Transcription inhibition using modified pentanucleotides

Article abstract of DOI:10.1016/S0968-0896(03)00071-3

Inhibition of gene expression was recently achieved by targeting the transcriptionally competent open complex using relatively short, pentameric modified oligonucleotides at ～60 μM. Corroborative affinity cleavage experiments using the copper complex of a phenanthroline conjugate provided the impetus to synthesize additional analogues containing substituents at the 2′-position of uridine in a derivative of 5′-GUGGA (-4 to +1), with the purpose of inhibiting transcription at lower concentrations. Conjugates of 5′-GUGGA modified at the 2′-position of uridine were convergently synthesized using a recently reported method. Seven analogues based upon the 5′-GUGGA scaffold were tested for their ability to inhibit transcription of the lac UV-5 operon. The conjugate containing a tethered pyrene showed 70% inhibition at 20 μM, and modest inhibition at as low as 5 μM. This is a significant improvement over previously tested pentanucleotides and provides direction for the preparation of a next generation of inhibitors.

Full text of DOI:10.1016/S0968-0896(03)00071-3

Bioorganic & Medicinal Chemistry 11 (2003) 2321–2328
Transcription Inhibition Using Modified Pentanucleotides
Jae-Taeg Hwang,a Francis E. Baltasar,b Daniel L. Cole,b David S. Sigman,,b
Chi-hong B. Chenb,* and Marc M. Greenbergc,*
aDepartment of Chemistry, Colorado State University, Fort Collins, CO 80523, USA
bDepartment of Biological Chemistry, School of Medicine, Department of Chemistry and Biochemistry,
Molecular Biology Institute, University of California at Los Angeles, Los Angeles, CA 90095-1570, USA
cDepartment of Chemistry, Johns Hopkins University, 3400 N. Charles St., Baltimore, MD 21218, USA
Received 8 December 2002; revised 20 January 2003; accepted 30 January 2003
Abstract—Inhibition of gene expression was recently achieved by targeting the transcriptionally competent open complex using
relatively short, pentameric modified oligonucleotides at ꢀ60 mM. Corroborative affinity cleavage experiments using the copper
complex of a phenanthroline conjugate provided the impetus to synthesize additional analogues containing substituents at the
20-position of uridine in a derivative of 50-GUGGA (ꢁ4 to +1), with the purpose of inhibiting transcription at lower concentra-
tions. Conjugates of 50-GUGGA modified at the 20-position of uridine were convergently synthesized using a recently reported
method. Seven analogues based upon the 50-GUGGA scaffold were tested for their ability to inhibit transcription of the lacUV-5
operon. The conjugate containing a tethered pyrene showed 70% inhibition at 20 mM, and modest inhibition at as low as 5 mM.
This is a significant improvement over previously tested pentanucleotides and provides direction for the preparation of a next
generation of inhibitors.
# 2003 Elsevier Science Ltd. All rights reserved.
Inhibitors targeted for gene expression have received
considerable attention in the past decades for their
potential as therapeutic agents. Strategies available for
inhibiting gene expression include either targeting
mRNA (antisense) using oligonucleotides or double
stranded siRNA, or targeting DNA (antigene), such as
triple helix formation.1ꢁ7 Specificity is often achieved by
designing molecules that form specific hydrogen bonds
with the donors and acceptors presented by the nucleo-
bases of the target. Modified nucleic acids that enhance
binding and other properties are often used as tools in
these endeavors. Typically, nucleic acid based inhibitors
need to be approximately 15 nucleotides long in order
to ensure high sequence selectivity. Oligonucleotides of
this length can be difficult to transport across cell mem-
branes and are expensive to synthesize on a large scale.
complementary to the template strand of the open
complex in the lac UV5 and trp EDCBA systems at
ꢀ60 mM (Fig. 1). These oligonucleotides contained
20-O-methyl groups to enhance stability and 30-deoxy-
adenosine at the 30-termini to prevent their elongation.
Gel retardation assays showed that transcription inhib-
ition was not due to the dissociation of RNA poly-
merase. The specificity of these pentanucleotides as
inhibitors was demonstrated by the fact that molecules
targeted for lac UV5 and trp EDCBA systems were only
successful for the corresponding promoters, where the
transcription start sites differ from each other by only two
bases. Additional studies established that pentanucleotide
Our laboratory discovered a unique approach for the
design of gene specific transcription inhibition by utiliz-
ing the formation of an open complex during transcrip-
tion initiation (‘bubble formation’).8,9 Significant
inhibition can be achieved by pentanucleotides that are
Figure 1. Hybridization of 50-GUGGA (20-O-methylated, 30-deoxy-
adenosine, 2) (ꢁ4 to +1) to lac UV5 open complex. This drawing does
not depict actual formation of the open complex.
*Corresponding author. Tel.:+1-410-516-8095; fax:+1-410-516-8420;
e-mail: mgreenberg@jhu.edu
,Deceased
0968-0896/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0968-0896(03)00071-3

2322
J.-T. Hwang et al. / Bioorg. Med. Chem. 11 (2003) 2321–2328
inhibitors were optimal in length, but there was some
latitude with position relative to the transcription start
site (Fig. 1). Verification of binding site and orientation
was obtained using oligonucleotides (1) containing
covalently bound mono- and bis-1,10-phenanthroline-
copper chelates [(OP)2Cu+], which cleaved the template
strand inside the open complex at positions just upstream
from the transcription start site. Phenanthroline modifi-
cation also indicated that the 20-position of the uridine
was amenable to substitution with components that might
further stabilize its binding to the open complex.
synthesis and incorporation of only one modified phos-
phoramidite, or use of a photolabile solid phase sup-
port. Efficient use of the oligonucleotide is achieved by
using a single solid-phase synthesis to prepare multiple
conjugates. Homogeneous oligonucleotide conjugates
are obtained using mild conditions, and a modest excess
of reagents relative to oligonucleotide substrate.
Although conjugates can be prepared while the oligo-
nucleotide is in solution or bound to its solid phase
synthesis support, the latter approach is well suited for
preparing analogues of 1 (Scheme 1). Furthermore, we
previously utilized 4 to synthesize oligodeoxynucleo-
tide conjugates functionalized at the 20-position of
uridine.10
In this study, we wish to report our efforts to discover a
more potent transcription inhibitor by improving the
binding affinity of this pentanucleotide through altering
its potential intercalative ability. We synthesized seven
oligonucleotides based on 1 and 2 with a 20-modified
uridine.9 The substituents varied from alkyl amine to
polyaromatics, such as pyrene. The inhibitory effects of
these oligonucleotides were assayed, and their abilities
to inhibit were compared.
The pentameric scaffold containing 4 was prepared as
previously described with two exceptions.8ꢁ10 The
50-termini of the inhibitors contained hydroxyl groups
instead of phosphates. In addition, the pentanucleotide
inhibitors require 30-deoxyadenosine nucleotides at their
30-termini. Solid-phase support containing N-benzoyl
protected 30-deoxyadenosine linked via its 20-hydroxyl
group is commercially available. However, N-isobutyryl
protected adenines are preferred in the conjugation
method due to incompatibility of N-benzoyl groups
with the o-nitrobenzyl photochemistry, and possible
inopportune transamidation of phenoxyacetyl protected
nucleotides.13,14 Consequently, solid phase support
containing N-isobutyryl-30-deoxy-20-succinatoadenosine
was prepared from 30-deoxyadenosine.15 Execution of
the standard approach for converting the free nucleo-
side into needed solid-phase support was hampered by
the unexpected lability of the N-isobutyryl group
(Scheme 2). Regardless of whether the hydroxyl groups
were silylated or acylated, we were unable to remove
them without cleaving significant amounts of the iso-
butyryl amide. After extensive experimentation we
moved forward using a low yielding procedure for the
preparation of the exocyclic amine protected nucleoside
based on the knowledge that very small amounts of the
50-dimethoxytrityl 30-succinate (9) were needed for
loading the controlled pore glass (CPG) support.
Results and Discussion
Design and synthesis ofinhibitor candidates
The ability of 1 to bind and cleave the target strand of
the open complex provided the impetus to design mol-
ecules that might have higher affinity for the complex by
intercalation into the duplex structure formed between
the inhibitor and DNA template. Pentanucleotide 1 was
prepared via solid-phase synthesis, and required the cor-
responding phosphoramidite (3), which was obtained
from 20-amino-20-deoxyuridine. Synthesis of 1 was time
consuming and a more efficient process for synthesizing
oligonucleotide conjugates based on 1 was desired. We
utilized a convergent method for bioconjugate synthesis
recently reported in which a single functional group is
selectively revealed in an otherwise protected oligonu-
cleotide (Scheme 1).10ꢁ12 This strategy necessitates the

J.-T. Hwang et al. / Bioorg. Med. Chem. 11 (2003) 2321–2328
2323
Scheme 1.
Scheme 2. (a) (i) TMSCI, pyridine, then isobutyryl chloride; (ii) NH4F, MeOH (8%); (b) DMTCl, pyridine (86%); (c) succhinic anhydride, DMAP,
pyridine (65%); (d) PyBOP, Hunig’s base, DMAP, CH3CN, LCAA–CPG.
A number of points regarding the synthesis of oligo-
nucleotides and their conjugates on 5 deserve mention.
The photoprotected scaffold (6) was prepared using
modified ABI 394 RNA synthesis cycles in which the
20-O-methylguanosine phosphoramidite was coupled for
5 min and photolabile phosphoramidite (4) was double-
coupled (5 min/cycle). We also found that 4,5-dicyano-
imidazole was a more effective activating agent than
either tetrazole or 5-ethylthiotetrazole.16,17 Trimethyl-
acetic anhydride was used for capping in place of acetic
anhydride in order to prevent transacylation upon
photochemical unmasking of the primary amine.10 We

2324
J.-T. Hwang et al. / Bioorg. Med. Chem. 11 (2003) 2321–2328
Table 1. Transcription inhibitor candidates
also found that the photolysis time previously used to
unmask 4 after incorporation into oligonucleotides had
to be extended to three 40-min cycles. It is unclear what
the source of this perturbation was, but previous reports
suggest the efficiency of the o-nitrobenzyl redox reaction
increases when the photochemical substrate is further
removed from the surface of the solid-phase synthesis
support.18
The desired conjugates (8) were obtained via either direct
condensation of the support bound 50-dimethoxytritylated
oligonucleotide (7) with the appropriate aromatic isocy-
anate (8c–e) or PyBOP mediated coupling of a carboxylic
acid (8a, 8b). The proximity of the nucleophilic substrate
to the surface of the solid-phase support also affected
conjugation efficiency necessitating reacting 25–50 equiv
of the electrophiles for 3 h at either room temperature or
55 ꢂC. The amide and urea products, as well as the ori-
ginal carbamate in 8f were stable to the 6 h, 55 ꢂC con-
centrated aqueous ammonia treatment. The latter was
cleaved upon deprotection for 24 h. Short oligonucleo-
tide length, 20-O-methyl groups, and nonpolar con-
jugants that decreased water solubility made the
purification of inhibitor candidates by denaturing gel
electrophoresis difficult. Consequently, 8a–f were puri-
fied by anion exchange chromatography and obtained
in modest yield compared to those typically prepared
using this method (Table 1).10ꢁ12 All seven modified
oligonucleotides were characterized by ESI-MS.19
Compd
8a
X
% Yield
conjugation
HPLC ret.
time (min)a
60
65
29.6 (B)
36.1 (A)
8b
8c
45
28.3 (B)
8d
8e
58
40
29.2 (A)
44.9 (A)
Transcription inhibition by oligonucleotide conjugates
The relative binding affinities and inhibitory abilities of
the seven conjugates were conveniently screened by
carrying out transcription assays at varying inhibitor
concentrations in the lac UV5 system. Inhibitor and
template were incubated in transcription buffer with
RNA polymerase and ribonucleotide triphosphates.
a-[32P]-UTP was used in this transcription mixture for the
purposes of visualization and quantification of transcript.
Transcription was quenched after 1 min. Samples were
analyzed on a 10% denaturing polyacrylamide gel and
visualized by phosphorimaging. The amount of transcript
produced in the presence of each inhibitor was then
quantified and compared using 2 as a benchmark (Fig. 2).
8f
—
—
29. (A), 21.0 (B)
17.4 (A), 16.3 (B)
8g
H
aHPLC solvent system in parentheses. See Experimental for details.
Using 1 as a guide, we anticipated that only modified
oligonucleotides containing substituents capable of
intercalation would enhance inhibition. Oligonucleotide
conjugate 8g, which bears no intercalative substituent,
showed no inhibitory capability at 50 mM(data not
shown). Evidently, the positive charge does not improve
the stability of the complex. The other six conjugates
contain one or more aromatic rings tethered to the 20-O-
1-(4-aminobutyl) chain of uridine. The conjugates con-
taining aromatic groups are divided into two families,
depending upon whether or not the aromatic portion of
the molecule is separated from the electrophilic point of
attachment (carbonyl group) by a short alkyl chain, as it
is in 8a and 8b. There was a direct correlation between
inhibition and size of the aromatic system amongst the
conjugates that did not contain an additional chain. The
pyrene conjugate 8c demonstrated the most significant
Figure 2. Transcription inhibition of lac UV5 by oligonucleotide con-
jugates 8a–8g relative to 2 at 20 mM.

J.-T. Hwang et al. / Bioorg. Med. Chem. 11 (2003) 2321–2328
2325
inhibitory ability (Fig. 2). Approximately 70% inhibi-
tion was achieved at 20 mM. Further transcription
assays performed on 8c showed that inhibition occurred
even at 5 mM (Fig. 3). The other pyrene containing
conjugate (8a) was considerably less effective. Con-
jugates 8a and 8c differ in the length of the tether
between the aromatic system and the 20-oxygen of the
uridine in the scaffold and the chemical nature of the
linkage. Although a greater number of compounds are
necessary to substantiate this proposal, we believe that
the former is more significant. This suggestion is based
partly on the observation that 8b, which contains the
often used acridine intercalator, was a poor inhibitor.
One should also note that the proposed tether length
effect is consistent with the observation that 8a is a
stronger inhibitor than 8b.
Gel mobility assay
Figure 4. Gel mobility assay of lac UV5 open complex. Arrow denotes
position of shifted bands. Lane 1: labeled template only; lane 2:
labeled template+RNA polymerase; lane 3: labeled template+RNA
polymerase+8c; lane 4: template+RNA polymerase+labeled 8c; lane
5: RNA polymerase+labeled 8c; lane 6: labeled 8c.
To characterize further the binding of inhibitor 8c to the
open complex, we performed gel mobility assays (Fig.
4). Inhibitor 8c (50 mM) was incubated with lac UV5
template under conditions suitable for the formation of
the open complex, which occurs upon addition of RNA
polymerase. Samples were then loaded on a 5% non-
denaturing acrylamide gel for analysis. The single band
in lane 1 represents movement of template by itself. In
lane 2, labeled template was incubated with RNA poly-
merase prior to loading. The presence of a slower mov-
ing band indicates the formation of the open complex
(Fig. 4, see arrow). Retarded mobility was due to the
binding of RNA polymerase to the template. This
slower band is still present in lane 3, when the experi-
ment was carried out in the presence of oligonucleotide,
showing that open complex formation was retained, and
inhibition was not due to the dissociation of RNA
polymerase. We also observed this slower moving band
in lane 4, which contained unlabeled template and
labeled inhibitor. This signified that 8c hybridized to the
template strand within the open complex. Lanes 5 and 6
show movement of oligonucleotide conjugate with and
without RNA polymerase, respectively. The shifted
band of interest does not appear in either lane, proving
that the retardation in lane 4 was not due to the binding
of RNA polymerase directly to inhibitor. Bands
appearing at the very top of lanes 3–6 represent sample
that has remained in the well. This could be due to either
excess radiolabeled sample or non-specific protein bind-
ing. The large amount of radiation evident at the bottom
of lanes 4–6 is due to unincorporated g-32P-ATP. The
short length of the oligonucleotide conjugates prohibited
removal of the excess radionuclide via the typical
Sephadex and/or precipitation treatments.
To test whether or not specificity of 8c was affected by
its increased binding ability, we performed gel mobility
assays using the trp EDCBA system under the same
conditions (data not shown). No retarded band corre-
sponding to that observed in the lac UV5 system was
found. This study demonstrated that the specificity of 8c
for the lac UV5 promoter was retained.
Summary
Using 1 and 2 as a starting point, we convergently
synthesized oligonucleotide conjugates that inhibit in
vitro transcription. The small library of compounds
synthesized provided a molecule that inhibits tran-
scription 3- to 4-fold more strongly than 2. Moreover,
these molecules provide direction for synthesizing a next
generation of inhibitors. The next generation of mol-
ecules should include conjugates of polycyclic aromatic
systems and substituents on these rings that enhance
intercalation. Further refinements in the effects of tether
length are also warranted.
Figure 3. Comparison of transcription inhibition of lac UV5 by 2 and
8c at various concentrations. Lane 1: 10-base DNA marker (Gibco/
BRL); lane 2: no template; lane 3: no inhibitor; lane 4: 2 (50 mM); lane
5: 2 (20 mM); lane 6: 2 (10 mM); lane 7: 8c (20 mM); lane 8: 8c (10 mM);
lane 9: 8c (5 mM).
Experimental
All reactions were carried out in oven-dried glassware
under an atmosphere of argon or nitrogen unless

2326
J.-T. Hwang et al. / Bioorg. Med. Chem. 11 (2003) 2321–2328
otherwise noted. Diisopropylethylamine, DMF, tri-
methylchlorosilane, and pyridine were distilled from
CaH2. Acetonitrile was passed through CuSO4 and then
distilled from CaH2.
mL) and precipitated into rapidly stirred hexanes (30
mL). The mixture was centrifuged and the supernatant
was decanted to leave 9 as a white powder (45 mg,
1
65%). H NMR (CDCl3) d 9.20 (bs, 1H), 8.62 (s, 1H),
8.22 (s, 1H), 7.42–7.05 (m, 9H), 6.78 (d, 4H, J=8.4 Hz),
6.18 (s, 1H), 5.78 (d, 1H, J=3.6 Hz), 4.60 (m, 1H), 3.68
(s, 6H), 3.40 (m, 2H), 3.10 (m, 1H), 3.04 (q, 3H, J=8.4
Hz), 2.64 (m, 5H), 2.22 (dd, 1H, J=4.0 and 12 Hz), 1.25
(t, 6H, J=7.2 Hz); 13C NMR (CDCl3) d 176.4, 172.3,
158.6, 152.6 150.8, 149.5, 144.6, 142.0, 135.8, 135.7,
130.1, 128.2, 127.9, 127.0, 122.4, 113.2, 90.1, 86.5, 80.6,
78.1, 64.5, 55.4, 45.5, 36.2, 33.1, 30.1, 29.9, 19.5, 8.8; IR
(film) 3245, 3197, 2968, 2932, 2874, 2837, 2597, 2553,
2492, 1734, 1718, 1701, 1684, 1608, 1584, 1508, 1457,
1250, 1221, 1177, 1084, 1034, 830, 735 cmꢁ1; HRMS
(FAB); calcd for C39H41N5O9, 724.2983 (M++H),
found 724.2970.
N-Isobutyryl-50 -O-dimethoxytrityl-30 -deoxy-20 -succina-
toa-denosine (9). 30-Deoxyadenosine15 (0.4 g, 1.6 mmol)
was dried three times by azeotroping pyridine (3ꢃ10
mL) under vaccuum. Trimethylchlorosilane (1.1 mL, 8
mmol) was added to a solution of the dried nucleoside
in pyridine (15 mL) at 0 ꢂC, and then the reaction was
warmed to room temperature. After the reaction mix-
ture was stirred for 2 h, isobutyryl chloride (0.16 g, 1.5
mmol) in CH2Cl2 (5 mL) was added slowly over 30 min
at room temperature. After stirring for 3 h the solvent
was removed in vacuo. The crude product (0.18 mg,
ꢀ0.45 mmol) was treated with NH4F (86 mg, 2.3 mmol)
in CH3OH (10 mL) for 2 h. After removing the solvent
in vacuo, the crude material was chromatographed on
silica gel (CH2Cl2/CH3OH, 10:1) to yield impure com-
pound. The impure material was filtered through dried
neutral alumina with CHCl3, followed by CH3OH to
Preparation of N-isobutyryl-50-O-dimethoxytrityl-30-deoxy-
20-succinatoadenosine support (5). LCAA-CPG (0.1 g),
PyBOP (2.9 mg, 5.5 mmol), diisopropylethylamine (2
mL, 12 mmol), DMAP (0.7 mg, 5.7 mmol) and 9 (4 mg,
5.4 mmol) were combined in a screw capped vial with
acetonitrile (2 mL). The reaction mixture was shaken at
room temperature for 20 min. The resin was filtered,
washed with MeOH (10 mL), CH2Cl2 (10 mL), and then
dried under high vacuum. The amount of nucleoside
loading on the support (ꢀ40 mmol/g) was determined
by trityl analysis. Unreacted alkylamines were capped
by reacting the resin with trimethylacetyl chloride (21
mL, 113 mmol), DMAP (6.7 mg, 55 mmol), and pyridine
(18 mL, 220 mmol) in acetonitrile (2 mL) in a screw cap-
ped vial. The reaction mixture was shaken at room
temperature for 2 h. The resin was filtered, washed with
MeOH (10 mL), followed by CH2Cl2 (10 mL), and dried
under high vacuum.
1
give a pure product (40 mg, 8%). H NMR (CDCl3) d
9.65 (bs, 1H), 8.45 (s, 1H), 8.30 (s, 1H), 5.85 (d, 1H,
J=2.8 Hz), 4.96 (m, 1H), 4.68 (m, 1H), 3.96 (d, 1H,
J=8.8 Hz), 3.84 (dd, 1H, J=2.2 and 8.8 Hz), 3.02 (m,
1H), 2.45 (m, 1H), 2.20 (m, 1H), 1.25 (d, 6H, J=7.2
Hz).
Dimethoxytrityl chloride (51 mg, 0.15 mmol) was added
to a solution of N-isobutyryl-30-deoxyadenosine (40 mg,
0.12 mmol) in pyridine (5 mL). The reaction mixture
was stirred overnight at room temperature, and then
evaporated to give a crude product, which was purified
on silica gel (CH2Cl2 to CH2Cl2/CH3OH, 10:1, con-
taining 1% Et3N) to yield a pale brownish foam (64 mg,
1
86%). H NMR (CDCl3) d 9.73 (s, 1H), 8.65 (d, 1H,
J=3.3 Hz), 8.32 (s, 1H), 7.42–7.20 (m, 9H), 6.82 (d, 4H,
J=8.7 Hz), 6.05 (d, 1H, J=2.4 Hz), 5.45 (bs, 1H), 4.74
(bs, 1H), 3.81 (s, 6H), 3.44 (dd, 1H, J=2.7 ad 10.2 Hz),
3.40-3.15 (m, 2H), 2.35 (m, 1H), 2.24 (m, 1H), 1.38 (d,
6H, J=6.9 Hz); 13C NMR (CDCl3) d 176.1,158.6,
152.3, 150.5, 149.8, 149.4, 144.6, 141.1, 136.2, 135.8,
135.7, 130.1, 128.9, 128.1, 128.0, 127.0, 123.9, 122.7,
113.3, 93.2, 86.6, 80.6, 76.2, 65.0, 55.4, 36.4, 34.4, 19.4;
IR (film) 3270, 2924, 2836, 2245, 1725, 1608, 1583, 1510,
Photolytic deprotection
Pentanucleotide bound to support (30 mg, ꢀ0.9 mmol
oligonucleotide based upon trityl response) was added
to a Pyrex tube containing a stir bar constructed from a
standard (white) pipe cleaner and CH3CN (20 mL). The
tube was fitted with a rubber septum and the solution
was sparged with Ar for 20 min, after which the needle
was raised well above the surface of the solvent. Photo-
lyses were carried out with a VWR Chromato-Vue
transilluminator (lmax=365 nm) for 40 min. The resin
was filtered, washed with MeOH (10 mL), followed by
CH2Cl2 (10 mL), collected, and dried under high
vacuum. The resin was transferred to a tube in CH3CN
(20 mL) whereupon the sparging and photolysis proce-
dures were repeated two more times. The resin was fil-
tered, washed with MeOH (10 mL), followed by CH2Cl2
(10 mL), dried under vacuum, and placed in a screw
capped vial for storage. Note: It is important to maintain
the temperature during photolysis at ꢄ25 ꢂC using a fan.
1462, 1251, 1178, 1084 cmꢁ1
.
To a solution of N-Isobutyryl-50-O-dimethoxytrityl-30-
deoxyadenosine (60 mg, 0.1 mmol) in pyridine (5 mL)
was added succinic anhydride (15 mg, 0.15 mmol) and
DMAP (18 mg, 0.15 mmol). The reaction was stirred
overnight, after which additional succinic anhydride (20
mg, 0.3 mmol) was added. After stirring for 3 days, the
pyridine was removed, the residue was coevaporated
with toluene (3ꢃ15 mL). The residue was dissolved in
CH2Cl2 and washed with 5% Na2HPO4. The aqueous
layer was extracted with CH2Cl2 (3ꢃ30 mL). The com-
bined organic layers were washed with brine (20 mL)
and dried over Na2SO4. The solvent was removed to
give a crude product as a foam, which was purified on
silica gel (CH2Cl2/MeOH, 10:0.1 to 10:1, containing 1%
Et3N) to yield a foam which was dissolved in CH2Cl2 (2
General procedure for conjugation of protected resin
bound pentanucleotides
When conjugating to a carboxylic acid a solution
(0.15 M) of the coupling reagents [39 mg PyBOP, 16 mL

J.-T. Hwang et al. / Bioorg. Med. Chem. 11 (2003) 2321–2328
2327
diisopropylethylamine (2 molar equivalents)] in DMF
(500 mL) was prepared in an oven dried 1 dram vial
equipped with a septum. A solution (0.15 M) of car-
boxylic acid was prepared in a second oven dried vial.
The resin bound DNA (10 mg, containing ꢀ300 nmol
of oligonucleotide, based on trityl response) was treated
with 100 mL (25 molar equivalents) of a 1:1 mixture (by
volume) of the PyBOP and carboxylic acid solutions.
The reaction was capped and shaken at room tempera-
ture for 3 h. The resin was washed with CH3CN (3ꢃ2
mL), and dried in vacuo. Isocyanates were conjugated
using a DMF solution of isocyanate (0.15 M) prepared
in an oven dried vial. The resin bound DNA (10 mg,
containing ꢀ300 nmol of DNA, based on trityl
response) was treated with 100 mL (50 molar equiva-
lents) of isocyanate solutions. The reaction was capped
and shaken at 55 ꢂC for 3 h. The resin was washed with
CH3CN (3ꢃ2 mL), and dried in vacuo.
jected to PvuII restriction (GibcoBRL) and purified on
an 8% non-denaturing polyacrylamide gel.
In vitro transcription assay
The transcription mixture (10 mL) contained unlabeled
lac UV5 (EcoRI fragment, 0.1 mM) and 0.3 units/ mL
Escherichia coli RNA polymerase (Sigma) in a buffer of
40 mMTris–HCl (pH 7.9), 50 mMKCl, 10 mMMgCl
2,
0.1 mMDTT, 100 mg/mL BSA, and 5% glycerol. Oli-
gonucleotides were added to give the proper inhibitor
concentrations. No inhibitors were added to control
samples. The reaction mixture was incubated at 37 ꢂC
for 20 min. To initiate transcription, ribonucleotides
(final concentration 100 mMeach, with 1 mCi of a-[32P]-
UTP) were then added. The reaction proceeded for 1
min and was quenched with 10 mL of formamide and
dyes. Samples were analyzed on a 10% denaturing
polyacrylamide gel. The gel was dried under vacuum at
80 ꢂC. Transcription product was visualized and quan-
tified by phosphorimaging.
In either instance, detritylation was effected by trans-
ferring the resin to a standard oligonucleotide synthesis
column and passing the standard trichloroacetic acid
solution (3ꢃ20 s) through the column, followed by
CH3CN, and drying under vacuum. The free flowing
resin was treated with 28% aqueous ammonia (1 mL)
for 6 h at 55 ꢂC and concentrated under vacuum. Puri-
fication of conjugated pentaribonucleotide was carried
out using anion exchange HPLC column (Vydac
301VHP575). Solvent system A: A, 10 mMTris, pH 8.0,
10% CH3CN; B, 10 mMTris, 0.5 MNH 4Cl, pH 8.0,
10% CH3CN; 0–40% B linearly over 25 min; 40–80% B
linearly over 3 min; hold 80% B for 10 min; flow rate, 1
mL/min. Solvent system B: A, 10 mMTris, pH 8.0,
30% CH3CN; B, 10 mMTris, 0.5 MNH 4Cl, pH 8.0,
30% CH3CN; 0–40% B linearly over 25 min; 40–80% B
linearly over 3 min; 80–90% B linearly over 10 min.
Solvent system A was used for 3-nitrophenyl isocyanate,
9H-fluoren-2-yl isocyanate, and acridine pentanoic acid
conjugated oligonucleotides. Solvent system B was used
for pyrenebutyryl conjugated products. Isolated yields
were obtained (O.D. 264 nm; 352 nm for pyrenebutyryl
conjugated products) by comparing the amount of con-
jugated oligonucleotide to the amount of starting mat-
erial, as previously described.10ꢁ12
Gel mobility assay
Either oligonucleotide inhibitor or template was radio-
labeled, depending on the gel mobility assay. Radio-
labeled inhibitor was prepared in a kinase buffer with 2
nmol of dephosphorylated oligonucleotide, g-[32P]-ATP
(20 mCi/nmol), and T4 polynucleotide kinase (Gibco-
BRL). The reaction was quenched upon heat deactivation
of the enzyme at 80 ꢂC for 2 min, and radiolabeled inhi-
bitor was then used directly for gel mobility studies.
Inhibitor (50 mM) was incubated with lac UV5 tem-
plate (40 nmol) and RNA polymerase (0.2 units/mL,
USB) in transcription buffer for 25 min at 37 ꢂC. No
RNA polymerase was added to control samples. After
incubation, samples were subjected to heparin chal-
lenge (200 mg/mL) for 2 min at 37 ꢂC and immediately
loaded on 5% non-denaturing polyacrylamide gel run
at 200 volts in the cold (4 ꢂC). The gel was then dried
under vacuum at 80 ꢂC. Samples were visualized by
phosphorimaging.
Preparation oflac UV5 template strand
Supporting Information
The 211 bp lac UV5 promoter fragment (ꢁ144 to
+67) was prepared from a pUC-derived plasmid
restricted with EcoRI (GibcoBRL).20 The fragment
was purified on an 8% non-denaturing poly-
acrylamide gel. Following elution and ethanol pre-
cipitation, purified fragment was used directly for
transcription studies. A portion of the EcoRI frag-
ment to be used for gel mobility studies was dephos-
phorylated with calf intestinal alkaline phosphatase
(Promega) and then radiolabeled. Radiolabeling pro-
ceeded in a kinase buffer with 2 pmol of dephos-
phorylated lac UV5 fragment, g-[32P]-ATP (3000Ci/
mmol), and T4 polynucleotide kinase (GibcoBRL). The
reaction was quenched upon phenol/chloroform
extraction, and sample was passed through a G-50 spin
column to remove salts. The fragment was then sub-
Electrospray ionization mass spectra of oligonucleotides
8a–g.
Acknowledgements
Financial support of this work from the National Insti-
tutes of Health (GM-21199) to D.S.S. and C.B.C., and
the National Science Foundation (CHE-9732843) to
M.M.G. is appreciated.
References and Notes
1. Agrawal, S. Antisense Therapeutics; Humana: Totowa, NJ,
1996.

2328
J.-T. Hwang et al. / Bioorg. Med. Chem. 11 (2003) 2321–2328
2. Ma, D. D.; Rede, T.; Naqvi, N. A.; Cook, P. D. Biotechnol.
Annu. Rev. 2000, 5, 155.
12. McMinn, D. L.; Greenberg, M. M. J. Am. Chem. Soc.
1998, 120, 3289.
3. Praseuth, D.; Guieysse, A. L.; He
Acta 1999, 1489, 181.
´
lene, C. Biochim. Biophys.
13. McMinn, D. L.; Greenberg, M. M. Tetrahedron Lett.
1997, 38, 3123.
4. Wengel, J. Acc. Chem, Res. 1999, 32, 301.
5. Nielsen, P. E. Acc. Chem. Res. 1999, 32, 624.
6. Caplen, N. J.; Parrish, S.; Imani, F.; Fire, A.; Morgan,
R. A. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 9742.
7. Martinez, J.; Patkaniowska, A.; Urlaub, H.; Luhrmann, R.;
Tuschl, T. Cell 2002, 110, 563.
8. Perrin, D. M.; Chen, C. h. B.; Xu, Y.; Pearson, L.; Sigman,
D. S. J. Am. Chem. Soc. 1997, 119, 5746.
9. Milne, L.; Xu, Y.; Perrin, D. M.; Sigman, D. S. Proc. Natl.
Acad. Sci. U.S.A. 2000, 97, 3136.
14. Venkatesan, H.; Greenberg, M. M. J. Org. Chem. 1996,
61, 525.
15. Norman, D. G.; Reese, C. R. Synthesis 1983, 304.
16. Wincott, F. E.; Direnzo, A.; Shaffer, C.; Grimm, S.;
Tracz, D.; Workman, C.; Sweedler, D.; Gonzalez, C.; Scar-
inge, S.; Usman, N. Nucleic Acids Res. 1995, 23, 2677.
17. Vargeese, C.; Carter, J.; Yegge, J.; Krivjansky, S.; Settle,
A.; Kropp, E.; Peterson, K.; Pieken, W. Nucleic Acids Res.
1998, 26, 1046.
18. McGall, G. H.; Barone, A. D.; Diggelmann, M.; Fodor,
S. P. A.; Gentalen, E.; Ngo, N. J. Am. Chem. Soc. 1997, 119,
5081.
10. Hwang, J.-T.; Greenberg, M. M. J. Org. Chem. 2001, 66,
363.
11. Kahl, J. D.; Greenberg, M. M. J. Am. Chem. Soc. 1999,
121, 597.
19. See supporting information above.
20. Spassky, A.; Sigman, D. S. Biochemistry 1985, 24, 8050.