A solution to the component instability problem in the preparation of peptides containing C2-substituted cis-cyclobutane β-aminoacids: synthesis of a stable rhodopeptin analogue

Article abstract of DOI:10.1016/j.tetlet.2006.06.027

Despite the inherent instability of C2-substituted cis-cyclobutane β-aminoacids, incorporation of such residues into peptides is shown to be possible through use of a 1-amino-2-(hydroxymethyl)cyclobutane derivative as a stable β-aminoacid surrogate. This

Full text of DOI:10.1016/j.tetlet.2006.06.027

Tetrahedron Letters 47 (2006) 5981–5984
A solution to the component instability problem in the
preparation of peptides containing C2-substituted cis-cyclobutane
b-aminoacids: synthesis of a stable rhodopeptin analogue
Olivier Roy, Sophie Faure and David J. Aitken*
Laboratoire SEESIB-CNRS, De´partement de Chimie, Universite´ Blaise Pascal, Clermont-Ferrand II,
`
24 Avenue des Landais, 63177 Aubiere cedex, France
Received 14 April 2006; revised 31 May 2006; accepted 2 June 2006
Abstract—Despite the inherent instability of C2-substituted cis-cyclobutane b-aminoacids, incorporation of such residues into pep-
tides is shown to be possible through use of a 1-amino-2-(hydroxymethyl)cyclobutane derivative as a stable b-aminoacid surrogate.
This synthetic strategy was validated by the synthesis of a rhodopeptin analogue.
ꢀ 2006 Elsevier Ltd. All rights reserved.
O
O
R1
R2
R1
R2
The incorporation of conformationally restricted b-
aminoacids into linear or cyclic peptides can dramati-
cally influence their secondary and tertiary structures,
and therefore their biological activities.1 Alicyclic b-
aminoacids are particularly attractive building-blocks
for imposing constraints upon peptides.2 In contrast
with intensive work on cyclohexane, cyclopentane and
cyclopropane b-aminoacids,2,3 structural studies on pep-
tides containing cyclobutane b-aminoacids are in their
R1
R2
CO2H
O
ii
NH
NH
i
N
NH2
N
H
O
N
H
O
H
(
)-2
( )-3
1
iii
CO2H
O
O
O
OH
R1
retro-
Mannich
iminium
hydrolysis
infancy.4 Ortuno’s work on the incorporation of the
˜
OH
R1 = H
2 = Me, Ph
parent compound cis-2-aminocyclobutane-1-carboxylic
acid into small oligomers produced encouraging preli-
minary results and indicated a strong tendency to
impose well-defined secondary structure.4 This parent
molecule can be prepared in racemic5,4d or enantiomeri-
cally pure6,4d form, but the synthesis of ring-substituted
cyclobutane b-aminoacids remains rare.7
NH2
NH2
R
R2
R2
5
(
)-4
R2
Scheme 1. Reagents and conditions: (i) hm, ethylene, acetone/water
1:1, rt; (ii) NaOH 0.5 M, rt; (iii) NaNO2, HCl 3.5 M, rt.
occurs by a retro-Mannich reaction leading to the corre-
sponding acyclic d-ketoacids 5.9 In this letter, we pro-
vide a solution to this component instability problem
in the preparation of peptides containing cis-cyclo-
butane b-aminoacids, and illustrate its application to
the synthesis of a constrained analogue of a rhodopeptin.
We recently described the synthesis of a series of C1- or
C2-substituted
cis-2-aminocyclobutane-1-carboxylic
acids 4 based on the photochemical [2+2] cycloaddition
of substituted uracils 1 with ethylene, to give 2, followed
by two-step heterocycle transformation via a ureidoacid
intermediate 3 (Scheme 1).8 This work revealed the
inherent instability of cis-cyclobutane b-aminoacids
substituted at C2 with electron-donating substituents,
even weak ones such as alkyl or aryl. Rapid ring opening
Since C2-substituted cis-cyclobutane b-aminoacids can-
not be used as such in peptide synthesis, it was consid-
ered appropriate to use a stable surrogate. A primary
alcohol as a masked carboxylic acid function seems to
be the most convenient. 1-Amino-2-(hydroxymethyl)-
cyclobutane derivatives have been described with no
*
Corresponding author. Tel.: +33 4 73407184; fax: +33 4 73407717;
e-mail: David.Aitken@univ-bpclermont.fr
0040-4039/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.06.027

5982
O. Roy et al. / Tetrahedron Letters 47 (2006) 5981–5984
OH
O
O
H
N
O
H
H
H
N
i
NH
O
(CH2)nNH2
N
H
N
H
NH2
R
N
H
N
H
O
Me
Me
O
O
( )-7
( )-6
ii
Rhodopeptin C1: R = (CH2)6CH(CH3)CH2CH3; n = 3
Rhodopeptin B5: R = (CH2)8CH(CH3)2; n = 4
iii
OH
H
H
iv
O
O
H
N
O
H
N
NH2
N
H
O
(CH2)4NH2
Me
Me
N
H
N
H
R
O
( )-9
( )-8
O
Scheme 2. Reagents and conditions: (i) NaBH4, H2O, rt, then H+
resin, 72%; (ii) NaNO2, HCl 3.5 M, rt, 15%; (iii) NaNO2, HCl 3.5 M,
THF, rt, 76%; (iv) NaOH, EtOH/H2O, reflux, 67%.
R= C11H23
10
Figure 1. Structures of rhodopeptins and the cyclobutane analogue 10.
indication of instability.10 First, we verified using a
model compound that rapid access to the corresponding
aminoalcohol from a cyclobutane dihydrouracil deriva-
tive was feasible. Compound 6 was obtained by a [2+2]
photocyclization reaction between 6-methyluracil and
ethylene, as previously described.8 Reduction using an
excess of sodium borohydride in water at room
temperature,11 followed by acidification with an acidic
ion exchange resin, gave ureidoalcohol 7 in 72% yield
(Scheme 2). Unexpectedly, the diazotization reaction
under standard conditions (1 equiv NaNO2 in 3.5 M
HCl solution)8 gave aminoalcohol 9 in poor yield (15%)
together with cyclic carbamate 8. The best compromise
to obtain the cyclobutane aminoalcohol was to generate
selectively the cyclic intermediate 8 via a diazotization
reaction carried out in tetrahydrofuran as co-solvent
(76% isolated yield). The cyclic carbamate was hydro-
lyzed in basic conditions to provide aminoalcohol 9
in good yield (67%).12 In contrast to its b-aminoacid
homologue, 9 is perfectly stable.
The appropriate b-aminoalcohol bearing a linear carbon
chain (C11H23) 20 was prepared from 6-undecyluracil 14
(Scheme 3). This latter compound was synthesized in
three steps from ethyl acetoacetate 11, inspired by the
literature procedures.15 The [2+2] photocycloaddition
protocol provided the cis-cyclobutane adduct 15 from
6-undecyluracil 14 and ethylene in a satisfying 86% yield.16
Fears for the stability of 16 were confirmed: a retro-Man-
nich reaction occurs as previously mentioned. The treat-
ment of heterocycle 15 with 0.5 M aqueous sodium
hydroxide solution at reflux gave d-ketoacid 17 directly
and in quantitative yield.17 As an aside, it is interesting
to note that this whole sequence constitutes a double
homologation of b-ketoester 12 to give d-ketoacid 17.
The inherent instability of aminoacid 16 was clearly
demonstrated, so its surrogate, aminoalcohol 20, was
prepared in three steps. Bicyclic intermediate 15 was
reduced using an excess of NaBH4 in a THF/water
mixture (due to the poor solubility of compound 15 in
water alone) to give ureidoalcohol 18 in excellent yield
(98%). The diazotization reaction of urea 18 was carried
out in aqueous HCl/THF solution to provide cyclic car-
bamate 19 with a yield of 77%. Finally, aminoalcohol 20
was obtained by treatment of compound 19 with base
(93% yield).18
Rhodopeptins are cyclic lipopeptides recently isolated
from the bacterial species Rhodococcus sp. Mer-
N1033.13 They are composed of three regular a-amino
acids and one b-amino acid with a long lipophilic side
chain borne on the b-carbon (Fig. 1). These cyclic pep-
tides show antifungal activities against Candida albicans
and Cryptococcus neoformans. Because of the unique
structure and biological activity, several analogues and
peptidomimetics were synthesized for structure–activity
relationship (SAR) studies.14 Ohta et al. have shown
that a lipophilic side chain ranging from 9 to 11 carbons,
a bulky lipophilic group (Val) and a basic site (Lys or
Orn) are crucial for antifungal activity. The spatial
arrangement of these three side chains depends on the
conformation of the cyclopeptide. The replacement of
the natural b-aminoacid by a cyclobutane b-aminoacid
bearing a straight carbon chain is expected to confer
conformational rigidity upon the macrocycle, and possi-
bly facilitate the macrocyclization step of its synthesis.
We decided to test the hypothesis of using an amino-
alcohol derivative as a stable synthetic equivalent of a
C2-substituted cis-cyclobutane b-aminoacid by conduct-
ing the multi-step synthesis of the rhodopeptin analogue
10 (Fig. 1).
Racemic compound 20 was coupled with Boc-L-valine
using EDC/HOBt to provide compound 21 in 61% yield,
without recourse to primary alcohol protection (Scheme
4). This compound was obtained as a mixture of diaste-
reoisomers, which unfortunately could not be separated
by chromatography. With the lone pair on the nitrogen
effectively deactivated due to its participation in the pep-
tide bond, subsequent oxidation of the primary alcohol
of 21 should give a relatively stable dipeptide derivative
of 16. We selected NaIO4/cat. RuCl3 as the oxidation
system, since it had already been shown to facilitate
the formation of peptides containing 2,3-methanoaspar-
tic acid.19 In the event, dipeptide acid 22 was obtained
smoothly; no retro-Mannich type ring-opening reactiv-
ity of this product was observed. Coupling of 22 with
a Gly-Lys derivative using EDC/HOBt provided tetra-
peptide 23 (31% yield in 2 steps from compound 21),
again as a diastereoisomeric mixture. Methyl ester

O. Roy et al. / Tetrahedron Letters 47 (2006) 5981–5984
5983
CO2H
O
O
O
O
CO2Et
H
H
ii
iv
iii
v
O
O
H
NH
NH
NH
O
NH2
10
N
H
S
N
H
O
N
O
10
10
10
10H
10
12
17
16
14
13
15
i
OH
H
H
H
vi
viii
CO2Et
NH2
vii
O
OH
NH2
O
N
10 H
O
N
10H
O
10
11
18
19
20
Scheme 3. Reagents and conditions: (i) NaH, BuLi, C10H21Br, THF, 0 ꢁC; (ii) thiourea, EtONa, EtOH, reflux; (iii) ClCH2CO2H, H2O, reflux, then
concd HCl, reflux, 33% (3 steps); (iv) hm, ethylene, acetone, rt, 86%; (v) NaOH, 0.5 M, 100 ꢁC, 100%; (vi) NaBH4, H2O/THF 1:1, rt, 98%; (vii)
NaNO2, HCl 3.5 M, THF, rt, 77%; (viii) NaOH, EtOH/H2O, reflux, 93%.
O
NHCbz
OH
O
H
N
O
CO2H
O
N
CO2Me
i
OH
NH2
iii
ii
O
H
NHBoc
NHBoc
NHBoc
N
H
N
H
N
10H
10
10
10
20
21
22
23
O
O
H
N
O
H
H
N
O
H
N
vii
N
iv-vi
NH2.TFA
NHCbz
N
H
N
H
N
N
O
10H
O
10H
O
O
24
10
Scheme 4. Reagents and conditions: (i) Boc-Val-OH, EDC, HOBt, CH2Cl2, rt, 61%; (ii) cat. RuCl3, NaIO4, CCl4/CH3CN/H2O; (iii) H-Gly-Ne -Cbz-
Lys-OMeÆTFA salt, Et3N, EDC, HOBt, CH2Cl2, rt, 31% (2 steps); (iv) LiOH, H2O/THF/CH3OH, 0 ꢁC, then rt, 99%; (v) TFA, CH2Cl2, 0 ꢁC, then rt,
98%; (vi) DPPA, NaHCO3, DMF, 5 ꢁC, then rt, 71%; (vii) Pd/C 10%, TFA, CH2Cl2, 100%.
hydrolysis and Boc-group removal gave an intermediate
tetrapeptide, which was macrocyclized using DPPA to
the cyclic tetrapeptide 24 with a yield of 69% for three
steps.20 This is the first macrolactamization between
the Val and Lys residues of a rhodopeptin analogue,
and the hairpin-like turn imposed by the cis-cyclobutane
b-aminoacid component4c probably facilitates the reac-
tion. Quantitative removal of the Cbz group by hydro-
genolysis in the presence of palladium on carbon in
trifluoroacetic acid/dichloromethane media furnished
the target rhodopeptin analogue 10 as its TFA salt.21
References and notes
1. (a) North, M. J. Pept. Sci. 2000, 6, 301–313; (b) Cheng, R.
P.; Gellman, S. H.; DeGrado, W. F. Chem. Rev. 2001, 101,
3219–3232.
2. (a) Fulo¨p, F. Chem. Rev. 2001, 101, 2181–2204; (b) Kuhl,
¨
´
A.; Hahn, M. G.; Dumic, M.; Mittendorf, J. Amino Acids
2005, 29, 89–100; (c) Gnad, F.; Reiser, O. Chem. Rev.
2003, 103, 1603–1623.
3. (a) Appella, D. H.; Christianson, L. A.; Karle, I. L.;
Powell, D. R.; Gellman, S. H. J. Am. Chem. Soc. 1999,
121, 6206–6212; (b) Barchi, J. J., Jr.; Huang, X.; Appella,
D. H.; Christianson, L.; Durell, S. R.; Gellman, S. H. J.
Am. Chem. Soc. 2000, 122, 2711–2718; (c) Raguse, T. L.;
Porter, E. A.; Weisblum, B.; Gellman, S. H. J. Am. Chem.
Soc. 2002, 124, 12774–12785; (d) Appella, D. H.; Chris-
tianson, L. A.; Klein, D. A.; Powell, D. R.; Huang, X.;
Barchi, J. J.; Gellman, S. H. Nature (London) 1997, 387,
381–384; (e) Appella, D. H.; Christianson, L. A.; Klein, D.
A.; Richards, M. R.; Powell, D. R.; Gellman, S. H. J. Am.
In conclusion, this initial study shows that an inherently
unstable C2-substituted cis-cyclobutane b-aminoacid
can be incorporated into a stable peptide, through use
of its stable aminoalcohol surrogate. On this premise,
we are currently developing access to enantiomerically
pure rhodopeptin analogues for conformational studies
and determination of biological activity.
Chem. Soc. 1999, 121, 7574–7581; (f) Fulo¨p, F. Stud. Nat.
¨
Prod. Chem. 2000, 22, 273–306.
4. (a) Izquierdo, S.; Kogan, M. J.; Parella, T.; Moglioni, A.
Acknowledgments
G.; Branchadell, V.; Giralt, E.; Ortuno, R. M. J. Org.
˜
`
Chem. 2004, 69, 5093–5099; (b) Izquierdo, S.; Rua, F.;
We acknowledge B. Saby, beneficiary of a Pfizer Global
Research Summer Scholarship, for carrying out some
initial experiments.
Sbai, A.; Parella, T.; Alvarez-Larena, A.; Branchadell, V.;
Ortuno, R. M. J. Org. Chem. 2005, 70, 7963–7971; (c)
Martın-Vila, M.; Muray, E.; Aguado, G. P.; Alvarez-Larena,
˜
´
`

5984
O. Roy et al. / Tetrahedron Letters 47 (2006) 5981–5984
´
A.; Branchadell, V.; Minguillon, C.; Giralt, E.; Ortuno, R.
˜
Lett. 2000, 2, 977–980; (d) Nakayama, K.; Kawato, H. C.;
Inagaki, H.; Ohta, T. Org. Lett. 2001, 3, 3447–3450; (e)
Kawato, H. C.; Nakayama, K.; Inagaki, H.; Ohta, T. Org.
Lett. 2001, 3, 3451–3454; (f) Durham, T. B.; Miller, M. J.
J. Org. Chem. 2003, 68, 35–42.
M. Tetrahedron: Asymmetry 2000, 11, 3569–3584; (d)
Ortuno, R. M.; Moglioni, A. G.; Moltrasio, G. Y. Curr.
Org. Chem. 2005, 9, 237–259.
˜
5. (a) Kennewell, P. D.; Matharu, S. S.; Taylor, J. B.;
Westwood, R.; Sammes, P. G. J. Chem. Soc., Perkin
Trans. 1 1982, 2563–2570; (b) Aitken, D. J.; Gauzy, C.;
Pereira, E. Tetrahedron Lett. 2002, 43, 6177–6179.
15. (a) Beijer, F. H.; Sijbesma, R. P.; Vekemans, J. A. J. M.;
Meijer, E. W.; Kooijman, H.; Spek, A. L. J. Org. Chem.
1996, 61, 6371–6380; (b) Gershon, H.; Braun, R.; Scala, A.
J. Med. Chem. 1963, 6, 87–89.
´
`
´
6. (a) Martın-Vila, M.; Minguillon, C.; Ortuno, R. M.
˜
1
Tetrahedron: Asymmetry 1998, 9, 4291–4294; (b)
Izquierdo, S.; Martın-Vila, M.; Moglioni, A. G.;
Branchadell, V.; Ortuno, R. M. Tetrahedron: Asymmetry
2002, 13, 2403–2405; (c) Bolm, C.; Schiffers, I.; Atodi-
resei, I.; Hackenberger, C. P. R. Tetrahedron: Asymme-
try 2003, 14, 3455–3467; (d) Gauzy, C.; Pereira, E.;
Faure, S.; Aitken, D. J. Tetrahedron Lett. 2004, 45,
7095–7097.
16. Compound 15: White powder, mp 132–133 ꢁC, H NMR
´
`
(400 MHz, CDCl3) d 0.90 (t, J = 7.2 Hz, 3H), 1.30 (m,
18H), 1.66 (m, 2H), 2.17 (m, 2H), 2.30 (m, 1H), 2.39 (m,
1H), 3.02 (dd, J = 7.0, 9.4 Hz, 1H), 5.66 (s, 1H), 7.92 (s,
1H). 13C NMR (100 MHz, CDCl3) d 14.1 (CH3), 21.7
(CH2), 22.7 (CH2), 29.3 (2CH2), 29.5 (3CH2), 29.6 (2CH2),
31.9 (CH2), 34.1 (CH2), 40.3 (CH2), 40.7 (CH), 57.0 (C),
152.7 (C), 165.1 (C). ESMS m/z 317 [MNa+]. Elemental
analysis calcd for C17H30N2O2: C, 69.35; H, 10.27; N,
9.51. Found: C, 69.71; H, 10.27; N, 9.53.
˜
7. (a) Mittendorf, J.; Kunisch, F.; Matzke, M.; Militzer,
H.-C.; Schmidt, A.; Scho¨nfeld, W. Bioorg. Med. Chem.
Lett. 2003, 13, 433–436; (b) Mitsudo, T.-A.; Zhang, S.-W.;
Satake, N.; Kondo, T.; Watanabe, Y. Tetrahedron Lett.
1992, 33, 5533–5536; (c) Yuan, P.; Driscoll, M. R.;
Raymond, S. J.; Hansen, D. E.; Blatchly, R. A. Tetra-
hedron Lett. 1994, 35, 6195–6198.
8. Gauzy, C.; Saby, B.; Pereira, E.; Faure, S.; Aitken, D. J.
Synlett 2006, 1394–1398.
9. Aitken, D. J.; Gauzy, C.; Pereira, E. Tetrahedron Lett.
2004, 45, 2359–2361.
17. d-Ketoacid 17 was identified by comparison of 1H and 13
C
NMR data with the literature: Utaka, M.; Watabu, H.;
Takeda, A. J. Org. Chem. 1986, 51, 5423–5425.
18. Compound 20: Yellow oil, 1H NMR (400 MHz, CDCl3) d
0.87 (t, J = 7.0 Hz, 3H), 1.25 (m, 18H), 1.51 (m, 2H), 1.63
(m, 1H), 1.72–1.85 (m, 2H), 1.94 (m, 1H), 2.19 (m, 1H),
2.53 (ls, 3H), 3.76 (m, 2H). 13C NMR (100 MHz, CDCl3) d
13.9 (CH3), 15.7 (CH2), 22.5 (CH2), 23.2 (CH2), 29.2
(CH2), 29.5 (4CH2), 29.9 (CH2), 31.7 (CH2), 33.3 (CH2),
42.7 (CH2), 44.1 (CH), 57.0 (C), 63.2 (CH2). HRMS
(ES+): m/z calcd for C16H34NO [MH]+: 256.2640; found:
256.2641.
´
10. (a) Santana, L.; Teijeira, M.; Teran, C.; Uriarte, E. Bull.
Soc. Chim. Belg. 1997, 106, 227–228; (b) Koch, C.-J.;
Ho¨fner, G.; Polborn, K.; Wanner, K. T. Eur. J. Org.
Chem. 2003, 2233–2242; (c) Katagiri, N.; Sato, H.;
Kaneko, C. Chem. Pharm. Bull. 1990, 38, 288–290.
11. (a) Kunieda, T.; Witkop, B. J. Am. Chem. Soc. 1971, 93,
3493–3499; (b) Kunieda, T.; Witkop, B. J. Am. Chem. Soc.
1971, 93, 3478–3487.
19. Godier-Marc, E.; Aitken, D. J.; Husson, H.-P. Tetra-
hedron Lett. 1997, 38, 4065–4068.
20. Analytical HPLC analysis of cyclic tetrapeptide 24 gave
insufficient base-line separation, precluding isolation by
preparative HPLC. Column AcclaimTM 120 C18 3 lm
(2.1 · 100 mm); mobile phase CH3CN–H2O 70/30 +
HCO2H 1&; temperature 30 ꢁC; flow rate 0.25 mL/min;
UV detection at 210 nm. tR: 11.04 and 11.65 min.
1
12. Compound 9: Yellow oil, H NMR (400 MHz, CDCl3) d
1.28 (s, 3H), 1.47 (m, 1H), 1.76 (m, 1H), 1.86 (m, 2H), 2.16
(m, 1H), 2.98 (ls, 3H), 3.68 (dd, J = 11.4, 4.6 Hz, 1H), 3.77
(dd, J = 11.4, 8.5 Hz, 1H). 13C NMR (100 MHz, CDCl3) d
15.5 (CH2), 29.7 (CH3), 35.0 (CH2), 45.7 (CH), 54.4 (C),
63.4 (CH2). HRMS (ES+): m/z calcd for C6H14NO [MH]+:
116.1075; found: 116.1084.
21. Compound 10: White solid, mp 117–119 ꢁC, 13C NMR
(100 MHz, CD3OD) d 14.4 (CH3), 17.4 (CH3 Val), 18.0
(CH2), 19.1 (CH3 Val), 19.5 (CH2), 20.0, 20.3, 23.7, 24.0,
24.1, 24.6, 24.7, 27.9, 30.5, 30.7, 30.7, 30.9, 31.2, 31.6 (CH
Val), 31.9, 32.2, 33.1, 39.0, 40.4 (CH2 Lys), 40.6, 44.2 (CH2
Gly), 45.3 (CH2 Gly), 50.0 (CH), 50.3 (CH), 57.7 (CH
Lys), 57.9 (CH Lys), 59.1 (CH Val), 60.3 (C), 61.4 (CH
Val), 62.5 (C), 171.6, 172.3, 172.6, 173.7, 174.4, 174.6,
175.5, 176.0. HRMS (ES+): m/z calcd for C29H54N5O4
[MH]+: 536.4176; found: 536.4180. Analytical HPLC:
Column AcclaimTM 120 C18 3 lm (2.1 · 100 mm); mobile
phase CH3CN–H2O 45/55 + HCO2H 0.5&; temperature
30 ꢁC; flow rate 0.25 mL/min; UV detection at 210 nm.
Single peak observed, tR: 12.81 min.
13. (a) Chiba, H.; Agematu, H.; Kaneto, R.; Terasawa, T.;
Sakai, K.; Dobashi, K.; Yoshioka, T. J. Antibiot. 1999, 52,
695–699; (b) Chiba, H.; Agematu, H.; Dobashi, K.;
Yoshioka, T. J. Antibiot. 1999, 52, 700–709.
14. (a) Chiba, H.; Agematu, H.; Sakai, K.; Dobashi, K.;
Yoshioka, T. J. Antibiot. 1999, 52, 710–720; (b) Kawato,
H. C.; Nakayama, K.; Inagaki, H.; Nakajima, R.;
Kitamura, A.; Someya, K.; Ohta, T. Org. Lett. 2000, 2,
973–976; (c) Nakayama, K.; Kawato, H. C.; Inagaki, H.;
Nakajima, R.; Kitamura, A.; Someya, K.; Ohta, T. Org.