A new convergent approach to biphenomycin antibiotics

Article abstract of DOI:10.1055/s-2003-37506

A new, convergent approach to the biaryl key intermediate of Schmidt's biphenomycin B total synthesis has been accomplished via a palladacycle complex catalyzed Stille cross-coupling of two o-tyrosine building blocks.

Full text of DOI:10.1055/s-2003-37506

522
LETTER
A New Convergent Approach to Biphenomycin Antibiotics
N
ew
,
Converg
r
e
nt App
a
n
zA
ntibiotics F. Paintner,*a Klaus Görler,a Wolfgang Voelterb
a
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
b
Abteilung für Physikalische Biochemie, Physiologisch-chemisches Institut der Universität Tübingen, Hoppe-Seyler-Straße 4,
72076 Tübingen, Germany
Received 10 January 2003
Schmidt and co-workers. However, neither biphenomycin
derivatives nor close analogues thereof have been pre-
pared and biologically evaluated so far.
Abstract: A new, convergent approach to the biaryl key intermedi-
ate of Schmidt’s biphenomycin B total synthesis has been accom-
plished via a palladacycle complex catalyzed Stille cross-coupling
of two o-tyrosine building blocks.
Key words: biphenomycins, antibiotics, cyclopeptides, asymmet-
ric synthesis, Stille coupling
The biphenomycins are a small family of cyclopeptide an-
tibiotics isolated from the culture broths of Streptomyces
filipinensis1 and S. griseorubiginosus2. Their key struc-
tural feature is a 15-membered macrocycle containing a
biaryl moiety (Figure 1). Among biphenomycins, biphe-
nomycin A (1a) is the most abundant. It has been shown
to exhibit high antibacterial activity both in vitro and in vi-
vo, especially against Gram-positive bacteria. These in-
clude serious pathogens such as Staphylococcus aureus,
Enterococcus faecalis or Streptococcus pneumoniae. Bi-
phenomycin B (1b), the simplest representative of this
family of cyclopeptide antibiotics, shows an in vitro anti-
microbial activity closely related to that of 1a. However,
its potential in vivo activities are still undetected, presum-
ably due to the paucity of material obtained from the nat-
ural source, which has precluded its further biological
evaluation. In view of their promising antibacterial pro-
files coupled with their intriguing structural features, bi-
phenomycins had attracted extensive synthetic studies3
culminating in the stereoselective total syntheses of bi-
phenomycin A (1a)4 and biphenomycin B (1b)5 by
Figure 1
Therefore, we recently have been engaged with the devel-
opment of a new, convergent approach toward biphenom-
ycin B (1b)6 and related cyclopeptides 2 in general, that
would be readily adaptable to parallel synthesis of a di-
verse set of analogues with modified biaryl moiety. The
strategies of the total syntheses mentioned above are
largely consecutive, i.e. 11 linear steps for key intermedi-
ate 3 of Schmidt’s biphenomycin B total synthesis (vide
infra), and thus unfavourable to this end.
In our convergent strategy cyclopeptides 2 are separated
into a Nα-Boc amino acid 4 and key building block 3,
which is further subdivided into o-tyrosine derivatives 5
and 6 (Scheme 1). Subsequent segment coupling of these
Scheme 1 Retrosynthetic analysis of biphenomycin B analogues (2)
Synlett 2003, No. 4, Print: 12 03 2003.
Art Id.1437-2096,E;2003,0,04,0522,0526,ftx,en;D00403ST.pdf.
© Georg Thieme Verlag Stuttgart · New York
ISSN 0936-5214

LETTER
A New Convergent Approach to Biphenomycin Antibiotics
523
fragments involving a Stille cross-coupling reaction and a
final macrolactamisation as key steps was envisioned to
give the desired macrocycles. Herein we report a short and
efficient asymmetric approach to o-tyrosine derivatives 5
and 6 as well as their subsequent cross-coupling to afford
key building block 3.
Building blocks 5 and 6 are both derived from (S)-2-ben-
zyloxy-5-iodophenylalanine (11). This novel α-amino
acid was synthesized following a diastereoselective path-
way based on Oppolzer’s camphor sultam as chiral auxil-
iary (Scheme 2). Thus successive treatment of chiral
glycine equivalent 77 with n-BuLi (THF, DMPU, –78 °C)
and benzylbromide building block 88 (–78 °C to r.t.) pro-
vided the alkylation product 9 in 84% yield and with high
diastereoselectivity (ds ≥ 98:2 as determined by 500 MHz
1H NMR).
Scheme
3
(a) Z-ONSu, DMF, aq Na2CO3, r.t. (88%); (b)
TMSCH2CH2OH, EDC, DMAP, 0 °C to r.t. (74%); (c) Me3SnSnMe3
(1.4 equiv), Pd(OAc)2 (0.04 equiv), PPh3 (0.08 equiv), toluene, ∆
(80%); (d) Bu3SnSnBu3, cat. PdCl2(PMePh2)2, KOAc, NMP, r.t.
(68%); (e) 5b, 6, palladacycle-complex 23, LiCl, NMP, 90 °C (64%
from 6)
cat. DMAP, CH2Cl2, 0 °C to r.t.) gave 14 in 65% overall
yield.
Subsequently, aryl iodide 14 was converted to the corre-
sponding aryl stannanes 5a and 5b by treatment with ei-
ther hexamethylditin or hexabutylditin [R3SnSnR3, cat.
Pd(OAc)2, PPh3, toluene, 100 °C] according to Ortar et
al.12 However, whereas aryl trimethylstannane 5a could
be obtained in good yields (80%) the incorporation of the
tributylstannyl group proceeded unsatisfactory (41%
yield). Nevertheless preparation of 5b still could be im-
proved by stannylation according to a novel procedure re-
ported by Masuda et al.13 Thus treatment of 14 with
hexabutylditin in the presence of cat. PdCl2(PMePh2)2
(KOAc, NMP, r.t.) afforded 5b in 68% yield.
We next focused our attention on optimising reaction con-
ditions for the Stille cross-coupling of key building blocks
5 and 6. To this end, we studied the reaction of the closely
related model compounds 17 and 19, which are more
readily available than 5 and 6.
Scheme 2 (a) (i) n-BuLi (1.05 equiv), THF, –78 °C; (ii) 7, DMPU,
–78 °C to r.t. (84%); (b) 0.01 M HCl, THF, 0 °C to r.t. (95%); (c)
LiOH (2.0 equiv), H2O2 (4.0 equiv), THF, 0 °C to r.t. (81%); (d)
Boc2O, dioxane, aq NaHCO3, r.t. (85%); (e) BnOH, EDCI, DMAP,
CH2Cl2, 0 °C to r.t. (75%)
Thus, both compounds 17 and 19 were prepared from
readily accessible 6-iodochroman-2-one (15)14 in three
and four steps respectively, as indicated in Scheme 4.
Deprotection of 9 under mild acidic conditions (0.01 M
HCl, THF, 0 °C to r.t.) followed by lithium hydroperox-
ide-mediated cleavage of the chiral auxiliary (LiOH,
H2O2, THF, 0 °C to r.t.) gave enantiomerically pure (S)-2-
benzyloxy-5-iodophenylalanine (11)9–11 in 77% overall
yield. The chiral auxiliary could be recovered in 87%
yield. Starting from α-amino acid 11, building block 6
was readily obtained (64% overall yield) by introduction
of a Boc group (Boc2O, aq NaHCO3, dioxane, r.t.) and
subsequent esterification with benzyl alcohol (EDCl, cat.
DMAP, CH2Cl2, 0 °C to r.t.). In an analogous sequence
the N-Z protected amino acid 2-trimethylsilylethyl
(TMSE) ester 14 was prepared (Scheme 3). Thus, protec-
tion of the amino group (Z-ONSu, aq Na2CO3, DMF, r.t.)
and subsequent esterification (TMSCH2CH2OH, EDCl,
Stille cross-coupling of aryl iodide 19 with stannanes 17a
and 17b, respectively, to give biaryl 20 (Scheme 5) was
performed under several well-established reaction condi-
tions, as summarized in Table 1.
In situ generated Pd(AsPh3)4 (1 mol%) in the presence
of LiCl (3.0 equiv, NMP, 65 °C)15 or of cocatalytic CuI
(10 mol%, DMF, 65 °C)12 gave fair coupling yields (53–
60%), both with trimethylstannane 17a and tributylstan-
nane 17b (entries 1–4). Pd(dppf)Cl2 catalysed cross-cou-
pling of 17b and 19 in the presence of cocatalytic CuBr
(10 mol%), reaction conditions recently reported by Wil-
liams et al.,16 gave product 20 in a comparable yield
(49%) (entry 5). Appreciable amounts of biaryl 21 (8–
13%) and alkylarenes 22 (2–10%), which arise from com-
Synlett 2003, No. 4, 522–526 ISSN 0936-5214 © Thieme Stuttgart · New York

524
F. F. Paintner et al.
LETTER
Table 1 Reaction Conditions and Results for the Stille Coupling
Outlined in Scheme 5
Entry
R
Conditions
20
21
22
Yield
Yield
Yield
(%)a
(%)b
(%)c
1
2
3
4
5
6
7
Me
Bu
Me
Bu
Bu
Me
Bu
Pd2dba3, AsPh3,
CuI, DMF,
65 °C, 14 h
56
53
55
60
49
56
73
13
8
2
< 1
10
9
Pd2dba3, AsPh3,
CuI, DMF,
65 °C, 168 h
Pd2dba3, AsPh3,
LiCl, NMP,
65 °C, 14 h
10
8
Scheme 4 (a) NaOMe, MeOH, r.t.; (b) BnBr, K2CO3, acetone, ∆
(88% from 15); (c) Me3SnSnMe3, Pd(OAc)2, PPh3, toluene (83%);
(d) Bu3SnSnBu3, PdCl2(PMePh2)2, KOAc, NMP, r.t. (74%); (e) 16,
1M NaOH, MeOH, r.t. (83%); (f) diphenyl phosphorazidate, NEt3,
t-BuOH, toluene, ∆ (79%)
Pd2dba3, AsPh3,
LiCl, NMP,
65 °C, 168 h
Pd(dppf)Cl2,
MeCN, CuBr,
80 °C, 168 h
11
13
13
< 1
24
9
petitive stannane homocoupling and alkyl group transfer
respectively, reactions that are frequently encountered
with Stille couplings,17 were isolated as the main side
products.18
Palladacycle 23,
LiCl, NMP,
90 °C, 14 h
A markedly improved yield of cross-coupling product 20
could be accomplished, employing palladacycle catalyst
23,19 which is not yet widely applied for Stille cross-cou-
pling reactions.20
Palladacycle 23,
LiCl, NMP,
90 °C, 14 h
a Yield of isolated, purified product 20 based on 19.
b Yield of isolated, purified product 21 based on 17.
c Yield of isolated, purified product 22 based on 19.
Thus coupling of tributylstannane 17b with 19 [23 (5
mol%), LiCl (3.0 equiv), NMP, 90 C]20c proceeded
smoothly to give 20 in good yield (73%) (entry 7). Nota-
bly a distinct lower yield was observed for the coupling of
trimethylstannane 17a under analogous reaction condi-
tions (entry 6). Apparently, this is due to a differing extent
of alkyl group transfer occurring in either case (i.e. 9%
versus 24% of the respective alkylarene 22). Again, stan-
nane homocoupling product 21 was isolated as the other
main side product.
1
ly). The H NMR spectroscopic data of 3 are in full
accordance with those reported in the literature.5a Appar-
ently, neither racemisation during the preparation of
building blocks 5b and 6 nor epimerisation during their
Stille cross-coupling occurred to an appreciable extent,
since product 3 was attained as a single diastereomer [ds
> 98:2 as determined by 1H NMR (500 MHz) and HPLC
from the crude reaction product].
Finally, we applied these optimised reaction conditions to
the Stille cross-coupling of o-tyrosine building blocks 5b
and 6 (Scheme 3).21
In conclusion we disclose an efficient convergent ap-
proach to the key intermediate of Schmidt’s biphenomy-
cin B total synthesis. Efforts to adapt this strategy for the
preparation of biphenomycin B analogues with modified
biaryl moiety are currently in progress in our laboratory.
Indeed, the key intermediate 3 of Schmidt’s biphenomy-
cin B (1b) total synthesis could be obtained in good yield
(64%) in addition to the respective stannane homocou-
pling and butyl transfer products (5% and 10% respective-
Scheme 5
Synlett 2003, No. 4, 522–526 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
A New Convergent Approach to Biphenomycin Antibiotics
525
(10) The enantiomeric purity of 11 was determined via its Fmoc
derivative by HPLC (ChiraDex®Gamma (5-µm)
LiChroCART® 250 × 4 mm (Merck) MeCN–triethylamine–
AcOH, 1000:7:0.5, 1 mL/min) according to Armstrong et al.
to be > 98% ee. See: Tang, Y.; Zukowski, J.; Armstrong, D.
W. J. Chromatogr. A 1996, 743, 261.
(11) The absolute configuration of 11 was determined to be S by
conversion into (S)-o-tyrosine, whose absolute configuration
has been unambiguously established. Lit.: Dugave, C. J.
Org. Chem. 1995, 60, 601.
Acknowledgement
We are greatly indebted to Prof. Dr. Klaus Th. Wanner for his
generous support. Financial support of this work by Deutsche
Forschungsgemeinschaft is gratefully acknowledged. We would
also like to thank G. Bauschke and T. Hausmann for technical as-
sistance.
References
(12) Morera, E.; Ortar, G. Synlett 1997, 1403.
(13) Murata, M.; Watanabe, S.; Masuda, Y. Synlett 2000, 1043.
(14) Davies, S. G.; Pyatt, D. J. Orgmet. Chem. 1990, 387, 381.
(15) Farina, V.; Krishnan, B.; Marshall, D. R.; Roth, G. P. J. Org.
Chem. 1993, 58, 5434.
(16) Albrecht, B. K.; Williams, R. M. Tetrahedron Lett. 2001, 42,
2755.
(17) Farina, V.; Krishnamurthy, V.; Scott, W. J. Org. React.
1997, 50, 1.
(18) In addition to 20, 21 and 22 small amounts of the respective
aryl iodide homocoupling product i (1–7%) and in case of
entries 1–4 product ii (3–6%), which arises from aryl
transfer by the arsine, were also obtained (Figure 2).
(1) (a) Martin, J. H.; Mitscher, L. A.; Shu, P.; Porter, J. N.;
Bohonos, N.; DeVoe, S. E.; Patterson, E. L. Antimicrob.
Agents Chemother.-1967 1968, 422. (b) Chang, C. C.;
Morton, G. O.; James, J. C.; Siegel, M. M.; Kuck, N. A.;
Testa, R. T.; Borders, D. B. J. Antibiotics 1991, 44, 674.
(2) (a) Ezaki, M.; Iwami, M.; Yamashita, M.; Hashimoto, S.;
Komori, T.; Umehara, K.; Mine, Y.; Kohsaka, M.; Aoki, H.;
Imanaka, H. J. Antibiotics 1985, 38, 1453. (b) Ezaki, M.;
Shigematsu, N.; Yamashita, M.; Komori, T.; Umehara, K.;
Imanaka, H. J. Antibiotics 1993, 46, 135.
(3) (a) Brown, A. G.; Crimmin, M. J.; Edwards, P. D. J. Chem.
Soc. Perkin Trans. 1 1992, 123. (b) Carlström, A.-S.; Frejd,
T. J. Chem. Soc., Chem. Commun. 1991, 1216. (c) Brown,
A. G.; Edwards, P. D. Tetrahedron Lett. 1990, 31, 6581.
(d) Schmidt, U.; Meyer, R.; Leitenberger, V.; Lieberknecht,
A. Angew. Chem., Int. Ed. Engl. 1989, 28, 929; Angew.
Chem. 1989, 101, 946.
(4) (a) Schmidt, U.; Leitenberger, V.; Griesser, H.; Schmidt, J.;
Meyer, R. Synthesis 1992, 1248. (b) Schmidt, U.;
Leitenberger, V.; Meyer, R.; Griesser, H. J. Chem. Soc.,
Chem. Commun. 1992, 951.
(5) (a) Schmidt, U.; Meyer, R.; Leitenberger, V.; Griesser, H.;
Lieberknecht, A. Synthesis 1992, 1025. (b) Schmidt, U.;
Meyer, R.; Leitenberger, V.; Lieberknecht, A.; Griesser, H.
J. Chem. Soc., Chem. Commun. 1991, 275.
(6) For a recent convergent synthetic approach to related
cyclopeptides, see: Carbonnelle, A.-C.; Zhu, J. Org. Lett.
2000, 2, 3477.
Figure 2
(19) For the preparation of 23 (Figure 3), see: Herrmann, W. A.;
Broßmer, C.; Öfele, K.; Reisinger, C.-P.; Priermeier, T.;
Beller, M.; Fischer, H. Angew. Chem., Int. Ed. Engl. 1995,
34, 1844; Angew. Chem. 1995, 107, 1989.
(7) (a) Josien, H.; Martin, A.; Chassaing, G. Tetrahedron Lett.
1991, 32, 6547. (b) Oppolzer, W.; Moretti, R.; Thomi, S.
Tetrahedron Lett. 1989, 30, 6009.
(8) Benzylbromide 8 was prepared from 5-iodo-salicylic acid in
40% overall yield as indicated in Scheme 6.
Figure 3
(20) (a) Louie, J.; Hartwig, J. F. Angew. Chem. Int. Ed. Engl.
1996, 35, 2359; Angew. Chem. 1996, 108, 2531.
Scheme 6
(b) Herrmann, W. A.; Böhm, V. P. W.; Reisinger, C.-P. J.
Orgmet. Chem. 1999, 576, 23. (c) Brody, M. S.; Finn, M. G.
Tetrahedron Lett. 1999, 40, 415.
(9) (S)-2-Amino-3-(2-benzyloxy-5-iodophenyl)propionic
acid(11): Colourless crystals, mp 228–230 °C (decomp.).
[α]D20 = –11.8 (c 0.11, MeOH); 1H NMR (500 MHz,
CD3OD): δ = 2.89 (dd, J = 9.2/14.3 Hz, 1 H, CH2CH), 3.41
(dd, J = 4.7/14.3 Hz, 1 H, CH2CH), 3.89 (dd, J = 4.7/9.2 Hz,
1 H, CH2CH), 5.18 (s, 2 H, OCH2), 6.86 (d, J = 8.6 Hz, 1 H,
(21) Stille coupling of 5b and 6: A solution of 5b (29.5 mg, 37
mol) in degassed NMP (0.14 mL) was added to a stirred
mixture of anhydrous LiCl (4.3 mg, 102 mol), 6 (20.0 mg, 34
mol) and 23 (1.6 mg, 1.7 mol) in degassed NMP (0.2 mL) at
room temperature. The reaction mixture was heated at 90 °C
for 14 h. After cooling to room temperature, the reaction
mixture was diluted with CH2Cl2 and filtered through a pad
of silica gel. The filtrate was washed with aqueous 1.3 M
phosphate buffer pH 7.0 and water, dried (MgSO4) and
concentrated under reduced pressure. The resulting residue
was purified by prep. HPLC (LiChrosorb® Si 60 5 µm, n-
H
arom), 7.28–7.49 (m, 5 H, Harom), 7.54 (dd, J = 2.2/8.6 Hz, 1
H, Harom), 7.57 (d, J = 2.2 Hz, 1 H, Harom); 13C NMR (125
MHz, CD3OD): δ = 33.5, 56.2, 71.3, 83.9, 115.9, 128.7,
128.9, 129.2, 129.8, 138.2, 138.8, 141.0, 158.3, 173.6; MS
+
(MALDI): m/z = 398 [M + 1]. HRMS (CI, CH5 ): Anal.
Calcd for C16H16NO3I (M + H+): 398.0253; Found:
398.0253.
Synlett 2003, No. 4, 522–526 ISSN 0936-5214 © Thieme Stuttgart · New York

526
F. F. Paintner et al.
LETTER
heptane–EtOAc, 85:15) to give 20.9 mg (64%) 3 as a
colourless solid.
(m, 2 H, CH2CH2Si), 4.56–4.67 (m, 2 H, CHCH2), 4.95–5.17
(m, 8 H, OCH2Ph), 5.38 (d, 1 H, J = 7.8 Hz, NHBoc), 5.61
(d, 1 H, J = 7.7 Hz, NHZ), 6.91 (d, 1 H, J = 8.6 Hz, Harom),
6.94 (d, 1 H, J = 8.6 Hz, Harom), 7.19–7.38 (m, 20 H, Harom.),
7.42–7.48 (m, 4 H, Harom).
3: [α]D20 = +10.3 (c = 0.95, CHCl3); lit.5a: [α]D = +11.2. 1H
NMR (500 MHz, CDCl3) δ = –0.03 [s, 9 H, Si(CH3)3], 0.85–
0.91 (m, 2 H, CH2CH2Si), 1.35 [s, 9 H, C(CH3)3], 3.06–3.14
(m, 2 H, CHCH2), 3.18–3.26 (m, 2 H, CHCH2), 4.05–4.21
Synlett 2003, No. 4, 522–526 ISSN 0936-5214 © Thieme Stuttgart · New York