A convenient route to α-amino acids with β-alkyne substituents from a serine derived aziridine

Article abstract of DOI:10.1016/S0040-4039(01)01886-X

1-[(S)-1-(2-Nitrobenzenesulfonyl)-aziridin-2-yl]-4-methyl-2,6,7- trioxabicyclo[2.2.2]octane 5 was ring opened regioselectively with a variety of lithium acetylides to give α-amino acids bearing γ,δ-unsaturation in very good to excellent yields. The 2-nitrobenzenesulfonyl and OBO ester protecting groups were removed in excellent overall yield.

Full text of DOI:10.1016/S0040-4039(01)01886-X

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 8713–8716
A convenient route to a-amino acids with b-alkyne substituents
from a serine derived aziridine
John J. Turner, Michiel A. Leeuwenburgh, Gijs A. van der Marel and Jacques H. van Boom*
Leiden Institute of Chemistry, Gorlaeus Laboratories, PO Box 9502, 2300 RA Leiden, The Netherlands
Received 31 August 2001; accepted 5 October 2001
Abstract—1-[(S)-1-(2-Nitrobenzenesulfonyl)-aziridin-2-yl]-4-methyl-2,6,7-trioxabicyclo[2.2.2]octane 5 was ring opened regioselec-
tively with a variety of lithium acetylides to give a-amino acids bearing g,d-unsaturation in very good to excellent yields. The
2-nitrobenzenesulfonyl and OBO ester protecting groups were removed in excellent overall yield. © 2001 Elsevier Science Ltd. All
rights reserved.
The synthesis of a-amino acids with novel b-side chains
has aroused much interest over the last couple of
decades. Of the many modifications known, amino
acids with unsaturated side chains are of particular
interest due to their potential biological activity as
specific irreversible enzyme inhibitors, also known as
‘suicide substrates’.1 Specifically it has been found that
compounds containing carbonꢀcarbon triple bonds
serve to deactivate a variety of enzymes.2 Consequently,
there have been many reports on the synthesis of
a-amino acids with b-side chains containing alkyne
functionality;1,3 however, general stereoselective routes
to these compounds are not available. A potentially
general route to a-amino acids bearing g,d-unsaturation
might be realised via the ring opening of an aziridine-2-
carboxylate with lithium acetylides. Previous reports
concerning the ring opening of such compounds with
carbon nucleophiles indicate that the use of tosyl-N-
aziridine-2-carboxylate esters results in unwanted
attack at the ester functionality.4,5 However, it has been
shown that this problem can be circumvented by
employing no carboxylate protection.4a–b,6 On the other
hand, Lajoie and co-workers7 have demonstrated that
masking the carboxylic acid functionality utilising the
Scheme 1. Reagents and conditions: (i) NaI (0.2 equiv.), DMF, rt, 2 days, 45%; (ii) BF3·OEt2 cat., DCM, rt, 4 h, 78%; (iii) Ph3P
(1.2 equiv.), DEAD (1.2 equiv.), 0°C, 1 h, 98%.
Keywords: Mitsunobu reaction; nitrobenzenesulfonamides; carbon nucleophiles; aziridine; ring opening.
* Corresponding author. Tel.: +31 71 5274274; fax: +31 71 5274307; e-mail: j.boom@chem.leidenuniv.nl
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01886-X

8714
J. J. Turner et al. / Tetrahedron Letters 42 (2001) 8713–8716
base-stable 2,6,7-trioxabicyclo[2.2.2]octane (OBO-ester)
Corey protecting group also prevents attack on the
carboxylate by organometallic reagents. Crucially,
cleavage of the ortho-ester proceeded with minimal
racemisation. It was therefore envisaged that an attrac-
tive substrate for ring opening with lithium acetylides
would comprise of an aziridine with the carboxylic acid
protected as the OBO ester. In addition, replacement of
the tosyl group for the 2-nitrobenzenesulfonyl (oNs)
group would avoid the harsh conditions required4,6 for
the removal of the former. We here report a convenient
synthesis of a-amino acids bearing a b-alkyne sub-
stituent via regioselective ring opening of serine derived
aziridine 5 with lithium acetylides.
prop-1-ynyl)-lithium (6d) in the same amount, no reac-
tion was observed (entry 4). The desired product could
be obtained simply by increasing the amount of 6d used
(entries 5 and 6) with the optimum being 4 equiv.,
whereupon 7d was isolated in 98% yield. A similar trend
was also observed for the synthesis of the corresponding
homologue, 7e, where the same excess of reagent (6e)
was necessary for efficient ring opening (entry 7).
Having examples of the successful ring opening of 5
with simple lithium acetylides in hand, attention was
turned to lithio-alkynes bearing more functionality and
steric hindrance. Accordingly, an excess of the lithium
anion of fully benzylated methyl 6,7-dideoxy-a-D-gluco-
hept-6-ynopyranoside9 6f was reacted with 5, providing
the desired product 7f in 85% yield (entry 8). In con-
trast, only 2 equiv. of the anion of fully benzylated
The synthesis of aziridine 5 (Scheme 1) was realised by
Mitsunobu-mediated intramolecular condensation of 2-
b-
-galactopyranosyl-ethyne10 6g (entry 9) were needed
D
nitrobenzenesulfonyl protected
L
-serine OBO ester 4.8
to effect efficient ring opening of 5, furnishing 7g in 82%
yield. It is noteworthy (entry 8) that when ring opening
of 5 was attempted with 2 equiv. of 6f, formation of 7f
could not be observed.
Compound 4 was obtained following a similar proce-
dure to that described previously7 for the synthesis of
the corresponding Fmoc, Boc and benzyloxylcarbonyl
derivatives. Thus, reaction of the caesium salt of 2-
nitrobenzenesulfonyl
L
-serine 1 with 3-methyl-3-(tolu-
To demonstrate the feasibility of liberating the amino
and carboxylic acid functionalities of these ring opened
enesulfonyloxymethyl)oxetane 2 in the presence of a
catalytic amount of sodium iodide afforded the oxetane
ester 3. The latter compound was then rearranged
catalytically with boron trifluoride ethyl etherate to the
corresponding OBO ester 4. Treatment of 4 with
triphenyl phosphine and diethylazodicarboxylate in
THF afforded aziridine 5 in an overall yield of 34%.
Table 1. Ring-opening of aziridine 5
Entry
Lithium acetylide Equiv. (reagent) Yield (%)
6, R=
1
2
TMS (6a)
TMS (6a)
2
7a R=TMS, 42
7b R=H, 8
8, 30
The ring opening of aziridine 5 (Scheme 2) was first
attempted with an excess of lithium trimethylsilyl
acetylide 6a, in THF (entry 1, Table 1). The reaction
afforded the expected product 7a in 42% yield, but this
was also accompanied by 7b and dimer 8 in 8 and 30%
yields, respectively. The formation of 7b presumably
arises from in situ silyl–lithium exchange. The lithium
anion of 7b successfully competes with 6a in opening
aziridine 5 leading to dimer 8. However, formation of 8
could be suppressed by using a large excess of 6a, with
ring opened products 7a and 7b being obtained in 48
and 46% yields, respectively (entry 2).
10
7a, 48
7b, 46
3
4
5
6
7
Ph (6c)
2
2
3
4
4
7c, 84
0
7d, 49
7d, 98
CH2OTHP (6d)
CH2OTHP (6d)
CH2OTHP (6d)
CH2CH2OTHP
(6e)
7e, 86
8
(6f)
4a
7f, 85
7g, 82
In order to optimise reaction conditions and prevent
similar in situ decomposition and side product forma-
tion to that described above, the use of more stable
lithio-alkyne reagents was explored. To this end, lithium
acetylide 6c was reacted with aziridine 5 (entry 3)
providing a homogenous ring opened product, 7c, in
very good yield. However, when ring opening was
attempted with known (3-tetrahydropyran-2-yloxy-
9
(6g)
2
a Using 2 equiv. of 6f, only starting compounds were recovered.
Scheme 2.

J. J. Turner et al. / Tetrahedron Letters 42 (2001) 8713–8716
8715
Scheme 3. Reagents and conditions: (i) K2CO3 (4 equiv.), PhSH (3 equiv.), 0.05 M ACN/DMSO 49/1 (v/v), 50°C, 4 h; (ii) a.
TFA/DCM/H2O (v/v/v), rt, 30 min, b. Cs2CO3 (5 equiv.), dioxane/MeOH/H2O 1/1/2 (v/v/v), rt, 16 h, 89% (three steps from 7g);
(iii) Boc2O (1.5 equiv.), Na2CO3 (1.5 equiv.), dioxane/H2O 1/1 (v/v), rt, 5 h, 92%; (iv) Pd/C, H2, dioxane/EtOH/H2O 2/2/1 (v/v/v),
rt, 4 h; (v) CH2N2, Et2O/DCM, rt, 10 min 96% (two steps from 11); (vi) Pd(OH)2/C, H2, dioxane/EtOH/H2O 2/2/1 (v/v/v), rt, 16
h; (vii) allylbromide (2 equiv.), KHCO3 (2 equiv.), DMF, rt, 16 h; (viii) Ac2O (2 equiv. per OH), pyridine, rt 16 h, 81% (three steps
from 11).
Erkelens for recording NMR spectra, Hans van den
Elst for performing the mass spectrometric analyses and
Nico Meeuwenoord for conducting HPLC analyses.
products, ortho-nitrobenzenesulfonyl (oNs) cleavage
conditions were applied to 7g, as previously described for
primary nosyl sulfonamides,11 to give 9 (Scheme 3).
Subsequent acid and base treatment7 to cleave the OBO
ester in 9 furnished novel unsaturated C-glycosyl amino
acid 10 in excellent yield over the three steps. For
additional proof of the structure and stereochemistry, the
latter compound was converted to the corresponding Boc
protected derivative 11, subjected to partial hydrogena-
tion followed by esterification of the carboxyl group, to
References
1. Abdulganeeva, S. A.; Erzhanov, K. B. Russ. Chem. Rev.
1991, 60, 676–688 and references cited therein.
2. Rando, R. R. Methods in Enzymology; Academic Press:
New York, San Francisco, London, 1977.
furnish known fully benzylated b-D-galactose-(CH2)2-
3. For further examples, see: (a) Boyd, D. R.; Sharma, N.
D.; Bowers, N. I.; Duffy, J.; Harrison, J. S.; Dalton, H.
J. Chem. Soc., Perkin Trans. 1 2000, 1345–1350; (b) Wolf,
L. B.; Tjen, K. C. M. F.; Rutjes, F. P. J. T.; Hiemstra,
H.; Schoemaker, H. E. Tetrahedron Lett. 1998, 39, 5081–
5084; (c) Miura, T.; Masuo, N.; Fusamae, Y.; Kajimoto,
T.; Ida, Y. Synlett 1998, 631–633; (d) Soloshonok, V. A.;
Avilov, D. V.; Kukhar, V. P. Tetrahedron: Asymmetry
1996, 7, 1547–1550; (e) Roedbotten, S.; Benneche, T.;
Undheim, K. Acta Chem. Scand. 1997, 51, 873–880; (f)
Dunn, M. J.; Jackson, R. F. W.; Pietruszka, J.; Turner,
D. J. Org. Chem. 1995, 60, 2210–2215; (g) Echavarren, A.
M.; Cardenas, D. J.; Castano, A. M.; Cuerva, J. M.;
Mateo, C. Bull. Soc. Chim. Belg. 1994, 103, 549–558.
4. (a) Church, N. J.; Young, D. W. J. Chem. Soc., Perkin
Trans. 1 1998, 1475–1482; (b) Church, N. J.; Young, D.
W. J. Chem. Soc., Chem. Commun. 1994, 943–944; (c)
Baldwin, J. E.; Spivey, A. C.; Schofield, C. J.; Sweeney, J.
B. Tetrahedron 1993, 49, 6309–6330.
asparagine methyl ester 12 with full agreement of analyt-
ical data to that reported previously.12 Furthermore,
compound 11 could be readily converted to the corre-
sponding O-acetylated fully saturated C-glycosylated
amino acid derivative 13 by subsequent hydrogenation,
allyl ester formation and acetylation. It should be noted
that omission of carboxylic acid protection prior to
acetylation resulted in an inseparable mixture of prod-
ucts. This problem was circumvented by protection of the
carboxylic acid before acetylation.
In summary, the ring opening of aziridine 5 was shown
to proceed with complete regioselectivity and high yield
using lithium acetylides. The facile conversion of 7g via
novel unsaturated C-glycosylated amino acid 11 to
known C-glycosylated amino acid 12, and the corre-
sponding acetylated allyl ester derivative 13, demon-
strates the potential of this methodology towards the
construction of glycopeptide isosteres suitable for incor-
poration into peptide sequences.
5. (a) It should also be noted that attack on the ester
functionality was the only observed reaction when
lithium trimethylsilylacetylide was added to (S)-1-(2-
nitrobenzenesulfonyl)-aziridine-2-carboxylic acid tert-
butyl ester5b at low temperature; (b) Turner, J. J.;
Sikkema, F. D.; Filippov, D. V.; van der Marel, G. A.;
Boom, J. H. Synlett, in press.
Acknowledgements
This work was financially supported by Unilever. The
authors would like to thank Fons Lefeber and Kees
6. Church, N. J.; Young, D. W. Tetrahedron Lett. 1995, 36,
151–154.

8716
J. J. Turner et al. / Tetrahedron Letters 42 (2001) 8713–8716
7. (a) Blaskovich, M. A.; Evindar, G.; Rose, N. G. W.;
H1, J1,2=3.5 Hz), 4.29 (bd, 1H, H5, J5,4=9.7 Hz), 3.85
(m, 1H, H3), 3.72 (app. dt, 1H, H9), 3.60–3.49 (m, 8H,
3×CH2 OBO, H2, H4), 3.39 (s, 3H, OMe), 2.59 (ABXY,
2H, H8 and H8%, J8,9=4.5 Hz, J8,5=1.6Hz, J8%,9=8.8 Hz,
Wilkinson, S.; Luo, Y.; Lajoie, G. A. J. Org. Chem. 1998,
63, 3631–3646; (b) Blaskovich, M. A.; Lajoie, G. A. J.
Am. Chem. Soc. 1993, 115, 5021–5030; (c) Blaskovich, M.
A.; Lajoie, G. A. Tetrahedron Lett. 1993, 34, 3837–3840.
8. All new compounds were fully characterised by 1H and
13C NMR spectroscopy as well as mass spectrometry.
Data for selected examples are as follows.
J
8%,5=1.7 Hz), 0.62 (s, 3H, CH3 OBO). 13C NMR (100
MHz, CDCl3): l 147.2 (C-NO2), 138.9, 138.7, 138.2
(3×Cq Bn), 136.4 (C-SO2), 132.8, 132.3, 131.1 (3×CHarom
oNs), 128.4–127.4 (CHarom Bn), 124.8 (1×CHarom oNs),
107.4 (Cq OBO), 98.3 (C1), 82.4 (C4), 81.9 (Cq alkyne),
81.0 (C3), 79.6 (Cq alkyne), 79.3 (C2), 75.8, 75.3, 73.5
(3×CH2 Bn), 72.4 (3×CH2 OBO), 61.9 (C5), 57.3 (C9),
55.5 (OMe), 30.4 (Cq OBO), 21.5 (C8), 13.9 (CH3 OBO).
MS (ESI) m/z 837.4 [M+Na]+.
1-[(S)-1-(2-Nitrobenzenesulfonyl)-aziridin-2-yl]-4-methyl-
2,6,7-trioxabicyclo[2.2.2]octane 5: 1H NMR (300 MHz,
CDCl3): l 8.31–8.28 (m, 1H, oNs), 7.75–7.72 (m, 3H,
oNs), 3.90 (s, 6H, 3×CH2 OBO), 3.19 (dd, 1H, H2,
J2,3=7.2 Hz, J2,3%=4.6 Hz), 2.85 (d, 1H, H3), 2.69 (d, 1H,
H3%), 0.80 (s, 3H, CH3 OBO). 13C NMR (75 MHz,
CDCl3): l 147.9 (C-NO2), 134.4, 131.9, 130.8 (3×CHarom
oNs), 130.5 (C-SO2), 123.9 (1×CHarom oNs), 104.7 (Cq
OBO), 72.3 (3×CH2 OBO), 40.6 (C2), 31.2 (C3), 30.3 (Cq
OBO), 13.5 (CH3 OBO). MS (ESI) m/z 356.9 [M+H]+,
379.0 [M+Na]+.
(S)-2-tert-Butoxycarbonylamino-5-(2,3,4,6-tetra -O-ace-
tyl-b-
1H NMR (300 MHz, CDCl3): l 5.98–5.85 (m, 1H,
CH2ꢁCH), 5.41 (dd, 1H, H4%, J4%,3%=3.2 Hz, J4%,5%=0.9
Hz), 5.37–5.24 (m, 2H, CH2ꢁCH), 5.11–4.97 (m, 3H, NH,
H2%, H3%), 4.64–4.62 (m, 2H, OCH2 allyl), 4.33–4.20 (m,
D-galactopyranosyl)-pentanoic acid allyl ester 13:
6
6
6
1-[N-(2-Nitrobenzenesulfonyl)-(1S)-1-amino-4-trimethyl-
silanyl-but-3-ynyl]-4-methyl-2,6,7-trioxabicyclo[2.2.2]-
1H, H2), 4.15–4.03 (m, 2H, 2H6%), 3.86–3.81 (m, 1H, H5%),
3.40–3.33 (m, 1H, H1%), 2.15, 2.05, 2.05, 1.97 (4×s, 4×3H,
4×CH3 Ac), 1.81–1.38 (m, 6H, 2H3, 2H4, 2H5), 1.45 (s,
9H, 3×CH3 tBu). 13C NMR (75 MHz, CDCl3): l 172.4
(C1), 170.4, 170.2, 170.1, 169.8 (4×CꢁO Ac), 155.3
1
octane 7a: H NMR (200 MHz, CDCl3): l 8.16–8.12 (m,
1H, oNs), 7.97–7.93 (m, 1H, oNs), 7.78–7.74 (m, 2H,
oNs), 6.03 (d, 1H, NH, JNH,1=9.5 Hz), 3.77–3.52 (m,
7H, H1, 3×CH2 OBO), 2.59 (ABX, 2H, H2, J2,1=4.4 Hz,
(NHCꢁO), 131.6 (CH2ꢁC
6 H), 118.7 (C6 H2ꢁCH), 79.8 (Cq
J2%,1=9.1 Hz), 0.69 (s, 3H, CH3 OBO), 0.16 (s, 9H,
tBu), 77.9 (C1%), 74.1 (C5%), 72.1 (C3%), 69.3 (C2%), 67.7 (C4%),
3×CH3 TMS). 13C NMR (50 MHz, CDCl3): l 147.1
(C-NO2), 136.6 (C-SO2), 132.7, 132.4, 131.0, 124.8 (4×
CHarom oNs), 107.3 (Cq OBO), 102.4, 87.2 (2×Cq alkyne),
72.4 (3×CH2 OBO), 57.2 (C1), 30.4 (Cq OBO), 22.3 (C2),
14.0 (CH3 OBO), −0.1 (3×CH3 TMS). MS (ESI) m/z
477.1 [M+Na]+.
65.7 (OC6 H2 allyl), 61.5 (C6%), 53.4 (C2), 32.3 (C3), 30.8
(C5), 28.3 (3×CH3 tBu), 21.1 (C4), 20.7, 20.7, 20.6, 20.5
(4×CH3 Ac). MS (ESI) m/z 588.0 [M+H]+, 610.0 [M+
Na]+.
9. Skrydstrup, T.; Mazeas, D.; Elmouchir, M.; Doisneau,
G.; Riche, C. Chem. Eur. J. 1997, 3, 1342–1356.
10. Lowary, T.; Meldal, M.; Helmboldt, A.; Vasella, A.;
Bock, K. J. Org. Chem. 1998, 63, 9657–9668.
11. Maligres, P. E.; See, M. M.; Askin, D.; Reider, P. J.
Tetrahedron Lett. 1997, 38, 5253–5256.
(9S)-9-(4-Methyl-2,6,7-trioxa-bicyclo[2.2.2]oct-1-yl)-9-(2-
nitrobenzenesulfonylamino)-6,7,8,9-tetradeoxy-2,3,4-tri-
O-benzyl-a-D
-gluco-non-6-ynopyranoside 7f: 1H NMR
(400 MHz, CDCl3): l 8.13–8.10 (m, 1H, oNs), 7.90–7.88
(m, 1H, oNs), 7.67–7.59 (m, 2H, oNs), 7.41–7.24 (m,
15H, 3×Harom Bn), 5.92 (d, 1H, NH, JNH,9=9.5 Hz),
4.94, 4.84, 4.71 (3×AB, 3×2H, 3×CH2 Bn), 4.55 (d, 1H,
12. Dondoni, A.; Marra, A.; Massi, A. J. Org. Chem. 1999,
64, 933–944.