WO1995010767A1

WO1995010767A1 - Assay method

Info

Publication number: WO1995010767A1
Application number: PCT/GB1994/002224
Authority: WO
Inventors: Adrian Robert Ford; Jay Lewington
Original assignee: Merck Patent Gmbh
Priority date: 1993-10-15
Filing date: 1994-10-12
Publication date: 1995-04-20
Also published as: JPH09503666A; EP0723658A1; BR9407815A; AU7819494A

Abstract

An assay method involves mixing bioluminescent bacteria with a liquid sample possibly containing analyte and observing a reduction in light output caused by the analyte. Observation is started only 15 seconds to 5 minutes after initial formation of the test mixture so as to reduce interference from environmental effects.

Description

ASSAY METHOD

This invention concerns a method of assaying a liquid sample for analyte, preferably a toxicant or other anti-microbial agent, by the use of a liquid suspension of signal-generating micro-organisms. In a preferred embodiment, the signal-generating micro¬ organisms are bioluminescent bacteria. The method involves mixing together a liquid sample possibly containing the analyte and an aliquot of the liquid suspension. The analyte reduces the signal generated (e.g. light emitted) by the micro-organisms in the suspension, and this reduction in signal is observed as an assay for the toxicant (GB 2005018).

The use of bacterial bioluminescence as an indicator of toxicity is well established, and the kit marketed by Microbics Inc. under the Trade Mark Microtox is in routine use with a number of environmental laboratories. With current microbial toxicity tests there is a need for the bacterial reagent to be used according to particular restricted test conditions and procedures which, it is felt, may affect the measured toxicity of certain agents.

It would be desirable to develop a test which can be used over a wide pH range, at different temperatures, with no compensation for turbidity or colour in the sample, and most importantly without the use of osmotically potent material to compensate for osmotic shock caused by adding the sample to the reagent. There is need, in short, for a test which can be used in the environment of the sample, that is without any additions to the sample, and with the minimum of interfering agents in the bacterial reagent itself. It is an object of this invention to meet that need. But the invention is not confined to methods and kits which meet that need; it is of broader scope, as will hereafter be explained. The invention provides a method of assaying an analyte, by the use of a liquid suspension of signal-generating micro-organisms, which comprises mixing together a liquid sample possibly containing the analyte and an aliquot of the liquid suspension to form a test mixture and thereafter observing a signal generated by the micro-organisms, wherein the size of the signal generated by the micro-organisms in the test mixture changes at a first rate during a first period, and then changes at a second rate during a second period due to the effect of analyte, and observations of the size of the signal are made at different times during the second period and are compared as an assay for analyte.

For a better understanding of the invention, reference is directed to Figure 1 of the accompanying drawings, which is a graph of logarithmic reduction in light output against time. The lines on the graph have been generated by the following experiment. Four aliquots of an aqueous suspension of bioluminescent bacteria have been provided in vessels numbered 1, 2, 3 and 4. To each of vessels 2, 3 and 4 is added a fixed volume sample of a different analyte potentially containing toxicant. To vessel 1 is added an equivalent volume of distilled water. All additions are made at time 0. The numbered lines in Figure 1 correspond to the light output from the numbered vessels in the experiment:

The light output from vessel 1 decreases at a slow rate during a first period A which lasts 30 seconds, and thereafter remains constant for the duration of the experiment. The initial decrease in light output is believed due to various environmental factors: dilution of the bacterial suspension; pH change; osmotic change; colour and turbidity changes. The observation made by the inventors, which forms the basis of the present invention, is that the changes in light output caused by these environmental factors are quite short lived. After a first period, 30 seconds in this case, any further change in light output would have been due to the presence of toxicant. Since no toxicant was present in tube 1, the light output remained constant after 30 seconds.

In tube 2, environmental factors reducing light output during the first period A were the same as in tube 1. But the light output in tube 2 went on decreasing during a second period B, which lasted for the remainder of the experiment, albeit at a slower rate than during the first period A. This decrease in light output during the second period B was due to the presence of toxicant in tube 2. - In tube 3, the environmental effects were more severe, and resulted in a sharper reduction in light output during the first period A. There was no further reduction in light output during the second period B, indicating that no toxicant was present in vessel 3.

In vessel 4, environmental factors caused a sharp drop in light output during the first period A; and toxicant caused a further drop in light output during the second period B. Each of the four curves has an elbow at the transition from the first period A to the second period B. The idea of the invention is to measure changes (normally reductions) in light output only after the elbow and during the second period B as an assay for analyte. Preferably the light output from the test sample is compared to the light output from the control sample during this second period B; a significant change in the ratio of the two light outputs is indicative of analyte in the test sample.

Thus the assay may be performed to determine the toxic effect of a sample defined by a formula

Toxic Effect x 100

where

T_χ is the signal x minutes after addition of a toxic sample to a test tube,

T_y is the signal y minutes after addition of the toxic sample to the test tube,

C_χ is the signal x minutes after addition of a control sample to a control tube,

Cy is the signal y minutes after addition of the control sample to the control tube, x and y are times chosen within the said second period and x is greater than y.

The duration of the first period A is typically in the range of 10 or 15 seconds to 5 minutes. By continuously monitoring the light output of the test sample from time zero, the duration of the first period A is easily determined. Preferably determining the duration of the first period, and the time for starting to observe the light output as an assay for analyte, are controlled by a computer. Alternatively, these factors can be determined manually. Alternatively, if experience has shown what is the likely duration of the first period A, it is possible to start measuring light output at a fixed time, say 30 seconds or 3 minutes, after admixture of the reagents to form the test mixture. That timed start should be not earlier than the "tart of the second period B, otherwise inaccuracies due to environmental effects may creep into the measurement. On the other hand, the timed start should not be so much later than the start of the second period B that the light output of the test mixture has already dropped to a level that is measured only inaccurately or with difficulty, or is too low to detect any significant subsequent effect by the toxicant.

The discovery described above permits further improvements in the assay protocol, compared with currently available commercial assays. a) In a commercial assay, a relatively small volume of an aqueous suspension of bioluminescent bacteria is added to a relatively large volume of analyte containing toxicant. The bacterial concentration in the resulting test mixture is low, so the light output is low, and is easily reduced to unmeasurable levels unless precautions are taken. According to the present invention, by contrast, the volume of the liquid sample of analyte is preferably not more than ten times, and typically from 2.0 to 0.5 times, the volume of the aliquot of the liquid suspension of bioluminescent bacteria. This arrangement permits the use of enough bacteria to provide a bright easily measured light output, and of enough sample to permit ready and sensitive detection of any analyte present. b) in one commercial assay kit, the bioluminescent bacteria are provided in lyophilised form together with a reconstitution buffer. Because the system is based on a marine bacterium (Photobacterium fischeri. , the reconstitution buffer needs to contain an osmotically-potent compound such as salt. Because of the large volume of the sample used, salt (or other osmotically-potent compound) needs to be added to the sample prior to addition of the bioluminescent bacteria suspension. Typically, the analyte is assayed at several dilutions, and each needs to have the same salt concentration.

Stabilised micro-organisms such as bioluminescent bacteria are also preferably used in the present invention. Stabilisation is preferably achieved by lyophilisation. But by contrast with the currently available systems, the reconstitution buffer preferably contains an osmotically-potent non-salt compound such as sucrose, dextran or polyethylene glycol. Preferably also, the analyte is tested, optionally after dilution but without the addition of any chemical agent. Specifically, because the sample volume is relatively small and the volume of micro- organism suspension relatively large, the addition of an osmotically-potent compound to the sample is found not to be necessary to preserve the signal-generating micro-organisms, e.g. not necessary to prevent extinction of the bioluminescent bacteria. The assay method of the present invention can be performed with the test mixture (and the control mixture) at any reasonable temperature e.g. in the range 5'C - 35^*C. Preferably the assay temperature of both the sample and the control is 15°C. The above description has referred particularly to bioluminescent bacteria, and these are indeed preferred. But in principle any signal- generating micro-organisms can be used. On the basis of the results reported in the examples, it can be predicted with confidence that any signal from any signal-generating micro-organism will be affected, in a relatively short first period by environmental considerations, and in a relatively long second period by analyte in the sample under assay. During the aforesaid second period, measurements of light output (or signal generated) are made at predetermined intervals on both the test mixture and the control mixture. The observations may be plotted on a graph resembling Figure 1. It is found in practice that analyte (toxicant) in the sample causes a reduction in light output during the second period B and that the reduction is dose-responsive, i.e. greater analyte (toxicant) concentrations produce greater light reductions. If the nature of the analyte is known, the assay can thus be used to estimate concentration. If the nature of the analyte is not known, or if a mixture of analytes may be present, the assay is used in a more qualitative way. The discussion up to now has assumed that the relatively rapid reduction in light output during the first period A after forming the test mixture, is due to environmental effects, pH, osmotic shock, etc. And indeed that is usually the case. But one other possibility is also envisaged.

In some circumstances, the sample may contain a fast-acting analyte in addition to one or more other analytes. When fast acting analytes are present at concentrations which do not completely extinguish the bioluminescence in the initial period of action, then the remaining bioluminescence can be used to measure the presence of other analytes. The halogens are particular examples of fast-acting biocides used to treat industrial process waters. For example chlorine acts quickly on luminescent bacteria, to reduce the light output in a dose responsive manner often within the first one or two minutes after formation of a test mixture. In this aspect of the invention, the light output of the test mixture reduces during the first period A due to the presence in the liquid sample of a fast-acting analyte, such as a halogen. Thereafter, the light output of the test sample is observed during the second period B as an assay for other analytes possibly present.

In another circumstance where two or more analytes have significantly different rates of action on the bioluminescence, it is possible to quantify the analytes independently in the same assay by measuring the light reduction in period A, then period B, etc. and comparing these with a previously prepared calibration. This has been demonstrated by the inventors.

Alternatively, prior to performing the assay, a neutralising agent may be added to react with the fast-acting analyte, effectively rendering it inactive and preventing interference in the assay of the slow- acting analyte. As neutralising agents, a variety of materials can be employed typically reducing agents which react with halogen and which do not interfere with bioluminescent micro-organisms. A preferred neutralising agent is sodium thiosulphate.

According to the invention, the analyte acts to alter (usually reduce) the light output (or other signal generated) during a relatively long second period B after formation of the test mixture. The nature of the analyte is not material to the invention. A considerable number of analytes is known to reduce the light output of bioluminescent bacteria in a dose- responsive way; the majority are found to act in a persistent gradual manner. Examples of analytes, all of which have been successfully assayed by the method of the invention, are heavy metals, biocides, antibiotics, quaternary ammonium salts, ionic detergents, non-ionic detergents, phenolics, halogenated hydrocarbons, acids and alkalis. The analytes are usually toxic substances or toxicants. Toxicity is defined as "a potential or capacity to cause adverse effects on living organisms, generally a poison or mixture of poisons. Toxicity is a result of dose or exposure concentration and exposure time, modified by variables such as temperature, chemical form and availability". Biocides and antibiotics are examples of toxicants. But in theory any substance is toxic if present at a high enough dose.

The analyte is or may be present in solution or suspension in a sample of a liquid, generally an aqueous liquid, to be assayed. Examples are river water, sea water, industrial effluents, run-off from land-fills, drainage water, sewage, water from industrial processes (e.g. cooling towers, paper mills, air conditioning systems, humidifiers, lubricants), swimming pools, liquid holding tanks, water bottoms in oil tankers/storage tanks, drilling muds, in-process water (i.e. water used to make something as opposed to cool it), agricultural discharges (farmyard slurries, land drainage) , and drinking water.

The invention provides the following advantages:- i) The combination of various features discussed above permit the use of relatively high concentrations of bioluminescent bacteria and relatively low volumes of analyte-containing sample, without sacrificing sensitivity. ii) The assay configuration and the method of analysis proposed here reduce the effect of osmotic shock to an extent that enables the addition of salt to the assay sample to be eliminated. iii) Because the signal generated in the test is very high, the effect of colour and turbidity of the sample are readily compensated for by the time lag between formation of the test mixture and light measurement for assay purposes. iv) Variation in results, due to changes in the physical environment which do not affect toxicity, are also eliminated by the time lag. v) In respect of several toxicants, as demonstrated below in the Examples, the system is more sensitive than at least one of the commercially available systems. vi) Ability to differentiate and quantify a mixture of two or more known analytes with differing kinetics of action using a single test reaction.

Reference is directed to: Figures 2 to 6 of the accompanying drawings, each of which is a graph of log reduction of light output against time; and to Figure 7 to 9 of the accompanying drawings, each of which is a graph of % Toxic Effect against a specified variable.

EXPERIMENTAL

The test reagent used in these experiments is based on the marine bacterium Photobacterium phosphoreum NCIMB 844, publicly available. But other bioluminescent bacteria, either naturally occurring or recombinant, could have been used with essentially similar results. The bacteria were cultured in a turbidostat, resuspended in a lactose/potassium chloride cryoprotectant, dispensed in 100 μl aliquots and lyophilised.

Each vial of bacterial reagent is intended for one assay. 0.5 ml of a reconstitution buffer consisting of 7.5% w/v sucrose in Analar water, was added to the reagent. The resulting bacterial suspension produces a stable light signal after 20 minutes which lasts for at least 40 minutes. 0.5 ml of the analyte sample under test was added to the vial. Any osmotic shock or other sample effect was allowed to take place over a time period which is variable and dependent upon the sample under test. A base light reading was then taken to represent the status of the cells once osmotic shock and other environmental effects had occurred, and before significant toxic effect caused by analyte. Subsequent light readings were taken at intervals timed from the base light reading. Analyte was assayed by measuring the light loss between the first recorded reading and the reading at one or more of the time intervals.

EXAMPLE 1 Phenol was the analyte. 60 mg of crystalline phenol was weighed out by difference and made up to 1 litre in Analar water. This solution was further diluted to 40 mg/1 and 20 mg/1. 0.5 ml samples of these analyte solutions at ambient temperature were used to make test mixtures with 0.5 ml of the bioluminescent bacteria suspension of the system described above. Light output was measured at 0, 3, 5, 10 and 20 minutes after formation of the test mixture. Figure 2 shows reduction in light output at these timed measurements, compared to the initial light output of the newly formed test mixtures. The readings after 10 minutes are such that it is not possible to distinguish between analyte-containing samples and the control sample. After 20 minutes, samples containing 40 and 60 ppm of phenol (but not 20 ppm of phenol) are distinguishable from the control sample.

Figure 3 is a graph of log reduction in light output, corresponding to Figure 2 except that light readings 3 minutes after formation of test mixtures are used as the starting point. A much clearer picture emerges. The 60 ppm phenol sample was clearly distinguishable after 5 minutes. All three analyte samples were clearly distinguishable after 10 minutes from the control sample. In this and the following examples, no attempt has been made to measure the duration of the first period A (Figure 1 ) during which environmental effects predominate. Experience in this experiment has shown that the duration of that first period is always less than 3 minutes. So the light output measurement at 3 minutes has been used as a base and gives a clear indication of the presence of phenol in concentrations as low as 20 ppm.

EXAMPLE 2

Sodium dodecylsulphate (SDS) was used as the analyte. 20 mg of SDS was weighed out by difference and made up to 1 litre in Analar water. This solution was then further diluted to 1 mg/1, 0.5 mg/1 and

0.25 mg/1. The procedure was as described in Example 1, except that the timed measurements were made 0, 2.5, 5, 10 and 15 minutes after formation of the test mixtures. Figure 4 is a graph of log reduction in light output based on the readings made 0 minutes after formation of text mixtures. The figures are such that apparently different results are obtained from readings at different time intervals. Figure 5 is a corresponding graph, made using as a basis the light output figures obtained 2.5 minutes after formation of test mixtures. Now the results are clear, and SDS concentrations as low as 0.25 ppm are clearly distinguishable from the control after 10 minutes.

EXAMPLE 3

Cetyl trimethyl ammonium bromide (CTAB) was used as the analyte. 112 mg of CTAB was weighed out by difference and made up to 1 litre in Analar water. This solution as then further diluted to 11.2 mg/1, 5.51 mg/1 and 2.75 mg/1. The same procedure was followed as in Examples 1 and 2, and light output readings were taken 2.5, 5 and 10 minutes after formation of test mixtures.

Figure 6 is a graph of log reduction in light output using the 2.5 minute readings as base. All the samples containing CTAB at various dilutions were clearly distinguishable from the control and from each other on readings made 5 and 10 minutes after initial test mixture formation.

EXAMPLE 4

Zinc sulphate (Zn) was used as the analyte. 1g of zinc sulphate heptahydrate was weighed out by difference and made up to 1 litre in Analar water to give a working concentration of 220mg Zn⁺⁺/litre. This solution was then further diluted to 27.5mg Zn⁺⁺/litre, 13.75mg Zn⁺⁺/litre and 6.87mg Zn⁺⁺/litre.

The following Table shows light remaining 5 minutes after initial test mixture formation, expressed as a % of the light remaining 2.5 minutes after initial test mixture formation, at various Zn concentrations.

Zn cone , ( ppm) ( T₅/T₂ . ₅ ) ^{x 1 0}°

0 1 00

6 . 87 7 1 3 . 75 0 27 . 5 0

Even with such an aggressive analyte as zinc, it was still possible to detect the effect of the analyte using the method of the invention. EXAMPLE 5

Three vials of test reagent (described under Experimental above) were treated as described previously except that they were reconstituted with 1 ml of 1 \ % (w/v) sucrose solution, and left at 15'C for 20 minutes. The contents of the vials were then pooled and redistributed into six ml aliquots.

A series of dilutions of zinc sulphate were prepared in Analar water to final concentrations of 0.001, 0.01, 0.1, 1.0 and 10 mg/litre.

The six tubes of reconstituted test reagent were positioned in a temperature controlled assay rack (15'C) .

Using an appropriate time delay between additions to allow measurements of light values in a luminometer, $ ml amounts of water (control) and the five dilutions of zinc sulphate were added in ascending concentration to the six vials. The light output from each vial was measured 2, 5, 10, 15 and 20 minutes after the addition of sample.

Raw data for sensitivity of Photobacterium phosphoreum to Zinc Sulphate

Light Readings 20 min

2 mins 5 mins 10 mins 15 mins 20 mins Toxic Effect

Control 47000 48000 43000 38000 34000

0.001mg/l 47000 47000 42000 37000 33000 2.94

O.Olmg/l 46000 46000 41000 34000 31000 6.84

0.1 mg/1 46000 42000 31000 20000 15000 54.92 l.Omg/l 26000 8500 1500 340 95 99.49 lO.Omg/l 3000 270 16 1 99.95

³ This data is shown graphically in Figure 7 Toxicity is calculated from the formula

Toxic Effect = x 100

that is for 0.1 mg/1 at 20 minutes Effect = [1-{(15000 ÷ 46000) ÷ (34000 ÷ 47000)} ] x 100% = 54.92%

EXAMPLE 6 Compensation for the Effects of Turbidity The time delay part of the patent should allow rapid effects like turbidity to affect the reagent before the analysis of toxicity begins. This experiment is designed to show that increasing levels of turbidity do not change the observed toxic effect (calculated as detailed above) of a standard solution of zinc sulphate provided that analysis of toxicity is made after a time delay.

Method Amerlex latex solution was diluted in water to 1:100, 1:200, 1:400 and 0 (no latex) for use as a source of turbidity. The percentage light transmission at 490 nm (wavelength chosen to be close to the emission of test reagent) was measured for each dilution. AMERLEX is a Registered Trade Mark.

Zinc sulphate was added to the Amerlex solutions and the water control to a final concentration of 0.1 ppm, and mixed thoroughly.

Nine vials of test reagent were reconstituted as described in Example 5 above, and pooled.

Reconstituted reagent was divided into seventeen \ ml aliquots and placed into a temperature controlled assay rack at 15'C. 1 ml amounts of the dilutions of latex were added to separate vials in triplicate, using an appropriate time delay between additions to allow measurement of light values in a luminometer. A sample of water containing no zinc sulphate was used as a blank control. The light output from each vial was measured just before addition of sample and again 2 and 5 minutes after the addition of sample.

Results

Results are shown in the following Tables and in Figure 8.

a) Turbidity of latex solutions

% Latex % Transmission

(v/v) at 490 nm

0 100

0.125 21.5

0.25 6.3

0.5 1.3

1.0 0.4

b) Toxicity of Latex solutions

SAMPLE LIGHT READINGS % Toxic effect % latex (v/v) T₀ mins T₊2 mins T₊5 mins 2-5 mins 0-5 mins

Control 443.7* 227.9* 259.4* (No Zinc or latex)

0 (No 447.3 285.4 224.0 31.04 14.34 latex)

429.7 287.6 225.7 31.05 10.16

0.125 423.5 225.7 190.7 25.69 22.9

430.6 223.4 189.9 25.32 24.57

436.6 230.3 191.0 27.14 25.07

0.25 399.5 174.5 153.3 22.82 34.36

399.0 168.2 152.9 20.14 34.35

389.1 172.2 149.5 23.72 34.28

0.5 321.2 99.42 89.76 20.68 52.20

310.7 98.28 87.8 21.51 51.66

306.5 96.3 87.4 20.26 51.22

1.0 305.8 58.66 54.12 18.94 69.73

301.2 55.61 51.69 18.34 70.65

294.3 56.21 50.99 20.30 70.36

* Mean of three readings

Toxicity was calculated using the formula described above, where the control values are obtained from the tubes containing no zinc or latex, e.g. for 2-5 minutes toxic effect with no. latex % Toxic Effect = [1-{(224 ÷ 285.4) ÷ (259.4 ÷ 227.9)}] x 100% = 31.04% Conclusions and Discussion

Since all concentrations of latex contain the same amount of zinc sulphate it would be expected that the toxic effect for each vial would be the same. In the standard method where there is no time delay, the measured toxicity appears to increase, which indicates that the presence of latex affects the response of the test in a concentration dependent way. Since Amerlex is a non-toxic component of a number of biological assay systems, it is unlikely to be due to any direct effect on the viability of the organism, and is therefore more likely to be due to interference with the bioluminescent signal produced by the organism. When the time delay method is used the response is much flatter, showing the same measured toxicity in each sample as expected. There is a small reduction in measured toxicity in the presence of latex, but this does not significantly affect the measurement of toxicity and is probably due to a small buffering effect on the organism. A shorter time delay could be used to reduce/eliminate this effect.

It should also be noted that although the presence of latex causes an increasing reduction in raw light reading between 0 and 2 minutes, the time delay method can still be used successfully even when the amount of signal available for measurement of toxicity is only 20% of the original. That is, when the interfering effect accounts for 80% of the light lost between 0 and 2 minutes.

EXAMPLE 7 Compensation for the effects of colour

The use of a time delay should allow any absorption of signal due to colour in the sample to occur before toxicity is measured. This experiment is designed to show that a standard concentration of zinc produces the same toxic response in the reagent over a range of different interfering colour concentrations, provided that the time delay method is used.

Method

0.2% Phenol red solution (Sigma) was diluted 1:10, 1:1000, 1:10,000 and the transmission of light at 490 nm measured. Zinc sulphate was added to all of the solutions and to a control containing no phenol red (i.e. water), to a final concentration of 0.1 ppm and mixed thoroughly.

Eight vials of test reagent was reconstituted as described in Example 5 above and pooled, then distributed into { ml aliquots. All of the phenol red solutions were tested in triplicate by measuring the light reading in a \ ml aliquot of test reagent BEFORE addition of the coloured solution, and then at 2 and 5 minutes after the addition of ml of the coloured solution. Three control tubes (water with no zinc sulphate or phenol red) were also used.

The toxic response of the zinc sulphate in the tubes was assessed using the formula described above, using either 0 mins (i.e. before addition of coloured solution) or 2 minutes as the base light reading.

Results are shown in the two Tables below and in Figure 9.

Results

Transmission of light at 490 nm

Dilution of % Transmission phenol red at 490 nm

1 0.2

1 : 10 4.4

1 : 100 78.9

1 : 1000 98.3

1 : 10000 99.7

Water (no 100 phenol red)

DILUTION LIGHT READINGS % Toxic effect OF PHENOL RED T₊₀ mins ₊2 mins T₊5 mins 0-5 mins 2-5 mins

Control 2176.0* 1122.0* 1061.0*

0 (No 2772.0 1091.0 640.1 52.64 37.96 colour)

2694.0 1109.0 661.9 49.61 36.88

1914.0 1258.0 836.9 10.32 29.65

1:10.000 1763.0 970.8 603.2 29.83 34.29

1759.0 998.2 630.3 26.51 33.23

1794.0 1008.0 667.4 23.74 29.98

1:1000 1565.0 570.8 328.0 57.02 39.23

1546.0 686.4 396.5 47.44 38.91

1693.0 723.1 443.6 46.26 35.13

1:10 1844.0 180.33 121.9 86.44 28.50

2019.0 211.4 136.0 86.19 31.97

1853.0 205.4 148.6 83.55 23.49

* Mean of three, control contains no zinc sulphate or phenol red Conclusions and Discussion

As with the turbidity experiment described in Example 6, the level of toxicity in each concentration of phenol red should be the same, producing a flat response over the concentrations of colour tested. Clearly, the time delay method produces just such a response, whereas the standard 0-5 minute method indicates an increasing level of toxicity which is colour concentration dependent. In this instance, it is not possible to state that the phenol red is non-toxic, so that some of the increases in measured toxicity may be due to a true difference in the toxicity of the sample. However, no difference in toxicity is seen in the time delay method even though the concentration of colour increased by two logs, and the increase in measured toxicity in the standard method over the same concentration change is only some 35%, indicating that any toxicity attributable to the phenol red is insignificant, unless the toxicity is unusually rapid, occurring within two minutes of the addition of sample. Again, any toxicity due to the phenol red could be assessed by using a shorter time delay.

Claims

1. A method of assaying an analyte, by the use of a liquid suspension of signal-generating micro¬ organisms, which comprises mixing together a liquid sample possibly containing the analyte and an aliquot of the liquid suspension to form a test mixture and thereafter observing a signal generated by the micro¬ organisms, wherein the size of the signal generated by the micro-organisms in the test mixture changes at a first rate during a first period measured from the time of formation of the test mixture, and then changes at a second rate during a second period due to the effect of analyte, and observations of the size of the signal are made at different times during the second period and are compared as an assay for analyte.

2. A method as claimed in claim 1, wherein a control mixture is also made up, comprising a control sample and an aliquot of the signal-generating micro¬ organisms, and the signal generated by the micro- organisms in the test mixture is compared to the signal generated by the micro-organisms in the control mixture during the second period as an assay for analyte.

3. A method as claimed in claim 1 or claim 2, wherein the first period is from 10 seconds to 5 minutes.

4. A method as claimed in any one of claims 1 to

3, wherein determining the duration of the first period, and the time for starting to observe the signal as an assay for analyte, are controlled by a computer.

5. A method as claimed in any one of claims 1 to

4, wherein the volume of the liquid sample is not more than ten times the volume of the aliquot of the liquid suspension of signal-generating micro-organisms.

6. A method as claimed in any one of claims 1 to 5, wherein the signal-generating micro-organisms are bacteria and the liquid suspension is formed by reconstituting lyophilised bacteria using a reconstitution buffer containing an osmotically-potent non-salt compound.

7. A method as claimed in any one of claims 1 to 6, wherein the liquid sample is used, optionally after dilution but without the addition of any chemical agent.

8. A method as claimed in any one of claims 1 to

7, wherein the signal-generating micro-organisms are genetically engineered bioluminescent bacteria.

9. A method as claimed in any one of claims 1 to

8, wherein the assay is performed to determine the toxic effect of a sample defined by a formula

% Toxic Effect = 1 - x 100

where

T_χ is the signal x minutes after addition of a toxic sample to a test tube,

C_χ is the signal x minutes after addition of a control sample to a control tube, Cy is the signal y minutes after addition of the control sample to the control tube, x and y are times chosen within the said second period and x is greater than y.

10. A method as claimed in any one of claims 1 to 9, wherein the size of the signal generated by the micro-organisms in the test mixture changes at the first rate during the first period due to environmental effects.

11. A method as claimed in any one of claims 1 to 9, wherein the size of the signal generated by the micro-organisms in the test mixture changes at the first rate during the first period due to the presence in the liquid sample of a fast-acting analyte.

12. A method as claimed in any one of claims 1 to 11, wherein the analyte is a toxicant.

13. A method of analysing first and second analytes which have different kinetics, by the use of a liquid suspension of signal-generating micro-organisms, which comprises mixing together a liquid sample possibly containing the first and/or the second analyte, and an aliquot of the liquid suspension to form a test mixture, and thereafter observing a signal generated by the micro-organisms, wherein the size of the signal generated by the micro-organisms in the test mixture changes at a first rate during a first period due to the effect of the first analyte, and then changes at a second rate during a second period due to the effect of the second analyte, and observations of the size of the signal are made during both periods as an assay for both analytes.