Polytheonamides, Highly Cytotoxic Polypeptides
A R T I C L E S
a pair of mutually coupled methylenes and a ketone resonance.
On reduction with NaBH4, polytheonamide B afforded a 2:1
mixture of dihydropolytheonamides B1 and B2, both of which
exhibited an oxymethine signal. The COSY spectrum of
dihydropolytheonamide B1 displayed signals assignable to a
partial structure -CδH2-CγH2-CâH(OH)- [δH 3.80 (Hâ), 1.60,
1.40 (Hγ,γ′), 1.29, 1.17 (Hδ,δ′); δC 71.5 (Câ), 29.6 (Cγ), 39.0
(Cδ)]. Although the methyl region of the HMBC spectrum of
polytheonamide B was highly crowded, cross-peaks were
observed from a singlet methyl (δH 0.87) to a methylene carbon
at δ 36.6 (Cδ) and to a nonprotonated carbon at δ 29.6, thus
suggesting the linkage of Cδ to a tert-butyl group. HMBC cross-
peaks were also observed between the amide and methylene
protons of the N-terminal Gly residue and the CR carbon (δ
161.2) of the blocking group. Although no NOESY cross-peaks
between Gly1 and the blocking group were detected, probably
due to an unfavorable correlation time for this portion of the
molecule, the ROESY cross-peak between the amide proton
(Gly1) and the oxygenated methine proton in dihydropoly-
theonamide B1 was observed, thereby disclosing that the
N-terminal Gly of polytheonamide B was blocked with a 5,5-
dimethyl-2-oxo-hexanoyl group.
Figure 1. 1H (A) and 13C NMR spectra (B) of polytheonamide B in
DMSO-d6.
To identify the uncommon amino acids, the acid hydrolysate
was directly analyzed by 2D NMR; the hydrolysate exhibited
the much simpler NMR spectra than those of the intact peptide,
which was due to the convergence of signals. In addition to the
common amino acids mentioned above, aThr,13 threo-â-
hydroxyaspartic acid (OHAsp),14 â-methylglutamic acid (âMe-
Glu),15 t-Leu,16 â-hydroxyvaline (OHVal),17 and â-methyliso-
leucine (âMeIle)18 could be assigned. Since âMeIle was not
thoroughly assigned due to the heavily crowded aliphatic signals,
this amino acid was isolated and identified from the hydroly-
sate.19 Intense signals assignable to methylamine were also
observed.20
Structures of the N-Terminus and 44th Amino Acid
Residues. Previously, the N-terminal-blocking group and the
44th residue were assigned as carbamoyl and γ-hydroxy-t-
leucine, respectively, on the basis of NMR data.10,11 In addition,
polytheonamide B was assigned as a tripropylammonium salt
in order to account for the MS data and due to the NMR signals
for tripropylamine. Subsequently, tripropylamine was found to
be a contaminant from the n-PrOH used for HPLC. In the
meantime, we realized the presence of a functional group
reducible with NaBH4 which prompted the reinvestigation of
the total structure. Inspection of the 2D NMR spectra indicated
Since the N-terminal structure was revised and tripropylamine
was not contained in the molecule, the 44th residue, which had
previously been assigned as γ-OH-t-Leu, had to be revised.
Interpretation of the HMBC data showed a cross-peak between
a singlet methyl (δ 2.52) and the methylene carbon at δ 63.3,
which was originally assigned to an oxygenated methylene.
However, these chemical shift values were more consistent with
two carbons which were separated by a sulfoxide group, which
was also in accordance with MS data. In fact, the elemental
analysis of polytheonamide B demonstrated the presence of one
sulfur atom. Therefore, the 44th residue was â,â-dimethylme-
thionine sulfoxide [Me2Met(O)]. This amino acid and 5,5-
dimethyl-2-oxo-hexanoic acid were not detected in the acid
hydrolysate of polytheonamide B.
Sequencing of Amino Acids. To sequence the amino acids
in polytheonamide B using NMR spectroscopy solvents were
required which gave sharp and well-dispersed 1H NMR signals,
and high-quality NOESY spectra; CDCl3/CD3OH (1:1) and
DMSO-d6 satisfied these criteria. In CDCl3/CD3OH (1:1), the
NOESY cross-peaks were observed between adjacent residues
and between sequentially distant residues. However, this feature
also complicated the sequence analysis. By contrast, poly-
theonamide B appeared to adopt a random-coil conformation
in DMSO-d6, in which NOESY cross-peaks were observed only
between adjacent residues, thereby facilitating sequential signal
assignments. The number of amino acid residues was established
by interpretation of the HOHAHA spectrum as follows: OHAsx
(2), Asx (8), Thr/aThr (1 each), Ser (1), Glx (1), âMeGlx (1),
Ala (7), Val (3), and Ile (2). The residues that contained a
quaternary carbon in the side chain were analyzed on the basis
of intra-residual NOESY cross-peaks which were more intense
than inter-residual cross-peaks, allowing the determination of
eight t-Leu, one âMeIle, three OHVal, and one Me2Met(O) unit
(Figure 2). Sequence analysis of the peptide chain in DMSO-
d6 was unexceptional once the types and numbers of components
were established (Figure 3, Table 1).
(13) In addition to an A3MX system for Thr (δ 1.21, 3.82, and 4.28), there was
another A3MX system with similar chemical shift values (δ 1.15, 3.94,
and, 4.25), which could be assigned as aThr.
(14) An AB system at δ 4.75 and 4.28 indicated the presence of threo-â-
hydroxyAsp. [(a) Kato, T.; Hinoo, H.; Terui, Y.; Kikuchi, J.; Shoji, J. J.
Antibiot. 1988, 41, 719-725. (b) Tymiak, A. A.; McCormick, T. J.; Unger.
S. E. J. Org. Chem. 1989, 54, 1149-1157. (c) ElDin, A. L. M. S.; Kyslik,
P.; Stephan, D.; Abdallah, M. A. Tetrahedron, 1997, 53, 12539-12552.]
(15) A methine proton at δ 2.50 exhibited COSY cross-peaks with an R-methine
signal at δ 3.98, a methylene at δ 2.39 and 2.53, and a methyl at δ 0.95.
(Debono, M.; Barnhart, M.; Carrell, C. B.; Hoffmann, J. A.; Occolowitz,
J. L.; Abbott, B. J.; Fukuda, D. S. J. Antibiot. 1987, 40, 761-777.)
(16) There was a large methyl singlet at δ 0.96, which exhibited HMBC
correlations with carbons at δ 26.0, 33.0, and 62.5. The R-methine proton
attached to a carbon at δ 62.5 resonated at δ 3.32, as a singlet.
(17) The HMBC spectra revealed that both singlet methyls at δ 1.17 and 1.32
were correlated with carbons at δ 62.9 and 70.6. (Ikai, K.; Takesako, K.;
Shiomi, K.; Moriguchi, M.; Umeda, Y.; Yamamoto, J.; Kato, I. J. Antibiot.
1991, 44, 925-933.)
(18) (a) Gulavita, N. K.; Gunasekera, S. P.; Pomponi, S. A.; Robinson, E. V. J.
Org. Chem. 1992, 57, 1767-1772. (b) Li, H.; Matsunaga, S.; Fusetani, N.
J. Med. Chem. 1995, 38, 338-343. (c) Rashid, M. A.; Gustafson, K. R.;
Cartner, A. K.; Shigematsu, N.; Pannell, L. K.; Boyd, M. R. J. Nat. Prod.
2001, 64, 117-121.
(19) 1H NMR data for the isolated âMeIle: δ 0.78 (3H, t), 0.88 (3H, s), 0.89
(3H, s), 1.32 (2H, q), and 3.40 (1H, s).
There were eight doublet N-methyl resonances attributable
to N-methylation in the side-chain amide groups. All of the
(20) 1H NMR data for the isolated MeNH2: δ 2.44 (3H, s).
9
J. AM. CHEM. SOC. VOL. 127, NO. 1, 2005 111