Total synthesis of sperabillin A and C

Article abstract of DOI:10.1055/s-2005-917106

The first total synthesis of sperabillin A and an improved total synthesis of sperabillin C have been achieved in 11 steps from N-Boc-Omethyl-L-tyrosine. The stereoselective pathway to the core (3R,5R)-3,6-diamino-5-hydroxyhexanoic acid involves an Arndt-

Full text of DOI:10.1055/s-2005-917106

LETTER
2615
Total Synthesis of Sperabillin A and C
Sperabillins ars Allmendinger, Gerd Bauschke, Franz F. Paintner*
Department Pharmazie – Zentrum für Pharmaforschung, Ludwig-Maximilians-Universität München, Butenandtstrasse 5-13, Haus C,
81377 München, Germany
Fax +49(89)218077247; E-mail: franz.paintner@cup.uni-muenchen.de
Received 22 August 2005
Dedicated to Prof. Dr. Dr. h.c. Fritz Eiden on the occasion of his 80th birthday.
OH NH2
O
OH NH2
O
Me
N
O
Abstract: The first total synthesis of sperabillin A and an improved
total synthesis of sperabillin C have been achieved in 11 steps from
N-Boc-O-methyl-L-tyrosine. The stereoselective pathway to the
core (3R,5R)-3,6-diamino-5-hydroxyhexanoic acid involves an
Arndt–Eistert homologation, an asymmetric Henry reaction and a
ruthenium tetroxide-catalyzed oxidative degradation of a benzene
ring as key steps.
H2N
H2N
OH
N
OH
H
negamycin (2)
OH NH2
1
O
NH
H
N
3
R
NH2
N
H
1
2
Key words: sperabillins, antibiotics, pseudo-peptides, Henry reac-
R
O
R
tion, ruthenium tetroxide
sperabillin A (3a): R1 = H
sperabillin B (3b): R1 = Me
sperabillin C (3c): R1 = H
sperabillin D (3d): R1 = Me
R2 = Me R3 = H
2 = Me R3 = H
R2 = H R3 = Me
R2 = H R3 = Me
R
(3R,5R)-3,6-Diamino-5-hydroxyhexanoic
acid
(1,
Figure 1) is the core fragment of the pseudo-peptide anti-
biotics negamycin (2) and sperabillin A and C (3a and 3c,
respectively). Negamycin (2), isolated in 1970 from cul-
Figure 1
ture filtrates of strains closely related to Streptomyces pur- carboxylic acid, and subsequent reduction of the nitro
peofuscus by Umezawa et al., shows potent growth- group.
inhibitory activity against Gram-negative bacteria.1 By
contrast sperabillin A (3a), the main member of the sper-
abillin family of antibiotics, which was isolated in 1986
Aldehyde 6 was readily obtained from commercially
available N-Boc-O-methyl-L-tyrosine (7) in two steps
(Scheme 2). Arndt–Eistert homologation of the a-amino
from the culture broth of Pseudomonas fluorescens YK-
acid 7 following a standard protocol6 afforded b-amino
437 by researchers at Takeda Chemical Industries Ltd.,
ester 8 {[a]D23 –9.3 (c 0.8 in CH2Cl2); lit.7 (R)-8: [a]D +9.2
(c 1.2 in CH2Cl2)}, which subsequently was reduced with
DIBAL-H (toluene, –85 °C) to give 6 in 73% overall
yield. Previous studies conducted in this and other labora-
tories have indicated that it is difficult to achieve control
of stereochemistry for Henry reactions of nitromethane
with chiral aldehydes like 6 bearing the inducing stereo-
exerts promising in vitro and in vivo antibacterial activity
especially against Gram-positive pathogens, including
multiresistant strains of Staphylococcus aureus.2 While
there are numerous reports in the literature concerning the
synthesis of negamycin (2), both in racemic and optically
active form,3 little attention has been given to the synthe-
sis of sperabillin A (3a) and C (3c) incorporating the same
genic center in the b-position.8 Therefore we envisioned
core b,e-diamino acid. Only one total synthesis of spera-
the application of a chiral catalyst, namely Evans’ bis(ox-
billin C (3c; 15 steps starting from N-Boc-glycine, 2.5%
azoline) copper(II) acetate-based catalyst (+)-11,9 to con-
overall yield) has been reported to date.4,5
trol the configuration of the newly created stereogenic
In this communication, we describe the first total synthe- center at C-5.8c Indeed, Henry reaction of nitromethane
sis of sperabillin A (3a) as well as an improved total syn- with aldehyde 6 in the presence of 5 mol% of (+)-11
thesis of sperabillin C (3c). Our approach toward these
antibiotics and analogues with modified fragments at-
tached to the Ne- and C-terminal regions is based on the
orthogonally protected b,e-diamino acid 4. As outlined in
Scheme 1 building block 4 was envisioned to be readily
accessible from N-Boc-L-tyrosine derived aldehyde 6 by
two key transformations: (a) stereoselective Henry reac-
tion to give b-nitro alcohol 5 and (b) oxidative degrada-
tion of the phenyl ring, which serves as a masked
SYNLETT 2005, No. 17, pp 2615–2618
1
7
.1
0
.2
0
0
5
Advanced online publication: 05.10.2005
Scheme 1
DOI: 10.1055/s-2005-917106; Art ID: D24805ST
© Georg Thieme Verlag Stuttgart · New York

2616
L. Allmendinger et al.
LETTER
pared in four steps, starting with a Sonogashira coupling
of (Z)-1-bromoprop-1-ene and propargylic alcohol
[(Ph3P)2PdCl2 (0.25 mol%), CuI (1.5 mol%), Ph3P, n-
PrNH2, Et2O, D) to give (Z)-hex-4-en-2-yn-1-ol (12)16 in
69% yield (Z:E > 99:1; Scheme 3).17 This product was ste-
reoselectively reduced with LiAlH4 (THF, D)18 providing
(2E,4Z)-hexa-2,4-dien-1-ol (13) in 94% yield
[(2E,4Z):(2Z,4Z) > 99:1]. Oxidation of 13 was then ac-
complished by a one-pot, two-step procedure. First, the al-
lylic alcohol 13 was oxidized with manganese dioxide
(pentane, r.t.) to give the corresponding aldehyde, which
was further oxidized by the addition of silver nitrate
(KOH, EtOH–H2O, r.t.) to afford diastereoisomerically
pure (2E,4Z)-2,4-hexadienic acid (14) in 77% overall
yield. Finally, the carboxylic acid 14 was converted into
the corresponding pentafluorophenyl ester 15 (C6F5OH,
EDC, CH2Cl2, r.t.) in 94% yield.
Br
a
+
OH
OH
OH
OH
12
14
Scheme 2 Reagents and conditions: (a) i) isobutyl chloroformate,
NMM, DME, –20 °C; ii) CH2N2, –20 °C to r.t.; (b) silver benzoate,
Et3N, MeOH, r.t. (86%, two steps); (c) DIBAL-H, toluene, –85 °C
(85%); (d) MeNO2, (+)-11 (5 mol%), EtOH, r.t. (83%); (e) TB-
DMSCl, imidazole, DMF, r.t. (82%); (f) RuO2 (5 mol%), NaIO4,
NaHCO3, EtOAc–H2O, r.t. (67%); (g) HCO2NH4, Pd/C, MeOH, r.t.
(90%)
O
b
d
c
13
15
O
OC6F5
(EtOH, r.t., 3 d) afforded the desired b-nitro alcohol 5
with both high diastereoselectivity (95% de)10 and yield
(83%).
Scheme 3 Reagents and conditions: (a) (Ph3P)2PdCl2, CuI, Ph3P, n-
PrNH2, Et2O, D (69%); (b) LiAlH4, THF, D (94%); (c) (i) MnO2, pen-
tane, r.t. (ii) AgNO3, KOH, Et2O–H2O, r.t. (77%); (d) C6F5OH, EDC,
CH2Cl2, 0 °C (94%)
After protecting the secondary alcohol as a tert-butyl-
dimethylsilyl (TBDMS) ether we turned our attention to
the oxidative degradation of the 4-methoxyphenyl ring.
Ruthenium tetroxide-catalyzed degradation of aromatic
rings to carboxylic acids has been repeatedly applied as a
useful tool in the synthesis of amino acids and peptides.11
However, N-acylamines bearing a primary alkyl goup at
the nitrogen were reported to undergo facile oxidation at
the a-position to give the corresponding imides.12 Accord-
ingly, the aromatic ring has to be cleaved prior to the gen-
eration of the C-6 amino group assuming that the nitro
group would act as a ‘protective group’ for amines during
ruthenium tetroxide oxidations. Indeed ruthenium tetrox-
ide-catalyzed degradation of 9 afforded the desired e-nitro
carboxylic acid 10 in good yield (67%). Best results were
obtained using catalytic amounts of ruthenium dioxide (5
mol%) and an excess of sodium metaperiodate (30 equiv)
in a two-phase system of ethyl acetate and water at room
temperature according to Yoshifuji et al.12,13 To avoid
cleavage of the acid labile protecting groups due to pH
decrease during the oxidation sodium hydrogen carbonate
was added.14 Finally, catalytic reduction of the nitro group
with ammonium formate in the presence of Pd/C afforded
the desired amino acid 415 in 90% yield.
S
S
a
b
Boc
N
NH2
BocHN
BocHN
H
16
17
+
Cl–
NH
NH2
c
Boc
+H3N
NH2
Cl–
N
BocHN
H
18
19
Scheme 4 Reagents and conditions: (a) (i) NaH (2.1 equiv),
(Boc)2O, THF, –78 °C to –40 °C (88%); (b) NH3, EtOH, 0 °C (99%);
(c) HCl, Et2O, 0 °C to r.t. (quant.)
3-Aminopropionamidine, the fragment common to all
members of the sperabillin family of antibiotics and also
part of a number of other antibiotics (e.g. distamycin A or
amidinomycin), was prepared in three steps, starting from
readily available N-Boc-b-alanine thioamide (16,19
Scheme 4).20 First, the thioamide 16 was acylated with di-
tert-butyl dicarbonate in the presence of 2.1 equivalents of
sodium hydride (THF, –78 °C to –40 °C)21 to afford N-
Boc thioamide 17 in 88% yield. Secondly, 17 was readily
converted into the corresponding N-Boc amidine 18 by
(2E,4Z)-Hexa-2,4-dienoic acid (14), representing the Ne-
acyl residue of sperabillin A (3a) and B (3b), was pre-
Synlett 2005, No. 17, 2615–2618 © Thieme Stuttgart · New York

LETTER
Sperabillins
2617
treatment with NH3 in ethanol at 0 °C (99% yield). Re- In summary, the first total synthesis of the pseudopeptide
moval of the tert-butoxycarbonyl protecting groups with antibiotic sperabillin A (3a) as well as a markedly im-
2.0 M ethereal HCl then led to 3-aminopropionamidine proved total synthesis of sperabillin C (3c) have been ac-
dihydrochloride (19) in quantitative yield.
complished [11 steps from N-Boc-O-methyl-L-tyrosine
(7), 16% and 20% overall yield, respectively]. Our
concise and inherently flexible approach seems to be quite
attractive for the generation of libraries of sperabillin A
and C analogues for biological evaluation.
With the three building blocks 4, 15 and 19 in hand
sperabillin A (3a) was readily assembled in four steps
(Scheme 5). First, the e-amino group was acylated with
pentafluorophenyl ester 15 (Et3N, DMF, r.t.) to give
amide 20 in 71% yield. Subsequently, 20 was activated as
a pentafluorophenyl ester (C6F5OH, EDC, CH2Cl2, r.t.)
and then coupled with 3-aminopropionamidine dihydro-
chloride (19) (Et3N, DMF, r.t.) to afford 22 in 82% yield
(two steps). Finally, the acid-labile protecting groups
were removed by treatment with TFA (5% H2O, r.t.) to af-
ford, after anion exchange on Amberlite® IRA 402 Cl res-
in, sperabillin A dihydrochloride (3a)22 in 92% yield. The
spectroscopic data for 3a were in full agreement with
those reported for the natural product.2c In the end, spera-
billin C (3c)23 was prepared analogously to 3a in 20%
overall yield from 4, (2E,4E)-hexa-2,4-dienoic acid
pentafluorophenyl ester and 19 (Scheme 6).
Acknowledgment
We are greatly indebted to Prof. Dr. Klaus T. Wanner for his ge-
nerous support. We would also like to thank Michael Felkel and
Alexandra Schüler for technical assistance.
References
(1) Hamada, M.; Takeuchi, T.; Kondo, S.; Ikeda, Y.; Naganawa,
H.; Maeda, K.; Okami, Y.; Umezawa, H. J. Antibiot. 1970,
23, 170.
(2) (a) Harada, S.; Ono, H. Eur. Pat. Appl.; EP206068, 1986.
(b) Katayama, N.; Nozaki, Y.; Tsubotani, S.; Kondo, M.;
Harada, S.; Ono, H. J. Antibiot. 1992, 45, 10. (c) Hida, T.;
Tsubotani, S.; Funabashi, Y.; Ono, H.; Harada, S. Bull.
Chem. Soc. Jpn. 1993, 66, 863.
(3) For some recent examples see: (a) Raju, B.; Mortell, K.;
Anandan, S.; O’Dowd, H.; Gao, H.; Gomez, M.; Hackbarth,
C.; Wu, C.; Wang, W.; Yuan, Z.; White, R.; Trias, J.; Patel,
D. V. Bioorg. Med. Chem. Lett. 2003, 13, 2413. (b) Jain, R.
P.; Williams, R. M. J. Org. Chem. 2002, 67, 6361.
(c) Davies, S. G.; Ichihara, O. Tetrahedron: Asymmetry
1996, 7, 1919.
(4) Hashiguchi, S.; Kawada, A.; Natsugari, H. Synthesis 1992,
403.
(5) Highly efficient asymmetric total syntheses of sperabillin B
(3b) and D (3d) have been reported recently, see:
(a) Davies, S. G.; Kelly, R. J.; Price Mortimer, A. J. Chem.
Commun. 2003, 2132. (b) Davies, S. G.; Haggitt, J. R.;
Ichihara, O.; Kelly, R. J.; Leech, M. A.; Price Mortimer, A.
J.; Roberts, P. M.; Smith, A. D. Org. Biomol. Chem. 2004, 2,
2630. (c) For a previous chiral-pool synthesis of sperabillin
D (3d) see: Hashiguchi, S.; Kawada, A.; Natsugari, H. J.
Chem. Soc., Perkin Trans. 1 1991, 2435.
Scheme 5 Reagents and conditions: (a) 15, Et3N, DMF, r.t. (71%);
(b) C6F5OH, EDC, CH2Cl2, r.t.; (c) 19, Et3N, DMF, r.t. (82%, two
steps); (d) TFA, H2O (5%), r.t. (92%)
(6) Chen, H. G.; Tustin, J. M.; Wuts, P. G. M.; Sawyer, T. K.;
Smith, C. W. Int. J. Pept. Protein Res. 1995, 45, 1.
(7) Jackson, R. F. W.; Dexter, C. S. J. Org. Chem. 1999, 64,
7579.
(8) (a) Ohtake, N.; Okamoto, O.; Mitomo, R.; Kato, Y.;
Yamamoto, K.; Haga, Y.; Fukatsu, H.; Nakagawa, S. J.
Antibiot. 1997, 50, 598. (b) Rudolph, J.; Hanning, F.; Theis,
H.; Wischnat, R. Org. Lett. 2001, 3, 3153. (c) Paintner, F.
F.; Allmendinger, L.; Bauschke, G.; Klemann, P. Org. Lett.
2005, 7, 1423.
(9) Evans, D. A.; Seidel, D.; Rueping, M.; Lam, H. W.; Shaw, J.
T.; Downey, C. W. J. Am. Chem. Soc. 2003, 125, 12692.
(10) The reaction of chiral aldehyde 6 with nitromethane in the
presence of catalyst (+)-11 represents the matched case of
double diastereoselection, since application of the
enantiomeric catalyst (–)-11 gave the corresponding epimer
of 5 with slightly lower stereoselectivity (91% de).
(11) For recent examples see: (a) Emmer, G.; Grassberger, M.
A.; Meingassner, J. G.; Schulz, G.; Schaude, M. J. Med.
Chem. 1994, 37, 1908. (b) Matsuura, F.; Hamada, Y.;
Shioiri, T. Tetrahedron 1994, 50, 9457. (c) Georgiadis, D.;
Matziari, M.; Vassiliou, S.; Dive, V.; Yiotakis, A.
Scheme 6 Reagents and conditions: (a) (2E,4E)-hexa-2,4-dienoic
acid pentafluorophenyl ester, Et3N, DMF, r.t. (83%); (b) C6F5OH,
EDC, CH2Cl2, r.t.; (c) 19, Et3N, DMF, r.t. (87%, two steps); (d) TFA,
H2O (5%), r.t. (94%)
Synlett 2005, No. 17, 2615–2618 © Thieme Stuttgart · New York

2618
L. Allmendinger et al.
LETTER
Tetrahedron 1999, 55, 14635. (d) Walker, J. R.; Curley, R.
W. Jr. Tetrahedron 2001, 57, 6695. (e) Moutevelis-
Minakakis, P.; Sinanoglou, C.; Loukas, V.; Kokotos, G.
Synthesis 2005, 933. (f) For a recent review covering the
oxidative degradation of benzene rings see: Mander, L. N.;
Williams, C. M. Tetrahedron 2003, 59, 1105.
(17) For an alternative stereoselective approach to (2E,4Z)-hexa-
2,4-dienoic acid (14), see ref. 5b.
(18) Smith, A. B.; Pitram, S. M.; Boldi, A. M.; Gaunt, M. J.;
Sfouggatakis, C.; Moser, W. H. J. Am. Chem. Soc. 2003,
125, 14435.
(19) (a) Barker, P. L.; Gendler, P. L.; Rapoport, H. J. Org. Chem.
1981, 46, 2455. (b) Josse, O.; Labar, D.; Marchand-
Brynaert, J. Synthesis 1999, 404.
(20) For a previous synthesis of 3-aminopropionamidine, starting
from 3-aminopropionitril, see: Hilgetag, G.; Paul, H.;
Günther, J.; Witt, M. Chem. Ber. 1964, 97, 704.
(12) (a) Yoshifuji, S.; Tanaka, K.; Nitta, Y. Chem. Pharm. Bull.
1985, 33, 1749. (b) Tanaka, K.; Yoshifuji, S.; Nitta, Y.
Chem. Pharm. Bull. 1988, 36, 3125.
(13) (a) In addition to product 10, its N-formyl derivative 3-tert-
butoxycarbonylformylamino-5-tert-butyldimethylsilyloxy-
6-nitrohexanoic acid (5% yield) and trace amounts (<1%) of
4-tert-butoxycarbonylamino-2-tert-butyldimethyl-
silyloxyhexanedioic acid, the latter indicating a Nef-type
reaction at the nitromethyl group, were obtained.
(21) Lee, H. K.; Ten, L. N.; Pak, C. S. Bull. Korean Chem. Soc.
1998, 19, 1148.
(22) Spectroscopic data of sperabillin A (3a)2c: [a]D23 –11.4 (c 0.4
in H2O), lit.2c [a]D23 –11 (c 1.1 in H2O). 1H NMR (500 MHz,
D2O): d = 1.77 (1 H, ddd, J = 4.5, 10.0, 15.0 Hz), 1.87 (3 H,
d, J = 7.2 Hz), 1.87 (1 H, m), 2.68 (2 H, t, J = 6.6 Hz), 2.74
(2 H, m), 3.33 (1 H, dd, J = 6.6, 14.0 Hz), 3.39 (1 H, dd,
J = 4.5, 14.0 Hz), 3.55 (1 H, dt, J = 6.6, 14.0 Hz), 3.59 (1 H,
dt, J = 6.7, 14.0 Hz), 3.86 (1 H, m), 4.00 (1 H, m), 6.02 (1 H,
dq, J = 7.2, 10.8 Hz), 6.07 (1 H, d, J = 15.1 Hz), 6.23 (1 H,
t, J = 10.8 Hz), 7.56 (1 H, dd, J = 11.8, 15.1 Hz). 13C NMR
(100 MHz, D2O): d = 16.0, 35.3, 38.0, 39.2, 39.8, 47.8, 49.1,
69.1, 125.1, 129.6, 139.4, 139.5, 171.7, 172.4, 174.7.
(23) Spectroscopic data of sperabillin C (3c)2c: [a]D23 –10.2 (c 0.4
in H2O), lit.2c [a]D20 –11 (c 0.7 in H2O). 1H NMR (500 MHz,
D2O): d = 1.75 (1 H, ddd, J = 4.7, 10.0, 15.0 Hz), 1.83 (3 H,
d, J = 5.5 Hz), 1.87 (1 H, ddd, J = 3.1, 7.5, 15.0 Hz), 2.67 (2
H, t, J = 6.7 Hz), 2.73 (2 H, d, J = 7.0 Hz), 3.30 (1 H, dd,
J = 6.5, 14.0 Hz), 3.37 (1 H, dd, J = 4.7, 14.0 Hz), 3.57 (2 H,
m), 3.85 (1 H, m), 3.98 (1 H, m), 5.97 (1 H, d, J = 15.5 Hz),
6.26 (2 H, m), 7.13 (1 H, dd, J = 9.7, 15.5 Hz). 13C NMR
(100 MHz, D2O): d = 20.7, 35.3, 38.0, 39.2, 39.8, 47.8, 49.1,
69.1, 122.9, 132.0, 143.2, 145.4, 171.6, 172.5, 174.7.
Application of the reaction conditions [2.2 mol% RuCl,
NaIO4 (18 equiv), CCl4–MeCN–H2O = 2:2:3, r.t.]
previously developed by Sharpless and co-workers (see ref.
13b) gave the oxidation product 10 in a 41% yield together
with 24% of the N-formyl derivative mentioned above.
(b) Carlsen, P. H. J.; Katsuki, T.; Martin, V. S.; Sharpless, K.
B. J. Org. Chem. 1981, 46, 3936.
(14) In the absence of sodium hydrogen carbonate a somewhat
lower yield (56%) of product 10 was obtained.
(15) Spectroscopic data of compound 4: [a]D20 +20.8 (c 0.95 in
CH2Cl2). 1H NMR (500 MHz, MeOH-d4): d = 0.13 (3 H, s),
0.15 (3 H, s), 0.92 (9 H, s), 1.43 (9 H, sbr), 1.78 (2 H, m), 2.31
(1 H, dd, J = 14.9, 7.7 Hz), 2.39 (1 H, dd, J = 14.9, 4.7 Hz),
3.01 (1 H, dd, J = 13.0, 4.1 Hz), 3.06 (1 H, dd, J = 13.0, 4.9
Hz), 3.85 (1 H, m), 4.18 (1 H, m), 4.09 (1 H, m). 13C NMR
(100 MHz, MeOH-d4): d = –5.0, 18.4, 25.9, 28.3, 40.0, 43.5,
45.2, 45.8, 67.7, 79.6, 157.1, 178.6.
(16) Rossi, R.; Carpita, A.; Quirici, M. G.; Gaudenti, M. L.
Tetrahedron 1982, 38, 631.
Synlett 2005, No. 17, 2615–2618 © Thieme Stuttgart · New York