FRAGMENTATION OF CHAINS OF NUCLEIC ACIDS

Abstract

Disclosed are methods and devices for fragmenting chains of nucleic acids (such as DNA) in a liquid sample. A liquid sample is provided, comprising chains of nucleic acids. A sample treatment device has a sample treatment zone. The liquid sample is contacted with the sample treatment zone. Surface acoustic waves (SAWs) are propagated along a surface of the sample treatment zone, or more generated acoustic waves are propagated to couple with the sample, and/or the sample is subjected to freeze-thaw cycling, in order to cause fragmentation of said chains of nucleic acids in the sample.

Claims

1. A method of fragmenting chains of nucleic acids in a liquid sample, the method comprising: providing a liquid sample comprising chains of nucleic acids; providing a sample treatment device, the sample treatment device having a sample treatment zone; contacting said sample with said sample treatment zone; generating and propagating surface acoustic waves (SAWs) along a surface of the sample treatment zone, said SAWs coupling into the sample to cause fragmentation of said chains of nucleic acids in the sample.

2. The method according to claim 1 wherein the liquid sample has volume V, an area of an interface between the sample and the sample treatment zone being area A, wherein the ratio A/V is at least 1000 m.sup.2/m.sup.3.

3. The method according to claim 1 or wherein the sample treatment zone includes an area having roughness Rz at least 10 m.

4. The method according to claim 1 wherein the sample treatment zone includes an array of cavities, being ordered or non-ordered, the cavities cumulatively containing at least part of the sample, optionally all of the sample.

5. The method according to claim 1 wherein the sample treatment zone includes an array of pillars, being ordered or non-ordered.

6. The method according to claim 1 wherein the contact angle between the sample and the sample treatment zone is lower than between the sample and a remaining part of the SAW transmission surface, in order to locate the sample.

7. The method according to claim 1 wherein the sample has a volume of not more than 30 L.

8. The method according to claim 1 wherein the concentration of the chains of nucleic acids in the sample is in the range 5-100 ng/L.

9. The method according to claim 1 wherein the SAW transmission surface is a surface of a superstrate coupled to the SAW transducer.

10. The method according to claim 1 wherein the temperature of the sample is controlled so as not to exceed 37 C.

11. The method according to claim 1 wherein the sample is subjected to active cooling.

12. The method according to claim 1 wherein the sample is frozen, or partially frozen, before the start of coupling SAWs into the sample.

13. The method according to claim 1 wherein, when the sample treatment zone is considered as the first sample treatment zone, the device includes an opposing member providing a second sample treatment zone, adapted to be located in contact with the sample opposite the first sample treatment zone, so that the sample is sandwiched between the first and second sample treatment zones, the opposing member being operable to be reciprocated relative to the SAW transmission surface.

14. The method according to claim 13 wherein, when the SAW transducer is considered as the first SAW transducer and the SAW transmission surface is considered as the first SAW transmission surface, the opposing member provides a second SAW transducer adapted to generate and propagate SAWs along a second SAW transmission surface including the second sample treatment zone, for coupling said SAWs into the sample to cause fragmentation of said chains of nucleic acids in the sample.

15. (canceled)

16. A sample treatment device for fragmenting chains of nucleic acids in a liquid sample, the device comprising: a surface acoustic wave (SAW) transmission surface having a sample treatment zone; a SAW transducer adapted to generate and propagate SAWs along the SAW transmission surface including the sample treatment zone, for coupling said SAWs into the sample to cause fragmentation of said chains of nucleic acids in the sample; wherein the device includes an active cooling means in thermal contact with the sample treatment zone.

17. A sample treatment device for fragmenting chains of nucleic acids in a liquid sample, the device comprising: a surface acoustic wave (SAW) transmission surface having a sample treatment zone; a SAW transducer adapted to generate and propagate SAWs along the SAW transmission surface including the sample treatment zone, for coupling said SAWs into the sample to cause fragmentation of said chains of nucleic acids in the sample; wherein when the sample treatment zone is considered as the first sample treatment zone, the device includes an opposing member providing a second sample treatment zone, adapted to be located in contact with the sample opposite the first sample treatment zone, so that the sample is sandwiched between the first and second sample treatment zones, the opposing member being operable to be reciprocated relative to the SAW transmission surface.

18. The sample treatment device according to claim 17 wherein, when the SAW transducer is considered as the first SAW transducer and the SAW transmission surface is considered as the first SAW transmission surface, the opposing member provides a second SAW transducer adapted to generate and propagate SAWs along a second SAW transmission surface including the second sample treatment zone, for coupling said SAWs into the sample to cause fragmentation of said chains of nucleic acids in the sample.

19. The sample treatment device according to claim 16 wherein one or more phononic structures are provided in order to affect the SAW distribution at the sample treatment zone.

20-78. (canceled)

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0170] Embodiments of the invention will now be described by way of example with reference to the accompanying drawings in which:

[0171] FIG. 1 shows a schematic cross sectional view of an embodiment of the present invention in operation.

[0172] FIGS. 2, 3 and 4 show electrographs indicating the change in fragment length of DNA subjected to SAWs under different conditions using a device as shown in FIG. 1. For each electrograph, 9 L of sample containing Genomic DNA (Promega G3041) at a concentration of 25 ng/L was exposed to 4.86 MHz ultrasonic surface acoustic wave radiation. For FIG. 2 the sample was liquid and 2 W transmitted power was applied for 90 s (temperature less than or equal to 4 C.). For FIG. 3 the sample was liquid but a higher power of 5 W transmitted power was applied for 40 s (temperature less than or equal to 8 C.), the shorter time due to nebulisation of samplenote the appearance of fragments peaking at 1292 bp. For FIG. 4 the sample was partially liquid (i.e. partially frozen) while 2 W of transmitted power was applied for 90 s (temperature less than or equal to 2 C.)this condition resulting in a desired peak position of sub 1000 bp. Note that time is exponentially linked to size on the x-axis.

[0173] FIG. 5 shows a schematic cross sectional view of another embodiment of the present invention in operation, using a superstrate.

[0174] FIGS. 6A-6D show different superstrates are shown for use with the arrangement of FIG. 5.

[0175] FIGS. 7 and 8 show electrographs indicating the change in fragment length of DNA subjected to SAWs under different conditions using a device as shown in FIG. 5. The electrograph of FIG. 7 was obtained using a flat Si superstrate as in FIG. 6A. The electrograph of FIG. 8 was obtained using a patterned Si superstrate as in FIG. 6D. For each electrograph, 9 L of sample containing Genomic DNA at a concentration of 25 ng/L was exposed to 4.86 MHz ultrasonic surface acoustic wave radiation. For FIG. 7 the sample was liquid and 12 W transmitted power was applied for 90 s (temperature less than or equal to 30 C.), with the sample in contact with a flat planar silicon surface. For FIG. 8 the sample was liquid and 12 W transmitted power was applied for 90 s (temperature less than or equal to 30 C.), with the sample in contact with a roughened or patterned planar silicon surfacethis condition resulted in a desired peak position of sub 1000 bp.

[0176] FIG. 9 shows a schematic cross sectional view of an embodiment of the present invention in operation, in which the superstrate includes an array of cavities.

[0177] FIG. 10 shows a schematic cross sectional view of a modified embodiment compared with FIG. 9, in which the cavities include additional projections.

[0178] FIG. 11 shows a schematic cross sectional view of another embodiment of the invention, in which the liquid sample is held between the transducer and a superstrate.

[0179] FIG. 12 shows a schematic cross sectional view of another embodiment of the invention, in which the liquid sample is held between two transducers.

[0180] FIG. 13 shows another embodiment of the invention in which the sample is held in an enclosed chamber at the sample treatment zone.

[0181] FIG. 14 shows a modification of the embodiment of FIG. 13.

[0182] FIG. 15 shows a further modification of the embodiment of FIG. 13.

[0183] FIG. 16 shows another modification of the embodiment of FIG. 13.

[0184] FIGS. 17-20 each show an electrograph of samples treated under various different conditions of power, duty cycle and temperature using the embodiment of FIG. 16.

[0185] FIG. 21 shows another embodiment of the invention in which the sample is held in an enclosed chamber at the sample treatment zone, the sample being treated using bulk acoustic waves.

[0186] FIG. 22 shows a modification of the embodiment of FIG. 21.

[0187] FIG. 23 shows a further modification of the embodiment of FIG. 21.

[0188] FIG. 24 shows another modification of the embodiment of FIG. 21.

[0189] FIG. 25 shows another embodiment of the invention in which the sample is held in an enclosed chamber at the sample treatment zone, the sample being treated using bulk acoustic waves.

[0190] FIG. 26 shows a modification of the embodiment of FIG. 25.

[0191] FIG. 27 shows a further modification of the embodiment of FIG. 25.

[0192] FIG. 28 shows another of the embodiment of FIG. 25.

[0193] FIG. 29 shows a part of a sample treatment zone for use in a further embodiment of the invention in which the acoustic wave is a Bleustein-Gulyaev wave.

[0194] FIG. 30 shows a part of a sample treatment zone for use in a further embodiment of the invention in which the acoustic wave is a guided Love wave.

[0195] FIGS. 31-33 show SEM images of arrays of pits formed at the sample treatment zone of a SAW superstrate.

[0196] FIG. 34 shows a mode for using the superstrate of FIGS. 31-33.

[0197] FIG. 35 shows DNA fragment distributions for the superstrate of FIGS. 31-33 at different applied powers.

[0198] FIGS. 36-38 show SEM images of arrays of pillars formed at the sample treatment zone of a SAW superstrate.

[0199] FIG. 39 shows a mode for using the superstrate of FIGS. 36-38.

[0200] FIG. 40 shows DNA fragment distributions for the superstrate of FIGS. 36-38 at different applied powers.

[0201] FIG. 41 shows an SEM image of an array of pillars formed at the sample treatment zone of a SAW superstrate.

[0202] FIG. 42 shows DNA fragment distributions for the superstrate of FIG. 41 at different applied powers.

[0203] FIG. 43 shows a mode for using a further superstrate.

[0204] FIG. 44 shows DNA fragment distributions for a flat superstrate used as shown in FIG. 43 at different DNA sample concentrations.

[0205] FIGS. 45-47 show SEM images of different parts of a roughened Si superstrate.

[0206] FIG. 48 shows DNA fragment distributions for a roughened superstrate used as shown in FIG. 43 at different DNA sample concentrations.

[0207] FIGS. 49 and 50 show SEM images for pits formed in SU8 subjected to different processing conditions.

[0208] FIG. 51 shows an SEM image for a pillar of SU8.

[0209] FIG. 52 shows DNA fragment distribution for the superstrate of FIG. 51.

[0210] FIG. 53 shows an SEM image for a different pillar of SU8.

[0211] FIG. 54 shows DNA fragment distribution for the superstrate of FIG. 53.

[0212] FIGS. 55-57 show SEM images for an array of pillars formed of SU8.

[0213] FIG. 58 shows DNA fragment distributions for the superstrate of FIGS. 55-57 at different applied powers.

[0214] FIGS. 59-61 show SEM images for a trough formed in SU8.

[0215] FIG. 62 shows DNA fragment distributions for the superstrate of FIGS. 59-61 at different applied powers.

[0216] FIGS. 63-65 show SEM images for a strip formed in SU8.

[0217] FIG. 66 compares DNA fragment distributions for the superstrates formed using troughs and strips of different depths.

[0218] FIG. 67 compares DNA fragment distributions for the superstrates formed using troughs and strips subjected to different processing conditions.

[0219] FIG. 68 shows a flow chart outlining a DNA sequencing process, including a fragmentation step according to an embodiment of the invention.

[0220] FIG. 69 shows a plan view of the processing of an interdigitated electrode structure to form a freeze-thaw DNA fragmentation device.

[0221] FIG. 70 shows a perspective view of a sample droplet located at the sample treatment zone of a freeze-thaw DNA fragmentation device.

[0222] FIG. 71 shows a plot of temperature with position across a freeze-thaw DNA fragmentation device during heating.

[0223] FIG. 72 shows a frequency scan of the freeze-thaw DNA fragmentation device using an Agilent vector network analyser (S11 parameter). Marked on the scan is a small trough indicative of small resonance around 32 MHz.

[0224] FIG. 73 shows a screenshot from a Bruker Contour GT white light profilometer scan of the surface of the freeze-thaw DNA fragmentation device.

[0225] FIG. 74 shows the data of FIG. 73 in plan view.

[0226] FIG. 75 shows a plot generated by a Polytec GmbH single point vibrometer (range up to 24 MHz) showing the presence of the first sub harmonic due to the restricted range of the vibrometer used (up to 24 MHz) when excited by a 5V pkpk signal at 32 MHz, indicating some actuation of the surface.

[0227] FIG. 76 shows an electrograph of 9 L of Human DNA (Coriell NA12878) with a concentration of 38 ng\L placed directly onto the freeze-thaw DNA fragmentation device.

[0228] FIG. 77 shows an electrograph of 9 L of Human DNA (Coriell NA12878) with a concentration of 38 ng\L placed onto a smooth glass superstrate on the freeze-thaw DNA fragmentation device.

[0229] FIG. 78 shows an electrograph of 6 L of Genomic DNA (Promega G3041) with a concentration of 43 ng/L placed onto a smooth glass superstrate on a micro strip heater.

[0230] FIG. 79 shows an electrograph of 6 L of Genomic DNA (Promega G3041) with a concentration of 43 ng/L placed onto a structured silicon superstrate (pegs 130 m dia. 160 m high with a pitch of 230 m) on a micro strip heater.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS, AND FURTHER OPTIONAL FEATURES OF THE INVENTION

[0231] In the preferred embodiments of the present invention, DNA such as genomic DNA is subject to treatment using SAWs in order to generate DNA fragments of length particularly suitable for automated sequencing. The use of SAWs allows the use of lower sample volumes and lower powers. The typical size and configuration of SAW transducers also enables the integration of fragmentation into sequencing instrumentation. This enables the implementation of sample preparation pre-sequencing steps within the next generation of sequencing instruments. This allows sequencing to be carried out in one integrated instrument, rather than having a stand-alone fragmenting instrument and a stand-alone sequencer, with a skilled operator required to transfer the DNA fragment sample to the sequencer (as is currently the case). This results in reducing total costs for sequencing, increased automation leading to increased throughput, and a broader uptake of the technique across existing and new sectors. The disclosed approach to DNA fragmentation also enables field-based DNA sequencingas may be required for determining microbial resistance and informing the treatment of infectious disease in the face of the emergence of drug resistance (as seen in rare variants of HIV not identified by traditional genotyping techniques.

[0232] As will be discussed, the approaches disclosed herein allow the use of a planar geometry, which is of particular interest for the development of a cartridge-based approach to fragmentation. The cartridge can be formed, in part, using the transducer, but more preferably, the cartridge may provide the superstrate used in preferred embodiments of the invention, for coupling with a transducer which forms part of a fragmentation apparatus. In this case, it is preferred that the cartridge is disposable.

[0233] In the preferred embodiments of the invention, a liquid sample is placed onto a treatment zone of a SAW transmission surface. The SAW transmission surface supports SAWs in the form of harmonic surface displacements with a frequency of at least 100 kHz, preferably about 1 MHz, and at most 1 GHz or at most 100 MHz. In the most preferred embodiments, the SAW frequency used is in the range 4-10 MHz. SAWs such as Rayleigh waves exist on a solid half space and they exhibit the property of no dispersion. However, other SAWs such as Lamb type waves can be used. Lamb waves are dispersive and this property can be exploited to enhance the amplitude of the surface displacements. The transversal component of these vibrations couple to the liquid sample and radiate compressional waves into the liquid. Due to the difference between the speed of sound in the solid and that of the sample, the compression waves are radiated at an angle which obeys a Snellius type law of refraction. Where there is a free surface of the sample, the longitudinal pressure waves are trapped in the sample due to the acoustic impedance mismatch between air and the liquid and between the liquid and the SAW transmission surface. This is described in more detail in WO 2011/060369, WO 2012/114076, WO 2012/156755 and PCT/GB2014/052672 (WO 2015/033139), the contents of which are hereby incorporated by reference.

[0234] The liquid sample shapes the propagation of the sound energy as the air/water interface is a very good reflector of sound as the acoustic impedance mismatch means that 99.99% of the sound wave gets reflected at the interface. This strong reflection at the air/water interface also creates high pressure waves in the sample. Further to this, because a fluid can change shape in response to acoustic forcing, the pressure wave distribution can vary with time. This variation causes differential flows and enhances the shearing of chains of nucleic acids.

[0235] It is preferred that the liquid sample is cooled in order to suppress loss of material due to nebulization, if the system is an open system. Furthermore, the sample can be frozen prior to fragmentation using surface acoustic waves. A particularly suitable configuration uses a Peltier cooler, in order to maintain the planarity of the system. In the case where the sample includes a two phase system, the rheology of the two phase system may change with temperature, allowing emulsification to occur and/or increasing the miscibility of the two phases, which may be disadvantageous. Therefore, even in a closed system, cooling can still provide a useful additional effect.

[0236] The liquid sample is preferably aqueous. Water can dissolve quantities of gas. This property can be used in order to cause cavitation in the sample due the acoustic strain developed in the liquid. Such acoustic strain can enable the nucleation of bubbles which can then act as cavitation centres. The cavitation centres irradiated by harmonic pressure waves in the medium can be used to further assist in the fragmentation of the chains of nucleic acids. Suitable cavitation centres include surface roughness features. For example, an ordered or non-ordered array of cavities may be produced in the sample treatment zone using deep reactive ion etching, e.g. using the Bosch process. Such an etching process tends to form a scalloped surface to the sidewalls of the cavities, having a nanometre length scale. Such features are suitable nucleation centres. Additionally, the edge/corner where the sidewall meets the base of the cavity provide suitable nucleation centres.

[0237] The liquid sample may be treated in order to dissolve gas into it. The gas may be any suitable gas that dissolves in the liquid and which promotes bubble formation. The liquid may be saturated with the gas, or supersaturated with the gas.

[0238] The devices according to preferred embodiments of the invention use at least one interdigitated transducer (IDT) fabricated on a piezoelectric substrate to generate SAWs such as Rayleigh-Lamb type elastic waves. The SAWs propagate along a SAW transmission surface to a sample treatment zone to couple into the liquid sample. A compressional wave is radiated into the liquid, because the speed of sound in the liquid is slower than at the SAW transmission surface and the compressional wave is refracted at an angle relative to the normal from the SAW transmission surface. The long chain molecules in the sample interact with the compressional waves propagating in the liquid sample and absorb the mechanical energy. This heats the sample. To control the internal heating of the sample, and thus to control unwanted thermal damage to the DNA, an active cooling device with heat sink is used to extract excess heat from the irradiated liquid. Indeed such temperature control can be achieved by using a pulsed mode of operation where high peak powers are used over a short time period and the rest of the duty cycle is used to allow the sample to cool down.

[0239] The acoustic impedance mismatch between air and water is 99.99% implying that almost all the sound wave that impinges on the liquid/air interface is reflected or trapped. This fact allows for energy to be pumped into the liquid and to create pressure at a greater efficiency in the liquid.

[0240] The device may form all or part of a chamber that itself forms part of a cartridge or part of a larger microfluidic device. This is discussed in more detail below.

[0241] A harmonic signal is applied to the transducer, of frequency typically not less than 1 MHz and not greater than 1 GHz. This generates a harmonic surface displacement, this surface displacement will generate accelerations of the order 10.sup.6 ms.sup.2 or higher at the surface of the substrate dependent on the frequency used. Where multiple transducers are used, corresponding signals are applied to the transducers. The magnitude, shape and position of the resultant surface displacements can be controlled by corresponding control of the configuration, frequency, phase and number of transducers.

[0242] Turning now to an explanation of the embodiments shown in the drawings, it is believed that micro flows generated via acoustic streaming in liquid samples containing DNA or other such polymeric or long chain materials are drivers for fragmentation of the long chain material into smaller parts. In the simplest embodiment of the present invention, shown in FIG. 1, an open geometry is used, in which the liquid sample 12 is placed on a sample treatment zone 16 of a LiNbO.sub.3 SAW transducer 14. SAWs generated by the transducer electrodes 18 are transmitted along the SAW transmission surface to the sample treatment zone 16. The drawing is schematic in the sense that the sample droplet appears large. In practice, the sample droplet is much thinner in height, increasing A/V compared with the impression given by FIG. 1.

[0243] One of the drawbacks to using an open geometry for the fragmentation of chains of nucleic acids is the propensity of the sample to nebulise and with it loss of material from the system. This can be overcome to some extent by cooling the sample liquid prior to and subsequent to irradiation by surface acoustic waves. For this reason, as shown in FIG. 1, a cooling system 20 is provided in contact with the lower face of the transducer, in order to extract heat from the transducer and therefore also from the liquid sample.

[0244] Acoustic streaming is a second order effect caused by the propagation or the presence s of acoustic vibrations interacting with a fluid. The streaming can induce rapid counter propagating flows in the sample fluid which can tear apart the strands of DNA or other long chain structures of interest. Bubbles can be a localised source for acoustic streaming.

[0245] In one approach of an embodiment of the invention, the inventors take advantage of the thermodynamic properties of water where the sample is in a partially frozen state. This allows fragmentation to occur at lower powers than otherwise, and a possible mechanism for this is explained below.

[0246] In a specific example of the embodiment schematically shown in FIG. 1, the transducer was based on 1 mm thick Y cut black LiNbO.sub.3 with an electrode spacing to provide a working frequency of 4.86 MHz. The IDT had a dimension of 23 mm square. A Peltier cooler was attached to a fan assisted heatsink using heat sink compound with the transducer attached to the Peltier cooler using the same heat sink compound. This enabled the system to operate in normal ambient temperatures.

[0247] In a straightforward modification, the upper surface of the transducer except for the sample treatment zone was treated such that it became hydrophobic. This makes it easier to recover the sample or to process the sample further directly on the device.

[0248] With reference to FIG. 1, an aqueous liquid sample 12 of volume about 9 L was placed as a drop onto the sample treatment zone 16, located spaced apart from the electrodes 18. The liquid sample contained between 25 ng/L to 100 ng/L of genomic DNA. During operation, the Peltier cooling device was operated in order to control the temperature of the sample during irradiation with the SAWs generated by the IDT. In this work, the Peltier cooler was operated so that the temperature of the sample did not exceed 37 C.

[0249] The surface temperatures of the samples were measured with the aid of a Fluke Ti25 IR camera. The results of the fragmentation were analysed using an Agilent Bioanalyser 2100 with the 12 k kit.

[0250] In order to further control the temperature of the liquid sample, the user can pulse the SAW excitation of the liquid sample. By altering the duty cycle, the ratio of time spent on to the time spent off, the average power can be kept low but the peak powers can be allowed to become high without heating of the sample being a problem. However, this requires that the total time to be extended so that the sample sees the required amount of SAW irradiation. For example a duty cycle of 50% on and 50% off will require double the time of a continuously irradiated sample. Typical times used for a continuously irradiated sample were 30 s to 120 s whereas a pulsed 50:50 irradiated sample required at least 60 s to 240 s.

[0251] FIGS. 2, 3 and 4 show electrographs carried out on samples of volume 9 L containing 25 ng/L of Genomic DNA (Promega G3041). The samples were subjected to 4.86 MHz SAW radiation using the configuration of FIG. 1. For FIG. 2, the sample was liquid and 2 W transmitted power was applied for 90 s. The temperature was less than or equal to 4 C. For FIG. 3, the sample was liquid but a higher power of 5 W transmitted power was applied for 40 s. The temperature was less than or equal to 8 C. A shorter time was used compared with FIG. 2 due to nebulisation of sample. In FIG. 3, note the appearance of fragments peaking at 1292 bp. For FIG. 4, the sample was partially liquid and partially solid while 2 W of transmitted power was applied for 90 s temperature less than or equal to 2 C., this condition resulted in a desired peak position of sub 1000 bp. Note that time is exponentially linked to size on the x-axis.

[0252] The power and the amount of time for which the sample is subjected to SAWs has some influence on the position and shape of the fragmented material's distribution. At low power, fragmentation is not apparent. Only when a threshold power is achieved is suitable fragmentation observed. To illustrate this, consider FIGS. 2-4. FIG. 2 uses 2 W and shows no fragmentation (the temperature used for FIG. 2 was less than or equal to 4 C.). FIG. 3 uses 5 W and shows fragmentation (the temperature used for FIG. 3 was less than or equal to 8 C.). Once fragmentation is achieved, the position of the fragmentation peak is substantially insensitive to the power used and will typically remain at approximately 1200 bp (see FIGS. 2 and 3) until very high peak powers are used (>30 W). However, the duration of exposure can have an influence on the shape of the size distribution. With a short duration (e.g. less than 60 s) there are seen distributions with a symmetrical shape (FIG. 3). This morphs into a wider distribution for samples exposed for more than 90 s, even when the temperature is not so high (FIG. 4, in which the temperature was less than or equal to 2 C.). In sequencing applications, a tight size distribution is preferred.

[0253] The liquid can be frozen or super cooled prior irradiation with ultrasonic surface acoustic waves. The liquid can be cooled such that only partial melting of the frozen drop is achieved on its surface, when subjected to the ultrasonic actuation. An hypothesis as to the mechanism linked to this could be drawn from results discussed below on the use of superstrates, in that the partially melted liquid is subjected to a roughened interface with the frozen parts. Under such conditions, low powers such as 2 W can be used to achieve the desired fragmentation of below 1000 bp. In this context, the term liquid sample is to be understood to include samples which are liquid at room temperature but which may be solidified, e.g. by freezing, and which at least partially liquefy during the process of fragmentation.

[0254] Another embodiment of the invention is illustrated in FIG. 5, which is a modification of the embodiment of FIG. 1. Here, a superstrate 22 is coupled to a transducer 14 such that the mechanical wave can propagate from the transducer to the superstrate 22. In this case, it is the superstrate 22 which provides the SAW transmission surface of interest. The superstrate also provides the sample treatment zone. Parts of the superstrate (other than the sample treatment zone) can be treated to make then hydrophobic, in order to aid to collection of exposed sample. In the examples based on FIG. 5, the sample was cooled prior to treatment using SAWs. During the application of SAWs, the sample melts if frozen, heats up and spreads over the sample treatment zone surface. Depending on the volume of sample, the liquid can spread to form a thin film over the surface. Such a thin film is liable to nebulise and this is to be avoided as there will be loss of material if this is allowed to occur. Ways to avoid nebulisation are to use lower powers, larger volumes or control the temperature of the liquid sample during the fragmentation process by pulsing the excitation. A further way to avoid nebulisation is to use an enclosed sample chamber, which is discussed in more detail below.

[0255] FIG. 5 shows a superstrate 22 in direct contact with a surface acoustic wave transducer. Different variations for the superstrate are shown in FIGS. 6A-6D. FIG. 6A shows a planar superstrate 22A which is smooth and flat. FIG. 6B shows a roughened or patterned superstrate 22B where the sample is in contact with a smooth planar part of the superstrate. FIG. 6C shows a roughened or patterned superstrate 22C where the sample is in partial contact with the roughened or patterned part of the superstrate. FIG. 6D shows a roughened or patterned superstrate 22D where the sample is in contact with the roughened or patterned superstrate only. At the time of writing, it is considered that variations used in FIGS. 6C and 6D are particularly suitable.

[0256] At different operating frequencies, varying behaviour of the liquid sample can be observed. At lower frequencies, there is significant movement of the drop while at higher operating frequencies the drop can be made to vortex with less translational motion and a similar rate of heating in the liquid sample. This movement can be controlled by modifying the surface chemically by changing the surface chemistry locally or physically by introducing a surface topology which augments the planar surface. The present inventors have used a periodic arrays of pits. In one example a hexagonal array was used with pit diameter 70 m, depth 50 m and centre-to-centre spacing 200 m. In another example a hexagonal array was used with pit diameter 140 m, depth 70 m and centre-to-centre spacing 250 m. The inventors observe that they could have used a random array of such pits or even a mechanically roughened surface formed using abrasive techniques. It is considered that one effect of the surface treatment is that the contact line of the liquid is pinned such that enough power can be applied to fragment the material of interest while having control where the liquid goes. However, the liquid can be allowed to move to the far edge of the superstrate, distal from the electrodes where it will remain throughout the fragmentation process, in effect pinned at the edge.

[0257] FIGS. 7 and 8 show electrographs of 9 L of fragmented genomic DNA samples with a concentration of 25 ng/L exposed to 4.86 MHz ultrasonic surface acoustic wave radiation. For FIG. 7, the sample was liquid and 12 W transmitted power was applied for 90 s. The temperature was less than or equal to 30 C., with the sample in contact with a flat planar silicon surface. For FIG. 8, the sample was liquid and 12 W transmitted power was applied for 90 s. The temperature was less than or equal to 30 C., with the sample in contact with a roughened or patterned planar silicon surface. This arrangement resulted in a desired peak position of less than 1000 bp.

[0258] Another advantage of a roughened surface is that it increases the efficiency of the fragmentation process, achieving a desired distribution peak value less than 1000 bp at relatively low applied powers (12 W instead of >30 W). This has the advantage of controlling the temperature of the fragmentation process to be less than 37 C., preferably less than 20 C., thus avoiding any issues of heat stress for any biological samples. The efficiency is achieved by enabling streaming flows to occur adjacent to or on the pitted surface enabling higher shear to occur than would be present on an otherwise flat planar surface.

[0259] FIGS. 9 and 10 show modified embodiments in which the superstrate is adapted in different ways. The Peltier cooler is not shown, but can be incorporated as described above.

[0260] In FIG. 9, an array of cavities 30 is provided in superstrate 22E which hold the sample 12. In this embodiment, the level of the free surface of the sample is below the top of the cavities 30, but it is possible instead for the sample to overfill the cavities. In operation, the cavities are pumped with acoustic energy to amplify the pressures and streaming flows in the sample. Instead of cavities, the superstrate can employ an arrangement of pillars where the sample is free to flow around. The pillars act as scatter sites and as such can create areas of enhanced pressure gradients and hence streaming flows. These structured features of the sample treatment zone are considered to act as sites for the promotion of bubble nucleation.

[0261] As shown in FIG. 10, an array of structures 32 can be included on the side walls of the pillars or cavities 30 in order to induce more streaming flows in the fluid sample. FIG. 10 represents a combination of 2D phononic crystal structures to form a 3D phononic crystal structure. The 3D structure can be used to shape the sound field generated by the coupled SAWs into the structure. This embodiment promotes cavitation within the phononic crystal by increasing the sound amplitude and therefore the acoustic strain.

[0262] The superstrate may comprise one or more phononic structures in order to affect the distribution of SAWs at the sample treatment zone. A SAW phononic structure is a structure designed to influence the propagation or distribution of SAWs. These phononic structures may be provided as an array of scattering sites or as one scattering site. Details of different arrangements of phononic structures are set out in WO 2011/060369, WO 2012/114076, WO 2012/156755 and PCT/GB2014052672.

[0263] Suitable phononic structures can be incorporated in the device in a number of ways. For example, they may be directly attached to the surface of the transducer. They may be constructed using solid material or using a number of gas bubbles in the fluid sample containing biologically relevant polymeric material. The phononic structures may be formed directly onto a superstrate where the phononic structure can be constructed out of solid material or depending on the nature of the superstrate could constructed out of a number of gas bubbles held in place for example by capillary forces due to surface chemistry or surface geometry. The phononic strucures may be formed inside the transducer or superstrate, for example embedded into the superstrate as layers of material with different density and elastic modulus such as an array of fluidic channels.

[0264] A simple embodiment of a phononic structure is a thin layer of metal deposited on the surface of a piezoelectric material. The metal shorts out the electrical component associated with the harmonic mechanical deformation of a traveling wave and this has the effect of slowing down the propagation of the SAW. By slowing down the propagation of the leading edge of a traveling wave there rest of wave bunches up, increasing the amplitude of the displacement in an analogous manner to a tsunami. By using this effect the effectiveness of the fragmentation process can be improved by increasing the surface displacement over a small distance and which may cause faster acoustic streaming flows. These flows can be broken up into adjacent regions from an incoming SAW by simply using narrow strips of metal. Multiple transducers operating independently can be used to cause strong acoustic streaming counter flows within the sample. Further, it is known that the use of metallic patterning of the piezoelectric surface can be used to fabricate other dispersive structures such as lenses, beamsplitters and or prisms.

[0265] Such displacement increases can be achieved in other ways. For example, they can be achieved by the addition of material onto the surface of either a transducing substrate or the surface of a superstrate where the speed of sound of the added material is lower than that of substrate or superstrate. Polymers such as SU8, glass and/or aerogels can be considered. Such changes in phase velocity can be achieved by using the dispersive properties of a plate (superstrate) where the phase velocity is dependent on the frequency thickness product. By making the superstrate thicker the phase velocity of a particular mode of propagation can be made to slow down. Note however that the A0 mode is an exception to this rule. When an A0 mode is excited in the superstrate then if the thickness of the superstrate was gradually increased, the mode would propagate at progressively higher velocities until it reached the Rayleigh limit. Another simple way to increase the surface displacement is to position the sample at an edge of another discontinuity that causes a significant reflection.

[0266] Suitable phononic structures include phononic crystal (PhnC) structures or grating structures such as pits, holes, troughs, strips or pillars. Here holes, pits and troughs are considered as type 1 and pillars and strips as type 2. Such structures are dispersive and their transmission properties are frequency dependent. Suitable structures can be designed based on the behaviour required, such as reflection where no Bloch-Floquet modes exist or transmission where the frequency chosen to drive a transducer will couple to Bloch-Floquet modes that can exist or propagate in the PhC structure.

[0267] Type 1 PhnC structures comprise pits, holes or troughs in the transducer or superstrate for example, or in a layer which comprises part of a multilayer superstrate structure. The pits, holes or troughs can behave as cavities supporting particular modes of vibration on the structure or in a fluid and these can be considered closed cavities. The pressure in the cavity can be higher than that of a surrounding fluid. This can be used to improve the probability that fragmentation will occur. However, if desired the pits, holes or troughs can be arranged so as to create an acoustic cavity to enhance the sound field in a particular area. Thus, the whole phononic crystal array can support (Bloch-Floquet) modes and a cavity can be created from a PhnC and excited with SAW. In this manner, a structure can be designed to operate on different length scales, the holes being excited individually at a high frequency or the whole structure at a lower frequency.

[0268] Type 2 PhnC structures are those composed of pillars or strips positioned on a surface. The frequencies used can be chosen such that transduced waves from the transducer can couple to Bloch-Floquet modes of the PhnC structure where the scattering of mechanical waves combine to form high pressure point within the spaces between the pillars or strips. Again structures or frequencies can be chosen such that no Bloch-Floquet modes exist therefore reflecting the sound energy. The structures can be arranged to create a cavity to enhance the sound pressure field in a particular area.

[0269] There is an intermediate case of the above two where the PhnC structure is embedded into either a transducer or a superstrate. For example arrays of channels may be embedded into the transducer or superstrate. Pillars may be provided in such channels. Therefore, dispersive elements can be engineered to control the shape of the sound field in a similar manner to lenses, beam splitters and prisms.

[0270] One problem with using pillars is the high contact angle to water that the structures present. However, the inventors have shown through experiments that such high contact angles can be overcome by the use of SAWs transmissions on a suitable substrate or superstrate where the sound can couple into the liquid and within a short time the liquid wets the pillars and is subsequently dragged down to the structure via capillary action. The use of ultrasound causes the drop to cast micro-droplets which then change the wetting characteristics of the structure. Once the structure is wetted then the power applied can be increased.

[0271] Further embodiments of the invention are now described with reference to FIGS. 11 and 12.

[0272] In FIG. 11, a superstrate 22 is provided, but this sandwiches the sample 12 between the transducer 14 and the superstrate 22. The effect of this is that the sample can be more readily contained, and nebulisation reduced or prevented. As shown in FIG. 11, the superstrate 22 can be moved relative to the transducer. Containing the sample in this way allows the ultrasonic waves in the sample (coupled from the SAWs from the transducer) to pass into the superstrate 22. The provision of the additional surface for inducing streaming aids the fragmentation of the chains of nucleic acids. In a modification of this embodiment, the perimeter of the sample may be frozen in order to further limit nebulisation loss from the sample.

[0273] In FIG. 12, FIG. 11 is modified so that the superstrate is a transducer superstrate 34, having electrodes 36. This allows the phase, frequency, amplitude and duty cycle of each transducer 14, 34 to be altered in order to further control the fragmentation in the liquid sample 12.

[0274] In another embodiment (not illustrated), an intervening superstrate can be inserted between the opposing transducers. In this configuration, the intervening superstrate may be narrower or wider than one or both of the opposing transducers.

[0275] In FIG. 11, SAWs couple into the sample 12 and are then transmitted to the superstrate 22 where surface displacements at both the substrate and superstrate interact in order to fragment the chains of nucleic acids. As shown in FIG. 11, the superstrate 22 may be translated relative to the substrate (transducer 14) in order to improve the efficiency.

[0276] In a modification of FIG. 11, another embodiment (not illustrated) uses a circular substrate sandwiching the sample between the superstrate and the transducer. The circular substrate can be rotated relative to the transducer. Such rotation has been shown to be effective by Shilton et al (2012) in which SAWs drive rotation of the circular rotor. In effect, the rotor can be considered as a small milling stone that assists in DNA fragmentation by generating differential shear.

[0277] The arrangement in FIG. 11 serves to increase the shear to the DNA. Moving the superstrate relative to the substrate ensures that the sample is subjected to a larger range of flow intensities and directions, thus again maximising impact.

[0278] In FIG. 12, two IDTs 14, 34 are coupled together by a layer of liquid sample 12. Surface waves couple into the sample and are then transmitted to each transducer and back again where surface displacements at both transducers interact in order to fragment the chains of nucleic acids. The driving frequency of each transducer need not be the same and indeed in some cases it is advantageous if they are not. As in FIG. 11, the transducers may be translated relative to each other in order to improve the efficiency.

[0279] As will be clear from the disclosure above, the preferred embodiments of the present invention seek to provide efficient and effective DNA fragmentation at relatively low applied power. Existing literature points towards the fact that micro flows generated via acoustic streaming and/or cavitation in liquid samples containing DNA or other such polymeric or long chain materials are drivers for fragmentation of the long chain material into smaller parts. However, these drivers typically have not been accessible in the case of low ultrasonic radiation powers. The preferred embodiments of the invention enable fragmentation using electrical powers associated with portable hand held devices. This opens up the possibility for the field use of next generation sequencing.

[0280] The embodiments of the invention illustrated in FIGS. 1, 5, 6A-6E, 9 and 10 show DNA fragmentation in open systems, i.e. where a drop of liquid sample is manipulated at a surface of the device and the sample presents a free surface. Such embodiments are advantageous for their ease of access, their ease of and implementation and low costs (associated with simple planar geometries). One of the drawbacks to using an open geometry for the fragmentation of polymeric long chain materials such as DNA is the propensity of the sample to nebulise and with it loss of material from the sample. This also leads to partial denaturation, which creates DNA structures that are not useable by sequencing methodologies (e.g. asymmetric, single base pair mis-pairing). This can be overcome to some extent by cooling the sample prior to and during irradiation by surface acoustic waves.

[0281] Acoustic streaming is a second order effect caused by the propagation or the presence of acoustic vibrations interacting with a fluid. The streaming can induce rapid counter propagating flows in the sample fluid which can tear apart the strands of DNA or other long chain structures of interest. Also such streaming flows can be induced via cavitation where micro bubbles (of dissolved gas for example) oscillate due to the presence of ultrasonic pressure waves.

[0282] Preferred embodiments of the invention take advantage of the thermodynamic properties of water where the sample is in a partially frozen state to allow fragmentation to occur at lower powers than otherwise.

[0283] In one suitable embodiment, the invention uses an interdigitated transducer (IDT) on a piezoelectric material such as LiNbO.sub.3 (or any other suitable material that can produce surface vibrations of the desired amplitude and frequency). The results reported here were obtained using 1 mm thick Y cut black LiNbO.sub.3 with an electrode spacing to provide a working frequency of 4.86 MHz, the transducer having a dimension of 23 mm square. A Peltier cooler was attached to a fan assisted heatsink using heat sink compound with the transducer attached to the Peltier cooler using the same heat sink compound. This enabled the system to operate in normal ambient temperatures. The system is as illustrated in FIG. 1. The surface of the transducer can be treated such that it becomes hydrophobic. This makes the recovery of sample, or its further processing directly on the device, easier for a planar geometry.

[0284] In the disclosure above, a drop of liquid sample containing between 25 ng/L to 100 ng/L DNA onto the surface of the IDT a suitable distance away from the electrodes. In order to control the temperature of the sample during irradiation from the surface acoustic waves generated by the IDT a cooling device is used such as a Peltier type device. The Peltier device was operated so that the temperature of the sample did not exceed 37 C. The surface temperatures of the samples were measured with the aid of a Fluke Ti25 IR camera. The results of the fragmentation were analysed using an Agilent Bioanalyser 2100 with the 12 k DNA kit. The effect of providing active cooling was found to be that nebulisation of the sample was reduced, compared with the same conditions except without active cooling.

[0285] A further way to control the temperature of the liquid sample is to pulse the transducer. This has been disclosed also in Yeo et al, Lab Chip. 2014 14(11):1858-65. doi: 10.1039/c4lc00232f. By altering the duty cycle, the ratio of on to off, the average power can be kept low but the peak powers can be allowed to become high without heating of the sample being a problem. However, this requires that the total time to be extended so that the sample sees the same amount of ultrasonic wave irradiation. For example a duty cycle of 50% on and 50% off will require double the time of a continuously irradiated sample. Typical times used for a continuously irradiated sample would be 30 s to 120 s whereas a pulsed 50:50 irradiated sample would require at least 60 s to 240 s. Although not dramatically problematic for sequencing applications, the shorter the time, the higher the throughput of processing, which is one of the major parameters in sequencing applications.

[0286] The power and the amount of time taken for the fragmentation have some influence on the position and shape of the fragmented material's distribution. At low powers, fragmentation is not observed and only when a threshold power is used is fragmentation achieved. As mentioned above, FIG. 2 uses 2 W and shows no fragmentation (temperature less than or equal to 4 C.) while FIG. 3 uses 5 W and shows fragmentation (temperature less than or equal to 8 C.). Once fragmentation is achieved, in an open system, the position of the fragmentation peak appears to be insensitive to the power used and will typically remain at approximately 1200 bp (see FIGS. 2 and 3) until very high peak powers are used (>30 W). However, the duration of exposure can have an influence on the shape of the size distribution with short time exposures less than 60 s having a symmetrical shape (FIG. 3) with this morphing into a wider distribution for samples exposed for more than 90 s, as shown in FIG. 4 (temperature less than or equal to 2 C.). In sequencing applications, a tight distribution is preferred.

[0287] The liquid was frozen or super cooled prior irradiation with ultrasonic waves, cooled such that only partial melting of the frozen drop was achieved during the irradiation with ultrasonic waves. Under such conditions, low powers can be used to achieve the desired fragmentation of below 1000 bp. It is not known at this time the mechanism by which fragmentation is occurring at these low temperatures and applied powers, where sample surface temperature is below or around 4 C. and applied power can be less than 1 W.

[0288] Without wishing to be limited by theory, a possible mechanism for low temperature and power DNA fragmentation could be that the partially melted sample subjects the liquid phase to a roughened interface due to the presence of the frozen parts. Additionally or alternatively the frozen parts may be free to move within the liquid phase to mechanically break up the DNA in the presence of a harmonic forcing cause by the transducer. However, the explanation of the phenomenon might be due to thermodynamic considerations related to repeated cycles of crystallization, thawing and recrystallization or even a combination of these effects.

[0289] The article provided at: http://www.mlo-online.com/freeze-thaw-cycles-and-nucleic-acid-stability-whats-safe-for-your-samples.php [accessed 7 Oct. 2016] provides some disclosure on the effect of freeze-thaw cycles on DNA. However, this method is apparently not reliable and does not provide good sizes (there is no relevant data provided). See also: http://online.liebertpub.com/doi/pdf/10.1089/bio.2011.0016 [accessed 7 Oct. 2016]

[0290] The embodiments of the invention illustrated in FIGS. 5 and 6A-6E utilise a superstrate coupled to a transducer such that the mechanical wave can propagate from the transducer to the superstrate. The superstrate provides the sample treatment zone. The surface of the superstrate can be treated to make it hydrophobic as an aid to collection of exposed sample. The sample is preferably cooled prior to exposure to ultrasonic surface acoustic waves. The liquid sample will heat up (melt if frozen) and spread over the surface. Depending on the volume of sample used, the fluid can make a thin film over the surface. Such a thin film is liable to nebulise and this is preferably reduced or avoided as there will be loss of material if this is allowed to occur. Suitable approaches for reducing or avoiding nebulisation are to use lower powers, larger volumes or control the temperature of the liquid sample during the fragmentation process by pulsing the excitation.

[0291] At different operating frequencies, varying behaviour of the liquid sample can be observed. At lower frequencies there is significant movement of the drop (i.e. motion of the drop shape), while at higher operating frequencies the drop can be made to vortex (motion of the liquid inside the drop) with less translational motion and a similar rate of heating in the liquid sample. This movement can be controlled by modifying the surface chemically by changing the surface chemistry locally or physically by introducing a surface topology which augments the planar surface. We have chosen to use a periodic array of pits (about 186 m diameter, 203 m pitch) but we could have used a random array of such pits or even a mechanically roughened surface using abrasive techniques. What is considered to be important is that the liquid's contact line is pinned such that enough power can be applied to fragment the material of interest while having control where the liquid goes. In some embodiments the liquid can be allowed to move to the far edge of the superstrate where it will remain throughout the fragmentation process, in effect pinned at the edge.

[0292] The effect of using a roughened surface at the sample treatment zone is illustrated by comparing FIGS. 7 and 8.

[0293] The use of a roughened surface at the sample treatment zone appears to have the effect of increasing the efficiency of the fragmentation process, achieving a desired distribution peak value less than 1000 bp at relatively low applied powers (12 W instead of >30 W). This has the advantage of permitting control of the temperature of the fragmentation process to be less than 37 C., preferably less than 20 C., thus avoiding serious issues of heat stress for biological samples. The efficiency is achieved by enabling streaming flows to occur adjacent to or on the roughened (e.g. pitted) surface enabling higher shear to occur than would be present on an otherwise flat planar surface.

[0294] The embodiments discussed above use an open system in which the liquid sample presents a free surface. With power budgets in the region of 1 W, it would be advantageous for the sample to be enclosed in a microfluidic structure. It is preferred for example that the sample is located in a sample chamber 40, 50 (see FIGS. 13-16), in order to enclose it during the fragmentation process while reducing or avoiding sample loss due to nebulisation.

[0295] In the simplest approach, the sample chamber 40 may hold a single phase, i.e. the sample 12 (see FIG. 13). Alternatively, the sample chamber 40 may hold a two-phase system where a water based sample 12 is adjacent to and/or surrounded by an immiscible oil phase 42. Such a system will exhibit excessive damping and be prone to emulsification due to the interaction of an intense acoustic irradiation with the immiscible liquids. Indeed this is what we observe for high powers when it is difficult to keep the temperature suitably low as heat generation is a problem. However, it is advantageous that the preferred embodiments of the present invention are compatible with such encapsulation technologies, allowing suitable DNA fragmentation, in order that the preferred embodiments of the invention can be incorporated in existing sequencing workflows, in particular those which use electrowetting on dielectric (EWOD) techniques in order to manipulate the sample.

[0296] Suitable sample chambers 40, 50 were formed as microfluidic structures fabricated from glass and silicon where the glass was 1 mm thick and the silicon was 500 m thick. Two approaches were used. The first approach was one where the microfluidic structure was bonded directly to the surface of the transducer with epoxy (FIGS. 13 and 14). The second approach was one where the microfluidic structure comprising a glass top and sides had a silicon base was coupled to the transducer using a KY gel (FIGS. 15 and 16).

[0297] In the prior art, it is known to enclose the sample in the chamber where the sample is exposed to high power ultrasonic pressure waves. Part of the insight of the present invention is that where substantial powers are used, there is a need for pulsed irradiation and active cooling such that the temperature of the sample does not exceed 37 C., otherwise the sample will be subjected to temperatures that may denature or damage proteins or DNA.

[0298] It is clear from the work reported here that there is a minimum magnitude of ultrasonic excitation required to fragment DNA. However, we have shown that this can be dramatically reduced by choosing suitable conditions, namely control over temperature, for example to ensure that the maximum temperature experienced by the sample is approximately 4 C. We used a number of frequencies ranging from 7.38 to 9.156 MHz and different modes of operation initially pulsed to ensure peak pressures high enough, then continuously when suitable conditions were found. The concentration of DNA used in this work varied from 7 ng/L to 40 ng/L, results for 12 and 20 ng/L being shown here.

[0299] In FIGS. 13 and 14 the sample 12 is in direct contact with the piezoelectric surface, but elsewhere contained by the sample chamber 40. In FIGS. 15 and 16 a wall of the sample chamber 50 is interposed between the piezoelectric surface and the sample 12. In effect, FIGS. 13 and 14 correspond to treating the sample directly on the piezoelectric surface and FIGS. 15 and 16 correspond to treating the sample on a superstrate.

[0300] Note that in the present work no specific advantage was determined by fragmenting directly on the piezoelectric surface compared with fragmenting on a superstrate. However, it is possible that an effect is seen, and this may offer a route to further reducing acoustic powers required to fragment DNA.

[0301] In the embodiments shown in FIGS. 13-16, the depth of the sample chamber 40, 50 was chosen with respect to integration with existing digital microfluidic platforms. However it is envisaged that sample depth is not a critical factor and could be varied without strong dependence on choice of frequency. In FIGS. 14 and 16, the sample 12 is shown in contact with the top and bottom interior surfaces of the chamber, but this is not a fundamental requirement and the liquid can be surrounded by the oil/wax phase. One suitable way to achieve this is to freeze the liquid sample in order to ensure that it can be encapsulated in the oil/wax phase. If the liquids are immiscible, there may be no need to freeze in order to achieve encapsulation.

[0302] We can process a chamber filled with a single phase namely the sample, however, it is not entirely clear that such a system could handle a two phase system found in digital microfluidics platforms. One concern mentioned above is the preponderance for the two phases to mix and generate an emulsion which would be deleterious to the operation of an electro wetting on demand (EWOD) system. Emulsification was readily observed at elevated temperatures (>10 C.) and high peak powers (36 W corresponding to input of 400 mV pkpk) when using pulsed mode of excitation. This was suppressed when the device was suitably cooled allowing input signals of 600 mV pkpk to be used (corresponding to 290 W peak power). The electrographs shown in FIGS. 17-20 are based on samples treated in a two phase system where the water based sample is surrounded by an immiscible oil. Therefore we have shown that the present invention can work satisfactorily with a two phase system used in digital microfluidics platforms.

[0303] FIGS. 17-20 demonstrates the importance of temperature control when fragmenting DNA with relatively low applied average power. In FIG. 17, the SAW frequency was 9.03 MHz with 7 W transmitted power (pulsed 150 k cycles each 200 ms) for 130 s. The liquid sample had a DNA concentration of 12 ng/L, with the temperature of the sample controlled to be maximum 7 C. In FIG. 18, the SAW frequency was 9.03 MHz with 4 W transmitted power (pulsed 80 k cycles each 200 ms) for 130 s. The sample was partially frozen and had a DNA concentration of 12 ng/L, with the temperature of the sample controlled to be maximum 1 C.

[0304] As can be seen by comparing FIGS. 17 and 18, control of the temperature is important also in this enclosed system, as for the open system discussed above. When the sample temperature was allowed to raise above 7 C., no fragmentation was observed (FIG. 17), compared to FIG. 18. Note that in FIG. 18, the average transmitted power was less than for FIG. 17, and yet a greater degree of fragmentation is seen in FIG. 18.

[0305] FIGS. 19 and 20 show that suitable control of the treatment conditions permits enough power to couple from a transducer through a superstrate in order to fragment DNA in a device of the configuration shown in FIG. 16. In FIG. 19, the SAW frequency was 9.156 MHz and 2 W transmitted power was used (continuous 110 mV pkpk input signal) for 240 s. The sample was partially frozen with a DNA concentration of 20 ng/L. In FIG. 20, the SAW frequency was 7.85 MHz and 5 W transmitted power was used (continuous 110 mV pkpk input signal) for 133 s. The sample was partially frozen with a DNA concentration of 20 ng/L. The maximum temperature of the sample during the process was approximately 2 C. The amount of power coupled into the sample is shown to have a stronger influence on the resultant peak fragment distribution than time of exposure as evidenced by FIG. 20 compared to FIG. 19. Both samples were at approximately the same temperature, showing that the amplitude of the vibration (power applied) has a greater effect than the duration of treatment in order to promote reduced fragment size.

[0306] The configuration of FIGS. 15 and 16, where the sample is coupled to but not in direct contact with the transducer can be considered as bulk wave excitation. In this configuration it was noticed that there was more control over the final peak fragment size.

[0307] The embodiments illustrated so far use SAW transducers. However, embodiments of the invention also work satisfactorily with bulk acoustic waves, generated using bulk acoustic wave transducers 60 transmitted via waveguide 62. Suitable configurations for such devices are shown in FIGS. 21-24 (these configurations use Langevin type bulk wave transducers 60 to achieve fragmentation of DNA in sample 12) and FIGS. 25-28 (these alternative configurations also use Langevin type bulk wave transducers 60 and waveguides 62 to achieve fragmentation of DNA). In FIGS. 25-28, cooling systems 21 are disposed around the sample chamber 40, 50 holding sample 12 (optionally with immiscible phase 42).

[0308] Embodiments of the present invention also work satisfactorily using surface shear waves. This is illustrated using the embodiments shown in FIGS. 29 and 30. FIG. 29 shows a part of a sample treatment zone showing the transducing material 70 in which the acoustic wave is a Bleustein-Gulyaev wave 72, schematically illustrated.

[0309] FIG. 30 shows a part of a sample treatment zone showing the transducing material 70 and in which the acoustic wave is a guided Love wave 74, schematically illustrated.

[0310] In FIG. 29, the transducer includes a raised ridge 71. As shown, the raised ridge is formed of the piezoelectric material and is formed monolithically with the remainder of the piezoelectric material. Shear waves (Bleustein-Gulyaev) propagate along the transducer, including along the raised ridge 71. At the side walls of the ridge, the effect seen is similar to a Rayleigh or Lamb wave turned on its side. This can couple into the liquid sample as for a Rayleigh or Lamb wave propagating on a planar surface. Additionally, the movement of the side walls provides inertial forcing of the liquid sample and for the promotion of the onset of cavitation.

[0311] FIG. 30 shows a similar arrangement to FIG. 29 except that the raised ridge 73 is not formed of a piezoelectric material. Instead, it is formed as a waveguide, forced to oscillate as shown by Love waves, and provides similar effects to FIG. 29 in the liquid sample.

[0312] Further work has been completed by the inventors to investigate the effects of the shape of the interface between the sample and the sample treatment zone. This work has demonstrated that the use of an engineered structure at the interface can have a positive influence on DNA fragmentation via ultrasonic waves acting on a sample containing the DNA. This is the case even while the liquid temperature is kept many degrees centigrade above freezing.

[0313] In the work reported here, the structures used had varying forms (pit, trough, pillar and strip) and varying edge curvatures, depths and heights. Additionally, the structures were formed of different materials. Some superstrates were completely made out of silicon while others were made out of a patterned layer of SU8 (photresist) on a silicon superstrate.

[0314] All the samples were frozen prior to the application of SAW. This was to ensure that all runs had similar starting conditions.

[0315] In these experiments the applied power was approximately 5 W to 15 W but generally 13 W was used from an input to the amplifier of 190 mV pkpk. At this power it is found that the temperature of the liquid does not rise too quickly and fragmentation sizes sub 17000 bp can be produced reproducibly.

[0316] The LiNbO.sub.3 interdigitated transducer was driven at a frequency of approximately 7.3 MHz and was pasted to a Peltier heater/cooler which in turn was pasted to a heatsink used in conjunction with a 12V fan. Temperature of the liquid samples was measured remotely using a Fluke Ti 25 IR thermal imaging camera. The maximum temperature was 53 C. and the minimum 20 C. although generally the temperature was kept to below 40 C. and above 17 C.

[0317] The silicon superstrates were patterned using optical lithography, this photoresist pattern was then transferred into silicon via a dry etch process. The SU8 structures were fabricated using optical lithography. The freeze thaw device was made using metal lift off on a polymeric surface which could be SU8 on silicon or a piece of plastic such as PMMA. With respect to the topologies of the superstrates, the pits or posts were arranged in either triangular or square lattices.

[0318] A Vectawave 80 W RF amplifier used in conjunction with RF power meter. The range of temperatures that all samples shown reached during the application of SAW, was between 10 C. and 41 C. A number of various structures were used arrays of pits or posts and flat or roughened silicon. The source of the DNA used came from Corriel (NA12878) at a concentration of 38 ng/L or 76 ng/L with 9 L used for all runs.

[0319] The minimum depth of the Si pits was approximately 80 m and maximum was approximately 200 m. The minimum height of the posts was approximately 45 m and the maximum height was approximately 145 m. Smallest feature size was 20 m in diameter and the largest was 1000 m in diameter. All silicon superstrates were fabricated using (100) oriented 500 m thick four inch single side polished wafers. An example of the arrayed pit structures is shown in FIGS. 31-33. FIGS. 31 and 32 show perspective SEM views of a silicon superstrate with a triangular lattice of pits, the pits having diameter 75 m, pitch 120 m and depth 200 m. FIG. 33 shows a cross sectional SEM view of the pits.

[0320] With the embodiment of FIGS. 31-33, the sample was initially placed within a phononic cavity (at the flat surface between the two arrays of pits shown in FIG. 31). However, on application of the SAWs, the sample 12 spread to be in direct contact with the pits 80 in the superstrate 82, as shown in FIG. 34.

[0321] FIG. 35 shows fragment distributions for 9 L samples for different applied powers using the superstrate of FIGS. 31-33. The peak fragment size was approximately 1734 bp using approximately 7.3 MHz for 120 s. As can be seen, the DNA readily fragments with peak sizes below 2 kbp on the pit structures and above 1700 bp and giving rise to relatively sharp peaks with less than 13 W applied power.

[0322] FIGS. 36-38 show SEM images of a silicon superstrate with a square array of pillars. The pillars have diameter 80 m, pitch 125 m and depth 144 m.

[0323] The experimental set up for the superstrate 86 of FIGS. 36-38 is shown in FIG. 39, with the liquid sample placed on the pillars 84. FIG. 40 shows fragment distributions for 9 L samples for different applied powers. The peak fragment size was approximately 1587 bp using approximately 7.3 MHz for 120 s, note, the yield is higher than for the pit structure superstrate. Thus, it appears that the pillar structures performed better than the pit structures, with all runs producing fragment distribution peaks below 1600 bp and typically above 1100 bp, again giving rise to a relatively sharp peak for applied powers less than 13 W, as shown in FIG. 40.

[0324] FIG. 41 shows an SEM image of another engineered Si superstrate, with a square array of pillars of height 44 m, diameter 20 m and pitch 80 m. Using a similar arrangement to FIG. 39, a sample was placed on the pillars and subjected to SAWs. FIG. 42 shows the resultant DNA fragment distribution for different applied powers.

[0325] Without wishing to be bound by theory, the inventors speculate that the fragmentation yield from the pillar structures is higher than that from the pit structures perhaps because the pillars have a greater degree of freedom to move or provide a larger interaction area with the sample. Also the open structure of a lattice of pillars may promote streaming flows within the liquid sample more effectively that an array of pits.

[0326] The roughness of the superstrate surface, apart from the engineered surface structures (pits or pillars) did not appear to be critical to the fragmentation of DNA under the conditions used. However, as expected the intensity of the acoustic field (magnitude of the elastic waves) influences fragment sizes produced with higher powers typically producing smaller fragments.

[0327] This dependence on acoustic power is shown in the fragmentation carried out on a flat silicon superstrate 88 (a piece of unprocessed silicon wafer), in the arrangement shown in FIG. 43. In this arrangement, it was not possible to stop the 9 L sample 12 from moving to the far edge of the superstrate 88. The drop perched at the end where the surface displacement would be highest with respect to rest of the superstrate 88. With this arrangement peak fragment size distributions of around 2000 bp can be produced, as shown in FIG. 44. In FIG. 44, fragmentation size distributions are plotted for two concentrations of DNA, after a 7.3 MHz SAW with a power of 10 W for 120 s was applied to the superstrate.

[0328] As shown in FIG. 44 compared with FIGS. 40, 42 and 35, the peak fragment sizes are larger on the flat superstrate than that obtained by using structured silicon. However, although may not be desirable to have a drop perched on the end of the superstrate, the result informs us that structures designed to manipulate the acoustic field allows the reduction of the applied power and yet still induce useful fragmentation of DNA.

[0329] For the superstrate reported in FIG. 44, the surface roughness was gauged by white light profilometry (ContourGT Bruker), giving a measure the surface roughness of about Rz=800 nm.

[0330] It was initially considered by the inventors that that the structures needed to be ordered in order to be of benefit to DNA fragmentation performance. However, a silicon wafer that has been used as a backing wafer for a dry etch process develops an altered texture on the backside of the wafer at its periphery. The measured roughness (gauged by white light profilometry (ContourGT Bruker)) of this area of the wafer was about Rz=19 m. FIGS. 45-47 show SEM images of different parts of the Si superstrate. Fragmentation of DNA was carried out by placing the sample on the dry etch damages area and applying SAWs. It was found that this provided the best results for fragmentation without the need to have the sample temperature kept below 5 C. This is shown in FIG. 48, which shows the fragmentation size distribution for two concentrations of DNA, after exposure to 7.3 MHz SAW with a power of 10 W for 120 s, where a fragmentation peak of 355 bp was obtained for the higher concentration.

[0331] Comparison of FIGS. 44 and 48 illustrates the effect of non-ordered surface roughness on the DNA fragment size distribution.

[0332] SU8 is a negative photoresist that crosslinks a monomer after exposure and baking prior to development in EC solvent. The permanent crosslinking of the monomers in the resist provide resilient structures that will not completely deform when heated. Some reflow of the resist is expected when undergoing hard bakes (120 C. to 230 C.) as complete cross linkage occurs at 240 C. This ability of the polymeric coating to undergo plastic deformation was used to fabricate smooth micron scale structures.

[0333] Two formulations of SU8 were used (designed 3050 and 3025). These respectively provided layer thickness of about 45 m and 20 m. In order to obtain different sidewall angles the 3050 samples were exposed at 30 s, 120 s, 600 s and 900 s. After development of the optical lithography the various samples were hard baked at 120 C., 180 C. and 230 C. between 3 hrs and 20 hrs.

[0334] Higher temperature or longer times for the hard bake causes either more reflow or smoothing of the surface of the SU8.

[0335] FIG. 49 shows an SEM cross sectional perspective micrograph of a SU8 3050 structure in the form of 100 m diameter pits in a square array with 0.5 mm pitch. The structure was subjected to a 10 min exposure and subsequently a 230 C. 20 hrs hard bake. Note the smooth sidewalls and undulating shape and a thin layer approximately 1 m thick on the silicon superstrate inside the pit.

[0336] FIG. 50 shows an SEM cross sectional perspective micrograph of a SU8 3050 structure in the form of a 300 m pit. The structure was subjected to a 30 s exposure and subsequently a 120 C. bake for 4 h.

[0337] The structures in FIG. 49 did not produce DNA fragments using 13 W applied power. The structures in FIG. 50 showed little if any fragmentation of DNA. It is possible perhaps that this inefficiency could be attributed to the thin residual SU8 layer found in these structures. This may have hindered transmission of acoustic energy to the sample or that the SU8 coating hindered the transmission of acoustic energy to the sample. Compared will the pillar embodiments described later, it is possible that these structures have more SU8 covering the silicon, affecting their performance.

[0338] There were also manufactured arrays of pillars using SU8. Pillars with a diameter of 0.3 mm and pitch of 0.5 mm appeared to be less efficient than similar structures fabricated wholly in silicon.

[0339] FIG. 51 shows a perspective SEM micrograph of an SU8 3050 pillar after 120 s exposure and 230 C. hard bake for 3 hrs. FIG. 53 shows a perspective SEM micrograph of an SU 8 3050 pillar after 600 s exposure and 230 C. hard bake for 20 hrs. Both structures were 0.3 mm in diameter and were fabricated in a square array formation with a pitch of 0.5 mm.

[0340] FIGS. 52 and 54 respectively show the performance of the structures of FIGS. 51 and 53 for DNA fragmentation. Each structure appeared inefficient for the fragmentation of DNA (43 ng/L Promega) using 13 W applied power at approximately 7.3 MHz for 120 s. There also appeared to be some influence of the hard bake temperature and times used where higher temperature and longer time appear to lower the performance of the structures.

[0341] FIGS. 55-57 show SEM micrographs of other SU8 pillar structures. These were SU8 3050 exposed 900 s with a 3 hr hard bake at 180 C. The pillars were 0.3 mm in diameter placed in a square array with a pitch of 0.5 mm.

[0342] FIG. 58 shows plots of DNA (43 ng/L Promega) fragmentation after exposure to 7.3 MHz SAW for 120 s with different applied powers, placed on an array of pillars as shown in FIGS. 55-57. The plots do not provide any clear trend with applied power but imply an optimal applied power for a particular engineered structure. However, it is clear that the presence of the pillars do enable fragmentation to occur at applied powers much lower than 13 W with a high yield for fragmented DNA.

[0343] Thus, some pillar structures (having the same diameter of 0.3 mm and pitch of 0.5 mm) perform better than others, comparing FIG. 58 with FIGS. 54 and 52. In FIG. 58, peak fragment sizes of approximately 1313 bp were obtained at relatively low applied power (5 W) and where the sample temperature was kept at approximately 18 C.

[0344] It is possible to think of the array of pits as a continuous layer of SU8 on the silicon superstrate whereas an array of pillars can be considered a discontinuous layer of SU8. The mass loading on the silicon superstrate will be lower in the case of a discontinuous coating. This may be an additional reason for the poor performance of the pit structure devices compared to the good performance of the pillars structures.

[0345] SU8 device structures were manufactured that were 1D arrays of troughs or strips. These were formed using similar conditions to the structures shown in FIGS. 55-57. The arrays were placed such that they were normal to SAWs produced by the transducer. That is, the arrays were parallel to the electrodes of the IDT. Here we see that the sidewall of the SU8 structure influences the efficiency of the device for fragmentation.

[0346] FIGS. 59-61 show SEM micrographs of SU8 3050 strip structures exposed for 900 s and hard baked at 180 C. for 3 hrs.

[0347] FIG. 62 shows the DNA fragmentation performance of the structures of FIGS. 59-61, carried out on a 9 L sample of 43 ng/L (Promega) DNA at about 7.3 MHz at 11 W applied power for 120 s. Raising the applied power to 12 W gave some improvement. However, the undulating or sigmoidal profile of the side wall of the strip structure shown in FIGS. 59-61 appears to inhibit efficient fragmentation of DNA using SAW.

[0348] However, it is found that if the sidewall is linear and an angle of approximately 60 fragmentation could be achieved at 11 W.

[0349] FIGS. 63-65 show SEM micrographs of SU8 3050 trough structures which were formed by exposing for 900 s and hard baking for 3 hrs at 180 C.

[0350] It was found that the strip structures performed much better than their trough structure counterparts. SU8 strip structures 1 mm wide with a pitch of 4 mm appeared to perform the best while strip structures 0.5 mm wide and a pitch of 2.5 mm also gave good results. The SU8 3050 layers were approximately 45 m thick whereas the SU8 3025 layers were approximately 20 m thick. There appeared to be a slight bias towards a smaller step height of the structures with respect to yield of fragments whereas peak fragment size appeared to show the reciprocal relationship where bigger step height produced fragment distributions peaks with lower base pair number.

[0351] FIG. 66 shows the superior performance of strips with respect to their reciprocal structure (troughs) when a 9 L sample (43 ng/L Promega) is exposed to approximately 7.3 MHz with an applied power of 13 W for, in the case of the troughs 120 s, in the case of the strips 60 s. Note that the SU8 3050 device gave smaller peak fragment size while the SU8 3025 device had a higher yield. All devices used were hard baked at 120 C. for 4 hrs.

[0352] It was noticed that hard bake temperature and time influenced the performance of the SU8 devices. Devices that underwent a long 230 C. hard bake performed less well with respect to yield of material. This would imply that the degree of cross linking of the resist and hence elastic properties changes may be the cause of the difference.

[0353] In FIG. 67, a high temperature hard bake for different time periods are compared, where sample of 9 L volume were used exposed to 7.3 MHz with an applied power of 13 W, samples on the trough structures were fragmented for 120 s, while samples on strip structures were fragmented for 60 s. There is some increase in sidewall curvature for the longer baked devices as can be seen from the SEMs in FIG. 67 but this appears small. Again the trough devices perform poorly compared to their strip counterparts. Both strip devices gave fragment distributions below 2 kbp for the conditions used.

[0354] In more detail, FIG. 67 compares SU8 3050 strip and trough structures that have undergone different hard bake times at 230 C. The SEMs for the 20 hr device exhibited slightly more curvature of the sidewall. Approximately 7.3 MHz with an applied power of 13 W was used to fragment 9 L samples (43 ng/L Promega) in the case of the troughs 120 s, in the case of the strips 60 s. Again troughs did not perform well whereas strip structures gave a high yield with fragment distribution peak below 2 kbp.

[0355] We now explain in general terms how the embodiments of the invention can be integrated into a DNA sequencing process, for implementation for example using a sequencing apparatus.

[0356] There are known to the skilled person various approaches for DNA sequencing. FIG. 68 illustrates a series of step for treatment of a DNA sample before the actual sequencing operation is carried out.

[0357] At step A, DNA is extracted. This step is dependent upon the nature of the sample, and may require lysis and/or purification.

[0358] At step B, the DNA is fragmented. This is carried out as explained in detail above.

[0359] After fragmentation, usually the fragment ends have overhangs that need to be blunted (using enzymes). This is done at step C.

[0360] At step D, an A is added at the end to provide an anchor for the adapter and prevent concatenation during ligation.

[0361] At step E, adapters (short DNA fragments) are added at the end of the sample DNA fragments to enable their hybridization onto the sequencing surfaces.

[0362] At step F, the adequate constructs are selected using size.

[0363] At step G, because the amount of adequate DNA constructs at this stage can be limited, a PCR amplification step is carried out. This also has the effect of purifying the sample.

[0364] Step H is optional. Here, there is the step of validating, normalising and pooling libraries (concentrations, quality).

[0365] Next, at step I, the pre-processed sample is subjected to the DNA sequencing operation itself.

[0366] One or more of steps C to H may be carried out at the sample treatment device used to carry out step B. Alternatively, the sample may be moved using either robotics (pipetting) or microfluidics (pressure or EWOD (electrowetting on demand), for example).

[0367] There are now described approaches to the fragmentation of DNA using freeze-thaw cycling methods, without necessary applying SAWs or other acoustic waves to the sample. However, it is to be understood that these freeze-thaw methods may be used in combination with the SAW or acoustic wave methods set out above.

[0368] Considering first the background to freeze-thaw treatment of DNA, it is known that multiple freezing and thawing of stored sperm samples will cause degradation of chromosomal DNA [Kopeika et al (2015)]. This is a method that has been used to increase the efficiency of gene modifications in gametes [Ventura et al (2009)]. In most cases however, fragmentation of DNA from freeze-thawing is a negative effect that methods have been designed to avoid. For example, in mass spectrometry, samples are frozen and desorbed from a frozen state (without thawing) to avoid unwanted fragmentation, as disclosed in EP-B-0404934 and U.S. Pat. No. 4,920,264.

[0369] Freeze-thawing has been used in preparing libraries for DNA sequencing, but the control obtained on the size or the efficiencies have not made it a preferred methodology [Makarov and Langmore (1999)]. Indeed the sizes of specific protocols have been characterised and shown to be higher than 10 kb [Shao et al (2012)], too large for efficient library preparation.

[0370] EP-A-1752542 discloses a method of generating non-human transgenic animals, including a freeze thaw step to cause fragmentation. U.S. Pat. No. 4,920,264 discloses a method for preparing samples for mass analysis by desorption from a frozen solution. One of the stated aims of this document is to mitigate or minimize fragmentation, and aims to achieve this by freezing target molecules. U.S. Pat. No. 6,117,634 discloses nucleic acid sequencing and mapping and mentions fragmentation as a by-product of freeze-thawing but does not utilise this. WO 2011/031127 discloses a method of isolating DNA from cells, in which rapid freeze thaw cycles between 65 C. and 70 C. are used.

[0371] Here we provide a platform, preferably based on a structured surface, to enhance the efficiency of the fragmentation of DNA by using temperature cycling, which includes a step below the freezing temperature of water. Note that in some circumstance it would be possible to freeze a sample above the triple point of water by the addition of a suitable additive. See, for example, http://news.mitedu/2016/carbon-nanotubes-water-solid-boiling-1128 [accessed 10 Oct. 2017].

[0372] The freeze-thaw method is illustrated here through the use of different heating mechanisms. The efficiency of the method appears to be linked to the presence of a microstructured surface (with feature size in the range of tens of microns, as described above).

[0373] In the work described below, human genomic DNA obtained from Promega Corporation (G3041) and Coriell NA12878 DNA was used with a concentration of 43 ng/L and 38 ng/L respectively, where each sample had a volume of 6 to 9 L.

[0374] Two superstrates were used: a glass coverslip and structured silicon (consisting of pillars 130 m in diameter, 160 m high with a pitch of 230 m) physically connected to a heat source via a small volume of heatsink compound while some samples were carried out directly on PZT\SU8 composite devices (see below).

[0375] The heaters were also thermally connected to a Peltier cooler rated for 6 A via a small volume of heatsink compound.

[0376] The power applied to the heaters was modulated such that the drop could melt and refreeze during each cycle of the modulated applied power, enabling multiple freeze/thaw cycles. Typically the modulation consisted of a square wave with a frequency between 0.05 Hz to 0.5 Hz.

[0377] In a first approach, an RF heater was used. Lead Zirconate Titanate (PZT) is a ferroelectric ceramic which can be used in the fabrication of piezoelectric transducers. A composite material was formed, comprising Ferroperm Pz26 powder (PZT) added to

[0378] SU8 2050 negative tone photoresist at 30% by volume and mixed thoroughly. The mixture was then applied to interdigitated electrodes [obtained from Epigem UK] by first masking off the interdigitated electrodes with Sellotape, with the excess mixture scraped off using the edge of a glass slide. This is illustrated in FIG. 69, showing the interdigitated electrodes 102 on SU8 coated glass 104. An area over the interdigitated electrodes 102 was masked off using Sellotape 106. This area was then coated in the 30% by volume mix of PZT powder dispersed in SU8 photoresist. Excess applied mixture was scraped off leaving a film approximately 150 m thick after processing (corresponding to the thickness of the Sellotape). After pre baking the mixture at 95 C. for 20 min the device was exposed to 365 nm UV (23 mW/cm.sup.2) for 10 min then post exposure baked for 10 min at 95 C. in order to crosslink the photoresist.

[0379] In the present approach, we do not use the PZT/SU8 composite devices for their piezoelectric properties, but for their roughness and ability to heat up quickly on application of a suitable RF signal via the electrodes. To avoid any doubt, we include a characterisation in FIG. 72 and in FIG. 75.

[0380] FIG. 72 shows a frequency scan of the PZT/SU8 composite material on an Epigem interdigitated electrode using an Agilent vector network analyser (S11 parameter). Marked on the scan is a small trough indicative of resonance point around 32 MHz which, although very much smaller than may be expected, is the correct frequency for the device.

[0381] FIG. 75 shows a polytec GmbH single point vibrometer (range up to 24 MHz) showing the presence of the first sub harmonic due to the restricted range of the vibrometer used (up to 24 MHz) when excited by a 5V pkpk signal at 32 MHz, indicating some actuation of the surface.

[0382] FIGS. 72 and 75 therefore show very small piezoelectric actuation of the surface, which is considered not to be significant enough to contribute meaningfully to a DNA fragmentation process.

[0383] FIGS. 73 and 74 show screen shots from a Bruker Contour GT white light profilometer scan of the surface of the 30% by volume mix of PZT/SU8 composite. The surface is apparently non smooth appearance. Based on these results, the average roughness was found to be about 5 m.

[0384] Heating using RF was found to be efficient where only 0.1 W of applied power at 32.5 MHz was enough to obtain a temperature of approximately 77 C. after 5 s. FIG. 71 shows a temperature plot across the central part of the device corresponding to the centre of the electrodes 102. This was plotted from an IR image taken with a FLIR IR camera of the PZT/SU8 composite device driven at 32.5 MHz with an applied power of 0.1 W for 5 s. The hottest part of the image is situated at the centre of the interdigitated electrodes. The maximum measured temperature was 77 C., with a temperature of at least about 44 C. measured at all areas directly above the interdigitated electrodes.

[0385] FIG. 71 shows a schematic perspective view of the device in operation. Base 104 holds the interdigitated electrodes 102 and the PZT/SU8 composite 108 is formed over and in contact with the interdigitated electrodes 102. Sample 12 has a volume of about 9 L in this example.

[0386] Note that if the PZT/SU8 composite 108 was not present, no significant heating effect was observed when the RF signal was applied to the electrodes 102.

[0387] FIGS. 76 and 77 show electrographs of 9 L of Human DNA (Coriell NA12878) samples each with a concentration of 38 ng\L placed onto a PZT\SU8 composite device. In FIG. 76, the sample was in direct contact with the PZT\SU8 composite device. In FIG. 77, the sample was in contact with a smooth glass superstrate which itself was in direct contact with the PZT\SU8 composite device. The device was driven at 32.5 MHz with 0.2 W applied power modulated at a frequency of 0.05 Hz for FIG. 76 and driven at 32.5 MHz with 0.3 W applied power modulated at 0.05 Hz for FIG. 77. The starting temperature for FIG. 76 was 7 C. with the temperature ranging between 6 C. to 1 C. on the application of the RF signal which was applied for 8 min. The staring temperature for FIG. 77 was 6 C. with the temperature ranging between 5 C. to 1 C. on application of the RF signal which was applied for 5 min.

[0388] Based on FIG. 76, it can be seen that when cycling between solid and liquid states of water, DNA fragments below 1 kb can be obtained, when the sample is positioned directly on the rough surface (FIG. 76), while they do not form when the surface is smooth (FIG. 77).

[0389] Heating via a resistive heater was also found to be effective. Resistive thermal devices (RTD) were used as micro strip heaters. These devices were produce via a lift off technique, where the device pattern was created lithographically on 300 m thick Pyrex glass followed by evaporating 100 nm of a suitable metal (e.g. Pt or NiCr) to create a serpentine track. The single wire was used as a heating element when a current, as high as 2.6 A, was passed through the device. The current supplied to the heaters was modulated with frequency of 0.05 Hz. The devices were used in conjunction with a superstrate. The superstrate was either a smooth glass coverslip or a structured silicon superstrate. These were coupled to the heater with the aid of heatsink compound.

[0390] FIGS. 78 and 78 show electrographs of 6 L of Genomic DNA (Promega G3041) samples each with a concentration of 43 ng/L after treatment by heating with a micro strip heater. In FIG. 78, the sample was placed onto a coverslip (smooth glass superstrate). In FIG. 79, the sample was in contact with a structured silicon superstrate (pegs 130 m dia. 160 m high with a pitch of 230 m). The starting temperature for FIG. 78 was 7.3 C. with the temperature ranging between 2.1 C. to 6.7 C. on the application of the modulated dc current (0.05 Hz) applied for 4 min. The staring temperature for FIG. 79 was 7 C. with the temperature ranging between 3.2 C. to 4.7 C. on application of the modulated dc current (0.05 Hz) applied for 4 min.

[0391] It can be seen when comparing FIGS. 78 and 79 that only comparatively large fragments are produced when the sample is placed on a glass slide. However, there is a striking contrast with the structured superstrate, in which very short fragments (which are desirable) are produced with the silicon pillars.

[0392] The embodiments of the freeze-thaw approach tested here use open systems. However, further embodiments would adopt features disclosed with the respect to the SAW and acoustic wave embodiments. In particular, it is envisaged to use closed chambers and/or two phase enveloped samples. It is considered that the fragmentation efficiency found using the structured superstrate would follow the same behaviour found with respect to the dimensions and shapes of the structured superstrates investigated in the SAW and acoustic wave embodiments.

[0393] Different approaches can be taken for heating the sample. For example, the sample may be heated (directly or indirectly) by radiation (such as an IR source (e.g. a flame) or IR/terahertz/visible diode or other radiation source, or laser), convection (warm air) or by contact with a heated source.

[0394] It is also possible to include particles in the sample (e.g. magnetic beads or plasmonic particles) that preferentially would heat up upon activation (e.g. via radiation or other excitement). Such an approach may have additional advantages in the sense of providing additional surfaces for fragmentation. Furthermore, such an approach may provide a means for capture of fragments. For example, the beads may be coated with suitable capture molecules or surface. Note that in some embodiments it is preferable not to include such further particles, in view of the desirability of removing such particles before sequencing, which therefore adds an additional processing step.

[0395] While the invention has been described in conjunction with the exemplary embodiments described above, many equivalent modifications and variations will be apparent to those skilled in the art when given this disclosure. Accordingly, the exemplary embodiments of the invention set forth above are considered to be illustrative and not limiting. Various changes to the described embodiments may be made without departing from the spirit and scope of the invention.

[0396] All references referred to above and/or listed below are hereby incorporated by reference.

LIST OF NON-PATENT REFERENCES

[0397] Marine R. et al., Appl. Environ. Microbiol., 77 (2011), 8071 [0398] Q. Tseng, A Lomonosov, E. Furlonga and C. Merten Fragmentation of DNA in a sub-microliter microfluidic sonication device Lab Chip, 2012, 12, 4677-4682, DOI: 10.1039/C2LC40595D [0399] Y. Okabe and A. Lee LCAT DNA Shearing Journal of Laboratory Automation, 2014, Vol. 19(2) 163-170 [0400] Larguinho M., Santos H. M., Doria G., Scholz H., Baptista P. V., Capelo J. L. Development of a fast and efficient ultrasonic-based strategy for DNA fragmentation Talanta 81 (2010) 881-886 [0401] N. Nama, P. Huang, T. Huang, F. Costanzo Investigation of acoustic streaming patterns around oscillating sharp edges Lab Chip, 2014, 14, 2824 [0402] Wu, T. & Chang, I., 2005. Actuating and detecting of microdroplet using slanted finger interdigital transducers. Journal of Applied Physics, 98(2), 024903-7 [0403] Shao et al BIOPRESERVATION AND BIOBANKING Volume 10, Number 1, 2012, -10.1089/bio.2011.0016 [0404] Hedskog C. et al., PLOS One, 5 e11345 (2010) [0405] R. J. Shilton, S. M. Langelier, J. R. Friend and L. Y. Yeo Surface acoustic wave solid-state rotational micromotor Appl. Phys. Lett. 100, 033503 (2012) http://dx.doi.org/10.1063/1.3676660 [0406] http://www.mlo-online.com/freeze-thaw-cycles-and-nucleic-acid-stability-whats-safe-for-your-samples.php [0407] http://online.liebertpub.com/doi/pdf/10.1089/bio.2011.0016 [0408] Kopeika, J.; Thornhill, A.; Khalaf, Y. The Effect of Cryopreservation on the Genome of Gametes and Embryos: Principles of Cryobiology and Critical Appraisal of the Evidence. Hum. Reprod. Update 2015, 21 (2), 209-227 DOI: 10.1093/humupd/dmu063. [0409] Ventura, O. M.; Gutierrez, A.; Motoliu, J. Method of Generating Non-Human Transgenic Animals. EP1752542 (A1), May 27, 2009. [0410] Hunter, J. M.; Lin, H.; Becker, C. H. Cryogenic Frozen Solution Matrixes for Analysis of DNA by Time-of-Flight Mass Spectrometry. Anal. Chem. 1997, 69 (17), 3608-3612 DOI: 10.1021/ac9703764. [0411] Makarov, V. L.; Langmore, J. Compositions and Methods for Analysis of Nucleic Acids. WO9839485 (A2), Feb. 25, 1999. [0412] Shao, W.; Khin, S.; Kopp, W. C. Characterization of Effect of Repeated Freeze and Thaw Cycles on Stability of Genomic DNA Using Pulsed Field Gel Electrophoresis. Biopreservation Biobanking 2012, 10 (1), 4-11 DOI: 10.1089/bio.2011.0016.