U.S. patent application number 10/105935 was filed with the patent office on 2002-10-24 for method and device for rapid color detection.
Invention is credited to Kao, Pin.
Application Number | 20020155485 10/105935 |
Document ID | / |
Family ID | 26821460 |
Filed Date | 2002-10-24 |
United States Patent
Application |
20020155485 |
Kind Code |
A1 |
Kao, Pin |
October 24, 2002 |
Method and device for rapid color detection
Abstract
Methods and apparatus are provided for miniaturized DNA
sequencing using four-color detection. The method employs a
multichannel chip where the channels are simultaneously irradiated
to excite fluorophores in the channels. The emitted light is
divided into four different wavelength beams and read by CCDs. The
results are multiplexed for rapid sequencing of a large number of
samples, with high fidelity of the sequences.
Inventors: |
Kao, Pin; (Fremont,
CA) |
Correspondence
Address: |
MORRISON & FOERSTER LLP
755 PAGE MILL RD
PALO ALTO
CA
94304-1018
US
|
Family ID: |
26821460 |
Appl. No.: |
10/105935 |
Filed: |
March 20, 2002 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10105935 |
Mar 20, 2002 |
|
|
|
09520423 |
Mar 8, 2000 |
|
|
|
60123349 |
Mar 8, 1999 |
|
|
|
Current U.S.
Class: |
435/6.11 ;
382/128 |
Current CPC
Class: |
C12Q 1/6869 20130101;
Y10T 436/143333 20150115; G01N 27/44721 20130101; G01N 27/44791
20130101; G01N 27/44782 20130101; C12Q 2563/107 20130101; C12Q
2565/619 20130101; C12Q 1/6869 20130101 |
Class at
Publication: |
435/6 ;
382/128 |
International
Class: |
C12Q 001/68; G06K
009/00 |
Claims
What is claimed is:
1. In a method for rapidly detecting light from a channel of
capillary dimensions, where a plurality of individual components of
a mixture move along said channel at spaced apart intervals for an
exposure time of less than about 500 msec, where the light appears
as a band, and is scanned by an optical fiber, the improvement
which comprises: detecting the light from each of said bands with a
CCD camera, where the light illuminates a spot which includes from
2 to 8 pixels in a row and binning the signal received along the
rows of pixels to provide a total signal for said band; and
transmitting the signal to a data processor for detecting said
band.
2. A method according to claim 1, wherein the rate of scanning is
at least about 10 Hz.
3. A method according to claim 1, wherein said CCD camera has a
quantum efficiency of at least 85%.
4. A method according to claim 1, wherein said individual
components comprise four different fluorophores emitting at four
different wavelengths, said method further comprising: dividing the
light from said bands into four different wavelength ranges; and
separately detecting each wavelength range with a separate CCD
camera.
5. A method for performing DNA sequencing detecting four different
fluorophores emitting at four different wavelengths, said method
comprising: moving electrophoretically a mixture of
fluorophore-labeled oligonucleotides at a rate to provide an
exposure time of less than about 500 msec to detection and at a
spacing between oligonucleotides in the range of about 50 to 200
.mu.m; irradiating said fluorophore-labeled at a detection site
with excitation light to produce bands of emitted light; dividing
the light from said bands into four different wavelength ranges;
and detecting the light from each of said bands with a CCD camera,
where the light illuminates a spot which includes from 2 to 8
pixels in a row and binning the signal received along the rows of
pixels to provide a total signal for said band.
6. A method according to claim 5, wherein said moving
electrophoretically is in a channel of from about 50 to
20,000.mu..sup.2 cross-section under a voltage in the range of
about 500 to 4000V.
7. A method according to claim 6, wherein said channel is in a
poly(methyl methacrylate substrate and said channel contains linear
polyacrylamide as a sieving agent.
7. A method according to claim 6, wherein said fluorophore labels
are dR110, dR6G, dROX and dTAMRA.
8. A method according to claim 6, wherein said DNA is at least 500
nucleotides.
9. A method according to claim 6, wherein said binning is of 8 to
36 pixels.
10. A method for performing DNA sequencing detecting four different
fluorophores emitting at four different wavelengths, said method
comprising: moving electrophoretically, in a channel having a
cross-section in the range of from about 50 to 20,000.mu..sup.2, a
mixture of fluorophore-labeled oligonucleotides at a rate to
provide an exposure time of less than about 500 msec to detection
and at a spacing between oligonucleotides in the range of about 50
to 200 .mu.m; irradiating said fluorophore-labeled at a detection
site with an argon laser to produce bands of emitted light;
dividing the light from said bands into four different wavelength
ranges; and detecting the light from each of said bands with a CCD
camera, where the light illuminates a spot which includes from 2 to
8 pixels in a row and binning the signal received along the rows of
pixels of from 8 to 36 pixels to provide a total signal for said
band.
11. A method of irradiating a plurality of channels in a solid
substrate substantially transparent to a wavelength range of
interest, using said solid substrate where a plurality of channels
are bordered by reflective channels as a result of the difference
in refractive index between the solid substrate and the material in
the reflective channel, one reflective channel having a first wall
which at the angle of the incident light transmits the light
orthogonally through said channels, and the other reflective
channel having a second wall which reflects the received light and
transmits the light to a light dump, said method comprising:
irradiating said first wall with light of a wavelength range of
interest at an angle which results in reflection of a beam
orthogonal to said channels, whereby said light is transmitted
through said channels.
12. A method according to claim 11, wherein the number of channels
is in the range of 1-5.
13. A method according to claim 11, wherein said irradiating is
with an argon light source.
14. A method according to claim 11, wherein said solid substrate is
glass or plastic.
15. A method according to claim 14, wherein said plastic is
poly(methyl methacrylate.)
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a continuation of U.S. patent
application Ser. No. 09/520,423 filed on Mar. 8, 2000 which claims
priority to U.S. Provisional Patent Application No. 60/123,349
filed Mar. 8, 1999.
BACKGROUND ART
[0002] Microfluidic devices offer numerous opportunities for rapid
manipulations of small volumes. With the ability to move small
volumes rapidly from one site to another and perform reactions in
very small volumes, where the rate of reaction is greatly enhanced,
the opportunities exist to greatly accelerate various operations.
However, where the result requires detection of the occurrence of
an event or measurement of different species present in a mixture,
the detection of a signal from the medium may become the
rate-limiting event.
[0003] For example, the need to be able to identify a nucleic acid
sequence is increasingly important. In research, forensics,
identification, as well as for other reasons, one is interested in
determining the sequence of a nucleic acid. As the demand for these
determinations increases, there is an increasing need for high
throughput DNA analyses which provide the sequence with high
fidelity. Four-color detection, one color for each nucleotide has
become a mainstay of the sequencing process. For reading the
colors, various excitation and detection systems have been devised.
Each system must accommodate the need to accurately distinguish
between background noise and signal and the different wavelengths
for the different nucleotides. Furthermore, for rapid determination
of the sequence of a large number of nucleic acid samples, one
wishes to have numerous channels, with a different sample in each
channel being sequenced simultaneously. Such systems exacerbate the
problems associated with the various excitation and detection
systems.
[0004] One of the systems used for nucleic acid sequencing is
electrophoresis. The ability to use capillary electrophoresis has
greatly improved the opportunities for high throughput sequencing.
Initially, arrays of capillaries in close juxtaposition were
suggested to be able to do simultaneous determinations. More
recently, the opportunity to have a multiplicity of channels in a
block of plastic or glass has become available. Depending on the
material used for forming the channels, there can be a substantial
amount of autofluorescence, which can serve to obscure the signal.
In addition to autofluorescence, there is light scatter and
cross-talk in the optical system to further limit the sensitivity
of detection of the signal. For high throughput, the time of
collection of the excitation light is limited and the number of
fluorescers in the band being detected is relatively small.
[0005] While CCD (charge-coupled devices) detection has many
desirable features, such as low cost, ease of use, and large areas
of detection, these devices are generally thought to have low
sensitivity. The ability to use these detectors where one is
detecting fluorescer concentrations of 100 pM or lower at high
speed in a multiplex system has been excluded
[0006] There is therefore a substantial interest in developing
sequencing systems using small devices comprising large numbers of
channels so that multiplexed sequence determinations of a large
number of samples may be simultaneously, rapidly and accurately
performed.
[0007] Prior Art
[0008] U.S. Pat. No. 5,741,411 is an encyclopedic description of
the issues associated with capillary electrophoresis and describes
the use of CCD devices for detection of multichannel arrays of long
capillaries for DNA sequencing. U.S. Pat. No. 5,846,708 describes a
multichannel device for detecting molecular tagged targets. Methods
for detecting fluorescent signals may be found in U.S. Pat. Nos.
4,820,048; 5,543,026; 5,776,782; and 5,777,733. U.S. Pat. Nos.
5,162,022 and 5,570,015 describe capillary electrophoresis devices
in which one or more channels are embodied in a single card.
SUMMARY OF THE INVENTION
[0009] Methods and devices are provided for rapid, color detection
as exemplified by four-color detection in the sequencing of nucleic
acids using capillary electrophoresis. The capillary
electrophoresis uses a solid substrate comprising a plurality of
spaced apart channels with electrodes proximal to the end of each
of the channels and a separation medium in each of the channels.
Each of the channels is irradiated with a single wavelength
excitation light and the emission light divided into four paths by
an optical train. The signal in each path is detected with an
independent CCD and the signals from the CCD analyzed by a
computer. The system provides for accurate resolution of the
signals, with greater than 500 nucleotides capable of resolution at
high speed.
BRIEF DESCRIPTION OF THE DRAWINGS
[0010] FIG. 1 is a cross-sectional view of a chip exemplifying
light irradiation of the channels in the chip;
[0011] FIG. 2 is a plan view of a chip having a fan-shaped
arrangement of channels with a portion as an exploded view, and
[0012] FIG. 2A is an exploded view of a portion of the chip having
the ports and
[0013] FIG. 2B is a cross-sectional view along line 2B of the
detection region of the channels;
[0014] FIG. 3 is a diagrammatic view of a system according to this
invention; and
[0015] FIG. 4 is a diagrammatic view of the binning of the signal
by the CCD.
DESCRIPTION OF THE SPECIFIC EMBODIMENTS
[0016] High speed color detection is provided employing an optical
train for receiving light emitted from one or a plurality of
channels, where the fluorescent label is moving along the channel
and is irradiated at a detection site, detecting the light
intensity with a CCD by binning pixels and integrating the light
signal received along the channel by the pixels of the CCD. The
channels are present in a microfluidic device, employing one or a
plurality of channels of capillary dimensions in a solid substrate.
The method finds use for rapid detection of a light signal,
particularly a fluorescent signal, which is moving in a path,
conveniently linear, where the molecules providing the signal are
spread over a finite area, usually having a single peak. The
detection method integrates the light received as the sample to be
measured moves through the detection zone. The method finds
particular application for nucleic acid sequencing, but may be used
in any device where a transient light signal is detected and needs
to be resolved at high speed, with accurate detection.
[0017] Since DNA sequencing as performed today with electrophoresis
requires four color detection for high-speed throughput, DNA
sequencing will be used as paradigmatic of the use of the subject
invention. Since each color which is detected could be a single
distinct component to be measured having the same color, one would
have a simpler optical train, since the light received by the CCD
would not have to be divided into different wavelengths for
detection. Alternatively, as the number of entities to be
determined increases, one may wish to detect more than four colors,
usually not more than about eight colors, conveniently not more
than about six colors.
[0018] The DNA fragments provided for the sequencing are normally
provided by performing either the Sanger dideoxy method (Smith, A.
J. H., Methods in Enzymology 65, 56-580 (1980) or the Maxam-Gilbert
method (Gil, S. F. Aldrichimica Acta 16(3), 59-61 (1983). In each
case, fluorescent dyes are provided, which may be present in the
primer, in the terminating nucleotide, or may be added by reaction
with a functionality which is provided as part of the extension of
the primer. Various dyes and dye combinations (referred to
hereinafter as "dyes" collectively) are employed, which dyes can be
efficiently excited by a single light source, but will emit at
different wavelengths, usually separated by at least about 10 nm,
preferably separated by at least about 15 nm, and generally less
than about 30 nm will be acceptable. Desirably, one can use an
inexpensive laser, such as an argon laser to excite the dyes at 488
nm and/or at 514 nm, where the dyes to be excited have a
significant absorption effiency at that wavelength. Dyes of
interest include xanthene dyes, e.g. fluoresceins and rhodamines,
coumarins, e.g. umbelliferone, benzimide dyes, e.g. Hoechst 33258,
phenanthridine dyes, e.g. Texas red, ethidium dyes, acridine dyes,
cyanine dyes, such as thiazole orange, thiazole blue, Cy 5, Cyfr,
carbazole dyes, phenoxazine dyes, porphyrin dyes, quinoline dyes,
or the like. Thus the dyes may absorb in the ultraviolet, visible
or infrared, preferably the visible wavelength range. Various
combinations of dyes may be used. A common combination is: R110,
R6G, ROX, and TAMRA, which has been exemplified herein, or their
"d" analogs, e.g., dR110, dR6G, dROX and dTAMRA. Other combinations
include: FITC, NBD chloride
(4-chloro-7-nitrobenzo-2-oxa-1-diazole), TMRITC
(tetramethylrhodamine isothiocyanate) and Texas Red; energy
transfer combinations, such as FAM
(fluorescein-5-isothiocyanate)(common donor), JOE
(2',7'-dimethoxy-4',5'-dichloro-6-carboxyfluorescein), TAMRA
(N,N,N',N'tetramethylcarboxyrhodamine), and ROX
(6-carboxy-X-rhodamine) (See, U.S. Pat. No. 5,707,804); or
replacing NBD chloride with NBD fluoride.
[0019] After performing the sequencing process in accordance with
conventional techniques as described above, where all of the
nucleotide extensions occur in a single reaction mixture or are
performed in separate reaction mixtures and then combined for
analysis, the mixture is introduced into the subject system. The
system comprises a microfluidic device and a detector for the
microfluidic device.
[0020] The microfluidic device comprises a solid substrate,
referred to as a chip, having a plurality of channels of capillary
dimensions and ports or reservoirs for receiving the reaction
mixture. The chip may be a solid substrate having a thickness of
from about 0.5 mm to 5 mm or more, although there will usually be
no advantage in having a thickness of more than about 2.5 mm. A
thin cover of about 0.05 to 1 mm is used to enclose the channels in
the solid substrate, which cover is bonded to the solid substrate.
The cover and/or the solid substrate may serve to provide the ports
for access to the channels and chambers present in the device or
the chambers may be extrinsic to the device and in fluid connection
to the capillaries--through flexible or rigid conduits. For the
most part, there will be a sample reservoir and a waste reservoir
at opposite sites in relation to the channel, where the reservoirs
may be part of the device or connected to the channel through an
external conduit. Electrodes are provided for providing an electric
potential across the channel. The width of the device will vary
with the number of channels present and the pattern of the
channels, usually being at least 5 mm, more usually at least one
centimeter and may be 10 cm or more. The spacing of the capillaries
will usually be at least about 0.1 mm and not more than about 2 mm,
usually not more than about 1 mm, where the spacing may be
difficult to define for certain patterns for the capillary
distribution. The design for the device may have parallel
capillaries, fan-like capillaries, where the capillaries are spread
apart at the sample entry site and the spacing narrowing at the
detection site, e.g. where the capillaries define an arc of up to
about 180.degree. at the entry site and have a linear cross-section
of about 50 to 200 mm at the detection site.
[0021] Normally, the cross-sectional area of the capillaries will
be in the range of about 50 to 20,000.mu..sup.2, more usually from
about 50 to 10,000.mu..sup.2. The length of the capillary will
generally be at least about 2 cm and may be as long as 50 cm,
generally being from about 2 to 40 cm. The length of the capillary
will depend upon the sample being analysed, the length of the
largest fragments and the number of nucleotides to be sequenced.
Different materials may be used for the solid substrate, such as
plastics, e.g. acrylates, exemplified by poly(methyl methacrylate),
polycarbonate, poly(ethylene terephthalate), etc., silicon, silica,
glass, and the like. Depending on the nature of the substrate, the
surface of the channel may be coated with a polymer composition
and/or a medium may be introduced into the channel. Various media
have found use in capillary electrophoresis, particularly
compositions that form gels or enhance the viscosity in the channel
or provide a charged surface for movement by electroosmotic flow.
These materials include polyacrylamide and N-substituted
derivatives thereof, agarose, poly(ethylene oxide),
hydroxyethylcellulose (HEC), aminoethylcellulose or other
convenient separation gel or charged polymer.
[0022] The number of channels will generally range from about 1 to
2000, more usually from about 50 to 1500, frequently from about 75
to 500, in a single chip and instrumentation can be provided for
reading two or more chips simultaneously. Desirably, the subject
system will be able to read at least 20, more usually at least 50
and preferably at least 75 channels simultaneously with high
resolution and accuracy, generally being limited to about 1000
channels for a single detector. Each of the channels will have a
detection zone, where one or more windows are provided which are
substantially transparent to the excitation and emission light. The
windows may be a continuous window over all of the channels or an
intermittent window, which covers a single channel or a plurality
of channels, generally not greater than about 100. In one
embodiment, the excitation light enters from the side of the
channel and leaves from the other side of the channel, where the
composition of the chip allows for light transmission.
[0023] The sensitivity of the system is to be able to accurately
detect (<5% error rate in 500 nucleotides) the individual dyes,
with an exposure time for detection of less than 1 sec, usually
less than about 0.5 sec, and may be as little as 100 msec or less,
with a frequency of determination of at least 5 Hz and up to about
100 Hz, usually not more than about 50 Hz, frequently in the range
of about 10 to 20 Hz. The concentration of the dye to be detected
will usually be greater than about 5 pM, usually greater than about
10 pM and less than about 50 pM, usually less than about 40 pM. The
speed and sensitivity will, to varying degrees, depend upon the
dye, which is selected
[0024] In performing the determination, the sample containing the
extended primers will be placed in a channel directly or into a
reservoir to which the channel is connected. The volume in a
reservoir or chamber will generally range from about 0.1 .mu.l to
100 .mu.l . The volume of the sample in the channel will usually be
at least about 0.5 nl, more usually at least about 1 nl, and
usually not more than about 10 nl, more usually not more than about
5 nl. Different samples may be put into each channel or the same
sample in a number of-channels, where some redundancy is permitted,
to ensure the accuracy of the sequence. Electrodes are placed at
opposite ends of the channel to provide a voltage across the
channel of from about 500 to 4000V to provide for separation of the
sequences by their different length. Normally, the conditions of
the separation and the length of the channel in which the
separation occurs should allow for a spacing between
oligonucleotides of from about 50 to 200 .mu.m, which is sufficient
for the optical detection employed in this invention.
[0025] The detection system involves an excitation source, optics
to receive the fluorescent light, an optical train to define the
area from which the light from the channel is received and to
divide the light into the four different wavelength ranges for the
different dyes and CCD detectors for determining the intensity of
the light from the channel. These signals are then sent to a signal
processor for decoding to identify the individual nucleotides.
[0026] As already indicated, the excitation source may be any
source of radiation in the excitation wavelength of the dyes,
coherent or non-coherent light source, and for DNA sequence
analysis with the commonly used dyes, is conveniently an argon
laser. Depending on the number of channels, their separation, the
nature of the material of the substrate, and the separation between
channels more than one light source may be used. The light may be
incident to the channels, transverse to the channels, and with the
appropriate configuration can provide for total internal
reflection. Laser intensity would be about 750 to 2000 W/cm.sup.2,
using a 4 mW laser. By using the appropriate lens objectives, the
light can be directed to the desired number of channels and provide
the necessary intensity for excitation of the different dyes.
[0027] The light collection is dependent on minimizing background
associated with light scatter, autofluorescence of the material of
the chip, area of the channel covered to maximize the number of dye
molecules from which light is received and the like. For example,
for a 50 .mu.m deep channel, with a 10 .mu.m diameter beam (in air,
20.times. objective), at a 10 pM concentration, there will be 24
molecules in the detection region.
[0028] The subject system was demonstrated using confocal,
laser-induced epifluorescence detection. Fluorescence from a DNA
sequencing sample was excited by an argon ion laser (Omnichrome,
543-AP) operated at 488 nm. The laser light was directed into an
inverted, epifluorescence microscope (E300, Nikon Instruments,
Melville, N.Y.), passed through an excitation filter (485DF22,
Omega Optical, Brattleboro, Vt.), reflected off a dichroic miror
(500DRLP, Omega Optical) and focused into the channel using a
10.times./0.45 NA microscope objective. The laser formed an
approximately 20 .mu.m spot within the channel which contained the
DNA sample. Fluorescence was collected through the objective,
passed through the dichroic mirror and a long pass filter (515EFLP,
Omega Optical) and focused onto a 500 .mu.m circular confocal
aperture. The light passing through the aperture was collimated
using an achromatic lens and separated into 4 colors and detected
using the optical train and analysis unit 304 described in FIG.
3.
[0029] DNA sequencing was performed in a 50 .mu.m deep by 120 .mu.m
wide channel etched into a 5-cm by 22-cm glass plate. The plate was
bonded to a top glass cover plate to enclose the channels and
create the bonded glass chip. The channel pattern was a cross
injector comprised of an injection channel intersecting a
separation channel. The effective separation length was 18 cm as
measured from the channel intersection to the detection point. Each
arm of the cross injector was connected to a reservoir created by a
drilled hole in the top cover plate. Voltages to create
electrophoretic potentials in the channels were applied using a
platinum wire electrode placed in each reservoir. Reservoir
voltages were controlled by computer using a 4 channel, high
voltage power supply.
[0030] The DNA sequencing sample was prepared by ABI Bigdye-labeled
primer ready reaction chemistry. The template used in the reaction
was M13mp18 ssDNA template standard. The primer set is--21M13
universal primer. 2.times. reaction with 10 .mu.L is carried in an
ABI Thermal Cycler 2400 according to the manufacture
recommendation. The thermal cycling protocol includes two phases:
amplification and chasing cycles. 15 cycles of amplification are
performed with a protocol of 96.degree. C. for 10 s, 55.degree. C.
for 5 s, and 70.degree. C. for 60 s. 15 cycles of chasing are
carried with a protocol of 96.degree. C. for 10 s and 70.degree. C.
for 60 s. Two holding cycles are followed as 96.degree. C. for 7
min and 4.degree. C. forever. After reaction, the reaction
mixture-is passed through a Centra-Sep spin column to remove the
excessive salt components. 4 reactions with ddATP, ddTTP, ddGTP and
ddCTP terminator and corresponding bigdye-labeled primers are
preformed and pooled together. 10 .mu.L of pooled reaction mixture
are mixed with 10 .mu.L deionized formamide and then heated up to
90.degree. C. for 2 min. After chilling on the ice bath, the
mixture is loaded onto the chips.
[0031] The sieving media is 2% LDA010 polymer solution (linear
polyacrylamide) in 1.times. TTE buffer (50 mM Tris, 50 mM TAPS, and
2 mM EDTA with 7 M urea). The polymer solution is loaded through
reservoir #3 under a pressure of 200 psi for 14 min. The separation
is performed under 200 V/cm and 35.degree. C. The temperature of
the entire glass chip is controlled by a Peltier temperature
controller.
[0032] The fluorescence from each channel is received by an optical
fiber and transmitted to an optical train, which serves to divide
the light into 4 pathways, each with a different range of
wavelengths. The light may be divided by prisms, dichroic mirrors
or gratings. Prism structures for wavelength separations are
available from Richter Enterprises, Wayland, Mass., with a
wavelength range of 430-950 nm, >80% transmission and a clear
aperture of 30.times.8 mm. Dichroic mirrors are available from a
variety of sources, e.g. Omega Optical. Filters or gratings are
provided in each pathway to provide the desired wavelength signal
to the CCD. While either filters or gratings may be used, for the
purposes of the subject invention, filters are preferred, and may
include bandpass and cutoff filters.
[0033] The light is passed through the filter and received by the
CCD. Various CCDs are available from a number of sources, but the
low noise, high QE CCD available from Hamamatsu has been found to
be useful. Characteristics of the CCD for the subject invention are
the CCD is back thinned, has 85-90% quantum efficiency, has a fast
readout and can be binned along the linear direction.
[0034] A single CCD is used for each wavelength range of light to
be measured. The light is focused on an area of pixels, where the
signal from at least 5, usually at least 8 and not more than about
625, usually not more than about 400, preferably from about 8 to
225 pixels, more usually 8 to 144 pixels, is integrated in the area
for a single channel. Generally, there will be from 2 to 25,
usually 3 to 20 pixels, more usually about 9 to 100 pixels in a row
and approximately the same number in a column. The signal in each
row of pixels is dumped into the next row, where the number of rows
in the device employed is 64. Larger or smaller numbers of rows may
be used, although the minimum number of rows in a column will be at
least 2 and will usually be not more than about 10. The signals
from the pixels will be integrated and passed to the register,
which will send the result to a computer for analysis. The
detection bandwidth will generally be less than 10 Hz and more than
about 0.5 Hz, usually about 1 to 5 Hz, preferably about 2 Hz.
Sampling will be at a rate of at least about 5 Hz, usually at least
about 10 Hz and may be as high as 50 Hz or greater. The total peak
to base will usually be about 1 to 2 sec during the detection
period. Sensitivity for the system was in the range of 3-7 pM.
[0035] Using the procedure described above, where the spot covered
about 40-80 pixels, almost 6 to 9 rows and columns, the following
results were obtained. The DNA sequencing results from the 4-CCD
DNA sequencer can be described in terms of separation, speed and
base calling accuracy. The resolution of two neighboring bases in
an electropherogram is defined as the ratio of the mutual distance
to their mean width. Most base-calling software can easily identify
the peaks with resolution better than 0.5. For the DNA sequencing
conducted in our 2% LDA010 polymer solution, the range of DNA
fragments with separation resolution larger than 0.5 is from 50 bp
to 760 bp. This ensures the 4-CCD detection system for on-chip DNA
sequencing gives more than 700 bp. The fast separation requires
fast data acquisition. In this system, the first peak in the
electropherogram, which is the primer peak, comes out at 7.2 min;
the last peak, which is the bias reptation peak, elutes out at
.about.26 min. All the useful information is generated from 8 min
to 20 min. On the average, it is about 1.2-1.5 second per peak
eluted. 10 Hz data acquisition rate of the 4-CCD detection system
gives about 12 to 15 points to map out a peak, with the spot on
each CCD being about 25 pixels. This is very efficient and
effective to determine a Gaussian peak although it is at the low
end. Higher data acquisition is desirable but not necessary to
yield a better sequence result. BaseFinder software was used to
call the sequence of M13mp18 sample from the raw data generated
from CCD. Without any sophisticated mobility shift calibration, the
base-calling accuracy for 700 bp is 95.7%. If one manually
eliminates the errors caused by the mobility shift difference of
the fragment sets, the sequencing accuracy is 97.8%. This
demonstrates that the quality of the raw data produced from 4-CCD
can be used in DNA sequencing.
[0036] For further understanding of the invention the drawings will
now be considered. FIG. 1 is a diagrammatic cross-sectional view of
a chip. The chip 100 has a plurality of channels 102, which are
depicted as rectangular. In this design, the channels will be
rectangular, although they may be elliptical or round, and
conveniently for use in the subject invention will have a depth of
about 25 to 100 .mu.m and a width of about 10 to 50 .mu.m. The
channels will usually have separations greater than the width of
the channel and generally in the range of about 100 to 1000 .mu.m.
A laser source 106 is directed toward an angled wall 108 of air
filled channel 110 as beam 112. The beam 112 is reflected from the
wall 108 by total internal reflection and is directed transversely
as transverse beam 114 through the light transferring channels 102
and the separations 104 to the angled wall 116 of a second air
filled channel 118 as beam 120, where it is sent to a waste light
receiver, not shown. While conveniently the angle shown is
45.degree., which provides for orthogonal entry of the beam 112
into the channels 110, other angles could be used, where the angle
of the light beam would be modified to provide the orthogonal beam.
Materials, such as acrylic, can be prepared with comparatively low
autofluorescence and a high efficiency of transmission, so that the
chip material may be used for transmitting the beam. Other
materials which may find use include glass, organic polymers,
silicone, etc. Usually, the beam will be effective for from about 1
to 5 channels, more usually from about 1 to 3 channels.
[0037] An exemplary chip is shown in FIG. 2, with exploded portions
of the chip in FIG. 2A. The chip is intended to be introduced into
an instrument, which would provide the sample, reagents and voltage
supply for the channels. Thus the chip 200 would be inserted into a
receiving opening in an instrument, where the chip would be indexed
to be positioned in relation to the sample and reagent supply for
each channel, as appropriate, the electrodes for each channel and
the excitation light source and detector. Separate reservoirs may
be provided which have the electrode for providing the voltage
potential for the channels, where the reservoir is electrically
connected to the channel, but may or may not have free fluidic
connection with the channel. Once introduced into the instrument,
the process for separation and analysis would begin
automatically.
[0038] Each of the channels 202 has a separate port 204 for
receiving sample and any reagents, including buffer, and as
indicated, may also receive an electrode. The channels 202 are
depicted as parallel, although they could be configured in a
fan-like pattern and converge at the downstream side of the channel
in relation to the movement of the oligonucleotides. All of the
channels pass the detection window 206. Light from a laser 208
passes transversely through the channels 202 in the detection zone
and fluorescent molecules are excited to emit light, which is then
read by a detector.
[0039] In FIG. 3 an exemplary system is diagrammatically shown. The
system 300 has excitation and emission subunit 302 and optical
train and analysis unit 304. The excitation and emission subunit
302 comprises a laser source for the excitation light. While
collimated light, particularly laser light is preferred, it is not
required and any light source which provides for efficient
excitation of the fluorophores may be employed. While in the figure
the laser 306 is shown as providing a beam 308, which is reflected
by dichroic mirror 310, any arrangement may be employed which
allows for efficient transmission of the light and minimizes the
amount of light received in the optical train unit 304. The light
beam 308 is focused by lens 312 and is then distributed by
excitation optical fibers 314 to the channels of the chip 316. The
chip is indexed, so as to be properly positioned in relation to the
excitation optical fibers 314 and the emission optical fibers 318.
Thus, each channel of the chip has a pair of optical fibers for
delivering excitation light to the channel and receiving emission
light from the channel.
[0040] As discussed previously, the excitation light can be
provided as a beam orthogonal to the channels, where less of the
excitation light will enter the emission optical fibers 318. The
light from emission optical fibers 318 is filtered by filter 320 in
order to cut off any stray light from the excitation light source,
e.g. argon laser, where the cutoff is >515 nm. Lens 322 focuses
and collimates the light from the optical fibers 318, to have the
light from the individual channels in substantially parallel beams
324 for entry into the optical train 304. The optical train is
shown for four color detection, but as already indicated, only a
single color might need to be detected. In that case, the filter
322 could suffice to cut off light outside the wavelength range of
interest, for example, a bandpass filter which would only allow
light in the wavelength range of interest. One could then direct
the light to a CCD camera for detection.
[0041] In FIG. 3, the optical train 304 is designed to detect four
different colors, as is commonly used in DNA sequencing. The system
is described for the dyes dR110, dR6G, dTAMRA and dROX. The light
beams exiting from lens 322 impinge on dichroic mirror 326 and for
commonly used dyes, would be divided into two beams, beam 328 and
336, which would be light above 590 nm and beam 330 and 354, which
would be light below 590 nm. Beam 328 would be further divided by
dichroic mirror 332 into beam 334, which would be light below 610
and beam 336, which would be light above 610. Beam 334 would be
further restricted by filter 338 to a wavelength range of 590-610,
focused by lens 340 and detected by CCD camera 342. The signals
from CCD camera 342 would be sent to computer 344 for analysis.
Similarly, beam 336 would be filtered by filter 346 to a wavelength
range of 615-650, focused by lens 348 and detected by CCD camera
350. The signals from CCD camera 350 would be sent to computer 344
for analysis. In analogous fashion beam 330 would be split by
dichroic mirror 352 into beams 354 and 356, passing light greater
than 560 nm and reflecting light less than 560 nm. Filter 358 is a
bandpass filter, which passes light of wavelength range 560-580 nm,
while filter 360 is a bandpass filter which passes light in the
range of 510-550. Lenses 364 and 366 focus beams 354 and 356
respectively onto CCD cameras 368 and 370, respectively. The CCD
cameras 368 and 370 transmit their signals for analysis to computer
344.
[0042] FIG. 4 is a diagrammatic view of what the CCD camera sees.
In FIG. 4, each of the squares symbolizes a single pixel. In FIG.
4A the camera face 500 has a plurality of pixels which are present
in columns and rows. Light from the optical fibers as depicted in
FIG. 3, are focused to create a spot 504 on the camera face 500, so
that light emitted from the channel impinges on the pixels and is
counted by the pixels. As depicted in the figure, the diameter of
the spot 504 covers about 6 pixels. The photons counted by the
pixels 502 in row 506 is transferred to row 508, which is added to
the photons counted in row 508 and this total is transferred to row
510 and added to the photons counted by the pixels in row 510, and
so on. These counts are continuously transferred to successive rows
until the total is received by register 512 and the total value
sent to a data analyzer, e.g. a computer, for processing. In FIG.
4B this process is indicated by the shaded regions 514, which
symbolizes the blurring of the values through the pixels and the
counts are transferred from row to row and finally the total
received by register 512. One may consider the method of binning as
a bucket brigade, where each pixel dumps into the lower register
412 and is cleared. Charge is shifted down the column and the full
charge well is always available. In this case saturation of a pixel
is prevented. By binning of the pixels a more rapid and accurate
detection of peaks is achieved, allowing for high throughput
analysis of four colors for sequencing of DNA in a plurality of
channels.
[0043] It is evident from the above description that the subject
invention provides an efficient rapid way to multiplex assays and
obtain a large amount of data rapidly, with high precision and with
high efficiency. The methodology employs the conventional
methodology for performing DNA sequencing using four color
discrimination for the different nucleotides, while being able to
miniaturize the methodology, so as to use very small volume
samples, high rapid separation times and rapid detection times. In
this way a large number of determinations may be made within a
short period of time, with efficient use of the instrumentation and
laboratory personnel.
[0044] While the foregoing invention has been described in some
detail for purposes of clarity and understanding, it will be clear
to one skilled in the art from a reading of this disclosure that
various changes in form and detail can be made without departing
from the true scope of the invention. All publications and patent
documents cited in this application are incorporated by reference
in their entirety for all purposes to the same extent as if each
individual publication or patent document were so individually
denoted.
* * * * *