Progress toward the synthesis of piperazimycin A: exploration of the synthesis of γ-hydroxy and γ-chloropiperazic acids

Article abstract of DOI:10.1016/j.tetlet.2008.04.132

Employing Hamada's chemistry with MAOS optimization of several steps, an expedient route to key (3S,5S)- and (3R,5R)-γ-hydroxy and (3R,5S)-γ-chloropiperazic acids, was developed en route to a total synthesis of piperazimycin A.

Full text of DOI:10.1016/j.tetlet.2008.04.132

Tetrahedron Letters 49 (2008) 4116–4118
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: http://www.elsevier.com/locate/tetlet
Progress toward the synthesis of piperazimycin A: exploration of the
synthesis of c-hydroxy and c-chloropiperazic acids
*
J. Phillip Kennedy, John T. Brogan, Craig W. Lindsley
Department of Chemistry and Pharmacology, Vanderbilt University, Vanderbilt Medical Center, Vanderbilt Program in Drug Discovery,
Vanderbilt Institute of Chemical Biology, Nashville, TN 37232, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Employing Hamada’s chemistry with MAOS optimization of several steps, an expedient route to key
(3S,5S)- and (3R,5R)-c-hydroxy and (3R,5S)-c-chloropiperazic acids, was developed en route to a total syn-
thesis of piperazimycin A.
Received 4 March 2008
Revised 22 April 2008
Accepted 23 April 2008
Available online 27 April 2008
Ó 2008 Elsevier Ltd. All rights reserved.
Fenical and co-workers recently reported on the isolation of
piperazimycin A 1 from the fermentation broth of a Streptomyces
sp., cultivated from marine sediments near the island of Guam.1
A cyclic hexadepsipeptide, 1, is composed of rare amino acids such
as hydroxyacetic acid (HAA), a-methylserine, a novel (S)-2-amino-
8-methyl-4,6-nonadecadienoic acid (AMNA), two c-hydroxypiper-
azic acids (S,S)-cOHPip1 and (R,R)-cOHPip2 and one c-chloropiper-
azic acid (R,S)-cClPip (Fig. 1).1 Piperazimycin A (1) proved to be a
potent cancer cell cytotoxin which exhibited in vitro cytotoxicity
toward multiple tumor cell lines with a mean GI50 of 100 nM.1
Based on its novel molecular architecture, the diversity of non-pro-
teogenic amino acid building blocks, and its potent cytotoxicity, we
embarked on a total synthesis campaign aimed at delivering piper-
azimycin A (1) in sufficient quantities to elucidate the molecular
target(s) responsible for the biological activity. In this Letter, we
describe our progress towards and optimization of the key c-func-
tionalized piperazic acids.
(Scheme 1).2 Exposure of 6 to hydrazine hydrate in EtOH for 16 h
afforded the cyclic hydrazone 7, which was then mono-N-benzyl-
ated to provide 8 in 89% yield over 2 steps. A Lewis acid-promoted
Strecker reaction with stoichiometric Zn(OTf)2 results in an 81:19
(9:10) ratio of diastereomers in favor of the cis adduct 9, which
In order to complete a total synthesis of 1, we first had to syn-
thesize the requisite c-functionalized piperazic acids 2–4 (Fig. 2).
Upon examination of the literature, we were surprised that there
are very few syntheic routes to these unnatural amino acids, espe-
cially with mono-N-protection.2–8 A single report by Hamada and
co-workers demonstrated that the mono-N-Boc analogs 2 and 3
could be accessed, which we desired for the sequence of peptide
couplings en route to 1.2 In accord with a report from Hale et
al.,7 we anticipated that deprotection of the TBS group of 3 and
exposure to PPh3 and MeCN–CCl4 would install the chlorine with
inversion providing the corresponding c-ClPip 4; however, this
step had never been performed on a mono-N-protected piperazic
acid.3,7
Figure 1. Structure of piperazimycin A (1).
Hamada’s protocol begins with (R)-4-chloro-3-hydroxybutanoic
acid ethyl ester 5 which is silylated and reduced with DIBAL-H in
hexanes to deliver aldehyde 6 in 75% yield for the two steps
* Corresponding author. Tel.: +1 615 322 8700; fax: +1 615 343 6532.
E-mail address: craig.lindsley@vanderbilt.edu (C. W. Lindsley).
Figure 2. c-Functionalized piperazic acids target molecules 2–4.
0040-4039/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.04.132

J. Phillip Kennedy et al. / Tetrahedron Letters 49 (2008) 4116–4118
4117
N
N
HN
HN
Cl
O
Cl
O
Cl
O
Cl
O
b
b
a
a
H
H
OEt
OEt
OTBS
OTBS
OTBS
OTBS
OH
OH
5
6
7
5
6
7
H
H
H
H
N
N
CN
N
CN
N
N
CN
N
CN
c
c
d
d
BzN
BzN
BzN
BzN
BzN
BzN
OTBS
OTBS
OTBS
OTBS
OTBS
OTBS
8
9
10
8
9
10
H
N
H
N
H
H
N
H
N
H
N
CO2Me
CO2H
N
CO2Me
CO2Me
CO2H
CO2Me
HN
HN
BocN
HN
HN
e
f
BocN
e
f
OH
12
OH
11
OH
OH
12
OH
11
OH
13
13
Scheme 1. Reagents and conditions: (a) (i) TBSCl, imid., DMF, 0 °C, 98%, (ii) DIBAL-
H, hexanes, À78 °C, 77%; (b) NH2NH2, EtOH, 0 °C to rt, 16 h; (c) BzCl, pyr, 89%, 2
(b–c) steps; (c) TMSCN, Zn(OTf)2 (1.0 equiv), HOAc (1.0 equiv), 81:19 (9:10), 70%;
(d) 6 N HCl, reflux, 12 h (under argon); (e) cat. TsOH, MeOH, reflux, 19 h; (f) BOC2O,
1:1 diox–H2O, 46 h, 59% over 3 (d–e) steps.
Scheme 2. Reagents and conditions: (a) (i) TBSCl, imid., DMF, 0 °C, 98%, (ii) DIBAL-
H, toluene, À78 °C, 91%; (b) NH2NH2, EtOH, mw, 120 °C, 20 min; (c) BzCl, pyr, MW,
120 °C, 20 min 90%, 2 (b–c) steps; (c) TMSCN, Zn(OTf)2 (1.0 equiv), HOAc (1.0 equiv),
81:19 (9:10), 57%:13%; (d) 6 N HCl, mw, 120 °C, 10 min; (e) cat. TsOH, MeOH, MW,
120 °C, 30 min; 75% over 2 steps (d–e); (f) BOC2O, 1:1 diox–H2O, 46 h, 5–25%.
the t-butyl carbamate (BOC)2O and various conditions, BOC-ON,
etc. . ., but all attempts provided very little conversion to 13. Dani-
shevsky and co-workers previously demonstrated that the analo-
gous Teoc derivative of 13 could be prepared on a large scale,
which also worked well in our hands, but we required an alterna-
tive protecting group orthogonal to fluoride deprotection.4–6 Other
simple aliphatic carbamates (methyl, ethyl) worked equally well,
but deprotection conditions proved too harsh.
After many unsuccessful attempts, we developed a high, yield-
ing, and scalable route to an Alloc-protected congener of 13 as our
modified MAOS-route to 12 worked well on multi-gram scale, and
at this point, we did not further explore selective N1 protection. As
shown in Scheme 3, exposure of 12 to Alloc-Cl in pyridine under
microwave irradiation for 10 min at 120 °C provides the previously
undescribed N-Alloc congener (R,R)-14 in 90% yield. Another MAOS
mediated silyation provided 15 in 81% yield, followed by a sapon-
ification step to deliver the key N-Alloc-(R,R)-c-OTBSPip2 16 in
quantitative yield. Following Scheme 2, but substituting the (R)-
4-chloro-3-hydroxybutanoic acid ethyl ester 5 with the corre-
sponding (S)-enantiomer 17 provides 18 in equivalent yield and
diastereomeric ratio as 9 (Scheme 4). Deprotection and hydrolysis
can be easily separated by column chromatography. Deprotection
and hydrolysis with 6 N HCl at reflux for 12 h under argon led to
11 which is then esterified to 12 with catalytic TsOH in refluxing
methanol for 19 h. Finally, 12 is selectively mono-Boc protected
at the N1 position by treatment with BOC2O in 1:1 dioxane–water
for 46 h to deliver 13, two steps away from (R,R)-cOHPip2 (4).2
Moreover, the anlaogous precursor to (S,S)-cOHPip1 (3) could also
be accessed via this route by starting the sequence with (S)-4-
chloro-3-hydroxybutanoic acid ethyl ester in place of 5. While this
was an attractive route, we felt there was room to optimize as
many steps required long reaction times (12–46 h).2
Our modified Hamada protocol employed microwave-assisted
organic synthesis (MAOS) to dramatically reduce reaction times
for four key steps.9,10 In our hands, (R)-4-chloro-3-hydroxybuta-
noic acid ethyl ester 5 is silylated and reduced with DIBAL-H in tol-
uene, in place of hexanes, to deliver aldehyde 6 in an improved
yield of 91% for the two steps (Scheme 2). An MAOS condensation
of 6 with hydrazine provides the cyclic hydrazone 7 in only 20 min
(vs 16 h), which is then mono-N-benzylated under microwave irra-
diation (120 °C, 20 min) to provide 8 in 90% yield over 2 steps. The
key Strecker reaction was performed as prescribed and delivered
an 81:19 ratio of 9:10, which were easily separated by flash col-
umn chromatography to afford a 57% yield of the cis diastereomer
9. Deprotection and saponification with 6 N HCl in the microwave
(120 °C, 10 min) delivers 11 which is then subjected to another
MAOS reaction (120 °C, 30 min) with cat. TsOH in MeOH to provide
ester 12 in 75% yield for the two steps. At this point, the modified
reaction protocol reduced total reaction time for the synthesis
dramatically and provided improvements in the overall yield on
a
mutli-gram scale. Specifically, the reaction times for the
synthesis of 7, 8, 11, and 12 were reduced from 49 h to only
80 min (an ꢀ40-fold reduction) by virtue of microwave irradiation.
However, we were unable to access the mono-N-Boc derivative
13 in reasonable yield. While we could prepare 13 on an ꢀ25 mg
scale in yields ranging from 5% to 25%, the reaction failed on a lar-
ger, synthetically useful scale en route to a total synthesis of piper-
azimycin A (1). Revisiting Hamada’s paper indicated that they only
prepared 13 on a <30 mg scale.2 We explored alternative protocols
(MAOS and conventional) and alternative reagents for formation of
Scheme 3. Reagents and conditions: (a) Alloc-Cl, TEA, pyr, MW, 120 °C, 10 min 90%;
(b) TBS–OTf, 2,4,6-collidine, MW, 120 °C, 10 min, 81%; (c) LiOH, aq THF, 99%.

4118
J. Phillip Kennedy et al. / Tetrahedron Letters 49 (2008) 4116–4118
in hand, mild saponification conditions delivered previously
unknown N-Alloc-(R,S)-c-ClPip 25 in 89% yield (unoptimized).
In summary, we have modified Hamada’s original approach and
developed an accelerated, high yielding protocol, taking advantage
of the power of MAOS, for the synthesis of c-functionalized piper-
azic acids. We have also demonstrated that selective mono-Boc
protection at the N1 position of piperazic acids is a poor reaction
that proceeds only on small scale. Importantly, we have developed
a scalable MAOS protocol for selective Alloc protection at the N1
position of functionalized c-piperazic acids. Following this
new synthetic route, we have prepared previously unknown
N-Alloc-(S,S)-c-OTBSPip1 22, N-Alloc-(R,R)-c-OTBSPip2 16, and
N-Alloc-(R,S)-c-ClPip 25 and are now poised to initiate the first to-
tal synthesis of piperazimycin (A) 1. Further refinements and the
total synthesis of 1 will be reported in due course.
Acknowledgments
The authors thank the Department of Pharmacology at Vander-
bilt and the Vanderbilt Institute of Chemical Biology (VICB) for
support of this research. J.P.K. thanks the VICB for a generous pre-
doctoral fellowship.
Scheme 4. Reagents and conditions: (a) 6 N HCl, mw, 120 °C, 10 min; (b) cat. TsOH,
MeOH, MW, 120 °C, 30 min; (c) Alloc-Cl, TEA, pyr, MW, 120 °C, 10 min 74% over
three steps (a–c); (d) (i) TBS-OTf, 2,4,6-collidine, MW, 120 °C, 10 min; (ii) LiOH, aq
THF, 82% over 2 steps.
References and notes
1. Miller, E. D.; Kauffma, C. A.; Jensen, P. R.; Fenical, W. J. Org. Chem. 2007, 72, 323–
330.
2. Makino, K.; Jiang, H.; Suzuki, T.; Hamada, Y. Tetrahedron: Asymmetry 2006, 17,
1644–1649.
3. Ushiyama, R.; Yonezawa, Y.; Shin, C.-g. Chem. Lett. 2001, 11, 1172–1173.
4. Kamenecka, T. M.; Danishefsky, S. J. Chem. A. Eur J. 2001, 7, 41–63.
5. Depew, K. M.; Kamenecka, T. M.; Danishefsky, S. J. Tetrahedron Lett. 2000, 41,
289–292.
6. Kamenecka, T. M.; Danishefsky, S. J. Angew. Chem., Int. Ed. 1998, 37, 2995–2998.
7. Hale, K. J.; Jogiya, N.; Manaviazar, S. Tetrahedron Lett. 1998, 39, 7163–7166.
8. Hassall, C. H.; Ramachandran, K. L. Heterocycles 1977, 7, 119–122.
9. Shipe, W. D.; Yang, F.; Zhao, Z.; Wolkenberg, S. E.; Nolt, M. B.; Lindsley, C. W.
Heterocycles 2006, 70, 665–689.
10. Daniels, R. N.; Kim, K.; Lewis, J. A.; Lebois, E. P.; Muchalski, H.; Lindsley, C. W.
Tetrahedron Lett. 2008, 49, 305–310.
11. Representative experimental for the synthesis of (3R,5S)-1-(allooxycarbonyl)-
5-chloropiperazine-3-carboxylic acid (N-Alloc-(R,S)-c-ClPip) 26: Into a 5 mL
microwave reaction vessel was placed 12 (116 mg, 0.72 mmol), pyridine (4 mL)
and cooled to 0 °C. Allyl chloroformate (154 lL, 1.45 mmol) was slowly added
via syringe. The microwave vial was then sealed and irradiated for 10 min at
120 °C. After cooling, the reaction mixture was transferred to a separatory
funnel, extracted into EtOAc and washed with 100 mL satd NaHCO3. The
organic layer was washed with brine, dried over MgSO4 and concentrated. The
crude product was purified via silica gel chromatography with EtOAc–hexanes
(1:3 to 1:2 gradient) to afford pure 14 (130 mg, 74%) as a light brown oil. 1H
NMR (CDCl3, 400 MHz) d 5.94 (m, 1H), 5.33 (d, J = 17.2 Hz, 1H), 5.25 (d,
J = 10.4 Hz, 1H), 4.65 (d, J = 6.0 Hz, 2H), 4.07 (m, 1H), 3.87 (m, 1H), 3.78 (s, 3H),
3.15 (m, 1H), 2.36 (m, 1H), 1.76 (m, 1H); 13C NMR (CDCl3, 125 MHz) d 171.4,
155.4, 132.7, 118.4, 77.5, 67.1, 64.7, 57.1, 52.8, 51.4; LCMS, single peak,
1.41 min, m/e, 245.1 (M+1). In a 25 mL a round-bottomed flask was placed 14
(50 mg, 0.21 mmol) and it was dissolved in 5 mL of 1:1 MeCN:CCl4 and cooled
to 0 °C. Triphenylphosphine (81 mg, 0.31 mmol) was then added in one
portion. The reaction was allowed to slowly reach room temperature and
stirred overnight. Once complete by TLC, the reaction was concentrated and
purified via silica gel chromatography with EtOAc–hexanes (1:1) to afford pure
24 (54 mg, 40%) as a colorless oil. 1H NMR (CDCl3, 400 MHz) d 5.91 (m, 1H),
5.31 (d, J = 17.2 Hz, 1H), 5.21 (d, J = 10.4 Hz, 1H), 4.65 (d, J = 7.6 Hz, 2H), 4.33
(m, 1H), 4.02 (m, 2H), 3.73 (s, 3H), 3.69 (m, 1H), 2.21 (m, 2H), 1.76; 13C NMR
(CDCl3, 125 MHz) d 171, 155.7, 132.6, 118.2, 71.8, 67, 53.8, 52.6, 50.9, 35.8;
LCMS, single peak, 2.43 min, m/e, 263.1 (M+1). In 25 mL a round-bottomed
flask was placed 24 (65 mg, 0.25 mmol) and dissolved in 5 mL of THF and
cooled to 0 °C. One-hundred forty microliters of 2 M LiOH was slowly added via
pipette for 20 min. The reaction was then acidified to pH 3 with 2 M HCl, and
extracted two times with EtOAc. The organic layers were combined and
washed with brine, dried over MgSO4 and concentrated to provide a light
yellow oil (26) of analytical purity (55 mg, 90%). 1H NMR (CDCl3, 400 MHz) d
5.92 (m, 1H), 5.32 (d, J = 17.2 Hz, 1H), 5.25 (d,J = 10.4 Hz, 1H), 4.65 (m, 2H), 4.2
(m, 1H), 4.02 (m, 1H), 3.9 (m, 1H), 3.75 (m, 1H), 2.4 (m, 1H), 2.2 (m, 1H); 13C
NMR (CDCl3, 125 MHz) d 173, 155.5, 132.6, 118.9, 77.5, 72.1, 67.6, 52.1, 51.2,
35.3; LCMS, single peak, 2.11 min, m/e, 249.1 (M+1).
Scheme 5. Reagents and conditions: (a) PPh3, MeCN–CCl4 (1:1), rt, 4 h, 40% (23),
20% (24); (b) LiOH, aq THF, 0 °C, 89%.
with 6 N HCl in the microwave (120 °C, 10 min) delivers 19 which
is then subjected to another MAOS reaction (120 °C, 30 min) with
cat. TsOH in MeOH to provide ester 20. Finally, 20 is selectively
mono-Alloc protected at the N1 position by treatment with Al-
loc-Cl in pyridine at 0 degrees, then heated under microwave irra-
diation for 10 min at 120 °C providing the previously undescribed
N-Alloc congener (S,S)-21 in 75% yield for the 3 steps. Another
MAOS mediated silyation provided 22 in 83% yield, followed by a
saponification step to deliver the key N-Alloc-(S,S)-c-OTBSPip1 22
in quantitative yield.
With two of the three requisite piperazic acids in hand, effort
now focused on preparing the remaining N-Alloc analog of target
molecule (R,S)-c-ClPip 4.11 As shown in Scheme 5, the application
of the Hale protocol7 (PPh3, MeCN–CCl4) employing (R,R)-14 pro-
vided (R,S)-23 in 40% isolated yield along with a 20% yield of the
elimination product 24. We were pleased to see that the first appli-
cation of the Hale protocol7 with a mono-N-protected substrate
provided an equivalent yield to the di-N-protected piperazic acids
without greater propensity for elimination to form 24. With 23