Angewandte
Chemie
The synthesis of the second peptide building block,
tetrapeptide 8, commenced with the synthesis of a suitable
(R)-3-hydroxy-l-homotyrosine derivative.[5,19] Toward this
end, the commercially available primary alcohol 19 was first
oxidized to the corresponding aldehyde and then converted to
the known E-olefinic ester 20[20] by a Wittig olefination (81%;
Scheme 5). A SAD reaction with AD-mix-b proceeded
Finally, the N-terminal Boc-group was removed under
acidic conditions to provide the tetrapeptide building block 8.
The final assembly of the echinocandin C cyclopeptide
started with the C-terminal oxidation of dipeptide 18
(Scheme 7). Acidic removal of the N,O-acetal (TsOH/
MeOH) did not affect the hemiaminal and afforded primary
alcohol 25 in good yield (87%). Direct oxidation (TEMPO,
NaOCl, NaClO2) to the corresponding crude carboxylic acid 5
and coupling with amine 8 then provided the linear hexapep-
tide 26 (87%, 2 steps). To achieve good cyclization yields, it
was imperative to remove the Teoc group from the N-acyl
hemiaminal (TBAF, 98%) to afford hexapeptide 4. Libera-
tion of the N-terminus by removal of the Alloc group (Pd0,
thiosalicylic acid, 90%) and C-terminal ester hydrolysis with
LiOH in THF then set the stage for the cyclization. Here,
activation of the carboxylic acid with DEPBT in DMF and
addition of solid NaHCO3 as a base were found to provide
optimal cyclization conditions, which afforded cyclopeptide
27 in excellent yield (90%, 2 steps). This first total synthesis of
an echinocandin-type cyclopeptide with N-acyl hemiaminal
unit (longest linear sequence: 21 steps starting from d-serine,
4.2% overall yield) easily provides 27 in quantities of more
than 200 mg, which can be used for the synthesis of
echinocandin C (2) and various lipidated analogues.
Scheme 5. Synthesis of the protected (R)-3-hydroxyhomotyrosine deriv-
ative 22. a) DMP, CH2Cl2; b) Ph3P=CHCO2Et, THF, reflux, 81% (2
steps); c) AD-mix-b, CH3SO2NH2, tBuOH/H2O, 08C, 92%; d) SOCl2,
Et3N, CH2Cl2, 08C; e) NaIO4, RuCl3 (cat.), CH3CN/H2O; f) LiBr, THF,
then H2SO4 (20%), Et2O; g) NaN3, DMSO, 65% (4 steps).
DMP=Dess–Martin periodinane.
Toward this end, the removal of the Bn and Cbz groups
was initially attempted under standard conditions for hydro-
genation (H2 (1 bar, 10 bar or 48 bar), Pd/C, MeOH or Raney-
Nickel, EtOH) and transfer hydrogenation (NH4HCO2, Pd/C,
40–608C, MeOH). However, the global deprotection pro-
ceeded very slowly, which ultimately led to partial degrada-
tion of the cyclopeptide. A solution to this problem was
eventually found by using cyclohexene as the hydrogen source
in the transfer hydrogenation (EtOH, 508C), which acceler-
ated the deprotection substantially and provided the depro-
tected cyclopeptide after 24 h as the major product (as judged
by analytical HPLC). The final acylation at the a-amino
group of the Dho unit was achieved employing known
conditions for the lipidation of unprotected echinocandin
cyclopeptides.[23] In the event, reaction of the deprotected
cyclopeptide with the HOBt ester of linoleic acid and
subsequent purification by silica gel chromatography pro-
vided echinocandin C (2) as a colorless solid with a melting
point and an optical rotation that corresponds well to
published data.[3b] Furthermore, the 1H and 13C NMR spectra
in combination with extensive 2D NMR experiments proved
the correct structure of synthetic echinocandin C (2).
efficiently to afford 2,3-diol 21 (92%), which was then
transformed to the corresponding 2-azido derivative 22. To
achieve this goal under retention of the C2 configuration,
a double inversion strategy was employed, which features the
opening of the corresponding cyclic sulfate with LiBr and
a second substitution of the bromide with NaN3.[21] This
sequence conveniently afforded azide 22 in good overall yield
(65%) and in multigram quantities.
Following Staudinger reduction of azide 22 (97%,
Scheme 6), the amine thus obtained was coupled with Boc-
l-Hyp by activation with DEPBT[22] to afford dipeptide 23
(87%). Next, the fully protected tetrapeptide 24 was gen-
erated by coupling with appropriate l-Thr derivatives with
DEPBT activation following the removal of the N- and C-
terminal protecting groups, respectively (75%, 4 steps).
The small scale lipidation of unprotected echinocandins
with this type of reagent and subsequent chromatographic
purification is known to proceed with moderate yields,[23] but
affords various lipid derivatives in a highly convergent
fashion. Accordingly, we easily prepared a second echinocan-
din derivative, namely the known tetrahydro derivative 28 of
echinocandin C, by deprotection of 27 and subsequent
acylation with stearic acid HOBt ester (Scheme 7). The 13C
spectral data of this compound matched the reported data
completely.[3b]
Scheme 6. Synthesis of tetrapeptide 8. a) PPh3, THF/H2O, reflux, 97%;
b) Boc-l-Hyp, DEPBT, DIPEA, THF, 87%; c) TFA/CH2Cl2, 08C; d) Boc-
l-Thr, DEPBT, DIPEA, THF, 92% (2 steps); e) aq. LiOH, THF, 08C;
f) l-Thr-OMe, DEPBT, DIPEA, THF, 81% (2 steps); g) TFA/CH2Cl2,
08C, quant. DEPBT=3-(diethoxyphosphoryloxy)-1,2,3- benzotriazin-
4(3H)-one, TFA=trifluoroacetic acid.
In conclusion, we have developed an efficient synthesis of
echinocandin C (2), which features an early introduction of
the N-acyl hemiaminal by a Curtius rearrangement. The
Angew. Chem. Int. Ed. 2013, 52, 1 – 6
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3
These are not the final page numbers!