Content uploaded by Wilhelm Ansorge
Author content
All content in this area was uploaded by Wilhelm Ansorge on Jan 06, 2014
Content may be subject to copyright.
Nucleic
Acids
Research,
Vol.
18,
No.
11
3419
One
label,
one
tube,
Sanger
DNA
sequencing
in
one
and
two
lanes
on
a
gel
W.Ansorge*,
J.Zimmermann,
C.Schwager,
J.Stegemann,
H.Erfle
and
H.Voss
European
Molecular
Biology
Laboratory,
Postfach
102209,
D-6900
Heidelberg,
FRG
Submitted
April
24,
1990
Sequencing
in
less
than
four
lanes
on
a
gel,
using
only
one
fluorescent
label
for
the
four
bases
(without
variations
in
electrophoretic
mobilities
or
problems
with
spectral
overlap)
in
one
tube
was
introduced
in
(1).
In
that
process
T7
DNA
polymerase
and
buffer
containing
magnesium
ions
were
used.
By
adjusting
the
ddNTPs
ratios
the
sequence
is
deduced
from
the
peak
magnitudes
corresponding
to
the
four
bases.
The
resolution
was
limited
by
nonuniformity
of
the
peaks
and by
decrease
of
overall
labelling
due
to
the
two-step
process
used
in
that
sequencing
protocol.
In
the
method
described
in
this
note,
applicable
both
to
fluorescent
and
radioactive
endlabelling,
we
have
used
the
one-step
protocol
with
manganese
ions
in
one
tube
as
described
(2).
This
protocol
results
in
uniform
overall
labelling
(within
15%
up
to
base
300).
Biochemicals,
annealing,
termination
and
stop
mixes,
as
well
as
diluted
enzyme
were
as
described
in
(4).
Briefly,
1.25
itg
of
M13
ssDNA
in
5
Il
solution
were
combined
with
2
IL
fluorescently
labelled
primer
(2
,tM)
and
6
ILI
annealing
mix.
The
mixture
was
denatured
at
65°C
for
A
B
C
3
minutes
and
cooled
down
to
25°C.
Then
4U
of
T7
DNA
polymerase
in
2
Il
were
added
and
mixed.
Nucleotide
mixes,
as
described
below,
were
added
and
strand
synthesis
proceeded
for
10
minutes
at
37°C.
The
reaction
was
stopped with
20
IL
stop
mix
containing
10
mM
NaOH
for
complete
denaturation.
Before
loading,
samples
were
heated
at
65°C
for
3
minutes
and
cooled
down
on
ice.
The
amount
of
sample
is
sufficient
for
4
gel
runs
on
the
EMBL
automated
fluorescent
DNA
Sequencer
or
the
A.L.F.
Sequencer
(Pharmacia
LKB).
One
lane,
one
tube
reaction:
Nucleotide
mixes
were
prepared
from
the
termination
mixes
(4)
in
a
ratio
of
T:C:G:A
=
2:2:1:0.5
(6
jl
of
T
termination
mix,
6
Id
of
C-mix,
3
il
of
G-mix,
1.5
g1l
of
A-mix).
Two
lane
sequencing
reactions:
Nucleotide
mixes
were
prepared
as
above,
in
a
ratio
of
1:1
for
the
T>C
and
2:1
for
the
G>A
reaction.
Fig.
1
(6)
shows
raw
data
obtained
in
the
standard
(2)
four
lanes
one-step
protocol
(Fig.
IA),
in
two
lanes
(Fig.
iB)
and
Figure
1.
*
To
whom
correspondence
should
be
addressed
3420
Nucleic
Acids
Research,
Vol.
18,
No.
11
Figure
2.
one
lane
(Fig.
1C)
according
to
the
above
protocols.
Displayed
are
bases
22-53
of
the
M13mpl8
DNA
analysed
on
6%
gel
with
separation
distance
20
cm
(original
in
four
colours).
As
shown
in
Fig.
2
(displayed
bases
160-206),
the
resolution
in
the
one
lane
method
is
at
least
200
bases
(as
expected
it
is
higher
in
two
lanes,
around
300
bases,
data
not
shown).
Using
higher
gel
concentration
(8-10%)
or
longer
separation
distance,
we
believe
the
resolution
could
be
improved
to
around
300
bases
or
higher,
the
upper
limit
being
given
by-
resolution
of
single
bases
in
the
gel.
We
noticed
different
discrimination
levels
for
the
four
ddNTPs
by
the
T7
DNA
polymerase
in
the
presence
of
manganese
ions.
When
equal
molar
ratios
of
the
ddNTPs
(1:1:1:1)
are
used
in
the
one
tube
reaction,
the
resulting
peaks
are
not
of
the
same
magnitude,
but
increase
in
the
order
A
<
G
<
C
<
T.
This
'natural'
preference
of
the
T7
DNA
polymerase
is
respected
in
the
present
protocol,
in
which
the
dideoxy
molar
ratios
are
changed
slightly
to obtain
the
desired
magnitudes
of
peaks.
The
molarities
can
be
possibly
further
optimised.
The
discrimination
by
T7
DNA
polymerase
of
standard
unlabelled
ddNTPs,
present
in
equimolar
quantities
during
the
extension,
has
not
been
reported
previously.
As
in
any
method
using
only
one
lane
for
sequence
analysis,
the
accuracy
depends
critically
on
template
purity.
Further
parameters
of
importance
may
be
the
age
of
manganese
solutions
and
content
of
salts
in
buffers.
The
accuracy
is
highest
in
the
absence
of
secondary
peaks,
filllstops
or
high
noisy
background.
At
present
this
is
best
achieved
by
methods
for
purification
of
M13
ssDNA
templates
(3),
with
7deaza-dGTP
to
eliminate
compressions,
where
the
error
rate
is
below
1%.
Direct
sequencing
with
this
protocol
of
plasmid
or
cosmid
DNA,
where
the
background
may
often
be
quite
noisy,
would
result
in
higher
error
rate.
As
shown
(4,
5),
in
these
cases
the
four
lanes
method
gives
higher
accuracy,
since
it
is
possible
to
follow
each
base
lane
independently
and
discern
accurately
peaks
from
the
background.
However,
in
the
future
purification
methods
for
plasmids
and cosmids
will
be
further
improved
and
result
in
low
uniform
background,
makdng
their
sequencing
in
one
or
two
lanes
more
accurate.
The
use
of
only
one dye
may
be
significant
for
the
'walking
primer'
strategy.
At
present
we
use
this
method
for
mapping
of
sequences,
rapid
screening
of
clones
and
for
DNA
sequencing
in
thin
capillaries.
5.
Voss,H.
et
al.
(1990)
Nucd.
Acids
Res.
18,
1066.
6.
Figure
1
and
the
principal
of
the
methcod
was
presented
at
the
Genome
Sequencing
Workshop,
Wolf-Stradt,
October
1987.
REFERENCES
1.
Ansorge,W.
et
al.
(1989)
J.
Biochem.
Biophys.
Meth.
20,
47-52.
2.
Voss,H.
et
al.
(1989)
Meth.
Mol.
Cell.
Biol.
1,
155-159.
3.
Zimmennann,J.
et
al.
(1989)
Meth.
Mol.
Cell.
Biol.
1,
29-34.
4.
Zimmermann,J.
et
al.
(1990)
Nue.
Acids
Res.
18,
1067.
