volume 8 Number io 1980
Nucleic Acids Research
Synthesis of oligonucleotides on cellulose by a phosphotriester method
Roberto Crea and Thomas Horn
Department of Organic Chemistry, Genentech, Inc., 460 Point San Bruno Blvd., South San
Francisco, CA 94080, USA
Received 22 January 1980
ABSTRACT
The synthesis of oligothymidilic aoid3, (dT) m (where m = U, 7, 10, 13i
16, 19, 22, and 25), was carried out using a solid phase approach In combination with the modified phosphotriester methodology developed in solution.
Cellulose was used as the solid support after it3 functionalization with a
specially featured dlnucleoside diphosphate, 5'-0-p-chlorophenylphospho2'(3')_o-acetyluridilyl-[2'(3')-3']-5'-0-dimethoxytritylthymldine
p-chlorophenylester. The fully protected trideoxynucleoside triphosphate containing
only thymidine wa3 repeatedly used to elongate the oligonucleotide chain in
the 3'-5' direction. Individual coupling yields ranged from 45* to 75JThe total time needed to prepare (dTJjj was four days. Similarly, the
tridecanucleotide d(AGAAGGTACTTTT) was synthesized in good yield.
The
results show that this approach can be used for a fast and economic way to
synthesize oligodeoxynucleotides.
IHTRODUCTION
The synthesis in solution of oligodeoxynucleotides of defined sequence
by a modified phosphotriester method ha3 been shown to be very successful
and
its effectiveness
demonstrated
in the chemical
synthesis of
genes for human insulin . The use of protected trideoxynucleotides as
2 3
building block3 ' and the application of high performance liquid chromatography (h.p.l.c.) as a purification tool have dramatically increased the
speed by which the synthesis of DNA fragments, 12-15 bases long, can be
carried out.
However, the synthesis of the trinucleotide building blocks
on a large scale, required to maintain a complete "codon" library, is a
time consuming and costly process.
reduce
the amount
desirable.
of
trimer
used
Therefore a method which allows us to
in each
elongation
step
is highly
A way to reduce the losses of nucleotidic material and avoid
the time consuming purification of intermediates, would be the synthesis of
the
oligonucleotides
on
a
solid
polymer.
Several
approaches
to the
synthesis of DNA fragments on a solid support have been published (for
review
see t,5).
Although
a wide
range of polymers
© IRL Press Limited, 1 Falconborg Court, London W1V 5FG, U.K.
have
been
used,
2331
Nucleic Acids Research
chemical
with
synthesis of oligodeoxynucleotide3
the
method '
phO3phodiester
nucleotides
by
the
has been accomplished mainly
Attempts
phosphotrie3ter
method
to
synthesize
on
oligodeoxy-
3Olid-phase
have
been
Q
publishedd ,
and
We
reported.
recently
describe
oligodeoxynucleotides
on
method
a
here
the
cellulose
to
synthesis
using
d
synthesize
of
homo-
the
modified
18
methodology and trideoxynucleotides as building blocks.
'T^g
and
wa
*
hetero-
phosphotriester
RESULTS AND DISCUSSION
Cellulose was chosen as the 3olid support because of its inherent
polarity, its well known swelling properties in pola^ solvents (i.e. pyridlne) and
its mechanical
stability.
The polymer was functionalized by
reacting the cellulose with the dimer 51-0-p-chlorophenylphospho-2'(3')-0acetyluridilyl-[2'(3')-3' ]-5'-O-dlmethoxytritylthymidine p-chlorophenylester (Tb).
o
„
u
vy
^ r
HO
OH
1 \ r
CH cicrH i
?^±
HOV
. ,.,°
o/oVa a
u
k°y
THF.BSA
^
r
3
~
-
CNCH CH
i A|/>
a-ZoVo' so
———-
k^>
CH.CN, 1-Me-lmictazoJe
X
N
0
i
0
T
4
I 80% a
CNCH^O
c.-®-o'V
0
-
TPSTa.Py
2.
•
R=
EtjN: P y : H , o
Cf.CH,CH,
SCHEME I
Th» synthesis of (Jh_) is outlined in Scheme 1 and it3 features provide
the following advantages:
A) It can be synthesized from easily available starting materials.
3) Coupling of the ph03phodiester function to the cellulose with TPSTe
provide the attachment of (7b) to the polymer (see experimental part).
C) Deblocking of dimethoxytrityl group
2332
(DMT) from the deoxy- moiety in
Nucleic Acids Research
acidic medium affords the corresponding 5'-hydroxyl derivative for the
subsequent coupling step; therefore chain extension of the oligomer on
the cellulose occurs in the more favorable 3'—5" direction.
D) The release of the synthesized
oligonucleotide
from the polymer is
simple, fast, and essentially complete under the conditions used for
the removal of the ba3e-labile protecting groups of the oligonucleotidic chain.
The aqueous concentrated ammonia hydrolyses the acetyl
group of the uridine used as anchoring group
.
Subsequent nucleo-
1
philic attaok of the free 2'(3 )-hydroxyl group on the phosphate moiety
triggers the release of the oligonucleotidlc chain with a free 3' end
(scheme 2 ) .
7b
U- llrocyl
T" Thytnine
R" p-chloroph«ny1
TPSTe, Cellulose
TPSTe • Triitopropylbwzerntulfonylutrozoli
OMT • 4,4 '-dimtthoiytfityl
CE • 2-cyanoithyl
OMTO
I
2 % BenzenesuKonic Acid
T
T
T
2.
DMTO
nx
3.
I 9 I 9 I9
D hOPON LOPO, kOPO" , TPSTe
N OR
O B ^ OR
nnN OR
9
Ac 2 O/Py
DMT
ii ?ii
HO-kkOPOlkOH
V^ 6-/T
+
^
°*A ° i —
Px
OPOH Cellulose
-o- x o-l/ 6 -^
SCHEME 2
2333
Nucleic Acids Research
Uridine was protected at it3 2', 3'-diol group by treatment with trimethyl
orthoacetate
in
(BSA) as cataly3t.
THF
in
the
presence
of
The 2', 3'-cyclic derivative
benzenesulfonic
acid
(2) was phosphorylated
with the monofunctional phosphorylating agent 2-cyanoethyl-p-chlorophenyl2 14
pho3phorochloridate '
(^3_, 2 equivalents) in acetonitrile and 1-methylimidazole (1 equivalents).
aqueous acetic acid
isolated
by
The resulting compound (H) wa3 treated with 80S
(10 minutes) and the 2'(3')-0-acetyl derivatives (5_)
silica
gel
column
chromatography
as a mixture of isomers.
Finally, this material (5.) was reacted with (6), 5'-0-dimethoxytrityl-3'-0p-chlorophenylthymidilyl
phosphate,
(1.1
equivalents)
TPSTe (2.8 equivalents) a3 coupling agent.
in
pyridine
with
The dimer (2a) so obtained was
purified on 3ilica gel and precipitated from petroleum ether (32J yield).
The efficiency of the release mechanism was tested by treatment of (2a_, 10
rag) with concentrated ammonia.
material
(7a_) had
After 30 minutes at 50 °C, the
completely
degraded
5'-dimethoxytritylthymidine (TLC analysis)
into
a
base-line
starting
material
and
.
Coupling of the dinucleotide to cellulose was accomplished in the following way:
the 2-cyanoethyl group was selectively removed from the fully
protected dlmer (Ja, 0.2S mraole) by treatment with a solution of trlethylamine-pyridine-H 0
Under
these
(1:3:1
conditions
phosphodiester
v/v)
the
derivative
for
acetyl
25
minutes
group
was
(7b) so obtained
at
room
completely
temperature.
3table.
The
was dried with pyridine and
reacted with cellulose (Whatman CC31, 0.7 g) in anhydrous pyridine in the
P"es«mce of TPST? (O.o mmole).
followed
acidic
by
aomtorir.?
t-eatment
of
The attachment of (7_o_) to the polymer was
the release
of
DMT
"esidue
a sraall sample
of
polymer
J
at
194
nn
after
(33° experimental).
An
average incorpo-at'.on of ca. 100 umole dime"/? cellulose was obtained.
Elongation of the nucleotidic chain was carried out with the trideoxvnucleotide
(Q_) ; the synthesis of this compound
scribed .
The functionalized cellulose (J3) (100 mg, 9.7 umole of dimer)
was first treated with 21 benzenesulfonic acid
has previously
been de-
(BSA) in CH OH/CHC1
at
0° C in order to make the 5'-hydroxyl group of the thymidine available fothe chain extension.
After several coevaporations with pyridine, the dried
polvraer was reacted with (0) (50 |jmoles) in pyridine containing TPSTe (220
ptaole).
TPSTe
is known
to
be
a
very
powerful
condensing
therefore the coupling reaction was stopped after two hours.
agent '
;
The solid
support was collected by centrlfugatlon and the supernatant saved for the
recovery of unreacted trimer (£) (see below).
2334
The cellulose was washed
Nucleic Acids Research
with
pyridine
and
then
reacted 5'-hydroxyl
partially
groups.
treated with 2% BSA, and
trlmer {$).
four days.
acetylated
in order to block
The resulting cellulose was washed
then coupled
again with the
the un(MeOH),
same quantity of
Eight such cycles were repeated under the same conditions in
Table 1 lists the operations per coupling cycle.
After each cycle a simple (ca. 10 mg) of the polymer was treated with
cone. aq. NH
followed by 80% aq. AcOH to remove all protecting groups.
The unblocked oligonucleotidic material was then analyzed to calculate the
yields
per coupling.
For
this purpose
we used
two different
a) the release of DMT residue in acidic solution
a
semiquantitative
b) h.p.l.c.
estimate
analysis , which
of
the
amount
allowed
the
of
(494 nm)
trimer
accurate
(£)
methods:
, which gave
reacted, and
calculation
of
the
yields per coupling derived from the peak areas (254 nm) (figs. 1 and 2 ) .
Table 2 shows the yields based on h.p.l.c. analysis.
H.p.l.c. analysis of the products showed
were due to incomplete chain elongation.
that
the major
co-products
Small quantities of side product?
that could not be accounted for are probably due to direct reaction of the
trimer (£) with the hydroxyl groups of the cellulose or chain degradation
TABLE 1
STEP
OPERATION
CHEMICAL USED
1
Removal of DMT
2X BSA in CHCl /MeOH*
2
Washing
MeOH, Pyridine
3
Drying
Pyridine
A
Coupling
5
VOL.**
TIME
(min.)
10
10
50,20
5
10
10
Trimer, TPSTe, Py.
1
120
Washing
Pyridine
2
5
6
Acetylation
Ac.O/Pyridine
1
30
7
Washing
Pyridine
8
10
Methanol
30
10
* At 0 C ( i c e - w a t e r b a t h ) .
** M i l l i l i t e r s per 100 mg polymer.
2335
Nucleic Acids Research
TIME
(mtnuccs)
Fig.l. H.p.l.c. analysis of the oligothymidilic acids
(n=2), three (n=3), and four (n=i) coupling cycles.
during
deblocking.
functionalization
the
mechanical
A
method
to
block
reactive
obtained after two
hydroxyl
groups
after
(i.e. coupling of (Jb_) to the cellulose) without changing
properties
of
the
polyrasr
support
is
currently
under
h.p.l.c. on a
strong
investigation.
Purification
anionlc
exchange
corresponding
of
the
resin
to the
products
(Pennaphase
last eluted
unit at 25t nm) was concentrated
2336
was
AAX)
peak.
and
performed
by
by
collecting
The pooled
solution
the nucleotidlc
the
material
(ca. 1.0
material
O.D.
desalted
on
Nucleic Acids Research
16
18
16
Time (minutes)
Fig.2. H.p.l.c. analysis of dT
eight (n=8) coupling cycles.
Bio-gel
P-2.
The
phoretic analysis
size
obtained after five (n=5), six (n=6), and
of each compound
(fig. 3 ) .
wa3 confirmed
Two-dimensional
by gel electro-
sequence analysis of
(dT)
showed the expected pattern for this compound (results not shown).
The synthesis in solution of oligonucleotides of defined sequence using
the trinucleotide block approach
sible trlnucleotide triphosphates.
of trimers with different
required the use of mo3t of the 61 posIn order to investigate the reactivity
base composition, we synthesized
TABLE
2
Cycle
No.
Compound
Yield
%
2
(dT) 7
73
3
dT
70
< >,o
(dT) (3
62
6
<dT\6
<dT>|9
57
7
WT)
4
5
e
a
(dT)^
the following
60
46
48
2337
Nucleic Acids Research
oligonueleotide:
d(AGAAGGTACTTTT) (J_6) .
This oligoraer, 13-base3 long, wa3
selected because its synthesis requires the use of four trimer3, each having a different ba9e at their 3'-phosphate end (3ee scheme 3)Starting
from
the polymer
(1.0) (n=l, 120 mg), the synthesis of the
fully protected oligonueleotide derivative was accomplished
by sequential
addition of the th-ee trlmera (50 pinoles each): DMT-Tp A B p CBp (12),
o
o
o
—
SCHEME 3
10 ( n=l )
2% BSA I
HO-Tp Tp Tp Tp Ac Up -Cellulose
1) DMT-Tp o A B p o C B p o (12)
2) Ac O/Py
3) 2X BSA
( 40% )
Bz Bz
HO-Tp A p C p Tp
Tp Tp TpAc,,
Up -Cellulose
o
o
o ro o o o
o
Bz
DMT-A
c 2200/Py
/
2) AAc
3) 2 % BSA
iBu iBu
( 421 )
Bz iBu IBu
Bz Bz
HO-A p G p G Kp Tp A Kp CK p pTp Tp Tp TpAc,
Up -Cellulose
*o
o o o o
o o o o ^o K o
Bz iBu Bz
DMT-A p G p A p
Bt
iBu Bz
Bz
IBu iBu
Bz
(1_4)
( 451 )
Bz
Ac.,
DMT-A p G p A p A p G p G p T p A p C p T p TK p T p r T p
Upr - C e l l u l o s e
o
o *o
o
o
o o
o
o o
o o
o
o
(15)
1)
cone.
NH,
2)
80% AcOH
HO-ApGpApApGpGpTpApCpTpTpTpT-OH
(16)
2338
Nucleic Acids Research
Fig. 3. Gel electrophoretic
analysis on a 201 polyacrilatnide slab, after incubation
with [ T - 3 2 P ] ATP and polynucleotide ktnase.
dT 1 9 dT 2 2 dT 2 5
DMT-ABp
C^G18"
DMT-ABp CiB>UABp
V±) and
( U ) , where
p
p-chlorophenyl-
phosphate.
Tie trimers (_12), (1_3_), and (Vi) were prepared from the corresponding
fully
protected
derivatives
by
removal
of
the cyanoethyl
group from
the 3'-terminal phosphotriesters with Et N-Py-H 0 (1:3:1 v/v).
The three cycles were carried out under the same conditions as described in the synthesis of oligothymldilic acids.
synthesi3 of
(.16) and
Scheme 3 outlines the
the yields per coupling calculated
from h.p.l.c.
analysis (see fig. 5) are shown in parentheses.
The final polymer (1^, 85 mg) wa3 treated with base (cone. aq. NH )
and acid
(80* aq. AcOH), the cellulose residue was pelleted off and the
supernatant evaporated to dryness.
The residue, dissolved in t> aq. NH
(? ml), wa3 washed with ether and used for the isolation of (JJS).
accomplished
by repeated
the last eluted peak wa3 collected and pooled.
?2,
the product
acid3.
From
0.5
(1^6) was analyzed
ml
This was
runs on h.p.l.c; the material corresponding to
of
product (V5) was obtained.
cnjde
After desalting on Bio-Gel
as described
solution
ca.
10
for the
A__.,
oUgothymidilic
units
of
pure
Gel electrophoretic analysis showed that (16)
2339
Nucleic Acids Research
Fig. h.
Two-dimen-
sional analysis of a partial
venom phosphodiesterase digesc
of d(AGAAGGTACTTTT).Dimension 1,
electrophoresis on cellulose
acetate strip at pH 3.5, dimension 2, homochroraatography on
DEAE cellulose thin layer plate.
has the same mobility as (dT)
)
analysis
of the partial
(see fig. 5) and two-dimensional sequence
vanom phosphodiesterase digestt
l ft
confirmed
that
(16) had the expected sequence (fig. H ) .
The
recovery
of the excess
following procedures:
trimers was accomplished
by one of the
a) purification of the trimer 3'-phosphodiester by
17
on Sephadex LH-60 , or b) reuse of
gel exclusion column chromatog-aphy
the filtrate solution for further condensation either on polymer support or
in solution.
From the data obtained 30 far we calculated that 35-701 of
the excess trime^ could be recovered.
CONCLUSIONS
This study demonstrates that a solid phase approach combined with the
trinucleoti'le phosphotrlester methodology ha3 several advantages over the
synthesis
2340
of
oligonucleotides
in
solution .
Only
one
kind
of
Nucleic Acids Research
Fig.5. (a) Synthesis of
d(AGAAGGTACTTTT) monitored by
h.p.l.c. after the second, third,
and fourth coupling cycle.
(b) Gel electrophoretic analysis
of the product (P) purified by
h.p.l.c.
References are
and a dodecamer (Ref).
10
12
M
16
IB
20
Time tminultt)
trinucleotide building blocks (3'-phosphate) 19 necessary for the chain
elongation.
column
The synthesis is faster, because washing procedures substitute
chromatography
oligonucleotidic
on silica
material.
gel.
Because
This also
reduces
of the recovery
the loss of
of the unreacted
trinucleotide at the end of each coupling step, and the reduced scale of
synthesis on the cellulose, the amount of trimer used per coupling is
significantly
reduced.
Functionalization
of cellulose
with
a
ribo-
deoxvribo dinucleo3ide diphosphate containing uridine and one of the fou-
2341
Nucleic Acids Research
deoxyribonucleosides
provides
the
starting
polymer
necessary
for the
solid-phase synthesis of DNA fragments with different bases at the 3'-end.
In the synthesis of an oligonucleotlde 13 bases long with trinucleotides of different sequence, the yields per coupling were lower than the
ones
in the synthesis of homopolymer3
(dT) .
However, no base speci-
ficity has so far been observed.
MATERIALS AND GENERAL PROCEDURES
Analytical grade reagent3 were U3ed throughout thi3 synthetic study.
Pyridine was dried over KOH, distilled, and stored over molecular sieves
(type HA).
Thin layer chromatography
(TLC) was performed on silica gel
plates (Merck, Silica Gel 60 F254) with methanol/chloroform
solvent system.
(1:9 v/v) as
The presence of the DMT group wa3 monitored by 3praying
the TLC plate with 10J aq. ^ S O ^ and heating the plate to 60° C.
fication
by column
chromatography
was performed
on 3ilica
Puri-
gel (Merck,
Silica Gel 60H).
H.p.l.c. was performed with a Spectra-Physics SP 8000 Liquid Chromatograph on Permaphase AAX (du Pont) (50 cm x 0.1) cm I.D.) using a linear
gradient of water to 1M KC1, pH M.5, at a rate of 3* per minute.
The elu-
tion was performed at constant flow (3 ml/min.) at 60°C.
Cellulose (Whatman CC 3 D was washed with pyridine, rin3ed extensively
with
methanol,
and dried
in vacuo
over
p
A
2 °5"
^1
reactions
at the
polymer level were carried out under magnetic stirring.
The
fully
following
protected
the
benzenesulfonyltetrazole
dure
mono-
procedures
reported
in
the
and
trinucleotides
described
previously
were
.
synthesized
by
2,H,6-T^llsopropyl-
(TPSTe) wa3 synthesized according to the proceliterature
and
recrystallized
from
benzene-
petroleum ether.
2'(3')-0-Acetylurldlne-5'-0-2-eyanoethyl-p-chlorophenylphosphate
Uridine,
(200 ml)
Q),
utes
(61 mmole) was suspended in a solution of anhydrous THF
containing
zenesulfonlc
under
acid
trimethvl o r t h o a c e t a t e
(BSA)
exclusion
of
(0.6 g)
.
moi3ture,
After
the
showed complete conversion of u r i d i n e
(125 mmole) and anhydrous benstirring
solution
at
had
25° C for
become c l e a r .
methanol
saturated
(5 ml) and ths solvents removed under reduced p r e s s u r e .
2342
In CHC1
(300 ml)
ten minTLC
i n t o a new compound with higher R .
The reaccion was quenched by a d d i t i o n of
was dissolved
(5) .
and e x t r a c t e d
with NH
The oily
with saturated
residue
aq. NaHCO
Nucleic Acids Research
solution
(150
ml) and
water
(150 ml).
The
organic
layer
wa3
dried
(MgSO^) and evaporated to drynes3.
Without further purification, 2 1 ,3'-0-methoxyethylideneuridine (.2) (ca.
60 mmole) was
dissolved
1-N-methylimidazole
and
freshly
In anhydrous
(200 mmole).
acetonitrile
(200 ml)
containing
(ice-water)
2 14
2-cyanoethyl-p-chlorophenylphosphorochloridate '
prepared
The solution was chilled
(90 mmole) in CH CN (25 ml) was added dropwise over a 15 minute period.
The solution was then left to warm up to room temperature.
showed virtually complete phosphorylatlon of (2)under
(200
reduced
ml)
and
pressure, and
aq.
NaHCO
the
residual
(150 ml).
The
The solvent was removed
oil partitioned
CHC1
TLC analysis
between CHC1,
layer was washed
with
water (150 ml) and then concentrated to a small volume (50 ml). This solution was triturated
twice with pentane
(500 ml) to remove most of the
1-N-methylimidazole. The product obtained (^) was treated with 80t aqueous
acetic acid (200 ml) at room temperature for 10 minutes.
Finally, the ace-
tic acid was removed under reduced pressure by coevaporation with absolute
ethanol.
TLC of the glass residue showed only one product.
Purification
of (5_) was performed on a silica gel column (150 g ) . After washing with
CHC1, the pure product was eluted with a solution of MeOH in CHC1,
(1 to 4J v/v). The fractions containing the product were pooled and concentrated to a small volume.
(35°- 60° C) gave a white
Ft =0.31; UV (95* EtOH):
i
Anal:
C^H^N
Precipitation of (5) from petroleum ether
solid
X
max
(14.5 g; 50% yield
based on uridine).
258 nm ( E9,600), X . 226 nra ( E 1,800).
mln
O^PCl (529-5); calcd: C 45.32, H 3-96, Cl 6.70,
N 5-90, P 7.90; found: C 45.2, H 4.09, Cl 6.91, N 5.88, P 7.98.
5'-0-2-Cyanoethyl-p-chloropho;phophenyl-2'(3')-0-acetyluridllylC2'(3' )-3' ]-5'-0-dlmethoxytrltylthynildlne p-chloroptienyle3ter(7a).
5'-0-Dimethoxytritylthymidine-3'-0-2-cyanoethyl-p-chlorophenylphosphate
(7.0 mmole) was treated with a solution
of
triethylamlne/pyridine/water
(1:3:1 v/v) (50 ml) at room temperature for 25 minutes.
The phosphodiester
derivative (j>) so obtained was dried by coevaporation with anhydrous pyridlne (3 x 25 ml). 2'(31 )-0-Acetylu'-idine-5'-0-2-cyanoethyl-p-chlorophenylphosphate (5_) (6.6 mmole) was added to derivative (^) and the mixture d-ied
again with pyrldine (30 ml). TPSTe U0.5 mmole) in anhydrous pyidine (50
ml) was added to this solution.
The reaction mixture was kept in vacuo and
protected
16 hours.
from
the
light
for
The
reaction
was
stopped
addition of water, and the solvents removed under reduced pressure.
by
The
2343
Nucleic Acids Research
residual material
(100
ml), and
in CHC1
then
(250 ml) was extracted
washed
with
water
(100
with
aqueous
NaHCO
The
combined
CHC1
ml).
extracts were _e_vaporated to dryness.
Chromatography of the product on silica gel (75 g) was performed using
MeOH/CHCl
(1:99
v/v) as
product were evaporated
solvent
system.
to a glass.
The fractions containing
the
Precipitation of this material from
petroleum ether (35° - 60° C) gave a colorless solid (2.6 g, 32% yield).
Its homogeneity was confirmed by TLC analysis (R f =0.65).
Anal:
C
H^N 0
P 2 (1247); calcd: C 54.90, H 4.44, Cl 5.68,
N 5.61, P 4.96; found C 54.61, H 4.44, Cl 5.92, N 5-91, P 5.31.
To confirm the release mechanism, this compound (Ja, 10 mg) was treated
with cone, aqueous aramonia/pyridine (9:1 v/v; 2 ml) for 30 minutes at 60° C
TLC analysis showed a complete degradation of (Ja.) into a baseline material
and a compound containing the DMT group.
fied as 5'-DMT-thyraidine (TLC).
This latter compound was identi-
Removal of DMT with 2% BSA gave a compound
identical to thymidine.
FuncUonalizatlon of cellulose with (7b)
The fully protected dimer (7.3.) was treated with a solution of pyridinetriethvlamine-water (3:1:1 v/v) (20 nil) at room temperature.
After 25 min-
utes, TLC showed complete conversion of (J_a) into its phosphodiester derivative (base-line material (7b)).
oratlon with pyridine
(Whatman CC 3 D
TPSTe
(3 x 15 ml) and
then
reacted
with dry
cellulose
(0.7 g) in anhydrous pyridlne (3 ml) in the presence of
(0.23 g ) .
temperature.
The latter material was dried by coevap-
The mixture was
stirred
overnight
(16
hours) at
room
The polymer was then filtered (sintered glass filter M) and
washed with dry pyridine.
The filtrate was stored for the recovery of un-
reacted dime" (Jb). The cellulose was washed extensively with MeOH (60 ml)
and dried in a dessicator over KOH.
The extent of coupling of (£b) to the polymer was determined by treating a sample of cellulose (8, 10 mg) with 2% BSA in CHC1 -HeOH (7:3 v/v)
(2 ml) at 0°C for 10 minutes.
After centrifugation of the 3olid particles,
the ^bsorbancy of the released DMT group was measured at 494 nm.
dard curve was obtained by using different amounts of
thymidine In 2% BSA 3tock solution.
A stan-
5'-O-dimethoxyt^ityl
From the data obtained, we calculated
an average incorporation of ca 100 pmole3 dimer {Jb) per g cellulose.
Another sample (.&_, 10 mg) was treated with cone, aqueous ammonia for 8
hours at 60°C.
2344
The polymer was centrifuged off and the supernatant concen-
Nucleic Acids Research
trated to dryness.
The residual material was dissolved in CHC1
and analyzed by TLC.
thymidine.
CHC1
(1 ml)
The only product detected wa3 5'-O-dimethoxytrityl
Treatment of the chloroform
solution with
(1 ml) gave thymidine and DMT residue.
U% BSA in CH 0H-
From the total amount of
DMT (194 run) we calculated that the ammonia treatment had released ca 95$
of the 5'-0- DMT-thymidine from the polymer.
Synthesis of ollqothymidilic acids, (dT)m (m=4,7,10,13,16,19,22, and 25).
The functionalized polymer ((3, 100 mg; 9.7 pmole of dimer) was fir3t
treated with 2% BSA in CHC1 /MeOH (7:3 v/v, 10 ml) at 0°C for 10 minutes
a3 described above.
After washing the polymer on the filter with methanol
(50 ml) the cellulose was dried by flushing with anhydrous pyridine (20 ml)
The trimer (£) (50 u moles) was mixed with the polymer in a 10 ml roundbottom flask.
The mixture was dried by coevaporation with pyridine ( 2 x 5
ml) and finally resuspended in anhydrous pyridine (1.0 m l ) .
was added quickly and the reaction mixture kept in vacuo.
TPSTe (75 mg)
After two hours
of continuous stirring, pyridine (2 ml) wa3 added to the reaction vessel
and the polymer recovered by centrifugation.
The supernatant was collected
separately for the recovery of excess trimer (5.). The polymer so obtained
was resuspended in a solution of pyridine-acetic anhydride (10:1 v/v, i ml)
and stirred fo^ 30 minutes.
pyidine
After washing the cellulose on a filter with
(8 ml) and methanol (30 m l ) , the polymer was ready for the next
coupling step (see Table 1 ) .
The following cycle:
a) deblocking of DHT group from the polymer with
BSA, b) coupling of the polymer so obtained with trimer (9_) in pyridine
containing
TPSTe,
and
c)
acetylation
of
the
resulting
polymer
with
pyri.dine/Ac 0 was repeated seven more times under the same conditions.
Release of the ollaonucleotldes
The
polymer
(H), n=l-8,
r
rom the polymer and removal of protecting
10 mg) suspended
in pyridine
(0.5 ml) was
treated with concentrated aqueous ammonia (4 ml) at 60° C for 8 hours.
solid polymer was removed by centrifugation and discarded.
The
The supernatant
was evaporated to dryness, and the residue treated with 80? aq. acetic acid
at room temperature for 15 minutes.
The acetic acid was removed by evapor-
ation under reduced pressure and the solid residue redissolved in t% aqueous ammonia (2 m l ) .
tion was analyzed
After washing with ether ( 3 x 1 ml) the aqueous solu-
on h.p.l.c.
For each coupling step the chromatogram
2345
Nucleic Acids Research
showed the formation of a new product with higher retention time (3ee figs.
1 and 2 ) . The yields per coupling were calculated from the peak areas and
are based on the next shorter oligomer.
For each elongation step the final
product, after complete deblocking, was isolated by preparative h.p.l.c on
Permaphase AAX.
The fractions corresponding to the last peak were pooled
(0.5 - 1.0 O.D. units at 25*4 nm), the solution concentrated and desalted on
Bio-gel P-2.
Gel electrophoretic analysis on a 20$ acrylamide slab of the
oligomer3 (dT)
[y-
P]-ATP
released
[
(m=13,lb,19, 22, and 25) after their incubation with
and
T.
polynucleotide
kinase,
from the polymer had the expected
P] (dT)
partial
wa3 analyzed
venom
showed
that
each
3ize (fig. 3 ) .
by two dimensional
sequence
16
phosphodiesterase digestion products
.
fragment
Furthermore,
analysis
of its
The analysis con-
firmed the structure of this compound.
Synthesis of d(AGAAGGTACTTTT) (16).
The functionalized polymer {B_,
125 mg) was treated with 2$ BSA as des-
c-ibed above and used for the synthesis of the 13-base long oligonucleotide
of defined
sequence (1_6).
DMT-Tp o Tp o Tp o (9), DMT-Tp o A B p o C B p o (U),
trimers:
and DMT-A p G p A p
The
carried
This was accomplished by using/four dffferent
coupling
(_14), where p
of these
D M T - A ^ G ^ V p ^ (_13),
= p-chlorophenylphosphate.
triraers (50 umoles each) to the polymer was
out in the sequential order
shown
in scheme
3, under the same
conditions as reported above for the 3ynthe3is of oligothymidilates.
The
yields obtained for each coupling are shown in scheme 3 and were calculated
by h.p.l.c. analysis based on the next shorter fragment.
The fully protected cligomer obtained after the fourth cycle was released from the polymer and deblocked with ammonia and acetic acid as described above.
as
(dT)
The isolated product (_16) (h.p.l.c.) had the same mobility
on
two-dimensional
20$
polyacrylamide
slab
(see
sequence analysis of the partial
fig.
venom
5).
Furthermore,
phosphodiesterase
digest of (lj>) confirmed the expected sequence (fig. 4 ) .
Recovery of the exoes3 t r l n u c l e o t l d e s
A.
The
Chromatography on Sephadex LH-60.
filtrate
trinucleotide
pyridine
solution
(DMT-Tp Ap Cp , ca
(3
<10 umole)
ml)
containing
and
TPSTe was used
recovery of t h i 3 compound (\2) by gel exclusion chromatography.
2346
the
excess
for
the
Water (0.5
Nucleic Acids Research
ml) wa3 first added and the solution evaporated to dryness.
wa3 dissolved
Sephadex
in THF/MeOH
LH-fiO
(0.7
x
(95:5 v/v, 1 ml) and
100
cm)
applied
preequilibrated
in
to a column
the
same
Isocratic elution of the product was obtained with THF/MeOH
Fractions
of
(fraction
38 to
1
ml
were
50)
collected.
(detected
as
The
fractions
DMT-posltive
The residue
(95:5 v/v).
containing
baseline
of
solvent.
product
material) were
pooled, concentrated and the trimer precipitated from petroleum ether (25
mg, 35S recovery yield).
B.
Reuse of the excess trimer for synthesis in solution .
The
ymole)
pyrldlne
and
solution
TPSTe
was
containing
dried
by
the
excess
coevaporation
3'_0-Ani9oyl- 2-N-isobutrylguanosine
DMT-Ap o Ap Q Gp o
with
pyridine
(ca
(It mg, 29 umole) was added and the
mixture dried by coevaporation with pyridine (2x2 m l ) .
Additional TPSTe
(50 mg) was added, and the reactants dissolved in pyridine (1 m l ) .
reaction mixture wa3 kept
complete
conversion
iri vacuo
of the
for
nucleoside
latter compound was purified
MO
(2x2 m l ) .
6 hours.
TLC
analysis
into the product
on a silica gel column and
The
showed
tetramer.
This
obtained
as a
homogeneous solid after precipitation from petroleum ether (23 mg) .
C.
Reuse for solid-phase synthesis.
The pyidine solution (3 ml) recovered from a coupling step containing
the
t>-iraer DMT-Tp Tp Tp
(°_, ca
M0 uraole), was reused
oligonucleotidic chain on cellulose.
to elongate the
The trimer was added to the polymer
(10) (n=l, 100 mg) and the suspension dried by coevaporation with pyridine.
An additional portion of TPSTe (75 mg) was added and the reaction mixture
suspended
in pyridine (1 m l ) .
The fully protected ollgomer was released
from the polymer and deblocked with aq. ammonia and aq. acetic acid as
described
above.
H.p.l.c. analysis
showed
a coupling yield
of 43>
for
d(T) ? based on d t T ^ .
ACKNOWLEDGEMENTS
The authors are grateful to Dr. D.G. Kleid, Dr. P.M.J. Burgers and Dr.
J.H.
van Boom
for their valuable comments and
critical
reading
of the
manuscript.
REFEREHCES
Abbreviations:
DMT = M,1'-dimethoxytrityl, BSA = benzenesulfonic acid,
TPSTe = 2,1,6,-Triisopropylbenzenesulfonyltetrazole, CE = 2-cyanoethyl,
Bz = benzoyl, iBu = isobutryl, Ac = acetyl.
2347
Nucleic Acids Research
1.
2.
3.
U.
5.
6.
7.
8.
910.
11.
12.
1314.
15.
16.
17.
18.
a.
b.
2348
Crea, R., Hiro3e T. and Itakura, K. (1979) Tetrahedron Letters, 5,
395-398.
Crea, R., Kraszewski, A., Hirose, T. and I t a k u r a , K. (1978) Proc. N a t l .
Acad. S c i . ( U . S . A . ) , 75, 5765-5759.
Hiro3e, T., Crea, R. and I t a k u r a , K. (1978) Tetrahedron L e t t e r s , 28,
2449-2452.
Ko33el, H. and S e l i g e r , H. (1975) Fortsuch. Chem. Org. Natur3toffe 32,
297.
Gait, H.J. (1979) Polymer-Supported Reagents, Catalysts and Protecting
Groups (P. Hodge and D.C. Sherrington, Eds), Wiley, London, i n press.
Belagaje, R-, Brown, E.L., Gait, H . J . , Khorana, H.G. and N o r r i s , K.E.
(1979) J . 3 i o l . Chem. 254, 5754 and subsequent papers.
Narang, C.K., B r u n f e l d t , K. and N o r r i s , K.E. (1977) Tetrahedron Letters
2 1 , 1819-1822.
r.ait, M.J. and Sheppard, R.C. (1977) Nucleic Acids Res. 4, 1135-1158.
Potapov, V.K., Veiko, V.P., Koroleva, O.N. and Shabarova, Z.A. (1979)
Nucleic Acids Res. 6, 2041-2056.
Pless, R.C. and Letsinger, R.L. (1975) Nucleic Acid3 Research 2, 773S e l i g e r , H. (1975) Die Hacromoleculare Chemie 176, 1611-1627.
Miyosht, K. and I t a k u r a , K. (1979) Tetrahedron Letters 38, 3635-3638.
Stawinski, J . , Hozumi, T . , Narang, S.A., Bahl, C.P. and Wu, R. (1977)
Nucleic Acid3 Re3. 4, 353-371.
Koster, H. and Heyns, K. (1972) Tetrahedron L e t t e r s , 16, 1531-1534.
The three " r i b o - d e o x y r i b o " dinucleoside diphosphates made up by u r i d i n e
and
4-N-benzoylcytidlne,
6-N-benzoyladenosine,
and
2-N-isobutrylguanoslne, r s s p e c t i v e l y , were synthesized i n a s i m i l a r manner and
i s o l a t e d as homogeneous solid3 i n comparable y i e l d . The relevant data
w i l l be published elsewhere.
Griffin,
B.E.,
Jarraan,
H.,
Ree3e,
C.B.,
Sulstow,
J.E.
(1970)
Tetrahedron 26, 1023.
Crea, R., et a l , manuscript in p r e p a r a t i o n .
S e l i g e r , H. , Holupirek, M. and Gortz, H.H. (1973) Tetrahedron L e t t e r s ,
24, 2115-2118.
Jay, E., Bambara, R., Padmanabhan, P. and Wu, R. C1974) Nucleic Acids
Res. 1, 331-353-le R o o i j , J.F.M., Arentzsn, R., den Hartog, . I . A . J . , van der Marel, G.
and van Boom, J.H. (1979) J . Chromatography 171, 453-U59After t h i 3 paper was submitted f o r p u b l i c a t i o n methods f o r the
synthesis o f ollgodeoxynucleotides on s o l i d supports were reported
using a p h o s p h i t e - t r i e s t e r 1 " 8 and a phosphotriester 1 "* 5 approach,
respectively:
Matteucci, M.D. and Caruthera, M.H. (1980) Tetrahedron L e t t e r s ,
21, 719-722.
G a i t , M.J.,Singh, H. .Sheppard, R.C,Edge, M.D.,Green, A.R.,
H e a t h c l i f f e , G.R., Atkinson, T . C . , Newton, C.R., and Markham, A.F.
(1980) Nucleic Acids Research, 8, 1081-1096.
Nucleic Acids Research