The measurement of reduced nicotinamide adenine dinucleotide (NADH) is of importance in several bioanalytical applications. The amount of NADH which is produced or consumed during a cofactor-dependent enzymatic reaction forms the basis of many im unoassays and biosensors. Further, miniaturised fluorescent probes, based on the measurement of the intrinsic fluorescence of NADH (at about 460mm) or on a chemiluminescent reaction involving NADH as one of the reactants, have been used to study enzymatic reactions and suspended cell metabolism, to measure cell biomass concentrations and to detect cell toxic compounds. Another important application of particular interest is the measurement of NADH in the whole blood matrix. Currently, this is not possible due to strong fluorescence and absorption properties of the endogenous compounds in the blood. Since the present methods of measuring NADH concentration have limited sensitivity, need pretreatment of the sample <in most cases), or are time consuming, there is a need for a method to detect NADH which can overcome these difficulties. One such technique which can be adopted to measure NADH concentrations with minimum background interference is fluorescence quenching. This technique 1s highly sensitive and, in a sensor application, full reversibility can be achieved by the proper choice of an indicator fluorophore which undergoes dynamic quenching by the analyte. Therefore, a simple and straightforward assay scheme for measuring NADH can be developed by finding a suitable fluorophore whose fluorescence is quenched by NADH.

If a fluorophore solution is illuminated for longer than the fluorescence decay time of the molecule, a steady state is obtained in which the rate of production of excited molecules ⁽ in the first singlet state) is equal to their rate of decay through quenching. Quenching can be inherent or brought about by added quencher molecules (which may be of a solvent or a solute).

These decay processes affect the fluorescence quantum yield and the radiative life time of the fluorophore (dynamic quenching).

In some cases only the quantum yield is effected (static quenching). Quencher concentration and fluorescence intensities or decay times are related by the Stern-Vol er equation.

I ₀ /I = 1 + _S ,[Q3

= τ ₀ /τ

where I ₀ and I are the intensities of fluorescence in the absence and presence of the quencher concentration CQ] and τ _Q and τ are the fluorescence decay times in the absence and presence of the quencher. This relation suggests that a plot of measured intensity C(I ₀ /I)-1] against quencher concentration will give a straight line of slope equal to K _sv (the Stern-Volmer constant). This relation is often used to give a calibration curve (even if it is not a straight line) in quenching experiments. Suπroary of the nvention

It has now surprisingly been found that the fluorophore benzopurpurin 4B produces an unexpectedly large change in fluorescence intensity in the presence of NADH.

According to a first aspect of the invention there is provided the use of the fluorophore benzopurpurin 4B for the assay of NADH. This may also be expressed as a method of assay of NADH which comprises adding benzopurpurin 4B to a sample containing NADH and detecting or measuring the fluorescence quenching of benzopurpurin 4B by NADH. The quenching is normally measured by the change of fluorescence intensity of the

fluorophore benzopurpurin 4B before and after its addition to the NADH quencher and the change is then relatable to NADH concentration, e.g. by a calibration curve or standard table.

According to a second aspect of the invention there is provided a fibre optic sensor comprising a fibre optic, said fibre optic including benzopurpurin 4B immobilised on the distal (probe) end of the fibre optic.

The assay for NADH based on fluorescence quenching of benzopurpurin 4B is superior to other assay procedures, since benzopurpurin 4B has well separated visible excitation and emission (large Stokes' shift), good photostability and good quantum yield, coupled with extremely high sensitivity towards NADH. Its optical properties are favourable for the measurement of NADH in whole blood. Brief description of the drawings

Figure 1 is a Stern-Volmer plot for the quenching of Benzopurpurin 4B by NADH at 22°C; and

Figure 2 illustrates schematically, partly in section, a fibre-optic set-up for NADH sensing. Description of the preferred embodiments

The concentration of benzopurpurin 4B in the NADH analyte can be varied according to the suspected concentration of NADH in the analyte. Typically, in solution assays it will be in the range of 0.5 to 20, preferably 1 to 10, molar percent of the concentration of NADH in the analyte. Preferably the concentration of benzopurpurin 4B is in the range 0.01 to 10 μ for assaying concentrations of NADH in the range 0.1 to 500 μ , with the benzopurpurin 4B concentration being adjusted approximately in proportion to the expected NADH concentration within these ranges. In an optic fibre a layer containing 1-5 weight percent benzopurpurin 4B has been found suitable. The following Examples illustrate the invention. Example 1

Nine fluorophores listed in the Table below were tested for possible fluorescence quenching by NADH. Choice of an indicator

was dictated by (i) visible absorption, (ii) good Stokes' loss, (iii) good fluorescence quantum yield, (iv) photostability, (v) solubility in water and (vi) sensitivity towards NADH. The Table below summarises the results. TABLE

Effect of 100 μm NADH on the fluorescence of various indicators dissolved in pH7, 0.5M phosphate buffer at 22°C.

9-Aminoacridine hydrochloride 0.5

Benzopurpurin 4B 10

Fluorescein 10

α-Naphtholbenzein 10

Nile blue A 10

4-Nitrophenol 10

Nuclear Fast Red 10

Phloxine 5

Safranine 0 10

The fluorescence quenching by NADH was operative in many of the fluorophores investigated in the present study. Aminoacridine hydrochloride, Nile blue, Safranine-0 and 4-nitrophenol all showed fluorescence quenching to some extent. Most of these molecules fulfilled the requirements stated above with the exception of sensitivity.

Benzopurpurin 4B showed the maximum change in the fluorescence intensity in the presence of NADH. Therefore, this molecule was further investigated. Example 2

In a series of experiments, fluorescence intensity of the indicator benzopurpurin 4B dissolved in pH7, (0.5M) phosphate buffer was recorded both in the absence and presence of NADH. The concentration of benzopurpurin 4B was kept constant at 10 μM while the concentration of the quencher NADH was varied from 0 to 200 μM. The solutions were excited in a 1 x 1 cm quartz cell with a radiation of wavelength 458 nm which was first passed through a monochromator and then a filter. The resulting fluorescence at 680 nm was measured at 90° to the incident radiation (although other angles could be used). It was passed through a filter and monochromator to a detector and thence a signal processor. These measurements provided the necessary data for plotting a curve in accordance with the Stern-Volmer equation. Figure 1 gives the results. Quenching was extremely fast with a K _sv of 11 x 10 ³ M ^-1 . Of course, these results clearly demonstrate that fluorescence quenching of benzopurpurin 4B can be used to measure NADH concentration in a solution. The Stern-Volmer plot can be used as a calibration curve. Observed non-linearity in the Stern-Volmer plot is probably due to the existence of more than one quenching mechanism. Preliminary experiments using time-resolved laser spectroscopy suggest a decay time of 4 nanaseconds for benzopurpurin 4B, which indicates the existence of static quenching.

Example 3

Benzopurpurin 4B was immobilised onto each of glass, silica gel, polyvinyl chloride, cellulose and carbon black. An optical sensing layer was prepared by adsorbing benzopurpurin 4B on a 60 μM thick layer of silica gel layer (pore size: 10 to 40

Angstrom units = 0.1 to 0.4 micrometres) on a PVC support.

Immobilisation was achieved simply by dipping the silica gel coated surface for a period of one hour, in a 50 μM solution of benzopurpurin (in pH 7.0, 0.5M phosphate buffer). Sensing strips were then removed and rinsed by pH7 phosphate buffer. Drying did not affect the performance of these sensing layers in the fluorescence quenching of NADH.

Example 4

In order to investigate these sensing layers, a fibre optic was set up as shown in Fig.2. Excitation radiation (458 nm) from a light source 1 passes via monochromator 2, focusing lens 3 and filter 4 and one arm 5 of a bifurcated optical fibre bundle, to a

60 μm thick sensing layer 6 of benzopurpurin 4B in silica gel, retained by a rubber member 7a and support layer 7b 2mm. thick, in a chamber 8 at the distal end of the fibre optic. The NADH solutions to be assayed were injected into the sample chamber 8 through an aperture 9 by means of a syringe. Fluorescence was collected via the other arm 10 of the optical fibre bundle and carried via filter 11, focusing lens 12 and emission monochromator 13 (set at 680 nm) to fluorimeter detector 14.

The monochromator 2 and lens 3 can be omitted if desired, the filter 4 being sufficient.

The effect of NADH on the fluorescence of immobilised benzopurpurin 4B was found to be less than when the benzopurpurin 4B is in solution. One possible explanation lies in the slow diffusion of NADH in the immobilising matrix.

Other optical sensing formats are possible including evanescence, surface plasmon resonance and planar waveguide.