Structural revision of halipeptins: Synthesis of the thiazoline unit and isolation of halipeptin C

Article abstract of DOI:10.1016/S0040-4039(02)01209-1

The structural revision of the anti-inflammatory marine metabolites halipeptin A (1) and B (2) along with the isolation of the new related product halipeptin C (3) are reported. In particular, the heterocyclic portion of the molecule, incorrectly assigned as an oxazetidine ring, has now been characterised as a thiazoline unit by comparison of the spectral data of the natural products (1-3) with an appropriate synthetic model (10). GIAO calculated ¹³C NMR chemical shifts for oxazetidine and thiazoline model compounds provide additional support to the revised structure.

Full text of DOI:10.1016/S0040-4039(02)01209-1

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 5707–5710
Structural revision of halipeptins: synthesis of the thiazoline unit
and isolation of halipeptin C
Carmela Della Monica,c Antonio Randazzo,b Giuseppe Bifulco,a Paola Cimino,a Maurizio Aquino,c
Irene Izzo,c Francesco De Riccardisc,* and Luigi Gomez-Palomaa,*
aDipartimento di Scienze Farmaceutiche, Universita` di Salerno, via Ponte Don Melillo, 84084 Fisciano (SA), Italy
bDipartimento di Chimica delle Sostanze Naturali, Universita` di Napoli Federico II, via D. Montesano 49, 80131 Napoli, Italy
cDipartimento di Chimica, Universita` di Salerno, Via S. Allende, 84081 Baronissi (SA), Italy
Received 30 May 2002; revised 20 June 2002; accepted 25 June 2002
Abstract—The structural revision of the anti-inflammatory marine metabolites halipeptin A (1) and B (2) along with the isolation
of the new related product halipeptin C (3) are reported. In particular, the heterocyclic portion of the molecule, incorrectly
assigned as an oxazetidine ring, has now been characterised as a thiazoline unit by comparison of the spectral data of the natural
products (1–3) with an appropriate synthetic model (10). GIAO calculated 13C NMR chemical shifts for oxazetidine and thiazoline
model compounds provide additional support to the revised structure. © 2002 Elsevier Science Ltd. All rights reserved.
Halipeptins A (1) and B (2) are cyclic depsipeptides
displaying a potent in vivo anti-inflammatory activity
(60% of carrageenan induced edema reduction at a
intraperitoneal dose of 0.3 mg/Kg in mice).1 Their
structures feature the presence of common coded amino
acid residues (2×L-Ala) along with unusual units, such
as the polysubstituted decanoic acid HTMMD, N-
methyl-d-hydroxyisoleucine (NMe-dOH-Ile) and the
heterocyclic version of an a,a-disubstituted amino acid
which was incorrectly identified as a methyloxazetidine–
carboxylic acid residue (OMCA) mainly on the basis of
NMR and HRFABMS data.
The opportunity of re-examine the spectral data of
halipeptins came with the isolation, from the same
Vanuatu species of Haliclona, of a new minor related
compound, named halipeptin C (3). Although all the
NMR data (see Table 1 and HMBC correlations of 3)
could be readily interpreted assuming that halipeptin C
was a derivative of 2 bearing a L-NMeVal in place of
the NMe-dOH-Ile residue, HRESIMS data suggested
that its molecular formula contained an unexpected
sulphur atom. Indeed, the pseudomolecular ion peak of
3 at m/z 605.3360 (M+Na+, 605.33487 calculated for
C29H50N4NaO6S versus 605.35263 calculated for
C29H50N4NaO8) clearly favoured the molecular formula
containing a sulphur in place of two oxygen atoms.
* Corresponding authors. Tel.: +39-089-962811; fax: +39-089-962828
(L.G.P.); Tel.: +39-089-965230; fax: +39-089-965296; e-mail:
dericca@unisa.it; gomez@unisa.it
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)01209-1

5708
C. Della Monica et al. / Tetrahedron Letters 43 (2002) 5707–5710
Table 1. 1H, 13C and 15N NMR data of of halipeptin C
(3) (CDCl3, 500 MHz)
halipeptin family indicated that the original assignment
of the heterocycle portion as OMCA was incorrect.
Residue
lHa, mult., J in
lC
1H–13C HMBC
a
1
A literature search led to the conclusion that H and
Hz
13C NMR data of 1, 2 and 3 were consistent with the
presence of a methylthiazoline unit.2
Ala1
CO
a
169.6
49.9
18.2
4.71, quintet, 7.6
1.41, d, 7.6
7.01, d, 7.6
CO
Ca, CO
CO-NMeVal
In order to gather conclusive evidences for the struc-
tural revision of these molecules (see Table 2 for spec-
tral data) we synthesised a model thiazoline unit
(Schemes 1 and 2). The synthesis of D2-thiazoline frag-
ment was based on a four steps preparation of (R)-2-
methylcysteine hydrochloride (4), reported by
Pattenden and co-workers in 1993.3 Esterification of 4
with HCl/MeOH4 furnished a mixture the desired
methyl ester 6 and the corresponding dimeric amino
acid 5 in variable ratio (Scheme 1). Reduction5 of the
disulphide bond with PPh3 gave the expected (R)-2-
methylcysteine hydrochloride methyl ester (6) in 80%
overall yield from 4.
b
N
NMeVal
CO
a
b
Me-1
Me-2
NMe
169.9
64.8
26.4
17.7
21.0
30.5
4.98
2.56, m
0.93, d, 6.3
0.98, d, 6.3
2.81, s
Ca, Cb, Me-2
Ca, Cb, Me-1
Ca, CO-a-MeCys
a-MeCys
CO
a
Me-a
b1
171.9
83.6
23.0
43.6
1.49, s
3.31, d, 12.0
4.15, d, 12.0
Ca, Cb, CO
Coupling3 of
aminoproprionitrile (9), easily prepared in 74% yield
6 with the 2-(S)-t-butoxycarbonyl-
b2
N
from commercially available (L)-alaninamide hydro-
chloride (7) through N-Boc protection6 and efficient
dihydration of the amide moiety7 (Scheme 2), yielded
the desired 2-[1-(S)-tert-butoxycarbonylaminoethyl)]-
4 - (R) - methyl - 4,5 - dihydrothiazole - 4 - carboxylic acid
methyl ester 10.‡
Ala2
CN(S)
a
b
N
177.3
48.8
22.1
4.82, quintet, 7.4
1.50, d, 7.4
7.23, d, 7.4
CO-HTMHD
Ca, CO
CO-HTMHD
HTMHD
1
2
173.6
45.6
25.9
21.9
82.3
34.0
14.1
32.0
Final support to the revised structure came from the
comparison of GIAO (gauge including atomic orbitals)
calculated8 13C and 15N NMR chemical shift of four
model compounds, representing the two diastereomeric
couples of oxazetidine (11a and 11b) and thiazoline
(12a and 12b) units, respectively, with those of the
natural product 1 (see Table 2), following an approach
recently reported by our group.9
Me%-2
Me¦-2
3
1.12, s
1.20, s
4.71, d, 2.3
1.93, m
0.80, d, 6.5
1.30, m
Me¦-2, C-1, C-2, C-3
Me%-2, C-1, C-2, C-3
CO-Ala1
4
Me-4
5
C-3
1.34, m
1.45, m
6
34.1
1.50, m
3.55, quintet, 5.3
1.30, m
1.43, m
1.33
O
CH3
H
H
CH3
O
7
8
72.0
39.8
CH3
CH3
N
N
O
N
HN
HN
O
O
CH3
O
CH3
9
10
17.5
14.9
O
N
CH3
CH3
O
0.91, t, 6.0
C-8, C-9
CH3
H3C
CH3
11b
11a
a Chemical shift values are referred to CHCl3 (lH=7.26) and
13CHCl3 (lC=77.0) as internal standards.
O
CH
CH
3 S
3 S
H
CH3
H
N
Careful HRESIMS measurements revealed that this
was also the case for the parent compound halipeptin
A, for which a pseudomolecular ion peak could be
found at m/z 649.3628 (M+Na+, 649.3611 calculated
for C31H54N4NaO6S versus 649.3788 calculated for
C31H54N4NaO8).† The new molecular formulas of the
O
HN
N
N
HN
CH
3 CH3
12b
12a
N
CH3
O
O
CH3
CH3
H3C
‡ Physical data for compound 10: [h]D=−23 (c 1.0, CHCl3); 1H NMR
(400 MHz, CDCl3): l 1.40 (1 H, d, J=7.3 Hz, CHCH3), 1.42 (9 H,
s, (CH3)3), 1.50 (3 H, s, CH3), 3.14 (1 H, d, J=11.3 Hz, CHH), 3.75
(1 H, d, J=11.3 Hz, CHH), 3.76 (3 H, s, OCH3), 4.50 (1 H, m,
CHCH3), 5.23 (1 H, m, NH); 13C NMR (100 MHz, CDCl3): l
20.43 (CHCH3), 23.80 (CH3), 28.30 (×3, (CH3)3), 41.63 (CH2S),
49.18 (CH), 52.81 (OCH3), 79.76 (OC(CH3)3), 84.20
(C(CH3)CO2Me), 154.86 (NCO2), 173.53 and 174.55 (SCꢀN and
CO2Me); HRESIMS: m/z 303.1372 (303.1379 calculated for
C13H23N2O4S).
† Actually, our first HRMS measurements on 1 gave pseudomolecu-
lar ion peaks that were more in agreement with the molecular
formula C31H54N4O9 than with C31H54N4O7S, leading to a misin-
terpretation of its NMR data. However, new HRMS data on 1 and
3 were obtained on a superior instrumentation (API QSTAR
Pulsar) capable of reaching a resolution of about 20000 in that
particular mass range.

C. Della Monica et al. / Tetrahedron Letters 43 (2002) 5707–5710
5709
Table 2. Calculated and experimental 13C chemical shifts values (ppm) of natural and synthetic compounds
Residue
Calculated 13C chemical shifts
values
Residue
Calculated 13C chemical shifts
values
Experimental 13C chemical shifts
values
Oxazetidine unit
Thiazoline unit
Natural
compound
1
Synthetic
compound
10
11a/11b
12a/12b
Ala2
CO
a
Ala2
SCꢀN
a
182.2/180.4
44.2/47.8
18.1/18.8
178.2/178.1
48.0/47.8
22.1/21.5
177.3
48.5
22.0
173.5
49.2
20.4
–
b
b
N
−247.1/−247.4
N
−109.4/−107.2
−89.3
OMCA
CO
a
a-MeCys
168.4/170.6
77.8/76.1
57.0/55.0
23.0/22.0
CO
a
b
174.2/173.6
78.0/79.1
40.9/43.1
29.5/27.7
172.4
83.3
44.2
23.1
174.6
84.2
41.6
23.8
b
Me-a
Me-a
COOH
Cl- +H3N
HS
a
COOMe
Cl- +H3N
HS
COOMe
+
Cl- +H3N
5
S
4
6
2
b
Scheme 1. Reagents and conditions: (a) 5.0 equiv. of SOCl2, MeOH, reflux, 12 h, 80%; (b) 4.5 equiv. of PPh3, DME/MeOH/H2O,
7:2:1, 90°C, 12 h, 100%.
Notably, this set of data would slightly favour the R
absolute configuration at C-4 of the thiazoline unit, a
finding which is not surprising, considering that this
configuration is expected from the cyclisation of a
L-a-methylcystein amino acid residue and taking into
account that all the amino acid residues of halipeptins
appear to belong to the L series.
straightforward due to the very limited differences
existing between 2 and 3 (their H NMR spectra were
1
virtually superimposable, except for few signals). How-
ever, besides the opportunity to correct the assignment
of the OMCA unit, the structural study of 3 was useful
also for stereochemical reasons. In fact, HPLC Marfey
analysis10 of the acidic hydrolysate of 3 allowed us to
assign the N-MeVal residue to the L series, with the
important implication that the same absolute configura-
tion is very likely present in the NMe-dOH-Ile of 1 and
2.
Extraction and isolation of 3 followed the same proto-
col used for 1 and 2.1,§ Structure elucidation of 3 was
Acknowledgements
We thank Professor Shigeki Matsunaga (University of
Tokyo, Japan) for helpful discussions on the structural
revision of halipeptin A. Financial support was pro-
vided by the Universita` di Salerno, Consiglio Nazionale
delle Ricerche (CNR, Roma) and Ministero dell
Istruzione, dell’Universita` e della Ricerca Scientifica e
Tecnologica (MIUR, Roma) through the PRIN 1999–
2001 and 2001–2003 programs.
Scheme 2. Reagents and conditions: (a) 1.5 equiv. of Boc2O,
1.5 equiv. of Et3N, MeOH, 12 h, 90%; (b) 2.2 equiv. of
(CF3CO)2O, 4.4 equiv. of pyridine, THF, 0°C“rt, 3 h, 82%;
(c) 1.0 equiv. of 6, 1.0 equiv. of 9, 1.0 equiv. of Et3N, MeOH,
65°C, 12 h, 30%.
References
1. Randazzo, A.; Bifulco, G.; Giannini, C.; Bucci, M.; Deb-
itus, C.; Cirino, G.; Gomez-Paloma, L. J. Am. Chem.
Soc. 2001, 123, 10870–10876.
2. Carmeli, S.; Paik, S.; Moore, R. E.; Patterson, G. M. L.;
Yoshida, W. Y. Tetrahedron Lett. 1993, 34, 6681–6684.
3. Mulqueen, G. C.; Pattenden, G.; Whiting, D. A. Tetra-
hedron 1993, 49, 5359–5364.
§ Halipeptin C (3), white amorphous solid, [h]D=−30 (c 0.3, CHCl3),
was purified on a m-Bondapack C-18 column (7.8×300 mm) with
linear gradient elution, H2O/CH3OH, 75:25–0:100 in 30 min (tR=
20.5 min). HRESIMS: m/z 605.3360 (M+Na+, 605.33487 calculated
for C29H50N4NaO6S). For NMR data, see Table 1.

5710
C. Della Monica et al. / Tetrahedron Letters 43 (2002) 5707–5710
4. Ronald, G. W.; Malcolm, W. H.; Stammer, C. H. J. Org.
7. Kruijtzer, J. A. W.; Lefeber, D. J.; Liskamp, R. M. J.
Tetrahedron Lett. 1997, 38, 5335–5338.
Chem. 1969, 34, 576–580.
8. Ditchfield, R. Mol. Phys. 1974, 27, 789–807.
9. Barone, G.; Gomez-Paloma, L.; Duca, D.; Silvestri, A.;
Riccio, R.; Bifulco, G. Chem. Eur. J. 2000, 8, 3233–3239.
10. Marfey, P. Carlsberg Res. Commun. 1984, 49, 591–
596.
5. Corey, E. J.; Mehrotra, M. M. Tetrahedron Lett. 1988,
29, 57–60.
6. Yuste, F.; Ortiz, B.; Carrasco, A.; Peralta, M.; Quintero,
L.; Sanchez-Obregon, R.; Walls, F.; Garcia Ruano, J. L.
Tetrahedron: Asymmetry 2000, 11, 3079–3090.