- Email: [email protected]
Evaluation of lipid peroxidation as a toxicity bioassay for plants exposed to copper
Environmental Pollution 109 (2000) 131±135
www.elsevier.com/locate/envpol
Evaluation of lipid peroxidation as a toxicity bioassay for plants exposed to copper A. Baryla a,*, C. Laborde a, J.-L. Montillet b, C. TriantaphylideÁs b, P. Chagvardie a a CEA/Cadarache, DSV, DEVM, Laboratoire de Phytotechnologie, 13108 Saint-Paul-lez-Durance, France CEA/Cadarache, DSV, DEVM, Laboratoire de Radiobiologie VeÂgeÂtale, 13108 Saint-Paul-lez-Durance, France
b
Received 28 January 1999; accepted 15 June 1999
``Capsule'': Chemical analysis of soils may not be representative of copper availability to plants. Abstract Agricultural soils may contain toxic levels of copper (Cu) due to sewage sludge spreading or industrial pollution but chemical analyses may not be representative of Cu bioavailability, de®ned as the soil Cu fraction that plants can actually absorb (i.e. Cu fraction which is not strongly adsorbed to soil components). Lipid peroxidation caused by Cu in plants was investigated as a relevant bioassay of toxicity. Seven-day-old rapeseed plantlets were grown on Cu-supplemented medium in controlled conditions. Lipidperoxidation was assessed by measuring: (1) the 2-thiobarbituric acid (TBA)-reactive substances; (2) the hydroperoxy acids by HPLC analysis; and (3) the alkane outputs by gas chromatography. We ®rst veri®ed the correlation between the results obtained by each method and then discussed their advantages and disadvantages within the context of a bioassay, showing that the volatile alkane output measurement is the most precise and easy to perform method for this purpose. # 2000 Elsevier Science Ltd. All rights reserved. Keywords: Lipid peroxidation; Toxicity; Bioassay; Plants; Copper
1. Introduction Industrial pollution or sewage sludge spreading may have accumulated high quantities of toxic metals (Cd, Pb, Cu, Zn, etc) in agricultural soils (Tarradellas and Bitton, 1997). It is, therefore, of great concern for public health and plant growth to assess the toxicity of a soil. Numerous bioassays have been developed, allowing the detection of toxic substance eects on model organisms (which are sensitive species in general). For example, physiological (i.e. root elongation) or biochemical (enzymic inhibition or activation) parameters (Tsay et al., 1995), proven to be disturbed by the suspected pollution, have been analyzed. These bioassays are useful to de®ne the eects of dierent substances or to assess the toxicity of a complex matrix like a soil. Nevertheless, toxicity bioassays are never easy to perform. In the soil, root elongation is not easy to measure. Tests based on enzymic activities are often dicult and time consuming to perform and, moreover, they sometimes * Corresponding author.
need numerous reagents. It is important to establish rapid and simple new tests. Copper (Cu) is a microelement necessary for plant growth. Non-polluted soils contain 10±30 ppm Cu (dry wt.) but soils located near mining areas or metalprocessing industries may be contaminated by very large amounts of Cu: we can ®nd soils containing more than 3500 ppm. Its large use as a fungicide and in sewage sludge spreading may also have led to toxic levels of this metal in orchards or in wine-producing areas (KabataPendias and Pendias, 1992). Therefore, it becomes interesting to rapidly and easily assess the Cu toxicity of such soils. Due to the ability of Cu to act as an ecient generator of toxic oxygen species, it can initiate the lipid peroxidation process (Halliwell and Gutteridge, 1984; Aust et al., 1985; Girotti, 1985; Weckx and Clijster, 1996). The aim of our study was to determine if assessing the lipid peroxidation level of 7-day-old plantlets could be used as a Cu toxicity bioassay, as a ®rst step before testing presumed Cu-polluted soils. With this aim, we investigated three dierent methods for lipid peroxidation. Seven-day-old rapeseed plantlets were exposed to Cu and the plantlet malondialdehyde (MDA) level was
0269-7491/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved. PII: S0269-7491(99)00232-8
132
A. Baryla et al. / Environmental Pollution 109 (2000) 131±135
measured by the thiobarbituric acid (TBA) method which is relatively simple to carry out and has been used by many authors (Janero, 1990). We also determined the level of free and esteri®ed hydroperoxy fatty acids by high performance liquid chromatography (HPLC) analysis of the corresponding hydroxyacids. Finally, we assessed the plantlets' alkane outputs by gas chromatography. Ethane and n-pentane are actually well-known endproducts of o-3 and o-6-unsaturated hydroperoxy fatty acids and their production occurs via b-scission of the corresponding alkoxy radicals (DegouseÂe et al., 1995). The usefulness and validity of these methods as bioassays have been assessed and discussed.
Plantlets were washed with deionized water, dried at 90 C for 72 h and digested in a 1/3/3 (v/v/v) mixture of nitric and sulfuric acids and hydrogen peroxide according to Sterritt and Lester (1980). The samples were heated slowly to 80 C and the resulting solution analyzed by a ¯ame atomic absorption photometer (Perkin Elmer 3110). Three replicates per Cu treatment were made. 2.4. Lipid peroxidation evaluation
For this experiment, seeds were surface sterilized and sown in sterile conditions on the culture medium previously described along the diameter of Petri dishes (six seeds per dish). The Petri dishes were arranged with a 45 inclination in order to keep the roots straight. The root length was measured on day 3. Six replications, randomly disposed in the growth chamber, were performed for each Cu treatment. Eight Cu concentrations (from 0 to 500 mM) were tested.
The MDA content was determined by the TBA reaction with minor modi®cation of the method of Dhindsa et al. (1981). A 0.30-g crushed sample was homogenized in 1.25 ml trichloroacetic acid (TCA) (0.1%)±natrium dodecyl sulfate salt (SDS) (1%). The homogenate was centrifuged at 12,000 g for 5 min. To 300 ml aliquot of the supernatant was added 1 ml 20% TCA containing 0.5% TBA. The mixture was heated at 95 C for 30 min and then quickly cooled in an ice-bath. The absorbance at 532 nm was read spectrophotometrically (Uvikon 932, Kontron Instruments S. p. a., Italy) and the concentration of MDA was calculated using the extinction coecient of 155 mMÿ1 cmÿ1. TBA and 1,1,3,3-tetraethoxypropane (TEP) as a reference were supplied from Fluka (France), SDS from Merck and TCA from Prolabo, all products of analytic grade. The spectrophotometer was a model Uvikon 932 from Kontron Instruments S. p. a. (Italy). Free and esteri®ed hydroperoxy acids hydroperoxyoctadecadienoic acid (HPODE), hydroperoxyoctadecatrienoic acid (HPOTE) were reduced in their corresponding hydroxy derivatives hydroxyoctadecadienoic acid (HODE), hydroxyoctadecatrienoic acid (HOTE), and quanti®ed after alkaline hydrolysis by HPLC analysis as previously described (DegouseÂe et al., 1994). HPLC-grade solvents were supplied from Rathburn (Scotland, UK). For this method, 2.5 g fresh weight of plantlets was necessary for one measure. Thermal production of the volatile alkanes was carried out as previously described (DegouseÂe et al., 1995) from a sample of 1 g of plantlets. Ethane and n-pentane analyses were carried out separately by injection (120 C) of 1 ml of gas on a gas chromatogram (GC) (Delsi Di 200, France) run at 80 and 200 C, respectively, with He (30 ml minÿ1) as the carrier gas. The chromatography was performed on an aluminia column (1.5 m3.2 mm i.d.) and detection of the hydrocarbon gases was done by ¯ame ionization at 250 C. Calibrations were carried out by injection of 1 ml of 10 ppm ethane and 10.7 ppm n-pentane in N2 (Scotly Analyzed Gases, USA).
2.3. Determination of Cu in plantlets
2.5. Statistical analysis
Cu content was determined by atomic absorption spectroscopy (Perkin Elmer, Norwalk, CT, USA).
Statistical analysis were performed with the software StatITCF (Institut Technique des CeÂreÂales et Fourrages,
2. Material and methods 2.1. Plant material and growth conditions Two cultivars of rapeseed (Brassica napus var. oleifera from Rustica Prograin GeÂneÂtique, France), bv. OrpheÂe and bv. CP 205, were surface sterilized for 5 min in 1/1 (v/v) ethanol and hydrogen peroxide mixture and washed three times with deionized water (Petchurkin N., Somova L., personal communication). They were then sown in sterile conditions in Magenta1 boxes (7711 cm, 30 seeds per box) and grown under controlled lighting conditions (18/6 h day/night) and temperature (25/ 20 C day/night). After sowing and 48 h in the dark, the growth chambers were lightened by ¯uorescent tubes (GROLUX and COOL-WHITE types, 1/1) at a photosynthetic photon ¯ux of 40010 mmol mÿ2 sÿ1. The sterile culture medium for germination and growth was agar (9%) (Sigma, France) in deionized water supplemented with CuSO4 at four concentrations: 0, 40, 80 and 250 mM. Cu sulfate was of analytic grade (Prolabo, France). The pH of these media was adjusted to 6.00.1 with NaOH 0.05 M. Plantlets were integrally harvested at day 7. 2.2. Root elongation evaluation
A. Baryla et al. / Environmental Pollution 109 (2000) 131±135
France). In the following text a single asterisk denotes signi®cance at the 0.05% level and a double asterisk signi®cance at the 0.01% level. 3. Results and discussion 3.1. Cu toxicity and Cu absorption Root elongation was measured on day 3 after sowing (Fig. 1). A threshold was observed for the low doses (0±10 mM) and a signi®cant root growth inhibition appeared when Cu concentration in the culture medium was higher than 10 mM. The slope changed when Cu concentrations were higher than 80 mM CuSO4 in the culture medium and from 350 to 500 mM CuSO4; here the root extremity had turned brown and was necrosed. Above these Cu concentrations, root development was totally inhibited. This assay indicates that plantlet growth is disturbed by Cu because of the negative correlation between the Cu added and the root elongation Ð measuring the root elongation is a toxicity test. Up to 250 mM, Cu is linearly absorbed by the plantlets (Fig. 2). Cu concentration measured in the tissues, 7 days after sowing, was indeed directly proportional to the Cu concentration in the culture medium (r2 0:9897).
133
As a matter of fact, all peroxides are not able to form MDA. The tri-unsaturated fatty acids are the main species which can form MDA with very weak reaction yields. Moreover, MDA can appear from processes other than peroxidation and it can be combined with other compounds (water, proteins, sugars, etc.), which involve artifactual ®gures in the measures. The last problem lies in the method itself because the TBA test is not entirely MDA speci®c. Several factors (pH, presence of iron salts, etc.) of the sample preparation or composition may aect the TBA±MDA reaction in various ways (Asakawa and Matsushita, 1979; Bird and Draper, 1984). Janero (1990) and Fernandez et al.
3.2. MDA quanti®cation A parallel increase between MDA content of the plantlets and Cu concentration in the plantlets (r2 0:9244) (Fig. 3) was observed. Nonetheless, for various reasons, care must be taken when examining these results.
Fig. 2. Cu amounts of 7-day-old rape plantlets (bv. OrpheÂe) grown on 0, 40, 80 or 250 mM Cu sulfate. Values are meansSD of three replicates.
Fig. 1. Root elongation of 3-day-old rape plantlets (bv. OrpheÂe) grown on a Cu-enriched culture medium. Values are meansSD of six replicates of six plantlets.
Fig. 3. TBARS amounts of 7-day-old rape plantlets (bv. OrpheÂe) grown on 0, 40, 80 or 250 mM Cu sulfate. Values are meansSD of ®ve replicates.
134
A. Baryla et al. / Environmental Pollution 109 (2000) 131±135
(1997) emphasized that numerous other compounds deriving, or not, from the lipid peroxidation processes can form adducts with the TBA (amino acids, sugars, DNA, etc.). Their precise chemical characteristics may be obtained by HPLC analysis but, more simply, the comparison of our sample absorbance spectrum with the one of pure MDA±TBA adducts con®rms qualitatively that we were not measuring only MDA±TBA adducts (data not shown). When Cu is added to the medium, ascorbate itself can in fact be degraded into a compound able to complex with the TBA (Gutteridge and Wilkins, 1982). The quanti®cation of TBA-reactive substances (TBARS) at 532 nm was used as an index of the biological samples, lipid peroxidation level (Janero, 1990). At least because of Cu interferences in the assay, this TBARS quanti®cation can be considered more as a stress index than as a peroxidation level index. 3.3. Hydroperoxy fatty acids quanti®cation Four hydroxy acids derived from hydroperoxides of linoleic acid (9-trans-trans-HODE, 13-trans-transHODE, 9-cis-trans-HODE and 13-cis-trans-HODE) and four hydroxyacids formed from linolenic acid (9HOTE, 12-HOTE, 13-HOTE, 16-HOTE) were quanti®ed by HPLC. Fig. 4 presents the total amounts of HOTE and HODE in the plantlets (bv. OrpheÂe). We obtained a signi®cant correlation (r2 0:997** for HODE and HOTE) between these results and the Cu concentration in the plantlets. Ethane and n-pentane outputs by 7-day-old rape plantlets are presented in Fig. 5. They increase linearly with the Cu concentration in the plantlets (r2 0:899 for ethane and 0.998** for n-pentane), but the slope and the correlation are greater for n-pentane. The ethane
Fig. 4. HODE (*) and HOTE (&) amounts of 7-day-old rape plantlets (bv. OrpheÂe) grown on 0, 40, 80 or 250 mM Cu sulfate. Values are meansSD of ®ve replicates.
output by plantlets exposed to 250 mM CuSO4 is signi®cantly dierent from the one by plantlets exposed to 0 or 80 mM. The n-pentane outputs are signi®cantly dierent between plantlets exposed to 0 or 40 and 80 mM CuSO4 or between 40 or 80 and 250 mM CuSO4. Relationship between volatile alkane production and hydroxy fatty acids has previously been established (DegouseÂe et al., 1995). Endogenous hydroxy acids, curently present in the plantlets, are not distinguished from the hydroperoxides reduced during the sample preparation. The measured hydroxy acids correspond to the amount of the hydroxy acids and of the products resulting from the hydroperoxide reduction. This analysis can be considered to provide more integrated than instantaneous data related to lipoperoxidation. A signi®cant correlation (r2 0:9518*) was obtained between the pentane production and the HODE level, both representative of the linoleic acid. There is a nonsigni®cant correlation (r2 0:6832) between ethane production and HOTE level although they both derive from linolenic acid. This fact can be explained by the physiological stage of the plantlets: the linolenic acid is predominant in the chloroplasts and these organites are only very few in the 7-day-old plantlets. 4. Conclusion To be convenient as a toxicity bioassay, measurement of the selected parameter must be as rapid and easy as possible. Overall, two dierent levels of the selected parameter must ®t with two dierent levels of environmental toxicity: a correlation is necessary between the
Fig. 5. Ethane (*) and n-pentane (&) outputs of 7-day-old rape plantlets (bv. CP 205) grown on 0, 40, 80 or 250 mM Cu sulfate. Values are meansSD of ®ve replicates.
A. Baryla et al. / Environmental Pollution 109 (2000) 131±135
response of the exposed organism and the toxicty of the environment. The measurement of root elongation is a very sensitive method for assessing environmental toxicity but it is not easy to perform in an opaque matrix like soil. The quanti®cation of lipid peroxidation of 7-day-old plantlets could then be a relevant bio-assay for Cu toxicity according to our results because of using aerial parts of the plants. Measuring n-pentane output by rape plantlets is as rapid as the TBARS quanti®cation (1 h is sucient to prepare 20 samples) but more precise and much more speci®c. It is a shorter method to perform than the hydroxy fatty acids determination (2 days are needed for assessing hydroxy acids of 20 samples) and it needs less plant material, which is practical when using plantlets. In conclusion, the volatile alkane production determination is rapid, does not need much fresh matter (1 g is sucient) and the data obtained are precise enough to evaluate the plantlets' lipid peroxidation status. We developed our test on agar solution supplemented with 0 to 250 mM Cu. It is dicult to compare with ®eld conditions because soil concentrations of Cu described in the bibliography represent total quantities of the metal but plants can only absorb Cu present in the soil solution (the bioavailable Cu). Nevertheless, we can compare the amounts of Cu absorbed by our plantlets and the amounts of Cu in plants already described by dierent authors. Kabata-Pendias and Pendias (1992) summarized the Cu contents found in plants growing in contaminated sites: these quantities may reach 590 ppm in sugar beet leaves. We did not ®nd any value directly for rape but these ®gures show that Cu concentrations in our plantlets (0.25±49 ppm) were not exceptionally high. Numerous chemical substances (other metals, etc.) or physical factors (chilling, radiations, etc.) are known to cause lipid peroxidation. For example, transition metals which are known to induce this phenomenon will certainly be detected by our bioassay: Fe (Aust et al., 1985), Zn, Cd (Chaoui et al., 1997), etc. A bioassay based on this metabolic parameter could then detect a set of unfavorable conditions, but our method needs a short cultivation duration and it is then possible to avoid injuries like chilling. Nevertheless, a strong standardization should be necessary. Acknowledgements The authors wish to thank Fernand Arnaud (Rustica Prograin GeÂneÂtique) for kindly providing several varieties of rape seeds, and Dr. Claire Sahut and Guy
135
GeÂniaut (CEA, Cadarche, France) for helping to measure Cu by AAS. Christine Battesti eciently assisted us in the HPLC analysis. References Asakawa, T., Matsushita, S., 1979. Coloring conditions of thiobarbituric acid test for detecting lipid hydroperoxides. Lipids 15, 137± 140. Aust, S.D., Morehouse, L.A., Thomas, C.E., 1985. Role of metals in oxygen radical reactions. Journal of Free Radicals in Biology and Medicine 1, 3±25. Bird, R.P., Draper, H.H., 1984. Comparative studies on dierent methods of malonaldehyde determination. In: Packer, L. (Ed.), Methods in Enzymology, Vol. 105: Oxygen Radicals in Biological Systems. Academic Press, Orlando, FL, pp. 299±305. Chaoui, A., Mazhoudi, S., Ghorbal, M.H., El Ferjani, E., 1997. Cadmium and zinc induction of lipid peroxidation and eects on antioxidant enzyme activities in bean (Phaseolus vulgaris L.). Plant Science 127, 139±147. DegouseÂe, N., TriantaphylideÁs, C., Montillet, J.-L., 1994. Involvement of oxidative processes in the signaling mechanisms leading to the activation of glyceollin synthesis in soybean (Glycine max). Plant Physiology 104, 945±952. DegouseÂe, N., TriantaphylideÁs, C., Starek, S., Iacazio, G., Martini, D., Bladier, C., Voisine, R., Montillet, J.-L., 1995. Measurement of thermally produced volatile alkanes: an assay for plant hydroperoxy fatty acid evaluation. Analytical Biochemistry 224, 524±531. Dhindsa, R.S., Plumb-Dhindsa, P., Thorpe, T.A., 1981. Leaf senescence: correlated with increased levels of membrane permeability and lipid peroxidation, and decreased levels of superoxide dismutase and catalase. Journal of Experimental Botany 32, 93±101. Fernandez, J., Perez-Alvarez, J.A., Fernandez-Lopez, J., 1997. Thiobarbituric acid test for monitoring lipid oxidation in meat. Food Chemistry 59, 345±353. Girotti, A.W., 1985. Mechanisms of lipid peroxidation. Journal of Free Radicals in Biology and Medicine 1, 87±95. Gutteridge, J.M.C., Wilkins, S., 1982. Cu-dependent hydroxyl radical damage to ascorbic acid. Formation of a thiobarbituric acid-reactive product. FEBS Letters 137, 327±330. Halliwell, B., Gutteridge, J.M.C., 1984. Oxygen toxicity, oxygen radicals, transition metals and disease. Biochemistry Journal 219, 1±14. Janero, D.R., 1990. Malondialdehyde and thiobarbituric acid-reactivity as diagnostic indices of lipid peroxidation and peroxidative tissue injury. Free Radical Biology and Medicine 9, 515±540. Kabata-Pendias, A., Pendias, H., 1992. Trace Elements in Soils and Plants, 2nd Ed. CRC Press LLC, Boca Raton, FL, pp. 95±108. Sterritt, R.M., Lester, J.N., 1980. The value of sewage sludge to agriculture and eects of the agricultural use of sludges contaminated with toxic elements. Science of the Total Environment 16, 55±90. Tarradellas, J., Bitton, G., 1997. Chemical pollutants in soils. In: Tarradellas, J., Bitton, G., Rossel, D. (Eds.), Soil Ecotoxicology. CRC Press, Boca Raton, FL, pp. 3±32. Tsay, C.C., Wang, L.W., Chen, Y.R., 1995. Plant response to Cu toxicity. Taiwana 40, 173±181. Weckx, J.E.J., Clijster, H.M.M., 1996. Oxidative damage and defense mechanisms in primary leaves of Phaseolus vulgaris as a result of root assimilation of toxic amounts of Cu. Physiologia Plantarum 96, 506±512.
www.elsevier.com/locate/envpol
Evaluation of lipid peroxidation as a toxicity bioassay for plants exposed to copper A. Baryla a,*, C. Laborde a, J.-L. Montillet b, C. TriantaphylideÁs b, P. Chagvardie a a CEA/Cadarache, DSV, DEVM, Laboratoire de Phytotechnologie, 13108 Saint-Paul-lez-Durance, France CEA/Cadarache, DSV, DEVM, Laboratoire de Radiobiologie VeÂgeÂtale, 13108 Saint-Paul-lez-Durance, France
b
Received 28 January 1999; accepted 15 June 1999
``Capsule'': Chemical analysis of soils may not be representative of copper availability to plants. Abstract Agricultural soils may contain toxic levels of copper (Cu) due to sewage sludge spreading or industrial pollution but chemical analyses may not be representative of Cu bioavailability, de®ned as the soil Cu fraction that plants can actually absorb (i.e. Cu fraction which is not strongly adsorbed to soil components). Lipid peroxidation caused by Cu in plants was investigated as a relevant bioassay of toxicity. Seven-day-old rapeseed plantlets were grown on Cu-supplemented medium in controlled conditions. Lipidperoxidation was assessed by measuring: (1) the 2-thiobarbituric acid (TBA)-reactive substances; (2) the hydroperoxy acids by HPLC analysis; and (3) the alkane outputs by gas chromatography. We ®rst veri®ed the correlation between the results obtained by each method and then discussed their advantages and disadvantages within the context of a bioassay, showing that the volatile alkane output measurement is the most precise and easy to perform method for this purpose. # 2000 Elsevier Science Ltd. All rights reserved. Keywords: Lipid peroxidation; Toxicity; Bioassay; Plants; Copper
1. Introduction Industrial pollution or sewage sludge spreading may have accumulated high quantities of toxic metals (Cd, Pb, Cu, Zn, etc) in agricultural soils (Tarradellas and Bitton, 1997). It is, therefore, of great concern for public health and plant growth to assess the toxicity of a soil. Numerous bioassays have been developed, allowing the detection of toxic substance eects on model organisms (which are sensitive species in general). For example, physiological (i.e. root elongation) or biochemical (enzymic inhibition or activation) parameters (Tsay et al., 1995), proven to be disturbed by the suspected pollution, have been analyzed. These bioassays are useful to de®ne the eects of dierent substances or to assess the toxicity of a complex matrix like a soil. Nevertheless, toxicity bioassays are never easy to perform. In the soil, root elongation is not easy to measure. Tests based on enzymic activities are often dicult and time consuming to perform and, moreover, they sometimes * Corresponding author.
need numerous reagents. It is important to establish rapid and simple new tests. Copper (Cu) is a microelement necessary for plant growth. Non-polluted soils contain 10±30 ppm Cu (dry wt.) but soils located near mining areas or metalprocessing industries may be contaminated by very large amounts of Cu: we can ®nd soils containing more than 3500 ppm. Its large use as a fungicide and in sewage sludge spreading may also have led to toxic levels of this metal in orchards or in wine-producing areas (KabataPendias and Pendias, 1992). Therefore, it becomes interesting to rapidly and easily assess the Cu toxicity of such soils. Due to the ability of Cu to act as an ecient generator of toxic oxygen species, it can initiate the lipid peroxidation process (Halliwell and Gutteridge, 1984; Aust et al., 1985; Girotti, 1985; Weckx and Clijster, 1996). The aim of our study was to determine if assessing the lipid peroxidation level of 7-day-old plantlets could be used as a Cu toxicity bioassay, as a ®rst step before testing presumed Cu-polluted soils. With this aim, we investigated three dierent methods for lipid peroxidation. Seven-day-old rapeseed plantlets were exposed to Cu and the plantlet malondialdehyde (MDA) level was
0269-7491/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved. PII: S0269-7491(99)00232-8
132
A. Baryla et al. / Environmental Pollution 109 (2000) 131±135
measured by the thiobarbituric acid (TBA) method which is relatively simple to carry out and has been used by many authors (Janero, 1990). We also determined the level of free and esteri®ed hydroperoxy fatty acids by high performance liquid chromatography (HPLC) analysis of the corresponding hydroxyacids. Finally, we assessed the plantlets' alkane outputs by gas chromatography. Ethane and n-pentane are actually well-known endproducts of o-3 and o-6-unsaturated hydroperoxy fatty acids and their production occurs via b-scission of the corresponding alkoxy radicals (DegouseÂe et al., 1995). The usefulness and validity of these methods as bioassays have been assessed and discussed.
Plantlets were washed with deionized water, dried at 90 C for 72 h and digested in a 1/3/3 (v/v/v) mixture of nitric and sulfuric acids and hydrogen peroxide according to Sterritt and Lester (1980). The samples were heated slowly to 80 C and the resulting solution analyzed by a ¯ame atomic absorption photometer (Perkin Elmer 3110). Three replicates per Cu treatment were made. 2.4. Lipid peroxidation evaluation
For this experiment, seeds were surface sterilized and sown in sterile conditions on the culture medium previously described along the diameter of Petri dishes (six seeds per dish). The Petri dishes were arranged with a 45 inclination in order to keep the roots straight. The root length was measured on day 3. Six replications, randomly disposed in the growth chamber, were performed for each Cu treatment. Eight Cu concentrations (from 0 to 500 mM) were tested.
The MDA content was determined by the TBA reaction with minor modi®cation of the method of Dhindsa et al. (1981). A 0.30-g crushed sample was homogenized in 1.25 ml trichloroacetic acid (TCA) (0.1%)±natrium dodecyl sulfate salt (SDS) (1%). The homogenate was centrifuged at 12,000 g for 5 min. To 300 ml aliquot of the supernatant was added 1 ml 20% TCA containing 0.5% TBA. The mixture was heated at 95 C for 30 min and then quickly cooled in an ice-bath. The absorbance at 532 nm was read spectrophotometrically (Uvikon 932, Kontron Instruments S. p. a., Italy) and the concentration of MDA was calculated using the extinction coecient of 155 mMÿ1 cmÿ1. TBA and 1,1,3,3-tetraethoxypropane (TEP) as a reference were supplied from Fluka (France), SDS from Merck and TCA from Prolabo, all products of analytic grade. The spectrophotometer was a model Uvikon 932 from Kontron Instruments S. p. a. (Italy). Free and esteri®ed hydroperoxy acids hydroperoxyoctadecadienoic acid (HPODE), hydroperoxyoctadecatrienoic acid (HPOTE) were reduced in their corresponding hydroxy derivatives hydroxyoctadecadienoic acid (HODE), hydroxyoctadecatrienoic acid (HOTE), and quanti®ed after alkaline hydrolysis by HPLC analysis as previously described (DegouseÂe et al., 1994). HPLC-grade solvents were supplied from Rathburn (Scotland, UK). For this method, 2.5 g fresh weight of plantlets was necessary for one measure. Thermal production of the volatile alkanes was carried out as previously described (DegouseÂe et al., 1995) from a sample of 1 g of plantlets. Ethane and n-pentane analyses were carried out separately by injection (120 C) of 1 ml of gas on a gas chromatogram (GC) (Delsi Di 200, France) run at 80 and 200 C, respectively, with He (30 ml minÿ1) as the carrier gas. The chromatography was performed on an aluminia column (1.5 m3.2 mm i.d.) and detection of the hydrocarbon gases was done by ¯ame ionization at 250 C. Calibrations were carried out by injection of 1 ml of 10 ppm ethane and 10.7 ppm n-pentane in N2 (Scotly Analyzed Gases, USA).
2.3. Determination of Cu in plantlets
2.5. Statistical analysis
Cu content was determined by atomic absorption spectroscopy (Perkin Elmer, Norwalk, CT, USA).
Statistical analysis were performed with the software StatITCF (Institut Technique des CeÂreÂales et Fourrages,
2. Material and methods 2.1. Plant material and growth conditions Two cultivars of rapeseed (Brassica napus var. oleifera from Rustica Prograin GeÂneÂtique, France), bv. OrpheÂe and bv. CP 205, were surface sterilized for 5 min in 1/1 (v/v) ethanol and hydrogen peroxide mixture and washed three times with deionized water (Petchurkin N., Somova L., personal communication). They were then sown in sterile conditions in Magenta1 boxes (7711 cm, 30 seeds per box) and grown under controlled lighting conditions (18/6 h day/night) and temperature (25/ 20 C day/night). After sowing and 48 h in the dark, the growth chambers were lightened by ¯uorescent tubes (GROLUX and COOL-WHITE types, 1/1) at a photosynthetic photon ¯ux of 40010 mmol mÿ2 sÿ1. The sterile culture medium for germination and growth was agar (9%) (Sigma, France) in deionized water supplemented with CuSO4 at four concentrations: 0, 40, 80 and 250 mM. Cu sulfate was of analytic grade (Prolabo, France). The pH of these media was adjusted to 6.00.1 with NaOH 0.05 M. Plantlets were integrally harvested at day 7. 2.2. Root elongation evaluation
A. Baryla et al. / Environmental Pollution 109 (2000) 131±135
France). In the following text a single asterisk denotes signi®cance at the 0.05% level and a double asterisk signi®cance at the 0.01% level. 3. Results and discussion 3.1. Cu toxicity and Cu absorption Root elongation was measured on day 3 after sowing (Fig. 1). A threshold was observed for the low doses (0±10 mM) and a signi®cant root growth inhibition appeared when Cu concentration in the culture medium was higher than 10 mM. The slope changed when Cu concentrations were higher than 80 mM CuSO4 in the culture medium and from 350 to 500 mM CuSO4; here the root extremity had turned brown and was necrosed. Above these Cu concentrations, root development was totally inhibited. This assay indicates that plantlet growth is disturbed by Cu because of the negative correlation between the Cu added and the root elongation Ð measuring the root elongation is a toxicity test. Up to 250 mM, Cu is linearly absorbed by the plantlets (Fig. 2). Cu concentration measured in the tissues, 7 days after sowing, was indeed directly proportional to the Cu concentration in the culture medium (r2 0:9897).
133
As a matter of fact, all peroxides are not able to form MDA. The tri-unsaturated fatty acids are the main species which can form MDA with very weak reaction yields. Moreover, MDA can appear from processes other than peroxidation and it can be combined with other compounds (water, proteins, sugars, etc.), which involve artifactual ®gures in the measures. The last problem lies in the method itself because the TBA test is not entirely MDA speci®c. Several factors (pH, presence of iron salts, etc.) of the sample preparation or composition may aect the TBA±MDA reaction in various ways (Asakawa and Matsushita, 1979; Bird and Draper, 1984). Janero (1990) and Fernandez et al.
3.2. MDA quanti®cation A parallel increase between MDA content of the plantlets and Cu concentration in the plantlets (r2 0:9244) (Fig. 3) was observed. Nonetheless, for various reasons, care must be taken when examining these results.
Fig. 2. Cu amounts of 7-day-old rape plantlets (bv. OrpheÂe) grown on 0, 40, 80 or 250 mM Cu sulfate. Values are meansSD of three replicates.
Fig. 1. Root elongation of 3-day-old rape plantlets (bv. OrpheÂe) grown on a Cu-enriched culture medium. Values are meansSD of six replicates of six plantlets.
Fig. 3. TBARS amounts of 7-day-old rape plantlets (bv. OrpheÂe) grown on 0, 40, 80 or 250 mM Cu sulfate. Values are meansSD of ®ve replicates.
134
A. Baryla et al. / Environmental Pollution 109 (2000) 131±135
(1997) emphasized that numerous other compounds deriving, or not, from the lipid peroxidation processes can form adducts with the TBA (amino acids, sugars, DNA, etc.). Their precise chemical characteristics may be obtained by HPLC analysis but, more simply, the comparison of our sample absorbance spectrum with the one of pure MDA±TBA adducts con®rms qualitatively that we were not measuring only MDA±TBA adducts (data not shown). When Cu is added to the medium, ascorbate itself can in fact be degraded into a compound able to complex with the TBA (Gutteridge and Wilkins, 1982). The quanti®cation of TBA-reactive substances (TBARS) at 532 nm was used as an index of the biological samples, lipid peroxidation level (Janero, 1990). At least because of Cu interferences in the assay, this TBARS quanti®cation can be considered more as a stress index than as a peroxidation level index. 3.3. Hydroperoxy fatty acids quanti®cation Four hydroxy acids derived from hydroperoxides of linoleic acid (9-trans-trans-HODE, 13-trans-transHODE, 9-cis-trans-HODE and 13-cis-trans-HODE) and four hydroxyacids formed from linolenic acid (9HOTE, 12-HOTE, 13-HOTE, 16-HOTE) were quanti®ed by HPLC. Fig. 4 presents the total amounts of HOTE and HODE in the plantlets (bv. OrpheÂe). We obtained a signi®cant correlation (r2 0:997** for HODE and HOTE) between these results and the Cu concentration in the plantlets. Ethane and n-pentane outputs by 7-day-old rape plantlets are presented in Fig. 5. They increase linearly with the Cu concentration in the plantlets (r2 0:899 for ethane and 0.998** for n-pentane), but the slope and the correlation are greater for n-pentane. The ethane
Fig. 4. HODE (*) and HOTE (&) amounts of 7-day-old rape plantlets (bv. OrpheÂe) grown on 0, 40, 80 or 250 mM Cu sulfate. Values are meansSD of ®ve replicates.
output by plantlets exposed to 250 mM CuSO4 is signi®cantly dierent from the one by plantlets exposed to 0 or 80 mM. The n-pentane outputs are signi®cantly dierent between plantlets exposed to 0 or 40 and 80 mM CuSO4 or between 40 or 80 and 250 mM CuSO4. Relationship between volatile alkane production and hydroxy fatty acids has previously been established (DegouseÂe et al., 1995). Endogenous hydroxy acids, curently present in the plantlets, are not distinguished from the hydroperoxides reduced during the sample preparation. The measured hydroxy acids correspond to the amount of the hydroxy acids and of the products resulting from the hydroperoxide reduction. This analysis can be considered to provide more integrated than instantaneous data related to lipoperoxidation. A signi®cant correlation (r2 0:9518*) was obtained between the pentane production and the HODE level, both representative of the linoleic acid. There is a nonsigni®cant correlation (r2 0:6832) between ethane production and HOTE level although they both derive from linolenic acid. This fact can be explained by the physiological stage of the plantlets: the linolenic acid is predominant in the chloroplasts and these organites are only very few in the 7-day-old plantlets. 4. Conclusion To be convenient as a toxicity bioassay, measurement of the selected parameter must be as rapid and easy as possible. Overall, two dierent levels of the selected parameter must ®t with two dierent levels of environmental toxicity: a correlation is necessary between the
Fig. 5. Ethane (*) and n-pentane (&) outputs of 7-day-old rape plantlets (bv. CP 205) grown on 0, 40, 80 or 250 mM Cu sulfate. Values are meansSD of ®ve replicates.
A. Baryla et al. / Environmental Pollution 109 (2000) 131±135
response of the exposed organism and the toxicty of the environment. The measurement of root elongation is a very sensitive method for assessing environmental toxicity but it is not easy to perform in an opaque matrix like soil. The quanti®cation of lipid peroxidation of 7-day-old plantlets could then be a relevant bio-assay for Cu toxicity according to our results because of using aerial parts of the plants. Measuring n-pentane output by rape plantlets is as rapid as the TBARS quanti®cation (1 h is sucient to prepare 20 samples) but more precise and much more speci®c. It is a shorter method to perform than the hydroxy fatty acids determination (2 days are needed for assessing hydroxy acids of 20 samples) and it needs less plant material, which is practical when using plantlets. In conclusion, the volatile alkane production determination is rapid, does not need much fresh matter (1 g is sucient) and the data obtained are precise enough to evaluate the plantlets' lipid peroxidation status. We developed our test on agar solution supplemented with 0 to 250 mM Cu. It is dicult to compare with ®eld conditions because soil concentrations of Cu described in the bibliography represent total quantities of the metal but plants can only absorb Cu present in the soil solution (the bioavailable Cu). Nevertheless, we can compare the amounts of Cu absorbed by our plantlets and the amounts of Cu in plants already described by dierent authors. Kabata-Pendias and Pendias (1992) summarized the Cu contents found in plants growing in contaminated sites: these quantities may reach 590 ppm in sugar beet leaves. We did not ®nd any value directly for rape but these ®gures show that Cu concentrations in our plantlets (0.25±49 ppm) were not exceptionally high. Numerous chemical substances (other metals, etc.) or physical factors (chilling, radiations, etc.) are known to cause lipid peroxidation. For example, transition metals which are known to induce this phenomenon will certainly be detected by our bioassay: Fe (Aust et al., 1985), Zn, Cd (Chaoui et al., 1997), etc. A bioassay based on this metabolic parameter could then detect a set of unfavorable conditions, but our method needs a short cultivation duration and it is then possible to avoid injuries like chilling. Nevertheless, a strong standardization should be necessary. Acknowledgements The authors wish to thank Fernand Arnaud (Rustica Prograin GeÂneÂtique) for kindly providing several varieties of rape seeds, and Dr. Claire Sahut and Guy
135
GeÂniaut (CEA, Cadarche, France) for helping to measure Cu by AAS. Christine Battesti eciently assisted us in the HPLC analysis. References Asakawa, T., Matsushita, S., 1979. Coloring conditions of thiobarbituric acid test for detecting lipid hydroperoxides. Lipids 15, 137± 140. Aust, S.D., Morehouse, L.A., Thomas, C.E., 1985. Role of metals in oxygen radical reactions. Journal of Free Radicals in Biology and Medicine 1, 3±25. Bird, R.P., Draper, H.H., 1984. Comparative studies on dierent methods of malonaldehyde determination. In: Packer, L. (Ed.), Methods in Enzymology, Vol. 105: Oxygen Radicals in Biological Systems. Academic Press, Orlando, FL, pp. 299±305. Chaoui, A., Mazhoudi, S., Ghorbal, M.H., El Ferjani, E., 1997. Cadmium and zinc induction of lipid peroxidation and eects on antioxidant enzyme activities in bean (Phaseolus vulgaris L.). Plant Science 127, 139±147. DegouseÂe, N., TriantaphylideÁs, C., Montillet, J.-L., 1994. Involvement of oxidative processes in the signaling mechanisms leading to the activation of glyceollin synthesis in soybean (Glycine max). Plant Physiology 104, 945±952. DegouseÂe, N., TriantaphylideÁs, C., Starek, S., Iacazio, G., Martini, D., Bladier, C., Voisine, R., Montillet, J.-L., 1995. Measurement of thermally produced volatile alkanes: an assay for plant hydroperoxy fatty acid evaluation. Analytical Biochemistry 224, 524±531. Dhindsa, R.S., Plumb-Dhindsa, P., Thorpe, T.A., 1981. Leaf senescence: correlated with increased levels of membrane permeability and lipid peroxidation, and decreased levels of superoxide dismutase and catalase. Journal of Experimental Botany 32, 93±101. Fernandez, J., Perez-Alvarez, J.A., Fernandez-Lopez, J., 1997. Thiobarbituric acid test for monitoring lipid oxidation in meat. Food Chemistry 59, 345±353. Girotti, A.W., 1985. Mechanisms of lipid peroxidation. Journal of Free Radicals in Biology and Medicine 1, 87±95. Gutteridge, J.M.C., Wilkins, S., 1982. Cu-dependent hydroxyl radical damage to ascorbic acid. Formation of a thiobarbituric acid-reactive product. FEBS Letters 137, 327±330. Halliwell, B., Gutteridge, J.M.C., 1984. Oxygen toxicity, oxygen radicals, transition metals and disease. Biochemistry Journal 219, 1±14. Janero, D.R., 1990. Malondialdehyde and thiobarbituric acid-reactivity as diagnostic indices of lipid peroxidation and peroxidative tissue injury. Free Radical Biology and Medicine 9, 515±540. Kabata-Pendias, A., Pendias, H., 1992. Trace Elements in Soils and Plants, 2nd Ed. CRC Press LLC, Boca Raton, FL, pp. 95±108. Sterritt, R.M., Lester, J.N., 1980. The value of sewage sludge to agriculture and eects of the agricultural use of sludges contaminated with toxic elements. Science of the Total Environment 16, 55±90. Tarradellas, J., Bitton, G., 1997. Chemical pollutants in soils. In: Tarradellas, J., Bitton, G., Rossel, D. (Eds.), Soil Ecotoxicology. CRC Press, Boca Raton, FL, pp. 3±32. Tsay, C.C., Wang, L.W., Chen, Y.R., 1995. Plant response to Cu toxicity. Taiwana 40, 173±181. Weckx, J.E.J., Clijster, H.M.M., 1996. Oxidative damage and defense mechanisms in primary leaves of Phaseolus vulgaris as a result of root assimilation of toxic amounts of Cu. Physiologia Plantarum 96, 506±512.