WO2003023375A2

WO2003023375A2 - Toxicity monitoring

Info

Publication number: WO2003023375A2
Application number: PCT/GB2002/003964
Authority: WO
Inventors: Peter Robert Fielden; Nicholas John Goddard; Andrew William Knight; Richard Maurice Walmsley
Original assignee: Gentronix Limited
Priority date: 2001-09-13
Filing date: 2002-08-30
Publication date: 2003-03-20
Also published as: WO2003023375A3; GB0122219D0; AU2002321564A1

Abstract

A toxicity monitoring apparatus comprising a light emitting diode (7) for illuminating a sample, a first detector (13) arranged to defect fluorescent light emitted by the sample, and a second detector (19) arranged to detect light scattered by the sample, thereby allowing determination of the cell density of the sample and normalisation of the measured fluorescence.

Description

TOXICITY MONITORING

The present invention relates to toxicity monitoring, and particularly though not exclusively to monitoring fluorescence emitted by green fluorescent protein (GFP).

It has been known for several years that protein expression in a variety of cells may be monitored by using GFP as a marker.

Monitoring the presence of GFP has in the past been carried out by exciting the protein by using ultraviolet light or blue light, for example as generated by a laser, and detecting fluorescent light emitted by the GFP. Typically this is done in a laboratory using techniques, commercially available and otherwise, which are often complicated and expensive.

In recent years light emitting diodes (LED's) have been developed which emit blue light at high intensities. This light has been used to excite GFP fluorescence. Blue LED's are much cheaper to manufacture and use than ultraviolet or blue lasers. In one known apparatus [1] a blue LED generates light which passes via a bandpass filter to a fibre-optic bundle, via the fibre-optic bundle to a sample, and via a second fibre-optic bundle through an emission filter to a photomultiplier tube (PMT). A disadvantage of this apparatus is that it provides a total fluorescent measurement only, rather than a 'per cell' normalised measurement.

In a second known apparatus [2] a blue LED is used to detect GFP expressed by E.coli. during fermentation processes. The apparatus uses a set of filters to limit the wavelengths of light incident upon the sample, and to minimise the incidence of scattered light on the fluorescence detector. The apparatus provides a total fluorescent measurement only, and does not provide a 'per cell' normalised measurement. The total fluorescence measurement suffers from the disadvantage that it does not take account of the density to which the cells are able to grow during the assay. Where a sample is toxic (either cytotoxic or genotoxic) the density of cells is determined by the toxicity. The term 'cytotoxic' is used to describe substances which are toxic to cells by any mechanism which disrupts normal cellular activity and thus inhibits growth or decreases viability. The term 'genotoxic' is used to describe substances which are capable of specifically damaging genetic material in the cell, principally DNA, and may also be referred to as mutagenic. Such substances may be carcinogenic or teratogenic in higher organisms.

The overall fluorescence reduces as a function of the density of the cells in the sample, which is a function of the toxicity of the sample. Normalisation is needed to correct for the reduced fluorescence signal caused by sample toxicity.

It is an object of the present invention to provide toxicity monitoring which allows a normalised fluorescence measurement to be determined.

According to the invention there is provided a toxicity monitoring apparatus comprising a light emitting diode for illuminating a sample, a first detector arranged to detect fluorescent light emitted by the sample, and a second detector arranged to detect light scattered by the sample, thereby allowing determination of the cell density of the sample and normalisation of the measured fluorescence.

Suitably, the sample comprises an aqueous solution or suspension, or pure aliquot of a substance to be tested, combined with toxicity test organisms or cells.

The cell density of the sample provides an indication of the cyto-toxicity of the tested substance. The normalised fluorescence provides an indication of the genotoxicity of the tested substance. Suitably, the apparatus is provided with means for dividing the detected fluorescent light by the detected scattered light to normalise the measured fluorescence.

Suitably, the sample comprises genetically modified yeast cells which express fluorescent protein.

Suitably, the fluorescent protein is green fluorescent protein.

Suitably, the fluorescent light and the scattered light are detected substantially simultaneously.

Suitably, the sample is illuminated by more than one blue LED.

Suitably, an interference filter is used to select a required band of wavelengths of light for illuminating the sample.

Suitably, a short pass filter is used to suppress the transmission of light above a required wavelength into the sample.

Suitably, an interference filter is used to select a required band of wavelengths which for transmission to the fluorescent light detector.

Suitably, a short wave cut-off filter is used to restrict transmission to the photomultiplier tube of light having a wavelength below a required value.

Suitably, the fluorescent light detector is oriented in a direction which is substantially perpendicular to the direction in which illumination from the light emitting diode enters the sample.

Suitably, the fluorescent light detector is a photomultiplier tube. Suitably, an interference filter is used to select a required band of wavelengths of light for transmission to the scattered light detector.

Suitably, the scattered light detector is oriented in a direction which is substantially perpendicular to the direction in which illumination from the light emitting diode enters the sample.

Suitably, the scattered light detector is a photodiode.

Suitably, the apparatus includes a housing provided with an opening dimensioned to receive a sample held in a suitable container.

Suitably, the container is a cuvette.

Suitably, the container is a flow cell.

Suitably, the light emitting diode is a blue light emitting diode.

Suitably, the light emitting diode emits ultra-violet light.

Suitably, the sample is illuminated with light having a first plane polarisation, and is subsequently illuminated by light having a second substantially perpendicular plane polarisation, and the fluorescent light detector is arranged to detect only light having the first plane polarisation.

Suitably, light having the first plane polarisation is obtained by locating a polarising filter between a first LED and the sample, light having the second plane polarisation is obtained by locating a second polarising filter between a second LED and the sample, and light having the first plane polarisation is selected for transmission to the fluorescent light detector by a polarising filter located between the detector and the sample. The invention also provides a method of monitoring toxicity, the method comprising illuminating a sample using a blue light emitting diode, detecting fluorescent light emitted by the sample using a first detector, detecting light scattered by the sample using a second detector, and normalising the detected fluorescence by dividing the detected fluorescence by the detected scattered light to provide a brightness measurement indicative of the genotoxicity of the sample.

Suitably, measured scattered light from the sample is compared to the measured scattered light from a non-cytotoxic control to provide a measurement of the cytotoxicity of the sample. The comparison may be based upon a ratio, subtraction, or any other suitable method.

Suitably, the method further comprises determining a threshold level of measured scattered light based upon the measured scattered light from the non- cytotoxic control, and determining that the sample is cytotoxic if the measured scattered light from the sample is less than the threshold level.

Suitably, the sample comprises a substance to be monitored, growth medium, and yeast cells. The term growth medium is not intended to exclude a mixture of more than one growth medium.

Suitably, the non-cytotoxic control comprises growth medium and yeast cells, or growth medium and yeast cells diluted with a non-toxic liquid.

Suitably, the method further comprises subtracting measured light scattered from a scattering control containing the substance to be monitored without yeast cells, to correct for scattered light interference arising from particulate matter in the substance to be monitored.

Suitably, the scattering control further comprises the growth medium. Adding the growth medium has the advantage that it makes the scattering control resemble the sample more closely. This may be useful for example if microbial contamination is contained in a sample, since the micro-organisms will grow if the growth medium is present.

Suitably, a series of measurements of scattered light are performed for a series of scattering controls, each having different concentrations of the substance to be monitored, and are subtracted from the measurements of scattered light for a series of samples having the corresponding concentrations of the substance to be monitored.

Suitably, the method further comprises measuring scattered light for a series of samples each containing different concentrations of the substance to be monitored, defining a relationship between the measured scattered light and the concentration of the substance, and using the defined relationship to determine the concentration of the substance that would provide a given proportion of the scattered light detected for the non-cytotoxic control.

Suitably, the sample comprises a substance to be monitored, a growth medium and genetically modified yeast cells which express fluorescent protein.

Suitably, the fluorescent protein is green fluorescent protein.

Suitably, the method further comprises comparing the measured brightness of the sample with the measured brightness of a non-genotoxic control to provide a measurement of the genotoxicity of the sample.

Suitably, the method further comprises comparing the measured brightness of the sample with a previously recorded level of brightness to provide a measurement of the genotoxicity of the sample. The previously recorded level of brightness may be a previously measured brightness of a non-genotoxic control.

Suitably, the non-genotoxic control comprises the growth medium and the genetically modified yeast cells, or the growth medium and the genetically modified yeast cells diluted with a non-toxic liquid. Suitably, the measured fluorescence of an autofluorescence control containing the substance to be monitored, without yeast cells, is subtracted from the measured fluorescence of the sample, to correct for fluorescence interference arising from autofluorescent matter in the substance to be tested.

Suitably, the autofluorescence control further comprises the growth medium.

Suitably, the measured brightness of a yeast-autofluorescence control containing the substance to be tested, growth medium and genetically modified yeast cells which cannot express the fluorescent protein, is subtracted from the measured brightness of the sample, to correct for non-specific auto-fluorescence produced by the yeast cells upon exposure to the substance to be tested. The correction for yeast- autofluorescence is performed using the measured brightness rather than for the measured fluorescence since the yeast-autofluorescence control needs to account for differences in final cell density which may arise due to differences between the yeast cell strains.

Suitably, the method is performed for samples having a range of concentrations of the substance to be monitored.

Suitably, the measured fluorescence of a range of autofluorescence controls containing a range of concentrations of the substance to be monitored, is subtracted from the measured fluorescence of the range of samples.

Suitably, the measured brightness of a range of yeast-autofluorescence controls containing a range of concentrations of the substance to be monitored, is subtracted from the measured brightness of the range of samples.

Suitably, the method further comprises determining a threshold level of brightness based upon the measured brightness of the non-genotoxic control, and determining that the sample is genotoxic if the measured brightness of the sample is greater than the threshold level.

Suitably, the method further comprises comparing the measured brightness of the sample with the measured brightness of a sample having a known genotoxicity value, to determine a calibrated genotoxicity value for the sample.

Suitably, the cytotoxicity and the genotoxicity are measured for the same sample or samples.

Suitably, a single control is used as the scattering control and the autofluorescence control.

Suitably, a single control is used as the non-cytotoxic control and the non- genotoxic control.

Suitably, zero values of the fluorescent light and the scattered light are determined using a zero signal control sample comprising fluid with low scattering and low autofluorescence.

The method may incorporate any other suitable elements of the invention.

A specific embodiment of the invention will now be described by way of example only, with reference to the accompanying figures in which:

Figure 1 is a schematic illustration of a monitoring apparatus which embodies the invention;

Figure 2 is graph representing transmission curves for filters of the monitoring apparatus shown in figure 1; Figure 3 is a graph showing results obtained using the apparatus shown in figure 1;

Figure 4 is a second graph showing results obtained using the apparatus shown in figure 1;

Figure 5 is a third graph showing results obtained using the apparatus shown in figure 1 ;

Figure 6 is a fourth graph showing results obtained using the apparatus shown in figure 1;

Figure 7 is a schematic illustration of a second monitoring apparatus which embodies the invention;

Figure 8 is a fifth graph showing results obtained using the apparatus shown in figure 1; and

Figure 9 is a schematic illustration of a flow cell comprising part of the first or second embodiment of the invention.

Referring to Figure 1, a monitoring apparatus comprises a body 1 defining a chamber 2 which receives a cuvette 3. Three sides of the body 1 are provided with openings which allow the transmission of light between the chamber 2 and optical components located outside of the chamber 2.

An optical mounting tube 4 is held in a recess 5 located behind a first opening 6 in the body 1. A blue light emitting diode (LED) 7 is held in the optical mounting tube. The blue LED is a Nichia ultra-bright blue LED having a peak emission wavelength of 470nm. A short-pass filter 8 and a band-pass filter 9 are held in the optical mounting tube between the blue LED and the opening 6. The short-pass filter is BG3 blue glass, and the band-pass filter is a 475RDF40 filter.

A second optical mounting tube 10 is held in a recess 11 located behind the second opening 12. A photomultiplier tube 13 (PMT) is mounted in the optical mounting tube. The PMT is a Hamamatsu H5784 and is connected to a high voltage supply and an amplifier (not shown). A short-wave cut-off filter 14 and an interference filter 15 are located between the PMT 13 and the opening 12. The short- wave cut-off filter 14 is a 515nm OG515 Schott Glass (9mm) filter, and the interference filter 15 is centred on 515nm and has a lOnm half band- width.

A third optical mounting tube 16 is held in a recess 17 located behind the third opening 18. A silicon photodiode 19 is mounted in the optical mounting tube. The silicon photodiode is provided with an integral amplifier (not shown). A neutral density filter 20 and an interference filter 21 are mounted in the optical mounting tube between the silicon photodiode 19 and the opening 18. The neutral density filter is OD=l, and the interference filter is centred on 470nm and has a lOnm half bandwidth.

For optimum sensitivity the three optical tubes 4, 10, 16 are mounted such that both fluorescence and light scatter are measured at right angles (perpendicular) to the direction of illumination of the LED 7. It will be appreciated that it would be possible to arrange the optical tubes at other angles and still obtain reasonably good results.

In use a sample containing a substance to be tested is prepared and placed in the cuvette 3. The LED 7 is used to illuminate the sample, and the PMT 13 and silicon photodiode 19 are used to detect, respectively, fluorescent light emitted from the sample and light scattered by the sample.

The test 'reagent' is a suspension of a genetically modified yeast that expresses a green fluorescent protein in response to DNA damage repair. Upon exposure to a substance which is genotoxic the yeast cells become increasingly fluorescent. The genetically modified yeast is described in WO98/44149. The test reagent is produced by growing up a frozen or freeze dried stock standard, in medium, for more than 16 hours, or overnight. The reagent may be used for approximately 1 week. A volume of reagent is placed into a disposable plastic cuvette 3, and fresh growth medium is added along with a volume of a substance to be tested (the term growth medium is not intended to exclude a mixture of more than one growth medium). The resulting sample is then left to incubate for a period of 4 to 16 hours. The sample cuvette does not require special incubation conditions as the test works over a wide temperature range. The sample cuvette does not need to be agitated during incubation. After the incubation period the cuvette 3 is shaken to re-suspend settled cells and is placed within the chamber 2 of the apparatus.

Measurement is made by pushing a button (not shown) on the apparatus, which provides power to the LED thereby illuminating the LED. Fluorescent light emitted by the sample is detected by the PMT 13.

Upon exposure to a genotoxic agent the modified yeast sustains DNA damage. As they carry out repair of this damage, the cells become increasingly fluorescent due to the expression of GFP. Thus a fluorescence measurement allows assessment of the presence, concentration or potency of genotoxic agents in the sample.

Simultaneously with the detection of the excitation light, the light scattered by the sample is detected using silicon photodiode 19. The measurement of scattered light gives a nephelometric estimation of cell density. If cytotoxic species are present in the sample they are likely to disrupt cellular activity or cause damage to the cell, requiring the cell to expend energy on repairing damage that might otherwise have been directed towards growth and replication. In either case if the sample is cytotoxic the cells will grow to a lower density compared to a non-cytotoxic control such as water. Thus measurement of cell density gives a measure of the general cytotoxicity of the sample.

Since the cytotoxicity of the sample limits the cell density, the measured fluorescent light needs to normalised by dividing the fluorescence signal by the cell density as indicated by the measured scattered light, otherwise an underestimation of fluorescence and genotoxicity results. This provides a "fluorescence per cell" or "brightness" measurement which gives an accurate assessment of the genotoxicity of the sample. For example, 1000 weakly fluorescing cells (not a very cytotoxic environment), might produce the same total fluorescence as 100 brightly fluorescing cells (a very cytotoxic environment). These two extremes would be readily distinguished by the normalisation described. The measured scattered light and the measured fluorescence are output as voltages to an interface (not shown). The interface carries out the normalisation to provide a genotoxicity value which, together with the measured scattered light, is represented as a digital signal. The interface allows data acquisition and processing using a lap-top (or other) computer. The genotoxicity value and the measured scattered light may also be displayed on a digital readout (not shown).

The measured scattered light is compared with a non-toxic control to provide a measurement of the cytotoxicity of the sample. The control is illuminated by the apparatus as described above, and light scattered by the control is detected. Only a low level of fluorescent light is emitted by the non-toxic control, which is due to GFP expressed as a consequence of natural background DNA repair activities taking place in cells. The measured scattered light indicates the density of cells in the control. Comparing this with the measured scattered light from the sample provides a measurement of the cytotoxicity of the sample.

In order to correct for any effects the presence of the GFP reporter plasmid may have on cellular fluorescence and the extent of proliferation, a control strain is used, the control strain being produced by constructing a control plasmid and transforming it into yeast cells. The control plasmid is produced by subtly modifying the original GFP reporter plasmid such that it is unable to express yEGFP, despite being identical to the reporter plasmid in every other way. In assays, measurements of fluorescence and scattered light from the test strain can be accurately corrected for induced autofluorescence, or changes in growth rate due to the presence of the reporter plasmid, by comparison to the results from the control strain.

Standard chemicals may be used for calibration and to confirm that the yeast reagents are working as expected. Known genotoxic chemicals may be used to induce GFP expression in the yeast cells. A good example is methyl methanesulphonate, a genotoxic alkylating agent. Known cytotoxic chemicals may be used to demonstrate the limiting of cell proliferation and thus cytotoxicity. A well characterised example is 3,5-dichlorophenol.

A significant advantage of the apparatus is that since the scattering and fluorescence measurements are taken simultaneously, the normalisation is particularly accurate.

The operation of the filters shown in figure 1 will now be described with reference to figure 2. The luminosity spectrum of the LED 7 is shown by line 107. The fluorescence excitation spectrum for Yeast Enhanced (yE) green fluorescent protein is shown by curve 100, and the fluorescence emission spectrum for yE green fluorescent protein is shown by curve 101.

Of key importance is that light scattered by the sample should be detected separately from fluorescent light emitted by the sample. This means that the wavelengths of light detected by the photomultiplier tube should as far as possible not overlap with the wavelengths of light used to illuminated the sample. Similarly, the wavelengths of light detected by the silicon photodiode should as far as possible not overlap with the fluorescent light emitted by the sample.

It can be seen from figure 2 that there is a considerable overlap between the luminosity spectrum 107 of the LED 7 and the fluorescence emission spectrum 101. The interference filter 9 is used to limit the wavelength of illumination provided by the LED 7, as can be seen from the spectral curve 109 of the filter. Theoretically the interference filter 9 should prevent any overlap between light from the LED and the fluorescence emission spectrum. However, in practice the effectiveness of the interference filter is limited by its mode of operation, which is by providing destructive interference for wavelengths outside of the desired range. The interference filter is designed to filter light which is incident upon it from a perpendicular direction. Light which is not incident from a peφendicular direction will not experience 100% destructive interference, and some of that light will pass through the filter. Since the LED 7 provides diffuse light rather than collimated light, some light which falls outside of the spectral curve 109 of the interference filter 9 will be transmitted by the interference filter 9. For this reason, the BG3 short-pass filter 8 is used in combination with the interference filter 9. BG3 is a coloured glass filter and thus effectively filters for all orientations of incident light. The spectral curve 108 of the BG3 shortpass filter 8 is shown in figure 2. The short-pass filter 8 combines with the interference filter 9 to allow transmission of light from the LED as shown by shaded area A. Any longer wavelength light which is not stopped by the interference filter 9 is strongly suppressed by the short-pass filter 8. It can be seen that the transmission curve 108 of the short-pass filter overlaps only very slightly with the fluorescence emission spectrum 32 for yE GFP. The overlap is sufficiently small that it does not significantly affect fluorescence measurements.

In addition to ensuring that only desired wavelengths of light are passed into the sample, filters are used to ensure that the photomultiplier tube (PMT) 13 and the silicon photodiode 19 only detect wavelengths of interest. As described in relation to Figure 1, an interference filter 15 and a short wave cut-off filter 14 are located in front of the PMT 13. The spectral curve 115 of the interference filter 15 is centred on 515nm and has a lOnm half bandwidth. The interference filter 15 is augmented with the short wave cut-off filter 14 which has a spectral curve 114. The wavelengths of light transmitted by the filters to the photomultiplier tube 13 are shown by the shaded area B. The short wave cut-off filter 14 is used to prevent short wavelength diffuse light which is not perpendicularly incident upon the interference filter being transmitted to the photomultiplier tube 13.

As described in relation to Figure 1, an interference filter 21 and a neutral density filter 20 are located in front of the silicon photodiode 19. The spectral curve of the interference filter 21 which is centred on 470nm and has a lOnm half bandwidth is not represented in figure 2. The interference filter allows the transmission of scattered light to the silicon photodiode 19. A long wave cut-off filter is not required for the silicon photodiode 19 for two reasons. The first reason is that the intensity of scattered light is much greater than the intensity of fluorescent light, so that any fluorescent light which passes through the interference filter 21 will be of such low intensity that it is unlikely to significantly affect the measurement of scattered light. The second reason is that the interference filter, by virtue of its inherent properties, will not transmit light which has a wavelength longer than the spectral curve of the filter. The neutral density filter (OD = 1) 20 is used to reduce the intensity of light incident on the silicon photodiode to avoid saturation of the diode.

The apparatus is powered by four 9-volt PP3 batteries, which may be rechargeable and which, in combination with associated electronic circuitry, provide a supply voltage of + 12 N. This is sufficient to power all electrical components including the LED 7, photomultiplier tube 13 and silicon photodiode 19. Alternatively, the apparatus may be connected to an external power source.

Performance and calibration data is shown in Figures 3 and 4. Referring to figure 3, a range of MMS concentrations (a known genotoxin) were tested, each in duplicate using two separate cuvette cultures. The average 'brightness' (i.e. fluorescence divided by cell density) is plotted as a function of MMS concentration. The brightness is seen to increase with genotoxicity.

Figure 4 shows the final cell density (i.e. detected scattered light) achieved in the range of MMS concentrations tested. This measurement gives a measure of general cytotoxicity. It can be seen that the final cell density decreases as the general cytotoxicity of the sample increases.

The apparatus can be used to determine a qualitative result with respect to cytotoxicity, i.e. cytotoxic or non-cytotoxic, by comparison of the final cell density achieved to a predetermined threshold. Samples which have cell densities which fall below the threshold are deemed cytotoxic. The threshold may be a fixed percentage of the value of cell density obtained for a non-cytotoxic control. The threshold may be a previously recorded value based upon a previous non-cytotoxic control light scattering measurement. The threshold may be recorded in the apparatus or in a computer connected to the apparatus via the interface, to allow automated determination of whether or not the sample is cytotoxic. The apparatus can be used to determine a quantitative result with respect to cytotoxicity such as an EC50, the effective concentration of substance to be monitored which would give a 50 % response, i.e. reduces cell proliferation such that the final cell density is half that of a non-cytotoxic control. To achieve this a series of dilutions of the substance to be monitored are tested in separate cuvettes. A parametric curve is fitted to the data describing the variation in final cell density with substance concentration, and is used to determine the substance concentration corresponding to a cell density half that of the non-cytotoxic control. An example is shown in Figure 5 for copper(II) ions, which are cytotoxic to yeast with an EC50 = 0.052 mg/L. The cell density of the non-cytotoxic control may be a previously recorded value based upon a previous non-cytotoxic control light scattering measurement.

A lowest observable effect concentration (LOEC) can be determined corresponding to the effective concentration of substance to be monitored which gives a statistically significant reduction in cell density compared to a non-cytotoxic control. For example a reduction equivalent to three times the standard deviation of cell density measurements taken from a suitable number of replicate non-cytotoxic controls.

The apparatus can be used to determine a qualitative result with respect to genotoxicity, i.e. genotoxic or non-genotoxic, by comparison of induced brightness to a predetermined threshold. Substances which produce brightness values which are above the threshold are deemed genotoxic. The threshold may be a fixed percentage of the value obtained for the brightness of a non-genotoxic control. For example, the threshold for genotoxicity may be set at a brightness value 30 % higher than that of a non-genotoxic control, corresponding to an induction of 1.3 where:

Induction = Brightness of test yeast strain exposed to the substance to be monitored Brightness of test yeast strain exposed to a non-toxic control An example is shown in Figure 6 for the pesticide Paraquat which is genotoxic in yeast. A dose dependant response gives further evidence of genotoxicity. The measured brightness of the non-genotoxic control may be a previously recorded value, which may be recorded in the apparatus or in a computer connected to the apparatus via the interface, to allow automated determination of genotoxicity.

The apparatus may be used to determine a quantitative genotoxicity measurement, i.e. the degree of genotoxicity of a given substance, by comparing the induction in brightness to that produced by a genotoxic standard (this may be prerecorded).

Correction for fluorescence or scattering arising due to unwanted fluorescent material or particulate material in samples may be used. For a given sample, a control comprising the substance to be monitored and the growth medium only, excluding yeast cells, is prepared. The control is treated in the same way as the sample with regard to incubation. Fluorescence and light scatter readings are taken using the control, and the results are subtracted from measurements of the corresponding sample. For ease of terminology, the control is termed the 'scattering control' when it is used to correct a scattering measurement, and is termed the 'autofluorescence control' when it is used to correct a fluorescence measurement. It will be appreciated that in the majority of cases a single control is used to correct both measurements. In this way background autofluorescence, and light scatter due to particulate material, inherent in the sample are corrected for, reducing their interference in the assessment of cytotoxicity and genotoxicity. This is particularly relevant in the analysis of aqueous environmental samples which often contain contaminants, such as microbial or algal contamination. This is because the growth medium may cause cells arising from the contamination to grow in numbers, producing a significant cell density reading which in the absence of a scattering control measurement may be mistakenly presumed to be due to the growth of yeast cells.

The correction of detected scattered light using the scattering control (equivalently the correction of cell density) can be expressed as: Corrected Cell Density = D_T - Ds

Where D_T is the measured scattering from the sample, and Ds is the measured scattering from the scattering control.

Similarly, the correction of detected fluorescence using the autofluorescence control can be expressed as:

Corrected Fluorescence = F_T - Fs

Where F is the measured fluorescence from the sample, and Fs is the measured fluorescence from the autofluorescence control.

The corrected measured brightness maybe expressed as:

F -F

Corrected measured brightness = — —

D '_τT -D,

Any autofluorescence which is induced within the yeast cell itself due to the components of the sample, is corrected for by subtracting the corrected brightness value observed for a yeast-autofluorescence control containing yeast which does not express GFP. The correction is useful for samples in which auto-fluorescence is produced by the yeast cells upon exposure to the substance to be tested. The subtraction of the measured yeast-autofluorescence signal is performed using the measured brightness rather than for the measured fluorescence since the yeast- autofluorescence control needs to account for differences in final cell density which may arise due to differences between the yeast cell strains.

The yeast-autofluorescence control correction, when combined with the scattering control and autofluorescence controls can be expressed as: Corrected Brightness Value = ^Fτ ^Fs _ ^Fc ^Fs (including yeast auto- D_τ - D_s D_C ^~ D_S fluorescence correction)

Where Fc and Dc respectively denote the measured fluorescence and cell density of the yeast auto-fluorescence control.

It is appreciated, that in some instances reasonably good results for genotoxicity may be obtained using the test yeast strain only, without corrections as described, especially in cases where the sample is not significantly autofluorescent or optically dense.

The fluorescence and light scatter signals may also be zeroed by using a cuvette containing a clear liquid of low autofluorescence, such as water or growth medium.

More than one LED may be used to illuminate the sample, thereby providing increased power of illumination. For example, an LED array may be used.

The photomultiplier tube 13 may be replaced by a suitable solid state detector. Similarly, the silicon photodiode 19 used for scattered light detection may be replaced by a photomultiplier detector. The choice of detectors in the illustrated embodiment of the invention simply reflects the relative intensities of the signals produced. Fluorescence is often weak, especially after strict filtering to remove scattered excitation light from the signal and thus requires a sensitive PMT. Whereas the cells scatter a significant proportion of the excitation light and this is a large signal, which is relatively easy to detect using a less sensitive silicon photodiode.

Polarising filters may be used to provide discrimination between fluorescent light emitted by GFP, and interference from other fluorescent components of the sample or cells (the interference is referred to hereafter as autofluorescence). It is known that GFP demonstrates marked fluorescence anisotropy when illuminated with plane polarised light. Thus a significant proportion of the of the emitted fluorescence remains polarised with respect to the excitation light. In contrast to this, the polarisation of autofluorescence typically shows considerably less correlation with the polarisation of light used to excite the autofluorescence (the polarisation of the autofluorescence generally has components in all directions). The difference in polarisation dependency can be used to remove background autofluorescence from GFP fluorescence, by subtracting detected fluorescent light which has a polarisation perpendicular to illumination polarisation, from fluorescent light which has a polarisation parallel to illumination polarisation. [3]

A modified apparatus which uses polarising filters in this way is shown in figure 7. The modified apparatus corresponds in large part to that shown in figure 1. Only those parts which are different or particularly pertinent to the operation of the modified apparatus are described.

The apparatus is provided with Nichia Ultra-Bright blue LED's 7, 30. A horizontally oriented polarising filter 31 is located between the right hand LED 7 (as viewed in figure 5) and the chamber 2. The horizontally oriented polarising filter 31 allows only light which is polarised in the horizontal plane to pass into the chamber 2. A vertically oriented polarising filter 32 is located between the left hand LED (as viewed in figure 5) and the chamber 2. The vertically oriented polarising filter 32 allows only light which is polarised in the vertical plane to pass into the chamber 2. A second horizontally oriented polarising filter 33 is located between the photomultiplier tube (PMT) 13 and the chamber, so that the PMT detects only light which is polarised in the horizontal plane.

In use, the right hand LED 7 is turned on first, illuminating a sample with plane polarised light to excite the GFP and autofluorescent components of the sample. Fluorescent light polarised parallel with respect to the excitation source is detected by the PMT 13 as a first value. The right hand LED 7 is turned off. The left hand LED 30 is then turned on, illuminating the sample with light polarised perpendicular (90°) to that of LED 7. Fluorescent light polarised perpendicular with respect to the excitation source is then detected by the PMT 13 as a second value. Due to the anisotropic nature of GFP fluorescence the difference between the first and second values will be a large for GFP but much smaller for the more isotropic autofluorescence. Thus the second value is subtracted from the first value, providing a measure of GFP fluorescence from which interfering autofluorescence has been substantially removed.

The apparatus has been used to determine the fluorescence intensity from GFP expressed in Chinese Hamster Ovary (CHO) cells to a limit of detection of approximately 500 cells per ml, and this is shown in figure 8. A flow cell was used in place of a cuvette to generate the results shown in figure 8. The flow cell is shown in figure 9 and comprises a chamber 50 provided with three windows 51-53. The windows 51-53 are arranged to align with the openings 6, 12, 18 shown in figure 1. A clear glass block 54 is provided at an upper end of the chamber 50. Tubes 55, 56 are held in the glass block 54, and communicate with the chamber 50 to allow a sample to flow into and out of the chamber 50.

The flow cell is advantageous because it allows the apparatus to be used to determine small volumes of sample (< 1 ml), which would otherwise not be seen in the cuvette assay. Using the flow cell also enables the instrument to be used for bench-top flow-injection-analysis (FLA) assay work.

An ultra-violet light emitting LED or LED's may be used in place of the blue LED or LED's.

The apparatus may be used to examine fluorescent proteins other than GFP. Examples include, but are not limited to, yellow fluorescent protein, cyan fluorescent protein, blue fluorescent protein and the red fluorescent proteins DsRed and DsRed2. The apparatus may be used to examine fluorescent proteins which are expressed in other cells and organisms which are sufficiently small and transparent for the purposes of genotoxicity and cytotoxicity evaluation. Examples include other micro- organisms such as bacteria, other cell types such as mammalian cells and other species such as protozoa and daphnia. An clear example is the use of the mammalian Chinese hamster ovary cells previously discussed. In each of these cases the emission wavelengths of the LED's and the wavelength specifications of the filters should be adjusted accordingly.

REFERENCES

[1] On-line green fluorescent protein sensor with LED excitation. Randers- Eichhorn L, Albano CR, Sipior J, Bentley WE, Rao G. BIOTECHNOLOGY AND BIOENGLNEERLNG 55: (6) 921-926 SEP 20 1997

[2] All solid-state GFP sensor. Kostov Y, Albano CR, Rao G BIOTECHNOLOGY AND BIOENGINEERING 70: (4) 473-477 NOV 20 2000.

[3] Knight A.W., Goddard N.J., Fielden P.R., Gregson A.L., Billinton N., Barker M.G. and Walmsley R.M. 2000. The application of fluorescence polarisation for the enhanced detection of green fluorescent protein (GFP) in the presence of cellular auto-fluorescence and other green fluorescent compounds. ANALYST. 125: 499- 506.

Claims

1. A toxicity monitoring apparatus comprising a light emitting diode for illuminating a sample, a first detector arranged to detect fluorescent light emitted by the sample, and a second detector arranged to detect light scattered by the sample, thereby allowing determination of the cell density of the sample and normalisation of the measured fluorescence.

2. An apparatus according to claim 1, wherein the apparatus is provided with means for dividing the detected fluorescent light by the detected scattered light to normalise the measured fluorescence.

3. An apparatus according to claim 1 or claim 2, wherein the sample comprises genetically modified yeast cells which express fluorescent protein.

4. An apparatus according to claim 3, wherein the fluorescent protein is green fluorescent protein.

5. An apparatus according to any preceding claim, wherein the fluorescent light and the scattered light are detected substantially simultaneously.

6. An apparatus according to any preceding claim, wherein the sample is illuminated by more than one blue LED.

7. An apparatus according to any preceding claim, wherein an interference filter is used to select a required band of wavelengths of light for illuminating the sample.

8. An apparatus according to claim 7, wherein a short pass filter is used to suppress the transmission of light above a required wavelength into the sample.

9. An apparatus according to any preceding claim, wherein an interference filter is used to select a required band of wavelengths which for transmission to the fluorescent light detector.

10. An apparatus according to claim 9, wherein a short wave cut-off filter is used to restrict transmission to the photomultiplier tube of light having a wavelength below a required value.

11. An apparatus according to any preceding claim, wherein the fluorescent light detector is oriented in a direction which is substantially perpendicular to the direction in which illumination from the light emitting diode enters the sample.

12. An apparatus according to any preceding claim, wherein the fluorescent light detector is a photomultiplier tube.

13. An apparatus according to any preceding claim, wherein an interference filter is used to select a required band of wavelengths of light for transmission to the scattered light detector.

14. An apparatus according to any preceding claim, wherein the scattered light detector is oriented in a direction which is substantially perpendicular to the direction in which illumination from the light emitting diode enters the sample.

15. An apparatus according to any preceding claim, wherein the scattered light detector is a photodiode.

16. An apparatus according to any preceding claim, wherein the apparatus includes a housing provided with an opening dimensioned to receive a sample held in a suitable container.

17. An apparatus according to claim 16, wherein the container is a cuvette.

18. An apparatus according to claim 16, wherein the container is a flow cell.

19. An apparatus according to any preceding claim, wherein the light emitting diode is a blue light emitting diode.

20. An apparatus according to any of claims 1 to 19, wherein the light emitting diode emits ultra-violet light.

21. An apparatus according to any preceding claim, wherein the sample is illuminated with light having a first plane polarisation, and is subsequently illuminated by light having a second substantially perpendicular plane polarisation, and the fluorescent light detector is arranged to detect only light having the first plane polarisation.

22. An apparatus according to claim 21, wherein light having the first plane polarisation is obtained by locating a polarising filter between a first LED and the sample, light having the second plane polarisation is obtained by locating a second polarising filter between a second LED and the sample, and light having the first plane polarisation is selected for transmission to the fluorescent light detector by a polarising filter located between the detector and the sample.

23. An apparatus according to claim 1, wherein the sample comprises an aqueous solution or suspension, or pure aliquot of a substance to be tested, combined with toxicity test organisms or cells.

24. A method of monitoring toxicity, the method comprising illuminating a sample using a blue light emitting diode, detecting fluorescent light emitted by the sample using a first detector, detecting light scattered by the sample using a second detector, and normalising the detected fluorescence by dividing the detected fluorescence by the detected scattered light to provide a brightness measurement indicative of the genotoxicity of the sample.

25. A method according to claim 24, wherein measured scattered light from the sample is compared to the measured scattered light from a non-cytotoxic control to provide a measurement of the cytotoxicity of the sample.

26. A method according to claim 25, wherein the method further comprises determining a threshold level of measured scattered light based upon the measured scattered light from the non-cytotoxic control, and determining that the sample is cytotoxic if the measured scattered light from the sample is less than the threshold level.

27. A method according to any of claims 24 to 26, wherein the sample comprises a substance to be monitored, growth medium, and yeast cells.

28. A method according to claim 27, wherein the non-cytotoxic control comprises growth medium and yeast cells, or growth medium and yeast cells diluted with a non- toxic liquid.

29. A method according to claim 27 or claim 28, wherein the method further comprises subtracting measured light scattered from a scattering control containing the substance to be monitored without yeast cells, to correct for scattered light interference arising from particulate matter in the substance to be monitored.

30 A method according to claim 29, wherein the scattering control further comprises the growth medium. Adding the growth medium has the advantage that it makes the scattering control resemble the sample more closely.

31. A method according to claim 29 or 30, wherein a series of measurements of scattered light are performed for a series of scattering controls, each having different concentrations of the substance to be monitored, and are subtracted from the measurements of scattered light for a series of samples having the corresponding concentrations of the substance to be monitored.

32. A method according to any of claims 28 to 31, wherein the method further comprises measuring scattered light for a series of samples each containing different concentrations of the substance to be monitored, defining a relationship between the measured scattered light and the concentration of the substance, and using the defined relationship to determine the concentration of the substance that would provide a given proportion of the scattered light detected for the non-cytotoxic control.

33. A method according to claim 24, wherein the sample comprises a substance to be monitored, a growth medium and genetically modified yeast cells which express fluorescent protein.

34. A method according to claim 33, wherein the fluorescent protein is green fluorescent protein.

35. A method according to claim 33 or claim 34, wherein the method further comprises comparing the measured brightness of the sample with the measured brightness of a non-genotoxic control to provide a measurement of the genotoxicity of the sample.

36. A method according to claim 33 or claim 34, wherein the method further comprises comparing the measured brightness of the sample with a previously recorded level of brightness to provide a measurement of the genotoxicity of the sample.

37. A method according to claim 35 or claim 36, wherein the non-genotoxic control comprises the growth medium and the genetically modified yeast cells, or the growth medium and the genetically modified yeast cells diluted with a non-toxic liquid.

38. A method according to any of claims 33 to 37, wherein the measured fluorescence of an autofluorescence control containing the substance to be momtored, without yeast cells, is subtracted from the measured fluorescence of the sample, to correct for fluorescence interference arising from autofluorescent matter in the substance to be tested.

39. A method according to claim 38, wherein the autofluorescence control further comprises the growth medium.

40. A method according to any of claims 33 to 39, wherein the measured brightness of a yeast-autofluorescence control containing the substance to be tested, growth medium and genetically modified yeast cells which cannot express the fluorescent protein, is subtracted from the measured brightness of the sample, to correct for non-specific auto-fluorescence produced by the yeast cells upon exposure to the substance to be tested.

41. A method according to any of claims 33 to 40, wherein the method is performed for samples having a range of concentrations of the substance to be monitored.

42. A method according to claim 38 and to claim 41, wherein the measured fluorescence of a range of autofluorescence controls containing a range of concentrations of the substance to be monitored, is subtracted from the measured fluorescence of the range of samples.

43. A method according to claim 40 and to claim 41, wherein the measured brightness of a range of yeast-autofluorescence controls containing a range of concentrations of the substance to be monitored, is subtracted from the measured brightness of the range of samples.

44. A method according to any of claims 35 to 43, wherein the method further comprises determining a threshold level of brightness based upon the measured brightness of the non-genotoxic control, and determining that the sample is genotoxic if the measured brightness of the sample is greater than the threshold level.

45. A method according to claim 24 or any of claims 33 to 44, wherein the method further comprises comparing the measured brightness of the sample with the measured brightness of a sample having a known genotoxicity value, to determine a calibrated genotoxicity value for the sample.

46. A method according to any of claims 25 to 45, wherein the cytotoxicity and the genotoxicity are measured for the same sample or samples.

47. A method according to claim 29 and claim 38, wherein a single control is used as the scattering control and the autofluorescence confrol.

48. A method according to claim 25 and claim 35, wherein a single confrol is used as the non-cytotoxic control and the non-genotoxic control.

49. A method according to any of claims 24 to 48, wherein the fluorescent light and the scattered light are detected substantially simultaneously.

50. A method according to any of claims 24 to 49, wherein zero values of the fluorescent light and the scattered light are determined using a zero signal control sample comprising fluid with low scattering and low autofluorescence.

51. A toxicity monitoring apparatus substantially as hereinbefore described with reference to the accompanying figures.

52. A method of monitoring toxicity substantially as hereinbefore described with reference to the accompanying figures.