Fully Automated Synthesis of DNA-Binding Py-Im Polyamides Using a Triphosgene Coupling Strategy

Article abstract of DOI:10.1021/ol503388a

The fully automated solid-phase synthetic strategy of hairpin pyrrole-imidazole polyamides is described. A key advance is the development of methodology for the application of triphosgene as a coupling agent in the automated synthesis of hairpin polyamides without racemization. This automated methodology is compatible with all the typical building blocks, enabling the facile synthesis of polyamide libraries in good yield (9-15%) and crude purity. (Chemical Equation Presented).

Full text of DOI:10.1021/ol503388a

Letter
pubs.acs.org/OrgLett
Fully Automated Synthesis of DNA-Binding Py-Im Polyamides Using
a Triphosgene Coupling Strategy
Lijing Fang,† Guiyang Yao,‡ Zhengyin Pan,† Chunlei Wu,† Heng-Shan Wang,‡ Glenn A Burley,§
and Wu Su*,†
†Guangdong Key Laboratory of Nanomedicine, Institute of Biomedicine and Biotechnology, Shenzhen Institutes of Advanced
Technology, Chinese Academy of Sciences, Shenzhen, Guangdong 518055, P. R. China
‡Key Laboratory for the Chemistry and Molecular Engineer of Medicinal Resources, School of Chemistry & Pharmaceutical Sciences
of Guangxi Normal University, Guilin 541004, P. R. China
§Department of Pure & Applied Chemistry, University of Strathclyde, 295 Cathedral Street, Glasgow G1 1XL, U.K.
S
* Supporting Information
ABSTRACT: The fully automated solid-phase synthetic strategy of
hairpin pyrrole−imidazole polyamides is described. A key advance is
the development of methodology for the application of triphosgene as
a coupling agent in the automated synthesis of hairpin polyamides
without racemization. This automated methodology is compatible
with all the typical building blocks, enabling the facile synthesis of
polyamide libraries in good yield (9−15%) and crude purity.
binding affinity for target dsDNA sequences, Py-Im polymaides
have considerable potential as exogenous agents that can
regulate transcription by disrupting transcription factor−
dsDNA binding interactions in vivo.3 Additionally, the ability
of Py-Im polyamides to discriminate between match and
mismatched dsDNA sequences renders these molecules as
excellent molecular probes for diagnostic applications4 and in
the design of DNA-based nanophotonic materials.5
N-Methylpyrrole-N-methylimidazole (Py-Im) polyamides are
cell-permeable small molecules capable of recognizing
predetermined sequences of double-stranded DNA (dsDNA)
by binding in the minor groove (Figure 1).1 Sequence
selectivity is achieved by precisely oriented pairs of aromatic
Py and Im amino acids, such that an antiparallel Py/Py
preferentially targets A·T or T·A base pairs; Im/Py targets a C·
G base pair, while a Py/Im pair targets a C·G base pair.2 As a
consequence of their predictable binding and nanomolar
Over the past two decades, much attention has been devoted
to the total synthesis of Py-Im polyamides.6 The syntheses of
these molecules are challenging as a consequence of the low
inherent nucleophilicity of the immobilized aromatic amine.
The uses of conventional coupling reagents such as HATU or
HOAt/DCC produce the corresponding coupled product in
poor yield. This problem is exacerbated when Im building
blocks are used.6c,g To circumvent this issue, a number of
strategies have been employed such as the preparation of
dimeric, trimeric, and tetrameric carboxylic acids in solution
phase6a,d,j or using solution-phase total synthesis of the Py-Im
polyamide core.6e,f Recently, microwave-assisted synthesis was
utilized to facilitate the manual synthesis of Py-Im polyamide-
s.6h,i Although this method could significantly enhance the
coupling yields and reduce coupling times, the difficulty in
conducting immobilized imidazole-amine couplings was noted
even at elevated temperatures (60−80 °C). Machine-assisted
syntheses have been employed by the Dervan and Sugiyama
laboratories using Boc- and Fmoc- chemistry.6a,b,j In these
cases, the use of Boc-Py-Im-OH/Fmoc-Py-Im-OH dimers was
a necessary requirement to overcome the difficulties afore-
Received: November 21, 2014
Figure 1. Structure of Py-Im polyamides.
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ol503388a | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
mentioned.6a,b,i,j We previously reported a facile and highly
efficient solid-phase synthesis of hairpin DNA-binding poly-
amides using the cost-effective BTC [bis(trichloromethyl)
carbonate or triphosgene] activating agent.6g This strategy
was employed with great effect to produce Py-Im polyamides in
8−33% yield by manual solid-phase synthesis. At present, a
flexible, modular, and facile protocol for the fully automated
synthesis of Py-Im polyamides using monomeric building
blocks much akin to solid-phase peptide synthesis is not
available. In this paper, we report such a strategy to produce a
palette of hairpin Py-Im polyamides using BTC as the
condensing agent for all coupling steps.
BTC putatively forms highly electrophilic acid chlorides in
the presence of carboxylic acids in situ, which undergo facile
amidation of sterically hindered and electronically less reactive
amines.7 One limitation with current BTC-mediated coupling
protocols is the strong exothermic nature of the reaction,
resulting in the emission of CO2 accompanied by large amount
of collidine hydrochloride precipitate (Figure 2). In order to
Figure 3. Building blocks for the synthesis of Py-Im polyamides.
mediated solid-phase coupling of proteinogenic amino
acids.10 The key features of the procedure include the
following: (1) the preactivation of the carboxyl acid is carried
out with collidine in THF, racemization-free conditions
validated by the research of Jung and Falb;7 (2) the resulting
precipitation is converted to the soluble DIEA hydrochloride
salt by the addition of a DIEA/DMF solution. According to this
procedure, Boc-Py-OH (1.0 equiv) and BTC (0.33 equiv) were
dissolved in dry THF (1 mL) (Figure 2, A), collidine was added
slowly to form a slurry (Figure 2, B).11 Upon addition of a 10%
DIEA/DMF mixture (1.5 mL), the insoluble collidine hydro-
chloride disappeared very quickly and a clear solution was
generated (Figure 2, C). To avoid the potential clogging of the
peptide synthesizer pipelines, this special combination of bases
for the activation of amino acids could be utilized in the
automated synthesis of Py-Im polyamides.
We then investigated whether the optimized BTC activation
procedure could be conducted on a CS336X peptide
synthesizer with a computer-controlled operation system. The
synthesizer was programmed in the standard hardware
configuration for DIC/HOBt (or HBTU/DIEA) protocols.
To evaluate the automated activation efficiency of the BTC
method, Boc-Py-OH (1.0 equiv) and BTC (0.33 equiv) were
loaded as solids into an amino acid reservoir, which was set up
to be dissolved by dry THF (1.25 mL) and activated by 2,4,6-
collidine/dry THF (15%, 1.25 mL), followed by DIEA/dry
DMF (10%, 2.5 mL). During each step, nitrogen was bubbled
into the amino acid reservoir to blend the reaction mixture and
facilitate the emission of CO2 and heat. As a model study for
the HPLC analysis, excess 2-(4-methoxyphenyl)ethylamine was
added to generate the corresponding amide. As shown in
Figure 4, we found that 0.33 equiv of BTC, typically used in the
activation step to release equal molar of phosgene, was not
sufficient for full conversion of the acid chloride (trace a). We
assume that this might be due to the emission of phosgene in
Figure 2. Activation of Boc-Py-OH with BTC: (A) solution of Boc-Py-
OH and BTC in THF; (B) slurry generated by addition of 2,4,6-
collidine to A; (C) clear solution obtained by addition of a 10% DIEA/
DMF solution to B.
render BTC couplings amenable to automation, methods were
therefore required to mitigate the exotherm and to prevent
clogging of the lines of the automated synthesizer by the
precipitate. Preactivation of the amino acids with BTC outside
the peptide synthesizer was employed by Konig et al. in the
machine-assisted synthesis of aromatic oligoamides as the
solution could be freed of insoluble particles by syringe
filtration.8 However, the preactivated amino acid chloride has
limited stability, which is not feasible for use in the automated
synthesis of Py-Im polyamides. Therefore, development of a
milder and solid-free activation method is crucial for the
automation of the BTC strategy.
In a primary study, the reaction conditions of BTC-mediated
in situ activation were investigated in order to meet the
requirements of commercially available peptide synthesizers. As
observed by Fuse et al. in the formation of acid chloride in a
microflow system,9 the combinations of BTC, organic bases
(Et3N, DIEA, 2,4,6-collidine, pyridine), and solvents (THF,
1,4-dioxane, DMF) generated precipitations. The only
combination of BTC, DIEA, and DCM generating no
precipitation resulted in low coupling yields as well as
racemization of the chiral amino acids. Because Py-Im
polyamides syntheses usually involved the coupling of a chiral
amino acid as γ-turn unit, the retention of the chirality during
the incorporation of this amino acid had to be addressed
(Figure 3). Recently, we developed a reproducible, highly
efficient, and racemization free protocol for triphosgene-
Figure 4. HPLC analysis of automated activation of Boc-Py-OH with
BTC (285 nm): (a) 0.33 equiv of BTC; (b) 0.41 equiv of BTC; (c)
0.50 equiv of BTC. tr = 18.29 (Boc-Py-OH), tr = 23.16 (the resulting
amide).
B
dx.doi.org/10.1021/ol503388a | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
6h
the nitrogen bubbling step. The activation efficiency was
significantly improved when 0.50 equiv BTC was used (trace
c). Similar results were achieved when Boc-Im-OH or Fmoc-D-
Dab(Boc)-OH was used, except for Im-OH which has poor
solubility in THF. Therefore, 0.50 equiv of BTC is an optimal
ratio for the efficient activation of both aromatic amino acids
and aliphatic amino acids.
CCl3 coupling and DMF washes. For a typical synthesis on
resin (400 mg, 0.15 mmol/g), Boc-amino acid (4 equiv) and
BTC (2 equiv) were activated just prior to the coupling step
and transferred to the pretreated resin to generate the
corresponding polyamide. The automated solid-phase synthesis
of polyamide 1 involved eight coupling cycles plus a protecting
group exchange step. The program proceeded smoothly using
the optimized BTC protocol, requiring no special precautions
beyond those used for general solid-phase peptide synthesis.13
Polyamide 1 was then cleaved from the resin using 3,3′-
diamino-N-methyldipropylamine after oxidation of the hydra-
zide by N-bromosuccinimide or atmospheric O2 in the presence
of Cu(II) salts to form a transient diazene species.6g,14
However, better yields were obtained by direct aminolysis of
resin-bound polyamides under air at 55 °C for 16 h.6k After
semipreparative HPLC purification, polyamide 1 was obtained
in 15% yield with a purity of 95% (Scheme 1). Using the
In regard to the poor nucleophilicity of the resin-bound Im
amines (resin-Im-NH2), we anticipated that the successful
synthesis of Py-Im polyamides on a peptide synthesizer relied
on the efficient coupling of Boc-Py-OH, Boc-Im-OH, or Fmoc-
D-Dab(Boc)-OH (γ-turn) to Im amines. An oligomer core (4)
terminating with an Im nucleophile was used as a model study
on the aryl hydrazine resin. We found that 40 min coupling
time was optimal for complete coupling of the Boc-Im-OH
monomer to amine 4. To ensure complete coupling of Boc-Py-
OH to 4, two successive rounds of coupling were conducted
(25 min each). The coupling of Fmoc-D-Dab(Boc)-OH to the
Im nucleophile was surprisingly slow and required an extended
coupling step (200 min) to achieve acceptable yields. In
optimizing the γ-Im coupling conditions, we found that neither
the reagent set DIC/HOAt/DIEA nor PyBOP/DIEA, which
could be applied in the peptide synthesizer, were efficient in
this amide bond formation. Further research revealed that the
addition of one equivalent of HOAt to the BTC activated
intermediate dramatically accelerated the rate of coupling with
complete conversion observed by HPLC within 20 min. Using
the combination of BTC/HOAt, the Im-Im coupling was also
complete in 20 min (Table 1).
Scheme 1. Automated Synthesis of Py-Im Polyamide 1
Table 1. Optimized Conditions for the Coupling of
Monomers to Im Amine 4
a
monomer/BTC
equiv (equiv)
HOAt
coupling
b
entry
monomer
(equiv)
time (min)
c
1
2
3
Boc-Py-OH
Boc-Im-OH
4/2
4/2
4/2
50
40
Fmoc-D-
Dab(Boc)-OH
200
modular program described above, polyamide 2 comprising a
diverse range of couplings encountered in polyamide synthesis
was attempted. These included Py-Py, Im-Py, Im-Im, γ-Py, and
Im-γ couplings in addition to the known challenging Py-Im
(Boc-Py-OH/Resin-Im-NH2) coupling in the synthesis. After
cleavage from the resin using dimethylamino-propylamine
(Dp) and purification by semipreparative HPLC, polyamide
2, which was one of the most difficult sequences in polyamides
synthesis, was obtained in 9% yield.
To further confirm whether the stereochemical information
encoded in the turn unit would survive several additional
coupling steps in the automated synthesis, we synthesized the
two enantiomers of polyamide 66h from Fmoc-L-Dab(Boc)-OH
and Fmoc-D-Dab(Boc)-OH, respectively. Mosher amide
derivatives 7-(R,S) and 7-(S,S) were formed by the reaction
of both enantiomers of 6 with (S)-Mosher’s acid chloride. 7-
(R,S), 7-(S,S), and a mixture of them were subjected to HPLC
analysis using a CHIRALPAK ID column to separate the two
diastereomers. As shown in Figure 5, less than 1% epimer
(estimated by integration) could be detected for each
diastereomer (trace b and c), indicating the chirality of the γ-
4
5
Fmoc-D-
4/2
4/2
4
4
20
20
Dab(Boc)-OH
Boc-Im-OH
a
After Fmoc-D-Dab(Boc)-OH was activated using a standard BTC
procedure, a solution of HOAt/DMF/DIEA was added and the
resulting mixture was transferred to the pretreated resin for the
coupling reaction. The optimal coupling times were determined by
b
c
the chloranil test for free Im amines of 4. Two cycles, 25 min each.
Encouraged by this result, we then investigated the
automated preparation of a biologically relevant polyamide 1,
targeting the 5′-WGWWCW-3′ (where W = A/T) subset of
the consensus androgen and glucocorticoid response element.12
The synthesis program on the CS336X peptide synthesizer was
divided into three modules: (1) deprotection of the resin-
bound polyamide, including removal of the Boc group with
TFA/Phenol/H2O (92.5/5/2.5) cocktail, DCM washes, DIEA/
DMF wash, and DMF washes; (2) monomer coupling to the
resin-bound polyamide, including BTC activation (5 min),
delivery of the activated monomer, coupling (25 min), and
DMF washes; (3) the terminal Im capping, including Im-
C
dx.doi.org/10.1021/ol503388a | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Technology Innovation Council (CXZZ20120831175831692,
KQCX20130628112914285), and the National Natural Science
Foundation of China (Grant No. 21432003 and 21402232).
We are grateful for assistance at the mass facility at Peking
University Shenzhen Graduate School.
REFERENCES
■
(1) Dervan, P. B.; Edelson, B. S. Curr. Opin. Struct. Biol. 2003, 13,
284.
(2) White, S.; Szewczyk, J. W.; Turner, J. M.; Baird, E. E.; Dervan, P.
B. Nature 1998, 391, 468.
(3) (a) Blackledge, M. S.; Melander, C. Bioorg. Med. Chem. 2013, 21,
6101. (b) Baraldi, P. G.; Bovero, A.; Fruttarolo, F.; Preti, D.; Tabrizi,
M. A.; Pavani, M. G.; Romagnoli, R. Med. Res. Rev. 2004, 24, 475.
(c) Bando, T.; Sugiyama, H. Acc. Chem. Res. 2006, 39, 935. (d) Leung,
C. H.; Chan, D. S.; Ma, V. P.; Ma, D. L. Med. Res. Rev. 2013, 33, 823.
(4) (a) Chenoweth, D. M.; Viger, A.; Dervan, P. B. J. Am. Chem. Soc.
2007, 129, 2216. (b) Fechter, E. J.; Olenyuk, B.; Dervan, P. B. J. Am.
Chem. Soc. 2005, 127, 16685.
(5) (a) Cohen, J. D.; Sadowski, J. P.; Dervan, P. B. J. Am. Chem. Soc.
2008, 130, 402. (b) Dose, C.; Ho, D.; Gaub, H. E.; Dervan, P. B.;
Albrecht, C. H. Angew. Chem., Int. Ed. 2007, 46, 8384. (c) Su, W.;
Bagshaw, C. R.; Burley, G. A. Sci. Rep. 2013, 3, 1883. (d) Su, W.;
Schuster, M.; Bagshaw, C. R.; Rant, U.; Burley, G. A. Angew. Chem., Int.
Ed. 2011, 50, 2712.
Figure 5. HPLC analysis of polyamide diastereomers (310 nm): (a) a
mixture of 7-(R,S) and 7-(S,S); (b) 7-(R,S), tr = 10.90 min; (c) 7-
(S,S), tr = 13.25 min.
turn unit was retained during the automated Py-Im polyamides
synthesis.
Finally, a further application of the optimized BTC protocol
was explored via the preparation of polyamide 3 using an
Fmoc- strategy. The deprotection step was conducted by twice
deblocking (10 min) with 20% piperidine/DMF to remove
Fmoc group from the resin-bound polyamide. Fmoc-Py-OH,
Fmoc-Im-OH, and Fmoc-GABA-OH were activated by the
same BTC activation procedure. It should be noted that two
coupling cycles were needed to drive the reaction of Fmoc-Py-
OH/Resin-Im-NH2 to completion (60 min each). All other
couplings were carried out with a single-coupling cycle with
extended reaction time (40 min). After capping with Im-CCl3,
cleavage from the resin and semipreparative HPLC purification,
polyamide 3 was obtained in 12% yield.
In summary, an automated protocol to solid phase Py-Im
polyamide synthesis using BTC as the activating agent has been
developed. This protocol was facile and fully amenable for
automation on a CS336X peptide synthesizer using monomeric
N-protected carboxylic acids regardless of the required
sequence or N-protection strategy. The automation of the
BTC strategy represented a significant step forward for the
multiple, parallel synthesis of structurally diverse peptoids for
biomedical studies and also for the large-scale production of
this family of molecules for further exploration as therapeutic
scaffolds.
(6) (a) Baird, E. E.; Dervan, P. B. J. Am. Chem. Soc. 1996, 118, 6141.
(b) Wurtz, N. R.; Turner, J. M.; Baird, E. E.; Dervan, P. B. Org. Lett.
2001, 3, 1201. (c) Krutzik, P. O.; Chamberlin, A. R. Bioorg. Med.
Chem. Lett. 2002, 12, 2129. (d) Harki, D. A.; Satyamurthy, N.; Stout,
D. B.; Phelps, M. E.; Dervan, P. B. Proc. Natl. Acad. Sci. U.S.A. 2008,
105, 13039. (e) Chenoweth, D. M.; Harki, D. A.; Dervan, P. B. J. Am.
Chem. Soc. 2009, 131, 7175. (f) Chenoweth, D. M.; Harki, D. A.;
Phillips, J. W.; Dose, C.; Dervan, P. B. J. Am. Chem. Soc. 2009, 131,
7182. (g) Su, W.; Gray, S. J.; Dondi, R.; Burley, G. A. Org. Lett. 2009,
11, 3910. (h) Puckett, J. W.; Green, J. T.; Dervan, P. B. Org. Lett. 2012,
14, 2774. (i) Li, B. C.; Montgomery, D. C.; Puckett, J. W.; Dervan, P.
B. J. Org. Chem. 2013, 78, 124. (j) Kawamoto, Y.; Bando, T.; Kamada,
F.; Li, Y.; Hashiya, K.; Maeshima, K.; Sugiyama, H. J. Am. Chem. Soc.
2013, 135, 16468. (k) Fallows, A. J.; Singh, I.; Dondi, R.; Cullis, P. M.;
Burley, G. A. Org. Lett. 2014, 16, 4654.
(7) (a) Falb, E.; Yechezkel, T.; Salitra, Y.; Gilon, C. J. Pept. Res. 1999,
53, 507. (b) Thern, B.; Rudolph, J.; Jung, G. Angew. Chem., Int. Ed.
2002, 41, 2307.
(8) Konig, H. M.; Gorelik, T.; Kolb, U.; Kilbringer, A. F. M. J. Am.
Chem. Soc. 2007, 129, 704.
(9) (a) Fuse, S.; Tanable, N.; Takahashi, T. Chem. Commun. 2011,
47, 12661. (b) Fuse, S.; Mifune, Y.; Takahashi, T. Angew. Chem., Int.
Ed. 2014, 53, 851.
(10) Fang, L.; Wu, C.; Yu, Z.; Shang, P.; Cheng, Y.; Peng, Y.; Su, W.
Eur. J. Org. Chem. 2014, 34, 7572.
(11) HRMS analysis revealed that Boc-Py-OH was converted to Boc-
Py-Cl in the inert solvent (see the Supporting Information).
(12) (a) Nickols, N. G.; Dervan, P. B. Proc. Natl. Acad. Sci. U.S.A.
2007, 104, 10418. (b) Yang, F.; Nickols, N. G.; Li, B. C.; Marinov, G.
K.; Said, J. W.; Dervan, P. B. Proc. Natl. Acad. Sci. U.S.A. 2013, 110,
1863. (c) Muzikar, K. A.; Nickols, N. G.; Dervan, P. B. Proc. Natl.
Acad. Sci. U.S.A. 2009, 106, 16598.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental procedures and analysis data. This material is
available free of charge via the Internet at http://pubs.acs.org.
(13) Coin, I.; Beyermann, M.; Bienert, M. Nat. Protoc. 2007, 2, 3247.
AUTHOR INFORMATION
Corresponding Author
́
(14) (a) Gongora-Benítez, M.; Tulla-Puche, J.; Albericio, F. ACS
Comb. Sci. 2013, 15, 217. (b) Woo, Y.; Mitchell, A. R.; Camarero, J. A.
Int. J. Pept. Res. Ther. 2007, 13, 181.
■
*E-mail: wu.su@siat.ac.cn.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was supported by the “Hundred Talents Program” of
the Chinese Academy of Sciences, Shenzhen Sciences &
■
D
dx.doi.org/10.1021/ol503388a | Org. Lett. XXXX, XXX, XXX−XXX