Environ. Sci. Technol. 2009 43, 8895–8900
An Innovative Continuous Flow System for Monitoring Heavy Metal Pollution in Water Using Transgenic Xenopus laevis Tadpoles JEAN-BAPTISTE FINI,† ´ ,‡ SOPHIE PALLUD-MOTHRE ´ BASTIEN LE ME ´ VEL,† SE KARIMA PALMIER,† CHRISTOPHER M. HAVENS,| MATTHIEU LE BRUN,§ VINCENT MATAIX,§ GREGORY F. LEMKINE,‡ BARBARA A. DEMENEIX,† N A T H A L I E T U R Q U E , †,# A N D P A U L E . J O H N S O N * ,|,⊥ UMR CNRS 7221, Evolution des Re´gulations Endocriniennes; De´partement Re´gulations, De´veloppement et Diversite´ Mole´culaire, Muse´um National d’Histoire Naturelle 75231 Paris, France, WatchFrog SA, 4 rue Pierre Fontaine, 91000 Evry, France, Division Recherche et de´veloppement, Electricite´ de France, F-78401 Chatou Cedex, France, Department of Physics and Astronomy, University of Wyoming, Laramie, Wyoming, 82071, and SoftRay, Inc., Laramie, Wyoming
Received March 26, 2009. Revised manuscript received June 10, 2009. Accepted July 6, 2009.
While numerous detection methods exist for environmental heavy metal monitoring, easy-to-use technologies combining rapidity with in vivo measurements are lacking. Multiwell systems exploiting transgenic tadpoles are ideal but require timeconsuming placement of individuals in wells. We developed a realtime flow-through system, based on Fountain Flow cytometry, which measures in situ contaminant-induced fluorescence in transgenic amphibian larvae immersed in water samples. The system maintains the advantages of transgenic amphibians, but requires minimal human intervention. Portable and selfcontained, it allows on-site measurements. Optimization exploited a transgenic Xenopus laevis bearing a chimeric gene with metal responsive elements fused to eGFP. The transgene was selectively induced by 1 µM Zn2+. Using this tadpole we show the continuous flow method to be as rapid and sensitive as image analysis. Flow-through readings thus accelerate the overall process of data acquisition and render fluorescent monitoring of tadpoles suitable for on-site tracking of heavy metal pollution.
* Corresponding author phone: 307-766-6150; fax: 307-766-2652; e-mail:
[email protected]. † Muse´um National d’Histoire Naturelle. ‡ WatchFrog SA. | University of Wyoming. § Electricite´ de France. ⊥ SoftRay, Inc. # Present address:. Observatoire Oce´anologique, Laboratoire ARAGO; Universite´ Pierre et Marie Curie-Paris6; Centre National de la Recherche Scientifique, CNRS, FRE3247; Banyuls-sur-Mer, France. 10.1021/es9008954 CCC: $40.75
Published on Web 07/24/2009
2009 American Chemical Society
Introduction There is a pressing need to improve technologies for rapidly detecting physiological effects of environmental pollutants. This need is felt not only in the context of screening chemicals that might affect human health, but also to detect pollutants accumulating in the environment. In each case methods have to be developed that provide robust and reproducible readings obtained on model systems that reflect the full impact of a chemical on a given organism. Thus, the model should be an in vivo model that absorbs, metabolizes, and eventually excretes the product tested and its metabolites. Small model organisms such as Xenopus and zebrafish larvae used in multiwell screens are ideally suited to this end (1). However, one limitation of multiwell screening of tadpoles is the need to place and orientate each animal in a multiwell plate. This step is labor-intensive, time-consuming, and may require anesthetization to immobilize the tadpoles while they are being imaged. Moreover, in the context of following environmental pollution, multiwell readings are ill-adapted to on-site monitoring. It is with these multiple constraints in mind that we set out to develop a flow-through system for monitoring fluorescence emitted from transgenic tadpoles. This system is based on Fountain Flow cytometry (2-4), originally developed for the detection of bacteria and protozoa in aqueous samples. Hereafter, it will be referred to as FFC. In this method, fluorescently labeled target particles (in this case Xenopus tadpoles) flow into a chamber where they are illuminated by a blue light-emitting diode, which excites fluorescence of eGFP in the tadpoles (Figure 1). Fluorescence measurements are made with a digital camera continuously and in real time. The results are processed to yield the mean fluorescence of tadpoles in the sample, an indicator of the presence of contamination. For this research program, the FFC was optimized for on-site monitoring using a transgenic tadpole designed to detect micromolar concentrations of zinc in water. Transgenic, fluorescent animals can be used as sentinels for aquatic pollution; examples include zebrafish (5), mussels, and gudgeon. In addition, transgenic, fluorescent Xenopus larvae have been used for detection of endocrine disrupting chemicals. Here we used transgenic Xenopus larvae bearing various elements of the metallothionein gene placed upstream of a fluorescent reporter gene. Metallothioneins (MTs) constitute a superfamily of low molecular weight, cystein-rich, metal-binding proteins (6) which can be induced by heavy metal ions, cytokines, stress, and hormones (for a review, see refs 7 and 8). The potential importance of MTs in toxic responses to metals was recognized at the time of its initial discovery and has been shown to be a potential biomarker for metal pollution in aquatic environments (9, 10). MTs are responsible for protecting cells from the deleterious effects of exposure to elevated amounts of nonessential metals such as Cd2+ or Hg2+ ions and for regulating the intracellular supply of the biologically essential ions, Zn2+ and Cu2+. Zn2+ deprivation induces congenital malformations (11-13), whereas excess Zn2+ can be highly toxic to embryos (14, 15). In vertebrates all MT genes have a similar structure and MT transcription is regulated by short cis-acting elements named metal responsive elements (MRE), with consensus core TGGNCRCC, which are commonly present in multiple copies. The use of MRE sequences for zinc detection has been reviewed (16). VOL. 43, NO. 23, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
8895
FIGURE 1. Schematic diagram of the LED-illuminated Fountain Flow system for measuring Xenopus tadpoles. A sample of transgenic tadpoles flows through the flow cell toward the digital camera and its optics. The tadpoles are illuminated in the focal plane by an LED. When the tadpoles pass through the camera’s plane of focus they are imaged by the camera and lens assembly through the transparent flow cell window and a filter that isolates the wavelength of fluorescence emission. The tadpoles and the fluid in which they are suspended then pass by the window and out the flow cell drain tube. (From Johnson et al. (3), used with permission.) The research described here is composed of two parts. First, we created a transgenic tadpole bearing an MT heavy metal responsive zebrafish promoter linked to eGFP (MTZFeGFP). As it showed weak sensitivity to environmental Zn2+, we next used two copies of the consensus MRE derived from the human metallothionein promoter (17) fused to eGFP (2 × MRE-eGFP), which proved to be capable of detecting micromolar amounts of environmental Zn2+. Second, we used this model to optimize the Fountain Flow system. This continuous flow technology permits rapid quantification of fluorescent signals from aquatic small model organisms. Versatile, it can be adapted to different species and to different fluorescent signals. Moreover, animals can undergo multiple readings with no adverse effect.
Materials and Methods Plasmid Constructs. To obtain the MTZF-eGFP transgene, we proceeded in two steps. A SalI-ApaI eGFP cDNA and the SV40 polyA signal fragment were cloned into the pBluescript vector (Promega, Lyon, France). Next the sequence containing -850 to +1 bp of the zebrafish (Danio rerio) wild-type metallothionein promoter (18) was amplified by PCR from genomic DNA using the following primers: 5′-CAGAGCTCGAATTCCAGAGAGACACT-3′ and 3′-TGAAGCTTTTCCAGAGAGTATCCTCA-5′ and cloned into this vector. The transgene plasmid 2 × MRE-eGFP was constructed by inserting a dimer of a heavy metal responsive element (MRE) into a pBluescript vector (Promega) (17), containing both eGFP cDNA and SV40 polyA signal as described above, in which the SacI-BamHI E1b virus TATAbox sequence had been cloned. The MRE dimer was obtained by annealing of primers-containing two MRE consensus sequences TGCACAC (underlined). The primers used were: 2 × MRE SacIXbaI forward CAGACTCTGCACACGGGCCAATCGGAGC8896
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 23, 2009
CAGATCTCTGCACACGGGCCATT and 2 × MRE SacI-XbaI reverse CTAGAATGGCCCGTGTGCAGAGATCTGGCTCCGATTGGCCCGTGTGCAGAGTCTGAGCT. Restriction Enzyme-Mediated Integration Nuclear Transplantation Transgenesis. Germinally transgenic tadpoles were produced by restriction enzyme-mediated integration nuclear transplantation according to Kroll and Amaya (19) with the following modifications: sperm was purified by centrifugation on a two-layer discontinuous Percoll (Sigma, St. Quentin, Fallavier, France) gradient before the permeabilization step, which was performed with digitonin (Sigma) instead of lysolecithin (20). Transgenic tadpoles were created using SacI-linearized constructs and SacI restriction enzyme according to Kroll and Amaya (19). Animal Care. Sexually mature X. laevis frogs were supplied by Xenopus Express (Vernassal, Haute-Loire, France). Spawning of adult X. laevis was induced by injection of human chorionic gonadotropin (HCG, Chorulon). Males and females were injected with 300 and 700 IU of hCG, respectively, then placed in the same aquarium. Tadpoles were raised in dechlorinated and deiodinated tap water (1:2) and fed, with nettle powder (Valle´e, Chanzeaux, France), starting at five days post-fertilization. Tadpoles were staged according to the Normal Table of X. laevis (21). All rearing tanks were continuously aerated. The light-dark cycle was 12:12 h. F1 and F2 generations of germinally transgenic tadpoles were produced by crossing transgenic founders crossed with wild type animals. Heavy Metal Tadpole Treatment Protocol. Transgenic animals were sorted according to NF stage and fluorescence the day before treatment. Positive, transgenic animals at NF stage 47 were placed at 10 or 12 tadpoles per well in transparent, 6-well plates (TPP, Zurich, Switzerland) and housed in an incubator at 21.0° ( 0.5 °C. Chemicals were added directly to the well. All experimental aspects were conducted in compliance with the institutional guidelines for the care and use of animals (22). Water levels with controlled levels of contamination were produced by dissolving ZnCl2, and CuCl2 (Sigma) in water. Treated tadpoles were then used for classical imaging, fluorescence plate reading, and FFC. Classical Imaging. Images were captured using an Olympus fluorescent dissecting microscope equipped with an Olympus video camera DP50 (Olympus, Rungis, France) and a GFP filter unit (Olympus SVZ-MGFP). Before photography, germinally transgenic tadpoles at NF stage 47 were anaesthetized in 0.1% tricaine methanesulfonate (MS-222; Sigma). (Tadpoles remain in NF stage 47 for about 1 week, so they remain in the same stage throughout the time of measurements.) All pictures were taken with the same parameters (32× objective and 3-s exposure). Quantification was performed using ImageJ software (23). Data are expressed in relative units of fluorescence. Imaging System for Fluorescence Monitoring of a Flowing Tadpole Sample. The system is adapted from an epifluorescence FFC system previously developed for a sensitive and real-time detection of individual bacteria and protozoa in aqueous samples (2). This new approach consists of fluorescent tadpoles flowing up a tube, through a flow cell, toward a digital camera (Lumenera, Ottawa, Canada) and foreoptics. In the research described herein, tadpoles were pumped into the FFC and illuminated by an LED through the transparent flow cell window, and eGFP expression was measured. The FFC used a blue LED with an excitation filter for eGFP, and the digital camera used an emission filter for eGFP. The camera continuously measured the emission from the flow cell, at a rate of 2-60 frames/s. Each photometric measurement was made by summing the intensity of all of the pixels in the camera frame. Data was recorded as a time sequence of intensity measurements (Figure 2).
FIGURE 2. Typical results of FFC analysis. The digital output of an FFC data sampling sequence is shown. The output stream of the FFC is a continuous series of images without any time interval between images. Each data point in the graph represents the mean intensity of an image. Peaks occur as fluorescent tadpoles pass through the flow cell and are imaged. The horizontal line represents a 6 standard deviation detection threshold. Two images corresponding to peaks in the graph are shown for reference. The number of peaks above this threshold (37) are designated as “Xenope Counts.” The upper image corresponds to a relatively high peak (indicated by the red arrow) from a highly fluorescent tadpole (delineated by the red rectangle). The lower image corresponds to a more weakly fluorescing tadpole. Both images were contrast-enhanced.
TABLE 1. Classical Imaging Sensitivity of Two Lines of Transgenic Tadpoles for the Detection of Heavy Metal Contamination in Water
number of MRE in promoter sensitivity threshold to Cd+2 sensitivity threshold to Cu+2 sensitivity threshold to Zn+2 c
a Corresponds to p < 0.05. Corresponds to p < 0.001.
b
2× MRE-eGFP
Metallothionein GFP
2 5 µMa 2 µMa 1 µMa
2 5 µMc 2 µMb 50 µMb
Corresponds to p < 0.01.
Statistical Analysis. In vivo results are expressed as means ( SEM (standard error of the mean) per group and with p, the probability of insignificance. Differences between means were analyzed by Student’s t test or ANOVA and the Tukey-Kramer test where appropriate. Differences were considered significant at p < 0.05. The experiments were repeated at least twice (with n g 12 tadpoles/ experiment), providing the same results.
Results Low Concentrations of Zn2+ Do Not Affect Endogenous MT. As the overall aim of this series of experiments was to perform on-site detection of heavy metal contamination in environmental water samples, we set out to create a line of transgenic Xenopus tadpoles responsive to low concentrations of Zn2+ in water. Two constructs were tested and compared, each cloned upstream of eGFP: an 850-bp fragment of the metallothionein (MT) promoter from zebrafish (MTZF-eGFP) and a synthetic consensus sequence (2 × MRE-eGFP) containing a dimer of a MRE sequence from the human promoter (17). Table 1 shows the relative sensitivity of classical imaging detection of transgenic tadpoles bearing either one of these constructs to zinc, cadmium, and copper in the aquarium
water. Unexpectedly, in our in vivo system the 850-bp promoter fragment from zebrafish showed no measurable sensitivity to 5 µM Zn2+, although it exhibited a strong sensitivity to Cd2+. This raised the question of whether the animals were sensitive to the metal. This sensitivity could be dependent on both the rate of uptake of Zn2+ by the organism and/or the effectiveness of the MT promoter. Using qPCR we showed that endogeneous MT expression was induced by low levels of Cu2+ (2-fold with 8 µM) and Cd2+ (4-fold with 5 µM) (see Supporting Information (SI) Figure S-1). However, the endogenous MT gene was insensitive to 5 µM Zn2+ in the water, despite the fact that Zn2+ taken up dose-dependently by the animals (validated by the ICPMS technique as described in SI and shown in S-2). Thus, given the insensitivity of the MT gene to modest Zn2+ exposure, we chose to use the transgenic tadpole bearing a dimer of MRE from the human promoter, these consensus elements having been shown to be sensitive to low concentrations of Zn2+ (17). Two × MRE-eGFP Tadpoles Allow Sensitive Zn2+ Detection Using Conventional Methods. The 2 × MRE-eGFP transgenic tadpoles were tested for their sensitivity to Zn2+ and Cd2+. Exposure of NF 47 stage 2 × MRE-eGFP tadpoles to 5 µM ZnCl2 or 5 µM CdCl2 for 96 h resulted in 1.3- and 1.5-fold inductions in fluorescence respectively (Figure 3). Note that this transgene has the added advantage of being less sensitive to Cd2+ than MTZF-eGFP, making it more selective for Zn2+ detection (Table 1). These measurements were made by anaesthetizing tadpoles and measuring the fluorescent signal emitted from the tadpole head. Similar results were obtained using a plate reading system in which tadpoles were placed individually in the wells by the operator (data not shown). Given the time-consuming and laborious nature of these two approaches, we used the 2 × MRE-eGFP tadpole to optimize the continuous flow system. Fountain Flow is as Sensitive As Classical Imaging for Zn2+ Pollution Detection. Direct comparison of FFC and classical digital imaging of transgenic tadpoles (Figure 4) shows that FFC performs at least as well as classical imaging VOL. 43, NO. 23, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
8897
FIGURE 3. eGFP expression in 2 × MRE-eGFP tadpoles is induced by 5 µM Zn2+. (a) Schematic representation of the 2 × MRE-eGFP construct used for germinal transgenesis (b) Representative photograph of an untreated NF stage 47 control transgenic tadpole expressing basal fluorescence (under exciting light). The yellow emission is autofluorescence from internal organs. (c) and (d) Representative photographs of heads of NF stage 47 transgenic tadpoles treated for 96 h with 5 µM Zn2+ and 5 µM Cd2+, respectively. (e) Fluorescence measurements of transgenic tadpoles treated as in b, c, and d (n ) 15). This experiment was repeated three times. * corresponds to p < 0.05 and *** to p < 0.001.
FIGURE 4. FFC is as sensitive and reproducible as classical imaging. A comparison of fluorescence intensity measurements of transgenic Xenopus tadpoles between FFC and classic multiwell-plate imaging is shown. The intensity of both FFC and classic imaging were normalized to 1.0 for the control samples. Data are expressed as means (SEM. The two FFC measurements of samples containing Zn2+ yielded detections with an intra-assay variability of 25% (SEM/mean) whereas classical imaging gives a 50% intra-assay variation at 1 µM and a 33% variation at 5 µM. * corresponds to p < 0.05 and ** to p < 0.01. in multiwell plates. For the data shown here, the FFC obtained a 25% intra-assay variability (SEM/mean) for 1 µM Zn2+ while the classical imaging technique produced 50% intra-assay 8898
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 23, 2009
variability. Both techniques showed an increase in fluorescence response with dose from 1 to 5 µM. We determined the method detection limit (MDL) for the FFC detection of Zn2+ in water from the algorithm MDL ) s × t, as used by the U.S. Environmental Protection Agency (USEPA) (24), where s is the standard deviation for the set of measurements at the lowest concentration for which we have a 99% minimum probability of detection, and t is the t-value appropriate for this probability and the number of samples measured (Figure 4). To illustrate the potential for the measurement of other water contaminants, we quantitatively assessed the viability of using Xenopus tadpoles for detection of thyroid hormone. We used a thyroid responsive element-containing promoter fused to eGFP previously developed for assessing thyroid hormone disruptive pollutants in water (25, 26). The FFC system sensitivity and dose response was measured for thyroid hormone sensitivity using this transgenic line and the active form of thyroid hormone, tri-iodo thyronine (T3). Thyroid hormone transgene induction was detected with T3 concentrations as low as 10-9M (Figure 5), comparable to those obtained previously with classical imaging (25) and multiwell plates (26).
Discussion Transgenic tadpoles have been shown to be an effective means for detecting endocrine disrupting chemicals, deriving in vivo information from in vitro tools (1). When adapting this method to a specific type of pollution two main factors have to be considered, the choice of promoter driving the fluorescent signal and the method used for detecting the signal. For this program, first we adapted the X. laevis model
bioaccumulation in the tadpoles, and their physiological response. Tadpoles serve as a direct measure of the toxicity of heavy metal contaminants in waters to aquatic life, and a model for their potential to adversely affect humans. Our method detection limit of 3.6 µM for Zn2+ meets that goal, far exceeding the USEPA recommendation of
