Selective cleavage of the HIV-1 TAR-RNA with a peptide-cyclen conjugate

Article abstract of DOI:10.1002/(SICI)1521-3773(19990802)38:15<2243::AID-ANIE2243>3.0.CO;2-V

Covalent linkage of the arginine-rich fragment of the Tat protein to 1,4,7,10-tetraazacyclododecane (cyclen) results in the selective cleavage of the TAR-RNA of HIV-1 (see picture; the biotin at the 5' end acts as a label for the subsequent analysis of the cleavage fragments). The cleavage occurs at room temperature and is diminished when Eu(III) ions are present - at a concentration of about 1/10 of the concentration of the peptide-cyclen conjugate. The pH dependence indicates that two ammonium ions are responsible for the cleavage reaction. The white arrows in the schematic diagram mark the cleavage sites in RNaseT1, and the black arrows the sites in the peptide- cyclen conjugate.

Full text of DOI:10.1002/(SICI)1521-3773(19990802)38:15<2243::AID-ANIE2243>3.0.CO;2-V

COMMUNICATIONS
RNA.[2] It is known that the binding of the Tat protein to the
TAR-RNA is a crucial event in the transcription of the viral
DNA, because it stabilizes the active complex between the
polymerase, the DNA template, and the nascent mRNA.
When the Tat protein is not bound to the TAR-RNA
premature abortion of transcription takes place.[3] Even more
interesting is the fact that the arginine-rich region of the Tat
protein is sufficient to induce budding of the virus in vivo. The
fact that peptide fragments containing the basic region of Tat
show similar TAR binding as the wild-type Tat protein[4]
encouraged us to investigate the hydrolysis of this particular
RNA with metal complexes bound to the arginine-rich region
of the Tat protein.
Our initial idea was to use 1,4,7,10-tetraazacyclododecane
(cyclen) as the metal-coordinating ligand and to conjugate it
to the arginine-rich region of the Tat protein. This strategy
would then allow the cleavage of the TAR-RNA by various
metal ion complexes of cyclen to be examined. The arginine-
rich nonamer should specifically bind to the targeted RNA
and bring the metal complex in close proximity to phospho-
diester bonds. As a result of this event RNA strand cleavage
should occur.
[9] R. A. Pascal, Jr., W. D. McMillan, D. Van Engen, J. Am. Chem. Soc.
1998, 108, 5652.
[10] B. Solouki, P. Rosmus, H. Bock, G. Maier, Angew. Chem. 1980, 92, 56;
Angew. Chem. Int. Ed. Engl. 1980, 19, 51; see also reference [1a].
[11] Laser-spectroscopic measurements for the parent molecule cyclohexa-
1,4-diene: E. F. Cromwell, D.-J. Liu, M. J. J. Vrakking, A. H. Kung,
Y. T. Lee, J. Chem. Phys. 1990, 92, 3230; their quantum-chemical
discussion: R. J. Rico, M. Page, C. Doubleday, Jr., J. Am. Chem. Soc.
1992, 114, 1131.
Selective Cleavage of the HIV-1 TAR-RNA
with a Peptide ± Cyclen Conjugate**
Katrin Michaelis and Markus Kalesse*
The hydrolysis of phosphodiesters by small molecules is of
great interest since efficient catalysis of phosphodiester
hydrolysis can significantly change the life cycle of cells or
viruses. In order to assess the effect of metal ions in such
hydrolyses we investigated the hydrolysis of metal complexes
covalently linked to peptides. We chose the TAR-RNA of
HIV-1 as the target for our hydrolysis studies (Scheme 1). The
The nonamer 3 with the attached cyclen moiety was
synthesized by solid-phase synthesis using the Fmoc protocol
(Scheme 3). As the last coupling step, the Boc-protected
Scheme 1. The TAR-RNA of HIV-1 and the 31mer RNA used in the
cleavage experiments.
Scheme 3. Synthesis of the peptides 3 ± 6 by solid-phase synthesis. Boc ˆ
tert-butoxycarbonyl, Fmoc ˆ fluorene-9-ylmethoxycarbonyl, TFA ˆ tri-
fluoroacetic acid.
TAR-RNA of HIV-1 is recognized by the HIV-1 regulatory
protein Tat (Scheme 2). It has been found that the binding of
Tat to TAR-RNA up-regulates HIV-1 mRNA transcription.[1]
Detailed spectroscopic methods have been used to investigate
the Tat ± TAR interactions in which arginine residues (52 or
53) make specific contacts to the bulge region of the TAR-
cyclen acetic acid[5] (2) was coupled to the N-terminus of the
nonapeptide with subsequent cleavage from the resin by using
standard conditions (TFA) to yield the desired peptide ± cy-
clen conjugate. The peptides were purified by semi-prepara-
tive RP-HPLC and the molecular weight was determined by
matrix-assisted laser desorption/ionization time of flight mass
spectrometry (MALDI-TOF MS).[6] For the purpose of
control experiments we also synthesized the corresponding
nonamer peptide without the cyclen moiety (4) and the
hexamer peptides with and without the cyclen moiety. We
synthesized both hexamers 5 and 6 to probe whether the
distance of the cyclen moiety from the crucial arginine
residues (Arg52 and Arg53) would have an effect on the
cleavage of the RNA.
Scheme 2. The Tat protein. The arginine-rich region is written in bold
letters. The arginine residues Arg52 and Arg53 are responsible for the
selective binding to the bulge region.
[*] Priv.-Doz. Dr. M. Kalesse, K. Michaelis
Institut für Organische Chemie der Universität
Schneiderberg 1B, D-30167 Hannover (Germany)
Fax: (‡49)511-762-3011
E-mail: kalesse@mbox.oci.uni-hannover.de
The target TAR-RNA of HIV was purchased from Genset
with a slight modification at the termini. The unpaired
[**] This work was supported by the DFG (Ka 913) and the Fonds der
Chemischen Industrie.
Angew. Chem. Int. Ed. 1999, 38, No. 15
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3815-2243 $ 17.50+.50/0
2243

COMMUNICATIONS
adenine residue (A17) was not incorporated and the base pair
A15:U46 was changed into G15:C46. Biotin was introduced at
the 5'-end as a label for detecting the cleavage fragments. We
will refer to the labeled RNA as the 31mer RNA (Scheme 1).
The cleavage experiments were carried out in autoclaved
Eppendorf reaction vessels. In reactions with EuIII or ZnII the
peptides were preincubated with the metal salt for 15 min at
room temperature at three different pH values (pH ˆ 6, 7.4,
and 8). The 31mer RNA was added after the metal complexes
had been formed. The reaction was stopped by the addition of
fish sperm DNA and formamide loading buffer, and the
mixture was then analyzed on a 20% denaturing polyacryl-
amide gel (Scheme 4). The omission of the fish sperm DNA
Figure 1. Polyacrylamide gel electrophoresis (PAGE) for the sequence-
selective hydrolysis of the 31mer RNA (labeled with biotin at the 5'-end) by
peptides at room temperature and pH 7.4. Peptide concentration: 167.2 mm.
Lane 1: alkaline hydrolysis; lane 2: 3 for 3 h; lane 3: no treatment; lane 4: 3
and EDTA (0.5 mm) for 1 h; lane 5: hydrolysis by RNase T1 (G-specific);
lane 6: 4 for 1 h; lane 7: 3 for 2 h; lane 8: 3 for 1 h; lane 9: 3 for
approximately 3 min; lane 10: 5 for 1 h; lane 11: 6 for 1 h; lane 12: 3 and
EuIII (167.2 mm) for 1 h; lane 13: 6 167.2 mm and EuIII (167.2 mm) for 1 h;
lane 14: 4 167.2 mm and EuIII (167.2 mm) for 1 h.
binding constants with the RNA ± peptide complex or differ-
ent binding sites. It is noteworthy that EuIII inhibits the
cleavage in concentrations of approximately 1/10 of com-
pound 3. The exact mechanism for the inhibition of the
cleavage reaction is still under investigation, but it is
remarkable that metal ions diminish the hydrolysis of 3. Also
the control peptides without the cyclen moiety (4, 5), as well
as the hexamer with the cyclen moiety (6) gave no cleavage
(lanes 6, 10, 11).[10] In contrast, compound 3 cleaves very
efficiently at pH 6 and 7.4 at room temperature. Analysis of
the cleavage reaction immediately after the RNA and
compound 3 had been dissolved in Tris bufffer showed only
a small amount of cleavage. Incubation for 1, 2 or 3 hours at
room temperature resulted in increasing, but still selective,
cleavage. The degree of cleavage increases with time
(lanes 7 ± 9). On the other hand, the control experiments with
peptide 4, which lacks the cyclen moiety, (lane 6) as well as the
lanes with EuIII present (lanes 12 and 13) show no hydrolysis.
The control experiments in the presence of EDTA (lane 4)
showed that the observed hydrolysis was not catalyzed by
metal contamination, and in order to show that these results
were not artifacts, all experiments have been reproduced with
RNA purchased from a different source and with newly
synthesized peptides. The fact that no hydrolysis was observed
in cleavage experiments with peptides lacking the cyclen
moiety together with the fact that hydrolysis decreases at
higher pH values confirms the presence of an ammonium,
presumably even a bis(ammonium), species in the hydrolytic
reaction. Even though it has been shown that basic amines can
efficiently hydrolyze RNA, it was anticipated that the
cleavage of nuclease model systems by metal ion complexes
is more effective than cleavage by basic groups. The important
work done by Komiyama et al.[11] shows that bis(amines) can
efficiently cleave RNA by an acid/base mechanism that would
increase the nucleophilicity at the 2'-hydroxyl group and at
the same time increase its ability to function as a leaving
Scheme 4. Selective cleavage of the 31mer RNA by 3. Arrows indicate the
cleavage sites.
did not alter the cleavage results and was used to ensure that
the RNA was liberated from the peptide.
Following this general protocol we also used the nonapep-
tide ± cyclen conjugate 3 without addition of metal ions as a
control experiment to see whether the observed cleavage was
a result of the metal complex or if cleavage had taken place in
the absence of metal ions. Very much to our surprise the
nonamer± cyclen conjugate without any metal present gave
selective and efficient cleavage at pH 6 and 7.4. Only a small
amount of hydrolysis was observed at pH 8. These results
indicate a pH-dependent cleavage mechanism in which the
protonated nitrogen atoms on the cyclen residue[7] catalyze
the phosphodiester cleavage. The only cleavage site was
located between U31 and G32, as can be seen by comparing
the lanes of the alkaline hydrolysis and the RNase T1
digestion to the cleavage experiments (Figure 1).
It was even more surprising to us that upon addition of
EuIII [8] or ZnII,[9] which were used very successfully in RNA
cleavage reactions previously, the cleavage with the nona-
mer± cyclen conjugate diminished. The presence of 13.5 mm
Eu(NO3)3 in cleavage reactions with 0.135 mm RNA and
167.2 mm peptide conjugate 3 inhibited the hydrolysis of the
RNA. The presence of Zn(ClO4)2 also inhibited the RNA
cleavage but only in concentrations of 1.35 mm. Interestingly
both metal ions seem to interfere with the cleavage reaction in
different concentrations, which could be the result of different
2244
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3815-2244 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1999, 38, No. 15

COMMUNICATIONS
crystalline solid (0.6 g, 66%). 1H NMR (400 MHz, CDCl3): d ˆ
3.35 ± 3.62 (m, 14H), 2.8 ± 3.0 (m, 4H), 1.47 (s, 9H), 1.45 (s, 18H);
13C NMR (100 MHz, CDCl3): d ˆ 175.7, 173.6, 156.15, 155.43, 80.0,
79.6, 54.21, 52.1, 49.87, 47.55, 28.59, 28.39. Synthesis of 3: The
nonapeptide was synthesized following the Fmoc technique. Each
coupling step was monitored for completeness by the Kaiser Test. The
coupling between the nonapeptide and cyclen residue 2 was achieved
in CH2Cl2 with DCC, HOBT, and DMAP as the coupling reagents
group through interaction with the ammonium group. On the
other hand Goebel et al.[12] showed that RNA cleavage can
also be accomplished without the aid of an internal base. Since
the cyclen moiety exists as a bis(ammonium) salt at neutral
pH it can be envisioned that the remarkable hydrolysis is the
result of a mechanism that involves the formation of a bis-
cation on the cyclen moiety. Investigations addressing the role
of the cyclen subunit and the mode of action for this cleavage
are presently being performed in our laboratories.
The fact that the nonamer± cyclen conjugate cleaves
selectively and very efficiently at neutral pH and room
temperature and that these cleavage reactions are more
efficient in the absence than in the presence of EuIII, makes
them a promising tool for a selective interaction with the life
cycle of HIV-1.
(DCC ˆ dicyclohexylcarbodiimide,
HOBTˆ 1-hydroxy-1H-benzo-
triazole, DMAP ˆ 4-dimethylaminopyridine). The peptide ± cyclen
conjugate was cleaved off the resin with TFA under standard
conditions. These conditions cleaved off all protecting groups includ-
ing the Boc protecting groups on the cyclen. Compound 3 was purified
by semi-preparative RP-HPLC and analyzed by MALDI-TOF MS.[6]
[6] Spectrometer: Kompakt Maldi 3 from Kratos. The experiments were
performed in the positive-linear-high mode with a-cyanocinnamic
acid as the matrix. The matrix was saturated with a mixture of MeCN/
0.1% TFA (2/1). The calculated mass for compound 6 is 1552, the
‡
observed mass was 1552 and corresponds to [M ] of 6.
[7] K. C. Chang, E. Grunwald, L. R. Robinson, J. Am. Chem. Soc. 1977,
99, 3794 ± 3796; cyclen: pK1 < 1, pK2 < 2, pK3 ˆ 9.6, pK4 ˆ 10.53.
[8] J. Hall, D. Hüsken, U. Pieles, H. E. Moser, R. Häner, Chem. Biol. 1994,
1, 185.
[9] a) S. Matsuda, A. Ishikubo, A. Kuzuya, M. Yashiro, M. Komiyama,
Angew. Chem. 1998, 110, 3477 ± 3479; Angew. Chem. Int. Ed. 1998, 37,
3284 ± 3286; M. Yashiro, A. Ishikubo, M. Komiyama, J. Chem. Soc.
Chem. Commun. 1995, 1793 ± 1794.
Experimental Section
All experiments were performed in autoclaved Eppendorf reaction vessels.
Extreme precaution has been taken to avoid RNase contamination. Water
(Millipore quality) and all equipment had been treated with diethyl
pyrocarbonate (DEPC) and then autoclaved prior to used. The RNA
cleavage reaction was carried out in pH 7.4 Tris-HCl (20 mm) containing
RNA (75.4 nm), peptides (167.2 mm), and NaCl (20 mm). The reaction
mixture (5 mL) was incubated for 1 h at room temperature unless otherwise
stated. After reaction, fish sperm DNA (1 mg) and formamide gel loading
buffer (6 mL) was added to the reaction mixture. The mixture was then
heated to 858C for 5 min and then cooled in ice. 7 mL of each sample was
loaded onto a 20% denaturing polyacrylamide gel. After electrophoresis
the RNA was transferred onto a positively charged nylon membrane by
electro-blotting. After immobilization at 808C for 30 min followed by the
wash protocol from Ambion, the RNA was visualized with streptavidin/
alkaline phosphatase and CDP-star on Kodak film.
[10] Hydrolysis experiments with nonconjugated cyclen gave no RNA
cleavage under these conditions for 4 hours. The hexamer 6 did not
generate unspecific cleavage even after 6 h.
[11] a) M. Komiyama, T. Inokawa, K. Yoshinari, J. Chem. Soc. Chem.
Commun. 1995, 77; b) M. Endo, K. Hirata, T. Ihara, Shinji, M. Takagi,
M. Komijama, J. Am. Chem. Soc. 1996, 118, 5478 ± 5479; c) M.
Komijama, K. Yoshinari, J. Org. Chem. 1997, 62, 2155; d) For a recent
review of base-catalyzed RNA cleavage, see M. Oivanen, S. Kuusela,
H. Loennberg, Chem. Rev. 1998, 98, 961; for a recent review of basic
RNase model systems, see K. Kurz, Chem. Unserer Zeit 1998, 32, 94 ±
103.
[12] K. Kurz, M. W. Göbel, Helv. Chim. Acta 1996, 79, 1967.
Received: December 23, 1998 [Z12820IE]
German version: Angew. Chem. 1999, 111, 2382 ± 2385
Keywords: macrocycles ´ nucleic acids ´ peptides ´ RNA
[1] a) C. Dingwall, I. Ernberg, M. J. Gait, S. M. Green, S. Heaphy, J. Karn,
A. D. Lowe, M. Singh, M. A. Skinner, EMBO J. 1990, 9, 4145; b) U.
Delling, L. S. Reid, R. W. Barnett, M. Y.-X. Ma, S. Climie, M.
Summer-Smith, N. Sonnenberg, J. Virol. 1992, 65, 7012.
The First Synthesis of Organic Diselenolates:
Application to the Synthesis of Diorganyl
Diselenides**
[2] a) B. J. Calnan, B. Tidor, S. Biancalana, D. Hudson, A. D. Frankel,
Science 1991, 252, 1167; b) J. D. Puglisi, R. Tan, B. J. Calnan, A. D.
Frankel, J. R. Williamson, Science 1992, 257, 76; c) M. J. Churcher, C.
Lamont, F. Hamy, C. Dingwall, S. M. Green, A. D. Lowe, P. J. G.
Butler, M. J. Gait, J. Karn, J. Mol. Biol. 1993, 230, 90; d) F. Hamy, U.
Assline, J. Grasby, S. Iwai, C. Pritchard, G. Slim, P. J. G. Butler, J.
Karn, M. J. Gait, J. Mol. Biol. 1993, 230, 111; e) F. Aboul-ela, J. Karn,
G. Varani, J. Mol. Biol. 1995, 253, 313; f) M. A. Farrow, F. Aboul-ela,
D. Owen, A. Karpeisky, L. Beigelman, M. J. Gait, Biochemistry 1998,
37, 3096.
Alain Krief,* Thierry Van Wemmel, Martine Redon,
Willy Dumont, and Cathy Delmotte
Â
Dedicated to Professor Leon Ghosez
on the occasion of his 65th birthday
There has been a growth in interest[1] of organoselenium
chemistry over the last two decades and although several new
reactions have been described, a relatively small number of
[3] P. A. Sharp, R. A. Marciniak, Cell 1989, 59, 229.
[4] a) K. M. Weeks, C. Ampe, S. C. Schultz, T. A. Steitz, D. M. Crothers,
Science 1990, 249, 1281; b) A. D. Frankel, Protein Sci. 1992, 1, 1539.
[5] For the synthesis of 1, see E. Kimura, S. Aoki, T. Koike, M. Shiro, J.
Am. Chem. Soc. 1997, 119, 3068 ± 3076. Synthesis of 2: The Boc-
[*] Prof. A. Krief, T. Van Wemmel, Dr. M. Redon, Dr. W. Dumont,
Dipl.-Chem. C. Delmotte
protected cyclen
1 (950 mg, 1.7 mmol) and K2CO3 (300 mg,
Á
Laboratoire de Chimie Organique de Synthese
2.15 mmol) were dissolved in acetonitrile (10 mL). Methyl bromo-
acetate (0.20 mL, 2.15 mmol) was added over 4 h at RT. The reaction
mixture was stirred for 4 days, then filtered and concentrated. The
crude material was dissolved in methanol/water (3/1, 12 mL) and
LiOH (90 mg, 2.15 mmol) added. The reaction was stirred for 12 h,
neutralized with NH4Cl, and extracted with CH2Cl2. The organic layer
was dried over MgSO4, concentrated and purified by flash chroma-
tography (MeOH/ethyl acetate, 5/1) to yield compound 2 as a
Â
Departement de Chimie
Â
Facultes Universitaires Notre-Dame de la Paix
61 rue de Bruxelles, B-5000 Namur (Belgium)
Fax: (‡32)81-724536
E-mail: alain.krief@fundp.ac.be
Â
Â
[**] T. Van Wemmel, Memoire de Licence, Facultes Universitaires N.-D.
de la Paix, Namur, September 1991.
Angew. Chem. Int. Ed. 1999, 38, No. 15
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3815-2245 $ 17.50+.50/0
2245