Highly diastereoselective desymmetrizations of cyclo(Pro,Pro): an enantioselective strategy toward phakellstatin and phakellin.

Article abstract of DOI:10.1021/ol026140q

[reaction: see text] Monoenolates of C(2)-symmetric, proline-derived piperazine-2,5-diones were generated and trapped with a variety of electrophiles to produce, in a highly diastereoselective fashion, functionalized diketopiperazines (DKPs). These reacti

Full text of DOI:10.1021/ol026140q

ORGANIC
LETTERS
2002
Vol. 4, No. 16
2645-2648
Highly Diastereoselective
Desymmetrizations of Cyclo(Pro,Pro):
An Enantioselective Strategy toward
Phakellstatin and Phakellin
Karine G. Poullennec, Anna T. Kelly,† and Daniel Romo*
Department of Chemistry, Texas A&M UniVersity, P.O. Box 30012,
College Station, Texas 77842-3012
romo@mail.chem.tamu.edu
Received May 6, 2002
ABSTRACT
Monoenolates of C -symmetric, proline-derived piperazine-2,5-diones were generated and trapped with a variety of electrophiles to produce,
2
in a highly diastereoselective fashion, functionalized diketopiperazines (DKPs). These reactions provide the basis for an asymmetric,
desymmetrization strategy toward the marine alkaloids phakellstatin and phakellin. The relative stereochemistry of the functionalized DKPs
was confirmed by single-crystal X-ray analysis and/or NOE experiments. Bis-functionalization of the DKPs was also found to proceed with
high levels of diastereoselectivity.
Marine organisms continue to be a rich source of diverse
natural products that show promise for development into
pharmaceuticals and tools for biology.1 The oroidin-derived
secondary metabolites are prime examples of the immense
and true structural diversity that can only be found in Nature.2
Although phakellin (1a) has not been isolated, the isolation
of both mono- and dibromophakellin (1b and 1c)3 in addition
to palau’amine (3) suggests that phakellin is also likely a
natural product that may be present in only minute quantities.
The phakellins (1a-c) and related phakellstatins (2a,b)4
possess a unique array of functionality including a cyclic
guanidine or urea, a pyrrole carboxylic acid, a pyrrolidine,
and potentially delicate vicinal diaminal stereocenters. The
concise and elegant biomimetic synthesis of racemic di-
bromophakellin by Bu¨chi stands as a benchmark for synthetic
efforts in this area.5 In connection with our synthetic studies
toward the immunosuppressive agent palau’amine (3),6 we
became interested in developing strategies for annulation of
the phakellin substructure onto a palau’amine spirocyclic core
intermediate. In addition, phakellin (1a) presented itself as
a challenging and attractive synthetic target due to its
compact, heteroatom-dense structure and potential antibiotic
activity.7 For these reasons, we initiated a total synthesis of
these tetracyclic natural products.8Our asymmetric synthetic
strategy is premised on the recognition that a diketopiperazine
† NSF-REU participant, Summer 2001.
(1) (a) Faulkner, D. J. Nat. Prod. Rep. 2000, 17, 1-6. (b) Faulkner, D.
J. Nat. Prod. Rep. 2000, 17, 7-55. (c) Jaspars, M. Chem. Ind. 1999, 51-
55.
(2) Potier, P.; Mourabit, A. A. Eur. J. Org. 2001, 237-243.
(3) Sharma, G. M.; Burkholder, P. R. J. Chem. Soc., Chem. Commun.
1971, 151-152.
(4) Pettit, G. R.; McNulty, J.; Herald, D. L.; Doubek, D. L.; Chapuis,
J.-C.; Schmidt, J. M.; Tackett, L. P.; Boyd, M. R. J. Nat. Prod. 1997, 60,
180-183.
(5) Foley, L. H.; Buchi, G. J. Am. Chem. Soc. 1982, 104, 1777-1779.
(6) Dilley, A. S.; Romo, D. Org. Lett. 2001, 3, 1535-1538.
(7) Kinnel, R. B.; Gehrken, H.-P.; Swali, R.; Skoropowski, G.; Scheuer,
P. J. J. Org. Chem. 1998, 63, 3281-3286. These authors propose that
phakellin may be responsible for the potent but never identified antibiotic
activity in crude extracts of Phakellia flabellata.
(8) For another strategy to these compounds, see: Jacquot, D. E. N.;
Hoffmann, H.; Polborn, K.; Lindel, T. Tetrahedron Lett. 2002, 43, 3699-
3702.
10.1021/ol026140q CCC: $22.00 © 2002 American Chemical Society
Published on Web 07/10/2002

be expected to show only low degrees of facial diastereo-
selectivity as a result of the near planarity of the tricyclic
system. Scho¨llkopf reported high diastereoselectivity for
alkylations of amino acid derived bislactim ethers.13 His
rationale for the observed selectivity was based on steric
effects due to ring substituents. It has also been suggested
that facial selectivity in related alkylations of lactam enolates
is directed either by stereoelectronic,14 torsional effects15 or
by solvation of the enolate counterion.16 Seebach also
suggested the role of adjacent nitrogen and R-carbon
pyramidalization on facial selectivity.14d
We initiated our desymmetrization studies with the DKP
derived from the less expensive (S) enantiomer of proline
rather than the (R) enantiomer as required for the natural
product. Low-temperature deprotonation of DKP 5 with 1
equiv of either LDA or KHMDS produced the monoenolate
(∼90%, as evidenced by quenching with MeOD). Trapping
of the monoenolates with a variety of electrophiles occurred
with high diastereoselectivity (dr 11-42:1, GC) in all cases
studied (Table 1). Unsurprisingly, potassium enolates were
found to be slightly more reactive than lithium enolates
(entries 2, 6, and 8 versus 3, 5, and 7). Treatment of the
potassium enolate of DKP 5 with benzylchloroformate
initially gave only a 48% yield of monobenzylester DKP
6c. This acylation was improved by slow, inverse addition
of the enolate to the electrophile, which presumably lowers
the possibility of competing deprotonation of the slightly
more acidic, monoacylated DKP product 6c. However,
neither inverse addition nor an increase in base employed
improved the yields of 3-azido DKP 6a.
Figure 1. Structures of oroidin and oroidin-derived marine
alkaloids.
is embedded in the phakellin/phakellstatin structure (Scheme
1). We, thus, sought to employ D-prolyl-D-proline anhydride
Scheme 1. Retrosynthetic Analysis of Phakellstatin and
Phakellin
The stereochemical outcome of these electrophilic addi-
tions was determined by NOE and/or X-ray analysis. A key
NOE was observed for 3-methyl DKP 6d between the C6
hydrogen and the methyl group, confirming the cis arrange-
ment of these substituents (Figure 2a). Single-crystal X-ray
analysis of the crystalline benzyl DKP 6f confirmed the
relative stereochemistry of this alkylated DKP (Figure 2b).
The high degree of diastereoselectivity in these electrophilic
additions may be attributed to the slightly puckered nature
of the tricyclic system, in which the electrophile approaches
[(R,R)-cyclo(Pro,Pro)] (5)9 as an optically active starting
material and utilize a diastereoselective desymmetrization
process to functionalize this system. Subsequent oxidation
to a pyrrole, ring annulation, and guanidinylation would
deliver phakellin (1a) via phakellstatin (2a). While there are
reports of electrophilic additions to enolates derived from
diketopiperazines,10,11 there are relatively few examples of
functionalization of cyclo(Pro,Pro).12 At first inspection,
electrophilic addition to the enolate of cyclo(Pro,Pro) might
(11) For diastereoselective alkylations of related systems see: (a) Mart´ın-
Santamar´ıa, S.; Buenadicha, F. L.; Espada, M.; So¨llhuber, M.; Avendan˜o,
C. J. Org. Chem. 1997, 62, 6424-6428. (b) Mart´ın-Santamar´ıa, S.; Espada,
M.; Avendan˜o, C. Tetrahedron 1997, 53, 16795-16802.
(12) For alkylations of cyclo(Pro, Pro), see: (a) Schmidt, U. Nikiforov,
A. Monatsch. Chem. 1975, 106, 313-320. (b) Schmidt, U.; Poisel, H. Chem
Ber. 1973, 106, 3408-3420. (c) O¨ hler, E.; Poisel, H.; Tataruch, F.; Schmidt,
U. Chem. Ber. 1972, 105, 625-634. (d) Poisel, H.; Schmidt, U. Chem.
Ber. 1972, 105, 635-641.
(13) (a) Scho¨llkopf, U.; Tiller, T.; Bardengen, J. Tetrahedron 1988, 44,
5293-5305. (b) Rjappa, S.; Natekar, M. V. AdV. Het. Chem. 1993, 57,
187-289.
(14) (a) Groaning, M. D.; Meyers, A. I. Tetrahedron 2000, 56, 9843-
9873 and references cited herein. (b) Meyers, A. I.; Seefeld, M. A.; Lefker,
B. A.; Blake, J. R. J. Am. Chem. Soc. 1997, 119, 4565-4566. (c) Meyers,
A. I.; Seefeld, M. A.; Lefker, B. A.; Blake, J. F.; Willard, P. G. J. Am.
Chem. Soc. 1998, 120, 7429-7438. For studies on pyramidalization of
trigonal centers see: (d) Seebach, D.; Maetzke, T.; Petter, W.; Klotzer, B.;
Plattner, D. A. J. Am. Chem. Soc. 1991, 113, 1781-1786. (e) Seebach, D.;
Juaristi, E.; Miller, D. D.; Schickli, C.; Weber, T. HelV. Chim. Acta 1987,
70, 237-261.
(9) Ishibashi, N.; Kouge, K.; Shinoda, I.; Kanehisa, H.; Okai, H. Agric.
Biol. Chem. 1988, 3, 819-827.
(10) For electrophilic additions to DKP enolates see (a) O¨ hler, E.;
Tataruch, F.; Schmidt, U. Chem. Ber. 1972, 105, 3658-3661. (b) Williams,
R. M.; Kwast, A. J. Org. Chem. 1988, 53, 5787-5789. (c) Williams, R.
M.; Glinka, T.; Kwast, E. J. Am. Chem. Soc. 1988, 110, 5927-5929. (d)
Kishi, Y.; Fukuyama, T.; Nakatsuka, S. J. Am. Chem. Soc. 1973, 95, 6490-
6492. (e) Kishi, Y.; Fukuyama, T.; Nakatsuka, S. J. Am. Chem. Soc. 1973,
95, 6492-6493. (f) Kishi, Y.; Fukuyama, T.; Nakatsuka, S.; Havel, M. J.
Am. Chem. Soc. 1973, 95, 6493-6495. (g) Williams, R. M.; Anderson, O.
P.; Armstrong, R. W.; Josey, J.; Meyers, H.; Eriksson, C. J. Am. Chem.
Soc. 1982, 104, 6092-6099.
(15) Ando, K.; Green, N. S.; Li, Y.; Houk, K. N. J. Am. Chem. Soc.
1999, 121, 5334-5335.
(16) Durkin, K. A.; Liotta, D. J. Am. Chem. Soc. 1990, 112, 8162-
8163.
2646
Org. Lett., Vol. 4, No. 16, 2002

Scheme 2. Functionalization of Azide 6a (Observed NOEs
Table 1. Enolization and Diastereoselective, Electrophilic
Additions to Cyclo(Pro,Pro)
Indicated by Arrows)
entry
base
E-X
compd no.
% yielda
drb
1
2
3
4
5
6
7
8
9
KHMDSc
KHMDSc
LDAe
KHMDSf
LDAd
KHMDSc
LDAd
KHMDSc
KHMDSc
Tris-N3
ClCO2Me
ClCO2Me
ClCO2Bng
MeI
6a
6b
6b
6c
6d
6d
6e
6e
6f
42 (33)
82 (6)
80
70 (9)
55
65 (19)
54
66 (13)
75 (6)
ndd
11:1
nd
12:1
nd
27:1
nd
18:1
42:1
To confirm that this was not occurring, we investigated
methods to assay the enantiopurity of the products since
equilibration would manifest itself by loss of optical purity
(Scheme 3).
MeI
allyl-Br
allyl-Br
BnBr
Scheme 3. Mechanism for Potential Erosion of Enantiopurity
of Monofunctionalized DKPs
a Yields shown are for purified products (SiO2). Recovered starting
material is indicated in parentheses. b Determined on crude reaction mixtures
using chiral GC (see Supporting Information for details). c KHMDS is added
to a precooled solution of DKP 5. d nd ) not determined. e A solution of
DKP 5 is added to freshly prepared LDA. f A solution of DKP 5 is added
to a precooled solution of KHMDS in THF. g The enolate is added to a
precooled solution of benzyl chloroformate.
the enolate from the sterically less encumbered face.
However, this explanation does not adequately rationalize
the high diastereoselectivity observed and contributions from
stereoelectronic effects cannot be excluded.14
Unfortunately, after numerous studies, we were unable to
determine the enantiomeric purity of any of the monofunc-
tionalized DKPs or derivatives by use of chiral GC, HPLC,
or NMR shift reagents. Thus, we employed the chiral
electrophile (+)-menthyl chloroformate, as this would lead
to the formation of diastereomers, if indeed epimerization
was occurring. Only one diastereomer was formed (97% de)
based on comparative GC analysis with a 1:1 diastereomeric
mixture obtained from racemic 5 (Scheme 4). This provides
Figure 2. (a) NOEs observed for 3-methyl DKP 6d. (b) Single-
crystal X-ray structure of 3-benzyl DKP 6f (POVchem rendering).
Scheme 4. Diastereoselective Acylation with Menthyl
To determine the syn stereochemistry of the monofunc-
tionalized DKP 6a, the azide moiety was reduced and
subsequent guanidinylation by the method of Kim and Qian18
delivered the di-Boc-protected guanidine 7. The latter
compound exhibited a NOE between the tert-butyl of the
Boc group when H5 was irradiated (Scheme 2).
Chloroformate Confirming High Facial Selectivity
It is known that cyclo(Pro,Pro), unlike other cyclic
dipeptides, exists almost exclusively (99.5:0.5) as the cis
isomer under equilibrating conditions.13b Thus, the high
diastereomeric ratio of the alkylated and acylated DKPs
described above could be the result of a trans to cis
interconversion following the initial electrophilic addition.
evidence that diastereoselectivity in these alkylations and
acylations is solely due to the initial, high facial selectivity
during electrophilic additions to the DKP enolates, rather
than subsequent epimerization of an initial diastereomeric
mixture.
(17) Evans, D. A.; Britton, T. C.; Ellman, J. A.; Dorow, R. L. J. Am.
Chem. Soc. 1990, 112, 4011-4030.
(18) Kim, K.; Qian, L. Tetrahedron Lett. 1993, 34, 7677-7680.
We also studied a bis-functionalization of cyclo(Pro,Pro)
5. Interestingly, the 3-carbomethoxy DKP 6b (and also
Org. Lett., Vol. 4, No. 16, 2002
2647

3-carbobenzyloxy DKP 6c, not shown) underwent azidation
with high diastereoselectivity and in higher yields than the
parent DKP 5. The relative stereochemistry of azide 9 was
determined by reduction of the azide group to the free amine,
which upon reflux in benzene cyclized to the tetracyclic
system 10. This structure was further confirmed by single
crystal X-ray analysis (Scheme 5). This cyclization would
Scheme 6. Functionalization of Unsymmetric DKP 12a
Scheme 5. Diastereoselective Bisfunctionalization of DKP 5
(inset is POV chem rendering of X-ray crystal structure of
lactam 10)
a Inset is POV chem rendering of X-ray crystal structure of lactam
14.
halides and acyl halides proceed with high facial selectivity.
The selectivity is presumably due to steric biases inherent
to the tricyclic system. However, other factors such as
stereoelectronic effects cannot be excluded at this time.
Azidation of cyclo(Pro,Pro) and 3-carbomethoxy DKP 6b
as well as derivatization of the azide moiety proceed
smoothly. The electrophilic addition of acyl halides to DKP
5 represents the starting point for our total synthesis of
phakellstatin and phakellin. Further studies toward this end
will be reported in due course.20
only be possible for a syn arrangement of the intermediate
amine and ester, which could be attained either by direct
azide reduction of the syn isomer of 9 or by reduction of the
trans azide ester followed by reversible ring opening via a
presumed imine intermediate. Although it has not been
verified which pathway is followed, if a syn arrangement of
azide and ester is obtained, it is interesting to note that the
steric hindrance introduced by the C-3 substituent does not
invert the facial selectivity. This is true even when C3 is
substituted with the bulky carbobenzyloxy group, which is
presumably folded over the DKP ring system.19
Acknowledgment. Dedicated to Prof. A. I. Meyers on
the occasion of his 70th birthday. We thank the NIH (GM
52964-05), the Robert A. Welch Foundation (A-1280),
Zeneca, and Pfizer for support of these investigations. A.T.K.
was a Summer 2001 NSF-REU participant (CHE-0071889).
D.R. is an Alfred P. Sloan Fellow and a Camille-Henry
Dreyfus Teacher-Scholar. We thank Prof. Robert M. Wil-
liams (Colorado State) for helpful input. We thank Dr. Joe
Reibenspies for X-ray structure determination using instru-
ments obtained with funds from the NSF (CHE-9807975).
The NSF (CHE-0077917) also provided funds for purchase
of NMR instrumentation.
When the starting DKP is not C2-symmetric, i.e., TBS-
protected L-prolyl-L-4-hydroxy proline [cyclo(Pro,Hyp), 12],
electrophilic addition yields a 1:1 mixture of regioisomers
13 and 14, surprisingly with no change in facial selectivity.
The structure of DKP 14 was confirmed by single-crystal
X-ray crystallography (Scheme 6). Hence, introduction of
bulky groups to direct the electrophilic attack of the DKP
enolate does not invert the facial selectivity, suggesting that
other factors are involved in this diastereoselective process.
Supporting Information Available: Selected experi-
1
mental procedures and characterization data (including H
and 13C NMR spectra) for compounds 6a-f and 7-10. This
material is available free of charge via the Internet at
http://pubs.acs.org.
In conclusion, we have demonstrated that electrophilic
additions to proline-derived DKP enolates with both alkyl
OL026140Q
(19) It has been demonstrated using NMR and X-ray crystallography
that aromatic systems are folded over DKP ring system, see ref 13b.
Moreover, the upfield shift of H-6 in benzyl DKP 6f and carbobenzyloxy
DKP 6c suggests that the aromatic rings are indeed folded over the DKP
in solution.
(20) While this manuscript was in review, an approach to racemic
dibromophakellstatin and dibromoisophakellin was recently described:
Wiese, K. J.; Yakushijin, K.; Horne, D. A. Tetrahedron Lett. 2002, 43,
5135-5136.
2648
Org. Lett., Vol. 4, No. 16, 2002