Concise synthesis of AHMHA unit in perthamide C. Structural and stereochemical revision of perthamide C

Article abstract of DOI:10.1016/j.tet.2010.07.060

Diastereomers of 3-amino-2-hydroxy-6-methylheptanoic acid (AHMHA), a new amino acid unit in perthamides C and D, have been synthesized from commercially available 4-methylpentanol in a concise manner and 50% average overall yield. Comparison of the ¹H and ¹³C NMR data, optical rotation data and Marfey's analysis of the resulting isomers with the natural fragment unambiguously allowed the configurational assignment of the natural residue as (2R,3R). A structural revision of perthamides C and D is also reported.

Full text of DOI:10.1016/j.tet.2010.07.060

Tetrahedron 66 (2010) 7520e7526
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Concise synthesis of AHMHA unit in perthamide C. Structural and stereochemical
revision of perthamide C
,
Valentina Sepe a, Maria Valeria D’Auria a, Giuseppe Bifulco b, Raffaella Ummarino a, Angela Zampella a
*
a Dipartimento di Chimica delle Sostanze Naturali, Università di Napoli “Federico II”, via D. Montesano 49, 80131 Napoli, Italy
b Dipartimento di Scienze Farmaceutiche, Università di Salerno, via Ponte don Melillo, Fisciano (SA) 84084, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Diastereomers of 3-amino-2-hydroxy-6-methylheptanoic acid (AHMHA), a new amino acid unit in
perthamides C and D, have been synthesized from commercially available 4-methylpentanol in a concise
manner and 50% average overall yield. Comparison of the 1H and 13C NMR data, optical rotation data and
Marfey’s analysis of the resulting isomers with the natural fragment unambiguously allowed the con-
figurational assignment of the natural residue as (2R,3R). A structural revision of perthamides C and D is
also reported.
Received 25 May 2010
Received in revised form
2 July 2010
Accepted 20 July 2010
Available online 27 July 2010
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
therefore the structure of perthamide C must be revised by
replacing the
b
-hydroxyasparagine residue with the Nd-carbamoyl-
Recently we reported the isolation, from a Solomon collection of
the marine sponge Theonella swinhoei,1 of two new octacyclopep-
tides, which we named perthamides C and D. Perthamide C (1),
when tested in a well characterised model of inflammation in vivo,
i.e., mouse paw oedema,2 significantly reduced carrageenan-in-
duced paw oedema, displaying potent dose-dependent anti-in-
b-sulfated asparagine (Fig. 1).
dAbu
O
H2N
Asn
O
flammatory activity. Structurally, perthamide
a 25-membered macrocycle with several non-proteinogenic amino
acid residues, such as -methylproline, -hydroxyasparagine, o-ty-
C
comprises
O
O
O
N
H
N
H
g
b
NMeGly
NH
NH
OH
N
rosine and 3-amino-2-hydroxy-6-methylheptanoic acid (AHMHA).
Whereas the configuration of non-conventional residues was se-
cured through an integrated approach, which combined NMR
analysis, chemical degradation, stereoselective synthesis and LC/MS
analysis, the remaining stereochemical uncertainty of perthamide C
concerns the configuration of the two stereogenic centres in the
AHMHA unit. In the original paper we tentatively proposed a threo
O
O
HN
O
H
N
OH
oTyr
O
N
NH2
OH
O
MeO
ThrOMe
MePro
OHAsn
relative disposition of the hydroxyl and the amino group on the
a-
and -position, respectively, on the basis of the J coupling analysis.
b
1
However, due to the lack of adequate standards the absolute con-
figuration remained unassigned.
Figure 1. Perthamide C (1) with structural and stereochemical ambiguities.
In this paper we report the results of a stereochemical in-
vestigation on the AHMHA unit in perthamides, that allowed the
revision of the relative configuration originally proposed and
the unambiguous definition of the absolute configuration. Further,
the comparison of the data for mutremdamide A3 indicated that
mutremdamide A and perthamide C are the same compound, and
2. Results and discussion
In our previous paper1 the application of the conventional
J-based NMR method4 evidenced some intrinsic difficulties, arising
from the qualitative comparison of the predicted and some ap-
proximation in the measurement of the experimental coupling
constants, which had been determined on the basis of the sole
analysis of phase sensitive-HMBC spectra. For this reason we
* Corresponding author. E-mail address: azampell@unina.it (A. Zampella).
0040-4020/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2010.07.060

V. Sepe et al. / Tetrahedron 66 (2010) 7520e7526
7521
decided to acquire further NMR data, using the 2D-HETLOC5 ex-
periment for the accurate determination of the heteronuclear J
values.
In addition, for the comparison of the experimental couplings,
the more reliabledfrom a quantitative point of viewdQM-NMR
integrated approach6e8 was applied.
OH
COOH
NH2
OH
COOH
NH2
(R,S)-2b
(R,R)-2a
In particular, we considered for our calculations the threo and
the erythro arrangements of a simplified fragment representative of
the AHMHA unit (Fig. 2), containing, together with the b-amino
O
COOH
acid unit, a N-methyl group and an acetyl group in the carboxy and
3
amino terminal positions, respectively.
OH
Sharpless AE
3
2
CONHCH3
OH
NHAc
4
Figure 2. Molecular fragment representing AHMHA; calculations for erythro and threo
arrangements were performed varying C2 and C3 centres.
HWE
H
O
Following the protocol previously reported in literature6e8 we
optimised at quantum chemical level (see Experimental) the three
staggered arrangements (gauche þ, gauche ꢀ and anti) for both the
threo and erythro configurations. On the optimised geometries, we
calculated J values at the MPW1PW91 level, using the 6-31G(d,p)
basis set and considering the MeOH solvent using the IEF-PCM
model.9 The results of the calculation were compared to the ex-
perimental data. Surprisingly, differently from what was found
using the J-based method,1 the quantitative QM-J method evi-
denced a better superimposition between the experimental and
calculated data for the erythro arrangement. In fact, as displayed in
Table 1, the lowest total absolute deviation (TAD) is observed for the
erythro gauche þ arrangement (4.8 Hz), even though a comparable
value of 6.4 Hz is observed for the gauche-conformation of the threo
arrangement.
5
Figure 3. Retrosynthetic analysis of (2R,3R)- and (2R,3S)-3-amino-2-hydroxy-6-
methylheptanoic acids (2aeb).
to oxidation under Swern condition11 and the unpurified aldehyde
was subjected to standard HWE two-carbon homologation giving
(E)-unsaturated ester 7 in 72% yield (>98% de, as judged by NMR
data) over two steps.
a,b
c
OH
COOEt
6
7
O
d
e
OH
OH
Table 1
Calculated and experimental J values (Hz) for a fragment corresponding to AHMHA
in perthamide C (1)
4
8
Calcd
threo
gþ
Exp.
OH
erythro
gþ
f
O
anti
8.7
1.2
1.9
ꢀ1.5
ꢀ3.0
12.7
gꢀ
anti
7.8
4.3
0.6
ꢀ3.0
ꢀ3.7
12.6
gꢀ
3.4
1.0
6.7
COOH
NH2 (R,R)-2a
COOH
3JH2eH3
3JH2eC4
3JH3eC1
2JH2eC3
2JH3eC2
TADa
5.3
4.2
6.4
ꢀ1.2
ꢀ2.0
14.9
1.5
1.7
1.0
1.8
1.3
6.4
P
3.0
4.1
0.4
ꢀ3.1
ꢀ0.6
4.8
1.8
1.9
0.7
3
Scheme 1. Reagents and conditions: (a) COCl2, DMSO, TEA, CH2Cl2 dry, ꢀ78 ꢁC; (b)
TEPA, LiOH, THF dry, 72% two steps; (c) DIBAL-H, toluene dry, ꢀ78 ꢁC; (d) Ti(Oi-Pr)4,
0.5
ꢀ2.5
(þ)-DET, TBHP,4 A MS, CH2Cl2 dry, ꢀ20 ꢁC, 76% two steps; (e) NaIO4, RuCl3, CH3CN/
ꢀ
ꢀ2.3
0.0
CCl4/H2O 2:2:3; (f) NaN3, Cu(NO3)2, H2O, 65 ꢁC, then NaBH4, 0 ꢁC, 75% two steps.
13.7
a
Total absolute deviation values (
׀
JcalcdꢀJexp.׀).
Chemoselective reduction with DIBAL-H to allylic alcohol 4,
followed by AE, leads to epoxy alcohol 8 in 76% yield over two steps,
whose optical purity was judged to be >98% by the application of
the modified Mosher’s method.
Treatment of (2S,3S)-8 with ruthenium chloride and sodium
periodate15 produced carboxylic acid (2R,3S)-3 in nearly quantita-
tive yield. With optically active key intermediate 3 on hand, we
completed the synthesis of (2R,3R)-2a and (2R,3S)-2b.
In order of confirming the above results concerning the relative
arrangement discording from our previous indications,1 and for the
complete and unambiguous definition of the absolute configura-
tion of AHMHA residue, we undertook the stereoselective synthesis
of all diastereomeric possibilities for the AHMHA residue.
The synthesis of
a-hydroxy-
b
-amino acids has been extensively
-epoxy esters, which
investigated10 and, for our purposes, trans
a-b
are easily obtained in optically enriched form (ee >90%) by
Sharpless asymmetric epoxidation (AE), may represent valuable
intermediates to access to both erythro and threo adducts.
As shown in our retrosynthetical analysis (Fig. 3), the key in-
termediate epoxy acid 3 would arise from allylic alcohol 4, which in
turn could be prepared through a HWE olefination on aldehyde 5
and subsequent chemoselective reduction.
We envisaged that the erythro diastereoisomer could be
obtained through a procedure recently developed by Fringuelli
et al., relying on a one-pot metal catalyzed azidolysis/reduction of
a-b
-epoxycarboxylic acids.12 Treatment of 3 with sodium azide in
the presence of 10 mol % Cu(NO3)2, followed by in situ NaBH4 re-
duction furnished erythro -hydroxy- -amino-6-methylheptanoic
a
b
acid 2a in a 75% yield (Scheme 1). The reaction proceeded smoothly
with excellent regio- and stereoselectivity as judged by NMR data of
2a when compared with those reported for allo-ethylnorstatine.12
The synthetic sequence is outlined in Scheme 1, starting with
commercially available alcohol 6. Primary alcohol 6 was submitted

7522
V. Sepe et al. / Tetrahedron 66 (2010) 7520e7526
Using Bonini’s methodology,13 threo diastereomer was synthe-
sized with similar stereoselectivity and chemical yields starting
from -epoxycarboxylic acid 3 (Scheme 2). Diazomethane ester-
ification produced trans epoxy methyl ester 9 that in turn was
subjected to MgBr2-mediated epoxy-opening to give bromohydrin
10. Azide substitution followed by hydrogenation and acid hydro-
Also in this case all steps in the reaction sequence proceeded in
good yields and in highly regio- and stereoselective manner.
Comparison of NMR spectra of the AHMHA obtained from the
acid hydrolysis of perthamide C with those of the two synthetic
analogues unambiguously indicates the configurational assignment
of the natural fragment as either R,R or S,S (Table 2). Aside from
slight differences in the chemical shift values of the methyl signals
of (R,S)-2b and the natural fragment, the 1H NMR spectrum of (R,S)-
2b displays two well distinguished proton signals (1.58 and
1.80 ppm) for the diastereotopic methylene at C-4 where the natural
fragment displays a single broad 2H multiplet (1.62e1.69 ppm).
Also 13C NMR data confirmed the stereochemical assignment
made. Once again, while the spectrum of diastereoisomer 2a
completely matches that of the natural material, 13C NMR data for
threo stereoisomer differ especially in the chemical shift of carbon
at position-4.
a-b
lysis produced the targeted threo
a
-hydroxy-b-amino acid 2b.
O
O
b
a
COOH
COOMe
3
9
OH
OH
COOMe
c
COOMe
With two diastereoisomers of AHMHA in our hands, it was then
possible to proceed to the determination of the absolute configu-
ration of this unit in perthamide C using a pre-column de-
rivatization method.
10 Br
11
N3
d,e
A small sample (1 mg each) of synthetic diastereomers 2ae2b
was derivatized with both enantiomers of Marfey’s reagent (N-(3-
OH
COOH
fluoro-4,6-dinitrophenyl)-alaninamide;
L-FDAA and D-FDAA).
The - and -FDAA derivatives 12ae12d were analysed using
L
D
NH2
ESI LC/MS in the positive ion mode. By monitoring for FDAA/
AHMHA at m/z 428, they were detected as separate peaks at 48.50,
(R,S)-2b
39.40, 40.99, 48.03 min, respectively (Scheme 3). The
rivative of AHMHA unit in perthamide C (1) was co-eluted with
the -FDAA derivative of (2R,3R)-2-hydroxy-3-amino-6-methyl-
heptanoic acid 2a.
L-FDAA de-
Scheme 2. Reagents and conditions: (a) CH2N2, Et2O, 82% two steps from 8; (b)
MgBr2$OEt2, Et2O, 98%; (c) NaN3, DMF, 65 ꢁC, 92%; (d) H2, Pd/C; (e) 6 N HCl, 120 ꢁC, 75%
two steps.
L
Table 2
Comparison of the NMR data of synthetic 2-hydroxy-3-amino-6-methylheptanoic acid diastereomers 2a,2b with the natural fragmenta
2R,3R-2a
d
, multiplicity (J), H
2R,3S-2b
d
, multiplicity (J), H
Natural fragment
d
, multiplicity (J), H
dH
dC
dH
dC
dH
dC
1
2
3
4
5
6
7
172.7
72.5
55.7
26.4
35.6
28.8
22.5
22.2
172.1
71.3
55.4
28.4
35.4
29.1
22.5
22.5
172.5
4.01 m, 1H
3.35 m, 1H
1.63, 1.70 m, 2H
1.30 m, 1H
1.57 m, 1H
0.93 d (6.7), 3H
0.92 d (6.7), 3H
4.00 m, 1H
3.31 m, 1H
1.58, 1.80 m, 2H
1.34 m, 1H
1.60 m, 1H
0.95 d (6.7), 3H
0.95 d (6.7), 3H
4.04 m, 1H
3.34 m, 1H
1.62, 1.69 m, 2H
1.29 m, 1H
1.56 m, 1H
72.5
55.7
26.8
35.6
28.9
22.6
22.1
0.93 d (6.5), 3H
0.92 d (6.5), 3H
Me-6
a
All chemical shifts are reported in parts per million and were measured in CD3OD. Coupling constants (J) are expressed in hertz.
OH
OH
L-FDAA
COOH
COOH
OH
NH(L)-FDAA
NH(L)-FDAA
COOH
(2R, 3R)-12a
(2R, 3S)-12b
rt = 48.50
rt = 39.40
NH2
OH
OH
(2R, 3R)-2a
(2R, 3S)-2b
D-FDAA
COOH
COOH
NH(D)-FDAA
NH(D)-FDAA
(2R, 3R)-12c
(2R, 3S)-12d
rt = 40.99
rt = 48.03
OH
COOH
(i) 6N HCl 16 h
(ii) L-FDAA
(iii) LC-MS
Perthamide C ( 1)
NH(L)-FDAA
AHMHA
rt = 48.50
Scheme 3. HPLC retention times (rt) of FDAA derivatives of synthetic (2R, 3R)- and (2R, 3S)-3-amino-2-hydroxy-5-methylheptanoic acid (2a and 2b) and configurational assignment
of AHMHA in perthamide C (1).

V. Sepe et al. / Tetrahedron 66 (2010) 7520e7526
7523
Thus, the (2R,3R) configuration for the AHMHA residue in 1 was
unambiguously established (Scheme 3), as further confirmed by the
mixing times (100 and 200 ms). The analysis of the NOESY spectra
evidenced clear negative NOE effects due to the long correlation
time of the molecule, as expected for its relatively high molecular
weight, in particular between NH and H3-4 of 2-aminobutenoic
residue pointing towards a Z-geometry. The results above de-
scribed suggest that attention must be paid to the evaluation of ROE
effects24 and chemical shift trends20,23 in the assignment of con-
figuration of dAbu residue.
We also tried to further validate the configuration assignment of
the dAbu residue through alternative approaches, via the synthesis
of the two possible stereoisomers and comparison of the chemical
physical data of the synthetic 2-aminobutenoic acids with those of
the residue isolated from the acid hydrolyzate of perthamide C.
According to literature,25 the synthesis of protected (E)- and (Z)-2-
aminobutenoic acids is straightforward, but we failed to obtain an
appreciable recovery of the unprotected residue, due to the hy-
drolytic instability of dAbu residue,26 a problem that we also ex-
perienced in the isolation of the residue from the acid hydrolisate of
perthamide C (1).
optical rotation data [synthetic 2a: [
a
]
20 þ8.6 (c 0.5, MeOH); natural
D
20
AHMHA: [
a]
þ8.8 (c 0.55, MeOH)].
D
During the preparation of the present manuscript, Bewley et al.
reported the isolation of mutremdamide A from different deep
water collections of T. swinhoei and Theonella cupola.3 The close
analogy between mutremdamide A and perthamide C appeared at
once evident, prompting us to an in depth investigation of the
reported differences that concern both constitution and configu-
ration. The most significant discrepancy is relative to molecular
formula: our HRESIMS measurement (positive ion mode) of the
observed highest mass peak was consistent with a molecular for-
mula C43H64N10O14 that matched with the 13C NMR data (43 carbon
resonances evidenced by the 13C NMR spectrum and HMBC data).14
When, in the light of Bewley’s considerations, we run the mass
spectrum in negative ion mode we observed a molecular ion peak
at m/z 1066.4106 [MꢀH]ꢀ corresponding with the molecular for-
mula C44H65N11O18S reported for mutremdamide A. This datum, in
addition with the total correspondence of NMR data of two com-
pounds in DMSO and CD3OH, secured the identity of perthamide C
with mutremdamide A. Therefore the structure originally proposed
3. Conclusion
by us must be revised (Fig. 4) by replacing the
b
-hydroxyasparagine
An efficient synthesis of two possible diastereoisomers of
3-amino-2-hydroxy-6-methylheptanoic acid (AHMHA), a new
residue with the new Nd-carbamoyl-
b
-sulfated asparagine. It is
worthy to note how in this case the use of CD3OH as solvent was
misleading, as it not only gave a poor dispersion of signals in the 13C
NMR spectrum,14 but also caused such a lowering of the intensity of
amino acid residue of the marine cyclopetides perthamides, was
accomplished using Sharpless asymmetric epoxidation and regio-
selective epoxide opening as key steps. The synthesis allowed for an
unambiguous definition of the absolute configuration of the resi-
due in the natural product. Comparison of the spectral data of
perthamide C and mutremdamide, recently appeared in the liter-
ature,3 indicated that the two molecules share the same structure.
Accordingly, the structure originally proposed by us has been here
revised (Fig. 4).
NH-4 and carbamoyl protons relative to the Nd-carbamoyl-
b-sul-
fated asparagine residue that the carbamoyl residue could not be
actually detected.
O
H2N
O
O
O
4. Experimental
O
N
H
N
H
4.1. General experimental procedures
NH
NH
OH
N
Specific rotations were measured on a Jasco P-2000 polarimeter.
High-resolution ESI-MS spectra were performed with a Micromass
Q-TOF mass spectrometer. ESI-MS experiments were performed on
a Applied Biosystem API 2000 triple-quadrupole mass spectrome-
ter. HPLC was performed using a Waters Model 6000-A pump
equipped with U6 K injector and a differential refractometer, model
401. NMR spectra were obtained on a Varian Mercury-400 and
O
HN
O
O
H
N
OH
O
N
NHCONH2
OSO3H
O
MeO
Inova-500 NMR spectrometers,
d in parts per million, J in hertz,
Figure 4. Revised structure for perthamide C.
spectra referred to CD2HOD or CHCl3 as internal standards (dH¼3.31
There are two additional differences between mutremdamide A
and 7.26, respectively).
and perthamide C, that concern with the configuration of o-Tyr and
dAbu residues. We proposed the L-configuration for the o-Tyr res-
For an accurate measurement of the coupling constants, NMR
spectra were recorded on Varian 700 MHz and Bruker DRX-600
spectrometers, equipped with cryo-probe. All spectra were ac-
quired in the phase-sensitive mode and the TPPI method was used
idue on the basis of a non-empirical method, extensively used by
us15,16 and others,17 involving the chemical transformation of the
aromatic ring into a carboxy group by oxidative ozonolysis and
subsequent Marfey’s analysis on the obtained aspartate derivative.
As concerning the 2-aminobutenoic residue (dAbu), in the
original paper we assigned the E-configuration to the above residue
for quadrature detection in the
u1 dimension. NMR sample was
obtained dissolving 20 mg of perthamide C in DMSO-d6.
3JHeH values were extracted from 1D 1H NMR and 2D E.COSY
spectra. For the E.COSY spectrum 32 scans for t1 value were acquired
2,3
on the basis of a dipolar coupling between the eNH signal at
d
9.08
with a t1max of 65 ms.
J
values were obtained from phase-
CeH
and the olefinic proton H-3 at
d
5.96,1 observed in the ROESY
sensitive PFG-PS-HMBC spectra and 13C coupled and decoupled
HSQC-TOCSYaccording the following conditions. The PFG-PS-HMBC
spectrum (400 ms mixing time) acquired in CD3OH and on the basis
of chemical shift consideration.18e23 When, in the light of the re-
sults by Bewley et al.,3 we run the ROESY spectrum in DMSO using
400, 200 and 50 ms mixing times, the amidic proton of the dAbu
residue invariably gave a week dipolar coupling with both H-3 and
CH3-4 protons. These controversial ROE effects probably are due to
the interference in the spectra of NOE effects of opposite sign. For
this reason, we decided to acquire NOESY spectra at different
spectrum was recorded using 2 K points in
long-range coupling evolution ( ) at 50 ms, with 32 scans/t1 (t1max
15.2 ms). Zero-filling (8ꢂ1 K) was carried out in 2 and 1, re-
spectively, to obtain a digital resolution of 0.9 Hz in
2. The 13C
coupled and decoupled HSQC-TOCSY were recorded using 2 K
points in
2, setting the mixing time 80 ms and 1JCH of 140 Hz, with
120 scans/t1 (t1max 9.7 ms). Zero-filling (8Kꢂ1K) was carried out in
u2, setting the delay for
D
u
u
u
u

7524
V. Sepe et al. / Tetrahedron 66 (2010) 7520e7526
u
2 and
u
1, respectively, to obtain a digital resolution of 0.5 Hz in
u
2;
(hexane/EtOAc, 99:1) afforded pure 7 (1.21 g, 75%). dH (400 MHz,
CDCl3) 6.95 (1H, dt, J 15.4, 6.6, CH]CHCO), 5.80 (1H, d, J 15.4 Hz,
CH]CHCO), 4.17 (2H, q, J 7.4, OCH2), 2.20 (2H, m, CH2CH¼),1.56 (1H,
m, CHMe2),1.32 (2H, m, CH2),1.24 (3H, t, J 7.4, OCH2Me), 0.88 (3H, d, J
6.6 Hz, Me), 0.89 (3H, d, J 6.6 Hz, Me); dC (100 MHz CDCl3) 165.4,
150.8, 121.0, 60.4, 37.2, 30.4, 27.8, 22.5 (2C), 14.5; HRMS (ESI): calcd
for C10H19O2: 171.1382; found 171.1389 [MþH]þ.
the protonecarbon 2,3J coupling were extracted through a com-
puter-aided analysis of the heteronuclear coupled and decoupled
multiplets acquired in two separate experiments.
The reverse single-quantum heteronuclear correlation (HSQC)
spectra were recorded by using a pulse sequence with a Wurst
pulse 0.15 s before each scan to suppress the signal originating from
protons not directly bound to 15N; the interpulse delays were ad-
1
justed for an average JNH of 90 Hz.
4.2.2. (E)-6-Methylhept-2-en-1-ol (4). To a solution of compound 7
(1.10 g, 6.47 mmol) in toluene (35 mL) at ꢀ78 ꢁC, under a nitrogen
atmosphere and with stirring, was slowly added DIBAL-H (7.59 mL
of a 1.7 M solution in toluene, 12.9 mmol). After 1 h (TLC monitor-
ing), the reaction was quenched with saturated NH4C1 (25 mL) and
the mixture was extracted with Et2O (50 mL). Organic layer was
dried over Na2SO4, concentrated in vacuo and purified on SiO2
chromatography (hexane/EtOAc, 90:10) affording pure compound
4 (787 mg, 95%). dH (400 MHz, CDCl3) 5.67 (2H, m, CH]CH), 4.09
(2H, d, J 5.2 Hz, CH2OH), 2.05 (2H, m, CH2CH¼),1.63 (1H, m, CHMe2),
1.27 (2H, m, CH2), 0.82 (3H, d, J 6.6 Hz, Me), 0.81 (3H, d, J 6.6 Hz,
Me); dC (100 MHz CDCl3) 133.8., 128.6, 63.8, 38.3, 30.1, 27.5, 22.4
(2C); HRMS (ESI): calcd for C8H17O: 129.1279; found 129.1274
[MþH]þ.
For the phase-sensitive PFG-HETLOC spectrum, a total of
160 scans/t1 were acquired using 4 K points in u2, with a spin lock
of 50 ms and a t1max of 41.7 ms. The data matrices were zero-filled
to 4Kꢂ1k affording a digital resolution of 0.4 Hz in u2.
Through-space 1H connectivities were evidenced using a ROESY
experiment with mixing times of 400, 100 and 50 ms, respectively,
and NOESYexperiment acquired with 200 and 100 ms mixing times.
All reagents were commercially obtained (Aldrich, Fluka) at
highest quality and used without further purification except where
noted. Dichloromethane, ether, tetrahydrofuran and triethylamine
were distilled from calcium hydride immediately prior to use. All
reactions were monitored by TLC on silica gel plates (Macherey,
Nagel). Crude products were purified by column chromatography
on silica gel 70e230 mesh. All reactions were carried out under
argon atmosphere using flame-dried glassware.
4.2.3. (2S,3S)-2,3-Epoxy-6-methylheptan-1-ol (8). To a 100 mL round
bottom flask under argon equipped with a magnetic stirrer were
added molecular sieves (4 A, 164 mg) in CH2Cl2 (50 mL). At ꢀ23 ꢁC
ꢀ
4.1.1. Computational details. Molecular mechanics (MM) calcula-
tions were performed using the MacroModel 8.5 software package
and the MMFFs force fields. MonteCarlo Multiple Minimum
(MCMM) method (10,000 steps) of the MacroModel package was
used in order to allow a full exploration of the conformational space.
All the structures, so obtained, were optimised using the
PolakeRibier Coniugated Gradient algorithm (PRCG, 1000 steps,
maximum derivative less than 0.05 kcal/mol). The initial geometries
of the minimum energy conformers were optimised at the hybrid
DFT MPW1PW91 level using the 6-31G(d) basis set (Gaussian 03
software package). GIAO J-coupling calculations were performed
using the MPW1PW91 functional and the 6-31G(d) basis set, using
as the input the geometry previously optimised at MPW1PW91/6-
31G(d) level.
were then added Ti(Oi-Pr)4 (77.7 mg, 0.273 mmol) and
L-(þ)-DET
(67.7 mg, 0.328 mmol). The solution was allowed to stir for 5 min,
then compound 4 (0.700 g, 5.47 mmol) and TBHP (1.98 mL of a 5.5 M
solution in decane, 10.9 mmol) were added successively. After 24 h
a solution of tartaric acid (492 mg, 3.28 mmol) and FeSO4 (1.8 g,
6.56 mmol) in 20 mL of water was added and was stirred at ꢀ23 ꢁC.
After 30 min, the cooling bath was removed and stirring was con-
tinued at room temperature for 1 h until the aqueous layer become
clear. After separation of the aqueous layer, the organic layer was
washed oncewithwater, dried over Na2SO4 and concentrated. Thisoil
was diluted with Et2O (50 mL) and cooled in a ice bath, and then
NaOH (25 mL of 1 N solution in brine) was added; the two phase
mixture was stirred at 0 ꢁC for 0.5 h, and then the ether phase was
washed with brine, dried over Na2SO4 and concentrated. SiO2 chro-
4.2. Synthetic procedures for 2a and 2b
matography (hexane/EtOAc, 90:10) afforded pure compound 8
20
(709 mg, 90%). [
a
]
ꢀ10.7 (c 0.3, CHCl3); dH (400 MHz, CDCl3) 3.89
D
4.2.1. (E)-Ethyl
6-methylhept-2-enoate
(7). DMSO (4.18 mL,
(1H, br d, J 12.6 Hz, OeCHCH2OH), 3.60 (1H, br d, J 12.6 Hz, CHeO),
2.92 (2H, m, CH2OH), 1.58 (1H, m, CHMe2), 1.32 (2H, m, CH2CHeO),
1.25 (2H, m, CH2CH2CHeO), 0.88 (3H, d, J 6.4 Hz, Me), 0.87 (3H, d, J
6.4 Hz, Me); dC (100 MHz CDCl3) 61.9, 58.8, 53.7, 35.1, 29.6, 28.0, 22.7,
22.6; HRMS (ESI): calcd for C8H17O2: 145.1229; found 145.1224
[MþH]þ.
58.8 mmol) was added dropwise for 15 min to a solution of oxalyl
chloride (14.7 mL, 29.4 mmol) in dry dichloromethane (50 mL) at
ꢀ78 ꢁC under argon atmosphere. After 30 min a solution of the
alcohol 6 (1.00 g, 9.80 mmol) in dry CH2Cl2 was added via cannula
and the mixture was stirred at ꢀ78 ꢁC for 1 h. Et3N (6.83 mL,
49.0 mmol) was added dropwise and the mixture was allowed to
warm to room temperature. The reaction was quenched by addition
of aqueous NaHSO4 (1 M, 50 mL). The layers were separated and the
aqueous phase was extracted with CH2Cl2 (3ꢂ50 mL). The com-
bined organic layers were washed with saturated aqueous NaHSO4,
saturated aqueous NaHCO3 and brine. The organic phase was then
dried over Na2SO4 and concentrated to give the corresponding al-
dehyde 5 (0.951 g, 97%) as a colourless oil, which was used without
any further purification. dH (400 MHz, CDCl3) 9.70 (1H, s, CHO), 2.44
(2H, m, CH2CHO), 1.59 (1H, m, CHMe2), 1.25 (2H, m, CH2), 0.92 (3H,
d, J 6.6 Hz, Me), 0.91 (3H, d, J 6.6 Hz, Me); dC (100 MHz CDCl3) 203.5,
40.2, 31.1, 27.9, 22.7, 22.5.
Compound 8 (0.5e1.0 mg) was dissolved in freshly distilled
CH2Cl2 and treated with triethylamine (10 mL) and (þ)-
a-methoxy-
a
-(trifluoromethyl)phenylacetyl chloride [(S)-MTPA-Cl] and a cata-
lytic amount of 4-(dimethylamino)pyridine. The mixture was left to
stand at room temperature for 2 h. After this period, the mixture
was concentrated in vacuo affording pure (R)-MTPA ester in
quantitative yield. dH (400 MHz, CDCl3) 7.47 (2H, m, ArH), 7.36 (3H,
m, ArH), 4.47 (1H, dd, J 12.0, 3.0 Hz CH2O), 4.16 (1H, dd, J 12.0, 6.0 Hz
CH2O), 3.50 (3H, s, eOMe), 2.94 (1H, m, CHO), 2.76 (1H, m, CHO)
1.47 (4H, m, CH2CH2CHeO), 1.22 (1H, m, CHMe2), 0.81 (3H, d, J
6.6 Hz, Me), 0.80 (3H, d, J 6.6 Hz, Me); HRMS (ESI): calcd for
C18H24F3O4: 361.1627; found 361.1632 [MþH]þ.
To a solution of compound 5 (0.950 g, 9.50 mmol) and LiOH
(250 mg, 10.5 mmol) in THF (10 mL) was added TEPA (triethyl
phosphonoacetate, 2.07 mL, 10.5 mmol). The reaction mixture was
stirred for 24 h at room temperature and then quenched with water
(10 mL). The mixture was then extracted with EtOAc (3ꢂ30 mL), and
the organic phase was concentrated in vacuo. Flash chromatography
4.2.4. (2R,3S)-2,3-Epoxy-6-methylheptanoic acid (3). To a vigor-
ously stirred mixture of compound 8 (0.700 g, 4.86 mmol) were
added NaIO4 (4.25 g, 19.9 mmol) in CCl4 (18.5 mL), CH3CN
(18.5 mL), H2O (27.8 mL) and RuCl.3H2O (26.2 mg, 0.0972 mmol).
The mixture was stirred at 20 ꢁC for 2 h; then the acidic material

V. Sepe et al. / Tetrahedron 66 (2010) 7520e7526
7525
was carefully extracted at 0 ꢁC into ether, dried briefly over Na2SO4
and evaporated in vacuo to give a crude residue 3 (750 mg), that
was subjected to next steps without further purification. dH
(400 MHz, CDCl3) 3.50 (1H, m, OeCHCOOH), 3.19 (1H, m, CHeO),
1.62 (1H, m, CHMe2), 1.36 (2H, m, CH2CHeO), 1.26 (2H, m,
CH2CH2CHeO), 0.91 (3H, d, J 6.3 Hz, Me), 0.90 (3H, d, J 6.3 Hz, Me);
dC (100 MHz CDCl3) 173.0, 59.4, 52.8, 34.7, 29.6, 27.9, 22.6 (2C);
HRMS (ESI): calcd for C8H13O3: 157.0865; found 157.0861 [MꢀH]ꢀ.
5.5) the reaction mixture was cooled to 0 ꢁC, and NaBH4 was added
portion-wise (96.1 mg, 2.54 mmol). After 30 min at 0 ꢁC the re-
action mixture was filtered and concentrated in vacuo to give
a residue that, subjected to HPLC purification on the reversed-
phase Phenomenex Hydro (4
85% H2O/MeOH (0.1% TFA), furnished compound 2a (167 mg,
m
, 250ꢂ4.6 mm) column eluting with
rt¼21 min) in 75% yield over two steps. [
a
]
D
20 þ8.6 (c 0.5, MeOH); dH
(400 MHz, CD3OD) see Table 2 in the text; 13C NMR (100 MHz,
CD3OD) d: see Table 2 in the text; HRMS (ESI): calcd for C8H18NO3:
4.2.5. (2R,3S)-Methyl 2,3-epoxy-6-methylheptanoate (9). To an ali-
quot of 3 (200 mg) in CH2Cl2 was added an ethereal solution of
diazomethane until the solution become yellow. Evaporation and
176.1286; found 176.1291 [MþH]þ.
4.3. Determination of the absolute configuration
silica gel chromatography (hexane/EtOAc, 99:1) afforded pure 9
20
(196 mg, 82% over two steps from 8). [
a
]
þ27.0 (c 1.1, CHCl3); dH
D
4.3.1. Peptide hydrolysis and AHMHA isolation. A 30 mg sample of
perthamide C (1) was dissolved in 6 N HCl (3 mL) and heated in
vacuo at 130 ꢁC for 12 h. The crude residue was fractionated by
(400 MHz, CDCl3) 3.76 (3H, s, OMe), 3.50 (1H, m, OeCHCOOMe), 3.13
(1H, m, CHeO),1.63 (1H, m, CHMe2),1.36 (2H, m, CH2CHeO),1.26 (2H,
m, CH2CH2CHeO), 0.88 (3H, d, J 6.3 Hz, Me), 0.87 (3H, d, J 6.3 Hz, Me);
dC (100 MHz CDCl3) 170.8, 58.9, 53.2, 52.6, 34.7, 29.5, 27.9, 22.5, 22.6;
HRMS (ESI): calcd for C9H17O3: 173.1178; found 173.1172 [MþH]þ.
HPLC on the reversed-phase Phenomenex Hydro (4
m,
250ꢂ4.6 mm) column eluting with 85% H2O/MeOH (0.1% TFA) (flow
20
rate 0.5 mL/min) to give 3 mg of pure ADMHA (rt¼21 min). [
þ8.8 (c 0.55, MeOH); dH (400 MHz, CD3OD) see Table 2 in the text;
13C NMR (100 MHz, CD3OD)
: see Table 2 in the text.
a]
D
4.2.6. (2S,3R)-3-Bromo-2-hydroxy-6-methylheptanoate (10). To a so-
lution of compound 9 (160 mg, 0.930 mmol) in Et2O (6 mL) was
added MgBr2$Et2O (361 mg, 1.40 mmol). The solution was stirred at
room temperature for 1 h and then washed with brine, and the or-
ganic layers were evaporated in vacuo. The crude mixture was pu-
rified by SiO2 chromatography (hexane/EtOAc, 99:1), affording pure
d
4.3.2. LC-MS analysis of Marfey’s (FDAA) derivatives. A portion of
pure AHMHA or the amino acid standards 2ae2b (500 g) was
dissolved in 80 L of a 2:3 solution of TEA/MeCN and this solution
was then treated with 75 L of 1% 1-fluoro-2,4-dinitrophenyl-5-
alaninamide ( - or -FDAA) in 1:2 MeCN/acetone. The vials were
heated at 70 ꢁC for 1 h, and the contents were neutralised with
0.2 N HCl (50 L) after cooling to room temperature. An aliquot of
m
m
m
10 (230 mg, 98%). [
a]
20 ꢀ6.4 (c 0.3, CHCl3); dH (400 MHz, CDCl3) 4.40
D
L
D
(1H, d, J 3.3 Hz, CHOH), 4.16 (1H, m, CHeBr), 3.81 (3H, s, OMe), 1.90
(1H, m, CHaHbCHBr), 1.79 (1H, m, CHaHbCHBr), 1.55 (1H, m, CHMe2),
1.44 (1H, m, CHaHbCH2CHBr),1.29 (1H, m, CHaHbCH2CHBr), 0.88 (3H,
d, J 6.3 Hz; Me), 0.86 (3H, d, J 6.3 Hz, Me); dC (100 MHz CDCl3) 171.5,
74.7, 53.0, 37.0, 32.0, 27.7, 23.0, 22.4, 21.2; HRMS (ESI): calcd for
C9H1879BrO3: 253.0434; found 253.0439 [MþH]þ.
m
the FDAA derivative was dried under vacuum, diluted with 50%
aqueous acetonitrile containing 5% formic acid, and separated on
a Proteo C18 (25ꢂ1.8 mm i.d.) column by means a linear gradient
from 10% to 50% aqueous acetonitrile containing 5% formic acid and
0.05% trifluoroacetic acid, over 45 min at 0.15 mL/min. The RP-HPLC
system was connected to the electrospray ion source by inserting
a splitter valve and the flow going into the mass spectrometer
4.2.7. (2R,3S)-3-Azido-2-hydroxy-6-methylheptanoate (11). A mix-
ture of 10 (189 mg, 0.750 mmol) and NaN3 (195 mg, 3.00 mmol) in
DMF (5.0 mL) was stirred at 65 ꢁC for 24 h. The mixture was then
diluted with EtOAc (10.0 mL), washed with water (10.0 mL), and
source was set at a value of 100 mL/min. Mass spectra were acquired
in positive ion detection mode (m/z interval of 320e900) and the
data were analysed using the suite of programs Xcalibur; all masses
were reported as average values. Capillary temperature was set at
280 ꢁC, capillary voltage at 37 V, tube lens offset at 50 V and ion
spray voltage at 5 V.
concentrated in vacuo. Silica gel chromatography (hexane/ether,
20
99:1) afforded pure 11 (148 mg, 92%). [
a
]
þ1.5 (c 0.1, CHCl3); dH
D
(400 MHz, CDCl3) 4.25 (1H, d, J 5.6 Hz, CHOH), 3.82 (3H, s, OMe),
3.47 (1H, t, J 7.0, CHeN3), 3.03 (1H, d, J 5.6 Hz, CHOH), 1.83 (1H, m,
CHaHbCHN3), 1.81 (1H, m, CHaHbCHN3), 1.61 (1H, m, CHMe2), 1.34
(2H, m, CH2CH2CHN3), 0.93 (3H, d, J 6.3 Hz, Me), 0.91 (3H, d, J
6.3 Hz, Me); dC (100 MHz CDCl3) 172.6, 68.4, 59.0, 53.7, 30.6, 29.2,
24.0, 23.2 (2C); HRMS (ESI): calcd for C9H18N3O3: 216.1348; found
216.1352 [MþH]þ.
Retention times of
L-FDAA-3-amino-2-hydroxy-6-methyl-
heptanoic acid (min): (2R, 3R)-12a (48.50), (2R, 3S)-12b (39.40),
AHMHA-1a (48.50).
Retention times of
D-FDAA-3-amino-2-hydroxy-6-methylhept-
anoic acid (min): (2R, 3R)-12c (40.99), (2R, 3S)-12d (48.30).
4.2.8. (2R,3S)-3-Amino-2-hydroxy-6-methylheptanoic acid (2b). A
solution of 11 (130 mg, 0.604 mmol) in ethyl acetate was hydro-
genated in the presence of Pd/C catalyst (2 mg) for 24 h at room
temperature. The mixture was filtered through Celite and concen-
trated in vacuo to give a 125 mg residue that was subjected to
under vacuum vapour-phase hydrolysis (HCl 6 N, 110 ꢁC for 18 h).
Acknowledgements
This work was supported by grants from MIUR (PRIN 2007)
‘Sostanze Naturali ed Analoghi Sintetici con Attività Antitumorale’
Rome, Italy. NMR spectra were provided by CSIAS, Centro Inter-
dipartimentale di Analisi Strumentale, Faculty of Pharmacy, Uni-
versity of Naples.
HPLC purification on the reversed-phase Phenomenex Hydro (4 m,
250ꢂ4.6 mm) column eluting with 85% H2O/MeOH (0.1% TFA)
furnished compound 2b (79.3 mg, rt¼19.5 min) in 75% yield over
Supplementary data
two steps. [
a
]
20 þ3.3 (c 0.3, MeOH); dH (400 MHz, CD3OD) see Table
D
2 in the text; 13C NMR (100 MHz, CD3OD)
d
: see Table 2 in the text;
Supplementary data associated with this article can be found in
the online version at doi:10.1016/j.tet.2010.07.060.
HRMS (ESI): calcd for C8H18NO3: 176.1286; found 176.1293 [MþH]þ.
4.2.9. (2R,3R)-3-Amino-2-hydroxy-6-methylheptanoic acid (2a).
Compound 3 (200 mg, 1.27 mmol) was dissolved in water (2 mL).
Powdered NaN3 (124 mg, 1.91 mmol) and 1.3 mL of an aqueous
solution 0.1 M of Cu(NO3)2 were added under stirring (resulting pH
4.3e4.5), and the mixture was warmed to 65 ꢁC. After 1.5 h (ca. pH
References and notes
1. Festa, C.; De Marino, S.; Sepe, V.; Monti, M. C.; Luciano, P.; D’Auria, M. V.;
Debitus, C.; Bucci, M.; Vellecco, V.; Zampella, A. Tetrahedron 2009, 65,
10424e10429.

7526
V. Sepe et al. / Tetrahedron 66 (2010) 7520e7526
2. Posadas, I.; Bucci, M.; Roviezzo, F.; Rossi, A.; Parente, L.; Sautebin, L.; Cirino, G.
Br. J. Pharmacol. 2004, 142, 331e338.
15. Zampella, A.; D’Orsi, R.; Sepe, V.; Casapullo, A.; Monti, M. C.; D’Auria, M. V. Org.
Lett. 2005, 7, 3585e3588.
3. Plaza, A.; Bifulco, G.; Masullo, M.; Lloyd, J. R.; Keffer, J. L.; Colin, P. L.; Hooper, J.
N. A.; Bell, L. J.; Bewley, C. A. J. Org. Chem. 2010, 75, 4344e4355.
4. Matsumori, N.; Kaneno, D.; Murata, M.; Nakamura, H.; Tachibana, K. J. Org.
Chem. 1999, 64, 866e876.
16. Gala, F.; D’Auria, M. V.; De Marino, S.; Sepe, V.; Zollo, F.; Smith, C. D.; Keller, S.
N.; Zampella, A. Tetrahedron 2009, 65, 51e56.
17. Faulkner, J. D.; Harper, M. K.; Reddy, R. V. M. Tetrahedron 1998, 54,
10649e10656.
5. Kurz, M.; Schmieder, P.; Kessler, H. Angew. Chem., Int. Ed. Engl. 1991, 10,
1329e1331.
18. Sano, T.; Kaya, K. Tetrahedron Lett. 1995, 36, 8603e8606.
19. Sano, T.; Kaya, K. Tetrahedron 1998, 54, 463e470.
6. Bifulco, G.; Bassarello, C.; Riccio, R.; Gomez-Paloma, L. Org. Lett. 2004, 6,
1025e1028.
7. Bifulco, G.; Dambruoso, P.; Gomez-Paloma, L.; Riccio, R. Chem. Rev. 2007, 107,
3744e3779.
8. Di Micco, S.; Chini, M. G.; Riccio, R.; Bifulco, G. Eur. J. Org. Chem. 2010,
1411e1434.
9. Tomasi, J.; Mennucci, B.; Cances, E. THEOCHEM 1999, 464, 211e226.
10. Bonini, C.; Righi, G. Tetrahedron 2002, 58, 4981e5021.
11. Mancuso, A. J.; Swern, D. Synthesis 1981, 165e185.
12. Fringuelli, F.; Pizzo, F.; Rucci, M.; Vaccaro, L. J. Org. Chem. 2003, 68, 7041e7045.
13. Righi, G.; Rumboldt, G.; Bonini, C. J. Org. Chem. 1996, 61, 3557e3560.
14. In CD3OH the signal relative to carbamate carbon atom fortuitously overlapped
with the C20 signal (dC 155.7) of o-Tyr residue.
20. Sano, T.; Beattie, K. A.; Codd, G. A.; Kaya, K. J. Nat. Prod. 1998, 61, 851e853.
21. Bewley, C. A.; He, H.; Williams, D. H.; Faulkner, D. J. J. Am. Chem. Soc. 1996, 118,
4314e4321.
22. Sepe, V.; D’Orsi, R.; Borbone, N.; D’Auria, M. V.; Bifulco, G.; Monti, M. C.;
Catania, A.; Zampella, A. Tetrahedron 2006, 62, 833e840.
23. Plaza, A.; Bewley, C. A. J. Org. Chem. 2006, 71, 6898e6907.
24. Ford, P. W.; Gustafson, K. R.; McKee, T. C.; Shigematsu, N.; Maurizi, L. K.; Pan-
nell, L. K.; Williams, D. E.; Dilip de Silva, E.; Lassota, P.; Allen, T. M.; Van Soest,
R.; Andersen, R. J.; Boyd, M. R. J. Am. Chem. Soc. 1999, 121, 5899e5909.
25. Nakamura, K.; Tetsuya, I.; Toshima, H.; Kodaka, M. Tetrahedron Lett. 2004, 45,
7221e7224.
26. Goto, Y.; Iwasaki, K.; Torikai, K.; Murakami, H.; Suga, H. Chem. Commun. 2009,
23, 3419e3421.