Introduction to the Laboratory Studies

The Target Enzyme
The Laboratory Studies

As suggested in the preceding literature survey, an assay for glutathione transferase shows good potential as a marker for water pollution. Previous work at WRc has resulted in the development of an assay for GST based on the original work of Habig et al. (1974). Briefly the procedure is as follows:

A sample of the tissue of interest is blotted dry, weighed, and homogenised in 5ml of ice cold 0.1M phosphate buffer (pH 6.5). The homogenate is then centrifuged at 40,000 rpm (approx. 100,000g) for 30 minutes at 4°C. The supernatant is decanted into stoppered, polyethylene vials and stored at -20 to -30°C for subsequent analysis.

For analysis, the sample is allowed to defrost before a 2.5 to 5% dilution of the sample is made into a solution of 2.5mM 1-chloro-2,4-dinitrobenzene (CDNB) and 2mM glutathione. Activity is measured as the absorbance change per minute at 340nm.

From the literature review it is apparent that most studies of GST have been undertaken using vertebrates. However the use of invertebrates, and more importantly bivalves, as biomarkers has many advantages. These have been discussed by Golderg (1986) in the context of data from the U.S. mussel watch. Some of the more relevant points are;

o Bivalves are sedentary and thus effective in defining the pollution status for a given area.

o Bivalves appear to be resistant to pollutants and often survive in areas where other organisms are eliminated.

o Bivalves can be readily transplanted from one area to another and are relatively easy to culture in a laboratory environment.

Although Goldberg (1986) states that bivalves exhibit a low or non-detectable activity of enzymes involved in xenobiotic metabolism, an active GST system has been demonstrated in mussels, and levels are responsive to environmental stimuli. Most previous studies have involved the use of the pea mussel Sphaerium corneum, and have investigated a wide range of factors such as body weight, seasonal effects, reproductive status, handling stress and storage (Garrood, A.C., Beverley, M., Johnson, I., 1990).

The test system has been found to provide a good response in field trials. For example a 30 day deployment of the marine mussel Mytilus edulis at sites on the heavily industrialised Mersey estuary showed significant increases in both gill and digestive gland GST compared with controls (Johnson, I., Fleming, R., Garrood, A.C., Beverley, M., 1992). A 28 day deployment of the swan mussel Anodonta cygnae upstream and downstream of a sewage outfall showed a significantly reduced GST activity in the gills of animals downstream of the outfall, though not in the digestive gland (Polak, M., 1992). Similarly a 6 day deployment of swan mussels around a pharmaceutical discharge produces a reduction of activity relative to controls, although this was not significant (Johnson, I., et al., 1992)

It has proved more difficult to verify these results in laboratory studies. A significant enhancement of GST activities has been found in pooled tissue homogenates of Sphaerium exposed to the organochlorine compounds lindane, permethrin and tecnazene (Garrood, et al., 1990). However the percentage induction is small and to a large extent masked by fluctuations in control group activity. These factors also confounded a laboratory study involving lindane exposure in Anodonta (Polak, M., 1992).

An effective laboratory demonstration of the test is considered necessary to provide a paradigm for the interpretation of field studies. Optimisation of the assay been attempted by Johnson et al., (1992). They studied the modifications of the assay using differential substrates, more stringent purification procedures and the activity of microsomal vs. cytosolic extracts from Sphaerium. The measurement of GST activity in preparations of different tissues from Mytilus was also assessed. It was concluded that, where possible, gill extracts should be used preferentially as these show the highest levels of induction. They found that the use of different substrates did not provide an increase in discriminatory power, while the other procedures exacerbated the already time consuming nature of the procedures and the need for specialised equipment such as high speed centrifuges.

The target enzyme

The most influential of the early work on glutathione S-transferases has proved to be that of Habig et al. (1974) who provided an investigation of the activity towards different substrates of 5 rat isoenzymes, and categorised them according to their order of elution from CM-cellulose. Habig et al. (1974) also demonstrated that GST exists in its native state as a dimer. GST is principally cytosolic, but microsomal and mitochondrial GST has around 10 to 30% (depending on substrate) of the activity of cytosolic GST, when expressed as activity per milligram of protein (Mukhtar, H., Baars, A.J., Breimer, D.D., 1981).

As well as detoxication by GSH conjugation, at least one isoenzyme (GST-B, also known as ligandin) is believed to play a role in detoxication by covalent binding of certain reactive substances (Moron, M.S., Depierre, J.W., Mannervik, B., 1979). GST binds with high affinity a variety of hydrophobic compounds such as polycyclic aromatic hydrocarbons and dextramethasone (Pickett, C.B., Lu, A.Y.H., 1989).

The steady state kinetics of rat liver GST have been extensively studied, and the most recent work suggests a sequential mechanism. There is a very strict structural requirement for the binding of GSH to the active site, which is hydrophobic in nature (ibid, 1989).

The distribution of GST varies according to tissue. In plaice, it is predominantly expressed in the liver, with significant activity present in the kidney, intestine and gills (Leaver, M.J., Scott, K., George, S.G., 1992). A similar relationship has been found in Sphaerium, where GST activity is located predominantly in the digestive gland, with smaller amounts present in the gills, siphon and foot (Johnson et al., 1992).

At least three isoforms have been identified in the mussel Mytilus edulis (Sheenan, D., Crimmins, K.M., Burnell, G.M., 1991). Three isoforms have also been identified in an antarctic fish, Dissostichus mawsoni (Falkner, K.C., Clark, A.G., 1992). This compares with the 6 isoforms which have been isolated from rat liver by Habig and Jakoby (1981). These isoforms can be distinguished in terms of their substrate specificity.

In rat liver, the isoenzyme most active towards CDNB is GST-A. However Leaver et al. have found that CDNB-conjugating activity in plaice liver is not indicative of GST-A, although it is a good measure of GST-B. They concluded that the disparity between GST-A and GST-B distribution and CDNB-conjugating activity implied that other, unidentified, isoforms with comparitively higher CDNB conjugating activities are present, particularly in the liver (Leaver, M.J., Scott, K., George, S.G., 1992). There is considerable species variation in the activity of fish liver GST towards a range of substrates, as well as in overall activity (Donnarumma, l., De Angelis, G., Gramenzi, F., Vittozzi, L., 1987), suggesting that the relative expression of isoenzymes is dependent on species.

The induction of GST-A in flounder liver has been studied by Scott et al. (1992) who found that, 6 days after 3-methylchloranthrene treatment (3-MC - a polycylic aromatic hydrocarbon), GST-A mRNA was reduced by 75%, GST-A itself was reduced by 20%, and GST activity towards CDNB was halved (Scott, K., Leaver, M.J., George, S.G., 1992). The polychlorinated biphenol (PCB) Aroclor 1254 was found to have little effect on activity and protein levels, despite halving mRNA. Trans-stilbene oxide (TSOX) substantially increased activity, and protein and mRNA levels. A similar result was found in plaice hepatocytes, where 3-MC and Aroclor 1254 decreased GST-A mRNA and protein but increased CDNB conjugation, and TSOX substantially increased all measures (Leaver, M.J., Clarke, D.J., George, S.G., 1992). Given the discrepancy between GST-A induction and induction of CDNB activity, it is likely that the various isoforms are differentially regulated.

The mechanism of gene regulation have not been fully investigated. However it is clear that the induction of at least one important class of monomer (Y_a) is dependent on the presence of the Ah (dioxin) receptor (Pickett, C.B., Lu, A.Y.H., 1989). In view of this the inability of Hahn et al. (1992) to demonstrate the Ah receptor in invertebrates may have important consequences for the use of GST activity in bivalves as a biomarker.

The laboratory studies

The following reports detail a series of investigations which were initiated with the aim of optimising the current assay for bivalve GST activity. The ideal assay would have low intra-treatment variability, a clear dose-dependent reaction to a specific contaminant or contaminants, and be rapid and simple with a minimal dependence on specialised equipment.

The first five studies involve the re-analysis of previously prepared samples. The intial two studies look at the reaction over a combination of wavelengths, while the third and fourth examine the activity of gill GST with a variety of different substrates. The fifth study investigates the pH dependence of the enzyme. The sixth and final study forms the central focus of the project, and uses fresh preparations to analyse the response of GST and GSH to the toxicant, chlorothalonil.

All spectrophotometric work was conducted using a Beckman du-65 spectrophotometer. A series of programs was written for it to facilitate the investigations. These are detailed in Appendix A. The high speed centrifugation in study 6 was conducted using a Centrikon T-1065 centrifuge and polyallomer tubes. The low speed centrifugation used an MSE Mistral 2000 centrifuge with polycarbonate tubes.

The force used was calculated by assuming that;

F/m = w²r

Where; F/m = is acceleration in ms^-2

w = angular velocity in radians per second

r = radius of the zone of sedimentation in metres

and was converted into multiples of gravity by dividing by 9.81.

Chemicals were purchased from Sigma chemical company or BDH, and were of reagent grade or higher. Chemicals were prepared in distilled water or absolute ethanol, as appropriate, and stored in the dark at 5°C. Unstable chemicals, such as GSH, were prepared at least weekly (usually daily).

Unless otherwise stated, all experimental factors were pseudo-randomised using the BASIC program listed in Appendix A. Statistics were performed using PC Minitab, except Tukey's test (Study 6), which was performed using Genstat v5.5., and Dunnet's test (Study 2), which was performed manually. All statistical analysis was conducted using a critical significance level of 5%.