Synthesis of sterically hindered N-acylated amino acids from N-carboxyanhydrides

Article abstract of DOI:10.1021/ol500523n

Sterically hindered N-acyl, gem-disubstituted amino acids are easily prepared via the addition of organometallic reagents to N-carboxyanhydrides (NCA). The process tolerates a wide variety of functional groups and allows the synthesis of amide products not readily accessible by traditional acylation chemistry. The existence of an isocyanate intermediate was established by in situ IR spectroscopy.

Full text of DOI:10.1021/ol500523n

Letter
pubs.acs.org/OrgLett
Synthesis of Sterically Hindered N‑Acylated Amino Acids from
N‑Carboxyanhydrides
Gabriel Schafer and Jeffrey W. Bode*
̈
Laboratorium fur Organische Chemie, Department of Chemistry and Applied Biosciences, ETH Zurich, Vladimir Prelog Weg 1-5,
̈
̈
8093 Zurich, Switzerland
̈
S
* Supporting Information
ABSTRACT: Sterically hindered N-acyl, gem-disubstituted
amino acids are easily prepared via the addition of organo-
metallic reagents to N-carboxyanhydrides (NCA). The process
tolerates a wide variety of functional groups and allows the
synthesis of amide products not readily accessible by
traditional acylation chemistry. The existence of an isocyanate
intermediate was established by in situ IR spectroscopy.
mide bond formation is among the most prevalent and
react with the second equivalent of the organometallic reagent
to form the desired amide (Scheme 1).
A
important reactions in organic synthesis.1 The common
method of preparing amides is the dehydrative coupling of
amines with carboxylic acids by the action of a coupling agent.2
Although this approach is routinely used, it is not without
limitations.3 The synthesis of sterically hindered amides can be
cumbersome due to the slow attack of the amine onto the
activated carboxylate. To address the formation of these
important amide products, we recently reported the addition
of Grignard reagents to isocyanates. Using this methodology,
extremely hindered amides can be prepared with ease and in
high yield.4
Scheme 1. Acylation of Amino Acids
Our approach, however, was not suitable for the preparation
of a key class of hindered amides, those derived from gem-
disubstituted amino acids, due to the lack of a suitable
isocyanate starting material. Recent advances in the con-
struction of enantiomerically enriched disubstituted amino
acids improve access to these building blocks,5 but do not
address the challenging acylation of these sterically hindered
substrates. Traditionally, N-acyl substrates are prepared from
the corresponding amino acid and an acid chloride under basic
conditions.6 This approach works well for sterically unbiased
starting materials, but is usually low yielding for sterically
hindered examples and incompatible with sensitive functional
groups.7
We now report the clean and convenient formation of
sterically hindered N-acyl, gem-disubstituted amino acids by the
addition of Grignard or organolithium reagents to N-carboxy-
anhydrides (NCAs or Leuchs anhydrides). Our studies were
inspired by literature reports on the use and properties of N-
carboxyanhydrides in polymer chemistry.8 NCAs are routinely
used as starting materials in the preparation of polypeptides via
ring-opening polymerization under basic conditions.9 We
postulated that it should be possible to prepare N-acylated
amino acids in one step by adding 2 equiv of a Grignard reagent
to an NCA. The first equivalent of the Grignard reagent would
abstract the N−H proton leading to the formation of an α-
isocyanatocarboxylate.10 This transient intermediate would
We began our investigation by the addition of a slight excess
of mesitylmagnesium bromide to a solution of 3-oxa-1-
azaspiro[4.5]decane-2,4-dione 1a (cyclohexyl-NCA) in THF
(Scheme 2). Only a complex mixture of products could be
obtained when the Grignard reagent was introduced at 0 °C,
most likely due to polymerization of the NCA starting material.
Upon lowering the temperature to −78 °C, we obtained the
desired product 3a in very good yield. The isolation of the
Received: February 18, 2014
Published: February 26, 2014
© 2014 American Chemical Society
1526
dx.doi.org/10.1021/ol500523n | Org. Lett. 2014, 16, 1526−1529

Organic Letters
Letter
a
Scheme 2. Addition of Various Grignard Reagents to Cyclohexyl-NCA (1a)
a
Isolated yields. 1a (0.50 mmol), Grignard reagent 2 (1.05 mmol), THF (0.2 M), −78 °C, 15 min; rt, 1 h. b Grignard reagent prepared via Mg−Br
c
exchange at low temperature before the NCA in THF (0.2 M) was added. 3-[Bis(trimethylsilyl)amino]phenylmagnesium chloride was used.
product was simple; after aqueous workup, the crude material
was washed with Et2O to obtain the analytically pure N-
acylated pure amino acid 3a.
organometallic reagent. Indeed, when cyclohexyl-NCA 1a was
treated with 1.0 equiv of mesityl Grignard, followed by the
addition of 1.1 equiv of a second Grignard reagent 2, only the
product stemming from Grignard reagent 2 was observed.
Using this modified procedure, we could increase the yield of
several products derived from sterically unbiased Grignard
reagents (see Scheme 3 in comparison to Scheme 2).
Encouraged by the success of the addition of Grignard
reagents to NCAs, we also investigated the use of organo-
lithium reagents in this amide-forming methodology. Due to
the instability of most organolithium reagents at room
temperature, we modified the addition sequence: the organo-
lithium species was generated at low temperature, followed by
the addition of the NCA in THF at −78 °C (Scheme 4). Using
this reverse addition protocol several interesting heterocyclic
amides could be prepared in good yields (5a−e). In addition,
lithium acetylide (5f) and lithium enolate (5g) were feasible
substrates. It was also possible to use ortho-lithiated arenes
bearing protected alcohols, free acids, amides, and sulfonamides
as nucleophiles; the corresponding N-acyl amino acids (5h−k)
were isolated in high yields.
With the optimal conditions in hand, we explored the
substrate scope of the Grignard reagent (Scheme 2). As
anticipated from the results with mesitylmagnesium bromide,
the reaction worked well with standard aromatic Grignard
reagents to provide the corresponding products (3b−g) in high
yields. Products containing sensitive functional groups such as
nitrile (3h), ester (3m), or free amine (3j) were easily
prepared, as well as products derived from pyridinyl (2k) and
styrenyl (2i) Grignard reagents.11 Many of these examples
would be difficult to synthesize via traditional methods.
Hindered Grignard reagents including pentafluorophenyl, 2,6-
dichlorophenyl, isopropyl, or tert-butyl Grignard also provided
the desired products (3o−3r). Even the extremely bulky 2,4,6-
tri-tert-butylphenyl Grignard reagent could be used to cleanly
afford sterically congested amide 3n.
The reaction with 2 equiv of sterically unbiased, aliphatic
Grignard reagents was problematic; only moderate yields of the
corresponding products (3u−w) could be isolated. We
attributed the low yield to a competing nucleophilic attack of
the Grignard reagent to the anhydride carbonyl C5, which
would lead to ring opening of the NCA followed by
polymerization.
The scope of the NCA was found to be equally broad
(Scheme 5). NCAs derived from aliphatic, gem-disubstituted
amino acids were excellent substrates, and a wide range of
different products could be prepared (6a−f). Products
containing aromatic side chains (6h−i) were also accessible.
Furthermore, our methodology allowed for the synthesis of a
N-acylated α,β-dehydroamino acid (6g). Heteroatom contain-
ing NCAs also cleanly reacted with Grignard reagents to
We hypothesized that the addition of a sterically hindered
Grignard reagent to an NCA at low temperature could
selectively generate the intermediate isocyanate (vida infra),
which could then be trapped by a second, less hindered
1527
dx.doi.org/10.1021/ol500523n | Org. Lett. 2014, 16, 1526−1529

Organic Letters
Letter
Scheme 3. Addition of Grignard Reagents (2) to Cyclohexyl-
Scheme 5. Addition of Grignard Reagents (2) to gem-
a
NCA (1a) Using Mesitylmagnesium Bromide As a Sacrificial
Disubstituted NCAs (1)
a
Base
a
Isolated yields. 1a (0.50 mmol), Grignard reagent 2 (1.05 mmol),
THF (0.2 M), −78 °C, 15 min; rt, 1 h.
Scheme 4. Addition of Organolithium Reagents (4) to
a
Cyclohexyl-NCA (1a)
a
Isolated yields. 1 (0.50 mmol), Grignard reagent 2 (1.05 mmol),
b
THF (0.2 M), −78 °C, 15 min; rt, 1 h. TMSCl (1.05 equiv) was
added. −78 °C to rt, 5 h.
c
In order to establish the existence of an intermediate α-
isocyanatocarboxylate we followed the reaction by react-IR.
Isocyanates are known to have a strong, isolated absorption
band at 2250 cm−1. Upon addition of mesitylmagnesium
bromide (2.1 equiv) to a solution of cyclohexyl-NCA 1a (1.0
equiv) at −78 °C, an absorption band at 2250 cm−1 appeared
immediately, consistent with the formation of an intermediate
isocyanate (Figure 1). To our surprise, this isocyanate was
stable at −78 °C, even in the presence of unreacted Grignard
reagent. Only upon warming of the reaction mixture was the
isocyanate consumed by the second equivalent of the Grignard
reagent to form the amide product.12
a
Isolated yields. In situ generation of 4 (2.2 equiv), then addition of 1a
(0.50 mmol) in THF (0.2 M), −78 °C to rt, 5 h.
An advantage of this methodology is that the NCA substrates
are readily prepared in one step from the amino acid and
triphosgene and, if necessary, amenable to column chromatog-
raphy.13 The NCAs are easily handled compounds that can be
stored for months without any sign of decomposition (see
Supporting Information for further information).
provide the corresponding products in high yields (6j−n).
Interestingly, product 6j was prone to fluoride elimination. This
side reaction was suppressed by the addition of TMSCl prior to
the introduction of the Grignard reagent. Finally, we
investigated the addition of mesitylmagnesium bromide to an
enantiopure NCA and obtained N-acylated amino acid 6o
without racemization.
In summary, we have identified a new approach to the
synthesis of N-acyl, gem-disubstituted amino acids by the
1528
dx.doi.org/10.1021/ol500523n | Org. Lett. 2014, 16, 1526−1529

Organic Letters
Letter
(7) An alternative approach towards sterically hindered N-acylated
Aib-derivatives was disclosed by Obrecht and Heimgartner. Obrecht,
D.; Heimgartner, H. Helv. Chim. Acta 1987, 70, 102−115.
(8) Seminal work: (a) Leuchs, H. Ber. Dtsch. Chem. Ges. 1906, 39,
857−861. (b) Leuchs, H.; Manasse, W. Ber. Dtsch. Chem. Ges. 1907,
40, 3235−3249. (c) Leuchs, H.; Geiger, W. Ber. Dtsch. Chem. Ges.
1908, 41, 1721−1726.
(9) Selected reviews: (a) Kricheldorf, H. R. Angew. Chem., Int. Ed.
2006, 45, 5752−5784. (b) Hadjichristidis, N.; Iatrou, H.; Pitsikalis, M.;
Sakellariou, G. Chem. Rev. 2009, 109, 5528−5578. (c) Deming, T. J.
Adv. Mater. 1997, 9, 299−311.
(10) α-Isocyanatocarboxylic acids were speculated to be intermedi-
ates in base initiated NCA polymerizations. However, several studies
excluded their existence, e.g. Kricheldorf treated Bn-Glu NCA with the
strong base HMDS, but no isocyanate absorption could be observed
by FTIR. Kricheldorf, H. R.; Greber, G. Chem. Ber. 1971, 104, 3131−
3145.
Figure 1. React-IR differential spectrum.
(11) The majority of Grignard reagents bearing sensitive functional
groups were prepared according to: Krasovskiy, A.; Knochel, P. Angew.
Chem., Int. Ed. 2004, 43, 3333−3336.
(12) We cannot completely rule out the possibility that the
isocyanate is only a resting state of the reaction and the amide
formation proceeds via a different mechanism.
addition of organometallic reagents to N-carboxyanhydrides
(NCA). The reaction proceeds via an intermediate α-
isocyanatocarboxylate, whose existence was demonstrated by
react-IR. This method is compatible with a wide range of
functional groups and the use of enantiomerically pure amino
acid derived NCAs.
́
(13) Daly, W. H.; Poche, D. Tetrahedron Lett. 1988, 29, 5859−5862.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental procedures and characterization data for all
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: bode@org.chem.ethz.ch.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was supported by ETH Research Grant ETH-12 11-
1. We thank the ETH Mass Spectroscopy Service for high-
■
resolution mass spectrometry data, Claudio Grunenfelder
̈
(group Prof. Helma Wennemers, ETH) for assistance with
react-IR measurements, and Lukas Leu (ETH) for experimental
assistance.
REFERENCES
■
(1) Roughley, S. D.; Jordan, A. M. J. Med. Chem. 2011, 54, 3451−
3479.
(2) Valeur, E.; Bradley, M. Chem. Soc. Rev. 2009, 38, 606−631.
(3) Pattabiraman, V. R.; Bode, J. W. Nature 2011, 480, 471−479.
(4) Schafer, G.; Matthey, C.; Bode, J. W. Angew. Chem., Int. Ed. 2012,
̈
51, 9173−9175.
(5) (a) Reeves, J. T.; Tan, Z.; Herbage, M. A.; Han, Z. S.; Marsini, M.
A.; Li, Z.; Li, G.; Xu, Y.; Fandrick, K. R.; Gonnella, N. C.; Campbell,
S.; Ma, S.; Grinberg, N.; Lee, H.; Lu, B. Z.; Senanayake, C. H. J. Am.
Soc. Chem. 2013, 135, 5565−5568. (b) Ooi, T.; Takeuchi, M.;
Kameda, M.; Maruoka, K. J. Am. Chem. Soc. 2000, 122, 5228−5229.
(c) Vachal, P.; Jacobsen, E. N. Org. Lett. 2000, 2, 867−870.
(d) Williams, R. M.; Im, M. N. J. Am. Chem. Soc. 1991, 113, 9276−
9286. (e) Seebach, D.; Sting, A. R.; Hoffmann, M. Angew. Chem., Int.
Ed. 1996, 35, 2708−2748.
(6) (a) Fischer, E. Ber. Dtsch. Chem. Ges. 1899, 32, 2451−2471.
(b) Ronwin, E. J. Org. Chem. 1953, 18, 1546−1553.
1529
dx.doi.org/10.1021/ol500523n | Org. Lett. 2014, 16, 1526−1529