Environ. Sci. Technol. 2005, 39, 1757-1763
Cu and Cd Effects on the Earthworm Lumbricus rubellus in the Laboratory: Multivariate Statistical Analysis of Relationships between Exposure, Biomarkers, and Ecologically Relevant Parameters MALYKA GALAY BURGOS,* CAROLE WINTERS, STEPHEN R. STU ¨ RZENBAUM, PETER F. RANDERSON, PETER KILLE, AND A. JOHN MORGAN School of Biosciences, Cardiff University, Cardiff CF10 3US, U.K.
This study sets out to examine the potential of a suite of novel molecular biomarkers as early warning indicators of environmental state and damage. Transcriptional responses of four genes, metallothionein 1 and 2, amine oxidase, and the lysosomal associated glycoprotein, were measured in the earthworm Lumbricus rubellus exposed to increasing concentrations of cadmium and copper in OECD soil. These responses were compared to metal body concentrations and lifecycle parameters: survival, cocoon production, and growth. Adverse physiological effects were observed at concentrations 1/3rd to 1/10th those of the artificial soil LC50. Multivariate statistics, principal component analysis (PCA), was used to investigate the correlations between the different variables. Three key components were derived explaining 77.6% of the variance, with component 1 contributing 32.4%, component 2 contributing 26.7%, and component 3 contributing 18.5%. These components were interpreted in terms of population health, pollutant exposure, and detoxification pathways, respectively. It is proposed that the use of such a suite of biomarkers could serve as indicators of the “health” of the soil environment and provide early warning signals of potential danger to the biota or as a means of monitoring soil remediation.
cocktails of inorganic and organic residues (7-12). Two practical intrinsic constraints with such tests are that the exposure periods are relatively long in most cases, and second, they are not readily suited for in situ measurement in field populations. Much effort has been expended, therefore, to develop biomarkers of earthworm functional condition as sensitive, rapid-response, and predictive tools in environmental diagnostics (13, 14). Many earthworm biomarkers target responses at the biochemical (15) and cellular (16, 17) levels of biological organization. Recently, partly due to swift technological advances, the use of molecular genetic biomarkers is gaining acceptance (18-22). The advantages of this approach are that first, once a protocol has been established for one target gene, protocols for several other genes are easily acquired; this is not the case with enzyme assays, for example. Second, gene-expression perturbations often precede, and underpin, alterations in functional parameters at the cellular level, through the whole organism, to the ecological level of organization. Biomarkers provide measures of altered states in individual organisms evoked by exposure to bioactive fractions of toxicants. Environmental stress, whether considered at the physiological or ecological levels, should be viewed as a multidimensional entity (23), and it therefore follows that the disturbance of an individual’s functional state is best tracked by measuring several rather than single variables. Galay Burgos et al. (20) described a real-time quantitative polymerase chain reaction for measuring the transcription of two specific genes in the earthworm, Lumbricus rubellus, exposed to copper and cadmium in experimentally spiked soil using the biomarkers metallothionein isoform 2 (wMT2) and the mitochondrial large ribosomal subunit (MLRS). The main purpose of the present paper was 2-fold: first, to measure genetic responses to cadmium and copper in L. rubellus under laboratory conditions using an extended suite of biomarkers and second, to correlate the changes in biomarker expression with accumulated tissue-metal concentrations and life cycle parameters (survival, growth, reproductive output) through multivariate principal component analysis (PCA). The choice of biomarkers was based on previous work in this laboratory that identified a number of genes that were upregulated in response to long-term metal exposure: metallothionein isoforms 1 and 2 (wMT-1, wMT2), amine oxidase (AOX), and lysosomal glycoprotein (LYS) (18).
Materials and Methods Introduction The disposal of hazardous waste material presents major technical, environmental, and economic challenges (1). Pressures to release contaminated land for housing and amenities within the framework of legislation protecting human and environmental health demand the development of robust biological tools for assessing the usability of these soils and monitoring any attempts at remediation. Earthworms are important organisms in terrestrial ecotoxicology, with approved protocols available for acute and sublethal testing (2-6). A number of whole-worm performance indicators, including altered behavioral patterns, have been used under standardized laboratory conditions to evaluate the net toxicities of field soils contaminated with ill-defined * Corresponding author phone:
[email protected]. 10.1021/es049174x CCC: $30.25 Published on Web 01/27/2005
029 2087 6800; e-mail:
2005 American Chemical Society
Whole Earthworm Life Cycle Parameters. Only sexually mature (i.e. fully clitellate) earthworms, Lumbricus rubellus (Oligochaeta: Lumbricidae), were used in this study. Experiments in which mortality, adult growth, and number of cocoons produced per unit time were evaluated following the modified OECD protocol (24) described by van Gestel et al. (25) and Gibbs et al. (ref 7, see below). The Exposure Substrate. Artificial soil was made according to the OECD (24) protocol:10% dry weight of finely ground peat (pH 5.5-6.0); 20% dry weight kaolin clay (kaolinite content 100%); 70% dry weight washed, nonsharp industrial sand (50% of the particles between 50 and 200 microns); pulverized CaCO3 to achieve a pH of 6.0 ( 0.5. Deionized water was added to an overall moisture content of about 35-40% of the total dry weight of the substrate. Maintenance of the Colony of Earthworms Prior to Experiment. Prior to the metal exposure experiments, purchased earthworms (Neptune-Ecology, Ipswich, U.K.) VOL. 39, NO. 6, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1757
TABLE 1. Sequences of the Probes and Primers Used for the Real-Time Quantitative PCR Assaysa gene
primer/probe
β-actin
forward primer reverse primer TaqMan probe forward primer reverse primer TaqMan probe forward primer reverse primer TaqMan probe forward primer reverse primer TaqMan probe forward primer reverse primer TaqMan probe
wMT-1 wMT-2 LYS AOX
5′-CGTCGACAATGGATCTGGAA-3′ 5′-CGACCATCACACCCTGATGA-3′ 5′-TTTCGCCGGTGATGACGCTCC-3′ 5′-ACCAACTGCAGATGCCTGAAA-3′ 5′-CTTGGGAATCAGCACAGCAA-3′ 5′-CTTCCTGCAGTTTGGCGAACATTCATC-3′ 5′-CGATTTGATACATTTTTATATATGGACGA-3′ 5′-TGAATTTAGCTCTGAATGTTTCTTGC-3′ 5′-ACTTTACCACATACATCTTTTGATTTCGTTTGCCA-3′ 5′-TGCTTATGGTGTCCTCCTT-3′ 5′-CCATCAGAGCGAAGTTAGT-3′ 5′-CCAATCATCGTGGGCGCTTGTCTC-3′ 5′-TTCCGCAAGGGAAAGATTCA-3′ 5′-TCCGACGACCACCAGAATG-3′ 5′-CCGCATCCAACGACTCGATCTGATG-3′
a Unmodified oligonucleotides were used both for forward and reverse primers, while TaqMan probes incorporate the fluorescent molecules 6-carboxyfluorescein (FAM) at the 5′-end and the complimentary quencher 6-carboxytetramethyl-rhodamine (TAMRA) at the 3′-end.
were maintained in a soil made up of a 5:1:1 ratio of topsoil, composted bark, and peat (British Soil Ltd., Wenvoe, S. Wales), to which alfalfa and sheep manure were added as food sources as required. Healthy colonies of about 1000 individual adult Lumbricus rubellus were kept for several weeks in deep fiberglass tanks of 1 × 1.5 m at 19 °C and a 12:12 light:dark cycle. General Experimental Design. Cadmium and coppers CdCl2.21/2H2O and CuCl2‚2H2O in deionized H2Oswere added separately to give the following concentrations: 5, 25, 125, and 200 µg g-1 dry weight, to artificial soil using the OECD protocol. For each metal concentration, 100 g dry weight equivalent of OECD soil were placed in 20 polythene bags (16.5 × 15.0 cm) punctured with a template of 23 pins in each side for aeration. Besides the 20 bags for each of the concentrations of cadmium and copper, 20 bags containing unspiked OECD soil were allocated for control worms. The bags were left in the constant temperature room for 5 days equilibration prior to the introduction of earthworms. After this equilibration period, 4.5 g of dry, ground, sheep manure was added to each bag as food source. Two days before the end of the equilibration period, two adult worms of similar weights per bag ()180 pairs) were rinsed briefly in deionized water and placed on wet filter paper (changed once, after 24 h) in Petri dishes to fully hydrate and void their guts. The moisture content of each bag was checked weekly by weighing; losses exceeding 5% were replenished by the addition of the appropriate volume of deionized water. This exposure protocol was adapted from that described by Gibbs et al. (7). A preliminary experiment was carried out on a separate group of worms to calculate dry mass from hydrated mass. Linear regression yielded a fresh-to-dry mass conversion factor of 0.152. Adult weight change (growth) during the exposure period was based on dry mass estimates. The bags and their contents were incubated in a temperature-controlled room at 17 °C on a 12:12 h light:dark cycle. After 21 days the content of each bag was emptied into a tray. Surviving adults were removed and mortality recorded. Surviving worms were then placed (either as surviving pairs or as individuals) on moist filter paper in Petri dishes and left for 2 days to hydrate and void their guts and weighed (Mettler AE 163 balance). Earthworms were frozen in liquid nitrogen and stored either at -20 °C for whole body metal measurements or at -70 °C for biomarker analysis. Reproductive output was measured by counting the number of cocoons in each bag. pH measurements of the soil were taken before and after the experiment. 1758
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 6, 2005
Metal Analysis. Approximately 0.5 g of soil from each exposure bag and whole worms were dried to constant weight at 60 °C in preweighed, acid washed Pyrex tubes and digested with concentrated nitric acid (Aristar grade, BDH) on a sand bath at 80 °C as described by Galay Burgos et al. (20). Soil digests were filtered prior to analyzing by atomic absorption spectrophotometry (Varian Spectr AA-100). Analysis of standard samples (Marine muscle tissue, GBW08571, and polluted farmland soil, GBW08303; State Bureau of Technical Supervision, People’s Republic of China) indicated that the preparative and analytical procedures yielded values within 7% of certified metal concentrations. Isolation of Total RNA and mRNA. To study the sublethal and chronic effects on the “healthy” survivors following exposure, the method described in Galay Burgos et al. (20) was followed. Earthworm parameters were assessed by studying four molecular biomarkers isolated by cDNA PCR: two metallothioneins (wMT-1, wMT-2) (26), amine oxidase (AOX) (27), and the membrane bound lysosomal glycoprotein (LYS) (18). To minimize RNA degradation all preparations were performed in a room designated for RNA handling only. Where possible, all materials required, such as pipets, pipet tips, microcentrifuge tubes, and reagents, were freshly autoclaved twice at 120 °C and 15 psi for 20 min. Total RNA was isolated from 300 mg of tissue removed by direct caudal amputation of the worms frozen in liquid nitrogen using the TRI-Reagent (Sigma, Pool, Dorset, U.K.) according to the manufacturer’s protocol. Subsequent isolation of mRNA as well as transcription of cDNAs and the quantitative real time PCR using the TaqMan PCR reagent system (PE Biosystems, Warrington, U.K.) was performed on an ABI PRISM 7700 Sequence Detection System (PE Biosystems) as described in Galay Burgos et al. (20). To detect the expression profiles of the selected genes primers and probes, shown in Table 1, were designed and calibrated as described in Galay Burgos et al. (20). Standards were prepared by diluting each purified plasmid to a known concentration in the range of 1 ng to 1 fg. PCR was performed on duplicate samples of each standard. A regression line was generated by plotting cycle number needed to attain a threshold fluorescence pertaining to the mid logarithmic portion of the amplification against log [molecules of target gene]. Quantitative PCR was performed on 2.5 µL cDNA generated as described above. The amount of target present in each reverse transcribed cDNA was quantified by regression analysis, extrapolating each “unknown” over the standard range. All samples quantified were analyzed in triplicate. The
TABLE 2. Effect of Increasing Sublethal Concentrations of the Metals Cadmium and Copper in the Artificial Soil on the Accumulation of Metal, the Survival, Growth, and Reproduction of Lumbricus rubellus after 21 Days of Exposurea concentration (ug/g)
soil
worm mean SD
control 3.0 ( 2.1 Cu 5 13.5 ( 3.8 Cu 25 16.6 ( 4.5 Cu 125 43.6 ( 4.9 Cu 200 60.7 ( 28.4 Cd 5 19.5 ( 6.7 Cd 25 77.6 ( 13.9 Cd 125 228.0 ( 57.1 Cd 200 303.9 ( 44.9 a
worm survival
% 67.5 70.0 55.0 70.0 32.5 50.0 72.5 42.5 67.5
no. no. of of pairs singles total 12 12 10 10 3 8 11 7 11
3 4 2 8 7 4 7 3 5
27 28 22 28 13 20 29 17 27
worm growth
reproduction
initial Dwt (g) mean SD
final Dwt (g) mean SD
(g)
0.229 ( 0.075 0.210 ( 0.083 0.208 ( 0.028 0.252 ( 0.0116 0.181 ( 0.031 0.199 ( 0.40 0.232 ( 0.079 0.236 ( 0.116 0.206 ( 0.044
0.241 ( 0.065 0.209 ( 0.077 0.205 ( 0.026 0.250 ( 0.089 0.189 ( 0.033 0.208 ( 0.024 0.236 ( 0.073 0.252 ( 0.098 0.199 ( 0.029
0.0122 -0.0016 -0.0030 -0.0017 0.088 0.0098 0.0041 0.0164 -0.0073
no. pairs cocoons cocoons of reproper per worm/ cocoons ducing worm week
% 5.3486 -0.7488 -1.4265 -0.6598 4.8551 4.9244 1.7686 6.9521 -3.5590
]*
]]
b ns ns ns b
63 52 54 41 1 30 40 22 44
12 12 10 9 1 8 8 6 11
2.63 2.17 2.70 2.28 0.50 1.88 2.50 1.83 2.00
0.88 0.72 0.90 0.76 0.17 0.63 0.83 0.61 0.67
The results represent the mean ( standard deviation for the number of replicates (see Results). b p < 0.01 (Student’s t-test).
resulting absolute values were normalized for differences in reverse transcription efficiencies, which may vary between 5 and 90% (28), by using the level of β-actin expressed. Therefore, differences observed in the “invariant” message can be ascribed to variable efficiencies during reverse transcriptions and the amount of the “unknown” transcript reported in relative terms. The analysis of specific isoforms of wMT-1 and wMT-2 genes was facilitated by designing probe and primer sets located in unconserved areas of the sequence rendering them isoform-specific which in turn was validated by analysis of the cross-reactivity in the primer and probe set for wMT-1 and wMT-2 gene targets as described in Galay Burgos et al. (20). Statistics. Differences between pairs of sample means were tested using Student’s t-test, after initially checking for between sample homogeneity of variances using the F-test. For groups of more than 2 samples, one way analysis of variance (ANOVA) was used to assess differences between means, after testing for homogeneity of group variances and normality of residuals. When a functional relationship between two variables was suspected, the data were analyzed by regression analysis using the method of least squares. The multivariate analysis, Principal Component Analysis (PCA), was used to identify underlying variables (components) that explain the intercorrelations within a set of observed variables (measured data). PCA was performed on the standardized data (correlation matrix) of a total of 8 variables (4 biomarkers, 3 life cycle parameters, and tissue metal concentrations), incorporating both Cu and Cd data sets. Statistical models relating genetic and ecological parameters have rarely been reported. In this case the model includes 8 variables in total and is valid for the range of metal concentrations used in this study. The statistical package used throughout this study was SPSS 10 (SPSS Science Software, Erkath, Germany).
Results Accumulation of Cadmium and Copper. Table 2 shows the mean concentration of cadmium and copper for the control worms together with the concentrations in the tissues of Lumbricus rubellus exposed for 21 days to 5, 25, 125, and 200 µg g-1 (nominal concentrations) of each metal. These values are well below the reported 14-day OECD artificial soil LC50 for cadmium (643 µg‚g-1) and copper (1843 µg‚g-1) in the earthworm Eisenia fetida (29). The actual concentrations in the soil were found to be 0, 3.9, 18.0, 98.0, and 155.2 µg Cu g-1 and 0, 4.2, 21.2, 63.4, and 166.5 µg Cd g-1, respectively. The systematically lower ‘actual’ compared with ‘nominal’ concentrations indicates that the soils were not absolutely dry when spiked with metals.
The concentration of cadmium and copper in the earthworm tissue increased with increasing metal concentration in the artificial soil (ANOVA, p < 0.01 and p < 0.05 respectively). Mortality. There was no significant correlation (regression analysis, P > 0.05) between survival of worm pairs and initial dry weight of the replicates for any of the metal exposures nor for the controls. For the cadmium experiments, there was no additional mortality at the highest test concentration (200 µg Cd g-1) in which 67.5% survived. For the remaining concentrations 50%, 72%, and 42.5% of the worms exposed to 5, 25, and 125 µg Cd g-1 survived. Of the 20 pair-replicates exposed to 5 µg Cd g-1 50% survived, of which four were single survivors and 8 were pairs. With 125 µg Cd g-1 in the soil only 42.5% survived of which 3 were single worms and 7 pairs; at the highest concentration of 200 µg Cd g-1 65% survived, 5 singles and 11 pairs. For the worms exposed to the different concentrations of copper in the soil only 32.5% survived at the highest concentration (200 µg g-1), while 70%, 55%, and 67.5% survived the remaining exposures of 25, 100, and 125 µg g-1, respectively. Out of the 20 worm pair-replicates exposed to 5 µg Cu g-1 in the soil 70% survived, of which 4 were singles and 12 were pairs; for the next higher concentration (25 µg Cu g-1) 55% survived, of which 2 were singles and 10 pairs; for the 125 µg Cu g-1 67.5% survived, of which 8 were singles and 10 pairs, only 32.5% survival was found for the highest concentration (200 µg g-1) comprising 3 surviving pairs and 7 singles. Growth. Only surviving pairs and not single survivors were taken into account in the calculation of dry weight. In the control group, mean adult growth was 5.35% (Table 2). In the cadmium exposure experiments, those worms exposed to the three lowest concentrations increased in weight, while for the highest concentration (200 µg Cd g-1) the worms lost nearly 3.5% of their original weight (p < 0.01). For the copper exposures, there was a slight decrease in weight for worms exposed to the lowest three concentrations (5, 25, and 125 µg g-1), while an increase of nearly 5% was found for worms exposed to 200 µg Cu g-1 (p < 0.001). Cocoon Production. No significant correlation (regression analysis, P > 0.05) was found between cocoon production and initial dry weight. Cocoon production rate was not significantly lower (ANOVA, P > 0.05) than the control values in worms exposed to the two metals at any concentration, although for the highest concentration of copper there was a decrease (Table 2). This decrease may be misleading and can be explained by the fact that the number of worm pairs surviving was only three from which only one cocoon was produced. It can be noted that metal exposure caused an inverse relationship between reproduction and growth. For the control group there was a 5.35% increase in growth (dry VOL. 39, NO. 6, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1759
FIGURE 1. Relative expression of five target sequences measured in Lumbricus rubellus when exposed to different and increasing sublethal concentrations of cadmium and copper in the soil. The normalized expression (log10) of wMT-1 (A), wMT-2 (B), amine oxidase (AOX) (C), and lysosomal glycoprotein (LYS) (D) is shown in response to copper (circles, dashed lines) and cadmium (squares, solid lines) concentrations. The data represent the mean (( standard deviation) of the β-normalized expression of each biomarker and are the result of three replicates for each of the three worms. weight change), and they produced 2.63 cocoons per worm. For the three lower concentrations of copper in the soil the worms failed to grow, but the cocoon production was similar to the control group, except at the highest concentration of copper in which no cocoons were produced but the worms did grow (see Table 2). For cadmium exposure in the soil there is an inverse correlation between cocoon production and growth, for example, relatively high growth (4.93% dry weight change) at the lowest cadmium concentration in the soil resulted in a lower cocoon production per worm (1.88). At the highest concentration of cadmium in the soil there was a negative growth (-3.55%) although cocoon production per worm (2.00) was slightly lower than normal (2.63). Biomarkers. Metallothionein-1 and -2 (wMT-1 and wMT2) amine oxidase (AOX), lysosomal glycoprotein (LYS), and β-actin expression were analyzed in three different worms for each metal concentration and for each metal (Cu and Cd) using real-time PCR. β-Actin was selected as an invariant internal control since real-time quantitative PCR has already established this gene 1760
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 6, 2005
as invariant in worms exposed to metal contaminated soil (30). Biomarker transcript expression was therefore normalized by ratioing it to the expression of β-actin. wMT-1 expression (Figure 1A) remained more or less constant throughout the different concentrations of cadmium in the soil, although for copper an initial increase was followed by a decrease at the two highest concentrations. This was a general finding at high concentrations of copper for all biomarkers. On the other hand, analysis of wMT-2 showed (Figure 1B) an increasing expression with increasing cadmium concentration in the soil while for copper, after an initial increase for the first two metal concentrations to which the worms were exposed, there was a decline in expression. Amine oxidase (AOX) expression (Figure 1C) increased with increasing concentration of cadmium in the soil except for the highest concentration of cadmium in which the normalized expression was lower than 1. The response to copper exhibited the pattern of an initial increase followed by a decrease for the two highest concentrations. The lysosomal glycoprotein (LYS) expression did not change for any of the
TABLE 3. Component Matrixa components variables
1
total survival dry weight change cocoons/worm/week metallothionein 1 (wMT-1) metal concentration in worm metallothionein 2 (wMT-2) amine oxidase (AOX) lysosomal glycoprotein (LYS)
0.813 -0.804 0.785 0.544
2
3
0.945 0.919 0.485
0.906 0.677
a Components extracted by the multivariate statistical procedure, principal component analysis, used to investigate the correlations between the different variables. Each component is assigned to a physiological function corresponding to the likely role in the organism (see Discussion).
FIGURE 2. The plot represents the results of the principal component analysis (PCA). The analysis revealed three components which comprise the three axes in the figure. The first component consisted of a positive correlation between the variables total survival (TotSurv), cocoon production/worm/week (Cocww), metallothionein-1 (wMT-1), and LYS (see Table 3). All these variables are negatively correlated with dry weight change. The second component includes metallothionein-2 (wMT-2) and metal concentration in the worm, while the third component consists of amine oxidase (AOX) and lysosomal glycoprotein (LYS). The three component solution explained a total of 77.6% of the variance, with component 1 contributing 32.4%, component 2 contributing 26.7%, and component 3 contributing 18.5%. different concentration of cadmium exposures in the soil (Figure 1D). As for the copper expression, there was a large increase for the first two concentrations (nominally 5 and 25 µg/g in the soil), while for the higher concentrations (nominally 125 and 200 µg/g in the soil) the expression was not different from the control levels. Principal Component Analysis (PCA). Several of the correlation coefficients between the 8 variables were greater than 0.3 and Bartlett’s test of sphericity was significant (P ) 0.0002), both indications that PCA would be appropriate. PCA resulted in three groups with eigenvalues >1, whereas those of subsequent components were much less. The relationships between the variables are illustrated in Figure 2. Each variable is positioned in three-dimensional space in which the axes represent the three components. The three component solution explained a total of 77.6% of the variance, with component 1 contributing 32.4%, component 2 contributing 26.7%, and component 3 contributing 18.5%. This indicates that a simple 3-axis structure is adequate to explain the intercorrelations in the data set, as indicated by the component loadings (eigenvectors) of the measured variables on the components (Table 3). Each variable showed a strong loading on one component only, except LYS which loaded strongly onto both components 1 and 2.
Discussion Most organisms when exposed to pollutants initiate compensatory mechanisms to ensure that they continue to function normally. Biochemical and physiological tolerance depends, at least in part, on those genes responsible for these adaptive mechanisms being regulated or differentially expressed. Measuring the expression of biochemical markers involved in the handling, biotransformation, and/or excretion of the pollutant that provoked the response can help assess the extent to which the organism has been stressed by the
new environmental challenge. For a biomarker to be predictive of the consequences of the exposure, it must be possible to relate this response directly to a change in the growth, reproductive output, or energy utilization. This study was designed to link life cycle parameters (survival, growth, and reproduction) and gene expression profiles as biomarkers after an exposure to different and sublethal cadmium and copper concentrations in OECD soil. With the particular experimental design adopted, it was found that adverse physiological effects of significance to the survival of the population were observed at concentrations 1/3rd to 1/10th those of the artificial soil LC50 (7). Metal exposure caused an inverse relationship between reproduction and growth; if the worms did not grow, then they produced cocoons (see Table 2). If cocoon production dropped, they continued to grow even though, since they were mature when the experiments began, they would naturally be expected to reproduce rather than grow. Only the highest concentration of copper interfered with this relationship. Perhaps at this level of copper, the worms channelled energy into growth rather than reproduction, because the copper interfered with the reproductive system, although the mechanisms involved are not yet fully understood (see ref 31). Once the reproductive capacity of the worms is halted, the population of Lumbricus rubellus would die out in time. With cadmium, on the other hand, at the highest concentration (200 µg Cd g-1), there was negative growth, and, although still producing cocoons, it can be speculated that the population would nevertheless die out, since their juveniles will never attain the minimum weight (g200 mg) to be able to reproduce. A method for measuring alterations in the expression of two L. rubellus transcripts, metallothionein isoform-2 (wMT2) and mitochondrial large ribosomal subunit (MLRS), in response to copper and cadmium exposure in a spiked commercial soil was developed by Galay Burgos et al. (20). The ensemble of molecular genetic biomarkers was expanded in the present study to include a second metallothionein isoform (wMT-1), the membrane bound lysosomal glycoprotein (LYS), and amine oxidase (AOX). This illustrates clearly the versatility of the TaqMan real-time quantitative PCR methodology and confirms the claim (20) that it is readily applicable for monitoring stressor induced alterations in the gene-expression profile of an ecologically relevant soildwelling macroinvertebrate. The biomarkers used in this study were chosen for their potentially predictive nature. Metallothioneins are associated with metal handling, sequestration, and detoxification (32, 33). Stu ¨rzenbaum et al. (26, 34) and Morgan et al. (35) showed that earthworms have two functionally differentiated metallothionein isoforms, wMT-1 and wMT-2. Although most VOL. 39, NO. 6, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1761
work has concentrated on the isoform wMT-2 and consequently little is known about the physiological functions of wMT-1, there is increasing evidence of a contrasting relationship between the two cysteine-rich metallothionein isoforms. Stu¨rzenbaum et al. (26) showed that wMT-2 protein possesses two regions that are probable N-glycosylation sites, sites that are absent in wMT-1. This observation led to the hypothesis that wMT-2 can be delivered with precision to the lysosomal system thus bearing some of the hallmarks of metal detoxification. Subsequent findings (34) that, although both wMT-1 and wMT-2 bind 6 atoms of cadmium per molecule, the stability of the cadmium binding is significantly higher in the latter, provided further support for the hypothesis, as did the finding that wMT-2 expression is more cadmium responsive. From these observations it would be reasonable to predict that wMT-2 rather than wMT-1 would display a stronger correlation with the tissue burdens of two sulfur-seeking metals. A similar detoxification role is therefore attributable to the lysosomal system, such as membrane bound lysosomal protein (LYS) (26), and mixed function oxidases systems (AOX) (36) such as the amine oxidase. The dose-response curves (see Figure 1) recorded for each of the molecular-genetic biomarkers in the present study were found to display typical hormetic-like biphasic curves (37, 38), where a potential toxicant has stimulatory effects at low doses but inhibitory effects at higher doses. This effect, which was particularly pronounced during copper exposures, presents interpretational challenges, especially since the amplitudes of the curves and the relative widths of their stimulatory and inhibitory phases varies from biomarker to biomarker and from toxicant to toxicant. One of the key recommendations of the Third International Workshop on Earthworm Ecotoxicology (39) was as follows: “Nonunimodal data on biomarker responses may also be good data .... if dose-response relationships do not meet expectations, the use of multivariate statistics should be considered”, and indeed it has been used in aquatic monitoring studies (see ref 40). Principal Component Analysis (PCA) was used in order to link variables with statistical “components” that could signify, for example, relationships between the specific biomarkers body metal concentration and life cycle parameters (Table 3). Once these components emerge from the analysis they may be interpreted in relation to their physiological roles in earthworms. PCA identified three components in the multivariable data set. The first, which could be termed ‘population health component’, is characterized by the positive relationship between earthworm survival, cocoon production rate (i.e. reproductive output), wMT-1, and AOX; this group of variables was negatively correlated with earthworm weight change during metal exposure. The correlation of wMT-1 with cocoon production suggests that this isoform may be particularly active, possibly as a zinc chaperone and donator (see ref 41), in the reproductive organs of these hermaphrodite organisms. The second, a ‘metal exposure component’, is characterized by the positive relationship between accumulated tissue concentrations of both metals and the expression level of the biomarker wMT2, this being consistent with its role in the absorption, transport, and excretion of heavy metals (42). The third, a ‘detoxification pathways component’, is characterized by the positive relationship between molecular-genetic biomarkers, AOX and LYS. Amine oxidases are involved with detoxification (43) and the lysosomal system will be active during periods of cellular stress when there may be a requirement to recycle damaged organelles. Thus, from a complex set of measured variables, including life cycle parameters and the expression profile of four genes, determined under different levels of stress from heavy metals, PCA was able to identify biologically meaningful relationships 1762
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 6, 2005
between the variables. These biomarkers can serve as indicators of the physiological consequences of pollutant exposure on a population of earthworms. This approach enables the prediction of high level (population and perhaps community) responses from measurements of low level (cellular, physiological, and individual) stress-induced alterations. In future, assaying suites of biomarkers may enable scientists to identify which environmental contaminant (or combination of contaminants) is responsible for the detrimental effect and at what concentration. This information would enable environmental agencies and legislators to make rational decisions as to which chemical discharges to limit as well as informing decisions concerning the usage of contaminated soils. Biomarkers may also serve a role in postpollution monitoring and/or in soil bioremediation.
Acknowledgments This research was supported by the Natural Environment Research Council (Grant No. GST/02/1782) and the Royal Society of London.
Literature Cited (1) Alloway, B. J.; Ayres, D. C. Chemical Principles of Environmental Pollution; Chapman and Hall: London, U.K., 2nd ed.; 1997. (2) Kula H.; Larink, O. Tests on the earthworms E. fetida and A. caliginosa. In Handbook of Soil Invertebrate Toxicity Tests; Løkke, H., Van Gestel, C. A. M., Eds.; Wiley: Chichester, U.K., 1998; pp 95-112. (3) Spurgeon, D. J.; Weeks, J. M.; Van Gestel, C. A. M. A summary of eleven years progress in earthworm ecotoxicology. Pedobiologia 2003, 47, 588-606. (4) ISO. Soil Quality - Effects of Pollutants on Earthworms (Eisenia fetida, Eisenia fetida andrei). 1. Determination of Acute Toxicity Using Artificial Soil Substrate. Part 2. Determination of Effects on Reproduction Draft Guideline, ISO 11268-2.2, 1997. (5) Fairbrother, A.; Glazebrook, P. W.; Van Straalen, N.; Tarazona, J. Summary of the SETAC workshop on test methods for hazard determination of metals and sparingly soluble metal compounds in soils, June 19-23, San Lorenzo de El Escorial, Spain, SETAC, Pensacola, FL, 1999. (6) U.S.E.P.A. Ecological Soil Screening Level Guidance; Office of Emergency and Remedial Response, U.S. Environmental Protection Agency: Washington, DC, 2000. (7) Gibbs, M. H.; Wicker, L. F.; Stewart, A. J. A method for assessing sublethal effects of contaminants in soils to the earthworm. Eisenia foetida. Environ. Toxicol. Chem. 1996, 15, 360368. (8) Meier, J. R.; Chang, L. W.; Jacobs, S.; Torsella, J.; Meckes, M. C.; Smith, M. K. Use of plant and earthworm bioassays to evaluate remediation of soil from a site contaminated with polychlorinated biphenyls. Environ. Toxicol. Chem. 1996, 15, 928938. (9) Gunderson, C. A.; Kistuk, J. M.; Napolitano, G. E.; Wicker, L. F.; Richmond, J. E.; Stewart, A. J. Multispecies toxicity assessment of compost produced in bioremediation of an explosivescontaminated sediment. Environ. Toxicol. Chem. 1997, 16, 2529-2537. (10) Saterbak, A.; Toy, R. J.; Wong, D. C. L.; McBain, B. J.; Williams, M. P.; Dorn, P. B.; Brzuzy, L. P.; Chai, E. Y.; Salanitro, J. P. Ecotoxicological and analytical assessment of hydrocarboncontaminated soils and application to ecological risk assessment. Environ. Toxicol. Chem. 1999, 18, 1591-1607. (11) Hund-Rinke, K.; Wiechering, H. Earthworm avoidance test for soil assessments. An alternative for acute and reproduction tests. J. Soils Sediments 2001, 1, 15-20. (12) Schaefer, M. Assessing 2,4,6-trinitrotoluene (TNT)-contaminated soil using three different earthworm test methods. Ecotox. Environ. Safe. 2004, 57, 74-80. (13) Morgan, A. J.; Stu ¨ rzenbaum, S. R.; Kille, P. A short overview of molecular biomarker strategies with particular regard to recent developments in earthworms. Pedobiologia 1999, 43, 574584. (14) Scott-Fordsmand, J. J.; Weeks, J. M. Biomarkers in earthworms. Rev. Environ. Contam. Toxicol. 2000, 165, 117-159.
(15) Nadeau, D.; Corneau, S.; Plante, I.; Morrow, G.; Tanguay, R. M. Evaluation for Hsp70 as a biomarker of effect of pollutants on the earthworm Lumbricus terrestris. Cell Stress Chaperon. 2001, 6, 153-163. (16) Sauve´, S.; Hendawi, M.; Brousseau, P.; Fournier, M. Phagocytic response of terrestrial and aquatic invertebrates following in vitro exposure to trace elements. Ecotox. Environ. Safe. 2002, 52, 21-29. (17) Svendsen, C.; Spurgeon, D. J.; Hankard, P. K.; Weeks, J. M. A review of lysosomal membrane stability measured by neutral red retention: is it a workable earthworm biomarker? Ecotoxicol. Environ. Safe. 2004, 57, 20-29. (18) Stu ¨ rzenbaum, S. R.; Kille, P.; Morgan, A. J. The identification of new biomarkers as a response to heavy metal pollution. In Earthworm Toxicology; Sheppard, S., Bembridge, J., Homstrup, M., Posthuma, L., Eds.; SETAC: Amsterdam, The Netherlands, 1998; pp 215-224. (19) Kille, P.; Stu ¨ rzenbaum, S. R.; Galay, M.; Winters, C.; Morgan, A. J. Molecular diagnosis of pollution impact in earthworms towards integrated biomonitoring. Pedobiologia 1999, 43, 602607. (20) Galay Burgos, M.; Spurgeon, D. J.; Weeks, J. M.; Stu ¨ rzenbaum, S. R.; Morgan, A. J.; Kille, P. Developing a new method for soil pollution monitoring using molecular genetic biomarkers. Biomarkers 2003, 8, 229-239. (21) Ricketts, H. J.; Morgan, A. J.; Spurgeon, D. J.; Kille, P. Measurement of annetocin gene expression: a new reproductive biomarker in earthworm ecotoxicology. Ecotoxicol. Environ. Safe. 2004, 57, 4-10. (22) De Coen, W. M.; Janssen, C. R. A multivariate biomarker-based model for predicting population-level responses of Daphnia magna. Toxicol. Chem. 2003, 22, 2195-2003. (23) Van Straalen, N. M. Ecotoxicology becomes stress ecology. Environ. Sci. Technol. 2003, 37(17), 324A-330A. (24) OECD. Organization for Economic Cooperation and Development. OECD guidelines for testing of chemicals: earthworm acute toxicity test, 207; Paris, France, 1984. (25) Van Gestel, C. A. M.; van Dis, W. A.; van Breemen, E. M.; Sparenburg, P. M. Development of a standardised reproduction toxicity test with earthworm species Eisenia foetida andrei using copper, pentachlorophenol and 2,4-dichloroaniline. Ecotoxicol. Environ. Safe. 1989, 18, 305-312. (26) Stu ¨ rzenbaum, S. R.; Kille, P.; Morgan, A. J. The identification, cloning and characterisation of earthworm metallothionein. FEBS Lett. 1998, 431, 437-442. (27) Stu ¨ rzenbaum, S. R. Molecular genetic responses of earthworms to heavy metals, Ph.D. Thesis, University of Wales, Cardiff, 1997. (28) Ferre, F. Quantitative or semiquantitative PCR: reality versus myth. Methods Applications 1992, 2, 1-9. (29) Neuhauser, E. F.; Loehr, R. C.; Milligan, D. L.; Malecki, M. R. Toxicity of metals to the earthworm Eisenia fetida. Biol. Fert. Soils 1985, 1, 149-152.
(30) Stu ¨ rzenbaum, S. R.; Kille, P. Control genes in quantitative molecular biological techniques: the variability of invariance. Comp. Biochem. Phys. B 2001, 130, 281-289. (31) Thomas, T.; Schreiber, G. The expression of genes coding for positive acute phase proteins in the reproductive tract of female rat. High levels ceruloplasmin mRNA in the uterus. FEBS Lett. 1989, 243, 381-384. (32) Dallinger, R. Metallothionein research in terrestrial invertebrates: Synopsis and perspectives. Comp. Biochem. Phys. 1996, 113c, 125-133. (33) Bremner, I. In Metallothionein III; Suzuki, K. T., Imura, N., Kimura, M., Eds.; Birkhauser Verlag: Basel, Switzerland, 1993. (34) Stu ¨ rzenbaum, S. R.; Winters, C.; Galay, M.; Morgan, A. J.; Kille, P. Metal ion trafficking in earthworms. Identification of a cadmium-specific metallothionein. J. Biol. Chem. 2001, 276, 34013-34018. (35) Morgan, A. J.; Stu ¨ rzenbaum, S. R.; Winters, C.; Grime, G. W.; Kille, P. Differential metallothionein expression in earthworm (Lumbricus rubellus) tissues. Ecotoxicol. Environ. Safe. 2004, 57, 11-19. (36) Fossi, C.; Leonzio, C.; Focardi, S.; Peakall, D. B. Avian mixed function oxidase induction as a monitoring device: The influence of normal physiological functions. In Biomarkers of Environmental Contamination; McCarthy, J. F., Shugart, L. R., Eds.; Lewis Publishers: Boca Raton, FL, 1990; pp 143-149. (37) Calabrese, E. J.; Baldwin, L. A. Hormesis: the dose-response revolution. Annu. Rev. Pharmacol. Toxicol. 2003, 43, 175-197. (38) Renner, R. Redrawing the dose-response curve. Environ. Sci. Technol. 2004, 38(5), 90A-95A. (39) Van Gestel, C. A. M.; Weeks, J. M. Recommendations of the 3rd International Workshop on Earthworm Ecotoxicology, Aarhus, Denmark, August 2001. Ecotoxicol. Environ. Safe. 2004, 57, 100105. (40) Galloway, T. S.; Brown, R. J.; Browne, M. A.; Dissanayake, A.; Lowe, D.; Jones, M. B.; Depledge, M. H. A multibiomarker approach to environmental assessment. Environ. Sci. Technol. 2004, 38, 1723-1731. (41) Vallee, B. The function of metallothionein. Neurochem. Int. 1995, 27, 23-33. (42) Nordberg, M.; Nordberg, G. F.Toxicological aspects of metallothionein. Cell. Mol. Biol. 2000, 46, 451-463. (43) Padiglia, A.; Medda, R.; Lorrai, A.; Paci, M.; Pedersen, J. Z.; Boffi, A.; Bellelli, A.; Agro, A. F.; Floris, G. Irreversible inhibition of pig kidney copper-containing amine oxidase by sodium and lithium ions. Eur. J. Biochem. 2001, 268(17), 4686-4697.
Received for review June 3, 2004. Revised manuscript received November 8, 2004. Accepted November 12, 2004. ES049174X
VOL. 39, NO. 6, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1763
