- Email: [email protected]
Transgenic Caenorhabditis elegans strains as biosensors
125
Transgenic Caenorhabditis elegans strains as biosensors E. Peter M. Candid0 and Don Jones Toxicity bioassays rely largely on lethality measurements. Such assays are generally lengthy and expensive, and provide little information on mechanisms of toxicity. A desire to understand the mechanisms by which cells respond to physical and chemical stresses has led to interest in measuring stress proteins as toxicological endpoints. Transgenic strains of the nematode Caenorhabditis elegans that carry a reporter enzyme under control of a stress-inducible promoter have been created. The reporter is easily quantified in intact nematodes, and it responds to a wide range of chemical stressors. Therefore, transgenic C. elegans can provide the basis for a wide range of quick, simple and informative bioassays.
The cellular stress response involves the synthesis of specific sets of proteins by organisms as a result of exposure to physical (e.g. heat) or chemical stressesI. These heat shock or stress proteins confer a protective function on cells, through their ability to minimize, or even reverse, protein denaturation’. It has recently been suggested that the measurement of stress-protein levels is a potentially useful tool in environmental toxicology’. This article discusses progress in the use of transgenic strains of the nematode Cue~orlz&itis elegant as biosensors of the stress response”. Free-living C. eleg~ns is a widely used model organism for genetic, developmental and molecular biologcal studies”. This popularity arises from several desirable characteristics: it is easy to maintain in the laboratory; it is transparent; it has an invariant developmental cell-lineage and a small number of somatic cells; and it is amenable to genetic manipulation and has a small genome. Despite a complement of only 959 cells in its hermaphroditic form, the organism is a fully differentiated animal with a nervous system, specialized muscles, and digestive and reproductive systems. The complete embryonic cell lineageb, as well as the post-embryonic cell lineage of the hermaphrodite7 and the ‘wiring’ diagram of the nervous systems have been determined. A detailed genetic map and an almost complete physical map of the genome are available, and nucleotide sequencing of the entire genome is progressing rapidly’. Methods have been developed for transforming C. eleguns with cloned DNA sequences’“. E. P. M. Can&do and D. Jones aw at the Department of Biochemistry and Molecular Biology, University of British Columbia, 2146 Health Sciences &la/l, Vancouver, Cartada V6T 123. Copyright 0 1996, Elsewer Science Ltd. All rights reserved. 0167 - 7799/96/$15.00
When it is exposed to elevated temperatures, C. elegans undergoes a typical stress response, inducing the synthesis of a number of stress or heat shock proteins (hsps; Refs 11,12). Our laboratory has studied a family of small hsps with molecular masses of 16 kDa (hsp16s). An important feature of the four known hsp 16 genes is their strict inducibility, i.e. under normal culture conditions hsplb proteins are undetectable, but they are synthesized rapidly following heat shock’“,‘J. In order to examine the tissue specificity of the stress response, a number of transgenic strains carrying various fusions of hspl6 genes with the 1acZ gene of Escherichia coli were created15.
Transgenic stress-reporter strains and assay conditions Figure la illustrates the hsp 26-1ucZ fusion gene carried in strain PC72 of C. elegans; this gene consists of IacZ fused in-frame to the second exon of hsy,16- 1. The construct also includes a nuclear localization signal (NLS) from the SV40 virus, which targets the fusion protein to the nucleus, greatly facilitating the identification of expressing cellsi6. PC72 carries 70-80 integrated copies of the hspl64acZ fusion in mixed arrays with the marker plasmid pRF4, which confers a distinct rolling phenotype to the animal”. Figure lb illustrates the principal steps in assaying soluble agents for stress induction. The predominant form of C. elegans is a self-fertilizing hermaphrodite; cultures can be maintained by transferring one or more hermaphrodites to a fresh agar plate approximately once a week. Larger-scale cultures are grown in liquid, and a half-liter culture yields enough specimens to perform hundreds of tests. C. elegam is sufficiently small that suspensions can be aliquotted using standard micropipets. Routine stress tests made use of synchronized PII: SO167.7799(96)10016-O
TIBTECH APRIL 1996 (VOL 14)
126
reviews a
Heat shock elements (HSEs) ,/I \
SV40 ’ NLS /
HRhsp 16-48
hsp16-I-IacZfusion
b Exposure to stress @-7f-d
Heavy-metal
1 Permeabilize
1
1
1
W-W
Figure 1 (a) Structure of reporter gene in transgenic Caenorhabditis elegans strain PC72 (Ref. 15). Open boxes represent exons, lines represent noncoding sequences. The gray and black bars represent the SV40 nuclear localization signal (NLS) and the IacZ coding region from Escherichia coli, respectively, fused in-frame with the second exon of hspl6-1. The two hspl6 genes are transcribed in opposite directions under the control of heat shock elements in the intergenic region: hsp16-48 is oriented leftward, and hspldl rightward. Flanking sequences of the fusion gene, including the 3’ untranslated region, are derived from the hspl6-I gene. (b) Flow diagram indicating the steps involved in evaluating chemical stressors using transgenic nematode strains such as PC72. populations of nematodes at a particular stage of development, usually L3 larvae or adults. Synchronized cultures are readily obtained by inoculating media with embryos prepared by dissolving gravid adults in alkaline bleach’*. Mixed cultures may be preferable for some applications, as different life stages may respond to chemical stressors in varying ways. Stress assays are performed in small (l-3ml) volumes of liquid in multiwell dishes. Agar plates are not used because charged groups in the agar, as well as other components in the growth medium, may bind many chemical stressors, especially metal ions. After being exposed to the stressor, nematodes are permeabilized by a brief acetone treatment, and the whole organisms are assayed directly. P-Galactosidase activity is measured either calorimetrically using the soluble substrate ONPG (o-nitrophenyl P-D-galactopyranoside), or histochemically using the substrate x-gal (5-bromo-4-chloro-3-indolyl P-D-galactoside); TIBTECH APRIL 1996 (VOL 14)
both assays have similar sensitivity (Fig. 2). The colorimetric assay is rapid and yields a quantitative value of enzyme activity (Fig. 2a). Although the histochemical assay is more time-consuming (requiring counting of positive worms under a dissection microscope), it produces a different type of data, indicating tissue specificity, as well as giving an overall measure of stress induction (Fig. 2b). If necessary, levels of hsplh can also be measured by immunological methods.
stressors
Mercury, lead, copper(H), zinc and cadmium ions can induce a transgene stress-response” and, in the absence of bacteria as a source of food (bacterial cell walls may also bind metal ions), different patterns of tissue stress are seen with some of these metal ions (Fig. 3). In all cases, the levels of stress-reporter enzyme are a more sensitive indication of exposure than lethality assays. Guven et al.“’ have also reported stress-inducible 1acZ induction by several heavy metals using nematode strain CB4027, which has a Drosophila hsp70 promoterz2. The PC72 stress response to cadmium ions is shown in Fig. 2. Induction of the stress reporter enzyme can easily be detected calorimetrically or histochemically at a sensitivity of l-10 parts per million (ppm) of cadmium; lethality assays indicate an LC,, value of 120ppm under identical conditions. When the assays are carried out with no bacteria present, induction of the stress response is largely limited to the pharynx of the nematode, as previously described-‘. This is probably due to the fact that the pharynx is the first tissue that is exposed to the stressor; the nematodes rapidly cease feeding in the presence of stressor+, so other tissues have limited exposure to cadmium ions. If normal concentrations of bacteria are included as a food source for the nematodes, cadmium ions induce a stress response primarily in intestinal cells (Fig. 3).
Organic stressors Many organic compounds that cause protein damage also induce a stress response in transgenic nematodes; examples include paraquat, short-chain alcohols, malathion and fungicides of the captan family. The latter compounds illustrate the potential of transgenie nematodes in toxicity screens for protein damaging agents. Figure 3 compares the response of PC72 nematodes to the fungicides captan and folpet, and to the parent-ring compounds phthalimide and tetrahydrophthalimide. The induction of a stress response by these fungicides correlates with the presence of the N-chloroalkylthio group that yields products, when hydrolysis occurs, that are highly reactive with the sulfhydryl groups of proteins’“. Compounds that lack this group, but that contain the phthalimide ring structure, fail to induce the stress-reporter enzyme.
The ‘real world’ - soil and water samples The examples mentioned above illustrate the use of stress assays for homogeneous compounds. What is the potential of this system as a biomonitor of complex
127
reviews 1.21
environmental samples? We recently used strain PC72 to assay water samples taken from the Carnon River, Cornwall, UJS (M. H. A. Z. Mutwakil et al., unpublished). The rank order of toxicity (as measured by transgene induction) was an inverse reflection of macroinvertebrate diversity, and generally correlated with the concentrations of most or all metals assayed. Other experiments demonstrate that C. elegans can be exposed to soil and sediment samples, and then assayed for stress inductiona5, and a method for extracting adult nematodes from soils has been described2h. Furthermore, C. elegans tolerates salt water, and stress assays on marine sediments and water samples have been carried out. Using PC72 with unpolluted samples, background levels of P-galactosidase are either very low or nonexistent, so the system is robust enough to be useful for the biological testing of complex environmental samples.
a
Advantages and limitations biomonitors
0
80
I
I
I
I
I
100
120
140
160
180
Cadmium (ppm)
!
Figure 2 The response of transgenic Caenorhabditiselegans to cadmium. In each assay, young adult PC72 nematodes were exposed to CdCI, in 1 ml volumes of K medium (50mM NaCI, 30mM KCI; Ref. 19). (a) Assays with onitrophenyl p-o-galactopyranoside (ONPG)used approximately 5000 PC72 nematodes for each sample, and an exposure time of five hours to the test solutions. ONPGcalorimetry was performed as described in Ref. 20, and the data points are the average of two determinations. (b) Assays with 5-bromokhloro3indolyl p-ogalactoside (x-gal) were performed as described in Ref. 20: individual nematodes were examined under a dissection micro scope, and the number of blue-stainedindividuals was tallied. The exposure time to cadmium was five hours and the results are the average of four experiments (error bars represent the standard deviation of the means). (c) Lethality was scored after a 24hour exposure to cadmium using the exclusion of methylene blue as the test criterion. The averages of three separate determinations are shown, as are the standard deviations of the means.
of C. elegans
Transgene stress assays detect sublethal concentrations of toxicants within an exposure period of five hours; the turnaround time for calorimetry is seven hours, and that for histochemical detection is 24 hours or less. The nematodes are simple and inexpensive to rear in large quantities, and can be exposed to soil or sediment samples as well as to liquids. They can tolerate wide variations in salinity and pH without there being any effect on the stress response, and stress assays can easily be combined with lethality tests using the same test animals. Correlating the nematode stress bioassay with existing assays will establish its range of applicability to toxicological monitoring. Another approach to the design of stress-sensitive biomonitors involves the use of cell-culture systems carrying IacZ or chloramphenicol acetyl transferase (CAT) reporters; for example, the system described by Todd et al.27 uses human liver cell-lines carrying the gene for CAT driven by one of a series of inducible promoters. Activity is assayed for using an enzymelinked immunosorbent assay (ELISA) in a 96-well format. Although limited comparative data are presently available, the sensitivity (e.g. to cadmium ions) of homologous stress-inducible promoters is similar in the nematode and cell-culture systems. The times required to obtain results are comparable, e.g. 5-24 hours for the C. elegans system, and 24-48 hours for the CAT-Tax(L) assays”. Cultured cells are well suited for use in testing defined chemical compounds, but are likely to be much more sensitive to toxic side-effects that could mask the activity of the reporter enzyme being monitored; they are also unsuited to direct testing of soils and sediments. The ability of a freeliving organism such as C. e?egansto tolerate a wide range of growth conditions allows greater flexibility in dosage, especially when testing crude water samples or chemical mixtures. In addition, C. elegans is inexpensive to maintain, and can be readily used in general chemical laboratories that may lack cell-culture facilities. TIBTECH APRIL 1996 (VOL 14)
128
reviews
Pha
Nerve ring
Caenorhabditis elegans
Tetrahydrophthalamide
Figure 3 Histochemicaldetectionof p-galactosidaseactivity in transgenicCaenorhabdtiselegans.Ineach case, PC72 nematodeswere exposedto the test solutionsfor five hours and then prepared for histochemistryl5.Blue nuclei indicate the presence of active p-galactosidase.The enzyme is sequesteredin cell nucleias a result of the presenceof a nuclearlocalizationsequencein the enzyme. Heat stress (33°C for one hour,followed by a haifhour recovery period) induces a stress response in most cells of the nematode. Including5 parts per million (ppm) of cadmium in the presence of bacteria as a source of food for the nematodes induces a stress response in the gut cells. In the absence of bacteria, the response to cadmiumis largeiy restrictedto the pharynxa.Mercury(2 ppm) induceda stress responsein the pharynx and in the nerve ring. The panel on the right shows the effects of four related compounds, all tested at a concentration of 5Oppm. Phthalimideand tetrahydro phthalimidedo not induce a stress response, while the fungicidescaptan and folpet are effective stress inducers.Stressinductionby captan occurs primarilyin the anteriorportionof the pharynx,whilestress inductionby folpet is primarilyin the posteriorpharynx. TlBTECHAPRlL1996NOL
141
129
reviews A potential limitation of C. elegans as a biosensor would seem to be its largely impermeable cuticle, t?om which it follows that exposure to test substances generally requires ingestion via the pumping action of the pharynx. However, even with a stressor such as captan, which inhibits feeding and produces a response that is limited to the pharynx, stress-reporter assays (such as that for cadmium) provide a rapid and sensitive measure of exposure*s. Further evidence that the internalization of stressors from the medium is generally effective is provided by the fact that heavy-metal tolerances of this nematode are in the same range as those of rats and micez8.
Outlook Transgenic nematode strains provide a rapid assay for the stress response at the organismal level, and should prove useful for toxicological and environmental screening. The general approach outlined here can be extended to include other regulatory regions and reporter genes. Further improvements in the B-galactosidase assays are also possible, for example, by using more-sensitive chromophores or fluorogenic substrates. It is easy to envisage a series of strains based on different promoters, with a variety of reporter genes for different types of assays. Nucleotide sequencing of the C. elegans genome is progressing rapidly, and will probably be completed within three years. Thus, gene sequences with regulatory regions of interest will be readily accessible for the design of new reporter systems. This information, coupled with differential display technology’” to identify transcripts induced by different classes of compounds, will provide a rich source of data from which to create biomonitoring nematode strains.
Acknowledgements This work was supported by grants to EPMC from the Medical Research Council of Canada, the British Columbia Health Research Foundation, the Science Council of British Columbia and StressGen Biotechnologies Corporation.
References 1 Lindquist, S. (1986) Annu. Rev. Biockem.55, 1151-I 191 2 Craig, E. A., Cambill, B. D. and Nelson, R. J. (1993) 12fiuobml.Rev. 57,402314 3 Sanders,B. M. (1993) Crit. Rev. Toxicol. 23, 49-75 4 Stringham, E. G. and Candido, E. P. M. (1994) En&n. Torrcol. Chem. 13, 1211-1220 5 Wood, W. B., ed. (1988) T/w i%natod~ Caenorhabdltis elegans, Cold Spring Harbor Laboratory Press 6 Sukton, J. E., Schmnberg, E., Whm, J. G. and Thomson, J. N. (1983) Dw Biol. 100, 64-119 7 S&ton, J. E. and Horvitz, H. K. (1977) Dw. Biol. 56, 11+156 8 White, J. G., Southgate, E., Thomson, J. N. and Brenner, S. (1983) Pkilos. Trans.R. Sot. London Ser. B 314, l-340 9 Waterston, R. et al. (1993) Cold Spring Harbor Syw~p.Quarrt. Bin/. LVIII, 367-376 10 Mello, C. and Fire, A. (1995) MethodsCell Biol. 38, 452-482 11 Snutch, T. P. and Baillie, D. L. (1983) Cm]. Biurlwrn. Celi Biol. 61, 480-487 12 Russnak, R. H., Jones, 1). and Candido. E. P. M. (1983) h’uckic Atids Res. 11,3187-3205 13 Kussnak, R. H. and Candido, E. P. M. (1985) !Ilol. Cc/l. Biol. 5. 1268-1278 14 Jones, D., Russnak. R. H., Kay, R. J. and Can&do. E. P. M. (1986)
J Biol. Gem. 261, 12006-12015 15 Stringham, E. G., Dixon, 1). K., Jones, D. and Candido, E. P. M.
(1992) 1i40ol. Biol. 0113, 221-233 16 Fire, A., Whxe Harrison, S. and Dixon, D. K. (1990) C&w Y3,18%198 17 SW, E. G. and Candido, E. I’. M. (1993)]. &u. Zml. 266,227-233
18 Enmons, S. W., I&s, M. R. and Hirsh, D. (197’)) Pm ,%fl Acad. Sci. USA 76, 1333-1337 19 Williams, P. L. and Dusenbery, D. B. (199(l) Envim Tuxicoi. Ckem. 9, 1285-1290 20 Fire, A. (1992) Genet.Anal. Tech.Appl. 9, 151-15X 21 Gwen, K., Duce, J. A. and dr Pomerai, 11 I. (1994) Aquat. Toximol. 29, llY-137 22 Fire, A. (1986) EMBOJ. 5, 2673-2680 23 Jaws, D., Stringham, E. G., Babich, S. L and Candido, E. P. M. Toxicolog}~(in press) 24 Lukens, R. J. (1971) in Chemistr), of Fun,@cidaiAction, pp. 59-60, Springer-Verlag 25 Candido, E. P. M., Jones, D. and Stringham, E. G. (1995) 1. Cell. Biockem.(Suppl. lYB), p. 194 26 Do&in, S. G. and Dusenbery. D. B. (1993) Arch. Envrron. Cowtam. Toxicol. 25, 1455151 27 Todd, M. D. et al. (1995) Fundam.ilppl. Toxicol. 2X. 118-128 28 William, P. L. andDwnbery, D. B. (1988)Tmzirool. Ied Htvldz4, &&47X 29 L&kens, M. H. K. et al. (1995) ~\~udeic ilcidr Rex. 23, 32443251
V ilamoura
May 1996, Alganre, Portugal
20-24
* Harvey Lodish - Biogenesis of secretory proteins and cell-surface receptors in mammaliancells. l Charles Weissmann - Molecular biology of prion diseases. l Wolf-Dieter Schleur~g - Animakell technology in drug discovery and pharmacology. SmithKline Beacham session - Animakell vaccines: present and future. Bayer session - Recombinantproteins: biosynthesis and processing. Now Nordisk session - Cell and physiology engineering. Dr Karl Themae session- Cellsandvectorsfor geneticmedicine,Genanbch session- Animakell processengineering. Hoffman-La Roche session - Animal cells as tools for discovery and testing. Baxter session - Tissue engineering and biomedical devices. Meeting Secretary:
Manuel Carrondo, IBET, Apartado 12, P-2780 Oeiras, Portugal. Tel: +351 1442 7787. Fax: +351 1442 1161. TIBTECH APRIL 1996 (VOL 14)
Transgenic Caenorhabditis elegans strains as biosensors E. Peter M. Candid0 and Don Jones Toxicity bioassays rely largely on lethality measurements. Such assays are generally lengthy and expensive, and provide little information on mechanisms of toxicity. A desire to understand the mechanisms by which cells respond to physical and chemical stresses has led to interest in measuring stress proteins as toxicological endpoints. Transgenic strains of the nematode Caenorhabditis elegans that carry a reporter enzyme under control of a stress-inducible promoter have been created. The reporter is easily quantified in intact nematodes, and it responds to a wide range of chemical stressors. Therefore, transgenic C. elegans can provide the basis for a wide range of quick, simple and informative bioassays.
The cellular stress response involves the synthesis of specific sets of proteins by organisms as a result of exposure to physical (e.g. heat) or chemical stressesI. These heat shock or stress proteins confer a protective function on cells, through their ability to minimize, or even reverse, protein denaturation’. It has recently been suggested that the measurement of stress-protein levels is a potentially useful tool in environmental toxicology’. This article discusses progress in the use of transgenic strains of the nematode Cue~orlz&itis elegant as biosensors of the stress response”. Free-living C. eleg~ns is a widely used model organism for genetic, developmental and molecular biologcal studies”. This popularity arises from several desirable characteristics: it is easy to maintain in the laboratory; it is transparent; it has an invariant developmental cell-lineage and a small number of somatic cells; and it is amenable to genetic manipulation and has a small genome. Despite a complement of only 959 cells in its hermaphroditic form, the organism is a fully differentiated animal with a nervous system, specialized muscles, and digestive and reproductive systems. The complete embryonic cell lineageb, as well as the post-embryonic cell lineage of the hermaphrodite7 and the ‘wiring’ diagram of the nervous systems have been determined. A detailed genetic map and an almost complete physical map of the genome are available, and nucleotide sequencing of the entire genome is progressing rapidly’. Methods have been developed for transforming C. eleguns with cloned DNA sequences’“. E. P. M. Can&do and D. Jones aw at the Department of Biochemistry and Molecular Biology, University of British Columbia, 2146 Health Sciences &la/l, Vancouver, Cartada V6T 123. Copyright 0 1996, Elsewer Science Ltd. All rights reserved. 0167 - 7799/96/$15.00
When it is exposed to elevated temperatures, C. elegans undergoes a typical stress response, inducing the synthesis of a number of stress or heat shock proteins (hsps; Refs 11,12). Our laboratory has studied a family of small hsps with molecular masses of 16 kDa (hsp16s). An important feature of the four known hsp 16 genes is their strict inducibility, i.e. under normal culture conditions hsplb proteins are undetectable, but they are synthesized rapidly following heat shock’“,‘J. In order to examine the tissue specificity of the stress response, a number of transgenic strains carrying various fusions of hspl6 genes with the 1acZ gene of Escherichia coli were created15.
Transgenic stress-reporter strains and assay conditions Figure la illustrates the hsp 26-1ucZ fusion gene carried in strain PC72 of C. elegans; this gene consists of IacZ fused in-frame to the second exon of hsy,16- 1. The construct also includes a nuclear localization signal (NLS) from the SV40 virus, which targets the fusion protein to the nucleus, greatly facilitating the identification of expressing cellsi6. PC72 carries 70-80 integrated copies of the hspl64acZ fusion in mixed arrays with the marker plasmid pRF4, which confers a distinct rolling phenotype to the animal”. Figure lb illustrates the principal steps in assaying soluble agents for stress induction. The predominant form of C. elegans is a self-fertilizing hermaphrodite; cultures can be maintained by transferring one or more hermaphrodites to a fresh agar plate approximately once a week. Larger-scale cultures are grown in liquid, and a half-liter culture yields enough specimens to perform hundreds of tests. C. elegam is sufficiently small that suspensions can be aliquotted using standard micropipets. Routine stress tests made use of synchronized PII: SO167.7799(96)10016-O
TIBTECH APRIL 1996 (VOL 14)
126
reviews a
Heat shock elements (HSEs) ,/I \
SV40 ’ NLS /
HRhsp 16-48
hsp16-I-IacZfusion
b Exposure to stress @-7f-d
Heavy-metal
1 Permeabilize
1
1
1
W-W
Figure 1 (a) Structure of reporter gene in transgenic Caenorhabditis elegans strain PC72 (Ref. 15). Open boxes represent exons, lines represent noncoding sequences. The gray and black bars represent the SV40 nuclear localization signal (NLS) and the IacZ coding region from Escherichia coli, respectively, fused in-frame with the second exon of hspl6-1. The two hspl6 genes are transcribed in opposite directions under the control of heat shock elements in the intergenic region: hsp16-48 is oriented leftward, and hspldl rightward. Flanking sequences of the fusion gene, including the 3’ untranslated region, are derived from the hspl6-I gene. (b) Flow diagram indicating the steps involved in evaluating chemical stressors using transgenic nematode strains such as PC72. populations of nematodes at a particular stage of development, usually L3 larvae or adults. Synchronized cultures are readily obtained by inoculating media with embryos prepared by dissolving gravid adults in alkaline bleach’*. Mixed cultures may be preferable for some applications, as different life stages may respond to chemical stressors in varying ways. Stress assays are performed in small (l-3ml) volumes of liquid in multiwell dishes. Agar plates are not used because charged groups in the agar, as well as other components in the growth medium, may bind many chemical stressors, especially metal ions. After being exposed to the stressor, nematodes are permeabilized by a brief acetone treatment, and the whole organisms are assayed directly. P-Galactosidase activity is measured either calorimetrically using the soluble substrate ONPG (o-nitrophenyl P-D-galactopyranoside), or histochemically using the substrate x-gal (5-bromo-4-chloro-3-indolyl P-D-galactoside); TIBTECH APRIL 1996 (VOL 14)
both assays have similar sensitivity (Fig. 2). The colorimetric assay is rapid and yields a quantitative value of enzyme activity (Fig. 2a). Although the histochemical assay is more time-consuming (requiring counting of positive worms under a dissection microscope), it produces a different type of data, indicating tissue specificity, as well as giving an overall measure of stress induction (Fig. 2b). If necessary, levels of hsplh can also be measured by immunological methods.
stressors
Mercury, lead, copper(H), zinc and cadmium ions can induce a transgene stress-response” and, in the absence of bacteria as a source of food (bacterial cell walls may also bind metal ions), different patterns of tissue stress are seen with some of these metal ions (Fig. 3). In all cases, the levels of stress-reporter enzyme are a more sensitive indication of exposure than lethality assays. Guven et al.“’ have also reported stress-inducible 1acZ induction by several heavy metals using nematode strain CB4027, which has a Drosophila hsp70 promoterz2. The PC72 stress response to cadmium ions is shown in Fig. 2. Induction of the stress reporter enzyme can easily be detected calorimetrically or histochemically at a sensitivity of l-10 parts per million (ppm) of cadmium; lethality assays indicate an LC,, value of 120ppm under identical conditions. When the assays are carried out with no bacteria present, induction of the stress response is largely limited to the pharynx of the nematode, as previously described-‘. This is probably due to the fact that the pharynx is the first tissue that is exposed to the stressor; the nematodes rapidly cease feeding in the presence of stressor+, so other tissues have limited exposure to cadmium ions. If normal concentrations of bacteria are included as a food source for the nematodes, cadmium ions induce a stress response primarily in intestinal cells (Fig. 3).
Organic stressors Many organic compounds that cause protein damage also induce a stress response in transgenic nematodes; examples include paraquat, short-chain alcohols, malathion and fungicides of the captan family. The latter compounds illustrate the potential of transgenie nematodes in toxicity screens for protein damaging agents. Figure 3 compares the response of PC72 nematodes to the fungicides captan and folpet, and to the parent-ring compounds phthalimide and tetrahydrophthalimide. The induction of a stress response by these fungicides correlates with the presence of the N-chloroalkylthio group that yields products, when hydrolysis occurs, that are highly reactive with the sulfhydryl groups of proteins’“. Compounds that lack this group, but that contain the phthalimide ring structure, fail to induce the stress-reporter enzyme.
The ‘real world’ - soil and water samples The examples mentioned above illustrate the use of stress assays for homogeneous compounds. What is the potential of this system as a biomonitor of complex
127
reviews 1.21
environmental samples? We recently used strain PC72 to assay water samples taken from the Carnon River, Cornwall, UJS (M. H. A. Z. Mutwakil et al., unpublished). The rank order of toxicity (as measured by transgene induction) was an inverse reflection of macroinvertebrate diversity, and generally correlated with the concentrations of most or all metals assayed. Other experiments demonstrate that C. elegans can be exposed to soil and sediment samples, and then assayed for stress inductiona5, and a method for extracting adult nematodes from soils has been described2h. Furthermore, C. elegans tolerates salt water, and stress assays on marine sediments and water samples have been carried out. Using PC72 with unpolluted samples, background levels of P-galactosidase are either very low or nonexistent, so the system is robust enough to be useful for the biological testing of complex environmental samples.
a
Advantages and limitations biomonitors
0
80
I
I
I
I
I
100
120
140
160
180
Cadmium (ppm)
!
Figure 2 The response of transgenic Caenorhabditiselegans to cadmium. In each assay, young adult PC72 nematodes were exposed to CdCI, in 1 ml volumes of K medium (50mM NaCI, 30mM KCI; Ref. 19). (a) Assays with onitrophenyl p-o-galactopyranoside (ONPG)used approximately 5000 PC72 nematodes for each sample, and an exposure time of five hours to the test solutions. ONPGcalorimetry was performed as described in Ref. 20, and the data points are the average of two determinations. (b) Assays with 5-bromokhloro3indolyl p-ogalactoside (x-gal) were performed as described in Ref. 20: individual nematodes were examined under a dissection micro scope, and the number of blue-stainedindividuals was tallied. The exposure time to cadmium was five hours and the results are the average of four experiments (error bars represent the standard deviation of the means). (c) Lethality was scored after a 24hour exposure to cadmium using the exclusion of methylene blue as the test criterion. The averages of three separate determinations are shown, as are the standard deviations of the means.
of C. elegans
Transgene stress assays detect sublethal concentrations of toxicants within an exposure period of five hours; the turnaround time for calorimetry is seven hours, and that for histochemical detection is 24 hours or less. The nematodes are simple and inexpensive to rear in large quantities, and can be exposed to soil or sediment samples as well as to liquids. They can tolerate wide variations in salinity and pH without there being any effect on the stress response, and stress assays can easily be combined with lethality tests using the same test animals. Correlating the nematode stress bioassay with existing assays will establish its range of applicability to toxicological monitoring. Another approach to the design of stress-sensitive biomonitors involves the use of cell-culture systems carrying IacZ or chloramphenicol acetyl transferase (CAT) reporters; for example, the system described by Todd et al.27 uses human liver cell-lines carrying the gene for CAT driven by one of a series of inducible promoters. Activity is assayed for using an enzymelinked immunosorbent assay (ELISA) in a 96-well format. Although limited comparative data are presently available, the sensitivity (e.g. to cadmium ions) of homologous stress-inducible promoters is similar in the nematode and cell-culture systems. The times required to obtain results are comparable, e.g. 5-24 hours for the C. elegans system, and 24-48 hours for the CAT-Tax(L) assays”. Cultured cells are well suited for use in testing defined chemical compounds, but are likely to be much more sensitive to toxic side-effects that could mask the activity of the reporter enzyme being monitored; they are also unsuited to direct testing of soils and sediments. The ability of a freeliving organism such as C. e?egansto tolerate a wide range of growth conditions allows greater flexibility in dosage, especially when testing crude water samples or chemical mixtures. In addition, C. elegans is inexpensive to maintain, and can be readily used in general chemical laboratories that may lack cell-culture facilities. TIBTECH APRIL 1996 (VOL 14)
128
reviews
Pha
Nerve ring
Caenorhabditis elegans
Tetrahydrophthalamide
Figure 3 Histochemicaldetectionof p-galactosidaseactivity in transgenicCaenorhabdtiselegans.Ineach case, PC72 nematodeswere exposedto the test solutionsfor five hours and then prepared for histochemistryl5.Blue nuclei indicate the presence of active p-galactosidase.The enzyme is sequesteredin cell nucleias a result of the presenceof a nuclearlocalizationsequencein the enzyme. Heat stress (33°C for one hour,followed by a haifhour recovery period) induces a stress response in most cells of the nematode. Including5 parts per million (ppm) of cadmium in the presence of bacteria as a source of food for the nematodes induces a stress response in the gut cells. In the absence of bacteria, the response to cadmiumis largeiy restrictedto the pharynxa.Mercury(2 ppm) induceda stress responsein the pharynx and in the nerve ring. The panel on the right shows the effects of four related compounds, all tested at a concentration of 5Oppm. Phthalimideand tetrahydro phthalimidedo not induce a stress response, while the fungicidescaptan and folpet are effective stress inducers.Stressinductionby captan occurs primarilyin the anteriorportionof the pharynx,whilestress inductionby folpet is primarilyin the posteriorpharynx. TlBTECHAPRlL1996NOL
141
129
reviews A potential limitation of C. elegans as a biosensor would seem to be its largely impermeable cuticle, t?om which it follows that exposure to test substances generally requires ingestion via the pumping action of the pharynx. However, even with a stressor such as captan, which inhibits feeding and produces a response that is limited to the pharynx, stress-reporter assays (such as that for cadmium) provide a rapid and sensitive measure of exposure*s. Further evidence that the internalization of stressors from the medium is generally effective is provided by the fact that heavy-metal tolerances of this nematode are in the same range as those of rats and micez8.
Outlook Transgenic nematode strains provide a rapid assay for the stress response at the organismal level, and should prove useful for toxicological and environmental screening. The general approach outlined here can be extended to include other regulatory regions and reporter genes. Further improvements in the B-galactosidase assays are also possible, for example, by using more-sensitive chromophores or fluorogenic substrates. It is easy to envisage a series of strains based on different promoters, with a variety of reporter genes for different types of assays. Nucleotide sequencing of the C. elegans genome is progressing rapidly, and will probably be completed within three years. Thus, gene sequences with regulatory regions of interest will be readily accessible for the design of new reporter systems. This information, coupled with differential display technology’” to identify transcripts induced by different classes of compounds, will provide a rich source of data from which to create biomonitoring nematode strains.
Acknowledgements This work was supported by grants to EPMC from the Medical Research Council of Canada, the British Columbia Health Research Foundation, the Science Council of British Columbia and StressGen Biotechnologies Corporation.
References 1 Lindquist, S. (1986) Annu. Rev. Biockem.55, 1151-I 191 2 Craig, E. A., Cambill, B. D. and Nelson, R. J. (1993) 12fiuobml.Rev. 57,402314 3 Sanders,B. M. (1993) Crit. Rev. Toxicol. 23, 49-75 4 Stringham, E. G. and Candido, E. P. M. (1994) En&n. Torrcol. Chem. 13, 1211-1220 5 Wood, W. B., ed. (1988) T/w i%natod~ Caenorhabdltis elegans, Cold Spring Harbor Laboratory Press 6 Sukton, J. E., Schmnberg, E., Whm, J. G. and Thomson, J. N. (1983) Dw Biol. 100, 64-119 7 S&ton, J. E. and Horvitz, H. K. (1977) Dw. Biol. 56, 11+156 8 White, J. G., Southgate, E., Thomson, J. N. and Brenner, S. (1983) Pkilos. Trans.R. Sot. London Ser. B 314, l-340 9 Waterston, R. et al. (1993) Cold Spring Harbor Syw~p.Quarrt. Bin/. LVIII, 367-376 10 Mello, C. and Fire, A. (1995) MethodsCell Biol. 38, 452-482 11 Snutch, T. P. and Baillie, D. L. (1983) Cm]. Biurlwrn. Celi Biol. 61, 480-487 12 Russnak, R. H., Jones, 1). and Candido. E. P. M. (1983) h’uckic Atids Res. 11,3187-3205 13 Kussnak, R. H. and Candido, E. P. M. (1985) !Ilol. Cc/l. Biol. 5. 1268-1278 14 Jones, D., Russnak. R. H., Kay, R. J. and Can&do. E. P. M. (1986)
J Biol. Gem. 261, 12006-12015 15 Stringham, E. G., Dixon, 1). K., Jones, D. and Candido, E. P. M.
(1992) 1i40ol. Biol. 0113, 221-233 16 Fire, A., Whxe Harrison, S. and Dixon, D. K. (1990) C&w Y3,18%198 17 SW, E. G. and Candido, E. I’. M. (1993)]. &u. Zml. 266,227-233
18 Enmons, S. W., I&s, M. R. and Hirsh, D. (197’)) Pm ,%fl Acad. Sci. USA 76, 1333-1337 19 Williams, P. L. and Dusenbery, D. B. (199(l) Envim Tuxicoi. Ckem. 9, 1285-1290 20 Fire, A. (1992) Genet.Anal. Tech.Appl. 9, 151-15X 21 Gwen, K., Duce, J. A. and dr Pomerai, 11 I. (1994) Aquat. Toximol. 29, llY-137 22 Fire, A. (1986) EMBOJ. 5, 2673-2680 23 Jaws, D., Stringham, E. G., Babich, S. L and Candido, E. P. M. Toxicolog}~(in press) 24 Lukens, R. J. (1971) in Chemistr), of Fun,@cidaiAction, pp. 59-60, Springer-Verlag 25 Candido, E. P. M., Jones, D. and Stringham, E. G. (1995) 1. Cell. Biockem.(Suppl. lYB), p. 194 26 Do&in, S. G. and Dusenbery. D. B. (1993) Arch. Envrron. Cowtam. Toxicol. 25, 1455151 27 Todd, M. D. et al. (1995) Fundam.ilppl. Toxicol. 2X. 118-128 28 William, P. L. andDwnbery, D. B. (1988)Tmzirool. Ied Htvldz4, &&47X 29 L&kens, M. H. K. et al. (1995) ~\~udeic ilcidr Rex. 23, 32443251
V ilamoura
May 1996, Alganre, Portugal
20-24
* Harvey Lodish - Biogenesis of secretory proteins and cell-surface receptors in mammaliancells. l Charles Weissmann - Molecular biology of prion diseases. l Wolf-Dieter Schleur~g - Animakell technology in drug discovery and pharmacology. SmithKline Beacham session - Animakell vaccines: present and future. Bayer session - Recombinantproteins: biosynthesis and processing. Now Nordisk session - Cell and physiology engineering. Dr Karl Themae session- Cellsandvectorsfor geneticmedicine,Genanbch session- Animakell processengineering. Hoffman-La Roche session - Animal cells as tools for discovery and testing. Baxter session - Tissue engineering and biomedical devices. Meeting Secretary:
Manuel Carrondo, IBET, Apartado 12, P-2780 Oeiras, Portugal. Tel: +351 1442 7787. Fax: +351 1442 1161. TIBTECH APRIL 1996 (VOL 14)