ABSTRACT
DNA primer sets, labeled with two fluorescent dyes to exploit fluorescence
energy transfer (ET), can be efficiently excited with a single laser line and
emit strong fluorescence at distinctive wavelengths. Such ET primers are
superior to single fluorophore-labeled primers for DNA sequencing and other multiple color-based analyses [J. Ju, C. Ruan, C. W. Fuller, A. N. Glazer and R. A.
Mathies (1995)
Proc. Natl. Acad. Sci. USA
92, 4347-4351]. We describe here a novel method of constructing fluorescent
primers using a universal ET cassette that can be incorporated by conventional
synthesis at the 5
'
-end of an oligonucleotide primer of any sequence. In this cassette, the
donor and acceptor fluorophores are separated by a polymer spacer (S
6
) formed by six 1
'
,2
'
-dideoxyribose phosphate monomers (S). The donor is attached to the 5
'
side of the ribose spacer and the acceptor to a modified thymidine attached to the 3
'
end of the ribose spacer in the ET cassette. The resulting primers, labeled
with 6-carboxy- fluorescein as the donor and other fluorescein and rhodamine dyes
as acceptors, display well-separated acceptor emission spectra with 2-12-fold enhanced fluorescence intensity relative to that of the
corresponding single dye-labeled primers. With single-stranded M13mp18 DNA as the template, a typical run with these ET
primers on a capillary sequencer provides DNA sequences with 99% accuracy in the first 550 bases using the same amount of DNA
template as that typically required using a four-color slab gel automated sequencer.
Four-color DNA sequencing with fluorescent primers or terminators is the most
commonly used method in high-throughput DNA sequencing laboratories (
1
-
3
). The fluorophores used in the previous studies are typically fluorescein and
rhodamine derivatives. Recently, near-infrared fluorophores have also been developed for DNA sequencing in a one-color, four-lane format (
4
). In our studies, we have exploited fluorescence resonance energy transfer
(ET), a well known and useful spectroscopic phenomenon (
5
-
13
), to optimize the spectroscopic properties of the fluorescent tags. This led to
the development of multiple-color fluorescent ET primers which are superior for DNA sequencing and PCR
fragment analyses (
14
-
17
). The fluorescent ET primers we designed carry a fluorescein derivative at the
5'-end as a common energy donor and other fluorescein and rhodamine
derivatives attached to a modified thymidine within the primer sequence as
acceptors. These primers all have strong absorption at a common excitation
wavelength (488 nm) and well separated fluorescence emission maxima of 525,
555, 580 and 605 nm, respectively. The acceptor emission intensity of the ET
primers is dependent on the spacing between the donors and the acceptors. The
electrophoretic mobilities of the primers also vary with the spacing. Sets of
four different primers can be selected where the number of nucleotides between
the donor and acceptor is chosen to maximize the acceptor fluorescence emission
and where the electrophoretic mobilities of the single base extension DNA
fragments generated with the primers are closely matched. High precision
sequencing and mapping with the ET primers requires much less DNA template than
previously available procedures (
14
-
17
).
We describe here a method for the synthesis of fluorescent primers labeled with
an ET cassette and their application in DNA sequencing. The ET cassette is
constructed with a polymer spacer carrying donor and acceptor fluorophores,
which is not constrained by the requirement of complementarity to a particular target DNA sequence, or
by the spacing between a 5'-terminal base and a modified thymidine residue within the DNA
primer sequence. Sets of primers modified with such ET cassettes can be
efficiently excited with a single laser line and display well-separated acceptor emission spectra. Fluorescence emission intensities of
primers labeled with ET cassettes are 2-12-fold greater than those of corresponding primers labeled with only
the acceptor dye. These primer sets allow high precision DNA sequencing by the
Sanger dideoxy method (
18
). The development of these universal cassettes provides a simple and practical
method to fluorescently tag oligonucleotides of any sequence as well as other
target molecules with ET coupled dyes.
Chemicals were purchased from Applied Biosystems (Foster City, CA) unless
otherwise stated. Sequenase Version 2.0 T7 DNA polymerase and other DNA
sequencing reagents were obtained from Amersham Life Science (Cleveland, OH).
Oligodeoxynucleotides were synthesized by the phosphoramidite method on an Applied Biosystems 392 DNA synthesizer. Figure
1
presents a schematic drawing of the components of ET cassette-labeled primers. The structures of the four ET cassette labeled primers
used in this work and a representative synthetic reaction scheme are presented
in Figure
2
. The M13 (-21) universal primers containing 18 nucleotides (nt) and a spacer (S
6
), with the sequence 5'-S
6
T*GTAAAACGACGGCCAGT-3', were synthesized with donor-acceptor fluorophore pairs separated by a polymer linkage (S
6
) formed by six 1',2'-dideoxyribose phosphates (S). The 1',2'-dideoxyribose phosphates were introduced
using 5'-dimethoxytrityl-1',2'-dideoxyribose-3'-[(2-cyanoethyl)-(
N
,
N
-diisopropyl)]-phosphoramidite (dSpacer CE Phosphoramidite, Glen Research, Sterling, VA) (
19
,
20
). The oligomer contains a modified base T* introduced by the use of 5'-dimethoxytrityl-5- [
N
-(trifluoroacetyl aminohexyl)-3-acrylimido]-2'-deoxyuridine, 3'-[(2-cyano- ethyl)-(
N
,
N
-diisopropyl)]-phosphoramidite (Amino-Modifier C6 dT, Glen Research) which has a protected primary amine linker arm at the C-5 position. The T* was attached to the 5' end of the nucleotide sequence and 3' to the spacer. The donor dye was attached
to the end of the spacer on the 5' side of the oligomer, and the acceptor dye was attached to the primary
amine group on the T*. The ET primers were purified as previously described (
14
), quantified by their 260 nm absorbances corrected for the dye absorptions, and
then stored in 10 mM Tris-HCl, 1 mM EDTA, pH 8.0 at a final concentration of 0.4 pmol/[mu]l for DNA sequencing reactions.
Sequencing was performed using M13mp18 template DNA and modified T7 DNA
polymerase on a four-color capillary electrophoresis (CE) DNA sequencer designed in our
laboratory (
15
). Four reactions were run, one for each dye/ddNTP combination. The reactions
containing ddCTP were run with the FS
6
F primer, ddATP with the FS
6
R primer, ddGTP with the FS
6
T primer, and ddTTP with the FS
6
J primer. The preparation of the sample for sequencing is similar to that previously described with 7-deaza-dGTP or dITP in place of dGTP (
14
,
15
). The sequencing of samples using 7-deaza-dGTP was performed in Mn
2+
-containing buffer. Samples were introduced into a 50 cm long (30 cm
effective separation length) 3% T, 3% C polyacrylamide gel-filled capillary (J&W Scientific, Folsom, CA) by electrokinetically injecting for 45 s.
Electrophoresis was performed at 200 V/cm in 100 mM Tris-borate-7 M urea, pH 8.3 (room temperature). A matrix transformation was
performed on the raw data to correct for the crosstalk between the four
channels (
15
) to generate the four-color CE sequencing profile.
We have established previously that when regular nucleotides are used as spacers
between the donor and acceptor fluorophores the fluorescence emission intensity
of the ET primers increases as the number of spacing nucleotides is increased
in the order 1, 2, 3, 4, 10 (
15
). ET primers with a 6 nt spacing, designed for multiple color PCR-based DNA typing, exhibited up to 10-fold fluorescence enhancement compared with the corresponding single dye-labeled primers (
16
). On the basis of these observations, each of the ET cassette-labeled primers described here was synthesized with six 1',2'-dideoxyribose phosphates as a spacer. The fluorescence
intensities of these primers are comparable to those of the optimized set of ET
primers we previously described (
15
). The use of such spacers is advantageous in several aspects: (i) the spacer
will not hybridize to any sequences within the DNA template and therefore false
priming is avoided; (ii) the linkage of the spacer maintains the natural
nucleic acid phosphate functionality, which avoids possible anomalies in
electrophoretic mobility; and (iii) the elimination of the aromatic base groups
on the deoxyribose rings in the spacer may reduce the likelihood of
fluorescence quenching.
Figure
3
presents the absorption and emission spectra of the ET primers. Each primer
exhibits the characteristic absorption of FAM at 496 nm, as well as strong
absorption at 525 nm due to JOE in FS
6
J, at 555 nm due to TAMRA in FS
6
T, and at 585 nm due to ROX in FS
6
R. The fluorescence emission spectra of the ET primers are dominated by the
acceptor emission, indicative of efficient energy transfer. While the emission
maximum of FS
6
F is at 525 nm, the emission of FS
6
J with 488 nm excitation is Stokes-shifted to 555 nm, that of FS
6
T is shifted to 580 nm, and that of FS
6
R is shifted to 605 nm. For FS
6
R, the Stokes shift is >100 nm. Figure
3
also presents emission spectra of the single dye-labeled primers, measured at the same molar concentration as that of the
corresponding ET primers. Since the common donor (FAM) in all four ET primers
efficiently captures the excitation energy at 488 nm and then efficiently
transfers the energy to the long-wavelength absorbing acceptors, substantial enhancement of the ET primer
acceptor emission intensity is observed compared with that of corresponding
single dye-labeled primers. The fluorescence intensity improvements derived from the
data in Figure
3
are: FS
6
F = 1.7* FAM; FS
6
J = 3* JOE; FS
6
T = 10* TAMRA; FS
6
R = 12* ROX. In the case of FS
6
F, where the donor and acceptor are identical, the fluorescence intensity
increases ~2-fold compared with single FAM-labeled primers, as expected. The ET efficiency was calculated
to be 83% for FS
6
J, 84% for FS
6
R and 85% for FS
6
T.
We thank the members of the Berkeley High-Sensitivity DNA Analysis Project, John Chase and Carl W. Fuller from
Amersham Life Science Inc. for many valuable interactions. This research was
supported by the Director, Office of Energy Research, Office of Health and
Environmental Research of the US Department of Energy under contract DE-FG-91ER61125. J.J. was supported by a DOE Human Genome Distinguished
Postdoctoral Fellowship sponsored by the US Department of Energy, Office of
Health and Environmental Research and administered by the Oak Ridge Institute
for Science and Education. Financial support from Amersham Life Science Inc. is
also gratefully acknowledged.
REFERENCES
Return