Synthesis and antifungal activities of cyclic octa-lipopeptide burkholdine analogues

Article abstract of DOI:10.1016/j.bmcl.2013.04.091

Synthesis and antifungal activity of cyclic octapeptide derivatives of burkholdines are described. To construct cyclic octapeptides, the combination of Fmoc-SPPS and cyclization with DIC/HOBt in the solution phase was employed. Synthesized peptides were evaluated for antifungal activity with MIC values against Saccharomyces cerevisiae, Aspergillus oryzae, and Candida viswanathii. As a result, the lipid side chain and the stereochemistry of each amino acid of Bk-1097 analogues significantly affected antifungal activity.

Full text of DOI:10.1016/j.bmcl.2013.04.091

Bioorganic & Medicinal Chemistry Letters 23 (2013) 4244–4247
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and antifungal activities of cyclic octa-lipopeptide
burkholdine analogues
a
a
b
Hiroyuki Konno a, , Yusuke Otsuki , Kenta Matsuzaki , Kazuto Nosaka
⇑
a Department of Biological Engineering, Graduate School of Science and Technology, Yamagata University, 4-3-16 Jonan, Yonezawa, Yamagata 992-8510, Japan
b Department of Chemistry, Hyogo College of Medicine, Nishinomiya, Hyogo 663-8501, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Synthesis and antifungal activity of cyclic octapeptide derivatives of burkholdines are described. To con-
struct cyclic octapeptides, the combination of Fmoc-SPPS and cyclization with DIC/HOBt in the solution
phase was employed. Synthesized peptides were evaluated for antifungal activity with MIC values
against Saccharomyces cerevisiae, Aspergillus oryzae, and Candida viswanathii. As a result, the lipid side
chain and the stereochemistry of each amino acid of Bk-1097 analogues significantly affected antifungal
activity.
Received 8 April 2013
Revised 27 April 2013
Accepted 30 April 2013
Available online 14 May 2013
Keywords:
Antifungal
Ó 2013 Published by Elsevier Ltd.
Lipopeptide
Burkholdine
Structure activity relationship study
Burkholdines (Bks) were isolated from a culture of Burkholderia
ambifaria 2.2N by Schmidt’s group.1 Bk-1097 (1), one of the burk-
holdines, is a cyclic octapeptide composed of eight amino acids,
including b-hydroxytyrosine, b-hydroxyasparagine and new fatty
acyl amino acid (FAA). Although the absolute stereochemistry of
b-hydroxytyrosine residue and FAA moiety of Bk-1097 (1) has re-
cently been reported to be 2R,3R and 3R,5R,6S,7S, respectively,2
there is no complete assignment for identification using NMR
experimentation with synthetic compounds and/or X-ray analysis.
In contrast, potent biological activity of 1.6 lg/ml has been shown
against the yeast Saccharomyces cerevisiae and the fungus Aspergil-
lus niger. These activities are 16-fold more potent than amphoter-
icin B,3,4 which is used in antifungal therapy. Burkholdines,
together with xylocandins5 and occidiofungins,6,7 are promising
antifungal agents that produce fungal cell membrane defects,
although the inhibitory mechanism and target enzyme are unre-
solved. Since there are many peptides with a variety of bioactivi-
ties, we are interested in the appearance of potent antifungal
activities by the combination of amino acids.8,9 Especially, cyclic
peptides tend to show the potent activities because of employment
of the rigid framework and preservation of sidechain trajectories.
We herein planned to synthesize and evaluate various simplified
Bk-1097 analogues to find cyclic peptides with more potent anti-
fungal activity (Fig. 1).
Figure 1. Bk-1097 (1).
Bk-1097 analogues were synthesized via Fmoc solid phase pep-
tide synthesis (Fmoc-SPPS) using L- or D-usual amino acids in place
of b-hydroxytyrosine, b-hydroxyasparagine and FAA. Synthesized
peptides were evaluated by the minimum inhibitory concentration
(MIC) against Saccharomyces cerevisiae, Aspergillus oryzae and Can-
dida viswanathii (Fig. 2).
As depicted in Scheme 1, the designed cyclic peptides were syn-
thesized. Linear precursors were prepared by Fmoc-SPPS using 2-
chlorotrityl chloride resin10,11 followed by cleavage from the resin
with 20% HFIP/CH2Cl2. Intramolecular cyclization of the linear
peptide and then final deprotection yielded designed cyclic
lipopeptide after purification by preparative HPLC.
⇑
Corresponding author. Tel./fax: +81 (0)238 26 3131.
E-mail address: konno@yz.yamagata-u.ac.jp (H. Konno).
0960-894X/$ - see front matter Ó 2013 Published by Elsevier Ltd.
http://dx.doi.org/10.1016/j.bmcl.2013.04.091

H. Konno et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4244–4247
4245
condition of DIC/DIPEA/HOBtÁH2O for 20 h in 55% yield. The macro-
cyclization between N-terminus and C-terminus was monitored by
RP-HPLC and ESI TOF-MS. According to the HPLC profile, the hydro-
philic peak as the linear peptide gradually dissolved for 20 h and a
subsequent more hydrophobic peak appeared as a single com-
pound. It showed no by-products as dimeric or trimeric peptides. Fi-
nally, the desired cyclic peptide (7) was obtained by deprotection
using TFA/TIPS/H2O (95:5:5) and subsequent purification of pre-
parative RP-HPLC in 57% yield.12,13 In addition to these anticipated
issues, an unexpected problem arose when final deprotection with
TFA was attempted, namely, the observation of three abundant
by-products with masses 17, 34 and 51 daltons smaller than the de-
sired product. We anticipated that intramolecular lactonizations
between the amide groups of Asn side chains and the hydroxyl
groups of Ser or Tyr residues proceeded to afford the hydrophobic
products on a detectable scale. To attempt the several conditions
for deprotection of the cyclic peptides, we found that TIPS as a scav-
enger was effective to avoid side reactions. Preparation of cyclic
peptides (8–25) was similar to the method in Scheme 3 (Table 1).
As the sequence of octapeptides, intramolecular cyclization of
Figure 2. Bk-analogues with D/L-amino acids.
To replace the FAA residue with simplified lipo-b-amino acid,
we designed N-lauryl-3-amino-4-carbamoylpropanoic acid (LAP)
(2), which was coupled with linear peptides as the last component
of the solid support. To prepare both enantiomers of LAP residues,
Fmoc-L/D-Asp(tBu)-OH was condensed with dodecylamine using
BOP/Et3N in DMF to give N-lauryl aspartic acid derivatives, and
subsequent removal of tBu by TFA afforded both enantiomers of
Fmoc-LAP-OH (3) in moderate yield (Scheme 2).
Synthesis of cyclic peptide (7) is shown in Scheme 3. Fmoc-
Asp(Trt)-OH was linked onto 2-chlorotrityl chloride resin by ester-
ification with DIPEA to obtain resin (4). After removing the Fmoc
group by treatment with 20% piperidine/DMF, the remaining 7 ami-
no acids were coupled by the general conditions. Resin was treated
with 20% HFIP/CH2Cl2 to provide linear peptide (5) in 64% yield from
loading on the resin. Linear peptide (5) was cyclized under the
linear peptides was effective with
D
-Ser-
D
-Tyr (entries 8 and 10)
or -Tyr- -Asn (entry 15) motifs. A couple of D
D
D
-amino acids on
the peptide sequences induce the formation of the turn structure
and subsequent intra-molecular macro-cyclization proceeded
smoothly. However, cyclization of linear peptides with
D-Ser-D-
Tyr- -Asn sequences (entries 3, 17 and 19) gave moderate yield.
D
Stereochemistry of LAP units, which are the sole b-amino acid res-
idues, did not affect the cyclization processes (Table 1).
The antifungal activities of burkholdine analogues (7–25) against
S. cerevisiae, A. oryzae and C. viswanathii are shown in Table 214
Scheme 1. Synthetic plan for cyclic lipopeptides.
Scheme 2. Synthesis of lipo-amino acids, Fmoc-LAP-OH (3).
Scheme 3. Synthesis of cyclic octalipopeptide (7).

4246
H. Konno et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4244–4247
Table 1
Sequences of synthesized cyclic octapeptides and their chemical yields
Entry
Peptide
Xaa4
Xaa3
Ser
Xaa2
Xaa1
Asn
Fmoc-SPPS (%)
Cyclization
(h)
Deprotection (%)
Total yield (%)
(%)
1
2
8
9
b-ALa
b-ALa
Tyr
Tyr
86
55
62
30
55
45
25
44
12
11
D
D
D
-Ser
-Ser
-Ser
D-Asn
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
7
b-ALa
68
38
57
51
45
32
60
61
23
15
50
62
52
50
45
54
64
25
23
21
13
19
18
22
15
17
23
22
14
18
26
21
16
20
46
63
46
35
32
73
54
66
59
64
69
69
79
34
46
64
55
6
54
68
64
97
54
41
24
76
76
27
42
27
50
47
55
57
2
13
8
D-Tyr
D-Asn
4
Tyr
Tyr
Tyr
Tyr
Asn
Asn
Asn
Asn
Asn
Asn
Asn
Asn
D
-LAP
-LAP
5
Ser
D
6
4
L
-LAP
-LAP
D
-Ser
7
Ser
8
L
8
18
21
14
8
D
-LAP
-LAP
D
-Ser
D
D
D
D
-Tyr
9
Ser
D
-Tyr
-Tyr
-Tyr
10
11
12
13
14
15
16
17
18
19
L
-LAP
-LAP
D
-Ser
Ser
L
Ser
Ser
Ser
Ser
Tyr
21
30
18
11
5
D
-LAP
-LAP
D
D
D
D
D
D
D
D
-Asn
D
D-Tyr
-Asn
-Asn
-Asn
-Asn
-Asn
-Asn
-Asn
Tyr
L
-LAP
-LAP
L
D-Tyr
Tyr
D
-LAP
-LAP
D
D
D
D
-Ser
8
D
-Ser
-Ser
-Ser
D-Tyr
Tyr
19
20
L
-LAP
-LAP
L
D-Tyr
Analogues (8), (9) and (10) with the involvement of b-Ala shown no
antifungal activities against S. cerevisiae. Analogues (20), (25) and (7)
lower antifungal activities against Candida than against Saccharomy-
ces or Aspergillus (Table 2).
exhibited great potency with the MIC values of 50, 50, and 25
l
g/mL,
For more potent antifungal activity, replacement with
stead of b-Ala or -LAP gave antifungal compounds. In contrast,
comparison of -Tyr and -Tyr produced no significant difference
LLAP in-
respectively. It is suggested that D-form stereochemisty of Asn res-
idue in Xaa1 position was essential and, additionally, stereochemis-
try of Ser residue in Xaa3 and LAP was likely to show opposite
stereochemistry. Antifungal activities of other sequences were weak
and therefore we predicted that there was a correlation between the
stereochemistry of naturally produced burkholdines and those of
analogues. Although the MIC values of A. oryzae have the same ten-
dency as those of S. cerevisiae, antifungal activities against C. visw-
anathii were not shown. Burkholdines also tended to result in
D
L
D
in antifungal activities (Fig. 3).
In conclusion, we synthesized 29 Bk-1097 analogues through
the use of Fmoc-SPPS. From the results of the MIC assay, the lipid
chain at the Xaa4 position and stereochemistry at Xaa1, Xaa3,
and Xaa4 positions were found to be important for antifungal
activity. We will plan a structure–activity relationship study for
more potent antifungal activity.
Table 2
MIC values for cyclic peptides (7–25)
Peptide
MIC (
l
g/mL)a
Peptide
MIC (l
g/mL)a
S. cerevisiae
A. oryzae
C. viswanathii
S. cerevisiae
A. oryzae
C. viswanathii
8
9
10
>800
800
800
>800
>800
800
800
>800
>800
17
18
>800
>800
800
400
>800
800
19
20
21
22
800
50
>800
800
800
100
400
400
>800
>800
800
11
12
13
14
400
200
400
800
>800
>800
>800
200
>800
>800
>800
>800
>800
23
24
25
7
400
400
50
400
800
100
50
>800
>800
400
15
16
800
800
800
>800
>800
>800
25
800
a
MIC value of Amphotericin B control against S. cerevisiae was 0.8 lg/ml.

H. Konno et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4244–4247
4247
Figure 3. Structure–activity relationship study of Bk analogues.
(30
l
l, 0.192 mmol)/HOBt (29 mg, 0.192 mmol)/DIPEA (22
l
l, 0.192 mmol) for
Acknowledgments
2 h. The resulting resin was treated with HFIP/CH2Cl2 (1:4, 5 ml) for 2 h and the
mixture was filtrated. After evaporation of the HFIP, ether was added, and the
resulting precipitate was purified by preparative HPLC (CH3CN/H2O = 88:12) to
yield octapeptide (5) (79 mg, 0.0409 mmol, 64%) as a white powder. To a
solution of octapeptide (5) (79 mg, 0.0409 mmol) in DMF (4.5 ml) were added
We thank Professor Shigekazu Yano (Yamagata University) for
useful discussions and the gift of A. oryzae strain 1-Murasaki-1
(Bio’s) and C. viswanathii strain NBRC10321.
DIC (32 ll, 0.204 mmol), HOBt (31 mg, 0.204 mmol) and DIPEA (21 ll,
0.123 mmol) and the mixture was stirred for 20 h at room temperature.
After adding AcOEt and H2O, the mixture was separated. The organic layer was
washed with brine, dried over MgSO4, filtered, and concentrated in vacuo. The
residue was purified by column chromatography on silica gel (95:5 CHCl3/
MeOH) to give the cyclic peptide (6) as a white powder. Cyclicpeptide (6)
(43 mg, 0.0225 mmol) was added to TFA/TIPS/H2O (95:5:5; 2 ml) and the
mixture was stirred for 2 h at room temperature. After evaporation of TFA,
ether was added, and the resulting precipitate was purified by preparative
References and notes
1. Tawfik, K.; Jeffs, P.; Bray, B.; Dubay, G.; Falkinham, J. O., III; Mesbah, M.;
Youssef, D.; Khalifa, S.; Schmidt, E. W. Org. Lett. 2010, 12, 664.
2. Lin, Z.; Falkinham, J. O., III; Tawfik, K. A.; Jeffs, P.; Bray, B.; Dubay, G.; Cox, J. E.;
Schmidt, E. W. J. Nat. Prod. 2012, 75, 1518.
3. (a) Croatt, M. P.; Carreia, E. M. Org. Lett. 2011, 13, 1390; (b) Volmer, A. A.;
Szpilman, A. M.; Carreira, E. M. Nat. Prod. Rep. 2010, 27, 1329.
4. Ernst, B.; Magnani, J. L. Nat. Rev. Drug Disc. 2009, 8, 661.
HPLC (CH3CN/H2O = 55.8:44.2) to yield
0.0129 mmol, 57%) as a white powder.
a cyclicoctapeptide (7) (13 mg,
13. Analytical HPLC using a Cosmosil 5C18 (10 mm  250 mm) column: linear peptide
(5): rt 11.96 min (CH3CN gradient; 80–100% in 30 min). ESI TOF-MS. Calcd
1933.36 for C114H138N12O16Na; found: 1933.36 for [M+H]+. Cyclic peptide (6):
32.59 min (CH3CN gradient; 80–100% in 30 min and then 100% CH3CN for
10 min). ESI TOF-MS. Calcd 1936.01 for C114H136N12O15Na; found: 1936.34 for
[M+Na]+. Cyclic peptide (7): 7.06 min (CH3CN gradient; 40–70% in 30 min). ESI
TOF-MS. Calcd 1041.50 for C45H70N12O15Na; found: 1041.59 for [M+H]+.
14. Assay protocol: The minimal inhibitory concentration (MIC) of burkholdine
analogues was determined with broth by the usual twofold serial dilution
method. Growth media for S. cerevisiae strain X2180-1A, A. oryzae strain 1-
Murasaki-1 (Bio’s) and C. viswanathii strain NBRC10321 were YPD medium.
Before starting the MIC assay, the microbial cells were grown in broth for
2 days. The concentration of the burkholdine analogues started with 1 mg/L in
the medium containing 0.1% DMSO, and underwent twofold serial dilution.
Then, the serial twofold dilution solution was poured into a 96-well microtiter
plate and inoculated with microbial suspensions of about 5.0 Â 103 cells. The
MIC was defined as the lowest concentration of antimicrobial agent that
inhibited the development of visible growth after 24 h of incubation at 30 °C in
each case.
5. Bisacchi, G. S.; Hockstein, R. D.; Koster, W. H.; Parker, W. L.; Rathnum, M. L.;
Unger, S. E. J. Antibiot. 1987, 40, 1520.
6. Gu, G.; Smith, L.; Liu, A.; Lu, S. Appl. Environ. Microbiol. 2011, 77, 6189.
7. Lu, S.; Novak, J.; Austin, F. W.; Gu, G.; Ellis, D.; Kirk, M.; Wilson-Stanford, S.;
Tonelli, M.; Smith, L. Biochemistry 2009, 48, 8312.
8. Seto, Y.; Takahashi, K.; Matsuura, H.; Kogami, Y.; Yada, H.; Yoshihara, T.;
Nabeta, K. Biosci., Biotechnol., Biochem. 2007, 71, 1470.
9. Luo, J.; Wang, X.; Ma, L.; Kong, L. Bioorg. Med. Chem. Lett. 2007, 17, 4460.
10. Kikuchi, M.; Watanabe, Y.; Tanaka, M.; Akaji, K.; Konno, H. Bioorg. Med. Chem.
Lett. 2011, 21, 4865.
11. Kikuchi, M.; Nosaka, K.; Akaji, K.; Konno, H. Tetrahedron Lett. 2011, 52, 3872.
12. Synthesis of cyclic peptide (7): To a suspension of 2-Cl-trityl chloride resin
(40 mg, 1.6 mmol/g) in DMF (1 ml) were added Fmoc-Asn(Trt)-OH (115 mg,
0.192 mmol) and DIPEA (22 ll, 0.128 mmol). The mixture was stirred for 2 h at
25 °C. To a resultant resin was added 20% piperidine in DMF and the mixture
was stirred for 30 min at 25 °C. Fmoc-Ser(tBu)-OH (74 mg, 0.192 mmol), Fmoc-
Asn(Trt)-OH (115 mg, 0.192 mmol), Fmoc-Gly-OH (57 mg, 0.192 mmol), Fmoc-
D
-Asn(Trt)-OH
0.192 mmol), Fmoc-
(111 mg, 0.192 mmol) were successively condensed to this resin using DIC
(115 mg,
0.192 mmol),
Fmoc-
D
-Tyr(tBu)-OH
(88 mg,
D-Ser(tBu)-OH (74 mg, 0.192 mmol), Fmoc-LAP-OH