Communications to the Editor
J. Am. Chem. Soc., Vol. 123, No. 49, 2001 12413
CH2Cl2) and coupled with 917 (EDCI, HOAt, DMF) to afford
dipeptide 10 (Scheme 3). Installation of the N-trifluoroacetamide
protecting group (TFA, DMS, CH2Cl2; then TFAA, 2,6-lutidine,
CH2Cl2)18 afforded phenolic dipeptide 11, which was now
positioned for the Cu(II)-promoted phenolic arylation. Diaryl ether
coupling between 8 and 11 (Cu(OAc)2, pyridine, 4 Å sieves, O2,
CH2Cl2) proceeded smoothly to provide 12 in 80% yield. In accord
with our previous study,11a no epimerization of either arylglycine
residue was detected. Saponification of the methyl ester in 12
was accomplished with LiOH (3:1 MeOH:H2O, 0 °C), again
without any detectable nucleophilic aromatic substitution or
epimerization, providing macrocyclization precursor 13.
After Boc deprotection of 13 (TFA, DMS, CH2Cl2), initial
attempts at macrolactamization of the amino acid derived from
13 (HATU, HOAt, CH2Cl2-DMF) resulted in low isolated yields
(<5%) of macrolactam 14. It was quickly realized that the desired
macrolactam is almost completely insoluble in standard solvents
(including MeOH, CH2Cl2, THF, EtOAc, MeCN, H2O and
mixtures thereof), and could be manipulated only in DMF or
DMSO. We reasoned that the low yields resulted from material
loss during the isolation and purification. Indeed, amide 14
precipitated from the reaction mixture during amide formation
under high dilution (1 × 10-5 M in 19:1 CH2Cl2:DMF) and could
be isolated by simple filtration of the entire reaction mixture.
Purification was effected by dissolution of 14 in DMF followed
by precipitation of the desired material by the addition of H2O.
Mass recovery of over 90% was consistently obtained when using
this procedure. Further attempts to purify 14 by normal or reverse-
phase chromatography resulted in substantial material loss.
Deprotection of the N-methylamide moiety in monocycle 14,
in preparation for coupling with the M(4-6)(5-7) bicycle 16,
proved challenging. We had anticipated using our two-step
nitrosation/hydrolysis procedure,19 which had previously proven
successful for complex peptidic systems.20 Yet, nitrosation with
N2O4 in CH2Cl2 or MeCN failed, presumably due to the
insolubility of 14. We then turned to DMF as a nitrosation solvent
(N2O4, 0 °C) in the presence of sodium acetate as an acid
scavenger. These conditions led to sluggish nitrosation and
incomplete conversion. However, in the absence of added base,
very clean and complete mono-nitrosation could be effected in
DMF. Conversion of the intermediate nitrosamide to the car-
boxylic acid with LiOOH (3:1 THF:H2O, 0 °C) resulted in
extensive decomposition and apparent epimerization of residue-
3. On the other hand, clean hydrolysis was observed by heating
of the nitrosamide in 2:1 DMF:H2O (6 h, 60 °C). This procedure
resulted in quantitative mass recovery of unpurified 15. Macro-
cyclic acid 15 displayed solubility characteristics similar to those
exhibited by amide 14 and was used without purification.
In agreement with observations by Boger,6 peptide coupling
of 15 and 1621 utilizing DEPBT22 (DMF, -5 °C) in the absence
of base afforded tricycle 17 in good yield as an inseparable 12:1
mixture of position-3 epimers. This coupling procedure was far
superior to other coupling agents screened, such as HATU/2,6-
lutidine, which promoted extensive epimerization and provided
only a 3:1 mixture of position-3 epimers. Nucleophilic aromatic
substitution (CsF, DMF, 10 °C)6b,10 proceeded with high atrop-
diastereoselectivity (>15:1) to afford 18 as a single diastereomer
after purification containing the entire tetracyclic core of teico-
planin aglycon. The favorable selectivity noted here strongly
suggests that the M(1-3) diaryl ether macrocycle present in 17
enhances the atropdiastereoselectivity noted for closure of the
M(2-4) macrocycle. The analogous ring closure first precedented
in our vancomycin synthesis proceeded with only 5:1 atropdiastereo-
selectivity.5a,5b
Reduction of the nitro moiety in 18 (1 atm H2, 10% Pd/C, 6:1
EtOAc:EtOH) and Sandmeyer reaction (t-BuONO, HBF4, MeCN;
then CuCl, CuCl2, H2O) afforded 19 bearing the requisite chlorine
substituent on ring-2. Deprotection of the carboxy-terminal
N-methylamide 19 to acid 20 was then accomplished in 85% yield
by successive nitrosation (N2O4, DMF, 0 °C) and pH neutral
hydrolysis, as previously described in the transformation of 14
f 15 (2:1 DMF:H2O, 7 h, 60 °C). The high site selectivity and
yield of this amide deprotection sequence demonstrates that amidic
protection of carboxylic acids is a viable strategy for complex
molecules containing multiple amides. Finally, global demethy-
lation and N-terminal trifluoroacetamide hydrolysis were effected
by treatment with AlBr3 and EtSH (CH2Br2, 0 °C to room
temperature) to provide teicoplanin aglycon 2 that was spectro-
scopically and analytically identical with material derived from
natural sources.23
(15) Abbreviations: TFA ) trifluoroacetic acid; DMS ) dimethyl sulfide;
EDCI ) 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide; HOAt ) 1-hy-
droxy-7-azabenzotriazole; TFAA ) trifluoroacetic anhydride; HATU ) 2-(1-
H-7-azabenzotriazol)-1-1,1,3,3-tetramethyluronium hexafluorophosphate;
DEPBT ) 3-(diethyloxyphosphoryloxy)-1,2,3-benzotriazin-4(3H)-one.
(16) 3-bromo-5-methoxy-phenylglycine 7 was synthesized in 3 steps from
3-bromo-5-methoxy-styrene: i) Sharpless asymmetric amino-hydroxylation
(BocNClNa, K2OsO2(OH)4, (DHQ)2PHAL, n-PrOH, H2O, 80%), See: Reddy,
K. L.; Sharpless, K. B. J. Am. Chem. Soc. 1998, 120, 1207-1217; ii) oxidation
to the carboxylic acid (TEMPO, NaOCl, KBr, acetone, H2O); iii) protection
of the carboxylic acid as its N-methyl amide (i-BuOC(O)Cl, NMM, EtOAc;
then MeNH2, 60-65% for 2 steps).
(17) Compound 9 was synthesized in four steps from commercially
available 3-benzyloxy-4-methoxy benzaldehyde: i) Wittig olefination (Ph3-
PCH3Br, KHMDS, THF, 96%); ii) Sharpless AA (BocNClNa, K2OsO2(OH)4,
(DHQD)2PHAL, n-PrOH, H2O); iii) oxidation to the carboxylic acid (TEMPO,
NaOCl, KBr, acetone, H2O, 70-81% for 2 steps); iv) hydrogenolysis (1 atm
H2, 10% Pd/C, EtOH, quant.).
(18) Because urethanes are highly susceptible to nitrosation, a protecting
group change at this point is required in advance of the carboxyl deprotection
step (19f20).
(19) Evans, D. A.; Carter, P. H.; Dinsmore, C. J.; Barrow, J. C.; Katz, J.
L.; Kung, D. W. Tetrahedron Lett. 1997, 38, 4535-4538.
(20) Evans, D. A.; Barrow, J. C.; Watson, P. S.; Ratz, A. M.; Dinsmore,
C. J.; Evrard, D. A.; DeVries, K. M.; Ellman, J. A.; Rychnovsky, S. D.; Lacour,
J. J. Am. Chem. Soc. 1997, 119, 3419-3420. See also ref 5b.
Acknowledgment. Support for this research was provided by the NIH
(GM 33328-18 and GM 43912-11). G.S.P. acknowledges the DOD for a
predoctoral fellowship.
Supporting Information Available: Spectral data for all compounds
are provided (PDF). This material is available free of charge via the
JA011943E
(21) Bicycle 16 was synthesized in analogy with previously published M(4-
6)(5-7) bicyclic systems, see ref 5b.
(22) Li, H.; Jiang, X.; Ye, Y.-H.; Fan, C.; Romoff, T.; Goodman, M. Org.
Lett. 1999, 1, 91-93.
(23) Teicoplanin aglycon was obtained by acidic hydrolysis of natural
teicoplanin complex (80% H2SO4, DMSO, 85 °C, 48 h), See: Boger, D. L.;
Weng, J.-H.; Miyazaki, S.; McAtee, J. J.; Castle, S. L.; Kim, S. H.; Mori, Y.;
Rogel, O.; Strittmatter, H.; Jin, Q. J. Am. Chem. Soc. 2000, 122, 10047-
10055.