Probing host-selective phytotoxicity: Synthesis of destruxin B and several natural analogues

Article abstract of DOI:10.1021/jo015953+

The syntheses of the host-selective phytotoxin destruxin B [cyclo(βAla-Hmp-Pro-Ile-MeVal-MeAla), Hmp = (2R)-2-hydroxy-4-methylpentanoic acid], and the closely related natural analogues homodestruxin B (MeVal→MeIle), desmethyldestruxin B (MeVal→Val), hydroxydestruxin B (Hmp→Dhmp, Dhmp = (2R)-2,4-dihydroxy-4-methylpentanoic acid), and hydroxyhomodestruxin B (MeVal→MeIle, Hmp→Dhmp) are described. In each case, the MeAla-βAla linkage was formed by cyclization and the precursor linear hexadepsipeptides were formed by condensing two three-residue fragments. Radiolabeled samples of destruxin B, homodestruxin B, and hydroxydestruxin B were prepared by coupling [3-¹⁴C]-β-alanine to the appropriate pentadepsipeptide followed by cyclization. A noteworthy feature of the synthesis involves the novel use of a Boc-hydrazide protecting group on dipeptides with a C-terminal N-methylalanine residue to inhibit the otherwise facile dioxopiperazine formation during peptide coupling.

Full text of DOI:10.1021/jo015953+

7832
J . Org. Chem. 2001, 66, 7832-7840
P r obin g Host-Selective P h ytotoxicity: Syn th esis of Destr u xin B
a n d Sever a l Na tu r a l An a logu es
Dale E. Ward,* Yuanzhu Gai, Ryszard Lazny, and M. Soledade C. Pedras*
Department of Chemistry, University of Saskatchewan, 110 Science Place, Saskatoon SK S7N 5C9,
Canada
dale.ward@usask.ca
Received J uly 19, 2001
The syntheses of the host-selective phytotoxin destruxin B [cyclo(âAla-Hmp-Pro-Ile-MeVal-MeAla),
Hmp ) (2R)-2-hydroxy-4-methylpentanoic acid], and the closely related natural analogues
homodestruxin B (MeValfMeIle), desmethyldestruxin B (MeValfVal), hydroxydestruxin B
(HmpfDhmp, Dhmp ) (2R)-2,4-dihydroxy-4-methylpentanoic acid), and hydroxyhomodestruxin
B (MeValfMeIle, HmpfDhmp) are described. In each case, the MeAla-âAla linkage was formed
by cyclization and the precursor linear hexadepsipeptides were formed by condensing two three-
residue fragments. Radiolabeled samples of destruxin B, homodestruxin B, and hydroxydestruxin
B were prepared by coupling [3-14C]-â-alanine to the appropriate pentadepsipeptide followed by
cyclization. A noteworthy feature of the synthesis involves the novel use of a Boc-hydrazide
protecting group on dipeptides with a C-terminal N-methylalanine residue to inhibit the otherwise
facile dioxopiperazine formation during peptide coupling.
In tr od u ction
to develop a convergent and efficient synthesis of the host
selective phytotoxin destruxin B (1) that would be
versatile enough to provide access to analogues, radio-
labeled congeners, and possible biotransformation prod-
ucts. Herein we report practical and efficient syntheses
of 1-5, including 14C-labeled samples of 1, 2, and 4.
Although a number of economically important plant
diseases appear to be mediated by host-selective toxins,
their role in disease remains unknown in the majority
of the examples. Host-selective toxins are secondary
metabolites produced by plant pathogens, which facilitate
colonization of host plants by the producing microorgan-
ism, but do not significantly affect nonhost plants. The
large array of chemical structures of host-selective toxins
would suggest multiple and diverse mechanisms of
action; however, one of the major obstacles to investiga-
tion of such mechanisms is the availability of these
metabolites, which are typically produced in very small
amounts. For example, destruxin B (1)1 is a host-selective
toxin produced both in vitro2 and in planta3,4 by the
fungal pathogen Alternaria brassicae (Berk.) Sacc., the
causative agent of Alternaria blackspot of brassicas, but
despite significant work, its production in culture has not
been improved to an acceptable level.5 In addition, related
destruxins 22 and 32,6 produced by A. brassicae have not
been properly evaluated as they are produced in even
smaller amounts. Because Alternaria blackspot is the
most destructive fungal disease of the economically
important oilseeds rapeseed (Brassica napus and B. rapa)
and canola (B. napus and B. rapa),7 it was of much
interest to probe the host-selective phytotoxicity of the
destruxins 1-3. As part of a research program aimed at
understanding disease resistance mechanisms, we wished
Destruxin B (1) is a cyclic hexadepsipeptide composed
of one R-hydroxy acid [(2R)-2-hydroxy-4-methylpentanoic
acid (Hmp)], one â-amino acid [3-aminopropanoic acid
(âAla)], and four R-amino acid residues [L-proline (Pro),
L-isoleucine (Ile), N-methyl-L-valine (MeVal), and N-
methyl-L-alanine (MeAla)]. The structures of the two
other host-selective phytotoxins of interest are closely
related with only the MeVal residue being modified
(homodestruxin B (2), MeValfN-methyl-L-isoleucine
(MeIle); desmethyldestruxin B (3), MeValfvaline (Val)).
The typical synthetic strategy utilized for preparation of
cyclic depsipeptides (and peptides) involves linear or con-
vergent coupling of intact hydroxy acid and amino acid
fragments followed by cyclization.8 This approach focuses
the key strategic decisions to the site of cyclization and
* To whom correspondence should be addressed. E-mail for
M.S.C.P.: pedras@sask.usask.ca.
(1) Kodaira, Y. Agr. Biol. Chem. 1961, 25, 261-262.
(2) Ayer, W. A.; Pen˜a-Rodriguez, L. M. J . Nat. Prod. 1987, 50, 400-
(7) Saharan, G. S. In Breeding Oilseed Brassicas; Labana, K. S.,
Banga, S. S., Banga, S. K., Eds.; Springer-Verlag: Berlin, 1993; pp
181-205. Canola refers to varieties of rapeseed containing very low
amounts of erucic acid and glucosinolates.
407.
(3) Buchwaldt, L.; J ensen, J . S. Phytochemistry 1991, 30, 2311-
2316.
(4) Pedras, M. S. C.; Smith, K. C. Phytochemistry 1997, 46, 833-
(8) (a) Bodanszky, M. Principles of Peptide Synthesis, 2nd ed.;
Springer-Verlag: Berlin, 1993. (b) Izumiya, N.; Kato, T.; Aoyagi, H.;
Waki, M.; Kondo, M. Synthetic Aspects of Biologically Acitve Cyclic
Peptides; Halsted Press: New York, 1979. (c) Kopple, K. D. J . Pharm.
Sci. 1972, 61, 1345-1356.
837.
(5) Pedras and Smith4 reported the best yields (54 mg/L of culture).
(6) Suzuki. A.; Taguchi, H.; Tamura, S. Agr. Biol. Chem. 1970, 34,
813-816.
10.1021/jo015953+ CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/17/2001

Probing Host-Selective Phytotoxicity
J . Org. Chem., Vol. 66, No. 23, 2001 7833
Sch em e 1
the order of the residue coupling. The choice of cyclization
site can be crucial for successful synthesis of a cyclic
(depsi)peptide;8c,9,10 additionally, it has been established
that N-methylamino acid residues are difficult to incor-
porate by standard peptide-coupling methods.10
Syntheses of compounds 2-5 have not been reported
previously. A previous synthesis of 1, reported by Kuya-
ma and Tamura in 1965,11 involved preparation of the
protected hexadepsipeptide AcO-Hmp-Pro-Ile-MeVal-
MeAla-âAla-OMe by linear coupling of the six residues
starting with âAla-OMe using Cbz-protected amino acid
residues with DCC-mediated coupling (16% overall yield).
The yield for lactonization of this hexadepsipeptide (i,
NaOH; ii, DCC; iii, pyridine) was not reported but is
likely to be very low judging from similar examples.12
Thus, Lee and Izumiya failed to obtain protodestruxin
(6) by lactonization of the corresponding linear hexadep-
sipeptide at the âAla-Hmp site using this method; by
contrast, cyclizations at both the Hmp-Pro (26%) and Pro-
Ile (48%) sites were successful using the N-hydroxysuc-
cinimide-activated ester method.12b Following this lead,
Cavelier et al. successfully prepared a variety of de-
struxin analogues (i.e., the Hmp residue in 1 was
replaced by various R-hydroxy and amino acids, Xxx)
based on cyclization of the corresponding hexa(depsi)-
peptides at the Pro-Ile site.13 The linear hexa(depsi)-
peptides were prepared by coupling P-Ile-MeVal-MeAla-
âAla with the requisite Xxx-Pro-P′ fragment (P/P′ ) Cbz/
BnO or Boc/t-BuO), and cyclizations (after deprotection)
were very efficient (63-78%) under optimized conditions
using diphenylphosphoryl azide (DPPA) or benzotriazol-
1-yloxytri(dimethylamino)phosphonium hexafluorophos-
phate (BOP) as the coupling reagent.
resistant to isomerization during chain extension.8a This
synthetic strategy of coupling the three-residue frag-
ments 7 and 8 not only provides for maximum conver-
gence but, by localizing the difficult N-methyl amide
bond(s) to a single tripeptide fragment 7, also should
maximize the overall yield in the longest linear sequence.
Should cyclization at the MeAla-âAla site prove to be
ineffective, an additional advantage of this design is that
simply interchanging the sites designated for fragment
coupling and cyclization would allow for cyclization at
the “proven”12,13 Pro-Ile site. Finally, we considered the
preparation of radiolabeled analogues (vide infra) by
incorporation of a radiolabeled residue as late as possible
in the synthesis. In the present context, this implies
introduction of the labeled N-terminal residue of the
linear hexadepsipeptide in the last coupling step prior
to cyclization. Cyclization at the MeAla-âAla site would
be particularly well suited to this requirement because
the precursor hexadepsipeptide would be prepared by
coupling âAla and a pentadepsipeptide with formation
of the relatively simple âAla-Hmp ester linkage.
A potential disadvantage of the above strategy con-
cerns the synthesis of the fragment 7 by condensation of
an N-protected isoleucine (e.g., 14) with a C-protected
MeVal-MeAla dipeptide (e.g., 17). This is a particularly
difficult coupling step because both amino acid residues
are hindered, the amine is secondary, and dipeptides with
C-terminal N-methylamino acids (e.g., MeVal-MeAla) are
prone to cyclize to dioxopiperazines (Scheme 2).10 For
example, Tamura et al.15 found that attempted prepara-
tion of 16 from its hydrochloride by treatment with Et3N
instead gave 13a . We attempted to prepare 17 by
deprotection of the corresponding trichloroethylcarbam-
ate (Zn, HOAc); however, 1H NMR (CDCl3) of the product
obtained immediately after work up indicated the pres-
ence of only a small amount (20-35%) of 17 with the
major component being 13a . Further monitoring of this
Resu lts a n d Discu ssion
At the outset of our work,14 the published syntheses
of 111 and close analogues12,13 were deemed unsatisfactory
for our objectives. Our retrosynthetic analysis of de-
struxin B (1) is presented in Scheme 1. Various guidelines
have evolved for selection of the cyclization site for cyclo-
(depsi)peptide synthesis.8c,9,10a Lactamization is more
facile than lactonization, and of the five amide bonds in
1, we selected the MeAla-âAla linkage as the most
favorable site for cyclization on the basis of minimal steric
hindrance.8a The Pro-Ile bond in the requisite linear
hexadepsipeptide precursor for cyclization presents a
logical disconnection for fragment coupling because C-
terminal proline residues in oligopeptide chains are
1
(9) (a) Cavelier-Frontin, F.; Achmad, S.; Verducci, J .; J acquier, J .;
Pe`pe, G. J . Mol. Struct. 1993, 286, 125-130. (b) Cavelier-Frontin, F.;
Pe`pe, G.; Verducci, J .; Siri, D.; J acquier, J . J . Am. Chem. Soc. 1992,
114, 8885-8890. (c) Ueda, K.; Waki, M.; Izumiya, N. Int. J . Pept.
Protein Res. 1987, 30, 33-39 and references therein.
(10) Reviews: (a) Humphrey, J . M.; Chamberlin, A. R. Chem. Rev.
1997, 97, 2243-2266. (b) Ryakhovskii, V. V.; Agafonov, S. V.; Kosyrev,
Y. M. Russ. Chem. Rev. 1991, 60, 924-933.
(11) Kuyama, S.; Tamura, S. Agr. Biol. Chem. 1965, 29, 168-169.
(12) (a) Suzuki, A.; Kuyama, S.; Kodaira, Y.; Tamura, S. Agr. Biol.
Chem. 1966, 30, 517-518. (b) Lee, S.; Izumiya, N.; Suzuki, A.; Tamara,
S. Tetrahedron Lett. 1975, 883-886. (c) Lee, S.; Izumiya, N. Int. J .
Pept. Protein Res. 1977, 10, 206-218.
(13) (a) Calmes, M.; Cavelier-Frontin, F.; J acquier, R.; Mercadier,
J .-L.; Sabil, S.; Verducci, J .; Quiot, J .-M.; Vey, A. Int. J . Pept. Protein
Res. 1993, 41, 528-535. (b) Cavelier, F.; J acquier, R.; Mercadier, J .-
L.; Verducci, J . Tetrahedron 1996, 52, 6173-6186. (c) Cavelier, F.;
J acquier, R.; Mercadier, J .-L.; Verducci, J .; Traris, M.; Vey, A. J . Pept.
Res. 1997, 50, 94-101. (d) Cavelier, F.; Verducci, J .; Andre´, F.; Haraux,
F.; Sigalat, C.; Traris, M.; Vey, A. Pestic. Sci. 1998, 52, 81-89.
(14) Preliminary communication: Ward, D. E.; Lazny, R.; Pedras,
M. S. C. Tetrahedron Lett. 1997, 38, 339-342.
mixture by H NMR suggested that the half-life for the
conversion of 17 to 13a in CDCl3 solution was ca. 15-30
min. In an effort to attenuate the propensity of 17 to
cyclize, we considered alternatives to the benzyl ester
protecting group. The increased stability of the tert-butyl
esters in this context has been noted,16 and several
successful acylations of dipeptide tert-butyl esters con-
taining a C-terminal N-methylamino acid have been
reported;16-19 a competing cyclization to dioxopiperazines
was noted in only a few cases.18 By contrast, examples
(15) Chang, C.-F.; Myokei, R.; Sakurai, A.; Takahashi, N.; Tamura,
S. Agr. Biol. Chem. 1969, 33, 1501-1506.
(16) (a) McDermott, J . R.; Benoiton, N. L. Can. J . Chem. 1973, 51,
2562-2570. (b) Wenger, R. M. Helv. Chim. Acta 1983, 6, 2672-2702.
(c) Tung, R. D.; Dhaon, M. K.; Rich, D. H. J . Org. Chem. 1986, 51,
3350-3354. (d) Galpin, I. J .; Mohammed, A. K. A.; Patel, A.; Priestley,
G. Tetrahedron 1988, 44, 1763-1772.

7834 J . Org. Chem., Vol. 66, No. 23, 2001
Ward et al.
Ta ble 1. Effica cy of Va r iou s Cou p lin g Rea gen ts for th e
Syn th esis of 11a a n d 15a
Sch em e 2
time
reactants reagent additive stoichiometrya (h)b (% isomerization)d
% yieldc
9e + 10a f BOP-Cl i-Pr2EtN 1:1.1:1.2:3
PyBrop i-Pr2EtN 1:1.1:1.2:3
18
18
48
72
24
14
14
14
14
14
14
96
36
87 (<0.5)
95 (<0.5)
55 (<0.5)
37 (<0.5)
72 (37)
60 (0.7)
75 (1.5)
77 (nd)
84 (2.3)
60 (0.7)
73 (nd)
20 (<0.5)
21 (23)
DCC
HOBt
1:1.1:1.2:1.1
1:1.1:1.2:1.1
1:1.1:1.2:0.1
EDCg HOBt
DCC
DMAP
11a e + 14h BOP-Cl i-Pr2EtN 1:1.1:1.2:3
1:1.1:1.2:4
1:1.1:1.2:5
1:2:2:8
PyBrop i-Pr2EtN 1:1.1:1.2:3
1:1.5:1.5:4
DCC
DCC
HOBt
DMAP
1:1.1:1.2:1.1
1:1.1:1.2:0.1
a
Stoichiometry of 9 (or 11a ): 10a (or 14): reagent: additive.
Reaction in CH2Cl2 at 0 °C for 30 min and then ambient
b
temperature for the cited time. c Isolated yield. Determined by
HPLC: reversed-phase column (5 mm ODS, 4.6 × 200 mm) eluting
with 75-100% CH3CN/H2O (linear gradient for 35 min) at a flow
rate of 1 mL/min. The minor component from the DCC/DMAP
coupling was assumed to be the isomerized product. e Compound
was subjected to hydrogenolysis (to remove the Cbz group) prior
d
f
g
to coupling. Product is 11a . 1-(3-Dimethylaminopropyl)-3-eth-
h
ylcarbodiimide. Product is 15a .
benzotriazole (HOBt) was ineffectual and “specialized”
coupling reagents such as bis(2-oxo-3-oxazolidinyl)phos-
phinic chloride (BOP-Cl)23 or bromotri(pyrrolidino)phos-
phonium hexafluorophosphate (PyBrop)24 were required
to obtain good yields of 11a with negligible isomerization.
Hydrogenolysis of 11a gave 12a , and in contrast to 17,
cyclization of 12a to 13a was slow in CDCl3 solution (t1/2
≈ 16 h) and even slower in the presence of Et3N (t1/2
>
24 h). The stability of 12a boded well for successful
coupling with 14, and we examined several reagents and
conditions for this reaction (Table 1). The use of PyBrop
or BOP-Cl gave acceptable yields of 15a ; yields were
markedly improved in the presence of excess i-Pr2EtN
(albeit with a concomitant increase in isomerization)
presumably due to the increased stability of 12a under
these conditions.
of N-terminal chain extensions on other dipeptide esters
with C-terminal N-methylamino acids are rare.16a,19
Reasoning that a hydrazide protecting group not only
should slow the rate of dioxopiperazine formation20 but
also could serve as a latent activating group for the
eventual cyclization of the hexadepsipeptide (i.e., via the
azide method),8a we chose to examine the stability of
12a .21
The required tridepsipeptide fragment 2112c was readily
prepared in 90% yield by routine DCC-mediated conden-
sations of 18, 19, and 20 as shown in Scheme 3. Efficient
coupling of fragments 15a and 21 (after hydrogenolysis)
to give the hexadepsipeptide 23a resulted with a variety
of reagents (DCC/HOBt, PyBrop, or DPPA) (Scheme 4).
Cyclization of 23a was achieved in excellent yield in a
three-step one-pot procedure.8a,25 Thus, the Boc protecting
groups in 23a were removed by reaction with TFA and
the resulting amino hydrazide was treated with nitrous
acid to produce the acyl azide that, after neutralization
to free the amine, cyclized to give destruxin B (1; 65%
yield) identical in all respects to an authentic sample
(Scheme 4).
The dipeptide 11a was obtained from 9 and 10a under
a variety of conditions (Scheme 2, Table 1).22 As expected
for this type of coupling, the use of DCC with 1-hydroxy-
(17) (a) Vedejs, E.; Kongkittingam, C. J . Org. Chem. 2000, 65, 2309-
2318. (b) Li, P.; Xu, J . C. Tetrahedron Lett. 1999, 40, 8301-8304. (c)
Lifferth, A.; Bahner, I.; Lackner, H.; Schafer, M. Z. Naturforsch., B:
Chem. Sci. 1999, 54, 681-691. (d) Sivanandaiah, K. M.; Babu, V. V.;
Suresh; Shankaramma, S. C. Int. J . Pept. Protein Res. 1994, 44, 24-
30. (e) Wenger, R. M.; Martin, K.; Timbers, C.; Tromelin, A. Chimia
1992, 46, 314-22. (f) Yoshino, H.; Kaneko, T.; Arakawa, Y.; Nakazawa,
T.; Yamatsu, K.; Tachibana, S. Chem. Pharm. Bull. 1990, 38, 404-
406. (g) Anteunis, M. J . O.; Van der Auwera, C.; Vanfleteren, L.;
Borremans, F. Bull. Soc. Chim. Belg. 1988, 97, 135-1481. (h) Schmidt,
U.; Potzolli, B. Liebigs Ann. Chem. 1987, 935-942. (i) Kunz, H.;
Bechtolsheimer, H. H. Liebigs Ann. Chem. 1982, 2068-2078. (j)
Bechtolsheimer, H. H.; Kunz, H. Angew. Chem. nt. Ed. Engl. 1982,
21, 630. (k) Shin, M.; Inouye, K.; Otsuka, H. Bull. Chem. Soc. J pn.
1978, 51, 1501-1506. (l) Kinoshita, H.; Kotake, H. Bull. Chem. Soc.
J pn. 1977, 50, 280-284. (m) Losse, G.; Raue, H. Tetrahedron 1969,
25, 2677-2685.
(22) The choice of protecting groups for 9, 10, and 14 (i.e., Cbz/Boc)
was dictated by the requirement for orthogonality and the established
superiority of Cbz- vs Boc-protected amino acids in acylations of
N-methylamino acids: (a) Fre´rot, E.; Coste, J .; Poncet, J .; J ouin, P.;
Castro, B. Tetrahedron Lett. 1992, 33, 2815-2816. (b) Coste, J .; Fre´rot,
E.; J ouin, P. J . Org. Chem. 1994, 59, 2435-2446.
(18) Van der Auwera, C.; Anteunis, M. J . O. Int. J . Pept. Protein
Res. 1987, 29, 574-588.
(19) (a) Boger, D. L.; Chen, Y.; Foster, C. A. Bioorg. Med. Chem.
Lett. 2000, 10, 1741-1744. (b) Mendre, C.; Rodriguez, M.; Laur, J .;
Aumelas, A.; Martinez, J . Tetrahedron 1988, 44, 4415-4430.
(20) Hyrazides are very resistant to acid- and base-mediated hy-
drolysis: Smith, P. A. S. Derivatives of Hydrazine and Other Hydroni-
trogens Having N-N Bonds; Benjamin/Cummings: Reading, MA,
1983.
(23) (a) Tung, R. D.; Rich, D. H. J . Am. Chem. Soc. 1985, 107, 4342-
4343. (b) Colucci, W. J .; Tung, R. D.; Perti, J . A.; Rich, D. H. J . Org.
Chem. 1990, 55, 2895-2903.
(24) (a) Coste, J .; Dufour, M.-N.; Pantaloni, A.; Castro, B. Tetrahe-
dron Lett. 1990, 31, 669-672. (b) Coste, J .; Fre´rot, E.; J ouin, P.; Castro,
B. Tetrahedron Lett. 1991, 32, 1967-1970.
(21) For an example of a C-terminal N-methylamino acid dipeptide
hydrazide, see: Ankersen, M.; Nielsen, K. K.; Hansen, T. K.; Raun,
K.; Hansen, B. S. Eur. J . Med. Chem. 2000, 35, 487-497.
(25) Gerlach, H.; Owstschinnikow, J . A.; Prelog, V. Helv. Chim. Acta
1964, 47, 2294-2302.

Probing Host-Selective Phytotoxicity
J . Org. Chem., Vol. 66, No. 23, 2001 7835
Sch em e 3
Sch em e 5
Sch em e 4
hexadepsipeptide 23a in 88% yield (Scheme 5). For
radiolabeling purposes, we wished to keep the scale of
the synthesis as small as possible (to maximize the
specific activity of the product) while still producing
sufficient product to allow isolation and purification by
traditional methods. Cognizant of the form of radiola-
beled â-alanine available to us,29 we developed a proce-
dure without purification of the intermediates whereby
â-alanine (24) was converted into its Boc derivative and
coupled with 25a and the resulting hexadepsipeptide
subjected to the three-step one-pot cyclization protocol.
The procedure using 5-10 mg of â-alanine was optimized
and gave 30-45% overall yields of 1 in several “practice”
runs. In the event, radiolabeled 1 was obtained in 30%
yield; use of the same approach gave radiolabeled 2 in
35% yield (Scheme 5).
Investigation of the metabolism of radiolabeled de-
struxin B (*1) and homodestruxin B (*2) by plant tissues,
as previously described,26,27 established that hydroxyla-
tion of the Hmp residue (to give 4 and 5, respectively)
was the major pathway. Significantly, biotransformations
of 1 into 4 and of 2 into 5 were much faster with leaves
of plants resistant to A. brassicae (S. alba cultivar Ochre)
than with leaves of susceptible plants (B. napus cultivar
Westar and B. juncea cultivar Cutlass).
The syntheses of 4 and 5 were undertaken to cor-
roborate the proposed structures and provide material
to determine their phytotoxicities to plants resistant and
susceptible to A. brassicae. Thus, we required (2R)-2,4-
dihydroxy-4-methylpentanoic acid (Dhmp) to replace the
Hmp residue in 1 and 2. To the best of our knowledge,
the preparation of Dhmp has not been described. Mori
and Takigawa30 obtained an ammonium salt of Dhmp in
very low yield by resolution of the corresponding car-
boxylate (obtained by treatment of racemic 27 with
NaOH). This salt readily lactonized to 27 upon acidifica-
tion, and the lactone was highly prone to racemization
suggesting that protection of the tertiary hydroxy group
in Dhmp would be necessary for our purposes. Our
The syntheses of homodestruxin B (2) and desmeth-
yldestruxin B (3) were accomplished with the above
strategy, but Cbz-MeIle (10b) and Cbz-Val (10c), respec-
tively, were incorporated in place of Cbz-MeVal (10a )
(Scheme 2). The stability of dipeptide 12b was similar
to that of 12a (t1/2 ≈ 14 h for cyclization to 13b in CDCl3
solution), and its condensation with 14 to give 15b
proceeded analogously. However, the propensity for cy-
clization of dipeptide 12c (t1/2 ) 3.5 h for cyclization to
13c in CD3OD solution) was much higher than that for
12a or 12b, and this decreased stability was reflected in
a lower yield of tripeptide 15c under otherwise identical
coupling conditions. Coupling 21 with the tripeptides 15b
and 15c, respectively, followed by cyclization gave ho-
modestruxin B (2) and desmethyldestruxin B (3) with
overall yields similar to that for destruxin B (1) (Scheme
4).
As a prelude to an investigation of the metabolism of
the host-selective toxins by plants resistant and suscep-
tible to Alternaria blackspot, we required radiolabeled
samples of destruxin B (1) and homodestruxin B (2).26,27
To maximize the flexibility of this investigation, we chose
to incorporate 14C into the peptide backbone of the
phytotoxins, and to minimize handling of radioactive
intermediates, we wished to introduce the radiolabel as
late as possible in the synthesis. These objectives were
anticipated in our synthesis design (Scheme 1), which
provides for the introduction of a commercially available
14C-labeled amino acid residue (i.e., âAla) in the penul-
timate step. Thus, hydrogenolysis and coupling of 15a
with 22 gave the pentadepsipeptide 25a in 76% yield28
and condensation of 25a with Boc-âAla (20) gave the
(28) The modest yield is due to competing cyclization of the activated
Hmp-Pro fragment. This reaction was not optimized.
(29) [3-14C]-â-alanine (0.250 mCi, 55.5 × 107 dpm, 46.1 mCi/mmol,
0.48 mg, 5.4 µmol) in 2% aqueous EtOH (2.5 mL, sealed ampule) was
purchased from Sigma Chemical Company. In each case, this material
was diluted with “cold” â-alanine (6 mg). Specific activities: destruxin
B (*1), 1.28 × 107 dpm/mg; homodestruxin B (*2), 7.94 × 106 dpm/mg;
hydroxydestruxin B (*4), 1.07 × 107 dpm/mg.
(26) Preliminary communication: Pedras, M. S. C.; Zaharia, I. L.;
Gai, Y.; Smith, K. C.; Ward, D. E. Org. Lett. 1999, 1, 1655-1658.
(27) Pedras, M. S. C.; Zaharia, I. L.; Gai, Y.; Zhou, Y.; Ward, D. E.
Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 747-752.
(30) Mori, K.; Takigawa, T.; Matsuo, T. Tetrahedron 1979, 35, 933-
940.

7836 J . Org. Chem., Vol. 66, No. 23, 2001
Ward et al.
Sch em e 6
sis of the benzyl carbonate 34. The Boc-âAla ester 35 was
also effective in this regard, and hydrolysis of 35 followed
by condensation of the resulting acid with 18 gave the
desired tridepsipeptide fragment 37 in 62% overall yield.
Use of the fragment 37 in place of 21 allowed the
synthesis of hydroxydestruxin B (4) and hydroxyhomode-
struxin B (5) as described above for 1 and 2 (Scheme 4).
Thus, hexadepsipeptides 41a and 41b were prepared by
deprotection and coupling of 37 with tripeptides 15a and
15b, respectively. Cyclization of 41a and 41b gave 4 and
5, respectively, which were identical (1H and 13C NMR,
HRMS, IR, [R]D, and HPLC) to the isolated metabolites.
To facilitate an investigation of the metabolism of 4
by plants resistant and susceptible to Alternaria black-
spot, we also prepared a 14C-radiolabeled sample using
the (3 + 2) + 1 coupling strategy described above.27 The
required didepsipeptide fragment 40 was obtained by
condensation of 36 with proline allyl ester (38) followed
by Pd-catalyzed deprotection of the allyl ester (Scheme
6). Coupling 15a with 40 gave the pentadepsipeptide 42b
after hydrogenolysis of the benzyl carbonate (Scheme 5).
Without purification of the intermediates, 42b was
subjected to coupling with radiolabeled Boc-âAla and the
resulting hexadepsipeptide cyclized to give radiolabeled
hydroxydestruxin B (*4) (Scheme 5).
In summary, the synthesis of destruxin B (1) proceeds
in six operations from readily available starting materials
(i.e., 9, 10a , 14, and 18-20) using a highly convergent
strategy. The average yield per operation is >88%, and
the overall yields for the longest linear sequences (four
steps) are 55% (from 18) and 49% (from 9). This route
compares very favorably with other reported synthe-
ses11,12,13 of this type of cyclic depsipeptide in terms of
the degree of convergence, the efficiency of the cyclization,
and the overall yield. A noteworthy feature of the
synthesis is the novel use of a Boc-hydrazide protecting
group on an N-methylamino acid dipeptide to inhibit the
otherwise facile dioxopiperazine formation during peptide
coupling. Efficient syntheses of the natural analogues
2-5 were achieved using the same convergent strategy
by substituting the alternate residue(s) as needed into
the appropriate fragment(s). Importantly, the synthetic
approach was easily modified to allow for preparation of
14C-radiolabeled compounds *1, *2, and *4 by incorporat-
ing [3-14C]-â-alanine immediately prior to cyclization.
Ready access to compounds 1-5 has been crucial for
probing their host-selective phytotoxicity and investiga-
tion of the chemical basis of plant disease resistance.
Bioassays of 1-5 utilizing staining of plant cell suspen-
sion cultures of S. alba (resistant), B. juncea (susceptible),
and B. napus (susceptible), in planta leaf assays, and leaf
uptake of toxin solutions indicated that 2 was the most
toxic (EC50 3 × 10-4 M) followed by 1 (EC50 5 × 10-4 M)
and 3 (EC50 . 5 × 10-4 M); metabolites 4 and 5 showed
much lower phytotoxicities (EC50 . × 10-3 M) establish-
ing that hydroxylation is a detoxification step.37 The rate
of hydroxylation of 1 and 2 in cruciferous plants cor-
related with the resistance of each species to A. bras-
sica.26,27 Interestingly, hydroxydestruxin B (4) elicited the
biosynthesis of phytoalexins in the resistant plant species
(S. alba), whereas no phytoalexins were detected in the
susceptible species (B. napus and B. juncea). Although
it is well-known that phytoalexins can be biosynthesized
enantioselective synthesis of the doubly protected Dhmp
derivative 36 is shown in Scheme 6. Saponification of the
readily available lactone 2631 followed by careful acidi-
fication gave the unstable32 hydroxy acid, which was
immediately treated with diazomethane to give the
methyl ester 28 in 70-90% overall yield. The hydroxy
group in 28 was protected as its triethylsilyl (TES) ether,
and hydrolysis of the resulting ester 29 with LiOH gave
the acid 30. Enantioselective introduction of a (2R)-
hydroxyl group on 30 was achieved using the method of
Evans et al.33 Thus, acylation of the commercially avail-
able oxazolidinone 31 with 30 using the mixed anhydride
method34 gave 32 in excellent yield. Reaction of the
sodium enolate of 32 with 2-phenylsulfonyl-3-phenylox-
aziridine35 gave 33 in 84% yield with excellent diaste-
reoselectivity.36 Despite considerable experimentation, all
attempts to remove the oxazolidinone auxiliary in 33
without prior protection of the hydroxy group resulted
in extensive concomitant γ-lactonization to give 27. By
contrast, the desired 36 was readily obtained by hydroly-
(31) Stevens, C. M.; Tarbell, D. S. J . Org. Chem. 1954, 19, 1996-
2003.
(32) Wheeler, O. H.; Granell de Rodriguez, E. E. J . Org. Chem. 1964,
29, 1227-1229.
(33) Evans, D. A.; Morrissey, M. M.; Dorow, R. L. J . Am. Chem. Soc.
1985, 107, 4346-4348.
(34) Ager, D. J .; Allen, D. R.; Schaad, D. R. Synthesis 1996, 1283-
1285.
(35) Vishwakarma, L. C.; Stringer, O. D.; Davies, F. A. Org. Synth.
1987, 66, 203-210.
(36) The presence of other diastereomers could not be detected by
1H NMR of the crude product.
(37) Pedras, M. S. C.; Biesenthal, C. J .; Zaharia, I. L. Plant Sci. 2000,
156, 185-192.

Probing Host-Selective Phytotoxicity
J . Org. Chem., Vol. 66, No. 23, 2001 7837
in response to a variety of abiotic elicitors,38 it is remark-
able that the detoxification product of 1 (i.e., 4) could also
induce phytoalexin biosynthesis in the resistant species.
These results suggest that the blackspot resistance in S.
alba due to detoxification of destruxin B may be amplified
by well-synchronized phytoalexin elicitation.27
J ) 9 Hz, HN), 5.19 (1H, q, J ) 7 Hz), 4.95 (1H, d, J ) 11 Hz),
4.92-4.80 (2H, m), 4.68 (1H, d, J ) 7 Hz), 4.05-3.97 (1H, m),
3.92 (1H, dd, J ) 8, 8.5 Hz), 3.43 (1H, ddd, J ) 8, 9, 9.5 Hz),
3.23 (3H, s), 3.09 (1H, dd, J ) 12, 13.5 Hz), 2.73 (3H, s), 2.70-
2.45 (3H, m), 2.35-2.28 (1H, m), 2.13-1.72 (7H, m), 1.46-
1.20 (2H, m), 1.31 (3H, d, J ) 6.5 Hz), 0.95 (3H, d, J ) 6.5
Hz), 0.90 (3H, d, J ) 7 Hz), 0.89 (3H, d, J ) 7 Hz), 0.85 (3H,
d, J ) 6.5 Hz), 0.82 (3H, d, J ) 7 Hz), 0.81 (3H, t, J ) 7 Hz);
13C NMR (125 MHz, CDCl3) δ 173.7, 173.5, 175.1, 170.9, 169.7
(×2), 71.9, 60.7, 58.1, 55.4, 53.6, 46.4, 38.9, 37.5, 34.4, 33.2,
30.8, 28.8, 28.0, 27.2, 24.4, 24.4, 24.1, 23.3, 21.4, 20.0, 19.6,
15.4, 15.2, 11.3; HRMS m/z calcd for C30H51N5O7 593.3788,
found 593.3788 (EI). Anal. Calcd for C30H51N5O7: C, 60.67; H,
8.66; N, 11.79. Found: C, 60.62; H, 8.59; N, 11.53.
Exp er im en ta l Section
Gen er a l Meth od s. All solvents were distilled prior to use.
Unless otherwise noted, reactions were carried out under an
atmosphere of argon and reaction temperatures refer to that
of the bath. Concentration refers to removal of volatiles at the
water aspirator pressure on a rotary evaporator. Flash column
chromatography (FCC) and dry flash column chromatography
(DFC) were performed according to the procedures of Still et
al.39 and Harwood,40 respectively. All mixed solvent eluents
are reported as v/v solutions. Unless otherwise noted, NMR
spectra were measured in CDCl3 solution at 300 MHz for 1H
and 75 MHz for 13C. Coupling constants (J ) are reported to
the nearest 0.5 Hz. Optical rotations ([R]D) were determined
at ambient temperature using a 1 mL, 10 dm cell; the units
are 10-1 deg cm2 g-1, and the concentrations (c) are reported
in g/100 mL.
Ma ter ia ls. Cbz-Ala, Cbz-Ile, and Cbz-Val, and compounds
10c, 14, 18, 20, and 38 were prepared from the corresponding
amino acids according to standard procedures.41 Compounds
9,42 10a ,42 10b,42 19,43 26,31 and 2-phenylsulfonyl-3-phenylox-
aziridine35 were prepared by published methods. The syntheses
of *4, 38-40, and 42 have been described.27 All other reagents
were commercially available and, unless otherwise noted, were
used as received.
Gen er a l Meth od for Hyd r ogen olysis. A suspension of
the protected substrate and 10% Pd on carbon (100 mg/mmol
of substrate) in CH3OH (18 mL/mmol of substrate) was
vigorously stirred under an H2 atmosphere (balloon). After 2
h (reaction complete by TLC), the mixture was filtered with
the aid of Celite and the combined filtrate and washings were
concentrated. Residual CH3OH was removed by dissolving the
residue in CH2Cl2 followed by concentration and drying under
high vacuum (0.2 Torr). The resulting product (yield is
essentially quantitative) was used without further purification.
Destr u xin B (1). The hexadepsipeptide 23a (250 mg, 0.30
mmol) was dissolved in TFA/CH2Cl2 (1:1 v/v, 4 mL), and after
being stirred for 30 min at room temperature, the solution was
concentrated and reconcentrated from CH2Cl2. Trituration of
the residue with ether gave a white solid that was dissolved
in water (5 mL) and aqueous HCl (1.2 M, 2 mL). The stirred
solution was cooled to <5 °C, and a solution of NaNO2 (30 mg,
0.43 mmol) in water (1 mL) was added dropwise. After 15 min,
ice-cold CH2Cl2 (120 mL) and saturated aqueous NaHCO3 (20
mL) were added and the resulting biphasic mixture was
vigorously stirred at room temperature. After 15 h, the organic
layer was separated and the aqueous layer was extracted with
CH2Cl2 (four times). The combined organic layers were dried
over Na2SO4, concentrated, and fractionated by DFC (15-40%
acetone in hexane) to give destruxin B as a white solid (115
mg, 65%): mp 234-235 °C (acetone/hexane) (lit.2,11,44 234,
225-227 °C); [R]D -240 (c 0.5, CH3OH) (lit.2,11,44 -228, c 0.50,
CH3OH; -237, c 0.08, CH3OH; -241, c 0.5, CH3OH); 1H NMR
(500 MHz, CDCl3) δ 8.20 (1H, d, J ) 8.5 Hz, HN), 7.19 (1H, d,
[3-14C]-âAla Destr u xin B (*1). A solution of [3-14C]-â-
alanine (0.250 mCi, 46.1 mCi/mmol, 0.48 mg, 5.4 µmol) in 2%
aqueous EtOH (2.5 mL) was added to a stirred solution of
â-alanine (24, 6 mg, 0.07 mmol) in t-BuOH (1 mL). The
solution was cooled to 0 °C (ice bath), and NaOH (1 M, 0.074
mL, 0.074 mmol) and a solution of (Boc)2O (15 mg, 0.081 mmol)
in t-BuOH (1.5 mL) were added. After being stirred for 22 h
at room temperature, the reaction mixture was concentrated
to ca. 0.5 mL and diluted with aqueous KHSO4 (0.2 M, 1 mL).
The mixture was saturated with NaCl and extracted with ethyl
acetate (1 mL × 6). The combined organic layers were dried
over Na2SO4, concentrated, and dried under high vacuum for
30 min to give *20 (9.6 mg) as a white solid. DCC (14 mg, 0.068
mmol) and DMAP (0.6 mg, 0.005 mmol) were added to a stirred
solution of *20 (9.6 mg) and 25a (32 mg, 0.049 mmol) in CH2-
Cl2 (0.45 mL) at 0 °C under argon. After 1 h at 0 °C and 21 h
at room temperature, the mixture was filtered through Celite
and the combined filtrate and washings were concentrated and
filtered through a plug of SiO2 eluting first with 50% ethyl
acetate in hexane to remove DCU and then with 80% ethyl
acetate in hexane to obtain *23a (36 mg). TFA (0.35 mL) was
added to a solution of the *23a (36 mg) in CH2Cl2 (0.35 mL),
and after 20 min, the mixture was concentrated to dryness.
Water (1 mL) was added to the residue, and after the solution
was cooled to between 0 and -5 °C (ice-salt bath), HCl (1 M,
0.1 mL, 0.1 mmol) and NaNO2 (6.3 mg, 0.092 mmol) were
added sequentially to the stirred solution. After 20 min, ice-
cold saturated aqueous NaHCO3 (1 mL) and ice-cold CH2Cl2
(7.5 mL) were added and the biphasic mixture was stirred
vigorously at ambient temperature for 20 h. The aqueous layer
was saturated with NaCl and extracted with CH2Cl2 (three
times), and the combined organic layers were dried over Na2-
SO4, concentrated, and fractionated by FCC (60% acetone in
hexane) to give [3-14C]-âAla destruxin B (*1, 13 mg, 0.075 mCi,
specific activity ) 1.28 × 107 dpm/mg; 30% yield based on 24).
Cbz-MeAla -NHNH-Boc (9). A solution of DCC (2.680 g,
13 mmol) in CH2Cl2 (3 mL) was added in three portions to a
stirred mixture of Cbz-N-methylalanine (2.844 g, 12 mmol),
H2NNH-Boc (1.585 g, 12 mmol), and HOBt hydrate (1.85 g,
12 mmol) in CH2Cl2 (5 mL) at 0 °C. After 1 h at 0 °C, the
mixture was diluted with ethyl acetate and the precipitated
DCU was filtered off. The combined filtrate and washings were
washed sequentially with 0.5 M aqueous citric acid and
saturated aqueous NaHCO3, dried over MgSO4, concentrated,
and fractionated by DFC (10-50% ethyl acetate in hexane) to
give 9 as a white hygroscopic solid (3.750 g, 89%): mp 46-48
1
°C; [R]D -51 (c 2.0, CH3OH); H NMR δ 7.99 (1H, br s, HN),
7.48-7.25 (5H, m), 6.45 (1H, br s, HN), 5.25-5.05 (2H, m),
4.92-4.88 (1H, m), 2.92 (3H, s), 1.45 (9H, s), 1.39 (3H, d, J )
7.5 Hz); 13C NMR (125 MHz, CDCl3) δ 171.0, 157.0, 155.2,
136.2, 128.5, 128.1, 127.9, 81.6, 67.8, 52.7, 29.0, 28.1, 13.6;
HRMS m/z calcd for C17H25N3O5 352.1872 (M + H), found
352.1880 (FAB). Anal. Calcd for C17H25N3O5: C, 58.12; H, 7.17;
N, 11.96. Found: C, 58.30; H, 7.29; N, 11.89.
(38) Pedras, M. S. C.; Okanga, F. I.; Zaharia, I. L.; Khan, A. K.
Phytochemistry 2000, 53, 161-176.
(39) Still, W. C.; Kahn, M.; Mitra, A. J . Org. Chem. 1978, 43, 2923-
2925.
(40) Harwood, L. M. Aldrichimica Acta 1985, 18, 25-25.
(41) Bodanszky, M.; Bodanszky, A. The Practice of Peptide Synthesis,
2nd ed.; Springer: New York, 1994.
(42) Cheung, S. T.; Benoiton, N. L. Can. J . Chem. 1977, 55, 906-
910.
(43) Degerbeck, F.; Fransson, B.; Grehn, L.; Ragnarsson, U. J . Chem.
Soc., Perkin Trans. 1 1993, 11-14.
(44) Pa¨ıs, M.; Das, B. C.; Ferron, P. Phytochemistry 1981, 20, 715-
723.
Cbz-MeVa l-MeAla -NHNH-Boc (11a ). Cbz-MeAla-NHNH-
Boc (9, 400 mg, 1.14 mmol) was deprotected by hydrogenolysis
according to general procedure. PyBroP (600 mg, 1.28 mmol)
was added portionwise to a stirred solution of the crude
hydrogenolysis product, 10a (300 mg, 1.12 mmol), and DIEA
(0.58 mL, 0.43 g, 3.3 mmol) in CH2Cl2 (3 mL) at 0 °C. The
mixture was stirred for 10 min at 0 °C and then for 18 h at

7838 J . Org. Chem., Vol. 66, No. 23, 2001
Ward et al.
room temperature. The mixture was diluted with CH2Cl2,
washed sequentially with 0.5 M aqueous citric acid and
saturated aqueous NaHCO3, dried over MgSO4, concentrated,
and fractionated by FCC (45% ethyl acetate in hexane) to give
11a as a white solid (494 mg, 95%). A similar reaction on a
scale 4-times larger than that described gave 11a in 90%
578.3544 (FAB). Anal. Calcd for C30H51N5O7: C, 60.29; H, 8.20;
N, 12.12. Found: C, 60.17; H, 8.24; N, 11.95.
Boc-âAla -Hm p -P r o-OBn (21).12c DCC (566 mg, 2.75 mmol)
was added to a stirred solution of Hmp-Pro-OBn (22) (798 mg,
2.5 mmol), Boc-â-alanine (20) (520 mg, 2.75 mmol), and DMAP
(34 mg, 0.28 mmol) in CH2Cl2 (5 mL) at 0 °C. The mixture
was stirred at 0 °C for 1 h and at room temperature for 20 h
and then diluted with ethyl acetate, and the precipitated DCU
was filtered off. The combined filtrate and washings were
washed sequentially with 0.5 M aqueous citric acid and
saturated aqueous NaHCO3, dried over MgSO4, concentrated,
and fractionated by FCC (35% ethyl acetate in hexane) to give
the titled compound as an oil (1.20 g, 98%): [R]D -33 (c 1.3,
CH3OH); 1H NMR (500 MHz) δ (a 2:1* mixture of conformers)
7.40-7.20 (5H, m), 5.24 (dd, J ) 3.5, 10 Hz), 4.90* (dd, J ) 2,
8.5 Hz), 4.85* (dd, J ) 2.5, 11 Hz) and 4.47 (dd, J ) 3.5, 8.5
Hz) (2H), 5.16* (d, J ) 12 Hz), 5.11 (ap s) and 5.10* (d, J ) 12
Hz) (2H), 3.95-3.25 (4H, m), 2.55-2.45 (2H, m), 2.35-1.25
(8H, m), 1.39* and 1.38 (9H, 2 × s), 0.92, 0.90, 0.84* and 0.74*
(6H, 4 × d, J ) 6.5 Hz); 13C NMR (125 MHz) δ 172.5*, 171.9,
171.7*, 171.3, 169.6*, 169.0, 156.1, 155.9*, 135.7, 135.0*, 128.7,
128.6*, 128.4 (×2), 128.3* (×2), 128.2* (×2), 128.1 (×2), 79.0,
78.9*, 70.9*, 70.6, 67.3*, 66.8, 59.4, 59.2*, 46.8*, 46.5, 39.4,
36.4, 36.3*, 34.7, 34.5*, 31.5*, 28.7, 28.4 (×6), 24.7, 24.5, 24.3*,
23.2*, 23.1, 22.2*, 21.6, 21.1; HRMS m/z calcd for C26H38N2O7
491.2757 (M + H), found 491.2749 (FAB). Anal. Calcd for
1
yield: mp 53-55 °C; [R]D -130 (c 2.0, CH3OH); H NMR δ (a
2.5:1.5:1 mixture of three conformers) 8.92 (d, J ) 2.5 Hz),
7.97 (d, J ) 2.5 Hz) and 7.95 (d, J ) 2.5 Hz) (1H, HN), 7.45-
7.25 (5H, m), 5.28-5.00 (3H, m), 4.74 (d, J ) 10.5 Hz), 4.66
(d, J ) 10.5 Hz) and 4.50 (d, J ) 10.5 Hz) (1H), 3.00 (s), 2.99
(s), 2.84 (s), 2.82 (s) and 2.77 (s) (6H), 2.42-2.24 (1H, m), 1.41
(9H, s), 1.37 (d, J ) 7 Hz), 1.32 (d, J ) 7 Hz) and 1.24 (d, J )
7 Hz) (3H), 1.02-0.85 (6H, m); 13C NMR (125 MHz, CDCl3) δ
172.0, 170.5, 168.8, 158.2, 156.8, 154.9, 136.1, 136.5, 128.6,
128.5, 128.4, 128.2, 128.0, 127.8, 127.6, 81.5, 81.4, 68.0, 67.8,
67.5, 60.8, 60.5, 59.3, 53.5, 50.7, 50.4, 30.8, 30.2, 29.6, 29.2,
28.4, 28.1, 27.8, 27.5, 27.4, 19.6, 19.0, 18.4, 18.3, 15.5, 13.3,
13.2; HRMS m/z calcd for C23H36N4O6 465.2713 (M + H), found
465.2717 (FAB). Anal. Calcd for C23H36N4O6: C, 59.46; H, 7.81;
N, 12.06. Found: C, 59.24; H, 7.60; N, 11.87.
(3S,6S)-1,4,6-Tr im et h yl-3-(1-m et h yet h yl)-2,5-p ip er a -
zin ed ion e (13a ). The crude product obtained from hydro-
genolysis of 11a (50 mg, 0.11 mmol) as described above was
allowed to stand in CDCl3 solution for 3 days. The mixture
was concentrated and fractionated by FCC (60% acetone in
hexane) to give 13a (38 mg, 90%): mp 143-144 °C (lit.15,45
148, 140-142 °C); [R]D +16 (c 1.0, CH3OH) (lit.15 +13.5, c 5,
C
26H38N2O7: C, 63.65; H, 7.80; N, 5.71. Found: C, 63.49; H,
7.94; N, 5.73.
1
Hm p -P r o-OBn (22). DCC (2.16 g, 10.5 mmol) was added
to a stirred solution of Hmp (19) (0.95 g, 7.2 mmol), Pro-OBn
(18) (freshly prepared by washing the HCl salt (1.47 g, 7.2
mmol) with aqueous NaHCO3), and HOBt hydrate (1.43 g, 9.3
mmol) in CH2Cl2 (8 mL) at 0 °C. The mixture was stirred at 0
°C for 1 h and at room temperature for 20 h and then diluted
with ethyl acetate, and the precipitated dicyclohexyl urea
(DCU) was filtered off. The combined filtrate and washings
were washed sequentially with 0.5 M aqueous citric acid and
saturated aqueous NaHCO3, dried over MgSO4, concentrated,
and fractionated by FCC (40% ethyl acetate in hexane) to give
the titled compound as an oil (2.11 g, 92%): [R]D -55 (c 1.5,
CH3OH); 1H NMR δ 7.45-7.25 (5H, m), 5.16 (1H, d, J ) 12.5
Hz), 5.11 (1H, d, J ) 12.5 Hz), 4.53 (1H, m, J ) 1.5, 8.5 Hz),
4.27 (1H, m, J ) 2.5, 10 Hz), 3.72-3.60 (1H, m), 3.45-3.35
(1H, m), 3.20-3.10 (1H, br s), 2.22-1.90 (5H, m), 1.44 (1H,
ddd, J ) 4, 10, 14 Hz), 1.30 (1H, ddd, J ) 2.5, 9.5, 14 Hz),
0.96 (3H, d, J ) 6.5 Hz), 0.92 (3H, d, J ) 6.5 Hz. 13C NMR
(125 MHz) δ 174, 171.6, 135.6, 128.4, 128.3 (×2), 128.1 (×2),
68.0, 66.9, 59.4, 46.3, 43.6, 28.8 (×3), 24.5 (×2), 23.6, 21.3;
HRMS m/z calcd for C30H51N5O7 320.1862 (M + H), found
320.1851 (FAB). Anal. Calcd for C30H51N5O7: C, 67.69; H, 7.89;
N, 4.38. Found: C, 67.45; H, 8.04; N, 4.57.
Boc-âAla -Hm p -P r o-Ile-MeVa l-MeAla -NHNH-Boc (23a ).
Usin g P yBr op . Compound 21 (220 mg, 0.45 mmol) and the
tripeptide 15a (230 mg, 0.40 mmol) were individually depro-
tected according to the general procedure for hydrogenolysis.
PyBroP (280 mg, 0.60 mmol) was added to sitrred solution of
the crude deprotected products (180 mg from 21, 177 mg from
15a ) and DIEA (0.13 mL, 0.74 mmol) in CH2Cl2 (2 mL) at 0
°C. After 10 min at 0 °C and 17 h at room temperature, the
reaction mixture was diluted with CH2Cl2, washed sequentially
with 0.5 M aqueous citric acid and saturated aqueous NaHCO3,
dried over Na2SO4, concentrated, and fractionated by DFC
(30-65% ethyl acetate in hexane) to give 23a as a white solid
(430 mg, 94%). Usin g DCC/HOBt. DCC (69 mg, 0.33 mmol,
1.1 equiv) was added to a stirred solution of HOBt hydrate
(50 mg) and the crude products from hydrogenolysis of 21 (133
mg from 162 mg, 0.33 mmol of 21) and 15a (133 mg from 175
mg, 0.30 mmol of 15a ) in CH2Cl2 (0.5 mL) at 0 °C. After 1 h at
0 °C and 24 h at room temperature, the reaction mixture was
diluted with ethyl acetate (ca. 3 mL) and the precipitated DCU
was filtered off. The combined filtrate and washings were
processed as above to give 23a as a white solid (233 mg, 95%).
Usin g DP P A. A solution of DPPA (93 mg, 0.39 mmol, 1.3
equiv) in DMF (0.5 mL) and then Et3N (0.07 mL) were
sequentially added to a stirred solution of the crude products
CH3OH); H NMR δ 3.93 (1H, q, J ) 8 Hz), 3.75 (1H, d, J )
6.5 Hz), 3.01 (3H, s), 2.95 (3H, s), 2.29-2.12 (1H, m), 1.58 (3H,
d, J ) 6.5 Hz), 1.12 (3H, d, J ) 7.5 Hz), 1.03 (3H, d, J ) 7.5
Hz); 13C NMR δ 167.1, 164.6, 68.2, 57.8, 34.5, 32.2, 31.8, 19.8,
18.8, 18.7; HRMS m/z calcd for C10H18N2O2 198.1368, found
198.1369 (EI).
Cbz-Ile-MeVa l-MeAla -NHNH-Boc (15a ). Usin g BOP -Cl.
A mixture of 11a (330 mg, 0.71 mmol) and 10% Pd/C (200 mg)
in methanol (10 mL) was vigorously stirrred at room temper-
ature under H2 (1 atm) for 0.5 h. The mixture was quickly
filtered with the aid of Celite, and the combined filtrate and
washings were concentrated and then dried under high
vacuum (<0.5 Torr, 5 min). Without delay, the residue (12a )
was dissolved in a cold mixture of CH2CI2 (1.5 mL) and DIEA
(0.75 mL, 0.55 g, 4.3 mmol) and the solution was added to a
stirred solution of 14 (370 mg, 1.4 mmol), DIEA (0.25 mL, 1.4
mmol), and BOP-Cl (356 mg, 1.4 mmol) in CH2Cl2 (1.5 mL) at
0 °C. After 30 min at 0 °C and 14 h at room temperature, the
mixture was diluted with CH2Cl2, washed sequentially with
0.5 M aqueous citric acid and saturated aqueous NaHCO3,
dried over MgSO4, concentrated, and fractionated by FCC (45%
ethyl acetate in hexane) to give the titled compound as a white
solid (345 mg, 84%). Usin g P yBr op . Compound 11a (464 mg,
1.0 mmol) was deprotected by hydrogenolysis as above, and
without delay, an ice-cold solution of 14 (399 mg, 1.5 mmol)
and DIEA (0.070 mL, 4.0 mmol) in CH2Cl2 (3.0 mL) was added
to the above residue. PyBroP (699 mg, 1.5 mmol) was added,
and the mixture was stirred at 0 °C for 1 h and then at room
temperature overnight. The mixture was diluted with CH2Cl2
and processed as above to give the titled compound as a white
solid (421 mg, 73%): mp 79-81 °C; [R]D -15 (c 2.0, CH3OH);
1H NMR δ (a 1.5:1 mixture of two conformers) 9.26 (br s), 7.90
(br s) and 6.36 (br s) (2H, br s, HN), 5.46 (d, J ) 9 Hz) and
5.41 (d, J ) 9.5 Hz) (1H, HN), 5.25-5.05 (4H, m), 4.55-4.45
(1H, m), 3.38 (s), 3.05 (s), 3.02 (s) and 2.82 (s) (6H, H3CN ×
2), 2.44-2.24 (2H, m), 1.75-1.65 (1H, m), 1.42 (s) and 1.41 (s)
(9H, (H3C)3C), 1.36 (d, J ) 7 Hz) and 1.28 (d, J ) 7 Hz) (3H,
H3CC-3 Ala), 1.15-1.03 (1H, m), 0.95-0.75 (12H, m, H3C ×
4); 13C NMR (125 MHz, CDCl3) δ 173.1, 171.7, 170.5, 156.4,
154.9, 136.2, 123.5, 128.1, 127.9, 81.6, 66.9, 58.0, 57.7, 55.5,
55.3, 53.5, 50.5, 37.7, 37.6, 31.1, 30.8, 30.5, 28.1, 27.6, 27.0,
24.3, 24.1, 19.6, 19.4, 19.1, 18.1, 15.5, 15.2, 14.9, 13.1, 11.2,
10.9; HRMS m/z calcd for C30H51N5O7 578.3554 (M + H), found
(45) Mauger, A. B.; Desai, R. B.; Rittner, I.; Rzeszotarski, W. J . J .
Chem. Soc., Perkin Trans. 1 1972, 2146-2148.

Probing Host-Selective Phytotoxicity
J . Org. Chem., Vol. 66, No. 23, 2001 7839
from hydrogenolysis of 21 (122 mg from 150 mg, 0.30 mmol of
21) and 15a (135 mg from 175 mg, 0.30 mmol of 15a ) in DMF
(3 mL) at -10 °C. After 1 h at -10 °C, 1 h at 0 °C, and 24 h
at room temperature, the reaction mixture was diluted with
CH2Cl2 and processed as above to give 23a as a white solid
(241 mg, 97%). F r om 20 a n d 25a . DCC (27 mg, 0.13 mmol)
was added to a stirred solution of 25a (75 mg, 0.11 mmol), 20
(25 mg, 0.13 mmol), and DMAP (1.3 mg) in CH2Cl2 (0.5 mL)
at 0 °C. After 1 h at 0 °C and 24 h at room temperature, the
reaction mixture was processed as above to give 23a (84 mg,
88%): mp 84-86 °C; [R]D -150 (c 1.0, CH3OH); 1H NMR (500
MHz, CDCl3) δ (several rotamers, partial data only) 8.93, 7.91,
7.04, 6.47, 6.44, 5.80, and 5.45 (4H, 7 × br s, HN), 3.26, 3.03,
2.97, and 2.80 (6H, 4 × s, H3CN), 1.29 (18H, s, (H3C)3C × 2),
1.32 and 1.26 (3H, 2 × d, J ) 7 Hz, H3C-3 MeAla), 0.96-0.70
(18H, m, H3CC × 6); 13C NMR (125 MHz, CDCl3) δ 173.2,
172.3, 171.9, 170.6, 169.9, 169.8, 165.0, 157.9, 154.8, 81.5, 71.3,
71.2, 60.3, 60.2, 58.1, 53.5, 53.4, 46.6, 46.5, 39.0, 36.6, 34.4,
30.4, 28.4, 28.1, 27.1, 24.7, 24.4, 24.3, 23.2, 21.5, 18.0, 15.2,
13.1, 10.8, 10.6; HRMS m/z calcd for C40H71N7O11 826.5290 (M
+ H), found 826.5257 (FAB). Anal. Calcd for C40H71N7O11: C,
58.16; H, 8.66; N, 11.87. Found: C, 57.81; H, 8.93; N, 12.24.
Hm p -P r o-Ile-MeVa l-MeAla -NHNH-Boc (25a ). The ben-
zyl ester 22 (190 mg, 0.60 mmol) and 15a (290 mg, 0.50 mmol)
were individually deprotected according to the general proce-
dure for hydrogenolysis. DCC (134 mg, 0.65 mmol) was added
to a stirred solution of the crude deprotected products (138
mg from 22, 222 mg from 15a ) and HOBt hydrate (0.113 mg)
in CH2Cl2 (0.75 mL) at 0 °C. After 10 min at 0 °C and 24 h at
room temperature, the reaction mixture was diluted with ethyl
acetate (ca. 3 mL) and the precipitated DCU was filtered off.
The combined filtrate and washings were washed sequentially
with 0.5 M aqueous citric acid and saturated aqueous NaHCO3,
dried over Na2SO4, concentrated, and fractionated by DFC
(30-80% ethyl acetate in hexane) to give 25a as a white solid
(250 mg, 76%): mp 106-108 °C; [R]D -180 (c 1.0, CH3OH);
1H NMR (500 MHz, CDCl3) δ (several rotamers) 9.29 (br s),
7.92 (br s), 7.18 (d, J ) 9 Hz), 6.64 (br s) and 6.45 (br s) (3H,
HN), 5.25-5.05 (2H, m), 4.80-4.70 (1H, m), 4.57-4.47 (1H,
m), 4.35-4.24 (1H, m), 3.71-3.60 (1H, m), 3.46-3.35 (1 H,
m), 3.32, 3.09, 3.07, 2.87 (1 H, br s), 3.75-3.60 (6H, 4 × s,
H3CN), 2.40-1.65 (10H, m), 1.39 (9H, s, (H3C)3C), 1.34 and
1.26 (3H, 2 × d, J ) 6.5 Hz, H3C-3 MeAla), 1.50-1.00 (3H,
m), 1.00-0.70 (18H, m, H3C × 6); 13C NMR (125 MHz, CDCl3)
δ 174.5, 174.3, 172.8, 171.6, 170.9, 170.5, 168.6, 155.0, 81.5,
68.1, 60.3, 58.1, 57.8, 54.0, 53.6, 50.5, 46.5, 43.4, 37.6, 37.2,
30.5, 28.1, 28.0, 27.0, 24.7, 24.5, 24.2, 23.6, 21.3, 19.6, 19.0,
18.1, 15.5, 13.3, 11.2, 10.9; HRMS m/z calcd for C32H58N6O8
661.4476 (M + Li), found 661.4506 (FAB). Anal. Calcd for
(0.84 mL, 0.77 g, 7.3 mmol) in CH2Cl2 (5 mL) at 0 °C under
argon. After 30 min at 0 °C and 14 h at room temperature,
the mixture was diluted with ether, washed with water, dried
over Na2SO4, concentrated, and fractionated by FCC (5% ether
in hexane) to give the titled compound (0.850 g, 90%): 1H NMR
δ 3.66 (3H, s), 2.41 (2H, ap t, J ) 8 Hz), 1.75 (2H, tap, J ) 8
Hz), 1.20 (6H, s), 0.93 (9H, t, J ) 8 Hz), 0.56 (6H, q, J ) 8
Hz); 13C NMR δ 176.1 (s), 73.1 (s), 51.6 (q), 40.0 (t), 30.1 (q ×
2), 29.7 (t), 7.3 (q × 3), 6.9 (t × 3); HRMS m/z calcd for
C
13H28O3Si 261.1886 (M + H), found 261.1885 (CI).
4-Tr ieth ylsilyloxy-4-m eth ylp en ta n oic Acid (30). LiOH‚
H2O (1.55 g, 37 mmol) was added to a stirred solution of 29
(1.30 g, 5.00 mmol) in THF-CH3OH-H2O (2:2:1 v/v, 150 mL)
at 0 °C. After 15 min at 0 °C and 3 h at room temperature,
the organic solvents were removed on a rotary evaporator and
the remaining aqueous soution was acidfied to pH 4-5 by
addition of aqueous citric acid (1 M). Excess NaCl was added,
and the mixture was extracted with ether (3 × 50 mL). The
combined orgainc layers were dried over Na2SO4, concentrated,
and fractionated by FCC (30% ethyl acetate in hexane) to give
30 (1.12 g, 91%): 1H NMR δ 2.46 (2H, ap t, J ) 8 Hz), 1.76
(2H, ap t, J ) 8 Hz), 1.22 (6H, s), 0.93 (9H, t, J ) 8 Hz), 0.57
(6H, q, J ) 8 Hz); 13C NMR δ 181.0 (s), 72.7 (s), 39.5 (t), 29.9
(q × 2), 29.7 (t), 7.2 (q × 3), 6.8 (t × 3); HRMS m/z calcd for
C
12H26O3Si 247.1729 (M + H), found 247.1734 (FAB).
(4R,5S)-(+)-4-Met h yl-5-p h en yl-3-[4-t r iet h ylsilyloxy-4-
m eth ylp en ta n oyl]-2-oxa zolid in on e (32). Pivaloyl chloride
(0.39 mL, 0.38 g, 3.2 mmol) was slowly added via syringe to a
stirred solution of acid 30 (739 mg, 3.0 mmol) and Et3N (0.46
mL, 334 mg, 3.3 mmol) in THF (3 mL) at -78 °C under argon.
The mixture (a thick white paste formed) was allowed to warm
to 0 °C, and after 1 h, a solution of 31 (483 mg, 2.73 mmol),
DMAP (60 mg, 0.5 mmol), and Et3N (0.42 mL, 0.30 g, 3 mmol)
in THF (2.5 mL) was added. After 14 h at room temperature,
the reacton mixture was diluted with ethyl acetate and washed
sequentially with aqueous NaOH (1 M) and brine, dried over
Na2SO4, concentrated, and fractionated by FCC (10-60% ethyl
acetate in hexane) to give recovered oxazolidinone 31 (23 mg,
5%) and the titled compound (1.03 g, 93%): [R]D +20 (c 0.8,
CH2Cl2); 1H NMR δ 1.73 (2H, m), 1.23 (6H, s), 0.93 (9H, t, J )
7 Hz), 0.85 (3H, d, J ) 7 Hz), 0.56 (6H, q, J ) 7 Hz); 13C NMR
δ 173.8 (s), 153.2 (s), 133.7 (s), 128.7 (d × 3), 125.9 (d × 2),
79.0 (d), 72.7 (s), 54.9 (q), 39.1 (t), 31.4 (t), 30.1 (q), 29.9 (q),
14.7 (q), 7.2 (q × 3), 6.9 (t × 3); HRMS m/z calcd for C22H35
-
NO4Si 412.2495 (M + Li), found 412.2501 (FAB, LiBr). Anal.
Calcd for C22H35NO4Si: C, 65.15; H, 8.70; N, 3.45. Found: C,
64.96; H, 8.70; N, 3.58.
(4R,5S,2′R)-(+)-4-m et h yl-5-p h en yl-3-[2-h yd r oxy-4-t r i-
eth ylsilyloxy-4-m eth ylp en ta n oyl]-2-oxa zolid in on e (33).
A solution of 32 (920 mg, 2.27 mmol) in THF (3.4 mL) was
added dropwise via syringe to a stirred solution of NaHMDS
(1 M solution in THF; 2.72 mL, 2.72 mmol) in THF (3.4 mL)
at -78 °C under argon. After 30 min, a precooled (-78 °C)
solution of 2-(phenylsulfonyl)-3-phenyloxaziridine35 (839 mg,
3.18 mmol) in THF (3.4 mL) was added via cannula. After 10
min, the reactiion was quenched by addition of a solution of
HOAc (0.69 mL) in THF (13.5 mL). The mixture was diluted
with hexane/CH2Cl2 (4:1 v/v, 85 mL) and was washed sequen-
tially with 50 mL each of 5% NaHCO3, saturated Na2SO3, 1
M KHSO4, 5% NaHCO3, and brine. The organic layer was
dried, concentrated, and fractionated by FCC (15% ethyl
acetate in hexane) to give the crude product, which was further
fractionated by FCC (0.5% CH3OH in CH2Cl2) to give the titled
compound (802 mg, 84%) as a viscous oil: [R]D +24 (c 3.5, CH2-
Cl2); 1H NMR δ 7.45-7.15 (5H, m), 5.66 (1H, d, J ) 7 Hz),
5.47 (1H, br dd, J ) 3, 8.5 Hz), 4.71 (1H, dq, J ) 7, 7 Hz), 4.57
(1H, br s), 1.97-1.80 (2H, m), 1.41 (3H, s), 1.32 (3H, s), 0.93
(9H, t, J ) 8 Hz), 0.88 (3H, d, J ) 7 Hz), 0.59 (6H, q, J ) 8
Hz); 13C NMR δ 173.8 (s), 152.8 (s), 133.3 (s), 128.9 (d × 2),
128.8 (d), 125.7 (d × 2), 79.7 (d), 75.0 (s), 69.3 (d), 55.3 (d),
46.6 (t), 31.0 (q), 29.3 (q), 14.4 (q), 7.1 (q × 3), 6.7 (t × 3); HRMS
m/z calcd for C22H35NO5Si 428.2445 (M + Li), found 428.2442
(FAB, LiBr). Anal. Calcd for C22H35NO5Si: C, 62.67; H, 8.37;
N, 3.32. Found: C, 62.40; H, 8.41; N, 3.30.
C
32H58N6O8‚H2O: C, 57.12; H, 8.99; N, 12.49. Found: C, 57.41;
H, 8.93; N, 12.20.
Meth yl 4-Hyd r oxy-4-m eth ylp en ta n oa te (28). A solution
of 4,4-dimethyl-butyrolactone 2631 (7.90 g, 69 mmol) and NaOH
(3.3 g, 83 mmol) in ethanol/water (2:1 v/v, 100 mL) was heated
under reflux for 2 h. The reaction mixture was concentated to
dryness under high vacuum to give a white solid (11.2 g). A
solution of this solid (4.77 g) in water (20 mL) at 0 °C was
acidified to pH 4-5 by addition of citric acid (ca. 2.5 g). Excess
NaCl was added, and the mixture was extracted with ether
(four times). The combined organic layers were dried over Na2-
SO4 and concentrated to give the unstable 4-hydroxy-4-methyl-
pentanoic acid, which was immediately reacted with excess
CH2N2 in ether solution at 0 °C for 2 h. The resulting mixture
was concentrated and fractionated by FCC (30% ethyl acetate
in hexane) to give 28 (3.05 g, 71%): 1H NMR δ 3.46 (3H, s),
2.60 (2H, ap t, J ) 8 Hz), 2.04 (2H, ap t, J ) 8 Hz), 1.44 (6H,
s); 13C NMR δ 177.0 (s), 84.9 (s), 51.0 (q), 34.9 (t), 29.6 (t), 27.9
(q × 2); HRMS m/z calcd for C7H14O3 147.1021 (M + H), found
147.1019 (CI). In several smaller-scale experiments, the white
solid above was converted into the titled compound in 70-
90% yield. The variable yield is attributed to the instability of
the intermediate acid and the volatility of 28.
Meth yl 4-Tr ieth ylsilyloxy-4-m eth ylp en ta n oa te (29).
TESOTf (97%; 1.01 mL, 1.15 g, 4.36 mmol) was added to a
stirred solution of 28 (0.530 g, 3.63 mmol) and 2,6-lutidine

7840 J . Org. Chem., Vol. 66, No. 23, 2001
Ward et al.
(2′R,4R,5S)-(+)-4-Meth yl-5-p h en yl-3-[4-m eth yl-(2-(p h e-
n ylm et h oxyca r bon yloxy)-4-t r iet h ylsilyloxyp en t a n oyl)-
2-oxa zolid in on e (34). Benzyl chloroformate (0.15 mL, 1
mmol) and DMAP (122 mg, 1 mmol) were added to a stirred
solution of 33 (280 mg, 0.67 mmol) in CH2Cl2 (4 mL) at room
temperature. After 2 h, additional benzyl chloroformate (0.075
mL, 0.5 mmol) and DMAP (61 mg, 0.50 mmol) were added;
after being stirred for 2 h, the reaction mixture was diluted
with CH2Cl2, washed with water, dried over Na2SO4, concen-
trated, and fractionated by FCC (50-75% CH2Cl2 in hexane)
to give the titled compound as an oil (365 mg, 99%): [R]D +27
J ) 14.5 Hz), 1.90 (1H, dd, J ) 9, 14.5 Hz), 1.22 (6H, s), 0.88
(9H, t, J ) 8 Hz), 0.54 (6H, q, J ) 8 Hz); 13C NMR δ 177.4 (s),
155.5 (s), 135.8 (s), 128.7 (d × 2), 128.5 (d × 3), 76.0 (d), 72.7
(s), 69.8 (t), 45.8 (t), 31.2 (q), 30.2 (q), 7.3 (q), 6.8 (t); HRMS
m/z calcd for C20H32O6Si 403.2128 (M + Li), found 403.2141
(FAB, LiBr). Anal. Calcd for C20H32O6Si: C, 60.58; H, 8.15.
Found: C, 60.31; H, 8.15.
Boc-âAla-Dh m p(OTES)-P r o-OBn (Dh m p(OTES) ) (2R)-
2-Hydr oxy-4-tr ieth ylsiloxy-4-m eth yl-pen tan oic Acid) (37).
H2O2 (30%; 0.32 mL, 2.8 mmol), LiOH‚H2O (24 mg, 0.57 mmol),
and a solution of 35 (282 mg, 0.48 mmol) in THF/H2O (4:1 v/v,
3 mL) were added sequentially to a stirred mixture of THF/
H2O (4:1 v/v, 4 mL) at 0 °C under argon. After ca. 20 min, the
reaction was quenched by addition of aqueous NaHSO3 (10%
w/v, 7 mL) and the mixture was diluted with water and
extracted with ethyl acetate (three times). The combined
organic layers were washed sequentially with aqueous citric
acid (1 M) and brine, dried over Na2SO4, concentrated, and
fractionated by FCC (eluting first with 45:45:9:1 CH2Cl2:
hexane:iPrOH:NH4OH and then with 15% CH3OH in CH2Cl2)
to give 31 (84 mg, 100%) and desired didepsipeptide Boc-âAla-
Dhmp(OTES) (156 mg, 76%): 1H NMR δ 5.38 (1H, dd, J ) 5,
5.5 Hz), 5.26 (1H, br s), 3.55-3.35 (2H, m), 2.67-2.52 (2H,
m), 2.18 (1H, dd, J ) 5, 15 Hz), 1.90 (1H, dd, J ) 5.5, 15 Hz),
1.43 (9H, s), 1.34 (3H, s), 1.32 (3H, s), 0.95 (9H, t, J ) 8 Hz),
0.62 (6H, q, J ) 8 Hz). The aqueous layer from the above
extractions was acidified to pH 4 with aqueous citric acid (1
M) and extracted with ethyl acetate; concentration of the
extract gave 20 (18 mg, 20%). DCC (82 mg, 0.40 mmol) was
added to a stirred solution of the above didepsipeptide (134
mg, 0.31 mmol), 18 (76 mg, 0.37 mmol), and HOBt hydrate
(62 mg) in CH2Cl2 (1 mL) at 0 °C. The mixture was stirred at
0 °C for 1 h and at room temperature for 18 h and then diluted
with ethyl acetate, and the precipitated DCU was filtered off.
The combined filtrate and washings were washed sequentially
with 0.5 M aqueous citric acid and saturated aqueous NaHCO3,
dried over Na2SO4, concentrated, and fractionated by FCC
(30% ethyl acetate in hexane) to give the titled compound (156
mg, 81%) (62% overall yield from 35): [R]D -22 (c 1.4, CH3-
1
(c 1.5, CH2Cl2); H NMR δ 7.35 (10H, m), 6.22 (1H, dd, J )
5.5, 6 Hz), 5.70 (1H, d, J ) 7 Hz), 5.18 (1H, d, J ) 12.5 Hz),
5.13 (1H, d, J ) 12.5 Hz), 4.73 (1H, dq, J ) 7, 6.5 Hz), 2.04
(2H, m), 1.36 (3H, s), 1.34 (3H, s), 0.92 (12H, m), 0.58 (6H, q,
J ) 8 Hz); 13C NMR δ 170.9 (s), 154.8 (s), 152.3 (s), 135.3 (s),
133.3 (s), 129.1 (d), 129.0 (d × 2), 128.8 (d × 2), 128.7 (d), 128.4
(d × 2), 125.9 (d × 2), 79.7 (d), 73.9 (d), 72.7 (s), 70.1 (t), 55.5
(d), 45.1 (t), 31.4 (q), 30.4 (q), 14.2 (q), 7.3 (q × 3), 6.9 (t × 3);
HRMS m/z calcd for C30H41NO7Si 562.2812 (M + Li), found
562.2814 (FAB, LiBr). Anal. Calcd for C30H41NO7Si: C, 64.84;
H, 7.44; N, 2.52. Found: C, 64.65; H, 7.22; N, 2.78.
(4R,5S,2′R)-(+)-4-Meth yl-5-p h en yl-3-[3-(t-bu toxyca r bo-
n yla m in o)p r op a n oyloxy-4-tr ieth ylsilyloxy-4-m eth ylp en -
ta n oyl]-2-oxa zolid in on e (35). DCC (160 mg, 0.77 mmol) was
added to a stirred solution of 33 (270 mg, 0.64 mmol), 20 (133
mg, 0.71 mmol), and DMAP (17 mg, 0.14 mmol) in CH2Cl2 (1.8
mL) at 0 °C. The mixture was stirred at 0 °C for 1 h and at
room temperature for 20 h and then diluted with ethyl acetate,
and the precipitated DCU was filtered off. The combined
filtrate and washings were washed sequentially with aqueous
citric acid (0.5 M) and saturated aqueous NaHCO3, dried over
Na2SO4, concentrated, and fractionated by FCC (20% ethyl
acetate in hexane) to give the titled compound (374 mg, 99%),
which solidified upon standing in a refrigerator: mp 82-83
°C; [R]D +29 (c 1.2, CH2Cl2); 1H NMR δ 7.45-7.25 (5H, m),
6.23 (1H, dd, J ) 4, 7.5 Hz), 5.70 (1H, d, J ) 6.5 Hz), 5.21
(1H, br s), 4.72 (1H, dq, J ) 6.5, 7 Hz), 3.50-3.35 (2H, t), 2.62-
2.53 (2H, t), 2.07-1.95 (2H, m), 1.41 (9H, s), 1.35 (3H, s), 1.33
(3H, s), 0.93 (9H, t, J ) 8 Hz), 0.88 (3H, d, J ) 7 Hz), 0.59
(6H, q, J ) 8 Hz); 13C NMR δ 171.8 (s), 171.0 (s), 155.9 (s),
152.2 (s), 133.3 (s), 129.0 (d), 128.8 (d × 2), 125.8 (d × 2), 79.5
(s), 79.0 (s), 72.7 (s), 71.1 (d), 55.3 (d), 44.7 (t), 36.5 (t), 43.7
(t), 31.2 (q), 30.2 (q), 28.4 (q × 3), 14.0 (q), 7.2 (q), 6.8 (t); HRMS
m/z calcd for C30H48N2O8Si 599.3340 (M + Li), found 599.3344
(FAB, LiBr). Anal. Calcd for C30H48N2O8Si: C, 60.78; H, 8.16;
N, 4.73. Found: C, 60.85; H, 8.18; N, 4.69.
(2R)-4-Tr iet h ylsilyloxy-4-m et h yl-2-(p h en ylm et h oxy-
ca r bon yloxy)p en ta n oic Acid [Cbz-Dh m p (4-OTES)] (36).
Hydrogen peroxide (30%; 0.52 mL, 4.59 mmol) and LiOH (1
M in H2O; 0.9 mL, 0.9 mmol) were added to a stirred solution
of 34 (170 mg, 0.31 mmol) in THF (3.2 mL) at 0 °C under
argon. After 75 min, 10% aqueous Na2SO3 (10 mL) was added
and the mixture stirred at room temperature for 20 min. The
reaction mixture was diluted with ethyl acetate, washed
sequentially with citric acid (0.75 M) and water, dried over
Na2SO4, concentrated, and fractionated by FCC (eluting first
with 45:45:9:1 hexane:CH2Cl2:iPrOH:NH4OH and then with
15% CH3OH in CH2Cl2) to give 31 (50 mg, 92%) and the titled
compound (110 mg, 91%): [R]D +52 (c 0.89, CH2Cl2); 1H NMR
δ 8.45-7.95 (1H, br s), 7.31 (5H, m), 5.20 (1H, d, J ) 12 Hz),
5.08 (1H, br d, J ) 9 Hz), 5.07 (1H, d, J ) 12 Hz), 2.09 (1H, d,
1
OH); H NMR δ 7.42-7.25 (5H, m), 5.53-5.44 (2H, m), 5.20
(2H, ap s), 4.47 (1H, dd, J ) 4, 8 Hz), 3.89-3.76 (1H, m), 3.72-
3.60 (1H, m), 3.45-3.32 (2H, m), 2.57-2.47 (2H, m), 2.22-
1.75 (8H, m), 1.41 (9H, s), 1.27 (3H, s), 1.24 (3H, s), 0.93 (9H,
t, J ) 8 Hz), 0.58 (6H, q, J ) 8 Hz); 13C NMR δ 171.6 (s × 2),
169.7 (s), 156.2 (s), 136.0 (s), 128.6 (d × 2, C6H5), 128.3 (d ×
3), 79.1 (s), 72.6 (s), 69.5 (d), 66.8 (t), 59.5 (d), 46.6 (d), 45.7
(t), 36.6 (t), 35.0 (t), 31.8 (t), 29.1 (q × 2), 28.6 (q × 3), 24.8 (t),
7.2 (q × 3), 6.7 (t × 3); HRMS m/z calcd for C32H52N2O8Si
627.3653 (M + Li), found 627.3643 (FAB, LiBr). Anal. Calcd
for C32H52N2O8Si: C, 61.91; H, 8.44; N, 4.51. Found: C, 61.95;
H, 8.39; N, 4.59.
Ackn owledgm en t. Financial support from the Natu-
ral Sciences and Engineering Research Council (Canada)
is gratefully acknowledged.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and spectroscopic data for 2, *2, 3-5, 11b, 11c, 13b,
13c, 15b, 15c, 23b, 23c, 25b, 41a , and 41b. This material is
available free of charge via the Internet at http://pubs.acs.org.
J O015953+