 |
Claims  |
|
|
What is claimed is:
1. A method of determining the sequence of DNA which comprises the steps of
processing said DNA to form four sets of DNA sequencing fragments where one
set contains fragments terminating in G, a second set contains fragments
terminating in A, a third set contains fragments terminating in T, and a
fourth set contains fragments terminating in C, such that selected mole
fractions of each one of selected sets of sequencing fragments is tagged
with a first fluorophore which fluoresces at a first wavelength and such
that different selected mole fractions of each one of selected sets of
sequencing fragments is tagged with a second fluorophore which fluoresces
at a second wavelength and in which one of said sets of sequencing
fragments is comprised of a mixture of selected mole fractions of
fragments where one mole fraction is labeled with the first fluorophore
and the other mole fraction is labeled with the second fluorophore,
separating said sets of tagged DNA sequencing fragments in a single channel
or lane, and
determining the sequence of said sequencing fragments by detecting the
ratio of fluorescence intensity at the first and second wavelengths as a
function of time or position.
2. The method of claim 1 where the separation is performed
electrophoretically in a capillary, gel-filled capillary or slab gel.
3. The method of claim 1 where the tagging of the sets of DNA sequencing
fragments with the first and second fluorophores is performed by
synthesizing a selected set of sequencing fragments by using selected mole
fractions of two DNA primers, one of which is labeled with the first
fluorophore and the other of which is labeled with the second fluorophore.
4. The method of claim 1 where the tagging of the sets of DNA sequencing
fragments with the first and second fluorophores is performed by using
selected mole fractions of two different fluorescently labeled terminators
in the synthesis of the selected set of sequencing fragments, one of which
is labeled with the first fluorophore and the other of which is labeled
with the second fluorophore.
5. The method of claim 3 in which one of the sets of DNA sequencing
fragments is synthesized with equal mole fraction ratios of primers
labeled with the first and second fluorophores, a second one of the sets
of DNA sequencing fragments is synthesized with a unity mole fraction of a
primer labeled with the first fluorophore, a third one of the sets of DNA
sequencing fragments is synthesized with a unity mole fraction of a primer
labeled with the second fluorophore, and the fourth one of the sets of DNA
sequencing fragments is unlabeled.
6. The method of claim 4 in which one of the sets of DNA sequencing
fragments is synthesized with equal mole fractions of terminators labeled
with the first and second fluorophores, a second of the sets of DNA
sequencing fragments is synthesized with a unity mole fraction of a
terminator labeled with the first fluorophore, a third of the sets of DNA
sequencing fragments is synthesized with a unity mole fraction of a
terminator labeled with a second fluorophore, and the fourth of the sets
of DNA sequencing fragments is unlabeled.
7. The method of claims 5 or 6 in which the sequencing fragments
terminating in G, A, T and C are identified by determining the ratio of
intensity of the fluorescence at the first and second wavelength.
8. The method of claim 3 in which a first of said sets of DNA sequencing
fragments is synthesized with unity mole fraction of primer labeled with a
first fluorophore, a second of said sets of DNA sequencing fragments is
synthesized with unity mole fraction of primer labeled with a second
fluorophore, a third of said sets of DNA sequencing fragments is
synthesized with equal mole fractions of primers labeled with said first
and second fluorophores and the fourth set of DNA sequencing fragments is
synthesized with unequal mole fractions of primers labeled with said first
and second fluorophores.
9. The method of claim 4 in which a first of said sets of DNA sequencing
fragments is synthesized with unity mole fraction of terminator labeled
with a first fluorophore, a second of said sets of DNA sequencing
fragments is synthesized with unity mole fraction of a terminator labeled
with a second fluorophore, a third of said sets of DNA sequencing
fragments synthesized with equal mole fractions of terminators labeled
with said first and second fluorophores and the fourth set of DNA
sequencing fragments is synthesized with unequal mole fractions of
terminators labeled with said first and second fluorophores.
10. The method of claims 8 or 9 in which the sequencing fragments
terminating in G, A, T and C are identified by determining the ratio of
intensity of the fluorescence at the first and second wavelength.
11. The method as in claims 8 or 9 in which each of the sets of DNA
fragments is coded with a different mole fraction ratio of the
fluorophores that emit at first and second wavelengths.
12. A method of determining the sequence of DNA which comprises the steps
of
processing sets of DNA sequencing fragments such that different selected
mole fractions of each one of selected sets of sequencing fragments is
tagged with fluorophores which emit fluorescence at different wavelengths
with one of said sets of sequencing fragments is comprised of a mixture of
selected mole fractions of fragments where one mole fraction is labeled
with a fluorophore which fluoresces at a first wavelength and the other
mole fraction is labeled with a fluorophore which fluoresces at a second
wavelength,
separating said sets of tagged DNA sequencing fragments in a single channel
or lane, and
determining the sequence of each of said sequencing fragments by detecting
the ratio of fluorescence intensity at different wavelengths as a function
of time or position.
13. The method of claim 12 where the separation is performed
electrophoretically in a capillary, gel-filled capillary or slab gel.
14. The method of claim 12 in which the tagging is performed by
synthesizing the DNA sequencing fragments using labeled DNA primers.
15. The method of claim 12 in which the tagging is performed by terminating
the synthesis of the DNA sequencing fragments using dye-labeled
terminators.
16. The method of claims 14 or 15 in which the fluorophores are selected to
emit fluorescence at two wavelengths.
17. The method of claims 14 or 15 in which the fluorophores are selected to
emit fluorescence at three wavelengths.
18. A method of determining the sequence of DNA which comprises the steps
of
processing sets of DNA sequencing fragments terminating in G, A, T and C,
where each one of said selected sets of sequencing fragments is labeled
with a selected ratio of two or more radio-isotopes having the same
mobility shift and one set of sequencing fragments is comprised of a
mixture of selected mole fractions of fragments, one of which is labeled
with one radio-isotope and the other of which is labeled with a second
radio-isotope,
separating said sets of tagged DNA sequencing fragments in a single channel
or lane, and
determining the sequence of said sequencing fragments by detecting the
ratio of isotopes as a function of position or time.
19. The method of claim 18 where the separation is performed
electrophoretically in a capillary, gel-filled capillary, or slab gel.
20. The method of claim 18 where the isotopes are radioactive and they are
detected based on the different energy of their radioactive emission.
21. The method of claim 18 where the isotopes are stable and they are
detected by using a mass spectrometer.
22. A method of determining the sequence of DNA which comprises the steps
of
processing sets of DNA sequencing fragments where one set contains
fragments terminating in G, a second set contains fragments terminating in
A, a third set contains fragments terminating in T, and a fourth set
contains fragments terminating in C,
tagging selected mole fractions of each one of selected sets of sequencing
fragments with a first tag which provides a first signal,
tagging different selected mole fractions of each one of selected sets of
sequencing fragments with a second tag which provides a second signal,
tagging two different mole fractions of one set of sequencing fragment with
two different tags which simultaneously provide first and second signals,
separating said sets of tagged DNA sequencing fragments in a single channel
or lane, and
determining the sequence of said sequencing fragments by detecting the
ratio of intensity of the first and second signals as a function of time
or position. |
|
 |
|
 |
Claims |
|
|
|
Description |
|
|
BRIEF DESCRIPTION OF THE INVENTION
This invention relates generally to a multiple tag labeling method for DNA
sequencing and more specifically to a capillary array electrophoresis
apparatus that uses a two-channel fluorescence detection system employing
multiple dye labeling.
BACKGROUND OF THE INVENTION
Current automated DNA sequencing methods use fluorescence detection of
labeled DNA sequencing fragments. One method is to form four sets of DNA
sequencing fragments terminating in G, A, T or C where each set is labeled
with the same fluorophore, and then mn the sequencing fragment sets in
adjacent lanes in a slab gel electrophoresis apparatus. Various apparatus
have been suggested for scanning the gel to monitor said fragments as or
after they move through the gel. In copending application Ser. No.
07/531,900 filed Jun. 1, 1990, U.S. Pat. No. 5,091,652, incorporated
herein by reference, there is described a laser-excited confocal
fluorescence gel scanner which provides enhanced detection of
fluorescently labeled DNA sequencing fragments separated on a slab gel.
The detection system uses an epi-illumination format where the laser power
is focused on the sample by a microscope objective followed by confocal
detection. However, lane-to-lane variations in the migration velocity of
the DNA fragments make it difficult to deduce the correct alignment of the
bands in the four sequencing lanes. The throughput is reduced because of
the need for running four lanes to detect the four sets of DNA sequencing
fragments terminating in G, A, T and C.
A solution proposed to overcome these drawbacks is to label each sequencing
fragment set with a different fluorophore, and then to perform the
electrophoresis operation in only one lane. This requires a multi-color
detection system and dyes that do not alter the mobility of the fragments
relative to one another. A method and apparatus for sequentially scanning
four colors in multiple lanes in a slab gel is described in U.S. Pat. No.
4,811,218 and by Smith, et al., Nature 321,674 (1986). An alternative
method using four different dye labeled dideoxy terminators along with
two-color detection has been described in U.S. Pat. No. 4,833,332 and by
Probet et al. Science 238, 336 (1987).
Capillary electrophoresis is emerging as a high-speed DNA sequencing
method. In copending application Ser. No. 07/840,501 filed Feb. 24, 1992,
U.S. Pat. No. 5,274,240, there is described an automated sequencing
apparatus which employs an epi-illumination format where a laser is
focused to a small volume by a microscope objective and fluorescence
emitted from said volume is gathered by the same objective followed by
confocal detection. An array of side-by-side parallel capillaries is
sequentially and periodically moved past the focal volume or vice versa to
cause and detect fluorescence in labeled DNA sequencing fragments within
the capillaries. The capillary array electrophoresis scanner is described
in said application for use in a one-color, single-channel detection
system where each set of DNA sequencing fragments is separated in a
separate capillary or in a four-color, four-channel detection system where
each set of fragments is labeled with a different fluorophore for
separation and detection in only one capillary.
It is very difficult, in practice, to find four dyes of exactly the same
electrophoretic shift. Therefore, it becomes necessary to perform
complicated shift corrections before the sequence can be read. Four-color
detection has been described in connection with capillary electrophoresis
by Smith and coworkers using a simultaneous four-color detection system
where the signal is split between each of four channels (Nucleic Acids
Research 18, 4417-4421 (1990)). This is satisfactory, but the
signal-to-noise ratio is reduced because the signal is split between four
different channels, and the problem of maintaining equal band shifts for
each of the sets of labeled sequencing fragments using different dyes must
still be resolved.
OBJECTS AND SUMMARY OF THE INVENTION
It is an object of this invention to provide an improved DNA sequencing
apparatus and method.
It is another object of this invention to provide a DNA sequencing method
in which each set of sequencing fragments is labeled with two different
fluorescent dyes.
It is yet another object of this invention to provide a DNA sequencing
method in which the sets of sequencing fragments are labeled with
fluorescent dyes having substantially the same mobility shift.
It is another object of the invention to provide a sequencing method in
which the sequencing fragments are labeled with different fluorescent dyes
and the ratio of the fluorescent signals in two different detection
wavelength regions is employed to detect the fragments.
The foregoing and other objects of the invention are achieved by a
multi-color electrophoresis scanning apparatus which employs labeling
selected DNA sequencing fragments with different mole fractions of
fluorophores, electrophoresing the labeled sequencing fragments in a
single lane to cause separation by sizes and determining the position of
said fragments in the overall sequence by detecting the ratio of intensity
of fluoresence at two detection wavelengths from said labeled fragments.
BRIEF DESCRIPTION OF THE DRAWINGS
The foregoing and other objects of this invention will be more clearly
understood from the following description when read in conjunction with
the accompanying drawings, in which:
FIG. 1 shows a two-color, capillary array electrophoresis scanner in
accordance with the invention;
FIG. 2 shows a two-color DNA fragment fluorophore coding method;
FIG. 3 is an example of the output using the labeled coding method of FIG.
2 in the apparatus of FIG. 1;
FIG. 4 is a plot of the ratio of fluorescence intensity using the labeled
coding method of FIG. 2 and in the apparatus of FIG. 1; and
FIG. 5 shows another two-color DNA fragment fluorophore coding method.
DETAILED DESCRIPTION OF DRAWINGS
A suitable, two-color capillary array electrophoresis scanner for detecting
DNA sequencing is shown in FIG. 1. It consists of two confocal detection
channels that are coupled into the optical system using a dichroic
beamsplitter. One channel detects the yellow-green emission from labeled
fragments, and the other detects the red emission from labeled fragments.
The DNA fragments are labeled with only two dyes which have been selected
so that they have the same mobility shift during electrophoresis. In one
example, the fluorescently labeled primers used in the production of the
Sanger sequencing fragments were supplied by Applied Biosystems Inc.
(Foster City, Calif.). The 5'-end of these primers are covalently attached
to one of the two different fluorophores (FAM or JOE) which have similar
absorption spectra and different emission spectra. The exact structure of
these dyes is proprietary. However, the dyes fluorescein and NBD (Smith,
et al., Nature, 321,674, 1986) have optical properties that are similar to
those of FAM and JOE, respectively. Fragments labeled with JOE and FAM had
substantially the same mobility shift in the capillary electrophoresis
separation media.
The two-color capillary array electrophoresis scanner is shown in detail in
FIG. 1. Light (488 nm; 6 mW) from an argon ion laser (Spectra-Physics
2020, Mountain View, Calif.), not shown, is reflected off a long-pass
dichroic beamsplitter 11 (96% reflection for s-polarization at 488 nm,
Omega Optical 480 DM, Brattleboro, Vt.) and directed through a
32.times.0.40 N.A. infinite conjugate objective 12 (Carl Zeiss LD
Plan-Achromat 440850, Thornwood, N.Y.). The input beam diameter, 5 mm, is
selected to give a 9 .mu.m diameter spot in a given capillary. The
micrometer adjustment 13 of the objective z-position is used to center the
focused beam in the capillary. The fluorescence emission induced in the
fluorophore by the laser beam is collected by the objective and passed
back through the first beamsplitter (.about.92% transmission) to a second
beamsplitter 14 (Omega Optical 565 DRLP) which separates the JOE and FAM
emissions (fluorescence emission peaks at 557 nm and 530 nm,
respectively). The two resulting beams are separately focused with 100 mm
focal length achromat lens 17, 18 (Melles-Griot, Irvine, Calif.) through
400 .mu.m diameter spatial filters 21, 22 (Melles-Griot) to effect
confocal detection of the fluorescence emission. A bandpass discrimination
filter with a transmission window of 525.+-.5 nm (Omega Optical, 525ODF10)
and 488 nm rejection band filter (Omega Optical, 488 RB) shown at 23, are
placed in front of the photomultiplier 26 (RCA 31034A) dedicated to FAM
detection while a bandpass discrimination filter, shown at 24, with a
transmission window of 590.+-.17 nm (Omega Optical, 590DF35) was placed in
front of the photomultiplier 27 (RCA 31034A) dedicated to JOE detection.
The outputs of the cooled phototubes 26, 27 are terminated (1 M.OMEGA.),
amplified and filtered (bandwidth .about.DC to 300 Hz) with a low-pass
filter-amplifier and digitized with a 12-bit ADC (Metra Byte DASH16-F,
Taunton, Mass.) in an IBM PS2 computer. A computer-controlled dc servo
motor-driven translation stage 31 (DCI4000, Franklin, Mass.) with a 6"
travel and 2-5 .mu.m resolution is used to translate the capillary array
past the laser beam. Scanning of the capillary array is accomplished with
periodic sweeps (1.4 s) of the array while sampling data at 1,500 samples
per second per channel. With a line-scan rate of 2 cm/s, the physical
dimension of the pixels acquired represent 13.3 .mu.m. The computer is
used to control the translation stage and to acquire and display images in
a split screen format for the output of each detector. The fluorescence
images are displayed in real time in pseudo-color and stored for
processing.
Selected mole fractions of each of said sequencing fragment sets are
synthesized using a primer labeled with a first fluorophore and different
selected mole fractions of each of said fragment sets are synthesized
using a primer labeled with a second fluorophore and the combined mole
fractions are electrophoresed in a single capillary or lane and the
fluorescence intensity is detected as a function of time as the DNA
fragments move down the capillary.
One method for coding the sequencing fragments is binary coding. This is
shown schematically in FIG. 2. The fragments terminating in A are
synthesized using a 50/50 mixture of primers, half labeled with the red
emitting (JOE) and half labeled with the green emitting (FAM) dye, and
thus, carry the code (1,1). The G-fragments are synthesized using a primer
that is just labeled with the red dye and carry the code (0,1), the
T-fragments are synthesized with just the primer labeled with the green
dye and carry the code (1,0) and the C-fragments are not labeled at all,
carrying the code (0,0).
Binary coded DNA sequencing fragments were prepared through the following
procedure: M13mp18 DNA sequencing fragments were produced using a
Sequenase 2.0 kit (U.S. Biochemical Corp., Cleveland, Ohio). Commercially
available FAM and JOE-tagged primers (400 nM, Applied Byosystems, Foster
City, Calif.) were employed in the primer-template) annealing step. Three
annealing solutions were prepared:
1. 4 .mu.l of reaction buffer, 13 .mu.l of M13mp18 single-stranded DNA, and
3 .mu.l of FAM;
2. 6 .mu.l of reaction buffer, 20 .mu.l of M13mp18 DNA, 1.5 .mu.l of FAM,
and 3 .mu.l of JOE;
3. 6 .mu.l of reaction buffer, 20 .mu.l of M13mp18 DNA, and 4.5 .mu.l of
JOE.
The tubes were heated to 65.degree. C. for 3 minutes and then allowed to
cool to room temperature for 30 minutes. When the temperature of the
annealing reaction mixtures had dropped below 30.degree. C., 2 .mu.l of
0.1 M DTT solution, 4 .mu.l of reaction buffer, and 10 .mu.l of ddT
termination mixture were added in tube 1; 3 .mu.l of DTT solution, 6 .mu.l
of reaction buffer and 15 .mu.l of ddA termination mixture were added in
tube 2; and 3 .mu.l of DTT, 6 .mu.l of reaction buffer and 15 .mu.l of ddG
termination mixture were added in tube 3. Diluted Sequenase 2.0 (4 .mu.l)
was added in tube 1, and 6 .mu.l of diluted Sequenase were added in tubes
2 and 3. The mixtures were incubated at 37.degree. C. for 5 minutes.
Ethanol precipitation was used to terminate the reaction and recover the
DNA sequencing sample followed by resuspension and pooling in 6 .mu.l of
80% (v/v) formamide. The sample was heated at 90.degree. C. for 3 minutes
and then placed on ice until sample injection.
An example of the sequencing of DNA fragments employing this coding in the
apparatus of FIG. 1 is shown in FIG. 3 where the outputs for the
wavelengths in the above examples are overlapped. The sequence can be
easily read off by examining the ratio of the green to the red signal
intensity to determine the fragments G, A, T or C. When the red,
represented by dotted curve, is largest, the fragment terminates in G;
when the green, represented by the solid curve, is much bigger than the
red, it is a T; when the red and green are the same, it is an A; and when
there is a gap, it is a C.
The advantages of this labeling or coding method are: (1) The instrument
design is simplified. Since there are only two optical detection channels,
the optical efficiency is increased, giving a better signal-to-noise
ratio. (2) With just two carefully selected dyes, there is no mobility
shift of one set of base fragments relative to the other. This is clearly
seen in FIG. 3, where the precise registration of the peaks in the red and
green channels shows that the fragment migrations are essentially
identical. (3) In the foregoing example, only two dye-labeled primers are
needed. Thus, the number of labeled primers that must be synthesized is
reduced. (4) Since the ratio of the signal in the green and red channels
is used to identify the base, the base calling is not sensitive to changes
in the optical alignment, laser intensity, or to the amount of the DNA
fragments that migrate in a particular band. The latter point is very
important since the termination reaction has different efficiency
depending on where the termination occurs in the sequence. FIG. 4 is a
plot of the fluorescence intensity in the green channel divided by that in
the red channel for approximately 300 bases in an M13mp18 DNA sequencing
run. The T fragments were labeled solely with FAM, the G fragments were
labeled solely with JOE, and the A fragments were labeled with FAM and
JOE. The ratio was calculated based on peak maxima. The diamonds represent
labeled T fragments, the triangles represent labeled G fragments and the
dots represent labeled A fragments. It is seen that the ratio provides an
excellent determination of the identity of the fragments.
Of course it is to be realized that the labeling of the sets of DNA
sequencing fragments can also be performed by labeling the dideoxy
nucleotide terminator used in the sequencing reactions with a fluorescent
label (Prober et al., Science 238, 336 (1987)) as opposed to labeling the
primer. In this case, selected mole fractions of each of said sequencing
fragment sets are synthesized using a terminator labeled with a first
fluorophore and different selected mole fractions of each of said fragment
sets are synthesized using a terminator labeled with a second fluorophore
and the combined mole fractions are electrophoresed in a single lane or
capillary, and the fluorescence intensity is detected at the
characteristic emission wavelengths of the first and second fluorophores.
The ratio of the intensities at the two wavelengths then determines the
identity and sequence of the DNA. It is obvious that the different coding
methods developed with labeled primers can also be implemented using
dye-labeled terminators.
Of course, a variety of coding algorithms can be used along with this ratio
detection. For example, some workers might object to the binary coding
since the C-fragments are not explicitly detected. This can be resolved in
several ways.
1. A second sequencing run can be done on the same DNA strand where the
binary coding is simply permuted. Then A would be (0,0); G would be (0,1);
T would be (1,0) and C would be (1,1). Since we can run a very large
number of lanes on the capillaries, determining the sequence twice is not
a problem.
2. One could sequence the complementary strand using the binary coding
algorithm in FIG. 2. The presence of a C on the original strand would now
be detected as a G on the complementary strand using the (0,1) coding.
3. Finally, one could use a modified labeling algorithm where all the
fragments are labeled with a dye, but the relative amounts of the two dyes
are adjusted to give four distinctive ratios for the green to the red
channel. As depicted in FIG. 5, this coding would be specified by, for
example, A (1,0); G (1,1); T (1,2); and C (0,1). In this case, the A
fragments would only be labeled with the green dye; half of the G
fragments would be labeled with the green and half with red; 1/3 of the T
fragments would be labeled with the green dye and 2/3 with the red dye;
finally, all the C fragments would be labeled with the red dye.
Since the ratio of the signals is used, three dyes that have the same
mobility shift effect as one another can be used and various mixtures of
three dyes can be used to label the primers to produce four sets of DNA
sequencing fragments. The two-channel detection method with a ratio read
out would still be used but by using three dyes in mixtures. The important
point here is the concept of using ratio detection to code for all four
base fragments on one capillary using only two detector channels.
Finally, it is clear that this method is not limited to fluorescence. For
example, different ratios of two or more different isotopic labels could
be employed. That is, the DNA sequencing fragments terminating in G, A, T
and C can be coded by labeling the fragments with different ratios of
isotopes. The labeled fragments could then be detected through
measurements of radioactive emissions at two different energies, if the
isotopes were radioactive, or by using mass spectrometer detection of
ratios of stable isotopes.
* * * * *
|
|
|
|
|
|
|
Description |
|
