 |
Description  |
|
|
The present invention is directed to a method for sequencing DNA molecules.
BACKGROUND OF THE INVENTION
The present invention provides a method for determining the nucleotide
sequence of DNA molecules (referred to herein as the nucleotide base
sequence or simply the base sequence). Several methods are known for
sequencing DNA molecules such as methods of F. Sanger, S. Nicklen, A. R.
Coulson, Proc. Natl. Acad. Sci. U.S.A., 74, 5463 (1977), and A. M. Maxam
and W. Gilbert, Proc. Natl. Acad. Sci. U.S.A., 74, 560 (1977). These known
methods use various means for producing labeled fragments of DNA, each of
which terminates with a known base (A, G, C or T). These fragments are
then separated by length, typically by an electrophoretic gel, utilizing a
different gel strip for each type of terminal base. The DNA sequence is
then read from the gel strips. As a variation, instead of using the same
label for each fragment (such as a fluorescent dye or radioactive label)
J. M. Prober, et al., Science, 238, 336-341, Oct. 1987, and C. Connell et
al., BioTechniques, Vol. 5, No. 4, 342-348 (1987), use a different dye to
label each of the different base termination fragments so there is a
different dye associated with A, G, C and T termination. This modification
allows a single gel to be used, however, it also introduces new problems
due to the effect of the different dyes on fragment mobility.
A limitation of the prior methods is that they are apparently limited by
the rate at which the fragments may be separated and are also limited by
the number of bases that can be sequenced in a given run by the resolution
obtainable on the gel. The separation rate is inherently limited, for
example, by thermal distortion of the gel caused by electrical heating,
and thus the identification can only be obtained as often on average as
about a few bases per minute. Also the resolution on the gel is a maximum
of about 1,000 bases, with improvement in this resolution not being likely
because of band compression effects, and because there are interactions
between the DNA strands which dominate over the length effect of very long
strands, thus confusing the signal for long fragments.
The present invention provides an improvement over these prior art methods.
It is thus an object of the present invention to provide a method of DNA
identification in which the rate limiting step is essentially the rate of
a polymerase reaction, which is usually on the order of at least 60 bases
per second, or limited by the rate in which the reagents can be delivered
to the reaction site, whichever is slower.
It is another object of the present invention to provide a method of DNA
sequencing in which the accuracy does not depend upon the length of the
DNA molecule to be sequenced but, rather on the signal-to-noise ratio of
the detection means, which is very low using optical detection methods.
Such high sensitivity detection means provide the advantage that only very
small quantities of DNA are necessary, typically, less than a million
molecules.
It is yet another object of the present invention to provide a DNA
sequencing which is unambiguous even in short sequences of identical
bases, which are difficult to distinguish by prior art methods.
Another object of the present invention is to provide a novel method for
DNA sequencing in which the reagents for detection comprises a single
mixture of bases, and does not require four separate preparations (one for
each base) as required by methods of the prior art.
These and other objects of the present invention will be apparent from the
following description, the appended claims and from practice of the
invention.
SUMMARY OF THE INVENTION
The present invention provides a method for determining the nucleotide
sequence of a single strand DNA molecule comprising the steps of:
(a) providing a set of identical single strand DNA molecules (ssDNA)
comprising at the 3' end a leader sequence, the leader sequence comprising
a region recognizable by a DNA polymerase for initiation of replication;
(b) providing an oligonucleotide complementary to at least a portion of the
leader sequence, and capable of forming a stable double stranded DNA
hybrid therewith;
(c) covalently attaching the 3' end of the leader sequence, the 5' end of
the ssDNA or an end of the oligonucleotide to a solid support;
(d) forming a stable double strand DNA hybrid bound to the solid support,
the hybrid comprising the oligonucleotide and the single stranded DNA
molecule with the leader sequence and the bound hybrid acting as a primer
for DNA polymerase replication;
(e) exposing the hybrid bound to the solid support to a DNA polymerase in
the presence of fluorescently-labeled 3'-blocked derivatives of the four
nucleotide 5'-triphosphates of 2'-deoxyadenosine, 2'-deoxyguanosine,
2'-deoxycytidine and 2'-deoxythymidine, where each of the four nucleotide
5'-triphosphate (NTPs) derivatives is labeled with a fluorescent label
distinguishable by fluorescent detection means from the other three labels
on the other three nucleotide 5'-triphosphate derivatives, under
conditions whereby the polymerase will add the appropriate complementary
nucleotide 5'-triphosphate derivative to the oligonucleotide;
(f) separating any unused NTP derivatives from the solid supported DNA
hybrid and the support;
(g) identifying the labeled NTP derivative added to the double stranded DNA
by optical detection means; thereby identifying its complementary
deoxynucleotide present in the single stranded DNA molecule;
(h) removing the fluorescent label and 3' blocking group from the labeled
NTP derivative of step (g) to expose the normal OH group in the
3'-position;
(i) separating the freed blocking group and label (which may be associated
with the blocking group) from the solid supported double stranded DNA
hybrid;
(j) repeating steps (e) through (i) through a plurality of cycles until
labeled NTPs can no longer be added to the oligonucleotide; whereby the
result of each cycle identifies the next deoxynucleotide in sequence in
the single stranded DNA molecule.
BRIEF DESCRIPTION OF THE DRAWINGS
In the accompanying FIG. 1 there is schematically shown a double stranded
DNA hybrid bound to a solid support utilized in accordance with the
present invention.
In the accompanying FIG. 2 there is illustrated a photolytic removal of a
3'-blocking group from a 3'-blocked nucleotide in accordance with the
present invention.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
The purpose of the present invention is to determine the sequence of a set
of identical single stranded DNA molecules, therefore it will be assumed
that such strands are initially provided. These strands, hereinafter
called ssDNA, are adapted to be used in accordance with the present
invention. Various ways to adapt the ssDNA for attachment with a
complementary oligonucleotide to a solid support will be apparent to those
of ordinary skill in the art. Several methods are described herein as
preferred embodiments. One method is to modify the ssDNA with a known
leader. The purpose of the modification is to attach to the 3' end of the
ssDNA a known leader sequence which (when hybridized for form a duplex) is
recognizable by the polymerase to be utilized for the initiation of
replication unless the provided ssDNA already has a known leader sequence.
The 3'-end of the leader may also provide a handle which may be attached
to a solid support. The ssDNA may be modified at least in any of the
following ways.
One such method is to first amplify the quantity of the ssDNA by polymerase
chain reaction techniques (PCR, reference, R. Saiki et al., Science, Vol.
239, 487-491, January 1988). Prior to amplification, if necessary, a known
single stranded sequence comprising the single stranded sequence of a
sticky end may be added as a short, temporary leader to the ssDNA. Methods
of joining the 5' end of one oligonucleotide to the 3' end of another DNA
molecule are known in the art and can be routinely performed.
Amplification by PCR creates many such DNA molecules with a short leader.
An oligo complementary to the short leader may then be added so that there
is a short section of double stranded DNA at the leader, as shown in FIG.
1.
An alternative approach is to attach a double stranded oligo with a sticky
end to the solid support. This oligo will have the complementary sequence
of the same restriction site used to create the ssDNA. Then the two sticky
ends will be ligated to form a double stranded DNA molecule attached to a
solid support as shown in the FIG. 1.
Alternatively, a single strand leader may be ligated to the end of the
unknown ssDNA strand. The oligo containing a sequence complementary to the
leader (or portion thereof) may be bound through its 5' end to the
substrate. Then the ssDNA and the associated leader will be bound to the
solid support by hybridization to the bound oligonucleotide to result in
the identical situation shown in FIG. 1.
Alternatively instead of sticky ends as described above, blunt end ligation
may be utilized.
Initially, the double stranded portion of the bound molecule in FIG. 1 will
be a primer for a suitable DNA polymerase, preferably Taq polymerase,
which is operable at high temperature.
Another starting material required for the present invention is a mixture
of four fluorescently-labeled (or other optically labeled such as optical
absorption dyes, or chelated ions) 3'-blocked, NTPs (nucleotide
triphosphates). The preferred embodiment has the fluorescent label as part
of the 3'-blocking group. Each of the NTPs will be labeled by a different
label (i.e., each of the A, G, C and T NTPs will have different labels on
them) so as to be distinguishable by fluorescent spectroscopy or by other
optical means. Such labels are known in the art and are disclosed for
example in Prober, et al., Science, vol. 238, pp. 336-341 (1987) and
Connell et al., BioTechniques, Vol. 5, No. 4, 342-384 (1987); Ansorge, et
al., Nucleic Acids Research, vol. 15 (11) 4593-4602 (1987) and by Smith,
et al., Nature:321, 674 (Jun. 12, 1986). Each of the NTPs has a
3'-blocking group, so as to prevent the polymerase from continuing to
replicate once one base has been added. This is preferably accomplished by
having the dye attached through a covalently linking group to the
3'-position so that the dye moiety and the 3'-blocking group are contained
in the same substituent. Examples are shown below.
##STR1##
wherein: X is --O--A-- and
A is a functional group which is removable to expose the 3'-hydroxyl group.
For example, A may be
##STR2##
This may be prepared by treating any of the reactive dyes, such as
reactive fluorescent dyes functionalized with halo groups, with
N-trifluoroacetyl propargylamine under conditions described by Prober et
al., Science, 238: 336-341, Oct. 1987, then deacylating. This amino-dye
may then be coupled to a 3'-O-succinyl protected nucleotide to produce a
3'-O-protected nucleotide wherein the protecting group is
##STR3##
which is removable under conditions similar to that of the O-succinyl
protecting group commonly used in solid phase nucleotide synthesis.
Alternatively, an acid anhydride, to which the dye is attached, may be
directly condensed with the 3'--OH group. Thus to initiate the sequencing,
the bound double strand DNA molecule shown in FIG. 1 is exposed to the DNA
polymerase and a mixture of the four fluorescently-labeled 3'-blocked
NTPs. The polymerase will then add one of the four NTPs to the growing
oligonucleotide chain, whichever NTP is complementary to the next unpaired
base in the ssDNA. This step is rapid since the average reaction rate of
adding a base to an oligonucleotide with a polymerase is in the range of
at least 60 bases per second. Since only one base is being added this can
be accomplished in less than a second.
The next step is to separate the unused NTP's from the vicinity of the
support bound DNA by washing. Since it is possible for the free NTPs to
bind to the ssDNA, the wash should take place at a temperature so that the
free NTPs do not bind to the ssDNA, but not high enough to dehybridize the
double stranded DNA. In order not to deactivate the polymerase at such a
temperature, it is preferred that a high temperature polymerase be
utilized such as the aforementioned taq polymerase. Since the double
stranded DNA is in the environment of a solid support, it is also required
that the solid support surface not attract the NTPs, so Teflon or
similarly non-adhering lining for the solid support should be utilized.
The wash which is used to wash the free NTPs from the support bound double
stranded DNA may be any convenient wash such as buffered saline, or
polymerase buffer without the NTPs.
Washing is also a rapid step since it is contemplated that one would only
be using small quantities of DNA concentrated in a small area. The washing
should only take around a few seconds or less.
Since only one of the four types of NTPs will be added to the
oligonucleotide, it may be read by its fluorescent label using a suitable
fluorescent detection means. According to the preferred embodiment of the
present invention this may be done by making the solid support for the
bound DNA the end of an optical fiber. By transmitting the radiation of
appropriate exciting wavelength through the fiber, the labelled DNA at the
end of the fiber will fluoresce and emit the light of appropriate
fluorescent frequency. The emitted fluorescent light will be partially
transmitted back into the optical fiber in the reverse direction to the
exciting light.
This back propagating light can be separated spectrally by such means as an
etched diffraction grating on the fiber. The returned light spectrum
identifies the particularly bound NTP. It will be within the skill of
those of ordinary skill in the art to adjust the concentration of the
bound DNAs such that there will be a sufficient number of fluorescent
molecules present for optical detection by this means.
It is preferred that the fluorescent dye marker and the 3'-blocking group
be present in the same substituent, for example as shown below.
##STR4##
The identification of the bound NTP identifies its complementary base as
the first unpaired base in the unknown sequence of the ssDNA.
The next step is to remove the 3'-block and associated fluorescent label
from the bound NTP to prepare for the addition of the next NTP onto the
oligonucleotide. This removal may be accomplished by chemical means or
photochemical means. The chemical means of removing the 3'-blocker will of
course depend upon the nature of the 3'blocking groups, many of which are
known in the art as shown for example in Chapter 1, Organic Chemistry of
Nucleic Acids, Part B, Eds. N. K. Kochetkov and E. I. Budovskii, Plenum
Press, 1972 and H. Weber and H. G. Khorana, J. Mol. Biol., 72, 219-249
(1972). Methods of their removal are therefore also known.
Preferably the fluorescent label and 3'-blocker may be removed quickly by
photolysis, as shown in FIG. 2. In the photolysis reaction, instead of the
absorbed light energy being re-emitted as fluorescent light, it is
occasionally conveyed to the 3' position by means of an alternating
single-double bonds hydrocarbon chain. The excitation energy can then
catalyse (or enable) the hydrolysis of the acyl group, as shown, Even
after cleavage of the dye label, it is still present in the vicinity for
optical identification. The dyes are chosen with fluorescent emission
spectra that are as well separated as possible but which can be activated
preferably by the same exciting source. Thus the only significant
limitation to detection is the signal-to-noise ratio due to the Rama
scattering of the other materials present, particularly water. By using
activation wavelengths which minimize Raman scattering from water the
signal-to-noise ratio will be improved.
The solid-supported ssDNA will be scanned for the optical labels preferably
using a optical fiber. The end of the fiber may simply be brought close to
the surface of the support for detection, or the end of the fiber may
itself be the solid support. An alternative means is to attach the DNA to
the sides of the core of an optical fiber (by removing the cladding from a
selected area). If attached to the sides of an optical fiber, the DNA is
illuminated by evanescent lighting surrounding the fiber, and only light
emission in this region can couple back into the fiber.
Concentration of the dye in a small area will also improve detection by
increasing the ratio of fluorescent material to background Rama scattering
solution. The optical fiber improves the signal-to-noise ratio compared to
methods of the prior art since the prior art illuminates comparatively
large volumes. Also the use of an optical fiber to bring the light to the
reaction area and to carry the fluorescent output away is simple and can
be used to separate the return light into spectral bands which can be then
detected by small solid state-like detection means such as PIN
photodiodes. Spectral separation may be accomplished for example., by
periodic diffraction grading etched into the fiber or by using a cladding
with a higher dispersion than the core so that different wavelengths will
or will not satisfy the critical angle condition. Wavelengths that do not
satisfy the critical angle condition will escape into the cladding where
they can be detected. Utilizing solid state components, the whole
sequencing cell, including the optical detection means may be provided in
a planar integrated optical system, which can thus be produced by
photolithographic means, as disclosed, for example in Planar Optical
Waveguides and Fibres, H. G. Unger, Clarendon Press, Oxford (1977). Planar
optics also allows many sequencing cells to be produced on the same
substrate and illuminated by the same light source. Signal-to-noise ratios
can be further improved by using time resolved fluorescence (TRF), with
relatively long-lived fluorescent species (e.g., chelated rare earth
ions). By this method the detection occurs after the Raman scattering has
subsided, but the fluorescent species are still in the emission mode, as
disclosed in I. Hemmila, et al., Anal. Biochem. 137, 335-343 (1984), for
example.
Once the fluorescent label and 3'-blocker (which are preferably one and the
same) are removed from the labeled bound NTP, they are separated from the
bound DNA by washing, using, for example, washing reagents described
above.
Then the sequence beginning with exposure to the polymerase and the four
labeled 3'-block NTPs is repeated, with each cycle adding another base to
the growing oligonucleotide, thereby identifying the next base in the
unknown ssDNA sequence and so on. The identification is completed when a
cycle fails to show that a label is present after the polymerase reaction
and washing step.
According to a preferred embodiment of the present invention, the double
stranded DNA is attached onto the end of an optical fiber, and the end of
the fiber may be dipped into each of the reagents in turn. By moving or
vibrating the fiber, there will be rapid mixing of the boundary layer in
the bulk reagent.
Simple flow systems may be used to bring the reagents to the active part of
the fiber, such as capillary electrophoresis or hydrostatic pressure,
connected to the appropriate reagents.
It is expected that all of the bound DNA molecules will start at the same
position on the ssDNA and add the same type of NTP during each cycle of
the process. Due to imperfect mixing or the dynamics of a densely
populated surface the rate of adding NTPs to each of the solid state-bound
DNAs may not be identical. Therefore some of the strands may fall behind
in the sequence of adding an NTP. For example, if a 3'-blocker remains
after the block removal step for a particular DNA unit, that blocker may
not be removed until the next cycle and therefore that particular strand
will fall behind by addition of one NTP unit. The symptoms of the loss of
synchronization of adding an NTP will be the appearance in the fluorescent
spectra of dyes other than the majority dye being detected for the NTP
which was added. As long as the interfering fluorescent spectra do not
interfere substantially with detection of the majority dye being added in
the cycle of interest, this need not be a concern. For sufficiently long
ssDNA, the accumulation of errors may require re-synchronization, forcing
the lagging DNA to catch up with the current DNA. This resynchronization
can be achieved by changing the reaction mixture of the NTPs, and having
two identical reaction cells, one running a few bases ahead of the other.
For example if the sequence in the leading cell is ACGTTC, and the
trailing cell is one base behind (i.e., ACGTT); then instead of adding the
usual four blocked NTPs to the trailing cell, blocked C, A, G may be added
with unblocked T. Those DNA units that are behind in the trailing cell
(e.g., ACGT or ACG) add the unblocked T (or TT), followed by a blocked C,
while the majority will just add the blocked C. The leading cell is
necessary so that the next base for the majority of the trailing cell is
known, and so an unblocked NTP can be added without interfering with the
process. The roles of leading and trailing cells can be switched so that
one can act as the predictor for the other, and vice versa.
Referring to FIG. 1 there is shown a solid substrate 10 covalently linked
through a linker 11. In the FIGURE the linker 11 is attached to the 5' end
of the oligonucleotide 12, however it may also be attached to the 3' end
of the ssDNA 13. The shaded area 14 is the DNA polymerase, which is primed
by the oligo bound to the ssDNA.
Having described the preferred embodiments of the present invention various
modifications will be evident to those of ordinary skill in the art from
the above description as well as from practice of the invention. These
modifications are intended to be within the scope of the invention and the
invention is not intended to be limited in any way except by the scope of
the following claims.
* * * * *
|
|
 |
|
 |
Description |
|
