Synthesis of Naphthyridine Carbamate Dimer (NCD) Derivatives Modified with Alkanethiol and Binding Properties of G-G Mismatch DNA

Article abstract of DOI:10.1021/acs.orglett.7b01632

A series of new DNA binding molecules NCD-Cn-SH (n = 3, 4, 5, and 6) is reported, which possesses the NCD (naphthyridine carbamate dimer) domain selectively binding to the G-G mismatch in the 5′-CGG-3′/5′-CGG-3′ sequence and a thiol moiety, which undergoes spontaneous dimerization to (NCD-Cn-S)₂ upon oxidation under aerobic conditions. The S-S dimer (NCD-Cn-S)₂ produced the 1:1 binding complex with improved thermal stability. The dimer binding to the CGG/CGG DNA showed higher positive cooperativity than the binding of monomer and previously synthesized NCTn derivative. The dimerization of NCD-Cn-SH was selectively accelerated on the CGG repeat DNA but not on the CAG repeat DNA.

Full text of DOI:10.1021/acs.orglett.7b01632

Letter
pubs.acs.org/OrgLett
Synthesis of Naphthyridine Carbamate Dimer (NCD) Derivatives
Modified with Alkanethiol and Binding Properties of G−G Mismatch
DNA
Takeshi Yamada, Shouta Miki, Anisa Ul’Husna, Akiko Michikawa, and Kazuhiko Nakatani*
Department of Regulatory Bioorganic Chemistry, The Institute of Scientific and Industrial Research (ISIR), Osaka University,
Mihogaoka 8-1, Ibaraki 567-0047, Japan
S
* Supporting Information
ABSTRACT: A series of new DNA binding molecules NCD−Cn−SH (n = 3, 4, 5, and 6) is reported, which possesses the NCD
(naphthyridine carbamate dimer) domain selectively binding to the G−G mismatch in the 5′-CGG-3′/5′-CGG-3′ sequence and
a thiol moiety, which undergoes spontaneous dimerization to (NCD−Cn−S)2 upon oxidation under aerobic conditions. The S−
S dimer (NCD−Cn−S)2 produced the 1:1 binding complex with improved thermal stability. The dimer binding to the CGG/
CGG DNA showed higher positive cooperativity than the binding of monomer and previously synthesized NCTn derivative. The
dimerization of NCD−Cn−SH was selectively accelerated on the CGG repeat DNA but not on the CAG repeat DNA.
rinucleotide repeat diseases are neurodegenerative dis-
orders caused by excessive expansion of genetically
molecular tool to investigate fragile X syndrome (FXS),19
FXTAS,20 and FXPOI,21 which were associated with patho-
logical expansion of CGG repeats in FMR1.22
T
unstable 5′-CNG-3′ (N = A, C, G, and T) sequences in
genome.1−3 Several small molecules binding to those
trinucleotide repeats DNAs and RNAs were reported.4−12 We
previously reported a small molecule named naphthyridine
carbamate dimer NCD (Figure 1a), which strongly bound to
the G−G mismatch in the 5′-CGG-3′/5′-CGG-3′ triad,13
possibly produced in a hairpin secondary structure of CGG
repeat DNAs.14 In the CGG/CGG triad, two NCD molecules
bound to four guanines by disrupting Watson−Crick hydrogen
bonding of two C−G base pairs and a G−G mismatch. NMR15
and cold spray ionization (CSI) TOF-MS analysis,16 and the
increased reactivity of flipped out cytosines toward hydroxyl-
amine17 revealed that two NCD domains bound to four
guanines in the CGG/CGG triad to force the widowed cytosine
to flip out. A series of naphthyridine tetramer NCTn16 and Z-
NCTS18 (Figure 1a), in which two NCD moieties were
connected by an aliphatic linker or Z-stilbene, were synthesized
to improve the affinity, and Z-NCTS was found to show
stronger binding to the G−G mismatch than NCD,18 most
likely due to the positive cooperative binding of four
naphthyridine units in two NCD domains to the CGG/CGG
triad by the spatial arrangement of two NCD domains in close
vicinity to the binding site. This speculation was supported by
the failure of detection of 1:1 bound complex, which underwent
subsequent binding of a second NCD molecule to form the 2:1
complex (Figure 1b). The strong binding property of NCD and
Z-NCTS to the CGG repeat DNAs implicates the potential as a
Despite the improved binding properties, the large molecular
size and hydrophobicity of Z-NCTS and other NCT derivatives
are disadvantages to utilize them for the biological studies. In
the development of low molecular weight drugs, it is known
that highly hydrophobic compounds often have problems on
ADME properties, particularly absorption.23,24 To take
advantage of Z-NCTS and NCTn derivatives25,26 with higher
affinity than the monomer and to circumvent the issues of large
molecular size and low hydrophilicity, we planned in situ
dimerization of thiol derivative of NCD by spontaneous
oxidation. Thus, we anticipated that the oxidation of thiol
derivatives of NCD under aerobic conditions in aqueous
solution should provide a NCD dimer through a disulfide
linkage, which might show comparable binding properties to
those characteristics of NCT derivatives. It was also anticipated
that CGG/CGG triad would serve as a reaction template to
accelerate the oxidative dimerization of NCD−Cn−SH (Figure
1b). Based on these assumptions, a series of NCD−Cn−SH (n
= 3, 4, 5, or 6) derivatives (Figure 1a) that have an alkyl thiol
group were designed. Molecular modeling simulations of the
complex consisted of two NCD−C4−SHs, and a G−G
mismatch DNA (Figure S1b) revealed that two thiol groups
Received: May 31, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.7b01632
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(retention time 15 min) (Figure S2) spontaneously started to
produce a product at the retention time of 18 min, which was
identified as the disulfide (NCD−C4−S)2 by MS (m/z =
1181.9, calcd [M + H]+ 1181.5). The rate for the dimerization
was determined by assuming the second order kinetics and
described in the Supporting Information. Among four
derivatives, NCD−C3−SH undergoes dimerization the most
rapidly and NCD−C5−SH was the slowest under the
conditions, although the difference in the rate is within 5-fold.
Next, the oxidative dimerization of NCD−C4−SH (40 μM)
was investigated in the presence of repeat DNAs 5′-(CXG)9-3′
(X = G or A, 5 μM), which could form a hairpin secondary
structure involving X−X mismatch in the CXG/CXG triad.14
The decrease in the amount of NCD−C4−SH was significantly
faster in the presence of the CGG repeat DNA (Figure 2c, blue
Figure 1. (a) Hydrogen bonding pattern of NCD derivatives to two
guanines in the G−G mismatch DNA and compounds used in these
studies. (b) Schematic illustration of cooperative binding of two NCD
derivatives and template assisted S−S dimer formation.
of two NCD−C4−SHs were close enough to form a disulfide
linkage, and the dimer (NCD−C4−S)2 is likely to form a stable
1:1 complex with the G−G mismatch DNA (Figure S1c). We
here describe the preparation and the binding properties of
these NCD−Cn−SH (n = 3, 4, 5, or 6) derivatives under redox
conditions. These studies revealed accelerated dimerization of
NCD−Cn−SH selectively on the CGG/CGG triad and the
formation of the 1:1 complex with high cooperativity as judged
by the tangent angle in the UV-melting profile. These in situ
dimerizations of ligands on the target repeat DNA could be a
potentially useful strategy for the molecular design in repeat
targeting.
Figure 2. RP-HPLC profiles of NCD−C4−SH in the presence of 5′-
(CGG)9-3′ repeat DNA in aerobic conditions at 37 °C after (a) 0 and
(b) 4 h detected at 260 nm. Reaction conditions: NCD−C4−SH (40
μM), 5′-(CGG)9-3′ (5 μM), sodium cacodylate buffer (10 mM, pH
7.0), NaCl (100 mM), internal standard (*, adenosine 20 μM), and
0.1% Tween20. HPLC: 0.1% TFA and MeCN were eluted 1 mL/min
through an ODS-end-capped column. The percentage of MeCN was
changed from 0 to 40 gradually for 20 min. The eluted compounds
were detected at 260 nm by PDA. (c) Plot of [NCD−C4−SH] in the
absence of DNA (red) and in the presence of NaCl (100 mM) and 5′-
(CGG)9-3′ (5 μM, blue) or 5′-(CAG)9-3′ (5 μM, green).
Preparation of NCD−Cn−SH is summarized in Scheme 1.
Aldehydes having trityl-protected thiol (see Supporting
Scheme 1
line) than the reaction without DNA (Figure 2c, red line;
Figure S2), while no significant acceleration was observed in
the presence of nontarget CAG repeat DNA (Figure 2c, green
line; Figure S5). In addition to the accelerated rate for the
dimerization, the efficiency of the reaction is also improved.
Most of NCD−C4−SH was transformed into the dimer after 6
h in the presence of the CGG repeat DNA, whereas about 50%
of the monomer remained without DNA and with the CAG
repeat DNA. These results clearly indicated that the oxidative
dimerization of NCD−C4−SH was accelerated selectively by
the presence of the CGG repeat DNA. It is highly likely that
the biding of two NCD−C4−SHs at the CGG/CGG site
arranged two thiols in close vicinity with the geometry suitable
for the oxidative dimerization.
Information) were coupled with NCD under reductive
amination conditions with NaBH3CN. Subsequent deprotec-
tion of the trityl group was carried out with triethylsilane to
finish the synthesis.27 First, the oxidation reactions of NCD−
C4−SHs leading to the corresponding S−S dimer were
investigated under aerobic conditions by reverse phase HPLC
(RP-HPLC) equipped with a photo diode array (PDA) and
analyzed by electrospray ionization mass spectrometry (ESI-
MS) or MALDI-TOF-MS. The solution of freshly prepared
NCD−C4−SH (40 μM, m/z = found 592.2694, calcd [M +
H]+ 592.2700) in an aqueous buffer solution (pH 7) was left
exposed to the air at 37 °C. Dimerization of NCD−C4−SH
The hydrophilicity of molecules was assessed by the
retention time on RP-HPLC using a ODS-end-capped column
and 0.1% TFA and CH3CN as eluents.28 The retention time of
B
DOI: 10.1021/acs.orglett.7b01632
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
NCD−C4−SH, (NCD−C4−S)2, NCT8, and Z-NCTS were
about 16, 18, 23, and 18 min, respectively (Figure S6), showing
that NCD−C4−SH is more hydrophilic than NCT8 and Z-
NCTS.
−S−S− dimer during the measurements (Figure 4). The
reported melting temperatures (tm) were the average of three
independent measurements and determined as the temperature
showing the highest tangent angle (TA), which was determined
by plotting the slope of the line connecting every two data
points (Table 1).
The stoichiometry for the binding of NCD−Cn−SH and
(NCD−Cn−S)2 to the G−G mismatch in the 5′-CGG-3′/5′-
CGG-3′ sequence was determined by CSI-TOF-MS. The CSI-
TOF-MS analysis of the mixture of NCD−C4−SH and the G−
G mismatch DNA 5′-d(TCA ACG GTT GA)-3′/3′-d(AGT
TGG CAA CT)-5′ showed the 5− ion of the ligand-bound
DNA complex in 2:1 binding stoichiometry [2NCD−C4−SH·
DNA]5− (m/z 1578.0593). The CSI-MS analysis of the
solution of the dimer (NCD−C4−S)2 and the DNA also
showed the 5− ion (m/z 1577.6612). The difference between
two 5− ions in m/z was 0.3981, suggesting the formation of the
1:1 complex (NCD−C4−S)2·DNA having a difference of two
hydrogens (in theory, 0.4 at the 5− state) from the 2:1 bound
complex 2NCD−C4−SH·DNA of the monomer (Figure 3).
Table 1. TA and Δtm Values for the DNA Containing a G−G
a
Mismatch in the Absence or the Presence of Compounds
b
c
compd
TA (°C−1
)
tm (°C)
SD
Δtm (°C)
0.010
0.013
0.013
0.031
0.015
0.011
0.012
0.027
0.013
0.019
0.014
0.027
32.6
54.6
53.0
54.7
51.6
34.3
55.7
58.0
51.4
59.1
51.2
59.1
2.5
0.3
2.7
0.2
0.1
0.8
0.1
0.7
0.3
1.3
0.3
1.1
NCD
22.0
20.4
22.1
19.0
1.7
d
NCT8
d
Z-NCTS
e
NCD−C3−SH
(NCD−C3−S)2
NCD−C4−SH
(NCD−C4−S)2
NCD−C5−SH
(NCD−C5−S)2
NCD−C6−SH
(NCD−C6−S)2
d
d
d
d
e
23.1
25.4
18.8
26.5
18.6
26.5
e
e
a
UV-melting profiles for the 5 μM DNA 5′-(TCAA CGG TTGA)-3′/
3′-(AGTT GGC AACT)-5′ in the absence or the presence of each
compound were measured in 10 mM sodium cacodylate buffer (pH
7.0) containing 100 mM NaCl, and 0.1% Tween 20. The calculation
b
method for tangent angle (TA) was described in Supporting
c
d
Information. Δtm = tm − tm (w/o ligand). Compound concentration
e
was 20 μM. Compound concentration was 40 and 60 μM TCEP was
Figure 3. CSI-TOF-MS analysis of the G−G mismatch DNA (5′-
TCAA CGG TTGA-3′/3′-AGTT GGC AACT-5′) (10 μM) in the
presence of NCD−C4−SH (40 μM) (red) and (NCD−C4−S)2 (20
μM) (blue) measured in 1:1 MeOH/H2O (v/v) containing NH4OAc
(100 mM). The region of the 5− ion of 2:1 adducts was shown. The
full range spectra are shown in Figure S6.
added in addition to the shown above.
The effects of each compound on the stability of the CGG/
CGG DNA were evaluated by the Δtm value. The Δtm value
obtained for NCD was 22.0 °C, and those for NCD−Cn−SH
were 19.0 (n = 3), 23.1 (n = 4), 18.8 (n = 5), and 18.6 (n = 6)
°C, respectively, suggesting that modification at the secondary
amino group in NCD showed small effects on the
thermodynamic stability of the 2:1 complex. In contrast to
the monomer, the dimer showed marked structure-dependent
stability of the 1:1 complex. While (NCD−C3−S)2 demon-
strated only a negligible Δtm value (1.7 °C), possibly due to the
thermal degradation of (NCD−C3−S)2 under the measure-
ments, the other three dimers showed higher Δtm values 25.4
(n = 4), 26.5 (n = 5), and 26.5 (n = 6) as compared with 22.0,
20.4, and 22.1 °C obtained for NCD, NCT8, and Z-NCTS,
respectively (Table 1).
NCD−Cn−SH and (NCD−Cn−S)2 were also determined by
CSI-TOF-MS to produce 2:1 and 1:1 binding complexes with
the CGG/CGG DNA, respectively (Figure S8). All m/z values
of complexes were summarized in Table S1.
Having confirmed the superior hydrophilicity of NCD−Cn−
SH and (NCD−Cn−S)2 to NCT8 and Z-NCTS and the 1:1
binding stoichiometry to the CGG/CGG DNA, we then
analyzed the cooperativity of four naphthyridine heterocycles in
the binding to the guanine base from the UV-melting profiles.
For the UV-melting measurements of NCD−Cn−SH, a
reducing agent tris(2-carboxyethyl)phosphine hydrochloride
(TCEP) was added to the solution to avoid the oxidation to the
Figure 4. UV-melting profiles of the G−G mismatch DNA (open circle) and tangent angle (filled circle) of the UV-melting profiles in the presence
or absence of the ligand. The highest TA is represented by a bar. (a) DNA only (black) and DNA with NCD (green); Z-NCTS (orange); (b)
NCD−C4−SH (blue), (NCD−C4−S)2 (red); (c) NCD−C5−SH (blue), (NCD−C5−S)2 (red); (d) NCD−C6−SH (blue), (NCD−C6−S)2
(red).
C
DOI: 10.1021/acs.orglett.7b01632
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(3) Orr, H. T.; Zoghbi, H. Y. Annu. Rev. Neurosci. 2007, 30, 575−623.
(4) Gareiss, P. C.; Sobczak, K.; McNaughton, B. R.; Palde, P. B.;
Thornton, C. a.; Miller, B. L. J. Am. Chem. Soc. 2008, 130, 16254−
16261.
(5) Ofori, L. O.; Hoskins, J.; Nakamori, M.; Thornton, C. A.; Miller,
B. L. Nucleic Acids Res. 2012, 40, 6380−6390.
(6) Arambula, J. F.; Ramisetty, S. R.; Baranger, A. M.; Zimmerman, S.
C. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 16068−16073.
(7) Nguyen, L.; Luu, L. M.; Peng, S.; Serrano, J. F.; Chan, H. Y. E.;
Zimmerman, S. C. J. Am. Chem. Soc. 2015, 137, 14180−14189.
(8) Pushechnikov, A.; Lee, M. M.; Childs-Disney, J. L.; Sobczak, K.;
French, J. M.; Thornton, C. A.; Disney, M. D. J. Am. Chem. Soc. 2009,
131, 9767−9779.
(9) Tran, T.; Childs-Disney, J. L.; Liu, B.; Guan, L.; Rzuczek, S.;
Disney, M. D. ACS Chem. Biol. 2014, 9, 904−912.
(10) Li, J.; Matsumoto, J.; Bai, L.-P.; Murata, A.; Dohno, C.;
Nakatani, K. Chem. - Eur. J. 2016, 22, 1−10.
(11) Nakatani, K.; Hagihara, S.; Goto, Y.; Kobori, A.; Hagihara, M.;
Hayashi, G.; Kyo, M.; Nomura, M.; Mishima, M.; Kojima, C. Nat.
Chem. Biol. 2005, 1, 39−43.
(12) Nakatani, K.; Sando, S.; Saito, I. Nat. Biotechnol. 2001, 19, 51−
55.
The characteristic aspects of the UV-melting profiles for the
S−S dimer gave steeper increase in the absorbance against
temperature than the monomer, suggesting positive coopera-
tivity for the binding of four naphthyridine units in two NCD
moieties. TAs for the monomer were almost the same as that
for NCD and NCT8 of 0.013, suggesting the mode of the
binding is not affected by the modification as we speculated
from stoichiometry analysis. In contrast, those for S−S dimers
were 0.027 (n = 4), 0.019 (n = 5), and 0.027 °C−1 (n = 6),
showing about 1.5−2.1 times larger TA than that for NCD,
NCT8, and monomers. Z-NCTS showed the largest TA value
(0.031 °C−1) among molecules we investigated. These data
strongly supported that the dimerization of NCD−Cn−SH is
positively effective for the binding of four naphthyridine units
to the four guanine bases in the CGG/CGG DNA.
In conclusion, NCD−Cn−SH showed superior hydro-
philicity to NCT8 and Z-NCTS and efficient binding to the
CGG/CGG DNA upon spontaneous dimerization under
aerobic conditions. The oxidative dimerization was selectively
accelerated in the presence of the CGG repeat DNA. While
these results would be one of marked examples of DNA−
template organic synthesis,29−31 much more important aspects
of NCD−Cn−SH were disclosed by these studies. Thus, the
monomeric NCD−Cn−SH may have a chance to accumulate
on the CGG repeat DNA by the accelerated oxidation to the
S−S dimer leading to the formation of the complex with higher
thermal stability. These scenario would be attractive for
chemical biology studies on the Fragile X syndrome and
related diseases caused by CGG repeat expansion not only in
vitro but also in more diseases relevant cell models.
(13) Peng, T.; Murase, T.; Goto, Y.; Kobori, A.; Nakatani, K. Bioorg.
Med. Chem. Lett. 2005, 15, 259−262.
(14) Nadel, Y.; Weisman-Shomer, P.; Fry, M. J. Biol. Chem. 1995,
270, 28970−28977.
(15) Nomura, M.; Hagihara, S.; Goto, Y.; Nakatani, K.; Kojima, C.
Nucleic Acids Symp. Ser. 2005, 49, 213−214.
(16) Peng, T.; Dohno, C.; Nakatani, K. Angew. Chem., Int. Ed. 2006,
45, 5623−5626.
(17) Oka, Y.; Peng, T.; Takei, F.; Nakatani, K. Org. Lett. 2009, 11,
1377−1379.
(18) Dohno, C.; Kohyama, I.; Hong, C.; Nakatani, K. Nucleic Acids
Res. 2012, 40, 2771−2781.
(19) Zhao, X. N.; Usdin, K. Genes 2016, 7, 1−14.
(20) Sellier, C.; Buijsen, R. A. M.; He, F.; Natla, S.; Jung, L.; Tropel,
P.; Gaucherot, A.; Jacobs, H.; Meziane, H.; Vincent, A.; Champy, M.
F.; Sorg, T.; Pavlovic, G.; Wattenhofer-Donze, M.; Birling, M. C.;
Oulad-Abdelghani, M.; Eberling, P.; Ruffenach, F.; Joint, M.; Anheim,
M.; Martinez-Cerdeno, V.; Tassone, F.; Willemsen, R.; Hukema, R. K.;
Viville, S.; Martinat, C.; Todd, P. K.; Charlet-Berguerand, N. Neuron
2017, 93, 331−347.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.7b01632.
Experimental section and additional figures, schemes, and
tables (PDF)
(21) Barasoain, M.; Barrenetxea, G.; Huerta, I.; Tel
Arrieta, I. Genes 2016, 7, 1−18.
́
ez, M.; Criado, B.;
AUTHOR INFORMATION
(22) Verkerk, A. J. M. H.; Pieretti, M.; Sutcliffe, J. S.; Fu, Y. H.; Kuhl,
D. P. A.; Pizzuti, A.; Reiner, O.; Richards, S.; Victoria, M. F.; Zhang, F.;
Eussen, B. E.; van Ommen, G. J. B.; Blonden, L. A. J.; Riggins, G. J.;
Chastain, J. L.; Kunst, C. B.; Galjaard, H.; Thomas Caskey, C.; Nelson,
D. L.; Oostra, B. A.; Warran, S. T. Cell 1991, 65, 905−914.
(23) Lipinski, C. A. J. Pharmacol. Toxicol. Methods 2000, 44, 235−
249.
■
Corresponding Author
*E-mail: nakatani@sanken.osaka-u.ac.jp.
ORCID
Takeshi Yamada: 0000-0003-3275-415X
Notes
(24) Dohta, Y.; Yamashita, T.; Horiike, S.; Nakamura, T.; Fukami, T.
Anal. Chem. 2007, 79, 8312−8315.
The authors declare no competing financial interest.
(25) Dohno, C.; Kohyama, I.; Kimura, M.; Hagihara, M.; Nakatani,
K. Angew. Chem., Int. Ed. 2013, 52, 9976−9979.
ACKNOWLEDGMENTS
■
(26) Matsumoto, S.; Hong, C.; Otabe, T.; Murata, A.; Nakatani, K.
Bioorg. Med. Chem. Lett. 2013, 23, 3539−3541.
This work was supported by JSPS KAKENHI Grant-in-Aid for
Specially Promoted Research (26000007) to K.N. and by
Young Scientist (B) (15K17885, 17K14516) to T.Y., performed
under the Research Program of “Dynamic Alliance for Open
Innovation Bridging Human, Environment and Materials” in
“Network Joint Research Center for Materials and Devices”.
The authors would like to thank Dr. Chikara Dohno and Mr.
Yihuan Lu for providing Z-NCTS.
(27) Pearson, D. A.; Blanchette, M.; Baker, M. L.; Guindon, C. A.
Tetrahedron Lett. 1989, 30, 2739−2742.
̂ ̆
(28) Tache, F.; Nascu-Briciu, R. D.; Sarbu, C.; Micale, F.;
Medvedovici, A. J. Pharm. Biomed. Anal. 2012, 57, 82−93.
(29) Sando, S.; Kool, E. T. J. Am. Chem. Soc. 2002, 124, 2096−2097.
(30) Li, X.; Liu, D. R. Angew. Chem., Int. Ed. 2004, 43, 4848−4870.
(31) Rzuczek, S. G.; Colgan, L. A.; Nakai, Y.; Cameron, M. D.;
Furling, D.; Yasuda, R.; Disney, M. D. Nat. Chem. Biol. 2016, 13, 188−
193.
REFERENCES
■
(1) Pearson, C. E.; Edamura, K. N.; Cleary, J. D. Nat. Rev. Genet.
2005, 6, 729−742.
(2) Mirkin, S. M. Nature 2007, 447, 932−940.
D
DOI: 10.1021/acs.orglett.7b01632
Org. Lett. XXXX, XXX, XXX−XXX