Synthesis of NP25302

Article abstract of DOI:10.1021/ol202381m

An efficient synthesis of NP25302 is presented that relies on 5-endo-dig N-cyclization to establish the bicyclic core and Curtius rearrangement to install the N-acyl vinylogous urea functionality.

Full text of DOI:10.1021/ol202381m

ORGANIC
LETTERS
2011
Vol. 13, No. 22
5964–5967
Synthesis of NP25302
Kiri Stevens,† Andrew J. Tyrrell,† Sarah Skerratt,‡ and Jeremy Robertson*,†
Department of Chemistry, Chemistry Research Laboratory, University of Oxford,
Mansfield Road, Oxford OX1 3TA, U.K., and Pfizer Global Research and Development,
Ramsgate Road, Sandwich CT13 9NJ, U.K.
jeremy.robertson@chem.ox.ac.uk
Received September 2, 2011
ABSTRACT
An efficient synthesis of NP25302 is presented that relies on 5-endo-dig N-cyclization to establish the bicyclic core and Curtius rearrangement to
install the N-acyl vinylogous urea functionality.
Bohemamine (1, Figure 1) was the first reported member
of a small class of alkaloids defined by a 3-N-acylamino-
tetrahydropyrrolizine core.1 Subsequent members, all of
bacterial origin, include jenamidines AꢀC 2ꢀ4,2 NP25302
5,3 bohemamines B 6 and C 7, and a chlorinated bohema-
mine 8.4 Jenamidine A inhibits the growth of chronic
myeloid leukemia K-562 cells (GI50 = 1.9 μg mLꢀ1),2
and both bohemamine and NP25302 inhibit the adhesion
of human promyelocytic leukemia HL-60 cells to Chinese
hamster ovary cells expressing intercellular adhesion mo-
lecule ICAM-1 (CD54) (IC50 = 24.3 and 27.2 μg mLꢀ1
,
respectively).3
With only one exception,5 the vinylogous urea func-
tionality in these natural products6 and their non-natural
† University of Oxford.
‡ Pfizer Global Research and Development.
(1) Doyle, T. W.; Nettleton, D. E.; Balitz, D. M.; Moseley, J. E.;
Grulich, R. E.; McCabe, T.; Clardy, J. J. Org. Chem. 1980, 45, 1324.
(2) (a) Isolation and proposed structure: Hu, J.-F.; Wunderlich, D.;
Thiericke, R.; Dahse, H.-M.; Grabley, S.; Feng, X.-Z.; Sattler, I.
J. Antibiot. 2003, 56, 747. (b) Structural reassignment: Snider, B. B.;
Duvall, J. R.; Sattler, I.; Huang, X. Tetrahedron Lett. 2004, 45, 6725.
(3) Zhang, Q.; Schrader, K. K.; ElSohly, H. N.; Takamatsu, S.
J. Antibiot. 2003, 56, 673.
Figure 1. Known alkaloids of the bohemamine type.
analogues7 has been established synthetically by either
N- or C-cyclization onto a tethered nitrile. We present
a fundamentally different approach, based on Curtius
rearrangement (Scheme 1), for which current literature
precedent is confined to pyridone8 and pyridazine9
(4) Bugni, T. S.; Woolery, M.; Kauffman, C. A.; Jensen, P. R.;
Fenical, W. J. Nat. Prod. 2006, 69, 1626.
(5) Contini, A.; Erba, E.; Trimarco, P. ARKIVOC 2008, (xii), 136.
(6) (a) Snider, B. B.; Duvall, J. R. Org. Lett. 2005, 7, 4519. (b) Duvall,
J. R.; Wu, F.; Snider, B. B. J. Org. Chem. 2006, 71, 8579.
(7) (a) Volovenko, Y. M. Chem. Heterocycl. Compd. 1998, 33, 854.
(b) Katritzky, A. R.; Vincek, A. S.; Steel, P. J. Heterocycles 2008, 76,
1401. (c) Tverdokhlebov, A. V.; Denisenko, A. V.; Tolmachev, A. A.;
Volovenko, Y. M.; Shishkina, S. V.; Shishkin, O. V. Synthesis 2009,
1265. (d) Nestorak, I. Y.; Tverdokhlebov, A. V.; Tolmachev, A. A.;
Volovenko, Y. M. Heterocycles 2010, 81, 611.
(8) (a) Heyns, K.; Vogelsang, G. Chem. Ber. 1954, 87, 1440. (b)
Cannon, W. N.; Jones, R. G. J. Org. Chem. 1957, 22, 1722.
(9) (a) Boger, D. L.; Coleman, R. S. J. Org. Chem. 1986, 51, 3250. (b)
Boger, D. L.; Coleman, R. S. J. Am. Chem. Soc. 1987, 109, 2717.
r
10.1021/ol202381m
Published on Web 10/18/2011
2011 American Chemical Society

a slightly improved yield, and for convenience, this method
is preferred.
Scheme 1. Key Stages in the Synthesis of NP25302 5 and nor-
NP25302 9 [R0 = Alkyl]
Carboxylic acid 18 was isolated following alkaline hy-
drolysis (LiOH, aq CH3CN) of ester 17, but standard
protocols20 for conversion of carboxylic acids into acyl
azides failed withthis compound. The intermediatelithium
carboxylate was also found to be rather unreactive toward
activation (e.g., with ethyl chloroformate), but the potas-
sium carboxylate was sufficiently nucleophilic in DMF to
react with diphenylphosphoryl azide21 and, following op-
timization, acyl azide 10 was obtained reproducibly in
excellent overall yield from ester 17.
substrates. In addition, we describe the first applica-
tion of a 5-endo-dig N-cyclization10ꢀ12 to the synthesis
of a pyrrolizidine-type ring system.13
Scheme 2. Synthesis of Curtius Rearrangement Precursor 10
Initially, the viability of the key 5-endo-dig and Cur-
tius steps was tested in a readily accessible system
leading to (þ)-nor-NP25302 9. N-Protection of (S)-
2-methylproline 1414 and an efficient reduction/oxidation
sequence furnished aldehyde 15 (Scheme 2). Building on
model studies,15 in which various functionalized alkynyl
organometallics were added to N-Boc-prolinal, the addi-
tion of lithiated methyl propiolate16 provided ketone 16
after oxidation with DessꢀMartin periodinane (DMP).
The sequence from (S)-methylproline to ketone 16 was
extremely efficient (71% over five steps), with no purifica-
tion of intermediates required.
The first key step to be addressed, the 5-endo-dig
N-cyclization (f 17), was unsuccessful using the Hg(II)-
mediated conditions described in the closest precedent
for such a process.12,17 However, mild N-deprotection
with p-toluenesulfonic acid,18 followed by addition of
solid NaHCO3, provided acceptable yields (up to 59%)
of the bicyclic product 17. Alternatively, use of TMSOTf
at low temperature to effect the Boc-deprotection19 gave
It was our intention to trap the isocyanate derived from
acyl azide 10 by addition of an isobutenyl organometallic to
provide nor-NP25302 directly. In practice, thermolysis of the
azide led to rapid polymerization, presumably of the inter-
mediate isocyanate, and compatible isobutenyl nucleophiles
for trapping in situ were not found. Instead, rearrangement
was effected in the presence of a variety of alcohols to provide
carbamates 19ꢀ22 (Scheme 3). Each of these was acylated
with 2,2-dimethylacryloyl chloride (28) but the so-formed
N-acyl carbamates 23ꢀ26 could not be deprotected (f 9)
without complication. In most attempts amine 27 was
formed. In one case (26), deprotection was accompanied
by migration of the dimethylacryloyl group to the C(2)
position; in other cases, the dimethylacryloyl fragment
was removed preferentially. Eventually, the Curtius re-
arrangement was performed in the presence of water to
provide amine 27 which was then acylated using the
conditions described in Snider’s synthesis of NP25302 to
afford (þ)-nor-NP25302 9.
(10) 5-Endo-dig N-cyclization onto alkynols: (a) Smith, C. R.; Bunnelle,
E. M.; Rhodes, A. J.; Sarpong, R. Org. Lett. 2007, 9, 1169. (b) Yan, B.;
Zhou, Y.; Zhang, H.; Chen, J.; Liu, Y. J. Org. Chem. 2007, 72, 7783. (c)
Bunnelle, E. M.; Smith, C. R.; Lee, S. K.; Singaram, S. W.; Rhodes, A. J.;
Sarpong, R. Tetrahedron 2008, 64, 7008. (d) Friel, D. K.; Snapper, M. L.;
Hoveyda, A. H. J. Am. Chem. Soc. 2008, 130, 9942.
(11) 5-Endo-dig N-cyclization onto alkynones: Gouault, N.; Le
Roch, M.; Cornee, C.; David, M.; Uriac, P. J. Org. Chem. 2009, 74, 5614.
(12) Application in natural product synthesis [(þ)-preussin from
(S)-phenylalanine]: Overhand, M.; Hecht, S. M. J. Org. Chem. 1994,
59, 4721.
(13) Related 6-endo-dig cyclizations: (a) Turunen, B. J.; Georg, G. I.
J. Am. Chem. Soc. 2006, 128, 8702. (b) Ward, T. R.; Turunen, B. J.;
Haack, T.; Neuenswander, B.; Shadrick, W.; Georg, G. I. Tetrahedron
Lett. 2009, 50, 6494. (c) Niphakis, M. J.; Georg, G. I. J. Org. Chem. 2010,
75, 6019. (d) Niphakis, M. J.; Turunen, B. J.; Georg, G. I. J. Org. Chem.
2010, 75, 6793. (e) Pepe, A.; Pamment, M.; Georg, G. I.; Malhotra, S. V.
J. Org. Chem. 2011, 76, 3527.
This model study had established an efficient nine-step
synthesis of nor-NP25302 from proline derivative 14, and it
was expected that the extra methyl group at C(5) in NP25302
would not complicate the route. (()-trans-2,5-Dimethylpro-
line derivative 29 (Scheme 4) was prepared in two high-
yielding steps from ethyl 2-nitropropionate in a modification
(14) Beck, A. K.; Blank, S.; Job, K.; Seebach, D.; Sommerfeld, Th.
Org. Synth. 1995, 72, 62.
(15) Tyrrell, A. J. D. Phil. Thesis, University of Oxford, Oxford, U.
K., 2008.
(16) Midland, M. M.; Tramontano, A.; Cable, J. R. J. Org. Chem.
1980, 45, 28.
(17) Use of AcOH as the solvent instead of CH3NO2 gave some
product (16%), and replacement of the reported Hg(OAc)2 with
Hg(OCOCF3)2 raised the yield to 38%.
(18) These conditions were found to be preferred in a related system:
Reed, P. E.; Katzenellenbogen, J. A. J. Org. Chem. 1991, 56, 2624.
(19) Cf. Bastiaans, H. M. M.; van der Baan, J. L.; Ottenheijm,
H. C. J. J. Org. Chem. 1997, 62, 3880.
(20) For example: (PhO)2PON3/Et3N in toluene or toluene/t-BuOH
was hampered by the limited solubility of acid 18 in these solvents; clean
preparation of the acid chloride, in readiness for substitution with azide,
could not be achieved.
(21) (a) Shioiri, T.; Ninomiya, K.; Yamada, S.-i. J. Am. Chem. Soc.
1972, 94, 6203. (b) Ninomiya, K.; Shioiri, T.; Yamada, S. Tetrahedron
1974, 30, 2151.
Org. Lett., Vol. 13, No. 22, 2011
5965

Scheme 3. Curtius Rearrangement Trials and Completion of the
Synthesis of (þ)-nor-NP25302
Scheme 4. Synthesis of (()-NP25302
of the literature procedure,6b and aldehyde 30 was obtained
straightforwardly in four further steps. In contrast to the
simpler aldehyde 15, the dimethyl variant 30 was unreactive
toward lithiated methyl propiolate at (the low) temperatures
at which the organometallic remained stable. Related alkynyl
organometallics were explored, including those derived from
tert-butyl propiolate and propiolic acid itself; although the
additions were successful, the derived ketones could not be
induced to cyclize in acceptable yields. Ultimately, the use of
lithiated isopropyl propiolate solved the problem. This
lithium acetylide derivative has not been reported, but it
proved reliable and sufficiently stable to effect efficient addi-
tion to hindered aldehyde 30. Subsequent 5-endo-dig cycliza-
tion of the derived ketone 31 proceeded without complication
under the preferred conditions for cyclization of model
substrate 16.22 The three final steps to complete the synthesis
were achieved under identical conditions and in comparable
yields to the model study.
the NMR spectra for NP25302 was affected by the quality
of the solvent and the concentration of the sample.24 For
example, when the 1H NMR spectrum was run in CDCl3
that had been filtered through basic alumina, the two
olefinic protons appeared as a single resonance [cf. lit.3
δH 5.75 (1 H, s) and 6.01 (1 H, s)], but addition of TFA
in 10 mol % aliquots led to splitting and shifting of the
olefinic resonances until the values approached those
reported. At the same time, colorless needles separated
out of the NMR solvent that proved suitable for single
crystal X-ray analysis. The structure (Figure 2);the
TFA salt of NP25302;confirmed the success of the
synthesis.
The NMR data for our synthetic NP25302 differed
sufficiently from those published (see Supporting Infor-
mation for details) to cast doubts on either the stereo-
chemistry or the connectivity in the final product. How-
ever, the trans-relationship of the C(5) and C(7a) methyls
was derived from amine 29 whose structure had been
confirmed by X-ray analysis of 29 HCl;23 in addition,
3
HMBC data for synthetic 5 showed an absence of H(5)ꢀ
C(3) and R-H(7)ꢀC(1) correlations, indicating approxi-
mately 90° relationships, consistent with the expected
stereochemistry. It became clear that the appearance of
ꢀ
Figure 2. X-ray molecular structure of 5 TFA [CF3CO2 ion
3
omitted for clarity].
This total synthesis of (()-NP25302 is longer than
Snider’s published route6b primarily due to the sequence
of functional group interconversions required to access a
suitably reactive acceptor 30 for alkynyl organometallic
reagents. Nevertheless, the key 5-endo-dig cyclization and
Curtius rearrangement steps successfully delivered the
defining structural features of the natural product in an
efficient overallsynthesis(28% from 29, average 88% yield
per step).
(22) During the course of this investigtion, we found that the
TMSOTf conditions for 5-endo-dig cyclization were at least as efficient
with less activated analogues of alkynyl ketone 31 bearing H, CH2OH,
or CH2OTBS in place of CO2i-Pr. Fuller details of the scope of this
cyclization will be described in due course.
(23) CCDC 834708 and 834707 contain crystallographic data for this
paper (for compounds 29 HCl and 5 TFA, respectively). These can be
3
3
obtained free of charge from the Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
(24) We are very grateful to Professor Barry Snider, Brandeis Uni-
versity, for informative correspondence concerning the NMR spectra of
natural and synthetic NP25302.
5966
Org. Lett., Vol. 13, No. 22, 2011

Acknowledgment. We thank Pfizer, the EPSRC, and the
University of Oxford for funding; Drs. T. D. W. Claridge
and B. Odell, University of Oxford, for extensive NMR
support; and Dr. A. L. Thompson, University of Oxford, for
help with the X-ray analyses.
Supporting Information Available. Experimental pro-
cedures, characterization data, and copies of 1H and 13
NMR spectra for new compounds. This material is
available free of charge via the Internet at http://pubs.
acs.org.
C
Org. Lett., Vol. 13, No. 22, 2011
5967