat’. Conclusions The ICP-MS method applied gave nearly identical results for the two isotopes 95Mo and 98Mo used for quantification of Mo concentration.No difference in the concentration was found as a function of age. A difference was found between the two genders, with lower values for men, but only after the concentration has been correction for the concentration of creatinine. No significant dependence of the Mo concentration in urine on beverage or food consumption was observed, with the exception of a positive correlation with butter. A fully statistically satisfactory description of the correlation between Mo concentration in urine and beverage/food intake is very difficult owing to the extremely complex nature of this subject. To complement the picture of Mo further, and assess the eventual toxicological risk or evaluate the nutritional status, a better knowledge of which human fluid best represents risk/ nutritional status is essential.Fluctuations in excretion with time after intake/exposure need to be examined for optimizing the time of sampling. References 1 Williams, R. J. P., in Molybdenum: an Outline of Its Chemistry and Uses, ed.Braithwaite, E. R., and Haber, J., Elsevier, Amsterdam, 1994, pp. 419–451. 2 Versieck, J., J. Micronutr. Anal., 1989, 6, 261. 3 Schramel, P., and Wendler, I., Fresenius’ J. Anal. Chem., 1995, 351, 567. 4 Caroli, S., Alimonti, A., Coni, E., Petrucci, F., Senofonte, O., and Violante, N., Crit. Rev. Anal. Chem., 1994, 24, 363. 5 Iyengar, I., and Woittez, J., Clin. Chem., 1988, 34, 474. 6 Calvo, C. P., Barrera, P. B., and Barrera, A.B., Anal. Chim. Acta, 1995, 310, 189. 7 Poulsen, O. M., Christensen, J. M., Sabbioni, E., and Van der Venne, M. T., Sci. Total Environ., 1994, 141, 197. 8 Minoia, C., Sabbioni, E., Apostoli, P., Pietra, R., Pozzoli, L., Gallorini, M., Nicolaou, G., Alessio, L., and Capodaglio, E., Sci. Total Environ., 1990, 95, 89. 9 Kucera, J., Bencko, V., Sabbioni, E., and Van der Venne, M. T., Sci. Total Environ., 1995, 166, 211. 10 Cornelis, R., Sabbioni, E., and Van der Venne, M. T., Sci.Total Environ., 1994, 158, 191. 11 Hamilton, E. I., Sabbioni, E., and Van der Venne, M. T., Sci. Total Environ., 1994, 158, 165. 12 Vanhoe, H., Vandecasteele, C., Versieck, J., and Dams, R., Anal. Chem., 1989, 61, 1851. 13 Cornelis, R., Versieck, J., Desmet, A., Mees, L., and Vanballenberghe, L., Bull. Soc. Chim. Belg., 1981, 90, 289. 14 Versieck, J., Vanballenberghe, L., De Kesel, A., Hoste, J., Wallaeys, B., Vandenhaute, J., Baeck, N., Steyaert, H., Byrne, A. R., and Sunderman, F.W., Anal. Chim. Acta, 1988, 204, 63. 15 Berg, T., Røyset, O., Steinnes, E., and Vadset, M., Environ. Pollut., 1995, 88, 67. 16 Solberg, H. E., Scand. J. Clin. Lab. Invest., 1986, 46, Suppl. 184, 125. 17 Solberg, H. E., Clin. Chim. Acta, 1987, 170, S13. 18 Mandel, J., J. Qual. Technol., 1984, 16, 1. 19 Sperlingova, I., Dabrowska, L., Kucera, J., and Tichy, M., Fresenius’ J. Anal. Chem., 1995, 352, 87. 20 Tsongas, T. A., Meglen, R. R., Walravens P. A., and Chappell, W. R., Am. J. Clin. Nutr., 1980, 33, 1103. 21 Minoia, C., Sabbioni, E., Ronchi, A., Gatti, A., Pietra, R., Nicolotti, A., Fortaner, S., Balducci, C., Fonte, A., and Roggi, C., Sci. Total Environ., 1994, 141, 181. 22 Hamilton, E. I., and Minski, M. J., Sci. Total Environ., 1973, 1, 375. 23 Ward, G. M., in Molybdenum: an Outline of Its Chemistry and Uses, ed. Braithwaite, E. R., and Haber, J., Elsevier, Amsterdam, 1994, pp. 452–476. Paper 7/06565E Accepted September 9, 1997 Table 5 Spearman Rank Order correlation between Mo concentration before (95Mo) and after (95MoCrea) correction for the concentration of creatinine Pairs of variables n Spearman R p-Level 95Mo–beer 117 0.02082 0.8237 95Mo–wine 114 20.03775 0.6901 95Mo–coffee 119 0.07703 0.4050 95Mo–tea 119 20.13173 0.1533 95MoCrea–beer 117 20.04813 0.6063 95MoCrea–wine 114 20.13124 0.1640 95MoCrea-coffee 119 0.05205 0.5740 95MoCrea–tea 119 20.09831 0.2875 Table 6 Multiple regression results with Mo (mg l21) as dependent variable, and 11 non-correlated factors found by PCA as regressors. For each factor the dominan variable is given in parentheses, (r2 = 0.13) Variable Regression coefficient p-Level Intercept 41.123 0.000 Factor 1 (pasta) 20.006 0.999 Factor 2 (egg) 1.571 0.635 Factor 3 (vegetables) 2.801 0.398 Factor 4 (rye bread) 22.075 0.531 Factor 5 (meat) 2.761 0.405 Factor 6 (cheese) 22.651 0.424 Factor 7 (fruit) 0.486 0.883 Factor 8 (butter) 6.924 0.040 Factor 9 (potato) 0.294 0.930 Factor 10 (fish) 22.728 0.411 Factor 11 (rice) 22.487 0.453 Analyst, January 1998, Vol. 123 85

ISSN:0003-2654    

DOI:10.1039/a706565e

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Strain dependence of steady-state retention and elimination of mercury in mice after prolonged exposure to mercury(II) chloride† Jesper B. Nielsen*a and Per Hultmanb a Department of Environmental Medicine, Odense University, DK-5000 Odense C, Denmark b Department of Health and Environment, Molecular and Immunological Pathology, Link�oping University, S-581 85 Link�oping, Sweden Previous studies have demonstrated that the toxicokinetics of a single oral dose of inorganic mercury in mice depends on the specific strain.The strain dependent kinetics were attributed primarily to differences in intestinal Hg2+ absorption. Elimination kinetics have been difficult to evaluate, however, as the majority of a single oral dose will pass through the gastrointestinal tract unabsorbed. Therefore, in contrast to previous studies, in this study exposure to inorganic mercury in drinking water for a prolonged time period was used in order to reach a steady state for whole-body retention of mercury.The exposure level (5 mg l21) was sufficiently low to exclude gastrointestinal toxicity. The steady-state retention of mercury was established in four inbred mouse strains (B10.S, DBA, A.SW and SJL). The DBA strain reached the highest whole-body steady state level of mercury (19 mg Hg) whereas B10.S mice, when considering the drinking water consumption, had the lowest steady-state retention of mercury (15 mg Hg). Analysis of the whole-body elimination of mercury after 12 weeks of drinking water exposure indicated that variations in the elimination kinetics could explain a large fraction of the observed strain differences in the steady-state retention of mercury.Thus, the approximate half-time for elimination of aged mercury depots was longest in DBA mice (83 d) and shortest in B10.S mice (44 d). Further, the observed organ depositions indicated that differences in the transport of mercury from the liver to kidney might also explain some of the differences in the elimination kinetics.Keywords: Mercury(II) chloride; mice; elimination; kinetics; strain dependence Previous studies have demonstrated that the toxicokinetics of a single oral dose of inorganic mercury in mice depends on the specific strain. Most of these studies have been observational, measuring whole-body retention and/or total urinary and faecal elimination after a single oral or parenteral dose of mercury(ii). The strain dependent kinetics have previously been attributed primarily to differences in intestinal Hg2+ absorption.The intestinal absorption of inorganic mercury has been described as consisting of different phases such as adsorption to the outside of the mucosal epithelial cell followed by absorption into the epithelial cells from where in a third step the mercury reaches the general circulation.1 Mercury will be considered to be absorbed only when it reaches the general circulation since part of the mercury in the epithelial cell will be lost due to shedding of the mucosal cells.A recent review2 concluded that this model is generally accepted for inorganic, non-essential metals and seems valid for those situations in which no intestinal toxicity is present, i.e., low doses. The intestinal absorption of inorganic mercury is not known to have any species or strain specific carrier in any mammalian species studied. Species differences in absorption rates following oral exposure to inorganic mercury(ii) compounds have not been observed.3–5 As no specific carrier or enzymatic processes seem to be involved at low doses, the absorption could be expected to depend primarily on the metal concentrations on both sides of the cellular membranes and follow FickAs law for passive diffusion.Therefore, differences in intestinal absorption between different strains would be marginal and differences in kinetics, which in relation to human risk assessment would illustrate inter-individual variation, could be expected to depend much more on elimination kinetics.However, elimination kinetics have been difficult to evaluate in previous single dose experiments as these studies reflect a situation in which no equilibrium exists between the mercury absorbed and the mercury eliminated, i.e., deposition (temporary) of absorbed mercury will tend to invalidate the determination of excretion and elimination at the steady state (intake = elimination of unabsorbed Hg + excretion + deposition).At the steady state, an equilibrium exists where the intake equals the sum of eliminated, unabsorbed mercury and excretion of absorbed mercury (intake = elimination of unabsorbed Hg + excretion). Elimination of unabsorbed mercury will depend on fractional absorption as this elimination equals the difference between intake and absorption (elimination of unabsorbed Hg = intake 2 absorption). Hence at equilibrium the absorption should equal the excretion of absorbed mercury and should not depend on the fraction of mercury eliminated unabsorbed (absorption = excretion).Further, as the excretion of absorbed and retained mercury approximates first-order kinetics, excretion of mercury is a fixed fraction (percentage) of the whole-body retention of mercury, which is inversely related to the elimination half-time of mercury in the body. Therefore, a longer half-time for mercury excretion will cause a decreased fractional excretion.This will lead to a higher steady-state deposition of mercury as the whole-body retention, due to the lower fractional excretion, must be higher in order to excrete sufficient mercury to equalize the absorption. In contrast to previous studies, the present study used exposure to inorganic mercury in drinking water for a prolonged time period in order to reach a steady-state for whole-body retention of mercury. After the apparent steady-state level of mercury whole-body retention was reached, mercury(ii) chloride was withdrawn from the drinking water and the elimination kinetics were followed for 10 weeks.The exposure level (5 mg l21) was sufficiently low to exclude gastrointestinal toxicity. The aim of this study was to demonstrate that strain dependent differences in whole-body retention and steady-state retention of mercury after prolonged mercury exposure are affected by elimination kinetics and by differences in mercury disposition of absorbed mercury.† Presented at The Sixth Nordic Symposium on Trace Elements in Human Health and Disease, Roskilde, Denmark, June 29–July 3, 1997. Analyst, January 1998, Vol. 123 (87–90) 87Experimental Female mice aged 10–12 weeks of four different inbred strains (A.SW, SJL, DBA and B10.S) were kept on beechwood bedding in a well controlled environment (50 + 5% relative humidity, 20 air changes h21, temperature 21 + 1 °C, light/dark periods 12/12 h with 0.5 h twilight) with free access to standard mouse pellets (Brogaarden, Chr.Pedersen, Ringsted, Denmark) and water. According to the manufacturer, the feed pellets contained < 0.01 mg kg21 of Hg. The group size was initially 20. At 12 weeks, half of the mice were killed by cervical dislocation to obtain organ disposition at the steady state, and elimination kinetics were followed in the remaining mice for 10 weeks before the experiment was terminated.The animals were marked and weighed before exposure to drinking water containing mercury(ii) chloride (HgCl2) was initiated. The initial median body masses at the start of exposure were 16.1, 16.4, 16,5, and 21.3 g for the mice belonging to the DBA, SJL, B10.S and A.SW strains, respectively. HgCl2 (5 mg l21) was administered through the drinking water for 12 weeks. At 12 weeks, mercury was removed from the drinking water and the elimination kinetics was followed for another 10 weeks before the remaining mice were killed.During the entire experimental period drinking water consumption was measured and used to calculate the intake of mercury. HgCl2 was labelled with the gamma-emitting isotope 203Hg (203HgCl2; Amersham, Amersham, UK) to allow the determination of the whole-body and organ deposition of mercury. Whole-body retention of mercury was measured at regular intervals in live animals throughout the experimental period of 22 weeks. Thimals were counted in a whole-body counter (NaI well crystal, diameter 50 mm, 125 mm deep).All countings were performed between 10 and 12 am to reduce the influence of diurnal rythms with respect to drinking water consumption and feeding. The amount of radioactivity added to the HgCl2 solution was adjusted so that 1 ml of drinking water initially had approximately 25 000 counts min21. The backgrounds in the whole-body counter and the organ counter were below 250 counts min21. The detection limit was defined as the mean background level + 3 s observed in 40 countings of blank solution.Five backgrounds and a standard of known intensity (1 ml of drinking water) were counted at the beginning and at the end of each counting session. Calculated from the specific activity, the counting efficiencies in the two counters were 45%, which was considered satisfactory. Validation of the experimental model has been published previously.6 From all mice the liver and kidneys were rapidly excised, weighed and counted in a Searle 1195R gamma counter.For each animal, the whole-body and organ counts were related to the 203Hg standard of known intensity, and mercury deposition was calculated in micrograms of mercury and related to body or organ mass. Results are stated as group medians. The elimination half-times were estimated on a group basis based on log-transformed whole-body retention curves from day 84 and onwards. For all statistical comparisons the non-parametric Mann– Whitney rank test was used.The level of rejection of the H0 hypothesis was set at 0.05. Results During the 84 d of exposure to mercury(ii) chloride through the drinking water, the intake of mercury as based on drinking water consumption was significantly lower in the SJL strain (Table 1). The average daily drinking water consumption varied between 1.8 ml (SJL mice) and 2.6 ml (B10.S mice). The DBA strain reached the highest whole-body steady-state level of mercury (19 mg Hg) whereas B10.S mice, when considering the drinking water consumption, had the lowest-steady state retention of mercury (15 mg Hg).This was also reflected in the lowest ratio between whole-body retention and intake of mercury in the B10.S strain (Table 1), a ratio that could be regarded as a proxy for the difference between intestinal absorption and excretion, i.e., the amount of the exposure that will be deposited in the body. Considering the body mass (bm) of the mice, the DBA strain had the highest whole-body retention of mercury (0.8 mg Hg g21 bm) after 84 d of exposure to drinking water.A steady-state retention of mercury was established within 4 weeks in all four mouse strains (Fig. 1). The variations in wholebody retention at the steady state were small considering that the daily intake through drinking water varied between 9 and 13 mg of mercury. The whole-body elimination was characterized by apparently three phases of elimination.An elimination halftime was calculated for each phase and each strain. The most significant finding was the considerably longer elimination half-time of mercury in phase 3 for DBA mice (Table 2). After 12 weeks of exposure to drinking water containing mercury(ii) chloride, the hepatic mercury deposition was almost identical in B10.S, A.SW and SJL mice, whereas the DBA strain had a significantly higher hepatic mercury deposition (Table 3). After 12 weeks, renal deposition of mercury was highest in the SJL strain, closely followed by A.SW and DBA, whereas renal deposition of mercury in B10.S mice was significantly lower (Table 3).After 10 weeks of elimination, mice belonging to the A.SW strain had renal concentrations of mercury more than twice as high as those in the other strains (Table 3). Table 1 Intake and whole-body retention of mercury and mass data from four groups of mice. The mice belonged to four different inbred strains and were offered drinking water containing mercury(ii) chloride (5 mg l21) ad libitum for 84 d.Calculation of intake was groupwise and based on drinking water consumption during 84 d of exposure (group size = 20). Results are given as medians with 25 and 75 percentiles WBR* WBR WBR/ Intake/ (day 84)/ Body mass g21 bm/ intake Strain mg Hg mg Hg (bm)/g mg Hg (%) B10.S 1110 14.8A† 21.7A 0.68A 1.33ABC (13.6–16.4) (21.3–22.5) DBA 1070 18.8AB 23.4A 0.80ABC 1.76A (16.9–21.1) (22.8–24.0) A.SW 970 16.8C 26.7A 0.63B 1.73B (14.6–18.1) (24.0–28.3) SJL 760 13.4B 20.3A 0.66C 1.76C (12.0–15.2) (19.9–21.2) * Whole-body retention.† Identical letters (A, B or C) indicate a significant difference between groups (Mann–Whitney, two-tailed, p < 0.05). Fig. 1 Whole-body retention of mercury (mg Hg) in four strains of inbred mice exposed ad libitum for 12 weeks through the drinking water to mercury(ii) chloride (5 mg l21) followed by a 10 week elimination period (n = 20 during exposure; n = 10 during the elimination period).Results are given as medians. 88 Analyst, January 1998, Vol. 123Discussion The elimination of mercury from a steady-state situation after prolonged exposure to drinking water was apparently divided into three phases. The initial phase would primarily reflect elimination of unabsorbed mercury from the gastrointestinal tract, the second phase would represent clearance from easily accessible deposits and the third phase would illustrate excretion from more aged deposits.The last elimination phase is likely to be the most relevant in relation to a chronic exposure situation. This is supported by the much higher elimination halftime for mercury during phase 3 in the DBA strain (Table 2). Despite having almost identical drinking water consumptions, and thus intakes of mercury, and also comparable body masses, B10.S and DBA mice demonstrated very different whole-body retentions at day 84.The differences in kinetics between the two strains are, however, not likely to be explained by differences in absorption but rather by differences in elimination kinetics, as illustrated by the considerably longer third elimination half-time in DBA mice. Human experience with elimination of mercury primarily originates from occupational studies in chloralkali plants. In a 3 week follow-up of six workers with long-term exposure to inorganic mercury (mercury vapour), a fast initial decrease in the urinary mercury concentration was observed with a halftime of approximately 3 d, followed by a slow excretion phase with an elimination rate constant of 0.01 d21.7 In another study of long-term elimination of mercury, a group of 17 workers were on average followed for more than a year and the median half-time for elimination of occupationally acquired mercury was found to be close to 75 d with a range from 27 to 96 d.8 It was not clear whether the variation in half-times depended on duration of exposure, which in this study group varied from 3 d to 35 years.Other authors, based on comparisons of elimination half-times in workers exposed for short or longer periods, have suggested the presence of different compartments for mercury deposition with different elimination half-times.9 More recent studies have described kinetic models with at least two inorganic mercury compartments.10 Workers exposed to high levels of inorganic mercury for a short period of time showed, after cessation of exposure, two phases of urinary mercury excretion,11 a fast phase of a few days was followed by another phase with half-times of 40–90 d. Likewise, in chloralkali workers exposed to mercury for extended time periods (median 10 years), a fast and a slow elimination half-life for mercury in blood were described, both with large inter-individual differences (range 0.1–7.9 and 14–77 d, respectively).9 Hence considerable inter-individual differences in the elimination kinetics of mercury after exposure of humans to inorganic mercury exist.The underlying mechanisms might be related to differences in hepatic or renal handling of mercury. Data on organ deposition from the present study in mice indicated that differences in hepatic and renal handling of mercury may explain some of the observations made on wholebody retention and elimination kinetics. Thus, a similar wholebody retention but 3–4 times higher renal deposition in A.SW mice compared with the other strains after 10 weeks without exposure combined with the lack of increased hepatic deposition indicates a decreased capacity for renal elimination.DBA mice, on the other hand, did not show an increased renal deposition 10 weeks after removal of mercury from the drinking water, despite a significantly higher whole-body retention at the end of exposure after 12 weeks. However, DBA mice had a considerably higher hepatic deposition immediately after exposure.This could indicate that DBA mice have a capacity for renal elimination as good as the other strains, but a lower capacity for transporting recently absorbed mercury from the liver to the kidney. These data illustrate that relevant information is gained by including an elimination period in the study design in order to obtain information on elimination kinetics from a steady-state situation in addition to disposition of mercury during the elimination phase.However, it is important to emphasize that these observations on strain differences in elimination kinetics need confirmation with respect to mechanistic explanations. Human epidemiology indicates large inter-individual differences in the elimination kinetics of mercury, which in the experimental situation may be mimicked by using different species and outbred strains and analysed by using inbred strains. It is concluded that the analysis of the whole-body elimination of mercury after 12 weeks of exposure through drinking water indicated that variations in elimination kinetics could explain a large fraction of the observed strain differences in the steady-state retention of mercury.Thus, the approximate halftimes for elimination of aged mercury depots was nearly twice as long in DBA mice (83 d) as in B10.S mice (44 d). Further, the observed organ depositions indicated that differences in transport of mercury from the liver to the kidney might also explain some of the differences in elimination kinetics.This study was supported by a grant from the Danish Medical Research Council to J.B.N. and from the Swedish Medical Research Council to P.H. (project 9453). References 1 Foulkes, E. C., and Bergman, D., Toxicol. Appl. Pharmacol., 1993, 120, 89. 2 Schoof, R. A., and Nielsen, J. B., Risk Anal., 1997, 17, 545. 3 Clarkson, T. W., Food Cosmet. Toxicol., 1971, 9, 229. 4 Friberg, L., and Nordberg, G., in Mercury, Mercurials, and Mercaptans, ed.Miller, M. W., and Clarkson, T. W., Charles C. Thomas, Springfield, IL, 1973, pp. 5–22. Table 2 Estimated elimination half-times of mercury in four strains of mice after prolonged drinking water exposure to mercury(ii) chloride (5 mg l21). The elimination half-times were calculated on a group basis based on logtransformed whole-body elimination curves t1 2/d Strain Phase 1 Phase 2 Phase 3 B10.S 2.0 12.0 44 DBA 1.5 10.0 83 A.SW 2.5 11.5 54 SJL 1.0 11.5 50 Table 3 Organ deposition of mercury (mg g21) after 84 d of exposure to HgCl2 (5 mg l21) and after another 10 weeks without exposure to HgCl2 (group size = 10).Results are given as medians with 25 and 75 percentiles Hepatic deposition Renal deposition 12 + 10 12 + 10 Strain 12 weeks weeks 12 weeks weeks B10.S 1.45A* 0.04A 14.9ABC 0.75AB (1.42–1.72) (0.02–0.06) (10.1–19.8) (0.52–0.92) DBA 2.70ABC 0.04B 22.8A 1.08A (2.36–2.94) (0.02–0.05) (19.1–27.6) (0.89–1.17) A.SW 1.72B 0.04C 24.4B 2.67AB (1.56–2.06) (0.03–0.04) (22.2–25.4) (2.02–3.05) SJL 1.65C 0.02D 27.7C 1.33B (1.46–2.31) (0.01–0.03) (23.1–34.1) (1.06–1.39) * Identical letters (A, B, C or D) indicate a significant difference between groups (Mann–Whitney, two-tailed, p < 0.05). Analyst, January 1998, Vol. 123 895 Nielsen, J. B., and Andersen, O., J. Toxicol. Environ. Health, 1990, 30, 167. 6 Nielsen, J. B., J. Toxicol. Environ. Health, 1992, 37, 85. 7 Piotrowski, J. K., Trojanowska, B., and Mogilnicka, E. M., Int. Arch. Occup. Environ. Health, 1975, 35, 245. 8 Ellingsen, D.G., Thomassen, Y., Langård, S., and Kjuus, H., Scand. J. Work Environ. Health, 1993, 19, 334. 9 S�allsten, G., Barregard, L., and Sch�utz, A., Br. J. Ind. Med., 1993, 50, 814. 10 Smith, J. C., Allen, P. V., Turner, M. D., Most, B., Fisher, H. L., and Hall, L. L., Toxicol. Appl. Pharmacol., 1994, 128, 251. 11 Barregard, L., Quelquejen, G., S�allsten, G., Haguenoer, J.-M., and Nisse, C., Int. Arch. Occup. Environ. Health, 1996, 68, 345. Paper 7/05215D Received July 21, 1997 Accepted September 3, 1977 90 Analyst, January 1998, Vo

ISSN:0003-2654    

DOI:10.1039/a705215d

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Comparison of enhanced elimination of bismuth in humans after treatment with meso-2,3-dimercaptosuccinic acid and D,L-2,3-dimercaptopropane-1-sulfonic acid Anja Slikkerveer*‡a, Leslie A. Noachb, Guido N. J. Tytgatb, Gijsbert B. Van der Voeta and Frederik A. De Wolffa a Toxicology Laboratory, Leiden University Medical Centre, P.O. Box 9600, 2300RC Leiden, The Netherlands b Department of Gastroenterology and Hepatology, Academic Medical Centre, Amsterdam, The Netherlands Two groups of 12 human volunteers, who had been treated with colloidal bismuth subcitrate, because of Helicobacter pylori-associated gastritis, participated in the study.The patients received a single dose of meso-2,3-dimercaptosuccinic acid (DMSA) or d,l-2,3-dimercaptopropane-1-sulfonic acid (DMPS) at a dose of 30 mg kg21 in a randomized single blind study. In contrast to DMPS, increasing concentrations of bismuth in blood were observed during the first 4 h after intake of DMSA. In urine, both chelators induced a 50-fold increase in urinary bismuth excretion compared with the control urines.The treatment was well tolerated. The results indicate that both DMSA and DMPS effectively increase the elimination of bismuth in human urine. Consequently, both chelators may be of benefit in the treatment of patients with bismuth intoxication. Keywords: Bismuth, meso-2,3-dimercaptosuccinic acid; d,l-dimercaptopropane-1-sulfonic acid; chelation, humans Bismuth compounds are used in a wide range of gastrointestinal complaints, with special emphasis on the treatment of peptic ulcer disease and Helicobacter pylori infection.Bismuth intoxication is a rare but well known phenomenon. The metal has been known to cause nephro- or neurotoxicity dependent on many factors such as dose and type of bismuth compound.1 Animal studies showed that treatment of mice with acute bismuth intoxication with dithiol group containing chelators was effective in preventing mortality.2 A study in bismuthloaded rats has shown that dithiol group containing chelators increased bismuth elimination in urine and reduced the bismuth body load significantly.3 Several case histories have been published4,5on the use of chelators in the treatment of bismuth intoxication, but a controlled experiment in humans is only available for d-penicillamine.6 This study was performed to confirm the potential of dithiol chelators in the treatment of human bismuth intoxication. An overdose with bismuth-containing drugs is rare and it is known that small amounts of bismuth are left in the body for about 3 months after a normal treatment course with colloidal bismuth subcitrate (CBS).Therefore, the study was performed in patients who just had finished such a course of CBS. Based on the in vivo results in rats and clinical feasibility, two dithiol chelators were selected: meso-2,3-dimercaptosuccinic acid (DMSA) and d,l-2,3-dimercaptopropane-1-sulfonic acid (DMPS).Study design A group of 24 volunteer patients (age 26–65 years; mean 43 years) who had been treated with CBS (480 mg d21) for 28 d for Helicobacter pylori-associated gastritis received a single dose of DMSA or DMPS (30 mg kg21) in a randomized single blind study (n = 12). Subjects received the medication in gelatine capsules of 300 mg each. The study was performed between 7 and 14 d after the last day of treatment with CBS, because the steep decline in bismuth elimination in the urine directly after the end of treatment might have concealed an increase in bismuth elimination.Before the administration of the chelator and 30, 60, 120 and 240 min thereafter, blood samples were collected. Urine was collected over 24 h before the chelator was given and 0–4, 4–8 and 8–24 h after administration. Bismuth in blood was determined by electrothermal atomic absorption spectrometry7 and bismuth in urine was determined with an FIAS-200 system and a Model 5100 AAS instrument with hydride formation.8 Volunteers were clinically observed for 4 h and adverse events occurring in the first 24 h were scored by questionnaire.Ethical approval for the study was granted by the hospital ethical committee and written informed consent from each volunteer patient was obtained. Data analysis The results are presented as means ± SEM. Bismuth concentrations in urine are expressed per mmole of creatinine (mg mmol21). The effects of the chelators on the bismuth concentration were analysed by repeated measures ANOVA (MANOVA) with paired Student’s t-tests for the detection of contrasts.Confidence limits of 95% were used. Results Before treatment with the chelators, no differences were present in bismuth concentrations in blood and urine between the groups. Blood levels of bismuth increased moderately after DMSA treatment. A small but significant decline (MANOVA, p = 0.002) in bismuth concentrations in blood was seen after intake of DMPS (Table 1).After 120 and 240 min blood levels of bismuth were significantly higher after treatment with DMSA (Student’s t-test, p = 0.018). In urine, both chelators induced a 50-fold increase in urinary bismuth excretion compared with the control urines (Fig. 1). The highest amounts of bismuth were excreted in the first 4 h † Presented at The Sixth Nordic Symposium on Trace Elements in Human Health and Disease, Roskilde, Denmark, June 29–July 3, 1997.‡ Present address: Yamanouchi Europe, Research and Development, Leiderdorp, The Netherlands. Analyst, January 1998, Vol. 123 (91–92) 91after ingestion of the chelators. A total amount of 2819 (± 268) mg of bismuth was excreted in 24 h after treatment with DMSA and 2700 (± 322) mg after treatment with DMPS. No significant differences in bismuth excretion were seen between the two chelators (MANOVA). In 7 of the 24 patients minor side effects were observed such as transient diarrhoea, headache and nausea.The treatment was generally well tolerated. Discussion The present study indicates that both DMSA and DMPS effectively increase the elimination of bismuth in human urine. Consequently, both chelators may be of benefit in the treatment of patients with bismuth intoxication. Although differences in intestinal absorption between the two chelators have been reported,9 no significant differences with regard to the efficacy of urinary excretion of bismuth have been found.In contrast to DMPS, increasing concentrations of bismuth in blood were observed during the first 4 h after intake of DMSA. This may be attributed to redistribution of especially the Bi– DMSA complex from tissue deposits to the vascular compartment. The increase in the blood concentrations of bismuth after treatment with DMSA is limited and causes no concern with respect to safety. Given the dose (30 mg kg21) used in this single dose study, the study medication was remarkably well tolerated. This is in line with earlier reports from clinical experience that side effects are usually mild.9–11 The effect of both chelators on essential metals, such as Cu and Zn, is relevant to further evaluation of the clinical safety and applicability of these chelators.DMPS has been known to increase Cu and Zn elimination in humans, whereas DMSA only has minor effects on the elimination of Cu in urine and Zn in plasma.9,12,13 Nevertheless, this study has confirmed the previous indications that dithiol chelators are effective in the treatment of human bismuth intoxication.References 1 Slikkerveer, A., and De Wolff, F. A., in: Toxicology of Metals, ed. Chang, L. W., Lewis, Boca Raton, FL, 1996, pp. 439–454. 2 Basinger, M. A., Jones, M. M., and McCroskey, S. A., J. Toxicol. Clin. Toxicol., 1983, 20, 159. 3 Slikkerveer, A., Jong, H. B., Helmich, R. B., and De Wolff, F. A., J. Lab. Clin. Med., 1992, 120, 529. 4 Molina, J. A., Calandre, L., and Bermejo, F., Acta Neurol. Scand., 1989, 79, 200. 5 Playford, R. J., Matthews, C. H., Campbell, M. J., Delves, H. T., Hla, K. K., Hodgson, H. G. F., and Calam, J., Gut, 1990, 31, 359. 6 Nwokolo, C. U., and Pounder, R. E., Br. J. Pharmacol., 1990, 30, 648. 7 Slikkerveer, A., Helmich, R. B., Edelbroek, P. M., Van der Voet, G. B., and De Wolff, F. A., Clin. Chim. Acta, 1991, 201, 17. 8 Slikkerveer, A., PhD Thesis, Leiden University, 1992. 9 Aposhian, H. V., Annu. Rev. Pharmacol. Toxicol., 1983, 23, 193. 10 Mann, K. V., and Travers, J. D., Clin. Pharm., 1991, 10, 914. 11 Graziano, J. H., Lolacono, N. J., and Meyer, P., J. Pediatr., 1988, 113, 751. 12 Hruby, K., and Donner, A., Med. Toxicol., 1987, 2, 317. 13 Fournier, L., Thomas, G., Garnier, R., Buisine, A., Houze, P., Pradier, F., and Dally, S., Med. Toxicol., 1099, 3, 499. Paper 7/04945E Received July 10, 1997 Accepted September 8, 1997 Table 1 Concentrations of bismuth in blood (mg l21) 0–4 h after oral intake of 30 mg kg21 of DMSA or DMPS by human volunteers after a course of CBS (mean ± SEM; n = 12). Time after intake/min Chelator 0 30 60 120 240 DMSA 15.8 ± 2.5 17.4 ± 3.3 21.2 ± 2.2 25.8 ± 3.0 33.4 ± 4.1 DMPS 13.9 ± 1.2 16.2 ± 2.6 16.2 ± 1.5 15.9 ± 2.4 15.5 ± 1.0 Fig. 1 Bismuth concentrations in urine after oral intake of DMSA or DMPS (30 mg kg21) by human volunteers (mean ± SEM; n = 12). 92 Analyst, January 1998, Vol. 123

ISSN:0003-2654    

DOI:10.1039/a704945e

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Decreased selenium concentration in maternal and cord blood in preterm compared with term delivery† Waldemar Dobrzynskia, Urszula Trafikowskab, Anna Trafikowskab, Adam Pileckib, Wies/law Szymanskia and Bronis/law A. Zachara*b a Department of Obstetrics and Gynaecology, University School of Medical Sciences, Bydgoszcz, 75 Ujejskiego Str., Poland b Department of Biochemistry, University School of Medical Sciences, 24 Karlowicza Str., 85-092 Bydgoszcz, Poland The Se concentration in maternal and cord whole blood and plasma was determined spectrofluorimetrically in: (1) 42 women at term and (2) 46 at preterm parturients, and in the placenta.The glutathione peroxidase (GSH-Px) activity was measured in red cells and plasma in maternal and cord blood of both groups. The Se concentrations and GSH-Px activities of the above-mentioned groups were compared with those of non-pregnant women. Whole blood and plasma Se concentration of parturients at term did not differ significantly from those of non-pregnant women (72.3 versus 80.3 ng ml21 whole blood and 48.7 versus 56.2 ng ml21 plasma). In preterm parturients, however, Se concentrations were significantly lower (61.1 ng ml21 whole blood and 39.2 ng ml21 plasma) when compared with term parturients. The Se levels in cord blood and plasma were similar to their mothers’ Se concentrations.No difference was observed in placenta Se levels (130 ng g21 wet weight in both groups). The same was true for glutathione (GSH): the concentration in maternal and cord blood of term and preterm parturients did not differ and varied from 2.43 to 2.50 mmol l21 red cells.Red cell GSH-Px activities were similar in maternal and cord blood of both term and preterm groups and ranged from 14.3 to 15.7 U g21 Hb. The plasma enzyme activity in the maternal blood of preterm parturients was significantly (p < 0.05) lower than that of mothers at term delivery. The GSH-Px activity in the plasma of cord blood was significantly (p < 0.001) lower in the preterm than in the term group. It is suggested that low Se levels in the blood of women at preterm delivery, as shown here, may be one of the causes of retinopathy and respiratory distress syndrome in preterm infants.Keywords: Cord blood; delivery; glutathione peroxidase; placenta; plasma; pregnancy; selenium There is substantial evidence that selenium (Se) is a micronutrient that is essential for foetal growth and development.During the last two decades there has been remarkable progress in establishing the human requirement for Se. Until 1980 the essential safe and adequate daily dietary Se intake proposed by the Food and Nutrition Board of the National Research Council was 50–200 mg.1 More recent studies have shown that in order to stay in a positive balance adults need about 1 mg of dietary Se per kilogram of body weight per day.2,3 Although this dose ensures the recommended intake for healthy adults, the Board remains silent about possible increased nutritional needs during periods of physiological stress such as pregnancy and puerperium. 4,5 Several workers have shown that in healthy pregnant women both whole blood and plasma Se concentrations decreased linearly with the progress of pregnancy, reached a nadir just before delivery6–11 and began to rise again post partum.12 It is therefore generally accepted that the requirement for Se of pregnant women increases as a result of Se transport to the foetus.6,13,14 Pregnant women tended to conserve Se by decreasing urinary Se excretion,10 which seems to be the primary regulatory mechanism of Se status.15 The concentration of Se in the placenta and amniotic membranes was much higher than in maternal blood,13,14 which suggests the presence of an active transport system for Se in the placenta.14 Although the optimum Se intake during pregnancy is not known it was calculated that at least 100 mg of dietary Se per day is needed to maintain an optimum balance.13 During lactation the requirement for Se is also high owing to the transport of this element to the infant via the milk.5,13,16 The biological role of Se was established in 1973 when it was shown that this element is an integral component of glutathione peroxidase (EC 1.11.1.9; GSH-Px).17,18 This enzyme plays a critical role in the control of oxygen metabolism, mainly in catalysing the breakdown of H2O2,19 lipid and other organic hydroperoxides.20 GSH-Px is one of the primary antioxidants that is present in tissues and that inactivates lipid peroxides within the cells.21,22 This enzyme uses glutathione (GSH) as its cofactor to detoxify organic peroxides into relatively harmless organic alcohols.23 The developing foetus accumulates Se in the tissues (mainly in the liver) and the accumulation is higher in late pregnancy. 14,24 Koller et al.25 have shown that Se readily crossed the placenta and entered the circulation of the progeny resulting in foetal blood concentrations equivalent to or greater than the maternal circulation.These workers have shown that the lower the amount of maternal blood Se, the greater the amount sequestered in the foetal blood. Thus, the foetus can partially compensate for low maternal blood Se and can store or compensate Se to meet the biological requirements of this element. The Se concentration in blood components of preterm newborns is usually lower than that of term infants.26–29 Owing to lower Se concentrations, preterm infants are considered to be at risk of Se deficiency. Low plasma Se concentrations ( < 40 ng ml21) are associated with erythrocyte fragility in preterm infants,30 and lower concentrations ( < 10 ng ml21) appear to be required for the appearance of the clinical symptoms of Se deficiency.31,32 The aim of this study was to measure the Se and GSH concentrations and the activities of GSH-Px in blood components of pregnant women at preterm and term deliveries, as well as in the cord blood components of their newborn infants.Se levels in the placenta of both groups were also studied. Experimental Subjects The study group consisted of 42 women aged 16–42 years (mean, 26.3) at term delivery at a gestational age of 38–43 † Presented at The Sixth Nordic Symposium on Trace Elements in Human Health and Disease, Roskilde, Denmark, June 29–July 3, 1997. Analyst, January 1998, Vol. 123 (93–97) 93weeks (mean, 39.6), 46 women aged 18–45 years (mean, 27.2) at preterm delivery at 25–37 weeks (mean, 32.8) and 34 healthy, non-pregnant women (control group) aged 19-41 years (mean, 28.3). Blood samples were taken from women and from umbilical cord in sterile plastic tubes containing lithium heparin (Vacuette, Geiner Labortechnik, Kremsm�unster, Austria) and were stored at 4 °C. Placental tissue pieces were washed in an excess of physiological saline solution and were stored at 4 °C.All analyses were performed within 1–2 d after collection. Subjects participating in the study gave written informed consent, and the study has been approved by our university’s Ethics Committee for Medical Research. Sample preparation and analyses The blood samples were divided into two portions, one of which was used for measurement of haematocrit value and for Hb, Se and GSH concentrations, and the second was centrifuged at 4 °C and plasma was harvested for further analysis.The buffy coat was removed, and the remaining red cells were washed twice with an excess of chilled physiological saline solution. The remaining cells were haemolyzed by two cycles of freezing and thawing. Haemoglobin concentration in whole blood and haemolysates was measured by the cyanmethaemoglobin method. Se concentration in whole blood, plasma and placenta was assayed by the fluorimetric method of Watkinson.33 Samples were digested overnight with HNO3–HClO4. 2,3-Diaminonaphthalene (DAN; Sigma, St. Louis, MO, USA) was used as the complexing reagent and cyclohexane as the extracting solvent for the Se–DAN complex formed. The fluorescence was measured with a Hitachi (Tokyo, Japan) F-4010 spectrofluorimeter. The values were expressed as ng ml21 whole blood and plasma or ng g21 wet placentaThe accuracy of the procedure was checked with whole blood (batch No. 205052) and serum (batch No. 311089) reference materials purchased from Seronorm (Nycomed, Pharma, Oslo, Norway) and by participating in interlaboratory comparison trials.34 The mean values were 81.7 ± 2.5 and 84.5 ± 2.4 mg l21, respectively, compared with certified values of 83 and 86 ± 1.09 mg l21.GSH-Px activity in red cell haemolysates and plasma was measured spectrophotometrically (Specord, Jena, Germany) at 25 °C by the method of Paglia and Valentine35 with tertbutylhydroperoxide (Fluka Chemica-BioChemica, Buchs, Switzerland) as substrate.One unit (U) of the enzyme activity was expressed as 1 mmol of NADPH oxidized per minute, and the results were expressed as U g21 Hb and U l21 plasma. The relative standard deviation for both materials was below 5%. GSH concentration in whole blood was assayed by the method of Beutler36 and the results were expressed as mmol l21 red cells. All data were subjected to statistical treatment using the Student’s t-test analysis of variance and calculation of correlation coefficients. Probability values below 0.05 were regarded as statistically significant.Results and discussion Mothers Se and GSH concentrations, and GSH-Px activities of the control group, both groups of women at delivery and cord blood data are presented in Table 1. Although the mean Se concentration in whole blood of mothers at term delivery was lower by 10% when compared with the non-pregnant women (72.3 and 80.3 ng ml21, respectively) the difference was statistically non-significant. The lower plasma Se level of term mothers was, however, statistically significant (p < 0.01).In preterm mothers the Se concentrations both in whole blood and in plasma were significantly lower than those of non-pregnant women (0.001 < p < 0.01) and women at term delivery (0.01 < p < 0.02). In a previous study6 we have shown that during non-complicated human pregnancy the Se concentration in blood components declined almost linearly, reaching the lowest values at delivery.In that study the levels in whole blood, red cells and plasma of women at delivery were significantly lower than those of non-pregnant females.6 The Se concentration in the blood components of women participating in this study is rather low. It appears that the Se concentration in the blood of the Polish population and thus in pregnant women is now the lowest in Europe.37 Higher serum Se levels in non-pregnant females as well as in mothers at term delivery and in cord blood were presented by Danish workers,27 who found a statistically significant difference (p < 0.001) between non-pregnant females (81.3 ng ml21) and mothers at term delivery (66.3 ng ml21 serum).Wasowicz et al.37 and Wilson et al.38 have shown that plasma Se levels of mothers at delivery were significantly lower than those of non-pregnant women. On the other hand, Mask and Lane39 studied the Se concentration in five groups, similar to our study, and found no significant differences in plasma Se levels between women who were not pregnant, and full term and preterm women at delivery.It is noteworthy, however, that the daily dietary Se intakes of the American women studied by Mask and Lane39 were 2–3 times higher (96–134 mg) than those of the women studied here (below 40 mg; unpublished data). Presumably, the lower blood/ serum Se concentration observed in several studies of pregnant women7–11 reflects the lower dietary Se intake and, in part, the uptake of Se by the foetus.27 In some states of the USA and in Japan, where the daily dietary Se intake is relatively high, the Se concentration in blood components during pregnancy and in mothers at delivery was the same as in non-pregnant females, 39,40 or even higher.41 The differences in blood Se concentrations between pregnant and non-pregnant women presented by various workers are not easy to account for.It is very likely that mothers with low blood Table 1 Se and GSH concentrations and GSH-Px activities in blood components of non-pregnant women and of pregnant women at term and at preterm deliveries and in respective cord blood components Se concentration/ng ml21 GSH-Px activity Subject Whole blood Plasma Red cells/U g Hb21 Plasma/U l21 GSH concentration/ mmol l21 Non-pregnant women 80.3 ± 20.0 56.2 ± 14.0 16.8 ± 6.3 222 ± 60.8 2.35 ± 0.35 Pregnant women at term 72.3 ± 15.8 48.7 ± 12.5 15.7 ± 4.5 201 ± 74.6 2.49 ± 0.37 Umbilical cord 73.8 ± 15.6 44.5 ± 12.4 14.3 ± 4.1 184 ± 57.2 2.43 ± 0.39 Preterm pregnant women 61.1 ± 16.4 39.2 ± 11.5 14.7 ± 4.3 181 ± 75.3 2.48 ± 0.40 Umbilical cord 61.9 ± 14.4 37.3 ± 11.3 14.3 ± 3.7 141 ± 47.2 2.50 ± 0.30 94 Analyst, January 1998, Vol. 123Se levels have a severely depleted Se depot in the body and, since the growing foetus sequesters this element, it decreases in maternal tissues.6 There are only a few publications where Se concentrations in the blood components of mothers at term and preterm delivery are compared. Wilson et al.38 and Mask and Lane39 have found no difference in plasma Se levels between the two groups, while Wasowicz et al.37 obtained non-significantly higher plasma Se concentrations in preterm mothers compared with mothers at term delivery.In the present study, Se concentrations in whole blood and plasma of preterm mothers were significantly lower than those of women at term delivery (p < 0.01 and p < 0.001, respectively).This result may reflect the accumulation of smaller amounts of Se during the shorter gestation period of these infants. Red cell GSH-Px activity of both groups of pregnant women was lower by 7–10% compared with the control group but the difference was statistically non-significant. Plasma GSH-Px activity of women at term delivery (201 ± 74.6 U l21) was lower by 10% compared with that of non-pregnant women (222 ± 60.8 U l21), but the difference was also statistically non-significant.In the plasma of pregnant women at preterm delivery (181 ± 75.3 U l21), however, the enzyme activity was significantly (p < 0.05) lower than that of the control group. The fact that there is no difference in red cell enzyme activity is not surprising. It is well known that red cells respond slowly and give information on longer term Se status. Plasma levels of elements and enzyme activities respond within a few days, so they can respond quickly to any change in body stores.42,43 No significant differences in red cell GSH-Px activity between women of both groups at delivery and non-pregnant females were observed by Mask and Lane.39 Similar to our results, Behne and Wolters8 did not find any difference in Se-dependent red cell GSH-Px activity between non-pregnant females and pregnant women in the third trimester of pregnancy.On the other hand, Rudolph and Wong44 found higher GSH-Px activities in red cells, and Butler et al.9 obtained higher activities in red cells and plasma at delivery compared with those of non-pregnant women.Our observation on the lower activity of plasma GSH-Px of mothers at delivery is in accord with other studies.8,44,45 No significant difference in blood GSH concentration of both groups of mothers at delivery and in cord blood was observed in our study. The Se levels in the placenta of both groups were the same (130 ± 30 ng g21 wet tissue). These levels are 2.7 and 3.3 times higher than those in the plasma of women at term and preterm delivery, respectively.Kasik and Rice14 have recently shown that during late pregnancy, selenoprotein P is synthesized in the placenta. Beginning 4 d before birth, the level of this protein increases and reaches its maximum near to term. These workers suggest that Se is incorporated in the protein for transport across the placenta and subsequent secretion in the foetal circulation. Infants The mean birth weight of term infants was 3370 ± 478 g (range, 2080–4300 g) and for preterm neonates it was 1947 ± 614 g (range, 550–2620 g).In the present study the mean Se concentration in whole blood and plasma of term and preterm umbilical cord did not differ significantly from maternal whole blood and plasma levels (Table 1). Plasma Se concentrations of the umbilical cord of both groups were, however, significantly lower compared with the levels of non-pregnant women (p < 0.01).Several workers have shown that the plasma/serum Se concentration in term newborn infants is approximately half the level seen in non-pregnant women.27,28,38,44–47 In the plasma of preterm cord blood the difference is even more pronounced. Wilson et al.38 found 77 ng ml21 Se in the plasma of non-pregnant women and only 30.8 ng ml21 in the plasma of preterm cord blood. Although the present study shows lower Se concentrations in cord plasma compared with non-pregnant females, the values are lower by only 21 and 34% for term and preterm samples, respectively.Whole blood and plasma Se levels of preterm umbilical cords (61.9 ± 14.4 and 37.3 ± 11.3 ng ml21, respectively) are significantly (p < 0.01) lower than those of term cord blood (73.8 ± 15.6 ng ml21) and plasma (44.5 ± 12.4 ng ml21). Lockitch et al.28 and Wilson et al.38 also found lower levels of Se in the plasma of preterm cord blood, while other investigators29,37 did not observe any difference between the term and preterm plasma Se levels of cord blood.The Se concentrations in the blood components of the umbilical cord depend on the levels in maternal blood. We have shown strong, highly significant correlations between Se levels in maternal and cord whole blood and plasma and between GSH-Px activities in red cells and in plasma in both term and preterm groups (Table 2). Several workers searched for the relationship between birth weight and gestational age and plasma Se concentrations of cord blood as well as between birth weight and gestational age and plasma GSH-Px activities.Dison et al.47 found positive and highly significant correlations between the above-mentioned parameters. All these parameters were even higher in infants’ blood collected on day 2 or 3 of life. Sluis et al.29 found weaker, but significant correlation between gestational age and plasma Se levels and also plasma GSH-Px activities. We did not find significant correlations between birth weight and whole blood and plasma Se levels nor between GSH-Px activities of cord blood components of preterm deliveries (Table 3).No significant correlation existed between the gestational age and whole blood Se concentration or between gestational age and red cell GSH-Px activity. The only significant correlation that we found in this comparison was Table 2 Correlation coefficients between Se concentration and GSH-Px activity of maternal and cord blood components Term deliveries Preterm deliveries Relationship between r p r p Maternal versus cord whole 0.610 < 0.0001 0.713 < 0.0001 blood Se level (n = 36) (n = 43) Maternal versus cord plasma 0.572 < 0.0001 0.720 < 0.0001 Se level (n = 37) (n = 45) Maternal versus cord red cell 0.853 < 0.0001 0.596 < 0.0001 GSH-Px activity (n = 34) (n = 41) Maternal versus cord plasma 0.734 < 0.0001 0.635 < 0.0001 GSH-Px activity (n = 34) (n = 40) Table 3 Correlations between birth weight and gestational age and preterm cord blood values Parameter r p Birth weight and: whole blood Se 0.148 NS* plasma Se 0.l66 NS red cell GSH-Px 0.138 NS plasma GSH-Px 0.089 NS Gestational age and: whole blood Se 0.189 NS plasma Se 0.254 0.052 red cell GSH-Px 0.178 NS plasma GSH-Px 0.302 < 0.05 * NS = Not significant.Analyst, January 1998, Vol. 123 95between gestational age of preterm newborns and plasma Se level (r = 0.254, p = 0.05) and between gestational age and plasma GSH-Px activity (r = 0.302, p < 0.05).Lockitch et al.28 divided infants in their first three days of life into three groups: < 1500 g, 1500–2499 g and > 2499 g, and have shown that the lower the birth weight, the lower the Se concentrations and GSH-Px activities in their plasma. We did not find any significant differences between birth weight and Se concentrations or between birth weight and GSH-Px activities in cord blood components of preterm newborns divided into three similar groups.In another trial, Lockitch et al.28 studied the Se levels in the plasma of term and preterm infants after birth. Interestingly, in the plasma of term newborn infants no significant change in Se levels was observed within 2 weeks after birth, while in preterm infants the level showed a slight but significant decrease at day 7 and 14 (70 ng ml21 in cord plasma, 59 ng ml21 at day 7, and 41 ng ml21 at day 14). This suggests that preterm infants had accumulated an inadequate amount of Se in their body.In this study, the GSH-Px activities of red cells of both term and preterm cord blood were at the same level (14.3 U g21 Hb) and did not differ between maternal (14.7–15.7 U g21 Hb) or between non-pregnant female (16.8 U g21 Hb) enzyme activities. The similarity of red cell GSH-Px activity in all five groups agrees fully with the results presented by Mask and Lane,39 who have shown that the activities of this enzyme in the control group, in term and preterm mothers and in both groups of cord blood ranged from 23 to 25 U g21 Hb.In contrast, the plasma GSH-Px activity in our study was lower in the cord blood of term infants (184 ± 57 U l21) than in the plasma of maternal (201 ± 74.6 U l21) and non-pregnant females (222 ± 61 U l21) but not at a statistically significant level. The plasma enzyme activity of preterm cord blood (141 ± 47 U l21) was, however, significantly lower than that of maternal (181 ± 75.3 U l21, p < 0.01) and term cord blood (p < 0.001).Wilson et al.38 showed significantly lower plasma GSH-Px activities in both term and preterm cord blood compared with the appropriate group of maternal samples, but they were at the same levels in the plasma of term and preterm cord blood. In contrast, Sluis et al.29 examined the plasma GSH-Px activity of newborn infants within 7 d of birth and showed that the activity in preterm infants was significantly lower than that of term infants.Our results on the differences in plasma GSH-Px activities between cord and maternal blood are, in part, similar to the data presented by Dison et al.,47 who showed that plasma GSH-Px activities in cord blood were 2.3 and 2.6 times lower (p < 0.001) compared with paired maternal and with non-pregnant female samples, respectively. In addition, they found that plasma GSHPx activities in cord blood were strongly influenced by birth weight (r = 0.48) and gestational age (r = 0.57, both p < 0.001).Low Se levels and low GSH-Px activity in the blood components of preterm infants may be important in the development or the enhancement of the severity of respiratory distress syndrome, retinopathy of prematurity, increased haemolysis or other diseases.26–30 Sluis et al.29 showed that in 12 infants with plasma Se levels below 15 ng ml21, eight were requiring oxygen at 28 d and there was X-ray evidence of bronchopulmonary dysplasia, five had evidence of acute retinopathy of prematurity and five showed haemorrhaging. Diminished defence against oxidative damage will probably impair the ability of preterm neonates to withstand the stress to which they are subjected.28 The dramatic decline in plasma Se in premature infants accompanied by the fall in GSH-Px activity leads to the question of whether their Se intake is adequate or whether they should be given dietary supplementation.29 More information is needed to give a reasonable answer to this question.We express our appreciation to H. Gwiazdowska and J. Rubach for their skilled technical assistance. This work is a part of a PhD thesis submitted by W.D. to the Medical Faculty, University School of Medical Sciences in Bydgoszcz. References 1 Levander, O. A., in Trace Elements in Human and Animal Nutrition, ed. Mertz, W., Academic Press, Orlando, FL, 1986, vol. 2, pp. 209–279. 2 Levander, O. A., J.Am. Diet. Assoc., 1991, 91, 1572. 3 Food and Drug Administration, Recommended Dietary Allowances, National Academy of Sciences, Washington, DC, 10th edn., 1989. 4 Levander, O. A., Moser, P. B., and Morris, V. C., Am. J. Clin. Nutr., 1987, 46, 694. 5 Trafikowska, U., Zachara, B. A., Wiacek, M., Sobkowiak, E., and Czerwionka-Szaflarska, M., Acta Paediatr., 1996, 85, 1143. 6 Zachara, B. A., Wardak, C., Didkowski, W., Maciag, A., and Marchaluk, E., Gynecol. Obstet. Invest., 1993, 35, 12. 7 Anttila, P., Salmela, S., Lehto, J., and Simell, O., in Vitamins and Minerals in Pregnancy and Lactation, ed. Berger, H., Nestle Nutr. Workshop Ser., Vevey/New York, Nestec/Raven Press, 1988, vol. 16, pp. 265–271. 8 Behne, D., and Wolters, W., J. Clin. Chem. Clin. Biochem., 1979, 17, 133. 9 Butler, J. A., Whanger, P. D., and Tripp, M. J., Am. Clin. Nutr., 1982, 36, 15. 10 Swanson, C., Reamer, D., Veillon, C., King, J., and Levander, O. A., Am. J. Clin. Nutr., 1983, 38, 169. 11 Lechner, W., Jenewein, I., Ritzberger, G., Solder, E., Waitz-Penz, H., Schirmer, M., and Abfelter, E., Wien. Med. Wochenschr. 1990, 140, 370. 12 Lockitch, G., Crit. Rev. Clin. Lab. Sci., 1989, 27, 483. 13 Korpela, H., Loneniva, R., Yrjanheikki, E., and Kauppila, A., Int. J. Vitam. Nutr. Res., 1989, 54, 257. 14 Kasik, J. W., and Rice, E. J., Placenta, 1995, 16, 67. 15 Campbell, D. M., Proc. Nutr. Soc., 1988, 47, 45. 16 Mangels, A. R., Moser-Veillon, P. B., Patterson, K. Y., Veillon, K., Am.J. Clin. Nutr., 1990, 52, 621. 17 Rotruck, J. T., Pope, A. L., Ganther, H. E., Swanson, A.B., Hafeman, D. G., and Hoekstra, W. G., Science, 1973, 179, 588. 18 Flohe, L., Gunzler, W. A., and Schock, H. H., FEBS Lett., 1973, 32, 132. 19 Lehr, D., J. Am. Coll. Nutr., 1994, 13, 496. 20 Zachara, B. A., J. Trace Elem. Electrolytes Health Dis., 1992, 6, 137. 21 Walsh, S. W., and Wang, Y., Am. J. Obstet. Gynecol., 1993, 169, 1456. 22 Krinsky, N. I., Proc. Soc. Exp. Biol.Med., 1992, 200, 248. 23 Walsh, S. W., Hypertension Pregn., 1994, 13, 1. 24 Neve, J., Rev. Fr. Gynecol. Obstet., 1990, 85, 29. 25 Koller, L. D., Whitbeck, G.A., and South, P. J., Am. J. Vet. Res., 1984, 45, 2507. 26 Amin, S., Chen, S. Y., Collip, P. J., Castro-Magana, M., Maddaiah, V. T., and Klein, S. W., Nutr. Metab., 1980, 24, 331. 27 Bro, S., Berendtsen, H., Norgaard, J., Host, A., and Jorgensen, P. J., J. Trace Elem. Electrolytes Health Dis., 1988, 2, 165. 28 Lockitch, G., Jacobson, B., Quigley, G., Dison, P., and Pendray, M., J.Pediatr., 1989, 114, 865. 29 Sluis, K. B., Darlow, B. A., George, P. M., Mogridge, N., Dolamore, B. A., and Winterbourn, C. C., Pediatr. Res., 1992, 32, 189. 30 Gross S., Semin. Hematol., 1976, 13, 187. 31 Smith, A. M., Chan, G. M., Moyer-Mileur, L. J., Johnson, C. E., and Gardnre, B. R., J. Pediatr. 1991, 119, 429. 32 Kien, C. L., and Ganther, H. E., Am. J. Clin. Nutr., 1983, 37, 319. 33 Watkinson, J. H., Anal. Chem., 1966, 38, 92. 34 Neve, J., Thomassen, Y., and van Damme, M., Pure Appl. Chem., 1992, 64, 765. 35 Paglia, D. E., and Valentine, W. N., J. Lab. Clin. Med., 1967, 70, 158. 36 Beutler, E., in A Manual of Biochemical Methods, ed. Beutler, E., Grune & Stratton, New York, 2nd edn., 1975, pp. 11–12. 37 Wasowicz, W., Wolkanin, P., Bednarski, M., Gromadzinska, J., Sklodowska, M., and Grzybowska, K., Biol. Trace Elem. Res., 1993, 38, 205. 96 Analyst, January 1998, Vol. 12338 Wilson, D. C., Tubman, R., Bell, N., Halliday, H. L., and McMaster, D., Early Hum. Devel., 1991, 26, 223. 39 Mask, G., and Lane, H. W., Nutr. Res., 1993, 13, 901. 40 Kundu, M., Parke, P., Petersen, L. P., Palmer, I. S., and Olson, O., Arch. Environ. Health, 1985, 40, 268. 41 Sekine, K., Kimura, M., and Itokawa, Y., Nippon Eiseigaku Zasshi, 1989, 44, 699. 42 Levander, O. A., Fed. Proc., 1985, 44, 2579. 43 Rea, H. M., Thomson, C. D., Campbell, D. R., and Robinson, M. F., Br. J. Nutr., 1979, 42, 201. 44 Rudolph, N., and Wong, S. L., Pediatr. Res., 1978, 12, 789. 45 Zachara, B. A., Wasowicz, W., Gromadzinska J., Sk/lodowska, M., and Krasomski G., Biol. Trace Elem. Res., 1986, 10, 175. 46 Lombeck, I., Kasperek, K., Harbisch, H. D., Feinendegen, L. E., and Bremer, H. J., Eur. J. Pediatr., 1977, 125, 81. 47 Dison, P. J., Lockitch, G., Halstead, A. C., Pendray, M. R., Macnab, A., and Wittmann, B. K., Am. J. Perinatol., 1993, 10, 30. Paper 7/04884J Received July 8, 1997 Accepted September 10, 1997 Analyst, January 1998, Vol. 123 97

ISSN:0003-2654    

DOI:10.1039/a704884j

出版商:RSC    

年代:1998

数据来源: RSC