(12) INTERNATIONAL APPLICATION PUBLISHED UNDER THE PATENT COOPERATION TREATY (PCT)
`
`eo) IMTLINN AMINA UN AAAMe
`
`
`ee
`sey
`Ly
`PCT
`
` .
`
`(19) World Intellectual Property Organization
`International Bureau
`.
`ous
`(43) International Publication Date
`18 February 2010 (18.02.2010)
`
`(10) International Publication Number
`WO 2010/018465 A2
`
`DZ, EC, EE, EG, ES, FI, GB, GD, GE, GH, GM, GT,
`HIN, HR, HU, ID, IL, IN,IS, JP, KE, KG, KM, KN, KP,
`KR, KZ, LA, LC, LK, LR, LS, LT, LU, LY, MA, MD,
`ME, MG, MK, MN, MW, MX, MY, MZ, NA, NG, NI,
`NO, NZ, OM,PE, PG, PH, PL, PT, RO, RS, RU, SC, SD,
`SE, SG, SK, SL, SM, ST, SV, SY, TJ, TM, TN, TR, TT,
`TZ, UA, UG, US, UZ, VC, VN, ZA, ZM, ZW.
`
`(51) International Patent Classification: Not classified
`;
`_
`(21) International Application Number:
`
`_
`PCT/IB2009/007568
`
`(22) International Filing Date:
`
`11 August 2009 (11.08.2009)
`
`(25) Filing Language:
`
`(26) Publication Language:
`(30) Priority Data:
`61/088,142
`
`12 August 2008 (12.08.2008)
`
`English
`.
`English
`
`US
`
`(72)
`(75)
`
`(54) Title: METHODS AND DEVICES FOR DIGITAL PCR
`
`(84) Designated States (unless otherwise indicated, for every
`kind ofregional protection available): ARIPO (BW, GH,
`GM,KE, LS, MW, MZ, NA, SD, SL, SZ, TZ, UG, ZM,
`ZW), Eurasian (AM, AZ, BY, KG, KZ, MD, RU, TJ,
`TM), European (AT, BE, BG, CH, CY, CZ, DE, DK, EE,
`ES, FI, FR, GB, GR, HR, HU, EE,IS, IT, LT, LU, LV,
`except US):
`all designated States
`(or
`(71) Applicant
`MC, MK, MT, NL, NO, PL, PT, RO, SE, SIL SK, SM,
`STOKES BIO LIMITED[IE/IE]; Shannon Arms Henry
`TR), OAPI (BF, BJ, CF, CG, CI, CM, GA, GN, GQ, GW,
`Street, Limerick (IE).
`ML, MR, NE, SN, TD, TG).
`Inventors; and
`Inventors/Applicants (for US only): DAVIES, Mark Declarations under Rule 4.17:
`[GB/IE]; 12 Harbour ViewTerrace South Circular Road,
`— as to applicant's entitlement to apply for and be granted
`Limerick (IE). DALTON, Tara [IE/IE]; Ashtfort Patrick-
`a patent (Rule 4.17(ii))
`swell, County, Limerick (IE).
`(81) Designated States (unless otherwise indicated, for every Published:
`kind of national protection available): AE, AG, AL, AM, — without international search report and to be republished
`AO, AT, AU, AZ, BA, BB, BG, BH, BR, BW, BY, BZ,
`upon receipt of that report (Rule 48.2(g))
`CA, CH, CL, CN, CO, CR, CU, CZ, DE, DK, DM, DO,
`
`GAPDH serialdilution
`T
`-T
`
`
`
`
`
` T
`
`t
`5X10? ng
`——}+—-
`—> X10" ng/yl
`seeKe5x40 ng/ul
`5X109 ng/l
`
`
`
`
`45
`
`(57) Abstract: The invention provides methods of conducting a nucleic acid reaction, including methods for performing digital
`PCR using a "droplct-in-oil" technology. In the methods, the starting sampled is segmented at lcast partially into a sect of sample
`droplets each containing on average about one or fewer copies of a target nucleic acid. The droplets are passed in a continuous
`flow of immiscible carrier fluid through a channel that passes througha thermal cycler, whereby the target is amplified. In one im-
`plementation, the droplets are about 350 nl each and the numberofpositively amplified droplets is counted at the near-saturation
`point.
`
`
`
`WO2010/018465A2|ANTIAIUMDIMTIANAAACAMAANITOUT
`
`
`
`
`
`

`

`WO 2010/018465
`
`PCT/IB2009/007568
`
`METHODS AND DEVICES FOR DIGITAL PCR
`
`Related Applications
`
`The present application claims priority to U.S. Provisional Application No. 61/088, 142,
`
`filed August 12, 2008, the contents of which are incorporated by reference.
`
`Technical Field
`
`[0001] The invention relates to methods and devices for conducting nucleic acid
`
`amplification reactions, including the polymerase chain reaction (PCR).
`
`Background
`
`[0002] PCR is a molecular amplification method routinely practiced in medical and
`
`bioresearch settings for a variety of tasks, such as the detection of hereditary diseases,
`
`the identification of genetic fingerprints, the diagnosis of infectious diseases, the cloning
`
`of genes, paternity testing, and other types of nucleic acid analysis. For a review of the
`
`PCR methodology, see, e.g., PCR Protocols (Methods in Molecular Biology) by Barlett
`
`and Stirling (eds.), Humana Press (2003); and PCR by McPherson and Moller, Taylor &
`
`Francis (2006)..
`
`[0003] Digital PCR is a technique that allows amplification of a single DNA template
`
`from a minimally diluted sample, thus, generating amplicons that are exclusively derived
`
`from one template and can be detected with different fluorophores or sequencing to
`
`discriminate different alleles (e.g., wild type vs. mutant or paternal vs. maternal alleles).
`
`For a review of the digital PCR methodology, see, e.g., Pohl et al., Expert Rev. Mol.
`
`Diagn., 4(1):41-7 (2004). The basic premise of the techniqueis to divide a large sample
`
`into a number of smaller subvolumes (segmented volumes), whereby the subvolumes
`
`contain on average a single copy of a target. Then, by counting the number of positives
`
`in the subvolumes, one may deducethe starting copy number of the target in the
`
`starting volume. Most commonly, multiple serial dilutions of a starting sample are used
`
`to arrive at the proper concentration in the subvolumes, the volumes of which are
`
`typically determined by a given PCR apparatus. This additional step increases the
`
`number of samples to be processed. A set of subvolumes maybe tested that
`
`statistically represents the entire sample to reduce that number. However, under certain
`
`conditions, it may be necessary to detect very lowly expressed genes, resulting ina
`
`

`

`WO 2010/018465
`
`PCT/IB2009/007568
`
`large number of blank segmented volumesand, thus, a large number of subvolumesto
`
`be evaluated. While making a sample more concentrated is a possibility in this case,
`
`doing so mayintroducesignificant variability and losses (see, e.g., N. Blow, Nature
`
`Methods, 4:869-875 (2007). In addition, a more concentrated sample means that more
`
`sample is necessary to begin with.
`
`[0004] Further considerations suggest that decreasing the volume of the amplification
`
`reaction might improve sensitivity for detecting a single molecule. For example, the
`
`TaqMan® assay requires near-saturating amounts of PCR amplification product to
`detect fluorescence. PCR reactions normally saturate at about 10"! product
`
`molecules/microliter due, in part, to reannealing of product strands. To reach this
`concentration of product after 30 cycles in a 10 pl PCR requires at least 10° starting
`
`template molecules. If the volume of the PCR were reduced to ~10 nanoliters, then a
`
`single molecule could generate the required product to be detected by the TaqMan®
`
`assay. Attempts have been made to miniaturize PCR volumes(for a review, see, e.g.,
`
`Zhang et al., Nucl. Acids Res., 35(13):4223-4237 (2007)). Nevertheless, as sample
`
`volumes decrease, amplification becomes increasingly more prone to biochemical
`
`surface absorption problems due to the increasing surface-to-volume ratio, as well as
`
`potential other sources of variability.
`
`[0005] Therefore, there exists a need for methods and devices for accurately detecting
`
`or quantifying target copy numbers, including by meansofthe digital PCR.
`
`SUMMARYOF THE INVENTION
`
`[0006] The invention provides methods of conducting a nucleic acid reaction, including
`
`methods for performing digital PCR using “droplet-in-oil”’ technology, wherein a sample
`
`is segmented into droplets placed to a continuousflow ofcarrier fluid through a
`
`microfluidic channel. One example of such technology is described in PCT Pat. Appin.
`
`Pubs. WO 2007/091228 (corresponding US Serial No. 12/092,261), WO 2007/091230
`
`(USSN 12/093,132); and WO 2008/038259, and in the Examples. In some of these
`
`systems, termed “continuous flow PCR,” the droplets are fully wrappedin the carrier
`
`fluid throughout the reaction and detection. The invention is based, at least in part, on
`
`

`

`WO 2010/018465
`
`PCT/IB2009/007568
`
`the realization that sample droplets of 10-500 nl provide advantages for a PCR analysis
`
`of lowly expressed targets. Various aspects of the invention are described below.
`
`[0007]
`
`In certain embodiments, the methods include:
`
`a) providing a starting sample comprising a target nucleic acid to be
`
`detected;
`
`b) segmenting at least part of the sample to provide a set of sample droplets
`
`in a continuous flow of immiscible carrier fluid (€.g.,
`
`oil) through a channel
`
`(e.g., a capillary), each of said droplets containing on average about one
`
`or fewer copies of a target nucleic acid and reagents sufficient to perform
`
`a polymerase chain reaction;
`
`Cc) passing the droplets through a plurality of thermal zones thereby allowing
`
`the target nucleic acid, if present, to be amplified in each droplet; and
`
`d) detecting the presence or absenceof, and/or determining the amountof,
`
`the amplified target nucleic acid in the droplets while in the flow.
`
`[0008] The sample droplets have volumes of 0.1 pl-500 nl, preferably, 10-500 nl, more
`
`preferably, 30-350 nl, while the starting sample volumes are 0.05-5000 ul, preferably, 5-
`
`3500 ul. These volumes may include volumesof reagents (e.g., primer solution) added
`
`prior to the detection step. In certain embodiments, the droplets are spherical. In some
`
`embodiments, the droplets created by segmenting the starting sample are merged with
`
`a second set of droplets comprising one or more primersfor the target nucleic acid,
`
`thereby producing the final droplets for the amplification reaction.
`
`[0009]
`
`In some embodiments(“real-time detection”), the step of detecting or
`
`determining the amount is performed at multiple thermal cycles, thereby monitoring the
`
`amount of amplified target nucleic acid throughout the cycles, for example, in
`
`performing qPCR or real time PCR. Typically, the thermocycling is performed for at
`
`least a number of cycles required to reach the near-saturation level of the amplification.
`
`The number of cycles depends on the concentration of the target and other conditions
`
`and is typically between 20 and 40.
`
`[0010]
`
`In preferred embodiments (“end-point detection”), the step of detecting or
`
`determining the amountis performed after the near-saturation point is reached. For the
`
`

`

`WO 2010/018465
`
`PCT/IB2009/007568
`
`end-point detection, the starting copy numberof the target nucleic acid may be
`
`determined by counting the number of positively amplified droplets in a given set of
`
`droplets.
`
`[0011] The number of droplets in a set being analyzed is such that their combined
`
`volume is representative of the starting sample. The set of droplets contain the entire
`
`starting sample or only its part, depending on the number of copies of the target nucleic
`
`acid present or suspected to be present in the starting sample. The starting
`
`concentration of a target nucleic acid may be adjusted by diluting or by concentrating
`
`the starting sample. In illustrative embodiments, the set of droplets contains a train of 10
`
`droplets, and 0.005 ng/ul cDNAin the starting sample.
`
`[0012] The invention further provides methods of processingaplurality of starting
`
`samplesin parallel, wherein at least some of the starting samples have a) a varying
`
`concentration of the target nucleic acid and/or b) varying target nucleic acids.
`
`In some embodiments, sets of droplets from different starting samples form a train of
`
`alternating droplets in the continuous flow of the carrier fluid in the channel. These and
`
`other embodiments of the invention are described in detail below.
`
`BRIEF DESCRIPTION OF THE FIGURES
`
`[0013] Figure 1 illustrates amplification curves for a serially diluted sample at four
`
`concentrations of GAPDH cDNA: 5, 0.5, 0.05 and 0.005 ng/ul. The amplification was
`
`performed as described in Example 1. The intersection of each amplification curve with
`
`the threshold, as set at 0.3 here, defines the Ct value.
`
`[0014] Figure 2 shows Ct value as a function of concentrations based on the results
`
`shownin Figure 1. The slope of the line allows the amplification efficiency to be
`
`measured.
`
`[0015] Figure 3 showsfluorescence signals from four sets of ten droplets with
`
`decreasing template concentrations as shownin Figure 1. The plots compare
`
`fluorescence measurements taken at cycle 7 (Figure 3A) and 42 (Figure 3B).
`
`[0016] Figure 4 shows amplification traces based on the study described in Example 2.
`
`Figure 4A showsdroplet fluorescence traces of No Template Control (NTC) droplets
`
`followed by seven increasing sample concentrations at cycle 42. Figure 4B showsan
`
`

`

`WO 2010/018465
`
`PCT/IB2009/007568
`
`amplification curve of ten 1650 pg/ul droplets. Figure 4C illustrates the standard curve
`
`generated from Ct data showing near single molecule detection in 300 nl droplets.
`
`[0017] Figure 5 demonstrates a typical PCR thermal cycling trace for a set of 3,000
`
`droplets from a 300 ul sample having 0.005 ng/ul cDNA, performed as described in
`
`Example 3.
`
`General methods
`
`DETAILED DESCRIPTION OF THE INVENTION
`
`[0018] The invention provides methods of conducting a nucleic acid amplification
`
`reaction, such as PCR in a sample containing or suspected to contain a target nucleic
`
`acid to be detected. Although the methods described here employ the PCR as an
`
`amplification method of choice, alternative techniques of nucleic acid amplification may
`
`similarly be used in place of the PCR. Such techniques include for example, the ligase
`
`chain reaction (LCR), the transcription based amplification system (TAS), the nucleic
`
`acid sequence-based amplification (NASBA), the strand displacement amplification
`
`(SDA), rolling circle amplification (RCA), hyper-branched RCA (HRCA), etc.
`
`[0019]
`
`In general, the invention relates to the so-called “digital PCR” and similar
`
`methods that allow one to quantify the starting copy number of a nucleic acid template,
`
`by segmenting the starting sample to smaller reaction volumes, most of which contain
`
`one copyof the target or fewer.
`
`[0020] Methods of the invention may be used for determining the presence of the
`
`amount of a nucleic acid target, and for example, in gene expression analysis and is
`
`especially useful for lowly expressed genes.
`
`[0021] Generally, the methods of the invention include at least the following steps:
`
`a) providing a starting sample comprising a target nucleic acid to be
`
`detected;
`
`b) segmenting at least part of the sample to provide a set of sample droplets
`
`in a continuous flow of immiscible carrier fluid (e.g., oil) through a channel
`
`(e.g., a capillary), each of said droplets containing on average about one
`
`copy or fewer copies of a target nucleic acid and reagents sufficient to
`
`perform a polymerase chain reaction;
`
`

`

`WO 2010/018465
`
`PCT/IB2009/007568
`
`Cc) passing the droplets through a plurality of thermal zones thereby allowing
`
`the target nucleic acid, if present, to be amplified in each droplet; and
`
`d) detecting the presence or absenceof, and/or determining the amountof,
`
`the amplified target nucleic acid in the droplets while in the flow.
`
`[0022]
`
`In some embodiments(“real-time detection”), the step of the detecting or
`
`determining the amountis performed at multiple thermal cycles, thereby monitoring the
`
`amount of amplified target nucleic acid throughout the cycles, for example, in
`
`performing qPCR or real time PCR. Typically, the thermocycling is performed for at
`
`least a number of cycles required to reach the near-saturation level of the amplification.
`
`The number of cycles depends on the concentration of the target and other conditions
`
`and is typically between 20 and 40. The dependenceof Ct on the target concentration is
`
`illustrated in Example 1. In some such embodiments, a Ct value for the target nucleic
`
`acid is determined by detecting the course of amplification at each cycle. The “real-time”
`
`detection may be used for constructing the standard curve as well as for quantifying
`
`targets in test samples.
`
`[0023]
`
`In preferred embodiments (“end-point detection”), the step of the detecting or
`
`determining the amountis performed after the near-saturation point is reached. For the
`
`end-point detection, the starting copy numberof the target nucleic acid may be
`
`determined by counting the number of positively amplified droplets in a given set of
`
`droplets.
`
`Droplets and samples
`
`[0024] The starting sample contains (or is suspected to contain) at least one target
`
`nucleic acid. As used herein, the term “starting sample” refers to the sample from which
`
`droplets are generated. For example, the starting sample may be placed in a wellina
`
`conventional 384-well place, from which sample droplets are drawn. The starting
`
`sample volumes may vary, and may be, for example, 0.05-5000 ul, preferably, 5-3500
`
`ul, e.g., 5-1000 ul, 50-500 pl, 100-350 ul. In some embodiments, droplets created by
`
`segmenting the starting sample are merged with a second set of droplets comprising
`
`one more primers for the target nucleic acid to producefinal droplets. In some
`
`

`

`WO 2010/018465
`
`PCT/IB2009/007568
`
`embodiments, the starting sample contains at least 2, 5, 10, 100, 500, 1000 or more
`
`copies of the target nucleic acid.
`
`[0025] The sample droplets have volumesof 0.1 pl-500 ni, preferably, 1 pl-500 nl, 10
`
`pl-500 nl, 100 pI-500 ni, 1-500 nl, or 10-500 nl, more preferably, 30-350 nl. In certain
`
`embodiments, sample droplets have volumes of 50-500, 100-500, 150-500, 200-500,
`
`50-400, 100-400, 150-400, 200-400, 50-300, 100-300, 150-300, 200-300, or 150-250nl.
`
`These volumes may include volumesof reagents (e.g., primer solution) addedprior to
`
`the detection step. In some embodiments, the droplets are spherical, while in other
`
`embodiments, the droplets are elongated along the axis of the channel.
`
`[0026] The number of droplets in a set being analyzed is such that their combined
`
`volume is representative of the starting sample. The set of droplets contain the entire
`
`starting sample or only its part, depending on the number of copiesof the target nucleic
`
`acid present, or suspected to be present, in the starting sample. For example, a set of
`
`droplets containing, in total, 10% of the starting sample volume may be considered
`
`representative for a starting sample containing 100 copies of the target nucleic acid.
`
`Therefore, depending on the expected number of the set of analyzed droplets, the sets
`
`may contain several droplets to several thousand droplets. In some embodiments, the
`
`set of droplets for a given target nucleic acid contains, e.g., 5-10000 droplets, e.g., 100-
`
`5000, 5-1000, 100-500, 5-50, 6-30, 10-25, 8 or more, or 10 or more droplets. In other
`
`embodiments, at least 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, or 90% of the
`
`starting sample, or the entire starting sample, is segmentedinto droplets.
`
`[0027] The standard digital PCR aims at determining the dilution of the sample at which
`
`only half of segmented volumes are positive. This dilution indicates that the target
`
`nucleic acid is diluted on average to 4 per segmented volume. The present method
`
`allows one to analyze a much greater number of segmented volumes, many of which
`
`maybe blank. In some embodiments, at least 30%, 40%, 50% 60%, 70%, 80%, 90%,
`
`95%, 99% or more droplets in a set do not contain a target nucleic acid. For instance,
`
`for end-point detection, fewer than 50% (e.g., 30%, 20%, 10%, 5% or less) of droplets in
`
`the set of 10 or more (e.g., 20, 50, 100, 1000, 5000, or 10000) are positively amplified.
`
`Accordingly, in some embodiments, regardless of volume, the starting sample contains
`
`

`

`WO 2010/018465
`
`PCT/IB2009/007568
`
`one genome equivalent of nucleic acid or less. In certain applications, it may be
`
`assumedthat the mass of DNA is ~3 pg per genome (e.g., 3.3 pg/genome). F
`
`[0028] A sample may bedivided into replicates (e.g., duplicates, triplicates, etc.), in
`
`which the expression levels are measured. The sample may be derived from the same
`
`source and split into replicates prior to amplification. Additionally, one may create serial
`
`dilutions of the sample. Replicate and dilution samples may be analyzed in a serial or a
`
`parallel manner. In a parallel processing system, sets of droplets corresponding to
`
`separate starting samples form a sequenceof alternating droplets which pass througha
`
`thermal cycler, where droplets are being amplified, for example, as described in WO
`
`2008/038259. A plurality of starting samples with varying concentrations of the target
`
`nucleic acid and/or varying target nucleic acids may be processed in this manner in
`
`parallel.
`
`[0029] A sample may contain material from obtained cells or tissues, e.g., a cell or
`
`tissue lysate or extract. Extracts may contain material enriched in sub-cellular elements
`
`such as that from the Golgi complex, mitochondria, lysosomes, the endoplasmic
`
`reticulum, cell membrane, and cytoskeleton, etc. In some embodiments, the biological
`
`sample contains materials obtained from a single cell. Biological samples may come
`
`from a variety of sources. For example, biological samples may be obtained from whole
`
`organisms, organs, tissues, or cells from different stages of development, differentiation,
`
`or disease state, and from different species (human and non-human, including bacteria
`
`and virus). The samples may representdifferent treatment conditions (e.g., test
`
`compounds from a chemical library), tissue or cell types, or source (e.g., blood, urine,
`
`cerebrospinal fluid, seminal fluid, saliva, soutum, stool), etc. Various methods for
`
`extraction of nucleic acids from biological samples are known(see, é.g., Nucleic Acids
`
`Isolation Methods, Bowein (ed.), American Scientific Publishers (2002). Typically,
`
`genomic DNAis obtained from nuclear extracts that are subjected to mechanical
`
`shearing to generate random long fragments. For example, genomic DNA may be
`
`extracted from tissue or cells using a Qiagen DNeasy Blood & Tissue Kit following the
`
`manufacturer’s protocols.
`
`[0030]
`
`In case of the RNAanalysis (e.g., MRNA, siRNA, etc.), for example, as in the
`
`case of gene expression analysis, the nucleic acid is initially reverse-transcribed into
`
`

`

`WO 2010/018465
`
`PCT/IB2009/007568
`
`cDNAprior to conducing the PCR. This type of PCR is commonly referred to as “RT-
`
`PCR’and is illustrated in the Examples.
`
`Droplet-in-oil systems
`
`[0031] Any suitable device may be usedto practice the methods of the invention.
`
`Generally, a PCR device contains a sample preparation system, a thermocycler, and a
`
`detection unit. During sample preparation, the sample is segmented into droplets which
`
`are wrappedin immiscible fluid (e.g., silicone oil, mineral oil) which continuously flows
`
`through the channel, such as a capillary having a circular cross-section. The oil,
`
`enveloping each droplet, avoids cross contamination between the sequential droplets
`
`and carry-over contamination. The sample may be pre-mixed with the primer, or the
`
`primer may be added to the droplet. In some embodiments, droplets created by
`
`segmenting the starting sample are merged with a second set of droplets comprising
`
`one or more primers for the target nucleic acid in order to produce final droplets. The
`
`merging of droplets can be accomplished using, for example, one or more liquid bridges
`
`as described in WO 2007/091230 (USSN 12/093,132) and WO 2008/038259.
`
`[0032] A queue of droplets from the preparation system may be passed through the
`
`thermal cycler. The velocity of the sample through the device is defined by the control of
`
`the velocity of the carrier fluid is controlled by an external pumping system. The sample
`
`undergoes the same thermal cycling and chemical reaction as it passes through N
`
`amplification cycles of the complete thermal device. This results in a maximum two-fold
`amplification after each cycle and a total amplification of I(1+E)‘ where| is theinitial
`
`product, E is the efficiency of the reaction and N is the number of cycles. Fluorescent
`
`probes are contained in each sample droplet. The fluorescence level is detected in each
`
`droplet at each cycle, e.g., in the case of real-time PCR. This mayinvolve the use of
`
`fluorescent probes, such as Taqman® probes, and intercollating fluorescent dyes, such
`
`as SYBR Green and LCGreen®,as described in, e.g., in US Patent Nos. 5,723,591 and
`
`5,928,907; www.idahotech.com; Gudnason et al., Nucleic Acids Res., 35(19):e127
`
`(2007); and in the Examples.
`
`[0033] An exemplary system for use with the method of the invention is described, for
`
`example, PCT Patent Application Pubs. WO 2007/091228 (USSN 12/092,261); WO
`
`

`

`WO 2010/018465
`
`PCT/IB2009/007568
`
`2007/091230 (USSN 12/093,132); and WO 2008/038259. One such system is made by
`
`Stokes Bio (www.stokebio.ie). Other exemplary systems suitable for use with the
`
`methods of the invention are described, for example, in Zhang et al. Nucleic Acids Res.,
`
`35(13):4223-4237 (2007) and include those made by Fluidigm (www.fluidigm.com),
`
`RainDance Technologies (www.raindancetechnologies.com), Microfluidic Systems
`
`(www. microfluidicsystems.com); Nanostream (www.nanostream.com); and Caliper Life
`
`Sciences (www.caliperls.com). For additional systems, see, é.g., Wang et al., J.
`
`Micromech. Microeng., 15:1369-1377 (2005); Jia et al., 38:2143-2149 (2005); Kim etal.,
`
`Biochem. Eng. J., 29:91-97; Chen et al., Anal. Chem., 77:658-666; Chen et al., Analyst,
`
`130:931-940 (2005); Munchowet al., Expert Rev. Mol. Diagn., 5:613-620 (2005); and
`
`Charbert et al., Anal. Chem., 78:7722-7728 (2006); and Dorfman etal., Anal. Chem,
`
`77:3700-3704 (2005).
`
`[0034] The following Examples provide illustrative embodiments of the invention and
`
`do not in any waylimit the invention.
`
`EXAMPLES
`
`Example 1: Measurement of qPCR Amplification Efficiency by Serial Dilution
`
`[0035] Total RNA is extracted from cultured cells, reverse transcribed into cDNA and
`
`used as the template for the qPCR reaction. The starting concentration of the template
`
`is 5 ng/pl which is then diluted 10-fold to 0.5 ng/ul. This 10-fold dilution is repeated
`
`yielding samples with four concentrations of cDNA template: 5 ng/ul, 0.5 ng/ul, 0.05
`
`ng/ul, and 0.005 ng/ul. The resulting amplification curves, obtained using a Stokes Bio
`
`device (www.stokesbio.ie), are shown in Figure 1. As seen from the figure, for lower
`
`starting template concentrations, more cycles of PCR are necessary to bring the
`
`fluorescence signal to the threshold level. In fact, if the PCR reaction was 100% efficient
`
`a 0.5 ng/ul sample would require 3.32 more cycles of PCR to reach the same threshold
`
`as a 5 ng/ul sample (10 times the 0.5 ng/ul sample). The fractional cycle at which the
`
`amplification curve reaches the threshold is called the Ct value: a measure of the gene
`
`expression of the sample. The threshold is set at 0.3 read off of the corresponding Ct
`
`values for each set of amplification curves in Figure 1. The Ct value is plotted against
`
`the cDNA concentration in Figure 2. Analysis of this data implies that the slope of the
`
`10
`
`

`

`WO 2010/018465
`
`PCT/IB2009/007568
`
`bestfit line is -3.53 + 0.13, estimating the amplification efficiency to be 92% + 4%, which
`
`is within the range expected of GPCR on a standard PCT system, such as Applied
`
`Biosystems’ 7900HT.
`
`[0036] Figure 3A showsthe fluorescence signal from cycle 7 of a 50 cycle amplification
`
`using the Stokes Bio device. Each of the small spikes represents a droplet passing. The
`
`far left spike represents the lead droplet. The concentrations of each set of droplets
`
`range from 0.005 to 5 ng/ul, as indicated in the figure. Two aspects of the data are
`
`clearly shown--there is background fluorescencefrom the oil filled capillary, and
`
`background signal from each droplet- both are constant. Figure 3B shows the same
`
`droplets at cycle 42 when amplification is complete. Note that the signal from each
`
`droplet in a set of ten varies from droplet-to-droplet. This is because the optical system
`
`for each channel is not identical. This effect is easily normalized out during data
`
`processing. At the highest level of dilution (0.005 ng/uL), only four out of ten droplets
`
`have amplified. This is a statistical effect, expected for this level of template dilution and
`
`droplet volume. The data processing algorithm is designed to take this into account.
`
`Example 2: High Throughput qPCR Performance Validation
`
`[0037] The analysis of gene expression is an essential element of functional genomics,
`
`and qPCR-based expression profiling is the gold standard for the precise monitoring of
`
`selected genes. Gene expression relies upon the reverse transcription of MRNAto
`
`cDNA. It is, however, generally not possible to use cDNAasa standardfor absolute
`
`quantification of mRNA becausethere is no control for the efficiency of the reverse
`
`transcription step. This Example presents Stokes Bio’s amplification representative
`
`performance data from genomic DNA (gDNA), a commonly usedin standard qPCR.
`
`[0038] A TaqMan® RNase P gene primer and probeset is used to evaluate instrument
`
`performance. The RNase P gene is a single-copy gene encoding the RNA moiety for
`
`the RNase P enzyme. Several two-fold dilutions are created from a stock of a known
`
`gDNA copy number. These dilutions were used to prepare complete qPCR reactions for
`
`amplification in the Stokes Bio instrument. Table 1 shows gDNA template
`
`concentrations with corresponding mean Cts and estimated starting copy numbersfor
`
`11
`
`

`

`WO 2010/018465
`
`PCT/IB2009/007568
`
`each of the seven reaction sets. A no template control (NTC) is also included and
`
`showed no amplification. Figure 4A shows droplet fluorescence traces of three NTC
`
`reactions with seven sets of ten 300 nl droplets. Each 10x replicate set was taken from
`
`a different concentration sample. The inset plot shows excellent data resolution for each
`
`300 ni droplet. Figure 4B shows a normalized amplification plot of the most
`
`concentrated reaction set with a Ct of 26.4. Figure 4B showsa standard curve taken
`
`from the Ct data in Table 1. Error bars indicate the Ct range in each of the 10x replicate
`
`sets. Ct variability is attributable to Poisson noise or a variation in the number of starting
`
`copies from one droplet to another at a given concentration.
`
`Table 1
`
`.
`
`Copies”
`
`26.4
`
`sDNA Stancerd|per 8001
`
`Pg'y
`Reaction
`1650
`
`4 f 1
`
`
`
`7
`
`5
`
`9 4|1
`
`03|309|34|94|
`|515|316|O76|47|
`
`*Determined using 1 copy per 3.3 pg
`
`Example 3: Digital PCR End-Point Detection
`
`[0039] The premise for this technique is to divide a large volume into a discreet number
`
`of smaller volumes reducing the number of copies in each sample, following the
`
`amplification process to perform fluorescence detection on the emerging droplets. The
`
`resulting total number of droplets with amplification can then be used to determine
`
`starting copy number. It also can be usedfor rare target detection wherein the statistic
`
`probability of amplification is increased for the rare target as the number of background
`
`molecules is reduced by the division.
`
`[0040] Using a statistical prediction model the probability of the distributed target
`
`molecules in the segment droplets can be generated. This is particularly of benefit for
`
`12
`
`6 |.75|
`
`

`

`WO 2010/018465
`
`PCT/IB2009/007568
`
`low concentration samples as it provides a prediction of the number of droplets
`
`containing molecules and thus the number of droplets expected to fluoresce.
`
`[0041] The binomial distribution model employedis a discrete probability distribution
`
`that arises in many common situations. The recognized example of binomial distribution
`
`is counting the number of heads in a fixed number of independent coin tosses, while in
`
`this case, it is counting the number of copies in a fixed known number of droplets
`
`created from the original sample. In a series of m independent trials, or n independent
`
`copies, each trial or copy results in a success (the outcome that is counted) orfailure. In
`
`other words, each copy has two possible outcomes, it enters the monitored droplet or
`
`not. For a sample segmented into 3000 droplets, each copy has the same probability, p,
`
`of success, 1
`
`in 3000, for each monitored droplet. The binomial distribution model
`
`counts the number of successesin a fixed number oftrials. Binomial distribution is
`
`completely determined by two parameters; n, the number of cDNA copies in the main
`
`volume, and p, the success probability common to each copy. As a consequence,
`
`knowing the number of copies and the number of droplets allows a binomial distribution
`
`to be used.
`
`[0042] Figure 5 demonstrates a typical trace for a 0.005 ng/ul sample. The data
`
`demonstrated here uses a 300 ul sample volume whichis divided into ~3000 droplets of
`
`100 nl droplet volumes.
`
`[0043]
`
`In another experiment, digital PCR involves amplifying a single DNA template
`
`from minimally diluted samples, generating amplicons that are exclusively derived from
`
`one template. It transforms an exponential, analog signal obtained from conventional
`
`PCR to linear, digital signals, thus allowing statistical analysis of the PCR product.
`
`[0044] To determine the mass of genomic DNA which corresponds to copy numbers of
`
`ta