WO2000062050A1

WO2000062050A1 - Electrophoresis method and apparatus with time- or space-modulated sample injection

Info

Publication number: WO2000062050A1
Application number: PCT/GB2000/001355
Authority: WO
Inventors: Kirat Singh; Nicholas John Goddard; Peter Robert Fielden
Original assignee: The University Of Manchester Institute Of Science & Technology
Priority date: 1999-04-09
Filing date: 2000-04-10
Publication date: 2000-10-19
Also published as: AU4128500A

Abstract

A detection apparatus comprising a separation channel (20) into which a sample comprising different components is introduced. The apparatus is provided with means (22, 24) for introducing the sample into the separation channel with a series of predetermined separations. The components are transported along the separation channel such that the different components travel at different speeds. A plurality of temporally or spatially distinct signals are generated, each representative of transport of one component past at least one fixed point in the channel. The signals are correlated to derive frequencies characteristic of the speed of travel of the components.

Description

ELECTROPHORESIS METHOD AND APPARATUS WITH TIME-OR SPACE- MODULATED SAMPLE INJECTION

The present invention relates to a detection method and apparatus.

The velocity of travel of a charged molecule through a separation medium towards an electrode of opposite charge is determined by the charge-to-size ratio of that molecule. This property is used in electrophoresis to provide molecular analysis.

In one known form of electrophoresis, a plate of gel is provided with a buffer containing a series of spaced apart samples at an upper end, and a second buffer at a lower end. A potential is applied across the ends of the plate, whereupon charged molecules contained within the samples commence migration through the gel.

In the most common arrangement for detecting sample molecules separated using the above apparatus, radioactive labels are attached to molecules of interest. Upon completion of electrophoresis, i.e. when the molecules have migrated through the gel for a predetermined period of time, the potential across the gel is removed, and the gel is placed against a piece of X-ray film. The radioactivity of the molecules causes the film located against the molecules to be exposed, and when the film is developed the positions of the labelled molecules appear as dark bands. A disadvantage of this arrangement is that the generation of an autoradiograph is expensive and time consuming. Furthermore, the radioactive labelled molecules are generally hazardous.

An alternative arrangement for detecting sample molecules separated using the above described apparatus is via the optical detection of fluorescent labelling. A laser beam is directed across the gel close to its lower end, such that fluorescent labelled molecules passing through the beam are excited and emit fluorescent light, which is detected by a photomultiplier tube. The velocity of migration of a given molecule may be calculated from the time taken for the molecule to migrate to the position of the laser beam. A disadvantage of this arrangement is that a fluorescent signal is only emitted when the molecule passes through the laser beam. The signal provided by a given molecule is thus emitted only briefly (typically for approximately 1/3000 of the total migration time), and as a consequence the signal to noise ratio of the detected fluorescence is limited. A second disadvantage of this arrangement is that it may require a fast detector, since fluorescence is only emitted briefly by each molecule (i.e. when it is in the laser beam), which will introduce further noise into any detected signal. A further disadvantage of this detection arrangement is that analysis cannot be completed until a slowest of the fluorescent labelled molecules comprising the sample crosses the laser beam.

US Patent No. 5,699,157 describes an electrophoresis apparatus having a channel along which species bands are made to move under electrophoretic forces. The channel is illuminated transversely through a mask, and a photo receptor is used to detect fluorescent light emitted by the species bands in the channel. This apparatus detects movement of the species bands at several points along the separation channel, and therefore provides a significantly enhanced ratio of signal to noise when compared to the above apparatus.

PCT/GB98/00645 describes an apparatus similar to that disclosed in US 5,699,157, with the notable difference that a series of detectors is used to detect fluorescent light, rather than a single detector.

A disadvantage of existing electrophoresis methods is the time required to carry out a separation of a sample. A typical DNA separation will take several hours in a gel electrophoresis system.

It is an object of the present invention to provide a detection method and apparatus which overcomes at least some of the above disadvantages.

According to a first aspect of the invention there is provided a detection apparatus comprising a separation channel into which a sample comprising different components is introduced, means for transporting the components along the separation channel such that the different components travel at different speeds along the separation channel, means for generating a plurality of temporally or spatially distinct signals each representative of transport of one component past at least one fixed point in the channel, and means for correlating the plurality of signals, wherein the apparatus is provided with introducing means for introducing the sample into the separation channel with a series of predetermined separations.

The inventors have realised that it is the velocity of travel of sample components, and not to their spatial separation, that is measured by an apparatus of the type described in US 5,699,157 or PCT/GB98/00645, and that it is therefore possible to introduce a sample into a separation channel with a series of predetermined separations. The invention is advantageous because it allows several sample separation measurements to be carried out simultaneously (each measurement being of the same sample). In a situation where more than one separation measurement is required in order to obtain a required ratio of signal to noise, the invention significantly reduces the total required measurement time.

Suitably, the series of predetermined separations is a series of temporal intervals, the introducing means comprises means for introducing the sample into the separation channel at predetermined temporal intervals, and the apparatus is provided with detection means arranged to generate a plurality of spatially distinct signals indicative of the presence of the sample at a plurality of points along the channel.

This is advantageous because it allows a measurement to continue indefinitely until a required ratio of signal to noise has been obtained. In prior art electrophoresis methods, a single electrophoresis measurement is taken, and then repeated as necessary. This suffers from the disadvantage that experimental conditions may vary between each measurement. For example, the temperature of apparatus may change, or a separation voltage may vary. It may also be very difficult to reproduce a gel for a separation channel without varying to some extent the properties of the gel. These disadvantages are avoided by the present invention firstly because an electrophoresis measurement is allowed to continue indefinitely until a required signal to noise ratio is obtained (there is no need to make replacement gel), and secondly because the measurement of a plurality of samples allows the measurement to be made much more quickly (this reduces the time span over which apparatus temperatures may change).

A further advantage of the use of temporally separated sample introduction is that the amount of sample that is wasted is reduced (i.e. that which does not contribute towards a measured signal). In known electrophoresis measurements, a sample is introduced into a separation channel via an introduction channel, for example in a 'dog leg' arrangement. The introduction channel must be filled with sample before any sample will reach the separation channel. If the measurement is to be repeated, the introduction channel must again be filled with sample before any sample will reach the separation channel. Thus, where a series of measurements is required to obtain a given signal to noise ratio, a significant amount of sample is wasted. Where temporally separated sample introduction is used, an introduction channel need be filled by the sample only once, thereby reducing significantly the amount of wasted sample.

Preferably, the sample and a buffer are alternately introduced into the separation channel, and the introducing means is arranged to allow an electrophoretic voltage to be continuously applied to the separation channel, such that electrophoresis along the separation channel will not be interrupted by introduction of the sample into the separation channel.

Preferably, the apparatus is arranged to apply the electrophoretic voltage across the separation channel and a buffer introducing means, or across the separation channel and a sample introducing means.

Suitably, the voltage is switched substantially instantaneously from being across the separation channel and the buffer introducing means to being across the separation channel and the sample introducing means, such that there is continuous flow into the separation channel, the flow comprising either the sample or the buffer.

Suitably, the voltage is switched gradually from being across the separation channel and the buffer introducing means to being across the separation channel and the sample introducing means, such that there is continuous flow of into the separation channel, the flow comprising either the sample or the buffer, or a combination of the sample and the buffer. Preferably, the voltage is switched in such a way that the total voltage applied across the separation channel remains fixed.

Preferably, the gradual switching of the voltage follows a sine wave. A sine wave is preferred because it simplifies any required processing of the detected signal. Other suitable waveforms may be used, for example a sawtooth waveform.

Preferably, the introducing means comprises a sample introducing channel and a buffer introducing channel which converge at or adjacent to the separation channel.

Suitably, the introducing means is arranged to introduce the sample at regularly spaced temporal intervals. Alternatively, the introducing means is arranged to introduce the sample at pseudo-randomly spaced temporal intervals.

Preferably, the introduction of the sample and the introduction of the buffer into the separation channel is controlled by electronics and/or software. Preferably, the apparatus includes control means arranged to continue introducing the sample and the buffer into the separation channel until an electrophoretic measurement of a sample has been obtained with a signal to noise ratio greater than a predetermined value.

Suitably, the introducing means comprises a series of channels arranged to introduce the sample at spaced introduction positions along the separation channel.

The introduction positions may be regularly spaced, or may be spaced in a pseudo-random sequence.

Preferably, the components are fluorescent molecules and the signal generating means is an optical detector.

A series of opaque regions may be interposed between the separation channel and the detector. This is one way in which the distinct signals required by the invention may be generated. The series of opaque regions may be located in the image plane of the separation channel. The detector may be a single large area detector. The single large area detector may only be used when the sample is introduced with a predetermined spatial separation, and is not appropriate for use when the sample is introduced with a predetermined temporal separation.

The detector may be divided into a series of independent detecting cells, such that signals from the cells may be combined to form a modulated time series. The detecting cells may comprise a charge coupled device (CCD) or other suitable detecting arrangement. This is an alternative way in which the distinct signals required by the invention may be generated. Where a CCD is used, it is possible to program cells of the CCD to act in the same way as the mask referred to above. This is advantageous because it allows the separation of mask elements to be tailored to suit specific applications, for example using appropriate software.

The detector may be spaced away from the separation channel, a focussing lens being interposed between the separation channel and the detector to focus light emitted from the channel onto the detector. The term light is intended to include electromagnetic radiation of any suitable wavelength, and is not limited to visible electromagnetic radiation.

Suitably, the separation channel is provided in a structure configured to support a propagating optical mode when filled with fluid or gel, thereby allowing fluorescent molecules to be excited continuously by said propagating optical mode during migration along the separation channel.

The term 'fluid or gel' is intended to encompass any medium suitable for the transport of fluorescent molecules.

Suitably, a grating is provided to couple light into the optical mode.

Suitably, the optical mode supporting structure comprises the separation channel for filling with fluid or gel, a lowermost surface of the separation channel being defined by a layer of material having a refractive index greater than that of the fluid or gel, and a substrate upon which the material is supported, wherein the structure defines a waveguide capable of supporting an optical mode confined in the separation channel and the material parameters are chosen such that an optical mode confined in the separation channel will suffer substantially anti-resonant reflection as a consequence of the interface between the material and the fluid or gel and the interface between the material and the substrate.

The reference above to a mode being confined in the separation channel of the waveguide structure is intended to mean that the mode is centred on the separation channel of the waveguide, and it will be appreciated that a proportion of the mode will extend beyond that layer.

Suitably, the optical mode supporting structure comprises the separation channel for filling with fluid or gel, a substrate forming a lower surface of the separation channel and having a refractive index greater that of the fluid or gel, and a superstrate forming an upper surface of the separation channel and having a refractive index greater than that of the fluid or gel.

The signal generating means may comprise a series of detectors spaced along the separation channel. The detectors may be regularly spaced, or may be spaced in a pseudo-random sequence.

The detectors may be arranged to detect an electrical property at points spaced along the separation channel.

Suitably, the signal generating means comprises a single detector together with means for injecting the molecules into the separation channel with a series of predetermined separations. According to a second aspect of the invention there is provided a method of detection comprising introducing a sample comprising different components into a separation channel, the sample being introduced with a series of predetermined separations, inducing the components to travel along the separation channel at different speeds, generating a plurality of temporally or spatially distinct signals each representative of transport of one component past at least one fixed point in the channel, and correlating the plurality of signals.

Suitably, the sample is introduced into the separation channel at a series predetermined temporal intervals, and a plurality of spatially distinct signals indicative of the presence of the sample at a plurality of points along the channel are measured.

Suitably, the components are fluorescent molecules, the separation channel is provided in a structure configured to support a propagating optical mode when filled with fluid or gel, thereby allowing fluorescent molecules to be excited continuously by said propagating optical mode during migration along the separation channel, and the signals are generated by optical detection of the fluorescent molecules.

Suitably, two propagating optical modes of different wavelengths are supported by the separation channel.

The method may incorporate any of the above features of the apparatus according to the first aspect of the invention.

Specific embodiments of the invention will now be described by way of example only, with reference to the accompanying drawings in which:

Figure 1 is a schematic illustration of a detection apparatus to which the invention may be applied;

Figure 2 is a schematic illustration of a detection apparatus according to a first embodiment of the invention;

Figure 3 is a schematic illustration of a detection apparatus according to a second embodiment of the invention;

Figure 4 is a schematic illustration of a second detection apparatus to which the invention may be applied;

Figure 5 is a schematic perspective view of a first waveguide compπsing part of a detection apparatus according to the invention; and Figure 6 is a schematic illustration of a second waveguide comprising part of a detection apparatus according to the invention.

Referring first to Figure 1, a spacer layer 1 into which a separation channel (not shown) is patterned is provided between a substrate 2 and superstrate 3, the entire structure being supported on a coupling prism 4 (the substrate 2 and prism 4 may be fabricated as a single entity). The separation channel within the spacer layer 1 is filled with liquid electrolyte, a molecular sample is provided at a first end of the channel, and a potential difference is applied across ends of the channel such that charged molecules in the sample will commence migration along the channel through the electrolyte.

A laser 5 provides a coherent beam of light 6 which is coupled via the prism 4 into the separation channel of the spacer layer 1. The separation channel and the substrate 2 and superstrate 3 have refractive indices chosen such that the beam is substantially confined within the separation channel. The beam propagates along the separation channel parallel to the direction of migration of the charged molecules of the sample. A proportion of the beam leaks from the separation channel back into the coupling prism 4, as shown by the arrows 7.

The charged molecules of the sample are provided with fluorescent labels, and the wavelength of the incident light 6 is chosen such that it will induce fluorescence of the labelled molecules. Fluorescence emitted by a labelled molecule is shown schematically by the arrows 8 in Figure 1. A focusing lens 9 is provided parallel to the spacer layer 1, spaced a suitable distance away from said spacer layer 1. A charged coupled device (CCD) track 10 (i.e. a linear series of CCD cells) is provided at an opposite side of the focusing lens 9, parallel to and spaced a suitable distance away from said lens 9 such that the spacer layer 1 is imaged onto the CCD track 10. The distances between the spacer layer 1 and the focusing lens 9, and between the CCD track 10 and the focusing lens 9 may be selected to magnify or de-magnify the spacer layer 1 image to fit the CCD.

As indicated by the arrows 8, fluorescent light emitted by a labelled molecule at a given position in the separation channel of the spacer layer 1 will be focused by the focusing lens 9 onto a cell at a corresponding position on the CCD track 10. As the fluorescent labelled molecule travels along the separation channel of the spacer layer 1, fluorescent light emitted by the molecule moves correspondingly along the CCD track 10.

A computer or other processor (not shown) is connected to the CCD track 10 to allow analysis of detected fluorescence. In one data analysis arrangement successive cells of the CCD track 10 are arranged to have alternately plus 1 and minus 1 response functions. A point of fluorescent light passing across such a CCD track 10 will produce an alternating signal, the frequency of which is determined by the velocity of the motion of the point of light across the CCD track 10. This velocity corresponds to the velocity of a fluorescent labelled molecule in the separation channel.

In an alternative data analysis arrangement, output from each element in the CCD track 10 is multiplied by a sinusoidal weighting factor^, the weighted output is summed, and a Fourier transform of a time record of the sums is calculated:

The transformed output includes distinct peaks, each of which is associated with a different velocity of migration of a fluorescent labelled molecule in the separation channel of the spacer label. It will be understood that any suitable function may be used to transform data into the frequency domain.

The intensity of light in the separation channel of the spacer layer 1 will decay along the channel, and this will be seen as low frequencies when the data is transformed to the frequency domain. If required, the CCD track 10 may be calibrated with an inverse function of that decay, to remove its effect from the detected signal.

The CCD track 10 may comprise a CCD array with a single track selected to be active, or a CCD array with rows lying transverse to the direction of molecule movement having their outputs summed together. A series of separation channels may be provided in combination with a series of appropriately positioned CCD aπays.

Although the apparatus illustrated in Figure 1 includes a laser 5 which produces a beam of coherent light 6, the invention does not require coherent light. The laser could be replaced with, for example, a light emitting diode (LED) collimated to allow coupling to the spacer layer 1 , or any other suitable source. The wavelength of the source is selected so as to induce fluorescence of molecules located within the spacer layer 1.

It is the velocity of travel of fluorescent labelled molecules, and not to their spatial separation that is measured by the apparatus shown in Figure 1. This allows a sample to be injected at different points along a separation channel. Figure 2 illustrates a first embodiment of the invention, which includes multiple injection points. A separation channel 20 is terminated at either end by buffer reservoirs 21 (an arrow A indicates the direction of separation). A sample reservoir 22 and a further buffer reservoir 23 are located on opposite sides of one end of the separation channel 20. The sample reservoir 22 and buffer reservoir 23 are each connected to the separation channel 20 by nine injection channels 24.

A lens 25 is arranged to image a portion of the separation channel 20 through a mask 26 and onto a large area detector 27. In use, a sample containing fluorescent labelled molecules is injected into the channel 20 simultaneously through the nine injection channels 24 for a limited period. The sample travels along the separation channel 20 as described above, and is made to emit fluorescence by a beam of light which is coupled along the separation channel 20. The mask 26 is separated into segments having a spacing which corresponds to the spacing of the injection channels 24. Since the injection of the sample and the mask 26 are in spatial phase, the light signal from each individual injected sample will rise and fall at the same time as the others. The output is collected over a suitable period of time and then Fourier transformed to obtain distinct peaks, each of which is associated with a different velocity of migration of a fluorescent labelled molecule in the separation channel.

Multiple spatial injection is advantageous because it allows several separation measurements of a sample to be carried out simultaneously.

The detector 27 and mask 26 may be replaced by a CCD track.

The injection channels 24 may be separated by a pseudo-random sequence of distances, for detection by correlation. Where this is done, an array of detection cells separated by the same pseudo-random sequence of distances is required.

Injection of a sample into the separation layer 20 may be made precise by metering the injection. Sensors may be placed in the injection channels 24 adjacent to the separation channel 20, to monitor the integrity of the sample at the injection points.

It is important to ensure that the sample is homogeneous when it enters the separation channel. The sample is drawn from the sample reservoir 22 towards the buffer reservoir 23 by an electric field, and there is a risk that the integrity of the sample will be compromised by different sample elements flowing from the sample reservoir 22 at different rates. In order to avoid this, the injection channels 24 are made from a material with a fixed surface charge such as silica or glass. The effect of this surface charge is to make the sample flow as a plug (electro-osmosis) from the sample reservoir 22, thereby ensuring that it is homogeneous when it enters the separation channel 20.

Figure 3 shows a second embodiment of the invention, which may be used to provide a series of temporally separated sample injections. A sample reservoir 28 and buffer reservoir 29 are provided at one end of a separation channel 30. A second buffer reservoir 31 is provided at the opposite end of the separation channel 30. A lens 32 is arranged to image a portion of the separation channel 30 onto a CCD track 33. In use, a sample is injected alternately with a buffer to provide a series of injections separated by equal time intervals, rather than equal spatial intervals. The injection is automated, for example by software and/or electronics arranged to inject from the sample reservoir 28 and buffer reservoir 29 according to a predetermined sequence. As before, fluorescent labelled molecules are made to emit fluorescence by a beam of light which is coupled along the separation channel 30. Since the mobility of different fluorescent labelled molecules varies, this will result in a series of peaks which are separated by a distance dependent upon their mobility, and are always separated by the same time between injections. A pattern of overlapping peaks will be imaged by the lens 32 onto the CCD track 33. The pattern can be Fourier transformed from the spatial to the spatial frequency domain to obtain peaks which correspond to velocities of migration of fluorescent labelled molecules in the separation channel 30. Some form of convolution may be required if the injections are temporally separated by a non-regular distribution, for example a predetermined pseudo-random distribution. The embodiment of the invention illustrated in Figure 3 is advantageous over the embodiment illustrated in Figure 2 because only a single injector is needed to perform multiple injections, but suffers from the disadvantage that some form of array detector is needed (a single detector with a mask cannot be used).

The embodiment of the invention illustrated in Figure 3 is advantageous because it allows a measurement to continue indefinitely until a required ratio of signal to noise has been obtained. In prior art electrophoresis methods, a single electrophoresis measurement is taken, and then repeated as necessary. This suffers from the disadvantage that experimental conditions may vary between each measurement. For example, the temperature of apparatus may change, or a separation voltage may vary. It may also be very difficult to reproduce a gel for a separation channel without varying to some extent the properties of the gel. These disadvantages are avoided by the embodiment of the invention illustrated in Figure 3 firstly because an electrophoresis measurement is allowed to continue indefinitely until a required signal to noise ratio is obtained (there is no need to make replacement gel), and secondly because the measurement of a plurality of samples allows the measurement to be made much more quickly (this reduces the time span over which apparatus temperatures may change).

A further advantage of the use of the embodiment of the invention illustrated in Figure 3 is that the amount of sample that is wasted is reduced (i.e. that which does not contribute towards a measured signal). In known electrophoresis measurements, a sample is introduced into a separation channel via an introduction channel, for example in a 'dog leg' arrangement. The introduction channel must be filled with sample before any sample will reach the separation channel. If the measurement is to be repeated, the introduction channel must again be filled with sample before any sample will reach the separation channel. Thus, where a series of measurements is required to obtain a given signal to noise ratio, a significant amount of sample is wasted. Where temporally separated sample introduction is used, an introduction channel need be filled by the sample only once, thereby reducing significantly the amount of wasted sample.

Sample from the sample reservoir 28 is introduced into the separation channel 30 by applying an electrophoretic voltage across the sample reservoir and the second buffer reservoir 31. Buffer from the first buffer reservoir 29 is introduced into the separation channel 30 by applying an electrophoretic voltage across the first buffer reservoir and the second buffer reservoir 31. Continuous electrophoresis in the separation channel 30 may be obtained by switching the electrophoretic voltage instantaneously from the sample reservoir 28 to the first buffer reservoir 29, and from the first buffer reservoir 29 to the sample reservoir 28.

Alternatively, continuous electrophoresis may be obtained by switching the electrophoretic voltage gradually from being across the second buffer reservoir 31 and the first buffer reservoir 29 to being across the second buffer reservoir 31 and the sample reservoir 28. The voltage is switched in such a way that the total voltage applied across the separation channel 30 remains fixed.

The gradual switching of the voltage may follow a sine wave. A sine wave is preferred because it simplifies any required processing of the detected signal. Other suitable waveforms may be used, for example a sawtooth waveform.

The sample may be introduced into the separation channel at regularly spaced temporal intervals, or at pseudo-randomly spaced temporal intervals.

The introduction of the sample and the introduction of the buffer into the separation channel is controlled by electronics and/or software.

The apparatus includes control means arranged to continue introducing the sample and the buffer into the separation channel until an electrophoretic measurement of a sample has been obtained with a signal to noise ratio greater than a predetermined value.

Multiple injection of the sample, as illustrated in Figures 2 and 3, allows the invention to be implemented with a single point detector (not shown). This may be for example a conductivity sensor placed at a lower end of the separation channel (20, 30). If the sample were to be injected at several equally spaced points along the separation channel 20, 30 (as shown in Figures 2 and 3), then the signal seen by the single point detector would be a modulated signal, the rate of modulation being determined by the spacing of the injection points and the speed of travel of molecules comprising the sample. The detected modulated signal may be processed to determine the speed of travel of the molecules compπsmg the sample (for example by performing a Fouπer transform).

The invention is advantageous over the pnor art because it provides a higher signal to noise ratio than can be achieved using a single point detector and a single injection of sample.

A second apparatus to which the invention may be applied is illustrated in Figure 4. The components of the apparatus shown in Figure 4 are the same as shown in Figure 1, except that the CCD track 10 is replaced by a single large area detector 11, and a seπes of blacked-out regions 12 (for example having a thickness of 100 microns, and a corresponding spacing) are provided at an uppermost surface of the superstrate 3. The operation of the embodiment shown in Figure 4 is analogous to the operation of the apparatus shown in Figure 1. Fluorescent labelled molecules passing along the separation channel of the spacer layer 1 will emit fluorescent light when excited by the incident light 6. If a fluorescent molecule is located below a blacked- out region 12 of the superstrate 3, the light emitted by that molecule will be masked by the blacked-out region, and will not reach the large area detector 11 Thus, fluorescent light emitted by a molecule will only reach the large area detector 11 when that molecule is not located beneath a blacked-out region 12 of the superstrate 3. Each fluorescent labelled molecule will thus provide intermittent fluorescent light to the large area detector 11, at a characteπstic frequency determined by the velocity of migration of that molecule. The large area detector 11 may be replaced by a CCD array configured to sum signals detected withm a given area.

In practice, it may not be possible to prevent fluorescent light reaching the large area detector 11 when a fluorescent labelled molecule is, for example, located underneath and close to an edge of a blacked-out region 12 of the superstrate 3 However, this will not effect the charactenstic frequency of the detected light, but will merely modify the shape of the detected signal from an ideal case square waveform to a waveform having a degree of curvature.

The apparatus shown m Figure 4 suffers from the disadvantage that fluorescent light from a labelled fluorescent molecule will not provide a signal continuously, but only when it is not located beneath a blacked-out region 12 of the superstrate 3 Approximately 50% of the fluorescence emitted by the molecule is therefore lost, thereby reducing the maximum possible available signal to noise ratio

A further disadvantage of the apparatus shown in Figure 4 is that the detected light will contain harmonics of the characteπstic frequency, which will appear as extra frequency components m the detected signal The harmonics aπse because the detected signal is effectively a square wave, and therefore includes many frequency components The magnitude of the harmonics may be significant, and the harmonics may reduce the achievable signal to noise ratio of the second embodiment

The seπes of blacked out regions 12 shown in Figure 4 may be provided in the image plane of the focusing lens 9 (I e adjacent the detector 11) This configuration is easier to arrange than that shown in Figure 4

Parallel processing of molecular samples may be achieved by providing a seπes of separation channels in a spacer layer, the channels being spaced apart and parallel, and providing complementary positioned CCD tracks (or a two-dimensional CCD array)

The separation channel of the spacer layer 1 shown in Figures 1 and 4 is required to confine liquid electrolyte through which a molecular sample can migrate, and also to confine a beam of light to propagate in a direction parallel to the direction of migration One possible configuration of separation channel which will satisfy these requirements is a conventional waveguide configuration, wherein the separation channel is clad on one side by matenal having a greater refractive index than the separation channel The conventional waveguide structure will support an optical mode which is centred to the higher index matenal, with an evanescent wave which extends into the separation channel The conventional waveguide suffers from the disadvantage that the intensity of the evanescent wave in the separation channel \\ ill be low, and the intensity of fluorescence thus obtained will be correspondingly weak A second disadvantage of the conventional waveguide is that the optical mode w ill suffer heavy scatteπng and absorption losses in the higher index matenal, and the intensity of light will thus decrease rapidly as it propagates along the waveguide

An alternative configuration of separation channel comprises an Anti- Resonant Reflecting Optical Waveguide (ARROW) ARROW waveguides were first developed in 1986 at AT&T Bell labs, and are descπbed in the paper Duguay et al, Appl. Phys. Lett., 49 (1986) 13-15. An example of a suitable ARROW waveguide is shown in Figure 5, wherein the waveguide comprises a liquid electrolyte 13 (for example, having a refractive index of n = 1.335, and a thickness of d = 4 μm) situated on top of a thin high index layer (for example, n = 2 and d = 0.1 μm), which in turn is located on top of a layer of silica (for example, n = 1.46 and d = 1 μm). The entire structure is supported on a transmissive substrate (for example lead glass, n = 1.7). A further low index material (not shown) is provided above the liquid electrolyte 13.

The ARROW structure acts as a Fabry-Perot resonator at anti-resonant wavelengths, providing a very high degree of confinement of the optical mode within the liquid electrolyte 13. The optical mode is confined at the uppermost surface of the electrolyte 13 by total internal reflection, provided that a material located at that surface has a refractive index less than that of the electrolyte 13 (for example, the material may be air). The thin high index layer 14, layer of silica 15, and substrate 16 may be repeated above the electrolyte 13, such that the electrolyte is held between two ARROW structures.

The waveguide shown in Figure 5 exploits an important advantage of ARROW structures, namely that they allow concentration of an optical field in a low refractive index region of interest (i.e. the electrolyte 13). This feature is important because it allows a high intensity of fluorescence to be obtained.

The waveguide shown in Figure 5 is easy to fabricate, and dispersion characteristics of ARROW modes are such that even a quite large variation in waveguide parameters, i.e. layer thickness or refractive index, does not significantly affect the operation of the waveguide. This is a significant advantage of the invention, since conventional waveguides are very sensitive to variation of waveguide parameters.

The waveguide of Figures 1 and 4 is shown in Figure 6, and is referred to hereafter as a light condenser. The light condenser comprises a high index superstrate 17, a low index electrolyte 18 and a high-index substrate 19. One possible configuration of a light condenser comprises an electrolyte fluid with a refractive index of n = 1.333, sandwiched between two layers of glass. The light condenser is a simple waveguide configuration which enhances the confinement of a beam within the liquid electrolyte 18. Total internal reflection of light in the electrolyte 18 does not occur at the boundaπes with the superstrate 17 and the substrate 19 Instead, reflection is provided as a result of the refractive index mismatch between the electrolyte 18 and the superstrate 17 and substrate 19, thereby providing a mode centred on the electrolyte 18 The light condenser enjoys the same advantageous feature as the ARROW, namely that the mode is centred in the electrolyte, thereby providing efficient excitation of fluorescent molecules and minimising scattenng and absorption losses of the mode. This allows the intensity of excitation of molecules (and thus the intensity of their fluorescence) to be maintained at a relatively constant level along the length of the waveguide

The light condenser does not generally require lateral mode confinement, since light coupled to the light condenser is colhmated Where lateral confinement is required, this may be provided by etching into a substrate or superstrate a channel for receiving electrolyte.

The light condenser waveguide may be fabncated from injection moulded plastic Electrodes and a coupling pπsm may be formed together with the light condenser dunng fabπcation. For example, where a hole is provided in to allow electrolyte to be introduced into the light condenser, an electrode may be located along that hole

In an alternative waveguide (not shown) an ARROW structure may be provided with an upper surface compnsmg a high index superstrate, such that the waveguide compnses a combination of ARROW and light condenser structures

Although in the embodiments of the invention illustrated in Figures 1 and 2 light is coupled to the spacer layer of the separation channel 1 via a coupling prism 4, any other suitable form of coupling may be used, for example end-fire coupling or grating coupling

An advantage of the invention is that it allows the analysis of very small samples of DNA, since fluorescent labelled DNA fragments of the sample will provide a continuous detectable signal Conventional fluorescent electrophoresis requires a much larger DNA sample in order to provide a useful amount of fluorescence A DNA sample sufficiently large to be used for conventional fluorescent electrophoresis may be time consuming and expensive to produce In conventional electrophoresis arrangements, a thick channel of gel is required (typically, 5mm by 200 microns) due to the large quantity of DNA being sampled, since the gel must be capable of expanding to allow the DNA to pass through it. In the present invention, because a much smaller sample of DNA is used, a narrow separation channel, for example 10 microns by 100 microns may be used. The narrow channel allows the used of liquid electrolyte instead of gel, which in turn allows quicker diffusion of the DNA, a further advantage of the present invention. Although the use of liquid electrolyte provides significant advantages, gel may used provided that the separation channel is sufficiently broad.

The use of liquid electrolyte is known, with a correspondingly required small sample of DNA, from the prior art, but transverse temporary illumination of the DNA requires a very bright laser to obtain a detectable fluorescent signal. Furthermore, in prior art of this form filtering is often required to separate scattering of the high power laser from the fluorescence. A further disadvantage is that the laser introduces noise at high frequencies. The present invention does not require the use of a high power laser, gives less scattering, and therefore allows easier detection of fluorescent light.

The invention may be implemented using detection anays which do not require light to be guided down the separation channel. For example, a series of conductivity sensors may be placed at equally spaced intervals along the separation channel. The conductivity measured by a sensor will be modified as a molecule travels past it. The series of sensors will give rise to a modulated detected signal, the frequency of the modulation being determined by the speed of travel of the molecule and the separation of the sensors. Other electrical sensing arrangements may be used, for example permittivity sensing, impedance sensing, amperometry, voltammetry and potentiometry.

Each sensor may comprise two electrodes, and may be provided in a variety of configurations. For example, two parallel electrodes separated by a small gap may be provided across a lower surface of the separation channel. Alternatively, a first electrode may be provided across a lower surface of the separation channel and a second parallel electrode may be provided across an upper surface of the separation channel. In a further alternative arrangement, each sensor may comprise a pair of electrodes having spaced apart ends. Each sensor may comprise three electrodes spaced apart along the direction of travel along the separation channel. This arrangement allows the derivative of the conductivity (or other electrical property) to be measured.

Electro-chemical luminescence sensors may be used, wherein the electrodes are chemically modified to have a specificity to a particular interaction such that a target molecule in a sample will combine with the electrode and emit light.

It will be appreciated that any suitable known type of sensing may be used to provide an array of sensors, for example refractive index sensing, optical absorption sensing, thermal conductivity sensing, acoustic density sensing or acoustic viscosity sensing.

The sensors may be arranged as a pseudo-random series. The rate of travel of detected molecules may then be obtained by convoluting the detected signal with the pseudo-random series.

As is described above, a potential difference is applied across ends of the separation channel such that charged molecules in the sample will migrate along the separation channel. The electric potential difference may be applied continuously. Alternatively, the electric potential may be applied intermittently, with measurements of the positions of molecules being made when the field is turned off. This alternative arrangement allows longer integration times, and thus larger signals, when a CCD track detector is used.

Multiple Fourier transforms can be taken and averaged together to improve signal-to-noise.

Multiple wavelength excitation of fluorescent molecules may be obtained by coupling light of different wavelengths into a separation channel. This may conveniently achieved for example by providing a first wavelength of light at one end of the channel and a second wavelength of light at the opposite end of the channel. Multiple wavelength excitation is particularly suited to non-dispersive waveguide structures (for example, the light condenser or ARROW structures), since a given coupling arrangement will work over a wide range of wavelengths.

Correlation of detected signals may be achieved in any convenient manner, for example using Fourier transforms as described above.

Claims

1. A detection apparatus comprising a separation channel into which a sample comprising different components is introduced, means for transporting the components along the separation channel such that the different components travel at different speeds along the separation channel, means for generating a plurality of temporally or spatially distinct signals each representative of transport of one component past at least one fixed point in the channel, and means for correlating the plurality of signals, wherein the apparatus is provided with introducing means for introducing the sample into the separation channel with a series of predetermined separations.

2. A detection apparatus according to claim 1, wherein the series of predetermined separations is a series of temporal intervals, the introducing means comprises means for introducing the sample into the separation channel at predetermined temporal intervals, and the apparatus is provided with detection means arranged to generate a plurality of spatially distinct signals indicative of the presence of the sample at a plurality of points along the channel.

3. A detection apparatus according to claim 1 or 2, wherein the sample and a buffer are alternately introduced into the separation channel, and the introducing means is arranged to allow an electrophoretic voltage to be continuously applied to the separation channel, such that electrophoresis along the separation channel will not be interrupted by introduction of the sample into the separation channel.

4. A detection apparatus according to claim 3, wherein the apparatus is aπanged to apply the electrophoretic voltage across the separation channel and a buffer introducing means, or across the separation channel and a sample introducing means.

5. A detection apparatus according to claim 4, wherein the voltage is switched substantially instantaneously from being across the separation channel and the buffer introducing means to being across the separation channel and the sample introducing means, such that there is continuous flow into the separation channel, the flow comprising either the sample or the buffer.

6. A detection apparatus according to claim 4, wherein the voltage is switched gradually from being across the separation channel and the buffer introducing means to being across the separation channel and the sample introducing means, such that there is continuous flow of into the separation channel, the flow comprising either the sample or the buffer, or a combination of the sample and the buffer.

7. A detection apparatus according to claim 6, wherein the gradual switching of the voltage follows a sine wave.

8. A detection apparatus according to any of claims 2 to 7, wherein the introducing means comprises a sample introducing channel and a buffer introducing channel which converge at or adjacent to the separation channel.

9. A detection apparatus according to any of claims 2 to 8, wherein the introducing means is arranged to introduce the sample at regularly spaced temporal intervals.

10. A detection apparatus according to any of claims 2 to 8, wherein the introducing means is arranged to introduce the sample at pseudo -randomly spaced temporal intervals.

11. A detection apparatus according to any of claims 2 to 10, wherein the introduction of the sample and the introduction of the buffer into the separation channel is controlled by electronics and/or software.

12. A detection apparatus according to claim 11, wherein the apparatus includes control means ananged to continue introducing the sample and the buffer into the separation channel until an electrophoretic measurement of a sample has been obtained with a signal to noise ratio greater than a predetermined value.

13 A detection apparatus according to claim 1, wherein the introducing means comprises a seπes of channels ananged to introduce the sample at spaced introduction positions along the separation channel

14 A detection apparatus according to claim 13, wherein the introduction positions are regularly spaced

15 A detection apparatus according to claim 13, wherein the positions are spaced m a pseudo-random sequence

16 A detection apparatus according to any preceding claim, wherein the components are fluorescent molecules and the signal generating means is an optical detector

17 A detection apparatus according to claim 16, wherein a seπes of opaque regions is interposed between the separation channel and the detector

18 A detection apparatus according to claim 16, wherein the detector is divided into a senes of independent detecting cells, such that signals from the cells

be combined to form a modulated time seπes

19 A detection apparatus according to any one of claims 16 to 18, wherein the separation channel is provided in a structure configured to support a propagating optical mode when filled with fluid or gel, thereby allowing fluorescent molecules to be excited continuously by said propagating optical mode dunng migration along the separation channel

20 A detection apparatus according to claim 19, wherein a grating is provided to couple light into the optical mode

21. A detection apparatus according to claim 19 or 20, wherein the optical mode supporting structure comprises the separation channel for filling with fluid or gel, a lowermost surface of the separation channel being defined by a layer of material having a refractive index greater than that of the fluid or gel, and a substrate upon which the material is supported, wherein the structure defines a waveguide capable of supporting an optical mode confined in the separation channel and the material parameters are chosen such that an optical mode confined in the separation channel will suffer substantially anti-resonant reflection as a consequence of the interface between the material and the fluid or gel and the interface between the material and the substrate.

22. A detection apparatus according to claim 19 or 20, wherein the optical mode supporting structure comprises the separation channel for filling with fluid or gel, a substrate forming a lower surface of the separation channel and having a refractive index greater that of the fluid or gel, and a superstrate forming an upper surface of the separation channel and having a refractive index greater than that of the fluid or gel.

23. A method of detection comprising introducing a sample comprising different components into a separation channel, the sample being introduced with a series of predetermined separations, inducing the components to travel along the separation channel at different speeds, generating a plurality of temporally or spatially distinct signals each representative of transport of one component past at least one fixed point in the channel, and coπelating the plurality of signals.

24. A method of detection according to claim 23, wherein the sample is introduced into the separation channel at a series predetermined temporal intervals, and a plurality of spatially distinct signals indicative of the presence of the sample at a plurality of points along the channel are measured.

25. A detection method according to claim 23 or 24, wherein the components are fluorescent molecules, the separation channel is provided in a structure configured to support a propagating optical mode when filled with fluid or gel, thereby allowing fluorescent molecules to be excited continuously by said propagating optical mode during migration along the separation channel, and the signals are generated by optical detection of the fluorescent molecules.

26. A detection method according to claim 25, wherein two propagating optical modes of different wavelengths are supported by the separation channel.

27. A detection method according to any of claims 23 to 26, and incorporating the apparatus according to any of claims 1 to 22.