Note: Descriptions are shown in the official language in which they were submitted.
CA 02816702 2013-05-01
WO 2012/061442 PCT/US2011/058854
1
ANALYSIS OF FRAGMENTED GENOMIC DNA IN DROPLETS
Cross-References to Priority Applications
This application claims the priority of the following earlier applications:
U.S. Provisional Patent Application Serial No. 61/409,106, filed November 1,
2010; and U.S. Patent Application Serial No. 12/976,827, filed December 22,
2010, published as U.S. Patent Application Publication No. 2011/0217712 Al
on September 8, 2011. Both of these patent applications are incorporated
herein by reference in their entireties for all purposes.
Cross-References to Additional Materials
This application incorporates herein by reference in their entirety for all
purposes the following materials: U.S. Patent No. 7,041,481, issued May 9,
2006; U.S. Patent Application Publication No. 2010/0173394 Al, published
July 8, 2010; PCT Patent Application No. WO 2011/120024, published
September 29, 2011; and Joseph R. Lakowicz, PRINCIPLES OF FLUORESCENCE
SPECTROSCOPY (2nd Ed. 1999).
Introduction
Many biomedical applications rely on high-throughput assays of
samples for nucleic acid targets. For example, in research and clinical
applications, high-throughput genetic tests using target-specific reagents can
provide accurate and precise quantification of nucleic acid targets for drug
discovery, biomarker discovery, and clinical diagnostics, among others.
Emulsions hold substantial promise for revolutionizing high-throughput
assays for targets. Emulsification techniques can create large numbers of
aqueous droplets that function as independent reaction chambers for
biochemical reactions. For example, an aqueous sample (e.g., 20 microliters)
can be partitioned into droplets (e.g., 20,000 droplets of one nanoliter each)
to
allow an individual test for the target to be performed with each of the
droplets.
Aqueous droplets can be suspended in oil to create a water-in-oil
emulsion (W/O). The emulsion can be stabilized with a surfactant to reduce
coalescence of droplets during heating, cooling, and transport, thereby
CA 02816702 2013-05-01
WO 2012/061442 PCT/US2011/058854
2
enabling thermal cycling to be performed. Accordingly, emulsions have been
used to perform single-copy amplification of nucleic acid target molecules in
droplets using the polymerase chain reaction (PCR). Digital assays are
enabled by the ability to detect the presence of individual molecules of a
target in droplets.
In an exemplary droplet-based digital assay, a sample is partitioned
into a set of droplets at a limiting dilution of a target (i.e., some of the
droplets
contain no molecules of the target). If molecules of the target are
distributed
randomly among the droplets, the probability of finding exactly 0, 1, 2, 3, or
more target molecules in a droplet, based on a given average concentration of
the target in the droplets, is described by a Poisson distribution.
Conversely,
the concentration of target molecules in the droplets (and thus in the sample)
may be calculated from the probability of finding a given number of molecules
in a droplet.
Estimates of the probability of finding no target molecules and of
finding one or more target molecules may be measured in the digital assay. In
a binary approach, each droplet can be tested to determine whether the
droplet is positive and contains at least one molecule of the target, or is
negative and contains no molecules of the target. The probability of finding
no
molecules of the target in a droplet can be approximated by the fraction of
droplets tested that are negative (the "negative fraction"), and the
probability
of finding at least one target molecule by the fraction of droplets tested
that
are positive (the "positive fraction"). The value of the positive fraction or
the
negative fraction then may be utilized in a Poisson algorithm to calculate the
concentration of the target in the droplets. In other cases, the digital assay
may generate data that is greater than binary. For example, the assay may
measure how many molecules of the target are present in each droplet with a
resolution greater than negative (0) or positive (>0) (e.g., 0, 1, or >1
molecules; 0, 1, 2, or >2 molecules; or the like).
For a combination of high throughput and accuracy in droplet-based
DNA assays of different samples, droplets should be generated rapidly and
with a uniform size (i.e., monodisperse droplets). However, sample
CA 02816702 2013-05-01
WO 2012/061442 PCT/US2011/058854
3
components can interfere with the ability of droplets to separate from the
bulk
sample phase, particularly as the frequency of droplet generation is
increased. As a result, the size of droplets formed, or even the ability to
form
droplets at all, can vary from sample to sample, diminishing the reliability
of
the assays. New approaches are needed to provide reliable and consistent
generation of droplets at a higher generation frequency.
Summary
The present disclosure provides a method of analyzing genomic DNA.
Genomic DNA including a target may be obtained. The genomic DNA may be
fragmented volitionally to produce fragmented DNA. The fragmented DNA
may be passed through a droplet generator to generate aqueous droplets
containing the fragmented DNA. A digital assay may be performed on the
droplets to determine a level of the target. In some embodiments, the droplets
may contain the genomic DNA at a concentration of at least about five
nanograms per microliter, the droplets may be generated at a droplet
generation frequency of at least about 50 droplets per second, the droplets
may have an average volume of less than about 10 nanoliters per droplet, the
droplets may generated at a sample flow rate of greater than about 50
nanoliters per second, or any combination thereof.
Brief Description of the Drawings
Figure 1 is a flowchart illustrating an exemplary method of analyzing
genomic DNA, in accordance with aspects of the present disclosure.
Figure 2 is a matrix of drawings made from photographs of a droplet
generator processing three different samples at each of four different driving
pressures.
Figure 3 is a graph of droplet volume plotted as a function of droplet
generation frequency for samples containing no genomic DNA or genomic
DNA (Raji or Coriell) that is digested (EcoRI) or undigested.
Figure 4 is a graph of droplet volume plotted as a function of sample
flow rate for the samples of Figure 3.
Figure 5 is a graph of maximum extension plotted as a function of
droplet generation frequency for the samples of Figure 3.
CA 02816702 2013-05-01
WO 2012/061442 PCT/US2011/058854
4
Figure 6 is a graph of maximum extension plotted as a function of
sample flow rate for the samples of Figure 3.
Detailed Description
The present disclosure provides a method of analyzing genomic DNA.
Genomic DNA including a target may be obtained. The genomic DNA may be
fragmented volitionally to produce fragmented DNA. The fragmented DNA
may be passed through a droplet generator to generate aqueous droplets
containing the fragmented DNA. An assay may be performed on the droplets
to determine a level of the target. In some embodiments, the droplets may
contain the genomic DNA at a concentration of at least about five nanograms
per microliter, the droplets may be generated at a droplet generation
frequency of at least about 50 droplets per second, the droplets may have an
average volume of less than about 10 nanoliters per droplet, the droplets may
generated at a flow rate of greater than about 50 nanoliters per second, or
any combination thereof.
The method of analyzing genomic DNA in droplets, as disclosed
herein, has substantial advantages over other droplet-based approaches. The
advantages may include generating droplets at a higher frequency, with
greater monodispersity, with a higher load of DNA, and/or with substantially
less interference from genomic DNA.
These and other aspects of the present disclosure are described in the
following sections: (I) overview of an exemplary method of genomic DNA
analysis, (II) exemplary data from tests of droplet generation, and (III)
selected embodiments.
I. Overview of an Exemplary Method of Genomic DNA Analysis
Figure 1 shows a flowchart illustrating an exemplary method 20 of
analyzing genomic DNA. The steps presented may be performed in any
suitable order and in any suitable combination.
Genomic DNA may be obtained, indicated at 22. The DNA may be
obtained from any suitable organism, such as a mammal (e.g., human,
mouse, rat, monkey, etc.), a non-mammalian vertebrate, an invertebrate, a
yeast or fungus, a plant, a protozoan, a bacterium, or the like. The DNA may
CA 02816702 2013-05-01
WO 2012/061442 PCT/US2011/058854
be obtained by any suitable process, such as purchased commercially,
received as a gift, acquired by extraction from cells or fluid, received as a
clinical sample, or the like. The DNA may be obtained in a relatively high
molecular weight form, such as having a molecular weight of at least about
5 104, 105 or 106 kilodaltons, among others (e.g., having an average length
of at
least about 25, 50, 100, 200, 500, or 1,000 kilobases).
The genomic DNA may be fragmented, indicated at 24, before droplet
generation. Fragmentation may be a volitional act, that is, performed
deliberately. Fragmentation generally involves any procedure that
substantially reduces the molecular weight of the genomic DNA, such as by
cutting or breaking DNA strands. The fragmentation may reduce the average
molecular weight and/or length by any suitable amount, such as at least about
5, 10, 20, 50, or 100-fold, among others. An exemplary approach to
fragmenting genomic DNA includes digestion with a restriction enzyme (e.g.,
an enzyme having a 4, 5, 6 or 8 nucleotide recognition site, among others).
The target may contain no recognition sites for the restriction enzyme, to
avoid any cleavage of target molecules. The restriction enzyme digestion may
be performed to completion or may be a partial digestion. Alternatively, or in
addition, an aqueous sample of the genomic DNA may be heated to fragment
the DNA. Exemplary heating that fragments the DNA may be performed at a
temperature of at least 9500, for at least about 10, 15, 20, or 30 minutes,
among others. In other cases, the DNA may be fragmented by shearing,
sonicating, nebulizing, irradiating, or the like.
The genomic DNA may include a target, generally a sequence of
interest to be tested. Fragmentation of the genomic DNA may be performed
without substantially disrupting the target, meaning that less than one-half
of
target sequences in the genomic DNA are disrupted (e.g., broken or cut) by
the fragmentation process.
Droplets containing the fragmented DNA may be generated, indicated
at 26. The droplets may be generated serially with each of one or more
droplet generators. The fragmented DNA may be passed through at least one
droplet generator to generate droplets. Generally, the fragmented DNA is
CA 02816702 2013-05-01
WO 2012/061442 PCT/US2011/058854
6
disposed in an aqueous sample, and the aqueous sample and an immiscible
continuous phase are passed through the droplet generator to form aqueous
droplets containing the fragmented DNA and disposed in the continuous
phase. Further aspects of droplet generators and emulsion phases that may
be suitable are described in the documents listed above under Cross-
References, which are incorporated herein by reference, particularly U.S.
Patent Application Publication No. 2010/0173394 Al, published July 8, 2010;
and PCT Patent Application No. WO 2011/120024, published September 29,
2011.
The droplets may have any suitable size. For example, the droplets
may have an average volume of less than about 1 pL, 100 nL, 10 nL, 1 nL,
100 pL, 10 pL, or 1 pL, among others. Alternatively, or in addition, the
droplets
may have an average volume of greater than about 10 fL, 100 fL, 1 pL, 10 pL,
or 100 pL, among others. In some cases, the droplets may have an average
volume of about 1 pL to 100 nL, 1 pL to 10 nL, or 0.1 to 10 nL, among others.
The droplets may be monodisperse.
The droplets may contain any suitable concentration of fragmented
DNA. For example, the fragmented DNA may be disposed in the droplets at a
concentration of at least about 0.1, 0.2, 0.5, 1, 2, 5, 10, 20, 50 ng/pL,
among
others. In some cases, the concentration may be about 0.1-50 or 0.2-20
ng/pL. Fragmenting the DNA allows a higher DNA load to be incorporated into
droplets. The fragmented DNA may be present at an average of less than
about two genome-equivalents per droplet. The target may be present at an
average of less than about two molecules per droplet.
The droplets may be formed at any suitable droplet generation
frequency, such as at least about 10, 20, 50, 100, 200, 500, or 1,000 Hz
(droplets/second), among others. Generally, the droplet generation frequency
is inversely related to the size of droplets being generated, with smaller
droplets allowing a higher droplet generation frequency.
An aqueous sample used to form the droplets (and containing the
fragmented DNA) may be passed through the droplet generator and/or
converted into droplets at any suitable flow rate. Exemplary flow rates that
CA 02816702 2013-05-01
WO 2012/061442 PCT/US2011/058854
7
may be suitable include at least about 1, 5, 10, 20, 50, 100, 200, 500, 1,000,
5,000, or 10,000 nL/second, among others. Generally, the sample flow rate is
directly related to the size of droplets being generated, with larger droplets
permitting a higher flow rate.
The exemplary values (or ranges) for droplet volumes, DNA
concentrations, droplet generation frequencies, and flow rates listed above
may be combined in any suitable combination(s).
An assay may be performed on the droplets, indicated at 28. The
assay may be a digital assay that detects individual target molecules in the
droplets. The digital assay may involve amplifying target molecules, such as
by PCR or a ligase chain reaction, among others. The digital assay also may
involve detecting fluorescence from the droplets. The assay further may
involve determining a level (e.g., a concentration) of the target in the
droplets
with a Poisson algorithm.
II. Exemplary Data from Tests of Droplet Generation
This section present exemplary data from tests of droplet generation
with genomic DNA, with or without fragmentation; see Figures 2-6.
Droplet generation in a microfluidic device may depend on the flow rate
at which the sample travels to a droplet generator, and on the frequency with
which droplets are generated. At high flow rates, a sample stream may jet into
the immiscible continuous phase, no longer generating droplets. At generation
rates close to the jetting limit, the sample starts to extend deeper into the
outlet channel before droplets are generated. This extension length can be
used to see how close a set of generation conditions is to the jetting limit.
Figure 2 shows a matrix of drawings made from photographs of a
droplet generator 30 processing three different aqueous samples 32-36 at
each of four different driving pressures and thus flow rates. The three
samples
are (a) a control sample 32 containing no DNA (PCR buffer with no template),
(b) an aqueous sample 34 of human genomic DNA (Raji, 18.75 ng/pL) that is
undigested and has a high molecular weight (MW), and (c) an aqueous
sample 36 of human genomic DNA (Raji, 18.75 ng/pL) that has been digested
with a restriction enzyme and has a reduced molecular weight. The four
CA 02816702 2013-05-01
WO 2012/061442 PCT/US2011/058854
8
driving pressures (1, 2, 3, and 4) are negative (vacuum) pressures applied
downstream of the droplet generator and are expressed in pounds per square
inch (psi), with 1 psi equal to about 6.9 kilopascals. The vacuum level may
control the sample flow rate, the total flow rate, and the droplet generation
frequency.
Droplet generator 30 may be formed by a channel network composed
of a sample inlet channel 38, at least one or a pair of oil inlet channels 40,
42,
and an outlet channel 44. Inlet channel 38 carries a bulk aqueous phase 46 of
aqueous sample 32, 34, or 36 to the droplet generator. Inlet channels 40, 42
carry a continuous phase 48 (e.g., oil with a surfactant) to the droplet
generator. Outlet channel 44 carries droplets 50 in continuous phase 48 away
from a channel intersection 52.
The top row shows droplet generation of control sample 32, which
contains no genomic DNA. Droplets 50 are approximately 1 nL and do not
vary much in size with the different vacuum levels.
The middle row shows droplet generation with sample 34 containing
undigested genomic DNA. The genomic DNA strongly impairs droplet
generation, which occurs only at the lowest vacuum level tested (1 psi). Even
at this lowest level, there is a considerable extension of bulk aqueous phase
46 past channel intersection 52, indicated by an arrow at 54, and droplets 50
are larger. At higher vacuum levels (e.g., compare 1 psi with 2-4 psi) and
flow
rates, no droplets are generated, because the sample stream jets into outlet
channel 44, indicated by an arrow at 56, without breaking up into droplets.
Accordingly, the presence of human genomic DNA can strongly interfere with
droplet generation, and may require use of lower DNA concentrations, flow
rates, and droplet generation frequencies. As a consequence, sample
processing may be slowed considerably. Also, the frequency of target-positive
droplets in the emulsion may be reduced substantially (due to the lower DNA
concentration), which would require more droplets to be analyzed to achieve
to the same confidence for the target level determined.
The bottom row shows droplet generation with sample 36 containing
the same concentration (mass per unit volume) of genomic DNA as sample
CA 02816702 2013-05-01
WO 2012/061442 PCT/US2011/058854
9
34, but after the DNA has been digested with a restriction enzyme into shorter
fragments. At the pressures (and flow rates) shown here, the genomic DNA in
fragmented form does not detectably impair droplet generation. The sample
produces droplets 50 that are similar to control sample 32 lacking DNA.
Further studies were conducted to quantitatively measure relationships
among the generation vacuum, sample flow rate, droplet generation
frequency, velocity in the outlet channel, droplet size, and maximum sample
extension during droplet generation. The aqueous samples used were
Spectral Dye Buffer (the same control sample as in Figure 2, but without DNA
polymerase), Raji human genomic DNA (Loftstrand Laboratories) ("Raji"), and
19205 human DNA (Coriell Institute) ("CoreII"). DNA samples were either
undigested or digested with a restriction enzyme, EcoRl. DNA digestion was
performed with a 20 U/pL concentration of EcoRI (New England Biolabs) in
NEB #4 buffer, with the genomic DNA at a final concentration of 200 ng/pL.
The mixture was incubated at 37 C for one hour, and then was diluted to
various final concentrations.
The graphs of Figures 3-6 show the results of droplet generation
experiments performed with samples of Master Mix (no DNA), EcoRI-digested
Raji DNA at 18.75 ng/pL, EcoRI-digested Coriell 19205 DNA at 18.75 ng/pL,
undigested Raji DNA at 18.75 ng/pL, and undigested Coriell 19205 DNA at
18.75 ng/pL. Figure 3 shows a graph of droplet volume plotted as a function of
droplet generation frequency for the samples. Figure 4 shows a graph of
droplet volume plotted as a function of sample flow rate for the samples.
Figure 5 shows a graph of maximum extension plotted as a function of droplet
generation frequency for the samples. Figure 6 shows a graph of maximum
extension plotted as a function of sample flow rate for the samples of Figure
3.
For the undigested human genomic DNA, only very low droplet
generation frequencies or sample flow rates were possible before jetting
occurred. For instance, for Coriell 19205 DNA, the maximum was 60 Hz, and
83 nL/sec. For Raji DNA, the maximum was 120 Hz, and 162 nL/sec.
However, even below these limits the generated droplets had a higher volume
CA 02816702 2013-05-01
WO 2012/061442 PCT/US2011/058854
than in the absence of DNA, and generation occurred with much longer
sample extension into the output channel.
The graphs also show that for digested DNA at the same
concentration, the effects of DNA on droplet generation are not detectable. No
5 jetting
or long sample extensions were observed at any of the flow rates or
generation frequencies that were tested, and the droplet volumes are the
same as with the sample with no DNA present.
These results show that droplet generation is strongly impaired in the
presence of undigested human DNA, but not after digestion with a restriction
10 enzyme.
III. Selected Embodiments
This section describes selected embodiments of the present disclosure
as a series of indexed paragraphs. These embodiments should not limit the
entire scope of the present disclosure.
A. A method of
analyzing genomic DNA, comprising: (i) obtaining
genomic DNA including a target; (ii) fragmenting the genomic DNA volitionally
to produce fragmented DNA; (iii) passing the fragmented DNA through at
least one droplet generator to generate aqueous droplets containing the
fragmented DNA; and (iv) performing a digital assay on the droplets to
determine a level of the target.
B. The method of paragraph A, wherein the droplets have an
average volume of less than about 10 nanoliters.
C. The method of paragraph A, wherein the droplets contain the
genomic DNA at a concentration of at least about 5 nanog rams per microliter.
D. The method of
paragraph A, wherein the genomic DNA is
disposed in an aqueous sample, and wherein the droplets are generated at a
flow rate of greater than about 50 nanoliters per second of the aqueous
sample through the droplet generator.
E. The
method of any of paragraphs A to D, wherein the droplets
are generated at a droplet generation frequency of at least about 50 droplets
per second.
CA 02816702 2013-05-01
WO 2012/061442 PCT/US2011/058854
11
F. The method of paragraph A, wherein the droplets have an
average volume of less than about 10 nanoliters and contain the genomic
DNA at a concentration of at least about 5 nanog rams per microliter.
G. The method of paragraph A, wherein the droplets have an
average volume of less than about 10 nanoliters, wherein the genomic DNA is
disposed in an aqueous sample, and wherein the droplets are generated at a
flow rate of greater than about 50 nanoliters per second of the aqueous
sample through the droplet generator.
H. The method of paragraph A, wherein the droplets contain the
genomic DNA at a concentration of at least about 5 nanograms per microliter,
wherein the genomic DNA is disposed in an aqueous sample, and wherein
the droplets are generated at a flow rate of greater than about 50 nanoliters
per second of the aqueous sample through the droplet generator.
I. The method of paragraph F, wherein the genomic DNA is
disposed in an aqueous sample, and wherein the droplets are generated at a
flow rate of greater than about 50 nanoliters per second of the aqueous
sample through the droplet generator.
J. The method of paragraph F, wherein the droplets are generated
at a droplet generation frequency of at least about 50 droplets per second.
K. The method of
paragraph G, wherein the droplets are generated
at a droplet generation frequency of at least about 50 droplets per second.
L. The method of paragraph H, wherein the droplets are generated
at a droplet generation frequency of at least about 50 droplets per second.
M. The method of paragraph L, wherein the droplets have an
average volume of less than about 10 nanoliters.
N. The method of any of paragraphs A to M, wherein the step of
fragmenting includes a step of digesting the genomic DNA with a restriction
enzyme.
0. The
method of paragraph N, wherein the restriction enzyme cuts
the genomic DNA an average of less than about once every kilobase.
P. The
method of any of paragraphs A to M, wherein the step of
fragmenting includes a step of shearing the genomic DNA.
CA 02816702 2013-05-01
WO 2012/061442 PCT/US2011/058854
12
Q. The method of any of paragraphs A to M, wherein the step of
fragmenting includes a step of son icating the genomic DNA.
R. The method of any of paragraphs A to Q, wherein the droplets
contain an average of less than about two copies of the target per droplet.
S. The method of
any of paragraphs A to R, wherein the droplets
contain an average of less than about two genome-equivalents of the
genomic DNA per droplet.
T. The
method of any of paragraphs A to S, wherein the step of
fragmenting does not disrupt the target substantially.
U. The method of
any of paragraphs A to T, wherein the step of
performing a digital assay includes a step of amplifying the target in the
droplets.
V. The
method of paragraph U, wherein the target is amplified by
PCR.
W. The method of
any of paragraphs A to V, wherein the step of
performing a digital assay includes a step of detecting fluorescence of the
droplets.
X. The method of any of paragraphs A to W, wherein the step of
performing a digital assay includes a step of determining a level of the
target
with a Poisson algorithm.
Y. The method of any of paragraphs A to X, wherein the droplets
have an average volume of about 0.1 to 10 nanoliters.
Z. A method of partitioning an aqueous sample comprising DNA
into droplets, the method comprising: (i) obtaining a sample comprising DNA
at a concentration of at least about 5 ng per microliter; (ii) fragmenting the
DNA volitionally to produce fragmented DNA; and (iii) passing the sample
through a droplet generator, to generate aqueous droplets containing the
fragmented DNA, the droplets being generated at a droplet generation
frequency of at least about 50 droplets per second and having an average
volume of less than about 10 nanoliters.
Al. A
method of partitioning an aqueous sample comprising DNA
into droplets, the method comprising: (i) obtaining a sample comprising
CA 02816702 2013-05-01
WO 2012/061442 PCT/US2011/058854
13
genomic DNA; (ii) fragmenting the DNA volitionally to produce fragmented
DNA; and (iii) passing the sample through a droplet generator, to generate
aqueous droplets containing the fragmented DNA, the droplets being
generated at a droplet generation frequency of at least about 50 droplets per
The disclosure set forth above may encompass multiple distinct
15 subcombinations of the various elements, features, functions, and/or
properties disclosed herein. The following claims particularly point out
certain
combinations and subcombinations regarded as novel and nonobvious.
Inventions embodied in other combinations and subcombinations of features,
functions, elements, and/or properties may be claimed in applications claiming
