WO2005002730A1

WO2005002730A1 - Microfluidic method and device

Info

Publication number: WO2005002730A1
Application number: PCT/GB2004/002850
Authority: WO
Inventors: Stephan Mohr; Philip John Royle Day; Nicholas J. Goddard; Peter Robert Fielden
Original assignee: The University Of Manchester
Priority date: 2003-07-02
Filing date: 2004-07-01
Publication date: 2005-01-13
Also published as: GB0315438D0

Abstract

A microfluidic device and method for the monitoring, manipulation and/or detection of chemical and biological samples wherein an aqueous reaction mixture comprising a sample is introduced into a volume of a carrier fluid immiscible with the aqueous reaction mixture to form a droplet of the aqueous mixture supported by the carrier fluid. The sensitivity of the method is regulated by controlling the volume of aqueous mixture used to form the droplet and is particularly suitable for detecting analytes from a minor population of cells within mixed populations of cells.

Description

MICRO LUIDIC METHOD AND DEVICE The present invention relates to microfluidic methods and devices for the accurate and sensitive, quantitative detection and analysis of chemical entities, including chemical and biological samples such as a minor population of cells within mixed populations of cells. In particular the invention relates to the amplification and subsequent detection of nucleic acid sequences from the minor population of cells of interest.

The analysis and manipulation of small quantities of chemicals may be carried out using conventional techniques. Of particular interest of late is the use of nanoliter sized plugs in miniaturised chemical and biological reactions whereby each plug can act as a microreactor. However, the content of the plug has to be compatible with the device in which it is created due to its contact with the device. The plugs are also unsatisfactory for the analysis and manipulation of very small quantities of sample. It is desirable to provide a method and device that could allow extremely small known volumes of chemicals (pico or nanoliters) to be monitored and manipulated whilst being kept separate from any chemicals that may result in unwanted side reactions.

On a biological level, existing methods for detecting a minor population of cells within mixed populations of cells suffer from relatively low sensitivity. This represents a problem in a number of important situations. For instance, a clinician may wish to test a biopsy sample to evaluate whether or not the tissue is free of cancerous cells (a procedure known as minimal residual disease (MRD) in the case of the analysis of leukaemic cells from whole blood samples). Conventional methods of detection are unsatisfactory because they may fail to detect very low numbers of cancerous cells (e.g. 1 - 100 cells) amongst a high number of normal cells (e.g. 10⁶ cells or more). If the clinician is unable to detect a low number of cancerous cells the patient may believe they are clear of cancer but could face the regrowth of a tumour from the "missed" cancerous cells. It this therefore clear that there is a need in the art to provide sensitive methods for detecting minor populations of cells (e.g. cancer cells) within mixed populations of cells. Indeed in time, retrospective analysis to clinical outcome will permit the definition of clinically meaningful gene expression thresholds to nurture the implementation of gene-assay directed therapy.

Therefore, a first object of the present mvention is to provide improved microfluidic methods and devices for the monitoring and manipulation of chemical entities, including biological samples.

A further object of the present invention is to provide improved methods for detecting a minor population of cells within mixed populations of cells.

Accordingly, a first aspect of the present mvention provides a method of monitoring and/or manipulating at least one chemical entity, the method comprising: (i) obtaining an aqueous reaction mixture comprising at least one chemical entity; (ii) introducing a volume of the aqueous reaction mixture into a volume of a carrier fluid immiscible with the aqueous reaction mixture to form a droplet of the aqueous reaction mixture supported by the carrier fluid; and (iii) monitoring and/or manipulating the chemical entity within the droplet.

The method of the first aspect provides a general schema to provide enhanced monitoring and/or manipulation of a chemical entity on a microscale. The chemical entity is kept within its own discrete environment and may be manipulated, for example by the addition of further reagents to the droplet and/or by movement ofthe droplet through area or zones of particular conditions, such as heat or light.

Further reagents may be added to the aqueous mixture prior to formation of the droplet or to the droplet itself. Droplets containing different chemicals may be merged together. Preferably, the droplet is transfeired within the carrier fluid, the droplet being surrounded by the fluid except for when it is desired to merge the droplet with another droplet.

The sensitivity of the method is regulated by controlling the volume of aqueous mixture used to form the droplet.

In order to expose the aqueous reaction mixture to different conditions for the manipulation thereof, (such as zones at different temperatures) it is preferable to introduce the aqueous reaction mixture into an immiscible moving carrier fluid that is supported in a microfluidic flow manifold or chip. Such devices represent an important aspect of the invention. The device may be fabricated by direct machining of polymeric substrates or by injection moulding. To this end, a second aspect of the present invention provides a microfluidic device comprising: a substrate with different zones that are subjectable to defined conditions, the substrate bearing a continuous conduit adapted to carry an immiscible carrier fluid that passes through the different zones; a reservoir which is comiected to the conduit with means to introduce droplets of an aqueous reaction mixture from within the reservoir into the carrier fluid within the conduit, the droplet being supported by the carrier fluid; and optionally a detector capable of identifying products within individual droplets which have travelled through the different zones within the conduit.

The method of the present invention is particularly applicable to the detection of minor population of cells within mixed population of cells.

Therefore, a third aspect of the present invention provides a method of detecting analytes from a minor population of cells within mixed populations of cells, the method comprising: i) obtaining an aqueous reaction mixture comprising analytes from a mixed population of cells; ii) adding assay reagents suitable for detecting analytes from the minor population of cells of interest to the aqueous reaction mixture; iii) introducing a volume of the aqueous reaction mixture into a volume of a carrier fluid immiscible with the aqueous reaction mixture to form a droplet of the aqueous reaction mixture supported by the carrier fluid; iv) performing some or all of an assay, in the droplet, wherein an assay product is representative ofthe presence ofthe minor population of cells of interest; and v) analysing the droplet for the presence ofthe assay product.

The method of the third aspect of the invention represents a general schema to increase sensitivity (and also specificity) of biological sample detection. It is particularly useful for aiding the determination of disease thresholds. The method may also be used to simplify current laboratory multi-step and contamination prone processes with savings relating to time, money and increased assay throughput, quality and reliability. The method is also useful as a basis for improved clinical diagnosis and prognosis. Furthermore the method offers for the first time an ability to place accurate numbers to characterise the state or progression of disease, which could be used in a dynamic context if used in serial samples as a function of time.

The method ofthe third aspect ofthe invention represents an improved method for detection of analytes from minor populations of cells in a number of fields. These improvements include: (a) increasing the sensitivity of pre-existing assays; (b) potential application to the analysis of single cells which may be mandatory for stem cell analyses; (c) definitive assessment of thresholds for diseases and infectious agents, including the assessment of host response; (d) hastening of reaction and ease of interpretation; (e) reducing ambiguity of result assessment and implementing a new regime of assessment by sequential analysis; and (f) improved mass spectrometry delivery systems; (f) PCR in presence of inhibitory substances - crude extracts relying upon dilution to remove the effect ofthe inhibitory substance; and (g) general flow cytometric analyses.

The mixed population of cells may include cultured cells and biological tissue samples, such as biopsies or blood samples, and samples of other biological fluids such as lymph, sputum, cerebrospinal fluid, and the like. Additionally, non-clinically related applications in forensic sciences, genetic modified organisms, environmental and toxicological testing would further benefit from the invention.

The method of the third aspect of the invention is particularly applicable to analysis of biological samples of interest that comprise substantially single cells free in the aqueous reaction mixture. To this end this method of the invention may further comprise dispersing cells from a biological sample of interest to provide a population of substantially single cells.

It will be appreciated that droplets may be formed that contain whole cells (in which case subsequent lysis may be required to release assayable cell contents) or droplets may be formed from cells that have been previously ruptured in the aqueous reaction mixture. The aqueous reaction mixture may advantageously comprise an agent capable of lysing biological cells, in order to better permit entry of the other components ofthe reaction mixture into cell nuclei present in the sample. Simple alkali lysis followed by neutralisation, application of detergents, proteolytic enzymes or use of somcation or high voltage are procedures successfully applied to releasing the nucleic acids from whole cells. The procedure is also amenable to accepting nuclear material from fixed paraffin wax embedded tissues.

It will also be appreciated that a droplet may be formed which contains a cell (or cell contents) and none, some or all of the assay reagents required to perform the assay. When some or all of the assay reagents are absent, it may be necessary to introduce further reagents into the droplet. This may be achieved by introducing a further droplet carrying the necessary missing reagents into the carrier fluid and allowing the first and further droplets to fuse to allow mixing of a full complement of the essential assay reagents and cells (or cell contents). By doing so, specific numbers and even types of cells can be isolated and then subjected to lysis and subsequent PCR amplification and analysis.

It is preferred that the methods of the first and third aspects of the invention represent the application of discrete microfluidic technology to control the amount of a chemical or chemicals or number of cells, or amount of cell content, contained within each droplet. Since the aqueous reaction mixture is present in the form of discrete volumes supported in an immiscible carrier fluid each droplet provides reaction products that may be analysed independently from the products of other droplets. Alternatively the droplets may be coalesced for combined analysis (e.g. on a down-stream gene array). The carriage of droplets within an immiscible carrier fluid prevents merger of discrete volumes of the reaction mixture, and hence ensures that there is no contamination (for example by diffusion or other means) between a given volume of the aqueous reaction mixture and similar preceding, or following, droplets. The sensitivity of the method of the invention is regulated by controlling the volume of the aqueous mixture used to form the droplet. With the third aspect of the invention, it is preferred that each droplet only contains a few cells and most preferred that each droplet contains a single cell. Since the contents of each droplet remain separate during subsequent assay procedure each assay result will be representative of the characteristics of the single cell. Furthermore the reaction products in each droplet may be specific to a single cell from a minor population. Accordingly the minor population may be differentiated from the other cells types on a cell by cell basis. Accordingly the prior art failings of sensitivity (e.g. resulting from cross-contamination by other cells which reduces the capability to detect populations of cells present at low copy number within a tissue) is obviated.

The concentration of reaction mixture and cells (or cell contents) in each droplet and also the volume of each droplet introduced into the carrier fluid may be controlled to permit the separation of cells by type andor into clusters of defined numbers. Whilst it is preferred that each droplet contains a single cell, it will be appreciated that the third aspect of the invention provides considerable benefits over the prior art even in the case that the droplet contains up to ten cells, up to a hundred cells, or more. It will be appreciated that pre-testing for a specific assay set-up will confirm the detection limit of one test cell amongst a maximum excess of other non-target cells.

When the method involves the generation of cell lysates that are subsequently made into droplets, the cells may be treated to release their contents within a reservoir containing the aqueous reaction mixture. The dilution of cells into the solution in the reservoir will be known and hence the expected concentration of cell contents calculated. The dilution will be controlled such that one copy (or at least a defined number) of a particular cell content will be present in a droplet. Hence alterations in copy number will be amenable to detection. Thus aqueous reaction mixtures comprising cell lysates, rather than whole cells, may be used to generate droplets according to the invention.

It will be appreciated that, since the droplet is transported within the carrier fluid, the analysis of the reaction mixture for the presence of the assay product may be performed at a site remote from the site at which the droplet is introduced into the carrier fluid.

Transport of the droplet may be achieved either by movement of the carrier fluid (such that it transports the droplet) or by independent movement of the reaction mixture with a static carrier fluid.

The droplets may be introduced into an immiscible moving carrier fluid, supported in a microfluidic flow manifold, by a number of means. The simplest procedure is to induce outflow of the aqueous fluid from a channel branch, or jet, or other orifice that feeds into the carrier fluid. Pressure, or positive volumetric displacement are well known means of inducing flow of the droplet into the carrier fluid. When this process is controlled, both the volume of the droplet and the frequency of successive droplets may be achieved with a level of precision that ensures that the discrete droplets/droplets avoid cross-contamination, with either each other or through intermittent conduit wall contact. The inventors have demonstrated that the droplets so formed will tend to travel at the centre of a microfluidic conduit filled with a carrier fluid, thus avoiding wall contact.

A preferred method by which such outflow may be achieved is through the application of pressure, or positive volumetric displacement, to the aqueous reaction mixture in a reservoir. This may be acliieved by means such as displacement of a plunger in a barrel containing the aqueous reaction mixture. Alternatively the displacement may be achieved by changing the volume of a reservoir containing the aqueous reaction mixture, for example by applying an electric current to a piezo electric material in which a channel (suitable for holding the aqueous reaction mixture) has been formed.

When the droplets are moved through a static immiscible carrier fluid, means of transporting droplets in the static immiscible carrier fluid include physical phenomena such as electrostatics, dielectrophoresis, ultrasonic agitation and thermally-induced convection.

An alternative method of inducing the requisite generation of droplets is by electrical droplet generation. The presence of buffer salts in the aqueous sample phase will mean that the sample has moderately high conductivity, in contrast with the carrier fluid which should have significantly lower conductivity than the sample, and preferably be a good insulator. According to this embodiment two or more side channels will be disposed opposite or adjacent to each other across the main conduit containing a flowing carrier fluid. One will contain the aqueous sample, while the other will be filled with a solid conductor, which may be a metal, metal-filled polymer, conducting carbon or graphite or a conducting carbon-filled polymer. When an electrical potential difference is applied between the two side channels, an attractive electrostatic force will be generated between the conducting contents of the two side channels. Since the solid conductor cannot move, only the aqueous sample will move into the main channel, where it will be subject to shear forces from the flow ofthe carrier fluid. When enough sample has entered the main channel, it will shear off from the end ofthe side channel and be carried along as a charged droplet.

The force experienced by the sample at the end ofthe side channel is given by:

ε_rε₀AV² d²

Where F is the force, ε_r is the dielectric constant ofthe carrier fluid, ε₀ is the permittivity

ofthe vacuum, A is the cross-sectional area ofthe side channels, Fis the applied potential difference and d is the distance across the main conduit between the ends of the side channels. We can also derive the pressure exerted at the end ofthe side channel:

A d²

Where P is the pressure. It can be seen from this equation that the pressure is independent of the cross-sectional area of the side channel. To generate a droplet, it is necessary that the pressure exerted on the aqueous sample exceeds the internal pressure of the carrier fluid. If we take a typical fluid pressure of 0.1 atmospheres (10,000 Pa) and a fluid with a dielectric constant of 2.5 and a side channel separation of 100 μm, we find that to balance the internal pressure will require an applied voltage of -2377 volts. Any higher voltage will result in the formation of droplets. The applied voltage can be applied continuously, which will result in a continual stream of droplets at a rate controlled by the separation of the aqueous sample from the end of the side channel (the maximum droplet generation rate), or can be pulsed to provide droplets on demand.

For ease of droplet generation it is preferred that two power supplies be used. One will provide the bias voltage, which will just balance the internal pressure and stop the carrier fluid from entering the side channel, and one that will be pulsed to generate droplets. It should be noted that the pressure depends on the square ofthe applied voltage, so to double the pressure will require only a 41% increase in applied voltage. An adjustable 2.5kN power supply may be used to provide bias and a pulsed lkN power supply to generate the droplets. Alternatively, similar results may be achieved through the application of a suitable high voltage AC waveform.

In addition to simple droplet generation, this scheme can be extended to provide droplet mixing. If the solid conductor in one side channel is replaced with a second aqueous phase, two oppositely-charged droplets will be generated, leading to rapid merging/fusing ofthe two droplets into one. This allows reagents to be added to a sample after the primary droplet has been generated. Similarly, if additional sets of side channels are added to the conduit further downstream, two or more reagents may be added to the primary droplet. It is preferred that the assay procedure according to the third aspect of the invention involves some sort of amplification ofthe analyte to be detected from the minor cell population. In this respect polymerase chain reaction (PCR) amplification and detection ofthe amplified PCR products (the analyte according to the invention) represent a preferred assay procedure according to the method ofthe third aspect ofthe mvention.

The quantitative measurement of minor nucleic acid populations within large heterogeneous nucleic acid populations cannot be accurately achieved using conventional PCR. The inventors have demonstrated that, if too highly diluted, PCR cannot detect low copy number aberrant nucleic acid sequences, and therefore, for example, current minimal residual disease thresholds are related to and limited by assay sensitivity rather than being a true reflection ofthe threshold level above which a cancerous clonal population requires remedial action. One of the great benefits of preferred, PCR based, methods according to the third aspect of the invention is that the methods allow true and reproducible quantitative measurement of nucleic acids from minor cell populations and thus reflects actual levels of minimal residual disease.

It will be appreciated that PCR primers may be selected that only amplify nuclear material (such as fusion gene transcripts) found in the minor cell population.

The primers may be designed to specifically lead to the amplification of a mutation in an oncogene that transforms a cell in cancer. The primers may also be designed to allow amplification of a transcript for a protein that is only expressed in the cancerous state. It will be appreciated that methods of the third aspect of the invention that utilise PCR and such primers are particularly useful for making a prognosis for cancer patients or in the diagnosis of cancer. Such primers are particularly useful for making assessments of minimal residual disease (MRD) in cancer patients (e.g. clinical assessment of reoccurring clonal populations in leukaemia).

Alternatively the primers may be designed to amplify a microbial gene foimd in a pathogen but not in a host organism. It will be appreciated that methods ofthe third aspect of the invention employing such primers can be used to detect a pathogen in a multi- cellular orgamsm. Accordingly the method is useful in the diagnosis of microbial infections (whether protozoa, fungal, bacterial or even viral). For instance preferred methods ofthe third aspect ofthe invention may be used to diagnose malaria.

The aqueous reaction mixture may also contain nucleotides, primers and a DNA polymerase (e.g. Taq polymerase). Droplets comprising the aqueous reaction mixture may then be injected into the carrier fluid which then flows into a PCR reactor. Real-time optical detection may be performed on the outflow from the PCR reactor to enable the dynamics ofthe reaction to be recorded.

The use of limited cell numbers in each PCR reaction is of great importance as this enables a minor nucleic acid component to be detectible from within a high number of other nucleic acid types.

In order to expose the aqueous reaction mixture to the temperature cycling required to perform the PCR reaction the temperature of the bulk carrier fluid may be varied thereby inducing PCR in the aqueous reaction mixture. However, it is preferred that the droplets are introduced into a PCR reactor wherein they are transported through zones of different temperature in order to perform the polymerase chain reaction. Such zones of different temperature may be formed within a "PCR chip". In such chips a circuitous conduit, carrying the carrier fluid and aqueous reaction mixture, passes through a solid substrate. Different zones ofthe substrate may be heated to different temperatures, and the temperature of the aqueous reaction mixture controlled by controlling the length of time that the reaction mixture spends in the different heated zones. Kopp et al. (Science Nol 280 p 1046 - 1048 , 1998) disclose a PCR chip that may be adapted for use according to the invention. It will be appreciated that the Kopp PCR chip should be adapted, as discussed in more detail below, such that there are: (a) means for introducing droplets of aqueous reaction mixture into the carrier fluid (the carrier fluid may then flow through the chip to allow PCR reactions to occur in each droplet); and (b) means for detecting PCR product in each droplet.

Such devices represent an important aspect of the invention. Therefore according to a fourth aspect ofthe invention there is provided a PCR reactor comprising: a substrate with different zones that are heatable to defined temperatures and the substrate bears a circuitous conduit adapted to carry an aqueous-immiscible carrier fluid that passes through the different zones; a reservoir which is connected to the conduit with means to introduce droplets of an aqueous reaction mixture from within the reservoir into the carrier fluid within the conduit, the droplet being supported by the carrier fluid; and a detector capable of identifying PCR products within individual droplets which have travelled through the different zones within the conduit.

The PCR reactor according to the fourth aspect of the invention is particularly useful for carrying out the method according to the third aspect ofthe invention.

The PCR reactor may serve as either a standalone device or may be used for integrated applications in both routine and research fields. For instance, the device may be used in the fields of prognostics, diagnostics, toxicology, security systems and forensic sciences. The device is highly suited to rapid, sterile, reproducible, integrated and cost effective multi-parallel or serial sensitive quantitative analyses.

A prefened embodiment of the PCR reactor is adapted for application to a high throughput experimental scheme, to enable many discrete PCR amplifications and associated diagnostic assays. This may be achieved by discrete microfluidics, where the PCR reaction, sample and detection components are contained within single droplets supported in the carrier fluid (e.g. an immiscible fluid phase, such as an oil). The droplets can be generated such that their diameter is significantly less than the cross-section ofthe conduit into which they are introduced, hi this way, the droplets may be transported along the channel structure defined on the substrate without touching the conduit walls and without being forced to merge with similar preceding or following droplets. This feature is particularly important in ensuring that there is no contamination of one PCR reaction (within each droplet) by another, and thereby no cross-talk between the discrete PCR samples. It is most preferred that the PCR reactor according to the fourth aspect of the invention is adapted for use in miniaturised integrated analysis systems and particularly in Micro Total Analytical Systems ( TAS). The term " TAS" encompasses miniaturised integrated analysis systems. These systems allow the production of meaningful data from raw biological samples without the requirement for manipulation by an operator.

Preferred PCR reactors according to the fourth aspect of the invention allow the introduction of a biological sample into the reactor according to the principles of μTAS. Cell or nucleic acid dilution may be regulated in the reservoir and the volume of droplet introduced into the carrier fluid such that enhanced sensitivity of detection and increased assay quantitation is achieved according to the method ofthe third aspect ofthe invention. This contrasts markedly with conventional quantitative PCR procedures biasing large quantities of input template to increase detection ofthe marker nucleic acid. The inventors believe that the conventional approach is incorrect because they have found that sensitivity increases with respect to detecting a minor nucleic acid population within an excess of other nucleic acids when the method of the third aspect of the invention is followed. The procedure is able to also eliminate the averaging effect of PCR to permit the detection of cells possessing different levels of a nucleic acid that is common to all cells, for example a proto-oncogene that becomes chromosomally amplified, or disease- related de-regulated or de-no vo expression of a transcript.

The cells may be treated to release their nucleic acid contents within a reservoir of the PCR Reactor containing the aqueous reaction mixture. The dilution of the solution in the reservoir is known and hence the expected concentration. The dilution is controlled such that one copy of a particular gene will be present in a droplet. Hence alterations in copy number will be detected easily. Furthermore RNA may be quantified (rather than DNA) and therefore the magnitude of change i.e. copy number of a particular transcript, will be amplified. RNA may be quantified using the same methodology as used to quantify DNA (e.g. a modified Kopp Chip) but will also employ an upstream reverse transcriptase to convert mRNA into cDNA. For example Obeid et al. (J.Anal Chem 2003 Jan 15;75(2):288-95)) disclose a direct development of the Kopp chip incorporating an initial reverse transcription phase. Thus aqueous reaction mixtures comprising cell lysates, rather than whole cells, may be used to generate droplets according to the invention.

The substrate may be comprised of messenger RNA or (mRNA) added to a reverse transcriptase enzyme, nucleotide triphosphates and an appropriate buffer. One temperature incubation (typically 42°C) and then straight into PCR (see Obeid et al supra).

Droplets within the conduit of the Reactor must be subjected to temperature cycling for the PCR process to achieve the desired amplification. This may be achieved through known means of bulk temperature cycling, or through transport of the droplet stream, along the conduit, through a meandering channel that re-enters the temperature cycle as many times as required (in this case, the temperature cycle is realised through differential heating of two, or three, zones within the supporting substrate structure). When the PCR Reactor has at least two zones it is preferred that a first zone is kept at about 95°C and a second zone kept at about 60°C. When the PCR Reactor has three zones it is preferred that the first zone is kept at about 95°C; the second zone kept at about 77°C; and the third zone kept at about 60°C.The zones may be heated by any means that generates a stable temperature, but preferably through the electrical heating of a conductive element which incorporates a feedback mechanism to ensure maintenance of a stable and fixed pre-selected temperature.

In operation, the temperature zones may be supplied with the same PCR aqueous fluid with common sample. The droplets may be continually generated for a fixed time period or for the generation of a pre-determined droplet number. An alternative to this is to introduce a statistical criterion that forces the prolonged production of droplets until sufficient data have been collected to realise a preset data quality of the data ensemble. This approach has the distinct advantage of economy of both PCR sample and assay time, and ensures the achievement of a pre-determined data quality. There are other possible means of application whereby the discrete droplets comprise different samples, thus providing a significantly high throughput of different samples. Clearly, it is also possible to provide any combination of multiple samples and multiples of the same sample to achieve the desired PCR-related data. The droplet outlet ofthe conduit manifold, either in single or multiparallel embodiment, may be manipulated to waste, or may be collected, sorted, combined or stored in suitable array formats.

It is preferred that the detector detects fluorescence of PCR reaction products. An epifluorescence configuration may be used, providing a set of detection zones (e.g. one per thermal cycle). In one of the heater zones on the substrate a set of holes may be provided for GRIN (gradient index) lenses. These lenses may be approximately 2mm in diameter. GRIN lenses are convenient for this application in that they are cylindrical in shape and do not require complex machining to provide a secure mount. The lenses will be chosen to bring collimated light incident on the outside face of the GRDST lens to a focus just beyond the inner face ofthe lens, in the main conduit. As droplets pass through the focus, fluorescence will be excited. Any fluorescence emission falling on the end of the GRIN lens will be collimated on exit from the outer face of the lens. A dichromic filter may be fitted to direct only the fluorescence emission onto a multi-channel photomultiplier. Up to 32 channels of fluorescence can then be monitored simultaneously. It should be noted that the PCR Reactor of this embodiment of the invention is limited to 32 thermal cycles. It may be useful to omit the first few thermal cycles from the detection system, as there will generally be too little PCR product to detect.

For parallel multi-channel PCR systems, a different approach may be taken. Since there will be many more detection zones, the use of individual GRIN lenses may be considered by some to be prohibitively expensive and difficult to fabricate. In this case, a thin transparent heater plate would be used, allowing the whole of the device to be monitored optically. An electron-multiplying single photon counting CCD camera or other suitable detector would be used to detect the droplet fluorescence at each detection zone. A holographic beam splitter would be used to direct the laser to the required detection zones. A most preferred PCR reactor according to the invention comprises a single channel device with on-chip droplet generation, off-chip carrier and sample reservoirs and pumps, two-sided heating (two-zone PCR) and GRIN lens array for detection.

According to a fifth aspect of the present invention there is provided a method of analysing for the presence of a nucleic acid sequence of interest within a tissue of interest, the method comprising: i) obtaining an aqueous reaction mixture comprising nucleic acids from biological cells from the sample of interest; ii) adding polymerase chain reaction (PCR) reagents suitable for amplifying the nucleic acid sequence of interest; iii) introducing droplets of the aqueous reaction mixture into a carrier fluid immiscible with the aqueous reaction mixture such that a single copy of the nucleic acid of interest is contained within the droplet, the droplet being supported by the carrier fluid; iv) performing the polymerase chain reaction to produce, in the droplet, an amplification product representative of the presence of the nucleic acid sequence of interest; and analysing the aqueous reaction mixture for the presence ofthe amplification product.

PCR based methods according to the third aspect of the invention are particularly useful in analysing for the presence of cancer cells within a biological tissue. At the early stages of disease, cancer cells may represent only a small proportion of the total number of cells present in a tissue, the majority of cells being non-transformed and healthy. The ability to analyse tissue samples, such as biopsies, for the presence of cancer cells is important in diagnosis and treatment of the disease. It is desirable to be able to detect cancer cells at very low levels within tissues, as found at the early stages ofthe disease, in order to allow timely administration of appropriate therapy. PCR based methods are also particularly useful for helping a clinician to test a biopsy sample to evaluate whether or not the tissue is free of cancerous cells (i.e. minimal residual disease (MRD) analysis). Conventional methods of amplifying nucleic acid sequences lack the requisite sensitivity to detect relatively low numbers of aberrant cells, a problem which is overcome by the method ofthe invention.

Indeed, so useful is the method ofthe mvention in the detection of cancer cells in a biological sample that, according to a sixth aspect of the invention there is provided a method of analysing for the presence of cancer cells in a biological sample comprising: i) obtaining an aqueous reaction mixture comprising the biological sample; ii) adding polymerase chain reaction (PCR) reagents suitable for amplifying a nucleic acid sequence indicative of cancer cells; iii) introducing a volume of the aqueous reaction mixture comprising a defined number of cells into a volume of a carrier fluid immiscible with the aqueous reaction mixture to form a droplet of the aqueous reaction mixture in the carrier fluid, the droplet being supported by the carrier fluid; iv) performing the polymerase chain reaction to produce an amplification product representative ofthe presence ofthe nucleic acid sequence indicative of cancer cells; and v) analysing the droplet for the presence of the amplification product; wherein the presence of the amplification product indicates that the biological sample contains cancer cells.

PCR based methods according to the third or fifth aspects ofthe invention are also able to detect the presence of pathogens, such protoza, fungi, bacteria or viruses, which may also be present at relatively low numbers in a biological sample.

Thus, according to a seventh aspect ofthe invention there is provided a method of analysing for the presence of a pathogen in a biological sample comprising: i) obtaining an aqueous reaction mixture comprising the biological sample; ii) adding polymerase chain reaction (PCR) reagents suitable for amplifying a nucleic acid sequence indicative ofthe pathogen; iii) introducing a volume of the aqueous reaction mixture comprising a defined number of cells into a volume of a carrier fluid immiscible with the aqueous reaction mixture to form a droplet of the aqueous reaction mixture in the carrier fluid, the droplet being supported by the carrier fluid; iv) performing the polymerase chain reaction to produce an amplification product representative of the presence of the nucleic acid sequence indicative of the pathogen; and v) analysing the droplet for the presence ofthe amplification product; wherein the presence of the amplification product indicates that the biological sample contains pathogens. The methods according to the fifth, sixth or seventh aspects of the invention are preferably carried out using the PCR reactor according to the fourth aspect of the invention.

The present invention will be further illustrated with reference to the following Examples in which Example 1 illustrates that a mixture of cells in a sample is detrimental to the sensitivity of conventional assays for detecting a minor population of cells in the sample, Example 2 illustrates droplet formation by volumetric displacement of fluid at a microfluidic junction, Example 3 investigates droplet formation on demand using a high voltage pulse, Example 4 illustrates a microfluidic device according to one embodiment ofthe present invention, its incorporation into a PCR microreactor and investigates the flow behavior of droplets through the device, Example 5 investigates droplet merging, Example 6 illustrates a PCR microreactor according to another embodiment ofthe invention and studies fluorescent detection ofthe contents ofthe droplets and Figure 7 illustrates the use ofthe device in carrying out PCR on a DNA sample, and with reference to accompanying drawings for which: Figure 1 is a schematic illustrating the importance of analysing from lαiown cell populations when using PCR; Figure 2 is a schematic illustrating that quantitative analysis of nucleic acids from heterogeneous cell populations is required for the identification of disease thresholds; Figure 3 is a schematic illustrating the use of flow cytometry to preselect cells types for nucleic acid purification prior to PCR; Figure 4 is a schematic illustrating preferred means for generating droplets according to the invention; Figure 5 is a schematic diagram of droplet formation at a T-junction; Figure 6 shows images of droplet formation at a T-junction; Figure 7 is a graph illustrating the effect of varying oil flow rate on droplet formation; Figure 8 is a graph illustrating the effect of varying water flow rate on droplet production; Figure 9 is graph illustrating the effect of varying oil flow rate on average droplet volume; Figure 10 is a graph illustrating the effect of varying water flow rate on average droplet volume; Figures 11a, 1 lb and 1 lc are respectively images of water droplets produced with an oil flow rate of 1041.67 nl/s and water flow rate of (a) 83.33 nl/s (b) 166.67 and (c) 250.00 nl/s; Figures 12a and 12b are respectively an image and a schematic diagram of a two- phase microfluidic chip; Figures 13 and 14 are images of a fluidic chip with a tapered channel; Figure 15 is an image of a fluidic chip having the water pressure balanced with the pressure ofthe flowing oil; Figure 16 illustrates droplet formations using a voltage of +500v on a water inlet; Figure 17 is a schematic drawing of a microfluidic chip according to another embodiment ofthe present invention, showing the oil and sample inlets; Figure 18 illustrates the flow behaviour of droplets at the corners ofthe channels ofthe microfluidic chip shown in Figure 17; Figure 19 is an exploded drawing of a fabricated PCR reactor according to one embodiment ofthe present invention; Figure 20 shows a video sequence ofthe thermal behaviour of Chromazone droplets in contact with different heat zones, the top surface is at 60°C and the bottom surface at 90°C with a flow speed of 2cm/sec; Figure 21 is a schematic diagram of a microfluid chip for droplet merging according to another embodiment ofthe present mvention; Figure 22 is a video sequence illustrating droplets forming in a micochannel from two inlets provided in a microfluidic chip; Figure 23 is a video sequence illustrating droplets merging in the microchannel of the microfluidic chip; Figure 24a illustrates the channel design of a chip according to a further embodiment ofthe invention; Figure 24b illustrates the chip shown in Figure 24a with heater blocks, GRIN lenses dichroic mirror and 16 channel PMT; Figure 25 is a single frame of a video show showing a string of droplets meandering through a chip ofthe present invention. Figure 26 is a schematic diagram ofthe optical set up for each channel ofthe PCR chip; Figure 27 is a plot of fluorescence intensity (a.u) versus time (10^"4 sec) for fluorescin containing droplets at the end of cycle 7 and 8; and Figure 28 is a schematic illustrating the steps ofthe polymerase chain reaction. EXAMPLE 1 Experiments were conducted to demonstrate that a mixture of cells in a sample is detrimental to the sensitivity of conventional assays for detecting a minor population of cells in the sample. This illustrates that the usefulness of the methods and PCR reactor according to the invention for increasing the sensitivity of detection of a minor population of cells within a heterogeneous mixture of cells.

Biological material typically comprises several differentiated cell types, and many cell types may be present in such low numbers that they become difficult to detect. By way of example, in the case of cancer, if a particular oncogene is to be analysed as either aneuploid or as a transcriptional product, Fig 1 shows schematically how different mixed populations of cells might be represented with respect to a specific oncogene. Fig 1 also indicates how PCR tends to provide an averaging effect. Whilst this phenomenon is satisfactory for correctly ascertaining gene copy number from within homogeneous cell populations, current applications of PCR are unable to detect minor cell populations within heterogeneous cell populations and tend to deliver a result resembling the average gene content per cell. Even if a unique sequence is to be analysed it is highly likely that reaction sensitivity and sample error will produce a false low or even negative reaction result.

The methods of the present invention demonstrate that following careful sample handling and treatment prior to PCR, in-vitro amplified results can be more meaningfully interpreted. Fig 2 further demonstrates the concept. The upper section of the figure represents how present working practice attempts to determine the presence of a sub- population of cells (via a marker gene) from an unknown total population of mixed cell types. This situation contrasts markedly with the one portrayed in the lower part of figure 2 where the total mixed cell population entering the reaction is known, leaving only the amount of cells comprising the sub-population to be determined by PCR experimentation. If the total number of cells per reaction is set to not exceed the detection sensitivity to permit nuclear material from a single cell to be detected within the excess of non-target cell nucleic acids, the invention suggests that a dilution of target sequences will effectively increase assay sensitivity. The invention is therefore able to alter assay sensitivity by both a change in concentration of nucleic acid offered to the reaction and the volume used per reaction, and the whole assembly is in dynamic flow and is amenable to parallelisation.

1.1 Methods

1.1.1 Mixing of cell types and pre-PCR cell preparation. An oncogene, MYCN was chosen as a model system as it is amplified at the DNA level in cancer cell lines and therefore PCR assays do not need to undertake in vitro transcription.

Suspensions ofthe neuroblastoma tumour cell line LAN 1 which possess some 80 copies of the oncogene MYCN DNA were used either directly from cell culture or following storage at -80°C in 10% DMSO. Cells were prepared for FACS using

propidium iodide nuclear staining (CycleTEST PLUS^®, Becton Dicldnson, San Jose,

CA), and were sorted according to nuclear content using a FACStar PLUS^® (Becton

Dickinson, San Jose, CA). 10 cells were collected directly into 96 well PCR plates and 1 ul of 0.1 M NaOH was added for 6 minutes at 75°C followed by 9 ul of 0.02 M Tris pH 7.5. As shown in Table 1, DNA known to be normal for the analyte oncogene was added from 0 to 10,000 equivalents. A schematic of the extraction procedure is shown in Fig 3 which had been previously validated (low Standard deviation value) using replicate analysis of 8 or more cells per PCR.

1.1.2 Real-time PCR Primer sequences and TaqMan™fluorescence resonance energy transfer probes for

MYCN and β globin are shown in Table 1, and were designed using Primer Express™ software (PE/ABI, Foster City, CA).

Table 1. Sequences of PCR amplimers and TaqMan probes

Name Sequence (5' - 3') Orientation Description

MYCN2F GCCGAGCTGCTCCACGT (+) MYCN exon 2 forward

MYCN2R TCAAACTCGAGGTCTGGGTTCT (-) MYCN exon 2 reverse

MYCN2P -ACCATGCCGGGCATGATCTGC-r -^ (+) MYCN exon 2 TaqMan probe βF CCCATCACTTTGGCAAAGAATT (+) β globin forward βR CACCAGCCACCACTTTCTGA (-) β globin reverse βP F/C-CCCCACCAGTGCAGG-CTGCCT-K4 &4 (+) β globin TaqMan probe

Firstly suitable FRET probe sequences were determined then the flanking forward and reverse amplimers were selected by applying the criteria suggested in the TaqMan^® Universal 2 x PCR Master Mix handbook (Perkin Elmer Applied Biosystems P/N 4304437). MYCN probes used FAM as the 5' reporter dye and the β globin probe used NIC as the 5' reporter dye, the 3' quenching dye was TAMRA or QSY-7 (Molecular Probes, Germany), β globin was chosen as a reference gene as it is present on chromosome 11 which is highly stable within cells of neuroblastoma tissues. The recommended rapid reaction optimisation procedure was adopted for all real-time PCRs which used 2x master mix. A final reaction volume of 25 ul and the standard temperature cycling profile (50°C for 2 min, 95°C for 10 min, and 40 x 95°C for 15 sec and 60oC for 1 min) was used in all investigations. Real-time PCR data were collected using an ABI Prism 7700 SDS (Sequence Detection System, Perkin Elmer Applied Biosystems, Foster City, CA). Analysis was perfonned using the ROX passive reference dye for SDS software to calculate the threshold cycle (C_T) values and also the standard deviation

values from replicate samples. From the duplexed reactions the C_T for β globin reference

gene value was subtracted from the C MYCN value to give the log value ΔCT- The

conversion into a linear value was achieved by applying the formula 2^"ACτ (ABI PRISM

7700 User Bulletin #2). The size of real-time PCR products was assessed following 2% agarose gel (EP -0010-05, Eurogentec, Geneva, CH) end point measurement of 8 ul of reaction mixture after 40 PCR cycles.

1.2 Result/Discussion

The data for the mixing of neuroblastoma MYCN-amplified and DNA normal for MYCN are shown in Table 2.

Table 2. Analysis of a minor cell population within mixed cell populations; a necessity to analyse reduced cell numbers

Detection of MYCN amplified cells within populations of MYCN normal cells

SK LAN 1 cells have approx 80 DNA copies of MYCN per cell

A calculation ofthe difference between the amount of MYCN relative to β globin

is derived by applying the formula 2^" _T to the cycle threshold (C_T) values obtained following PCR. The analysis of only normal cells give a value of 0.32 and 10 cancer cells selected from a homogenous population show that the relative oncogene abundance rises 500 fold to present a value of 162.02. The addition of 1 normal cell to the 10 cancer cells could not be detected, but when an equivalence of 10 normal and 10 cancer cells were analysed the calculated oncogene increase dropped to 21.56, and again to 1.12 when 100 normal cells were added to 10 cancer cells, at which point the influence ofthe cancer cell sub-population was over 3 fold compared to the normal homogeneous population (1.12 vs 0.32). Increasing the abundance of normal cells relative to the fixed cancer cell population produced a result that made the presence of the cancer cell population indiscernible when compared to the homogeneous normal population.

This model study demonstrates that somewhere between 100:10 and 1,000:10 (100 to 10: 1) normal to cancer cells that the real-time PCR assay is no longer able to distinguish the presence of the amplified oncogene. These are relatively modest ratio differences and are cause for concern as many current testing may be yielding too low results. Typically quantitative gene assessment work of this type is based on transcriptional studies where the abundance of for example oncogene transcripts can be at the several 1,000's per cell. These studies are often referred to as minimal residual disease investigations and currently therapy in leukaemia is in part directed by the analysis of abenant transcripts detected within a population of 10,000 cells. These studies suggest that this may well be the detection limit imposed by dilution/inhibition of the target sequences. It will therefore be appreciated that the scheme of sample dilution and packaging in droplets according to the methods of the invention improve sensitivity and provide for a more quantitative approach to diagnostiscs and other assay requirements.

EXAMPLE 2 Figure 4 is a schematic illustrating various preferred means for vesicle or droplet generation, namely pressure driven, pressure, electrically driven and electrically driven droplet merging. The formation of droplets by volumetric displacement at a microfluidic junction, as shown in the first schematic of Figure 4, was investigated. Kloehn N6 syringe pumps were used to inject a dispersed and continuous phase flow of water C into an oil flow A B, as shown in Figure 5 of the accompanying drawings. The pumps had a resolution of 48000 steps and a minimum pump speed of 40 steps per second. A 100 μl syringe was used for the water and a 250 μl for the oil.

A standard CCD video camera fitted with a 4x microscope objective was used to image the droplets fonned. The camera had a frame rate of 30 fps and the frame size was 320 x 240 pixels. The droplet volume was estimated by counting the number of droplets produced in 5 seconds and calculating the average droplet volume based on the flow rate ofthe water.

Droplets were produced at the T-junction as shown in Figure 6.

The droplet production rate was measured with varying oil and water flow rates (Figure 7 and Figure 8). This data was then used to calculate the average volume of the droplets (Figure 9 and Figure 10). It was found that droplet production rate increased with both increasing water flow rate and increasing oil flow rate. The average droplet volume however increased with increasing water flow rate and decreased with increasing oil flow rate. Some examples of droplets produced are shown in Figures 11a, lib and lie. Note that each droplet is fully supported by the carrier fluid.

EXAMPLE 3 Whilst the method of Example 2 did generate droplets of controllable size, it only produced them as a constant stream. A method that could produce droplets into the continuous phase on demand is desirable as it would offer more confrol over the manipulation of the droplets. The droplets could be produced at different points in a microchannel network in a time fashion so that they collide and coalesce, before being transported to a different area of a microchip to be reacted further. For this reason a method was investigated to produce droplets on demand by using a high voltage pulse to push a controlled volume of water into the channel containing the flowing oil.

A chip was fabricated as in Example 2 but with an extra channel, filled with conducting epoxy, directly opposite the smaller water inlet.

The fluidic structures were fabricated on a 75 x 75 x 6mm sheets of polmethyl methacrylate (PMMA) using a precision CNC machine (Datron CAT3D), Datron GmbH, Germany) with 100 and 200 um diameter end mills. The cross section of the channels were rectangular and between 100 and 200 um deep and between 100 and 500 um wide.

Figures 12a and 12b shows a two-phase flow microfluidic chip. The chip consists of an oil inlet 10, a water inlet 12 and an added electrode 13. All inlets are comiected via a 100 um wide and 100 um deep channel to the main channel that is 30mm long and 200 um deep. The electrode 13 consists of a thermosetting silver loaded epoxy (RS Ltd, UK) that is filled into a milled channel opposite the water inlet. This channel was milled and filled with silver epoxy before the main channel was milled to produce a smooth transition between the electrode and the main channel. Further downstream two more inlets 14 and 15 intersect the main channel for creating droplets of different reagent. The fluid connectors were made by 1mm access holes drilled through the chip. Threads were cut to fit tube connectors (062 Minstan tube fitting, The Lee Company, USA) for easy connection to the pumps.

Figures 13 and 14 show a fluidic chip with a tapered main channel. The main channel, which is 200 um deep, is narrowed down from 300 um to 150 um and is then intersected by the 100 um deep and 100 um wide water inlet and the silver epoxy electrode. The chip was sealed with a second sheet of Perspex in that a mirror image of the main channel is milled but which is only 100 um deep forming a total of 300 um. This method ensured that the droplets injected from the water inlet (100 um deep) flowed into the centre ofthe main channel to form spheres and were kept away from the channel walls.

hi this experiment while a syringe pump was used to inject the oil as before, it was not used to supply the water. Instead the water reservoir was placed at such a height so that the water pressure is balanced with the pressure of the flowing oil (see Figure 15). This was to ensure that the oil did not flow into the water inlet but also so that the water did not flow into the oil channel. The water inlet and conducting epoxy were then connected to a high voltage supply. It is thought that the application of a high voltage pulse of around 2kN would overcome the dielectric force of the oil, causing a volume of water to be ejected into the flowing oil, creating a droplet on demand. The silver electrode was attached to a Brandenburg High Voltage Power supply and the water inlet was attached to ground via a piece of metal HPLC tubing which was inserted into the water inlet tube. It was hoped that applying a suitably high voltage to the silver electrode would overcome the dielectric forces of the oil and cause the water to be pushed into the oil channel.

A standard CCD video camera fitted with a 4x microscope objective was used to image the voltage effects. The camera had a frame rate of 30 φs and the frame size was 320 x 240 pixels.

The effect of applying voltages from 200 N to 500N for 10 seconds was investigated. This was carried out with the silver electrode set as both the anode and the cathode. The results are detailed in Table 3 below: Table 3

Figure 16 illustrates the effect of a voltage of +500N on the water inlet.

As shown in the table, the water was drawn into the oil channel when voltages of 300N or greater were applied to the silver electrode.

This demonstrated that water could be drawn into a flowing oil channel using an electrode connected to a high voltage power supply. However, the power supply controls were not sensitive enough to allow droplets to be created on demand. Therefore, a computer controlled electronic set-up should be employed which will produce high voltage pulses. These will be timed so that only enough water to form one droplet will be pulled into the oil channel, thus creating a droplet on demand.

EXAMPLE 4 A microfluidic device for enabling droplets to meander through different temperature zones was developed and the flow behaviour of droplets therethrough was studied.

A flow chip 38 was fabricated by milling into polycarbonate sheets of 4mm thickness and an overall dimension of 75 by 75 mm.

In order to achieve a good thermal conductivity between heater elements 42, 44

and sample the chips were successfully sealed with a thin acetate sheet 46 (100 μm

thickness) and a thin layer (approx. 5 μm) of casted epoxy (Norland 68, US). This UN

curable epoxy shows very good adhesion and solvent resistance and withstands temperatures from -150°C to +125°C. Casting epoxy between two sheets of acetate 46 and separating them at dry ice temperatures results in a good surface quality important for subsequent scattering and fluorescent measurements. Thermal experiments were conducted to test the behaviour ofthe seal at high temperatures (up to 100°C).

The aqueous sample is injected from a 100 x 100 μm injector into the oil

containing channel with the dimension of 200 x 200 μm forming droplets of certain sizes

according to relative pressures at the inlets.

In the prototype chip 38 this main channel opens up to 400 x 500 μm via a tapered

region in order to slow down the flow (see Fig. 17). Initial concerns that droplets might touch the channel walls at the 90 degree corners were not a problem. Droplets of different sizes and with different flow speeds were generated and no contact with the channel walls were observed, as shown in Figure 18. hi this image, the droplet volume is approx. 1 nl and the flow speed is 2cm/sec.

The droplets were subjected to temperature cycling achieved through the meandering channel 40 between the heated top 42 and bottom surfaces 44 ofthe flow chip, see Figure 19. In this three dimensional channel structure the droplets oscillate between the two heated zones, whereby the number of cycles is determined by the number of complete oscillations between the two temperature zones.

For visualisation purposes a thermal ink Chromazone (TMC, Flintshire, UK) with a transition temperature of T= 66+/-2 °C was injected into the oil to form the droplets. Fig.20 shows one meander in a sequence of video frames. The top surface is heated to 60 °C and the bottom surface heated to 90 °C. In this sequence black droplets (flowing from left to right) turn into white ones on the hot surface and turn back to black subjected to the colder surface. Experiments were carried out to investigate the thermal characteristics of the droplets due to temperature gradients and flow speeds. It was noted that the flow behaviour of the droplets under heated conditions did not change despite the change in viscosity ofthe oil.

EXAMPLE 5 In a different experiment a flow chip was developed to allow droplets to be merged together as they flow through a microchannel network (see Fig. 21). The water and oil were gravity fed from open syringes held on clamp stands. Short pieces of HPLC tubing, of i.d. 1mm and approximately 1cm long, were inserted into both water inlet tubes, approximately 2cm from the Lee fittings connecting them to the chip. One inlet was connected to the ground and one to a High Voltage Power Supply Unit via crocodile clips attached to these pieces of metal.

The two water inlets were then balanced so that the oil was flowing freely but the water was stationary at the edge ofthe main channel, as shown in the first frame of Figure 22.

Droplets were formed from each inlet by switching on the Power Supply Unit as has been reported previously. However, as the inlets were separated by the lOOμm wall they were not able to merge immediately (Fig 22.) It would appear that the droplets become either positively or negatively charged, depending on the inlet that they are produced from. This results in an attractive force strong enough to overcome the surface tension of the droplets once they reach the end of the wall and are able to merge (Fig. 23).

EXAMPLE 6 An alternative prototype PCR chip 50 was made and tested. It comprises a 16 thermo-cycle channel system and was fabricated as described in Example 4. The main channel meanders 52 between two different temperature zones provided by a heater block 54 at 95°C at the bottom and a heater block 56 at 60°C at the top ofthe chip and has a total length of almost 380 mm (see Figures 24a and 24b). This channel is made using a 500 um end mill tool creating a width of 500um and a depth of 300um. Injecting the sample from a 100 x 100 um channel into the oil filled channel produces the droplets. The droplet volume could be varied between lnl and 10 nl. Droplet speeds were measured from approx. 5mm/sec to several cm/sec. During thermo cycling the sample enclosed in the droplets will therefore be exposed approx 3 sec to one temperature with the current design of 16mm channel length at the bottom and the top. Figure 25 shows droplets flowing through the PCR chip, at the bottom and the top ofthe chip. The flow of droplets proved to be very stable, tested in a continuous run for more than 2 hours. Also, due to laminar flow conditions in the main channel, the droplets stay centred in the channel all along the way tlirough the chip, a very important fact that enables us the fluorescence of every individual droplet to be monitored at the end of each thermo cycle.

The optical set up for performing the fluorescence measurement was as follows: The lower temperature block 54 was equipped with 16 GRIN (Gradient Refractive Index, Newport, UK) lenses 58 to focus excitation light into the flow channel (see Figure 26). These lenses have a radially variable index of refraction that causes an optical ray to follow a sinusoidal propagation path through the lens. They offer a high Numerical Aperture (light collection ability) of NA=0.46 but with very small dimensions (Diameter: 1.8mm, Length: 4.26mm) and can be directly inserted into the heater block avoiding a complicated lens setup. The correct working distance ofthe lenses have been calculated to take the refractive index steps ofthe air/seal and the seal/oil/sample into account. Reflection losses, which occur on any interface of different refractive index as in this case the oil/water interface, are kept minimal since the rays ofthe focused laser light are striking the droplet surface perpendicular as illustrated in Fig.26. Emission from a fluorophore-containing droplet is collected and collimated with the same GRIN lens. The emitted light is than separated from the excitation light through a 45° dichroic mirror (Comar, UK) and finally imaged onto one ofthe 16 Photomultiplier elements (Hamamatsu, Japan). Imaging is realised by an "achromised" lens system containing two convex and one concave lens which has been designed using WinLens 4.2 (Linos, Goettingen, Germany)

The laser beam itself is split into 16 individual beams 60 of almost equal intensity (in line) via an acrylic holographic beam splitter 62 (Photonics & Analytical Marketing, Leeds, UK). The beams are collimated to match the GRIN lens arcay with a pitch of 2 mm. Apart from the Ar+ -Laser (6mW @ 488nm, LaserGraphics, Germany) and the dual syringe pump (Harvard, US), the setup is housed in a box with all the crucial optics parts mounted onto an optical rail. Initial fluorescence measurements were performed in order to test and calibrate the system. Sample solutions of fluorescin were prepared to create droplets of well-defined amounts of a known flourophore.

Data was recorded using a 16 channel A D converter 64 (National Instruments, US) and sfreamed to the computer hard drive in order to be analyzed. A comprehensive software (C++, Borland Builder) will be developed to analyse each ofthe recorded channels for quality of experiment and intensity of fluorescent signal. A fast Fourier Transform (FFT) will be performed on every detector channel to ensure droplets are delivered with the same volume and frequency to insure quality is maintained during thermo-cycling and the ability to quantify the amount of fluorescent light after every thermo- cycle will enable the experiment to perform real-time PCR measurements.

Figure 27 shows an example of detected fluorescence of two neighboring detector elements for cycle 7 and 8. The sampling rate during recording was 10kHz for all 16 individual channels. This plot shows a scan of 4.5 sec, the droplets are produced with approx. 6 Hz. Droplet size was estimated to 4nl and the concentration ofthe sample fluorescin solution was lxl 0^"3 M.

Initial results were as follows:

1. The plot ofthe two neighboring channels showed the "phase shift". It was observed that the phase of droplets remains stable even till the end ofthe cycle. A FFT will give a real time control ofthe phase.

2. There was no "cross talk" noticeable in the signal due to improved imaging optics. 3. Intensity did vary, probably caused by the syringe pumps working at low speed.

Smaller diameter syringes will deliver a more constant flow and will eliminate fluctuations.

The system may be calibrated and its sensitivity enhanced to allow detecting fluorescence ofthe PCR reaction products per thermal cycle at real time.

EXAMPLE 7 A prefened application for the device and method of the present invention is the amplification of a DNA segment from a complex DNA mixture using the polymerase chain reaction which employs heat to produce the cyclic denaturation of double-stranded DNA into single-strands at high temperature, and the subsequent re-association of complementary sequences at lower temperature into double strands.

An aqueous sample containing the DNA to be amplified is injected into an oil- filled channel to produce a droplet that is fully supported by the oil. A further aqueous sample containing the PCR reagents, namely oligonucleotides (forward or reverse amplimers or primers), DNA polymerase and deoxynucleotides is then introduced into the oil-filled channel to merge with the droplet containing the DNA sample. The merged droplet is then passed through the different heat zones of the chip of allow denaturation, annealing and extension. Each of these steps is governed by temperature with denaturation being highest, annealing lowest and extension mid-way, although often the annealing and extension phases are combined, as illustrated in Figure 28. The process starts with the target DNA sample that is typically double-stranded and possesses the sequence of the gene of interest. Two synthetic oligonucleotides (termed forward and reverse amplimers or primers) are selected using software to ensure that they bind with appropriate specificity. They are chosen to each have complementary to opposite strands ofthe target gene, but also their orientation is such that uni-directional 5' to 3' DNA synthesis generates the DNA that exists between the two primers. The primer extension product from one primer serves as a template for the other primer to hybridise and initiate DNA synthesis. The main product from the PCR is thus the exponential production of a gene fragment flanked at both ends by the 5' sequence ofthe forward and reverse primers.

The determination of gene or transcript quantities within a sample is heavily compromised by the heterogeneity of biological tissues. Conventional quantitation of nucleic acids delivers the average gene value per cell, such that a minor cell population would produce an undetectable contribution to the overall measured signal.

By dividing a sample of known concentration into smaller aliquots a predetermined number of cells can thus be made to enter a droplet. When this droplet is mixed with appropriate PCR reagents by for example droplet fusion, effective quantitative in-vitro gene amplification can be monitored in real-time. Therefore a system of PCR based around droplet generation offers a means to integrate raw sample processing through to quantitative real-time PCR detection. The process has the advantage of continuous flow high throughput, is highly reproducible, quantitative, greatly reduces the risk of contamination in addition to time and cost savings.

Claims

1. A method of monitoring and/or manipulating at least one chemical entity, the method comprising: (i) obtaining an aqueous reaction mixture comprising at least one chemical entity; (ii) introducing a volume of the aqueous reaction mixture into a volume of a carrier fluid immiscible with the aqueous reaction mixture to form a droplet ofthe aqueous mixture supported by the carrier fluid; and (iii) monitoring and/or manipulating the chemical entity within the droplet.

2. A method as claimed in claim 1 wherein the droplet is transfened within the carrier fluid.

3. A method as claimed in claim 1 or claim 2 wherein additional reagents are added to the droplet by merging of droplets containing different reagents.

4. A method according to any preceding claim, wherein the droplet is introduced into the carrier fluid by positive displacement.

5. A method according to claim 4, wherein the droplet is formed by displacement of the aqueous reaction mixture from a reservoir by means of a plunger in the reservoir.

6. A method according to any of claims 1 to 3, wherein the droplet is introduced into the carrier fluid by electrostatic means (13).

7. A microfluidic device comprising: a substrate (38) with different zones (42, 44) that are subjectable to defined conditions, the substrate bearing a continuous conduit (40) adapted to carry an aqueous immiscible carrier fluid that passes through the different zones; a reservoir which is connected to the conduit with means to introduce droplets of an aqueous reaction mixture from within the reservoir into the carrier fluid within the conduit, the droplet being supported by the carrier fluid; and, optionally, a detector capable of identifying products within individual droplets which have travelled through the different zones within the conduit.

8. A device according to claim 7 wherein the means for introducing droplets of an aqueous reaction mixture from within the reservoir into the carrier fluid introduces the droplet by positive displacement.

9. A device according to claim 8, wherein the means is a plunger in the reservoir.

10. A device according to claim 7, wherein the means for introducing droplets of an aqueous reaction mixture from within the reservoir into the carrier fluid applies a potential difference between the reservoir connection with the conduit and the opposite wall ofthe conduit to allow introduction ofthe droplet by electrostatic means.

11. A device according to any one of claims 7 to 10 wherein the detector is an optical detector in the wall ofthe conduit.

12. A device according to claim 11 wherein the optical detector comprises a Gradient Index (GRIN) lens capable of directing excitation light onto the droplets contained within the carrier fluid, and collecting any emitted fluorescence and directing it on to a suitable detector.

13. A device according to claim 11 wherein the detector is an electron-multiplying single photon counting CCD camera within a detection zone of the device and wherein the detection zone is defined by a thin transparent heater plate.

14. A method of detecting analytes from a minor population of cells within mixed populations of cells, the method comprising: i) obtaining an aqueous reaction mixture comprising analytes from a mixed population of cells; ii) adding assay reagents suitable for detecting analytes from the minor population of cells of interest to the aqueous reaction mixture; iii) introducing a volume of the aqueous reaction mixture into a volume of a carrier fluid immiscible with the aqueous reaction mixture to form a droplet ofthe aqueous reaction mixture supported by the carrier fluid; iv) performing some or all of an assay, in the droplet, wherein an assay product is representative of the presence of the minor population of cells of interest; and v) analysing the droplet for the presence ofthe assay product.

15. A method according to claim 14, wherein the mixed population of cells is derived from a biological sample of interest comprising substantially single cells.

16. A method according to claim 14, further comprising dispersing cells from a biological sample of interest to provide a population of substantially single cells to form the mixed population of cells.

17. A method according to any one of claims 14 to 16, wherein the volume of the droplet is regulated such that it contains a single cell.

18. A method according to any one of claims 14 to 17, wherein the aqueous reaction mixture further comprises an agent capable of permeabilising biological cells.

19. A method according to any one of claims 14 to 18, wherein the aqueous reaction mixture further comprises an agent capable of lysing biological cells.

20. A method according to any one of claims 14 to 19, wherein the droplet is introduced into the carrier fluid by positive displacement.

21. A method according to claim 20, wherein the droplet is formed by displacement of the aqueous reaction mixture from a reservoir by means of a plunger in the reservoir.

22. A method according to any of claims 14 to 19, wherein the droplet is introduced into the carrier fluid by electrostatic means.

23. A method according to any one of claims 14 to 22, wherein the assay is the polymerase chain reaction and the assay reagents comprise PCR primers, nucleotides and a DNA polymerase.

24. A method according to claim 23, wherein the carrier fluid containing the droplets is transported through zones of different temperature in order to perform the polymerase chain reaction.

25. A method according to any one of claims 14 to 24, wherein the aqueous reaction mixture further comprises signal molecules capable of indicating the presence ofthe assay product.

26. A method according to claim 25, wherein the signal molecules are capable of inducing fluorescence in the presence ofthe assay product.

27. A method according to claim 26, wherein the signal molecules are FRET partners.

28. A method according to any one of claims 14 to 27 wherein the assay reagents are added to the aqueous reaction mixture before droplet formation.

29. A method according to any one of claims 14 to 27 wherein the assay reagents are added to the aqueous reaction mixture after droplet formation.

30. A PCR reactor comprising: a substrate (38) with different zones (42, 44) that are heatable to defined temperatures and bearing a circuitous conduit (40) adapted to carry a carrier fluid that passes through the different zones; a reservoir which is connected to the conduit with means to introduce droplets of an aqueous reaction mixture from within the reservoir into a carrier fluid within the conduit, the droplet being supported by the carrier fluid; and a detector capable of identifying PCR products within individual droplets which have travelled through the different zones within the conduit.

31. The PCR Reactor according to claim 30 with at least two zones wherein a first zone may be kept at 95°C and a second zone may be kept at 60°C

32. The PCR Reactor according to claim 31 with three zones wherein the first zone may be kept at 95°C; the second zone may be kept at 77°C; and the third zone may be kept at 60°C.

33. The PCR Reactor according to claims 30 - 32 wherein the means for introducing droplets of an aqueous reaction mixture from within the reservoir into the carrier fluid introduces the droplet by positive displacement.

34. The PCR Reactor according to claim 33, wherein the means is a plunger in the reservoir.

35. The PCR Reactor according to any one of claims 30 - 34, wherein the means for introducing droplets of an aqueous reaction mixture from within the reservoir into the earner fluid applies a potential difference between the reservoir connection with the conduit and the opposite wall of the conduit to allow introduction of the droplet by electrostatic means.

36. The PCR Reactor according to any one of claims 30 - 35 wherein the detector is an optical detector in the wall ofthe conduit.

37. The PCR Reactor according to claim 36 wherein the optical detector comprises a Gradient Index (GRIN) lens (58) capable of directing excitation light onto the droplets contained within the carrier fluid, and collecting any emitted fluorescence and directing it on to a suitable detector.

38. The PCR Reactor according to claim 36 wherein the detector is an electron- multiplying single photon counting CCD camera within a detection zone of the Reactor and wherein the detection zone is defined by a thin transparent heater plate.

39. A method of analysing for the presence of a nucleic acid sequence of interest within a tissue of interest, the method comprising: i) obtaining an aqueous reaction mixture comprising nucleic acids from biological cells from a sample ofthe tissue of interest; ii) adding polymerase chain reaction (PCR) reagents suitable for amplifying the nucleic acid sequence of interest; iii) introducing droplets of the aqueous reaction mixture into a carrier fluid immiscible with the aqueous reaction mixture such that a single copy of the nucleic acid of interest is contained within the droplet that is supported by the carrier fluid; iv) performing the polymerase chain reaction to produce, in the droplet, an amplification product representative of the presence of the nucleic acid sequence of interest; and i) analysing the aqueous reaction mixture for the presence of the amplification product.

40. A method of analysing for the presence of cancer cells in a biological sample comprising: i) obtaining an aqueous reaction mixture comprising the biological sample; ii) adding polymerase chain reaction (PCR) reagents suitable for amplifying a nucleic acid sequence indicative of cancer cells; iii) introducing a volume ofthe aqueous reaction mixture comprising a defined number of cells into a volume of a carrier fluid immiscible with the aqueous reaction mixture to form a droplet of the aqueous reaction mixture in the carrier fluid, the droplet being supported by the carrier fluid; iv) performing the polymerase chain reaction to produce an amplification product representative of the presence of the nucleic acid sequence indicative of cancer cells; and v) analysing the droplet for the presence ofthe amplification product; wherein the presence of the amplification product indicates that the biological sample contains cancer cells.

41. A method of analysing for the presence of a pathogen in a biological sample comprising: i) obtaining an aqueous reaction mixture comprising the biological sample; ii) adding polymerase chain reaction (PCR) reagents suitable for amplifying a nucleic acid sequence indicative ofthe pathogen; iii) introducing a volume ofthe aqueous reaction mixture comprising a defined number of cells into a volume of a carrier fluid immiscible with the aqueous reaction mixture to form a droplet of the aqueous reaction mixture in the carrier fluid, the droplet being supported by the carrier fluid; iv) performing the polymerase chain reaction to produce an amplification product representative of the presence of the nucleic acid sequence indicative of the pathogen; and v) analysing the droplet for the presence ofthe amplification product; wherein the presence of the amplification product indicates that the biological sample contains pathogens.

42. A method according to any one of claims 39 - 41 conducted in a PCR reactor according to any one of claims 30 — 38.