Anal. Chem. 2001, 73, 5168-5171
Detection of Organomercurials with Sensor Bacteria Angela Ivask,†,‡ Kaisa Hakkila,‡ and Marko Virta*,‡
National Institute of Chemical Physics and Biophysics, Tallinn, Estonia, and Department of Biotechnology, University of Turku, Turku, Finland
The detection of mercury and especially of its organic species in the environment is of great interest, because of their very high toxicity1 and biomagnification in the food chain. Methylmercury (MeHg) is currently assumed to be the main biomagnifying species of mercury.2,3 It is present in seawater at very low concentration (nanograms per liter)4 and accumulated by plankton, which is consumed by fish. Fish are further consumed by mammals, which accumulate MeHg over time because most species are unable to efficiently eliminate it from the body. Currently, gas chromatography and liquid chromatography coupled to spectroscopic detection are used for the analysis of organomercurial species.5,6 The detection limits of those methods,
for example, for MeHg, are in the range of nanograms per liter.7 Labor-intensive sample preparation, relatively expensive instrumentation, and the requirement for trained personnel are drawbacks of above-mentioned analytical methods. Several bacteria are resistant toward mercury by a mechanism that is genetically encoded. The genes merTP(C)A(B)D responsible for the resistance are aggregated into mer operon.8 Genes in parentheses are not present in all mer operons. Two different types of resistance mechanism have been described; one of them yields resistance toward inorganic mercury (i.e., narrow-spectrum mercury resistance) and another toward inorganic as well as organic compounds of mercury (i.e., broad-spectrum mercury resistance).9 The resistance toward organic compounds of mercury in a broad-spectrum mercury resistance system is achieved due to organomercurial lyase (product of merB gene), the enzyme that catalyses the breakdown of the mercury-carbon bond, which is very stable.10 The released Hg2+ ions complex with the regulatory protein of the mer operon, MerR (product of merR gene), which in the absence of mercury ions represses the transcription of mer operon. After binding Hg2+ ions, the conformation of the mercuryMerR complex changes and the mer genes necessary for the mercury detoxification will be transcribed. Hg2+ ions are detoxified by mercuric reductase (product of merA gene), and metallic mercury (Hg0) volatilizes from the cells. The narrow-spectrum mercury resistance system lacks the merB gene, and thus, only ionic mercury can induce the transcription of the mer genes.9 The rest of the mer operon genes are involved in the transport of Hg2+ ions inside the cells (merTPC) or play a minor role in regulation (merD). The regulatory part of the mer operon (merR and operator/promoter part of mer operon) can be used for the construction of sensor bacteria. The whole bacterial cells or sensor bacteria have been used by various research groups for analyzing different compounds, for example, inorganic mercury,11,12 naphthalene,13 and arsenite.14 In a bacterial sensor, the expression of a reporter gene is
* Corresponding author: (tel) +358 2 333 8059; (fax) +358 2 333 8050; (email)
[email protected]. † National Institute of Chemical Physics and Biophysics. ‡ University of Turku. (1) Harada, M. Crit. Rev. Toxicol. 1995, 25, 1-24. (2) Morel, F. M. M.; Kraepiel, A. M. L.; Amyot, M. Annu. Rev. Ecol. Syst. 1998, 29, 543-66. (3) Baldi, F. Met. Ions Biol. Syst. 1997, 34, 213-57. (4) Kannan, K.; Smith, R. G.; Lee, R. F.; Windom, H. L.; Heitmuller, P. T.; Macauley, J. M.; Summers, J. K. Arch. Environ. Contam. Toxicol. 1998, 34, 109-18. (5) Clement, R. E.; Yang, P. W.; Koester, C. J. Anal. Chem. 1997, 69, 251R87R.
(6) Harrington, C. F. LC GC Eur. 2000, 420-7. (7) Bowles, K. C.; Apte, S. C. Anal. Chem. 1998, 70, 395-9. (8) Osborn, A. M.; Bruce, K. D.; Strike, P.; Ritchie, D. A. FEMS Microbiol. Rev. 1997, 19, 239-62. (9) Misra, T. K. Plasmid 1992, 27, 4-16. (10) Begley, T. P.; Walts, A. E.; Walsh, C. T. Biochemistry 1986, 25, 7192-200. (11) Selifonova, O.; Burlage, R.; Barkay, T. Appl. Environ. Microbiol. 1993, 59, 3083-90. (12) Virta, M.; Lampinen, J.; Karp, M. Anal. Chem. 1995, 67, 667-9. (13) King, J. M. H.; DiGrazia, P. M.; Applegate, B.; Burlage, R.; Sanseverino, J.; Dunbar, P.; Larimer, F.; Sayler, G. S. Science 1990, 249, 778-81. (14) Ramanathan, S.; Shi, W.; Rosen, E. P.; Daunert, S. Anal. Chem. 1997, 69, 3380-4.
Mercury and its organic compounds, especially methylmercury, are hazardous compounds that concentrate in biota via biomagnification and cause severe neurological disorders in animals. In this paper, a recombinant wholecell bacterial sensor for the detection of the organic compounds of mercury was constructed. The sensor carries firefly luciferase gene as a reporter under the control of the mercury-inducible regulatory part of broad spectrum mer operon from pDU1358. In addition, a gene-encoding organomercurial lyase (an enzyme necessary for cleavage of the mercury-carbon bond) was coexpressed in the sensor strain. The sensitivity of the sensor was evaluated on some environmentally important organomercurial compounds. The lowest detectable concentrations were 0.2 nM (50 ng/L), 1 nM (0.34 µg/L), and 10 µM (2.3 mg/L) for methylmercury chloride, phenylmercury acetate, and dimethylmercury, respectively. The sensor responded also to inorganic mercury and, therefore, using the sensor described here together with sensor bacteria responding only to inorganic mercury, it should be possible to characterize the mercury contamination, for example, in environmental samples.
5168 Analytical Chemistry, Vol. 73, No. 21, November 1, 2001
10.1021/ac010550v CCC: $20.00
© 2001 American Chemical Society Published on Web 09/26/2001
controlled by a genetic regulatory unit which responds to the given analyte (receptor-reporter concept).15 Most of the regulatory units used in the construction of sensor bacteria originate from bacteria that have a genetically encoded resistance system toward a toxic compound, like mercury. In the most simple case, the regulatory unit consists of a regulatory protein (and its corresponding gene), like MerR, that recognizes the analyte and controls the expression of a reporter gene. The sensitivity and specificity of the sensor bacteria toward given analyte are therefore mainly defined by the regulatory unit. The bacterial sensors are inexpensive to produce; they measure biological response (bioavailability) and require little (if any) pretreatment of sample. So far, no sensor bacteria for organomercurial have been described. In this study, we constructed a new whole-cell bacterial sensor for the detection of organic compounds of mercury by using the receptor-reporter concept. The receptor part of the sensor originated from a broad-spectrum mer operon from the plasmid pDU1358 (Serratia marcescens).16 merR and the operator/promoter part of mer operon were used to control the expression of firefly luciferase (lucFF) reporter gene. The merB gene was used to produce organomercurial lyase for the breakdown of those organomercurials that are not recognized by the broad-spectrum MerR protein. Therefore, the bacteria should be able to detect a wide range of the organic compounds of mercury. EXPERIMENTAL SECTION Materials. DNA-modifying enzymes were purchased either from New England Biolabs or Promega, and Vent DNA polymerase was purchased from New England Biolabs. For the DNA purification Qiagen (Hilden, Germany), PCR purification and gel purification kits were used. High-copy number vector pSL1190 was purchased from Amersham Pharmacia Biotech. Tryptone and yeast extract were purchased from Difco and casamino acids from Sigma. Mercury(II) chloride, methylmercury chloride (g98.0%), and dimethyl sulfoxide (DMSO) were of analytical grade and purchased from Riedel-de-Hae¨n. Phenylmercury acetate (PhHgOAc) (purum; g98.0%) was purchased from Fluka and dimethylmercury (DMeHg) (95%) from Aldrich. D-Luciferin was from BioNobile (Turku, Finland). All other reagents were of analytical grade. Water was purified with MilliQ equipment (Millipore, Bedford, MA). The work with organomercurial compounds was done in a ventilation hood, using the required safety equipment. The waste containing mercury was treated according to institutional guidelines. Bacterial Strains and Plasmids. Standard recombinant techniques17 were used for the construction of the sensor plasmid for the detection of organomercurials, pmerBRBSluc. First, lucFF gene (excluding the first 12 bases) was inserted into a 3.4-kbp BamHI-SmaI fragment of pSL1190 as a 1.8-kbp fragment (BamHIPvuII fragment from pTOO1112). The result plasmid pSLluc was used for the construction of the final sensor plasmid (pmerBRBS(15) Lewis, J. C.; Feltus, A.; Ensor, C. M.; Ramanathan, S.; Daunert, S. Anal. Chem. 1998, 70, 579A-85A. (16) Griffin, H. G.; Foster, T. J.; Silver, S.; Misra, T. K. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 3112-6. (17) Sambrook, J.; Fritsch, E. F.; Maniatis, T. Molecular cloning. A laboratory manual; Cold Spring Harbor Laboratory Press: New York, 1989.
luc). pGN120,18 a plasmid containing the broad-spectrum mer operon from pDU1358, was used as a template in PCR amplification of merB and the fragment containing the merR gene together with the operator/promoter part of the mer operon with primer pairs 5′-ATATCACGTGATGAAGCTCGCCCCATATAT-3′, 5′-AATTCTCGAGGCTGCGAATCCGATGCCGGT-3′ and 5′-ATATCTCGAGTCCTCAGCATAGTACCGGGA-3′, 5′-TTAAGGATCCCCTCATACGCTTGTCCTTTCAAA-3′, respectively. The sizes of the products were 607 and 706 bp for merB and merR, respectively. The PCR product of merR was inserted into pSLluc by XhoI and BamHI digestions and the structure of the resulting 5.6-kbp plasmid pmerRBSluc was confirmed by specific digestions. Then, the PCR product of merB was inserted into pmerRBSluc by PmlI and XhoI digestions. The structure of the resulting 6.2-kbp final sensor plasmid pmerBRBSluc was confirmed by sequencing the essential part of the plasmid. All three constructed plasmids were transformed to Escherichia coli MC106119 using electroporation.20 Two new strains, MC1061(pmerRBSluc) and MC1061(pmerBRBSluc), were used thereafter for measurements. We previously constructed a recombinant bacterial sensor for the detection of inorganic mercury,12 in which the expression of the lucFF gene is controlled by the merR and operator/promoter part of the narrow-spectrum mer operon from Tn21.21 This strain, MC1061(pTOO11), was used for the comparative measurements. Cultivation and Freeze-Drying of Sensor Bacteria. Bacteria were cultivated on a shaker (300 rpm) at 37 °C in LB medium17 supplemented with 100 µg/mL ampicillin in cloning phase and in M9 minimal medium17 supplemented with 0.5% casamino acids and 100 µg/mL ampicillin for freeze-drying. Before freeze-drying, bacteria were grown until an OD600 of 0.7 was reached, harvested by centrifugation, and resuspended into an equal volume of the M9 medium supplemented with 0.5% casamino acids and 10% lactose. Freeze-drying was performed according to a standard procedure22 in 0.5-mL aliquots using a Lyofast S 04 freeze-dryer (Edwards Kniese & Co Hochvacuum GmbH, Marburg, Germany). Measurements. Before measurements, the freeze-dried bacteria were rehydrated by adding 3 mL of the M9 medium with 0.5% casamino acids and the resultant mixture was incubated for 2 h at room temperature before use. Stock solutions (0.1 M) of organomercurials and HgCl2 were made in DMSO. Further dilutions were made in water (the final concentration of DMSO in the test was 1%), and 50 µL of the respective dilution was pipetted onto white 96-well Cliniplates (Thermo Labsystems, Helsinki, Finland). The same volume (50 µL) of cells was added, and plates were incubated without shaking at 37 °C for 2 h. After that, 100 µL of luciferase substrate (0.5 mM D-luciferin in 0.1 M sodium citrate buffer, pH 5.0) was added. Luminescence was measured using a microplate luminometer Wallac Victor (Perkin-Elmer Life Sciences, Turku, Finland) after 30-min incubation with D-luciferin at room temperature. (18) Nucifora, G.; Chu, L.; Silver, S.; Misra, T. K. J. Bacteriol. 1989, 171, 42417. (19) Casadaban, M. J.; Cohen, S. N. J. Mol. Biol. 1980, 138, 179-207. (20) Dower, W. J.; Miller, J. F.; Ragsdale, C. W. Nucleic Acids Res. 1988, 16, 6126-44. (21) Misra, T. K.; Brown, N. L.; Fritzinger, D.; Pridmore, R.; Barnes, W.; Silver, S. Proc. Natl. Acad. Sci. U.S.A. 1984, 84, 5975-9. (22) Sidyakina, T. M.; Golimbet, V. E. Cryobiology 1991, 28, 251-4.
Analytical Chemistry, Vol. 73, No. 21, November 1, 2001
5169
Figure 1. Structure of the plasmid pmerBRBSluc. Abbreviations used: amp, gene-encoding ampicillin resistance; Plac, lac promoter; merB, organomercurial lyase gene; merR, gene-encoding repressor/ activator of the mer promoter; PmerR, merR promoter; Pmer, promoter of mer operon; lucFF, gene-encoding firefly luciferase.
Calculation of Induction Coefficients. Induction coefficients were calculated using the formula IC ) Li/Lb, where IC is the induction coefficient, Li is the luminescence value of the sample, and Lb is the luminescence value of a blank solution (not containing mercury compounds). RESULTS Construction of the Plasmid pmerBRBSluc. The cloned regulatory unit in pmerBRBSluc contains merR gene and two promoters, merR promoter (PmerR) and promoter of mer operon (Pmer). In the natural resistance system, transcription of the merR is started from the PmerR and transcription of the other mer genes from the Pmer. In the plasmid pmerBRBSluc, Pmer promoter initiates transcription of the lucFF (Figure 1). In addition, merB gene is transcribed from lac promoter. Response of Sensor Bacteria to Different Compounds of Mercury. The sensor bacteria were functional after freeze-drying and no remarkable loss of activity was observed. The use of freezedried cells eliminated the need for routine cultivation of bacteria and resulted in good reproducibility (CV of different assays was 5.4%). That value was obtained in the measurements that were
conducted during six-month period, which shows that the bacteria are stable in the lyophilized form, at least when stored at -20 °C. In the absence of mercurial compounds, the expression of luciferase in MC1061(pmerBRBSluc) cells was strongly repressed whereas inorganic or organic mercuric compounds induced the bioluminescence even in relatively low concentrations. The lowest concentration that caused a noticeable response (induction coefficient, 2) was 10 nM (2.7 µg/L) for HgCl2, 0.2 nM (50 ng/L) for MeHgCl, and 1 nM (0.34 µg/L) for PhHgOAc (Figure 2A). For MC1061(pTOO11). the concentrations were 10 nM for HgCl2, 0.1 µM for MeHgCl, and PhHgOAc (Figure 2B). The luminescence increased with increasing concentration of mercury compound but fell rapidly to zero when the concentration of the compounds reached toxic level exceeding 0.5-1 µM. Response to DMeHg was measured with MC1061(pmerBRBSluc) and MC1061(pTOO11), but the increase of luminescence was seen only with MC1061(pmerBRBSluc) (Figure 3). However, the concentration needed for the induction with DMeHg was 103 times higher than that for HgCl2 and about 105 and 104 times higher than those for MeHgCl and PhHgOAc, respectively. The induction increased with increasing concentration of DMeHg until 1 mM after which a slight decrease of the light production was detected. The maximum induction coefficient was 20-30 with all tested compounds for MC1061(pmerBRBSluc) and 200 with HgCl2 and 20 with MeHgCl and PhHgOAc for MC1061(pTOO11). DISCUSSION In this work, a new recombinant whole-cell bacterial sensor was constructed by fusing genes of firefly luciferase and a regulatory region of the broad-spectrum mer operon from pDU1358.16 The constructed system is based on the natural bacterial resistance mechanism toward mercury and organomercurial compounds. In the presence of mercury, the resistance mechanism is activated by a regulatory protein and the Hg2+ ions are reduced to Hg0, which evaporates from the cells.9 In nature, the resistance to organic compounds of mercury is achieved due to organomercurial lyase, an enzyme that is produced by broadspectrum mer operon and that catalyses the breakdown of the mercury-carbon bond. According to previous studies,10 organo-
Figure 2. Response (induction coefficient) of sensor strains toward different compounds of mercury: (A) response of sensor strain for the detection of inorganic mercury and organomercurials [MC1061 (pmerBRBSluc)]; (B) response of sensor strain for the detection of inorganic mercury [MC1061 (pTOO11)] when incubated with HgCl2 ([), MeHgCl (0), and PhHgOAc (O). Data represent mean ( standard deviation of three independent measurements. 5170 Analytical Chemistry, Vol. 73, No. 21, November 1, 2001
Figure 3. Response (induction coefficient) of the sensor strains toward dimethylmercury. Symbols used: (9) MC1061(pmerBRBSluc) and (b) MC1061(pTOO11). Data represent mean ( standard deviation of two independent measurements.
mercurial lyase has a wide substrate range, from alkyl to aromatic organomercurial compounds. In addition, the regulatory protein of the broad-spectrum mer operon has been shown to react directly to some organomercurials.18 Therefore, to obtain a sensor responding to as many organomercurials as possible, we used the gene-encoding organomercurial lyase and the gene-encoding regulatory protein together with operator/promoter region of the broad-spectrum mer operon in the construction of the sensor plasmid. Accordingly, the resulting construct MC1061(pmerBRBSluc) gave a response to all four compounds of mercury tested in this study (Figures 2 and 3). In the breakdown of, for example, MeHgCl, one molecule of MeHgCl will produce one Hg2+ ion. Therefore, the concentration of Hg2+ should be identical with MeHgCl, if all the MeHgCl would have been broken down. However, the sensor bacteria responded to MeHgCl at 50 times and PhHgOAc at 10 times lower concentrations than to HgCl2. One explanation for that could be the higher lipophility of organomercurial compounds23 and their better ability to enter the cell. Previously described sensor strain MC1061(pTOO11), which has regulatory part from the narrowspectrum mer operon, responded to organomercurial compounds in 100-1000 times higher concentrations than MC1061(pmerBRBSluc). Moreover, the concentrations of organomercurials needed for the induction of MC1061(pTOO11) and the induction coefficients were different from the ICs for HgCl2. As the MC1061(pTOO11) does not contain the organomercurial lyase gene, the response toward MeHgCl and PhHgOAc by this system could be partly explained by the presence of Hg2+ as an impurity in the chemicals used. For HgCl2, the patterns of response obtained with MC1061(pmerBRBSluc) and MC1061(pTOO11) were very similar. DMeHg was previously shown not to be cleaved by the organomercurial lyase.10 In contrast, in this study, the MC1061(pmerBRBSluc) cells responded to DMeHg, although in 103, 104, and 105 times higher concentrations than to HgCl2, PhHgOAc, and MeHgCl, respectively (Figures 2 and 3). No response to DMeHg was obtained with MC1061(pTOO11) cells (Figure 3). Surprisingly, the toxic effect of DMeHg was not seen even in 100 (23) Halbach, S. Arch. Toxicol. 1985, 57, 139-41. (24) Toribara, T. Y.; Clarkson, T. W.; Nierenberg, D. W. Chem. Eng. News 1997, 75 (24), 6. (25) Gilmour, C. C.; Henry, E. A. Environ. Pollut. 1991, 71, 131-69.
mM concentration. One explanation to the extremely low apparent toxicity of DMeHg (104 times less toxic than HgCl2; see Figures 2 and 3) could be evaporation of DMeHg from the reaction mixture as this compound has found to be very volatile.24 Since the regulatory protein of broad-spectrum mer operon has been shown to react directly with some organomercurials, it could be possible that the response obtained was caused by DMeHg directly activating this protein. However, the response to DMeHg was measured also with a construct containing only the regulatory protein from broad-spectrum mer operon but not the organomercurial lyase [MC1061(pmerRBSluc)] and no induction of luminescence was obtained (data not shown). Therefore, the response to DMeHg has to result from the activity of the organomercurial lyase. The whole-cell sensor described here responds to inorganic and various organic compounds of mercury. Thus, if these sensor bacteria would be used together with sensor bacteria responding only to inorganic mercury, it should be possible to differentiate between inorganic and organic mercury contamination. The detection limit for organic compounds of mercury, for example, MeHg, obtained with the described whole-cell sensor was somewhat higher than that of the best chemical methods.7 Bowles and Apte7 obtained the detection limit of 0.024 ng/L with a 50-mL sample, whereas our detection limit was 48 ng/L with a 50-µL sample. Those values correspond 0.0012 and 0.0024 ng of Hg, respectively. Nevertheless, the detection limit of the sensor bacteria for MeHg should be enough to detect this compound in contaminated waters and sediments.25 Moreover, it should be possible to enhance the sensitivity of the sensor bacteria by optimizing the measurement protocol. Furthermore, the sensor bacteria are inexpensive and easy to produce and can also be used in field studies using portable instruments. One significant advantage of the sensor bacteria is a very high throughput of samples (hundreds per day) that is very difficult to achieve with existing analytical methods mostly due to the laborious sample preparation procedure. Additionally, the chemical methods measure the total amount of the chemical in contrast to whole-cell sensors that measure the biologically available fraction. The analysis of the bioavailable fraction is an important issue in environmental and human hazard assessment studies. Our further studies will involve the use of the sensors described in this paper for the analysis of the bioavailable fraction of different species of mercury in the environmental samples (e.g., polluted soils and sediments). ACKNOWLEDGMENT Plasmid pGN120 was a gift from Simon Silver and Amit Gupta (University of Chicago, IL). The study was financially supported by Academy of Finland (grant 45754). Additional funding was provided by CIMO and the Estonian Science Foundation (grant 3845). We thank Dr. Matti Karp (University of Turku, Finland) for his support and Drs. Anne Kahru (Institute of Chemical Physics and Biophysics, Tallinn, Estonia) and Stephen Forsythe (The Nottingham Trent University, Nottingham, U.K.) for improving the manuscript. Received for review May 15, 2001. Accepted July 16, 2001. AC010550V Analytical Chemistry, Vol. 73, No. 21, November 1, 2001
5171
