ANALYTICAL BIOCHEMISTRY
ARTICLE NO.
234, 166– 174 (1996)
0068
‘‘Doublex’’ Fluorescent DNA Sequencing: Two Independent
Sequences Obtained Simultaneously in One Reaction with
Internal Labeling and Unlabeled Primers
Stefan Wiemann,1 Josef Stegemann, Jürgen Zimmermann, Hartmut Voss,
Vladimir Benes, and Wilhelm Ansorge
Biochemical Instrumentation, European Molecular Biology Laboratory, D-69117 Heidelberg, Germany
Received July 11, 1995
The novel ‘‘doublex’’ DNA sequencing technique that
makes it possible to obtain simultaneously two independent sequences from one sequencing reaction with
the use of unlabeled primers and internal labeling is
described. The different sequencing products are labeled in parallel with fluorescein-15-dATP and Texas
red-5-dCTP present in the same tube. The characteristics of T7 DNA polymerase are exploited to ensure that
only either of the labeled dNTPs is incorporated into
the corresponding sequencing products. Specificity of
labeling is ensured by the selection of primers. One of
the unlabeled primers is chosen to be followed by an
‘‘A,’’ the other by a ‘‘C’’ to be incorporated immediately
downstream from the primer binding site. The doublex
sequencing technique is applicable to the simultaneous sequencing of either the same DNA template/
strand or a mixture of different templates. Combinations of unlabeled and labeled primers in the same sequencing reaction are also possible. The two sequences can be determined in parallel and on-line in
the same lanes of a gel with a novel automated DNA
sequencer, which was previously described for use
with labeled primers. q 1996 Academic Press, Inc.
In standard automated DNA sequencing technology,
detection of products is performed on-line and the output is one sequence per reaction (1 –3). To obtain sequence information also from the complementary
strand of double-stranded templates, a second sequenc1
Present address: Molecular Genome Analysis, German Cancer
Research Center, D-69120 Heidelberg, Germany.
2
Abbreviations used: (d)dNTPs, (di)deoxynucleoside triphosphates; A, dATP; C, dCTP; Fl, fluorescein; TR, Texas red; Ar, argon;
HeNe, helium –neon; DTT, dithiothreitol; DMSO, dimethyl sulfoxide;
TBE, Tris– borate– EDTA; EDTA, ethylenedinitrilo tetraacetic acid.
ing reaction was necessary which had to be loaded onto
separate lanes of a gel. We have recently introduced the
simultaneous automated sequencing of both strands
of double-stranded templates with labeled primers (4).
With this technique, the output of sequence is doubled
per reaction. Due to the high labeling cost of primers,
this approach is limited to the complete doublestranded sequencing of short DNA fragments of up to
1000 bases (e.g., small cDNAs) and to the sequencing
from both ends of longer fragments, in one sequencing
reaction with labeled standard primers.
Complete sequencing of large templates necessitates
strategies such as random or ordered subcloning,
nested sets of deletions (5), or primer walking (6). The
advantage of subcloning and nested deletion cloning is
the possibility of using the same standard primers in
all sequencing reactions. On the other hand, these
strategies result in high cloning effort, sequence redundancies, and costs. Currently, the lowest possible redundancies are achieved with the primer walking
strategy. Development of the EMBL multiple segmental DNA synthesizer (7) for the parallel synthesis of
10 primers in small amounts has reduced the cost of
unlabeled sequencing primers (to about $10) and made
primer walking an inexpensive alternative to the other
strategies. Primer design has been optimized (8) for
use in combination with automated DNA sequencers
and internal labeling (9); the success rate of walking
primer sequencing was increased to over 95%. With
primer walking, the effort for cloning and template
preparation is minimized because all sequencing reactions are performed on the same template. Yet, for every new sequence an individual sequencing reaction
was necessary.
The possibility of simultaneous sequencing with two
or even more unlabeled primers had not been investigated thus far, because the specific assignment and
incorporation of only one labeled dNTP2 to a corre-
166
0003-2697/96 $18.00
Copyright q 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.
/ m4927f9375
01-30-96 15:16:14
aba
AP-Anal Bio
SIMULTANEOUS DOUBLEX FLUORESCENT DNA SEQUENCING
sponding primer was thought to be not achievable. This
is now possible with the ‘‘doublex’’ sequencing technique. The advantages of simultaneous sequencing in
the same tubes and primer walking with unlabeled
primers have been combined to allow sequencing on
one or both strands of a double-stranded template or
on mixtures of different single-stranded or doublestranded templates. On-line detection of both sequences may also be performed in parallel or can be
done separately on two separate automated DNA sequencers for the detection of the differently labeled
products.
MATERIALS AND METHODS
Source of DNA and Sequencing Template
Preparations
Full-length cDNAs have been derived from a human
keratinocyte cDNA library cloned into pBluescript vector (Stratagene, La Jolla, CA). Plasmid DNAs were purified using Quiagen (Quiagen, Hilden, Germany) or
Nucleobond AX ion-exchange columns (Macherey-Nagel, Düren, Germany).
167
ing, 1 ml each of 10 mmol/liter fluorescein-15-dATP, 10
mmol/liter Texas red-5-dCTP, and either native T7
DNA polymerase (8 U/ml) or Sequenase (13 U/ml) were
added. Incorporation of labels was achieved in a labeling reaction at 377C for 10 min. Then, 1 ml of extension
buffer (304 mmol/liter NaCitrate, 40 mmol/liter MnCl2 ,
324 mmol/liter DTT) was added and mixed, and the
solution was divided into the four termination mixes
made up of 3 ml of the respective termination solutions
(40 mmol/liter Tris/HCl, pH 7.4, 50 mmol/liter NaCl, 5
mmol/liter of either ddNTP, 1 mmol/liter each of dATP,
dCTP, C7 dGTP, and dTTP) and 1 ml of DMSO. The
sequencing reaction was incubated at 377C for 5 min
and finally stopped by addition of 4 ml Stop solution (6
mg/ml dextran blue and 20 mmol/liter EDTA, pH 7.3, in
deionized formamide). Prior to loading, samples were
denatured at 907C for 4 min.
Automated DNA Sequencer
All primers were designed using either the GeneSkipper program package (EMBL, Heidelberg, Germany) or
the Oligo 4.1 program (Medprobe, Oslo, Norway). Care
was taken to choose only primer pairs that had no
primer dimer formation probability. Primers also had
neither secondary binding sites on the template nor
hairpin loop formation probabilities. One of the primers
was selected to have an ‘‘A’’ to be incorporated directly
downstream from the primer binding site, whereas the
second primer was immediately followed by a ‘‘C’’ to be
incorporated. Primers were synthesised on the EMBL
multiple segmental DNA synthesiser (7).
The setup of the novel two-dye DNA sequencer has
been described in detail elsewhere (4). In short, a helium–neon (HeNe) laser with respective detectors was
mounted into a prototype device additionally to the
standard argon (Ar) laser/detector system. The Ar laser
excites fluorescein-labeled samples, whereas the HeNe
laser is used to excite Texas red-labeled DNA fragments. Both laser beams are coupled into the gel with
help of a single light coupling plate placed between two
spacers on one side of the gel. The spatial distance of
0.7 cm between the lasers and the combination of two
different laser/detector and filter systems with corresponding fluorophores ensures that no cross detection
of the sequence signals occurs (4). The excitation and
emission spectra of fluorescein and Texas red (11, 12)
do only slightly overlap at the wavelengths of lasers
and detectors used in this sequencing system, ensuring
unambiguous sequencing results.
Sequencing Reactions
Gel Conditions
Sequencing reactions were carried out with unlabeled primers using the Autoread sequencing kit (Pharmacia, Uppsala, Sweden). Fluorescein-15-dATP was
obtained from Boehringer Mannheim (Mannheim, Germany), and Texas red-5-dCTP was from NEN-DuPont
(Boston, MA). Sequenase (Version 2.0) was obtained
from United States Biochemical (Cleveland, OH). Five
to 10 mg of double-stranded DNA (5 –8 kb) was mixed
with two unlabeled walking primers, 10 to 100 pmol of
each, in a total volume of 12 ml. Samples were denatured following the quick denaturation protocol (10) by
addition of 1 ml of 1 mol/liter NaOH solution and heating to 657C for 3 min. After cooling at 377C for 1 min,
condensed water was spun down. Then, 1 ml of 1 mol/
liter HCl and 2 ml of annealing buffer (1 mol/liter Tris/
HCl, pH 8, 0.1 mol/liter MgCl2) were added. After mix-
Separation of DNA fragments was done on 6.5% Duracryl (Millipore, Bedford, MA) 1.2 1 TBE denaturing
sequencing gels. Separation distance was 18.5 cm for
Texas red-labeled fragments and 19.2 cm for fluorescein-labeled fragments, using standard A.L.F. (Pharmacia, Uppsala, Sweden) glass plates. Electrophoresis
was limited to 29 W at a temperature of 507C. The
sequences of both reactions were detected on-line and
recorded in a computer. Base calling was done automatically after the end of the run.
Primer Design and Synthesis
/ m4927f9375
01-30-96 15:16:14
aba
RESULTS AND DISCUSSION
Principle of Doublex Sequencing
Simultaneous DNA sequencing reactions with two
unlabeled primers and internal labeling with fluores-
AP-Anal Bio
168
WIEMANN ET AL.
FIG. 1. Schematic diagram of the doublex sequencing reaction for
the simultaneous sequencing with two unlabeled primers and internal labeling. Sequencing reactions are performed in the same tubes.
After denaturation and neutralization, the two unlabeled primers
annealed to their template strands. In the labeling reaction, fluorescein-15-dATP (A-FL), and Texas red-5-dCTP (C-TR) are specifically
incorporated downstream from the corresponding primers. In the
extension/termination reaction, products are extended until a dideoxynucleotide terminates polymerization.
cently labeled dNTPs in the same tubes are described.
With this doublex sequencing technique the output of
each reaction is doubled. Figure 1 shows schematically
the principle of doublex sequencing. The denaturation,
annealing, and labeling reactions are carried out in one
tube. Templates can be the same strand of a single- or
a double-stranded DNA, different strands of the same
double stranded DNA, or from different DNAs. After
denaturation, the primers anneal to their template
strands. Two differently labeled dNTPs are used in the
sequencing reaction. Selective labeling of specific products with only one dye is achieved by the nucleotide
that is incorporated as the first base directly downstream the primer. In doublex sequencing one primer
is chosen to be directly followed by an ‘‘A’’ to be incorporated, the other by a ‘‘C.’’ Only the correct nucleotide
is incorporated in the labeling step of sequencing reactions under the conditions described above. After incorporation of the first labeled dNTP, the polymerase
pauses and does not proceed further during the labeling reaction owing to the low concentration of the two
labeled dNTPs and the lack of other dNTPs.
In the extension/termination reaction no more labeled dNTPs are incorporated due to the large excess of
unlabeled dNTPs over the labeled dNTPs and because
unlabeled dNTPs are prefered substrates for T7 DNA
polymerase compared to labeled dNTPs. The described
sequencing technique takes advantage of a novel automated DNA sequencing device capable of detecting simultaneously and on-line both sequencing products in
parallel (4) but can also be employed when the sequencing products are analyzed on separate DNA sequencers, one for each dye. With the combination of doublex
sequencing and the novel automated sequencer, not
only the sequencing reactions are performed simultaneously but also the on-line detection of the differently
labeled sequencing products is done in parallel. Figure
2 shows typical unprocessed raw sequence data from
doublex sequencing on both strands of a plasmid DNA.
Two unlabeled walking primers were used in the same
sequencing reaction with fluorescein-15-dATP (Fig. 2A)
and Texas red-5-dCTP (Fig. 2B). Both reactions had
been carried out simultaneously in the same tubes. The
pile-ups in both sequences around position 38 bp originate from an impurity in the conmmercially available
labeled dNTPs and do appear consistently.
Effects of Two Differently Labeled dNTPs Present in
One Sequencing Reaction
Depending on the template sequence, two major effects are possible when two differently labeled dNTPs
are combined in one sequencing reaction that is carried
out in one tube: (i) After incorporation of the first labeled dNTP, a second dNTP, labeled with the other
dye, could be incorporated. For example, after incorporation of a Texas red-labeled dCTP, a fluorescein dATP
might be incorporated if the sequence downstream
from the primer binding site was accordingly. If incorporation of labeled dATP did also occur downstream
from the second primer where it was intended, this
might lead to ambiguous sequence in the fluorescein
channel. (ii) Labeling of the primer might occur and
obscure the sequencing results if the last nucleotide
FIG. 2. Double-stranded plasmid DNA containing a human cDNA was sequenced simultaneously with two unlabeled primers in the
presence of each fluorescein-15-dATP and Texas red-5-dCTP, as described under Materials and Methods. Sequencing products were loaded
into the same lanes of a sequencing gel, separated, and detected on-line and independently with the Ar and HeNe laser/detector systems
of an automated DNA sequencer. Shown are unprocessed raw sequence data. (A) Sequence generated with fluorescein-15-dATP. (B) Sequence
generated with Texas red-5-dCTP.
/ m4927f9375
01-30-96 15:16:14
aba
AP-Anal Bio
169
/ m4927f9375
01-30-96 15:16:14
aba
AP-Anal Bio
170
WIEMANN ET AL.
FIG. 3. Test for multiple incorporation of labeled dNTPs. Sequencing primers were selected that were followed by a C and then an A,
and primers that were followed by an A and then a C to be incorporated at positions /1 and /2. Sequencing reactions were carried out in
presence of fluorescein-15-dATP and Texas red-5-dCTP, with native T7 DNA polymerase. Products were separated on a sequencing gel and
data were recorded. Circles around the labels (TR, Texas red; Fl, fluorescein) represent the signals of the respective dyes that are recorded.
(A) Texas red signals from a primer that had a C at position /1 and an A at position /2. (B) Fluorescein signals obtained with the same
primer. (C) Texas red signals from a primer that had an A at position /1 and a C at position /2. (D) Fluorescein signals obtained with
the same primer.
of one primer was the same as the first base to be
incorporated downstream from the other primer. Incorporation of fluorescein-15-dATP into the primer by the
subsequent action of 3* –5* exonuclease and polymerase activities of native T7 DNA polymerase has been
demonstrated (8).
To test for possible multiple incorporation of labels,
sequencing primers were selected that were followed
by an A and then a C, and primers that were followed
by a C and then an A to be incorporated at positions
/1 and /2. Sequencing reactions were performed with
native T7 DNA polymerase and a mixture of fluorescein-15-dATP and Texas red-5-dCTP. If incorporation
of one labeled dNTP did inhibit the polymerase from
incorporating a second labeled dNTP, only the labeled
base immediately downstream from the primer, but not
the base at the second position, should be incorporated.
This was found with the primers that were followed by
a C and then an ‘‘A’’ at position /2. Only products
that were labeled with Texas red-labeled dCTP were
detectable (Fig. 3A). No products were detectable in
/ m4927f9375
01-30-96 15:16:14
aba
the fluorescein channel (Fig. 3 B); consequently, no
fluorescein-15-dATP had been incorporated.
When the primers were followed by an A and when
the C was the second base that could be incorporated,
labeling with Texas red-5-dCTP at position /2 was also
observed (Fig. 3C). These sequencing products carry
two labels, both fluorescein and Texas red. The mobility
shift caused by incorporation of the Texas red dye does,
however, not spoil the results obtained in the fluorescein channel (Fig. 3D). The amount of sequencing products that carries two labels is too small to be detectable
above the signals from products that carry only the
fluorescein label. Fluorescein-15-dATP seems to be not
as potent for inhibiting multiple incorporation of labels
as Texas red-5-dCTP. This labeling at position 2 does,
however, not influence the usefulness of the doublex
sequencing method. The sequence obtained with the
second primer is also unambiguous because labeling
with Texas red-5-dCTP in position /2 is much weaker
compared to the labeling in position /1 (compare Fig.
3A). The results are summarized in Table 1. The major
AP-Anal Bio
SIMULTANEOUS DOUBLEX FLUORESCENT DNA SEQUENCING
TABLE 1
Incorporation of Labeled dNTPs in the Second Position
Downstream from the Primer
/1/2
PrimerÉAC
/1/2
PrimerÉCA
Fl-dATP
TR-dCTP
///
/
0
///
Note. Primers that were followed by an A to be incorporated at
position /1 and a C at position /2 (PrimerÉAC) were sequenced with
fluorescein-15-dATP and Texas red-5-dCTP simultaneously. Primers
that were followed by a C to be incorporated at position /1 and
an A at position /2 (PrimerÉCA) were sequenced under the same
conditions. The strength of sequencing signals detected in the fluorescein (Fl-dATP) and Texas red channels (TR-dCTP) was evaluated
and classified from no signal (0) to very strong signals (///).
difference between fluorescein-15-dATP and Texas red5-dCTP is that the latter is incorporated at a very low
rate even if a fluorescein-15-dATP had been incorporated previously. Incorporation of Texas red-5-dCTP
inhibits the polymerase from incorporating a fluorescein-15-dATP. The same results were obtained when
the experiments were repeated with Sequenase instead
of native T7 DNA polymerase.
Nevertheless, even primers that are followed by several A’s which could be incorporated downstream from
the primer lead to a clear sequence under the sequencing conditions described above (8). Even if incorporation of a second labeled dATP does occur, any obscuring
signals that would originate from the mobility shift
caused by this second label are not detectable above
the noise level (4). If a high proportion of products did
carry varying numbers of label, the sequence would be
ambiguous owing to the different mobility shifts caused
by the variety in the number of labels.
171
next base. The second base to be incorporated following
the primer was neither A nor C. Both fluorescein-labeled dATP and Texas red-labeled dCTP were included
in the sequencing reaction. When the reaction was performed with Sequenase (Version 2.0), sequencing products were formed only with the Texas red-labeled dCTP
that was incorporated downstream from the primer
(Fig. 4A) whereas the fluorescein-labeled dATP was not
incorporated into the primer (Fig. 4B). When the reaction was performed with native T7 DNA polymerase,
products labeled with fluorescein dATP were also generated (Fig. 4D) in addition to the expected sequencing
products labeled with Texas red dCTP (Fig. 4C). These
additional signals could obscure the results obtained
in a simultaneous sequencing reaction with a second
primer that was followed by an A. The results are summarized in Table 2. Comparable results were obtained
when primers were used that carried an A as 3*-terminal base and that had a C as the first base to be incorporated downstream from the primer (data not shown).
Choice of Enzyme
In conclusion from the experiments described above,
Sequenase is best suited for simultaneous sequencing
with two unlabeled primers and internal labeling due
to its lack of a 3* –5* exonuclease activity. Selection of
primers is simplified when using this enzyme because
only the sequence downstream from the primer needs
to be considered for the labeling step. Nevertheless, all
sequencing primers still need to be selected carefully
to avoid primer dimer formation, hairpin loops, and
secondary binding of primers. In combination with native T7 DNA polymerase, doublex sequencing is only
possible with primers that do not end with a C when
the next base to be incorporated would be an A and
vice versa; otherwise, an ambiguous sequence might
result.
Influence of the 3* – 5* Exonuclease Activity of Native
T7 DNA Polymerase on the Simultaneous
Sequencing Reaction
Further Improvements and Applications of Doublex
Sequencing
We have recently shown that native T7 DNA polymerase but not Sequenase (Version 2.0) incorporates a
labeled dATP into primers if the 3*-terminal base of
these primers is an A (8). The same reaction takes place
when labeled dATP is replaced with fluorescently labeled dCTP and the primers end with a C. The lack of
a 3* –5* exonuclease activity in Sequenase is the major
difference of this enzyme compared to native T7 DNA
polymerase (13).
A possible influence of the exonuclease activity of
native T7 DNA polymerase on simultaneous sequencing was tested. Sequencing reactions were performed
with a primer that carried an A as 3*-terminal base
and that was followed by a C to be incorporated as the
Texas red-labeled dCTP has recently become available as a second dye for internal labeling and can now
be used in combination with fluorescein-15-dATP making the doublex sequencing technique possible. Using a
stable fluorescently labeled dGTP and a corresponding
third primer, doublex sequencing may even be expanded to ‘‘triplex’’ sequencing with the output of then
three independent sequences from one sequencing reaction. Fluorescently labeled dUTP, however, which is
the fourth labeled nucleotide and is already available,
is at present not applicable for simultaneous sequencing under the reaction conditions described above.
After incorporation of labeled dUTP, T7 DNA polymerase incorporates a second labeled dNTP downstream
/ m4927f9375
01-30-96 15:16:14
aba
AP-Anal Bio
172
WIEMANN ET AL.
FIG. 4. Test for incorporation of label into the primer. Sequencing primers were selected that had an A as the 3 *-terminal base and were
followed by a C to be incorporated as the next base. Sequencing reactions were carried out in the presence of both fluorescein-15-dATP and
Texas red-5-dCTP and with either Sequenase (Version 2.0) or native T7 DNA polymerase. Sequencing products were separated on a
sequencing gel and data were recorded. (A) Texas red signals from sequencing with Sequenase. (B) Fluorescein signals obtained with
Sequenase. (C) Texas red signals from sequencing with native T7 DNA polymerase. (D) Fluorescein signals obtained with native T7 DNA
polymerase.
from the dUTP leading to ambiguous sequence due to
multiple incorporation of labels.
The number of sequencing reactions that can be carried out simultaneously may be further enhanced, by
a combination of different fluorescently labeled primers
and unlabeled walking primers provided that the internal labels are specifically incorporated downstream
from their corresponding unlabeled primers according
TABLE 2
Comparison of Native T7 DNA Polymerase and Sequenase
in Doublex Sequencing Reactions
Native T7 DNA polymerase
Sequenase (Version 2.0)
Fl-dATP
TR-dCTP
/
0
///
///
Note. Primers that carried an A as 3*-terminal base (position -1)
and that were directly followed by a C (position /1) to be incorporated
as the next base were sequenced with both fluorescein-15-dATP and
Texas red-5-dCTP present in the labeling step. The strength of sequencing signals detected in the fluorescein (Fl-dATP) and Texas red
channels TR-dCTP) was evaluated and classified from no signal (0)
to very strong signals (///).
/ m4927f9375
01-30-96 15:16:14
aba
to the rules described above. The number of laser/detector/dye combinations that can be used without spectral
overlaps is the primary limitation for the number of
sequencing reactions that can be carried out in parallel.
Doublex sequencing is not limited to the sequencing
from one template, but the reactions can as well be
performed on mixtures of single-stranded or doublestranded templates. Care must be taken that different
templates have about the same molar concentration in
the reaction (not differing by more than a factor of 5)
or that the concentration of primers is limiting for the
production of template/primer complexes. If one of the
templates was in great excess, this template might outcompete the other template for the generation of sequencing products even though the polymerase is present in huge amounts and should not be limiting in the
described protocol.
This technique can be used in any primer walking
approach provided that the primer pairs are designed
according to the rules described above. Currently, doublex sequencing is routinely applied to the sequencing
of full-length human cDNAs and polymorphic microsatellites in the course of the EU Human Genome Analysis
Program BIOMED1 (14 –16).
AP-Anal Bio
SIMULTANEOUS DOUBLEX FLUORESCENT DNA SEQUENCING
We have previously introduced simultaneous sequencing of double-stranded templates in one sequencing reaction from both ends with the use of labeled
standard primers (4). The sequencing approach with
labeled primers is helpful for sequencing small fragments as short cDNAs in the double-stranded form
with standard primers. For the complete sequencing of
longer fragments, however, either subcloning or primer
walking (6) strategies are necessary. In strategies that
involve subcloning and sequencing with standard primers (e.g., random shotgun sequencing and multiplex
sequencing; 17, 18), from every clone the output of new
sequence is at most two sequences. A great advantage
of primer walking over subcloning strategies is the
need for only one template DNA for the whole sequencing project, cloning effort is minimized. Low redundancy sequencing with about three readings per base
pair (6, 19) is only achieved with the primer walking
strategy compared to redundancies of greater than
eight with random shotgun approaches (20). The lower
the redundancy, the lower the cost for template preparation and sequencing reactions. By applying doublex
sequencing in primer walking the cost for template
preparations as well as for sequencing reactions is further reduced.
In computer simulations we and others have delineated that inserts of cosmid size (about 40 kb) are most
efficiently sequenced when a limited random sequencing phase is carried out (21, 22). A cost minimum is
achieved with the proposed RANDI strategy (21) if 150
to 200 reactions are performed with random clones and
simultaneous sequencing with two labeled primers in
one reaction. Gaps are closed and single-stranded regions are made double-stranded in a finishing phase
which employs about 50 walking primers. The doublex
sequencing technique makes RANDI even more efficient by allowing simultaneous sequencing from the
starting point to the finishing phase, reducing the cost
of sequencing reactions by 50%.
Another possible application of doublex sequencing
is the double-stranded sequencing of PCR products in
in the growing field of clinical diagnostics. High sequencing accuracy in combination with low cost are
major demands for routine analyses such as the screening for mutations and heterozygosities in genes such as
p53 (23), the differentiation of bacterial or viral strains
(e.g., mycobacteria, 24; hepatitis C, 25), or in HLA typing (26). By using two unlabeled sequencing primers
and internal labeling in one sequencing reaction, the
sequence can now be obtained in the double-stranded
form at low cost. The sequence information is gathered
with the lowest possible sequencing redundancy of one
sequencing reaction, while the highest possible accuracy is achieved by the sequencing of both strands.
The major reason that the simultaneous sequencing
with more than one unlabeled primer had thus far been
/ m4927f9375
01-30-96 15:16:14
aba
173
regarded as not being possible is the necessity for specific incorporation of only one of the differently labeled
dNTPs into the corresponding products, in order to obtain specifically and differently labeled sequencing
products. Here we describe for the first time a technique that gives a solution to the problem of specific
incorporation of different labels. In combination with
the novel two-dye DNA sequencing device (4) the value
of this technique is further enhanced since two different sequencing products are not only generated in one
sequencing reaction but these products are also analyzed on-line and in parallel in the same lanes on a
gel. The output of each sequencing reaction and gel is
doubled compared to standard automated sequencing
technology. The products of a doublex sequencing reaction could, however, also be loaded on two separate
DNA sequencers, one for detection of the products labeled with fluorescein and the other for products carrying the Texas red label.
ACKNOWLEDGMENTS
This work was supported by the European Union in the course of
the Human Genome Analysis Program (BIOMED 1).
REFERENCES
1. Ansorge, W., Sproat, B., Stegemann, J., and Schwager, C. (1986)
J. Biochem. Biophys. Methods 13, 315–323.
2. Smith, L. M., Sanders, J. Z., Kaiser, R. J., Hughes, P., Dodd, C.,
Connel, C. R., Heiner, C., Kent, S. B. H., and Hood, L. E. (1986)
Nature 321, 674–679.
3. Ansorge, W., Sproat, B., Stegemann, J., Schwager, C., and Zenke,
M. (1987) Nucleic Acids Res. 15, 4593 – 4602.
4. Wiemann, S., Stegemann, J., Grothues, D., Bosch, A., Estivill,
X., Schwager, C., Zimmermann, J., Voss, H., and Ansorge, W.
(1995) Anal. Biochem. 224, 117– 121.
5. Henikoff, S. (1984) Gene 28, 351.
6. Voss, H., Wiemann, S., Grothues, D., Sensen, C., Zimmermann,
J., Schwager, C., Stegemann, J., Erfle, H., Rupp, T., and Ansorge,
W. (1993) BioTechniques 15, 714– 721.
7. Ansorge, W., Voss, H., Wiemann, S., Schwager, C., Sproat, B.,
Zimmermann, J., Stegemann, J., Erfle, H., Hewitt, N., and Rupp,
T. (1992) Electrophoresis 13, 616–619.
8. Wiemann, S., Rupp, T., Zimmermann, J., Voss, H., Schwager,
C., and Ansorge, W. (1995) BioTechniques, 18, 688– 697.
9. Voss, H., Wiemann, S., Wirkner, U., Schwager, C., Zimmermann,
J., Stegemann, J., Erfle, H., Hewitt, N., Rupp, T., and Ansorge,
W. (1992) Methods Mol. Cell. Biol. 3, 153–155.
10. Zimmermann, J., Voss, H., Schwager, C., Stegemann, J., Erfle,
H., Stucky, K., Kristensen, T., and Ansorge, W. (1990) Nucleic
Acids Res. 18, 1067.
11. Smith, L. M., Sanders, J. Z., Kaiser, R. J., Hughes, P., Dodd, C.,
Connel, C. R., Heiner, C., Kent, S. B. H., and Hood, L. E. (1986)
Nature 321, 674–679.
12. Titus, J. A., Haugland, R., Sharrow, S. O., and Segal, D. M.
(1982) J. Immunol. Methods 50, 193–204.
13. Tabor, S., and Richardson, C. C. (1989) J. Biol. Chem. 264, 6447 –
6458.
AP-Anal Bio
174
WIEMANN ET AL.
14. Bosch, A., Wiemann, S., Guimerà, J., Ansorge, W., Patterson,
D., and Estivill, X. (1993) Hum. Mol. Genet. 2, 1744.
15. Bosch, A., Wiemann, S., Ansorge, W., Patterson, D., and Estivill,
X. (1994) Hum. Genet. 93, 359–360.
16. Bosch, A., Wiemann, S., Guimerà, J., Ansorge, W., Patterson,
D., and Estivill, X. (1995) Hum. Genet. 95, 119–122.
17. Church, G. M., and Kieffer-Higgins, S. (1988) Science 240, 185–
188.
18. Cherry, J. L., Young, H., Di Sera, L. J., Ferguson, F. M., Kimball,
A. W., Dunn, D. M., Gesteland, R. F., and Weiss, R. B. (1994)
Genomics 20, 68–74.
19. Wiemann, S., Voss, H., Schwager, C., Rupp, T., Stegemann, J.,
Zimmermann, J., Grothues, D., Sensen, C., Erfle, H., Hewitt, N.,
Banrevi, A., and Ansorge, W. (1993) Yeast 9, 1343 –1348.
20. Koop, B. F., Rowan, L., Chen, W.-Q., Deshpande, P., Lee, H.,
and Hood, L. (1993) BioTechniques 14, 442– 447.
/ m4927f9375
01-30-96 15:16:14
aba
21. Voss, H., Schwager, C., Wiemann, S., Zimmermann, J., Stegemann, J., Erfle, H., Voie, A-M., Drzonek, H., and Ansorge, W.
(1995) J. Biotechnol. 41, 121–129.
22. Roach, J., Boysen, C., Wang, K., and Hood, L. (1995) Genomics
26, 345–353.
23. Cripps, K. J., Purdie, C. A., Carder, P. J., White, S., Komine, K.,
Bird, C. C., and Wyllie, A. H. (1994) Oncogene 9, 2739 –2743.
24. Frothingham, R., and Wilson, K. H. (1993) J. Bacteriol. 175,
2818 –2825.
25. Tokita, H., Okamoto, H., Tsuda, F., Song, P., Nakata, S., Chosa,
T., Iizuka, H., Mishiro, S., Miyakawa, Y., and Mayumi, M. (1994)
Proc. Natl. Acad. Sci. USA 91, 11022 –11026.
26. Versluis, L. F., Rozemuller, E., Tonks, S., Marsh, S. G., Bouwens,
A. G., Bodmer, J. G., and Tilanus, M. G. (1993) Hum. Immunol.
38, 277–283.
AP-Anal Bio