Asymmetric synthesis of Boc-β2-homophenylglycine

Article abstract of DOI:10.1016/S0957-4166(98)00036-6

Enantiomerically pure Boc-β²-homophenylglycine has been prepared in five steps starting from Oppolzer's sultam. The key step of this route is the acylation of metalated phenylacetonitrile with sultam carbonyl chloride. Subsequent reactions (reduction, N-Boc protection, oxidation of the Boc- amino alcohol) led to Boc-(S)-β²-HPhg starting from (+)-sultam.

Full text of DOI:10.1016/S0957-4166(98)00036-6

Pergamon
Tetrahedron: Asymmetry 9 (1998) 865–871
2
Asymmetric synthesis of Boc-β -homophenylglycine
∗
Rachel Ponsinet, Gérard Chassaing and Solange Lavielle
Laboratoire de Chimie Organique Biologique, Université Paris VI, CNRS UMR 7613, 4 Place Jussieu, 75252 Paris Cedex 05,
France
Received 15 January 1998; accepted 27 January 1998
Abstract
2
Enantiomerically pure Boc-β -homophenylglycine has been prepared in five steps starting from Oppolzer’s
sultam. The key step of this route is the acylation of metalated phenylacetonitrile with sultam carbonyl chloride.
2
Subsequent reactions (reduction, N-Boc protection, oxidation of the Boc-amino alcohol) led to Boc-(S)-β -HPhg
starting from (+)-sultam. © 1998 Elsevier Science Ltd. All rights reserved.
1. Introduction
In the past few years, there has been increasing interest in non-proteinogenic amino acids for mainly
two reasons: as secondary-structure inducers (i.e. constrained analogues of α-amino acids) and more
recently as starting materials for the syntheses of scaffolds in combinatorial chemistry. β-Amino acids,
1
far less abundant than α-amino acids, are however present in natural products. β-Amino α-hydroxy
2
amino acids have deserved special interest as enzyme inhibitors, since the discovery of bestatin. With the
exception of β-alanine, β-amino acids have rarely been incorporated into peptides for structure–activity
3
relationships. Indeed, current strategy for such an analysis is to constrain conformational mobility of
4
peptides and the extra methylene group in β-amino acids should confer an extra flexibility to peptides
5
6
incorporating β-amino acids. The recent work of two Swiss groups and Gellman show that this a
priori statement turns out to be incorrect. They have demonstrated that, unexpectedly, β-amino acids
induce highly stable helical structures, i.e. 14-helix or 12-helix for penta-, hexa- or heptamers composed
3
of β-amino acids. The stability of the left-handed helical structure made of β -substituted amino acids
3
(
β that is β to COOH) seemed to be slightly less affected by changing the polarity of the solvent than
2
5c
2
that of the right-handed helix corresponding to a hexamer made of β -substituted amino acids (β that
is α to COOH).
7
Three recent reports review the various strategies developed for the syntheses of these β-amino acids.
Even though diazomethane is the key reagent in the Arndt–Eistert homologation, this procedure is up
∗
Corresponding author. Fax: 33 1 44 27 71 50; e-mail: lavielle@ccr.jussieu.fr
0957-4166/98/$19.00 © 1998 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(98)00036-6

866
R. Ponsinet et al. / Tetrahedron: Asymmetry 9 (1998) 865–871
3
until now the easiest way to prepare enantioselectively β -substituted amino acids starting from the
2
corresponding commercially available α-amino acids. The synthesis of β -substituted amino acids still
7
–9
8
remains a challenge,
usually involving diastereoselective alkylation of an enolate, whose synthesis
2
is the limiting step. Another way of stereoselectively constructing the stereogenic center of a β -amino
acid may come from an electrophilic carboxylation α to a nitrile function. Herein, we report the one-
pot synthesis of an electrophilic sultam derivative 1 and its use in the acylation of phenylacetonitrile
which led selectively to one diastereoisomer 3a. Subsequent reactions (reduction, N-Boc protection and
oxidation) of 3a afforded N-Boc-β -homophenylglycine (Boc-β -HPhg), in five steps with an excellent
enantiomeric excess (99%), Scheme 1.
2
2
Scheme 1.
2. Results and discussion
Both enantiomers of the electrophilic sultam carbonyl chloride 1 were obtained in 80% yield by
10
reacting one enantiomer of sultam with triphosgene in the presence of triethylamine. The crystallized
1
sultam carbonyl chloride 1 was contaminated with small amounts (<10%, as estimated by H-NMR) of
unreacted sultam and triethylammonium chloride salt and used without further purification; it may be
kept at 0°C for two to three months without further decomposition into free sultam. A pure sample of 1
may be obtained after flash chromatography.
Complex TLC profiles and NMR spectra of the crude extracts were observed when this electrophile
was tentatively added to lithiated phenylacetonitrile. These data showed the presence of significant
amounts of sultam in the mixture, whatever the conditions (base, solvents, temperature) used for the
preparation of lithiated phenylacetonitrile i.e.: BuLi or LDA in THF, THF/HMPT or THF/DMSO at
−
78°C, 0°C or room temperature. When metallated phenylacetonitrile was prepared at room temperature
with sodium hydride in DMSO, acylated species 3a and 3b were identified in the crude product by NMR,

R. Ponsinet et al. / Tetrahedron: Asymmetry 9 (1998) 865–871
867
Table 1
as well as sultam which was still the major product. The diastereoisomeric excess 3a:3b was also poor,
entry 1, Table 1. In Table 1 are recorded representative experiments of the attempts performed with (−)-
or (+)-sultam carbonyl chloride 1, to improve both the yield and diastereoselectivity of this acylation
reaction.
We found that the amount of sultam in the crude material decreased by increasing the concentration
of phenylacetonitrile 2 from one to two equivalents (comparison of entries 1 and 2). Addition of THF in
DMSO, which allowed us to conduct the experiment at lower temperature (−10°C), similarly decreased
the amount of sultam in the crude product (comparison of entries 1 and 3, and entries 2 and 4). Increasing
the concentration of 1 (from 0.08 to 0.25 M) did not significantly modify this ratio. Entry 6 shows that
the minimal percentage of sultam expected in the crude product, after extraction, could be reached (10%)
considering the purity of sultam carbonyl chloride 1 (90%). Therefore, in the other experiments, free
sultam (19 to 68%) must not come from partial hydrolysis of 1 in the reaction medium. We might envisage
−
that sultam might be generated via the pathway in Scheme 2, that is S* being the leaving group in the
tetrahedral intermediate. Sultam derivatives may be readily ‘transesterified’ by benzyl alcoholate showing
11
that sultam may be a good leaving group. This result might originate from the pyramidalization of the
nitrogen atom in cyclic sultam as observed by X-ray.12
The diastereoisomeric excess 3a:3b was independent of the solvent, temperature, concentration and
also of the equivalent amounts of phenylacetonitrile 2. The differences in 3a:3b ratios reflected in fact
kinetic reprotonation (using independently one or two equivalents of acetic acid or aqueous NH Cl)
4

868
R. Ponsinet et al. / Tetrahedron: Asymmetry 9 (1998) 865–871
Scheme 2.
versus thermodynamic equilibrium of the acylated species. If the reaction mixture is not rapidly extracted
after quenching and/or the medium buffered with NH Cl, the acylated products equilibrated to a 25:75
4
1
ratio (thermodynamic products, as established by H-NMR with a sample of pure 3a in CDCl with trace
3
amounts of SiO ). The acidity of the remaining Hα in 3a precluded purification by flash chromatography
2
1
on silica gel; 3a was obtained pure, with no contamination of 3b as ascertained by H-NMR, in acceptable
yield after a single crystallization of the crude material obtained after extraction and washings (50% yield
for two steps, calculated from starting sultam).
In order to avoid equilibration of 3a in either acidic or basic conditions required for the reduction of the
nitrile function or the deprotection of the sultam, respectively, 3a was reduced by LiAlH into the amino
4
alcohol which was not isolated but N-Boc protected with ditertbutyldicarbonate. This N-Boc-protected
β-amino alcohol was oxidized by pyridinium dichromate (3.5 equiv.) in dimethylformamide to N-Boc-3-
2
2
amino-2-phenylpropionic acid 4, that is N-Boc-β -homophenylglycine (Boc-β -HPhg); 4 was obtained
with an overall yield for these three steps of 20%. The absolute configuration and the enantiomeric excess
of 4 were attributed by comparison of its specific rotation with literature data. Compound (S)-N-Boc-
β -HPhg was obtained starting from (+)-sultam with excellent enantioselectivity (specific rotation [α]
86 (c 1.25, CHCl ) compared to +88 for (R)-Boc-β -HPhg, whose enantiomeric excess was ascertained
9
2
20
D
2
−
3
9
2
by chiral HPLC ). Conversely, (−)-sultam yielded (R)-N-Boc-β -HPhg.
These data demonstrated firstly that reduction of 3a by LiAlH into the corresponding amino alcohol
4
did not alter the stereochemical integrity of the initially pure diastereoisomer 3a, and secondly allowed
the orientation of the kinetic reprotonation of the enolate to be determined. Interestingly, identical results
13
have been previously obtained with sultam-derived aldimines, reprotonation (NaH/DMSO, acetic acid,
14
room temperature) of these Schiff bases resulted in the formation of (S)-amino acid starting from (+)-
sultam and (R)-amino acid starting from (−)-sultam. This stereochemistry was in contrast to that found
15
13
for the alkylation of these ketimine or aldimine derivatives which is a kinetically controlled (si)-
approach of a (Z)-sultam-enolate. With these aldimines the diastereoisomeric ratio of the reprotonation
14
reaction never exceeded 20:80 (in contrast, alkylations were highly diastereoselective). In the present
study, the enolate derived from sultam-phenylacetonitrile led to a high diastereoisomeric ratio 95:5,
whatever the quenching agent used (NH Cl, or equivalent amount of an acid such as acetic acid or
4
2,5-ditertbutylphenol). This high diastereoselectivity is however restricted to the particular structure of
the sultam-phenylacetonitrile enolate, since attempts to use other nitriles such as phenylpropionitrile,
propionitrile, 3,3-dimethylpropionitrile were unsuccessful. In all cases, both the diastereoselectivity and
the yields of acylated species versus recovered free sultam were rather low (around 65:35 and <30%
in the crude material, respectively). The lower acidity of the Hα with these nitriles required the use of
lithiated base (BuLi or LDA) and as for phenylacetonitrile (see above) the presence of large amounts of

R. Ponsinet et al. / Tetrahedron: Asymmetry 9 (1998) 865–871
869
free sultam (40 to 50%), besides degradation products coming from both nitrile and THF, were detected
by NMR and TLC of the crude material.
2
2
We must emphasize that this strategy allowed us to prepare Boc-β -HPhg. Other β -substituted amino
acids must be obtained by alkylation of judicious sultam-derived enolate(s). We are currently examining
the alkylation of new sultam-derived precursor(s), a strategy which will be inappropriate for obtaining
2
β -homophenylglycine.
3. Experimental
Melting points were determined with a Reichert Heizbank (system Kofler) apparatus and are uncor-
rected. Optical rotations were determined at 20°C on a Perkin–Elmer Model 141 polarimeter equipped
with a 10 cm path length cell. TLC was performed on precoated silica gel plates (Merck 60F, 0.25 mm
thick). The spots were detected by UV, molybdophosphoric acid oxidation or ninhydrin reagent. Silica
1
gel (0.040–0.063 mm, 60F) supplied by Merck was used for flash chromatography. H-NMR spectra
were recorded on Brucker ARX-400 or AC-200 instruments. Chemical shifts are reported in δ (ppm)
with residual chloroform as an internal reference. Microanalyses and mass spectra were obtained from
the University Pierre et Marie Curie.
Tetrahydrofuran (THF) was distilled from sodium benzophenone and kept over 4 Å molecular sieves,
dimethylsulfoxide (DMSO) was distilled from CaH and then dried by azeotropic distillation of toluene
2
prior to the experiment. Anhydrous reaction were performed under an argon atmosphere, glassware was
flame-dried prior to use under a stream of nitrogen. The course of the reactions was monitored by TLC.
3
.1. (−)-10,10-Dimethyl-3,3-dioxo-3-thia-4-aza-tricyclo[5.2.1.01,5] decane-4-carbonyl chloride 1
To a solution of (1S)-(−)-2,10-sultam, previously dried by azeotropic distillation of toluene (3.7 g; 17.2
mmoles), in CH Cl (70 mL), were added first NEt (2 equiv., 5 mL; 36 mmoles) and after cooling at 4°C,
2
2
3
triphosgene, (CCl O) CO, (0.37 equiv., 1.9 g; 6.4 mmoles). After 16 hours stirring at room temperature,
3
2
the solution was concentrated in vacuo. The residual yellow solid was suspended in Et O (50 mL), this
2
suspension was stirred vigorously for one hour and then filtered, the solid was washed with Et O. The
2
combined filtrates were concentrated to afford 1 as a slightly yellow powder (3.76 g, 79% yield), which
is kept under nitrogen at 4°C; m.p. 119–120°C; [α]D20 −133 (c 1, CHCl ); MNH : 295; H-NMR (400
+
1
3
4
MHz, CDCl ): δ 3.96 (q, 1H), 3.55 (q, 2H), 2.4–1.9 (m, 5H), 1.5–1.3 (m, 2H), 1.22 (s, 3H), 1.02 (s, 3H);
3
trace amounts of sultam (7%), characteristic signals, δ 4.0 (NH), 3.08 (AB, 2Hα to SO ), 1.11 and 0.93
2
(
two s, CH ); After purification by flash chromatography a pure sample may be obtained. Anal. calcd for
3
C H NO SCl: C, 47.57; H, 5.76; N, 5.04. Found: C, 47.65; H, 5.77; N, 4.97
11
16
3
3
.2. Acylation procedure: 3-(10,10-dimethyl-3,3-dioxo-3-thia-4-aza-tricyclo[5.2.1.01,5 ] dec-4-yl)-3-
oxo-2-phenylpropionitrile 3a+3b
A flame-dried 100 mL round-bottomed flask was capped with a septum, purged with argon, charged
with DMSO:THF (40 mL, 3:1) containing a crystal of o-phenanthroline and phenyl propionitrile (18
mmoles, 2.1 mL). After addition of sodium hydride (19.8 mmoles, 0.6 g) under an argon stream and four
hours stirring at room temperature, the flask was cooled at −10°C and the sultam carbonyl chloride 1
(9 mmoles, 2.47 g) was added under an argon stream. Acetic acid (10 mmoles, 0.6 mL) was added after
two hours stirring at −10°C and then ethyl acetate. The organic layer was washed with saturated aqueous

870
R. Ponsinet et al. / Tetrahedron: Asymmetry 9 (1998) 865–871
NH Cl (3×) and the combined aqueous layers were extracted (2×) with CH Cl . The combined organic
4
2
2
solutions were dried (MgSO ) and concentrated in vacuo to yield an orange oil (crude product) which
4
was precipitated by adding Et O; filtration led to a slightly yellow solid 3a (1.6 g, 50% yield); R : 0.64
2
f
20
+
1
(
CH Cl –MeOH 1%); m.p. 203–204°C; [α]
−90 (c 1, CHCl ); MNH : 376; H-NMR (400 MHz,
2
2
D
3
4
CDCl ): δ 7.57 (m, 2H), 7.40 (m, 3H), 5.46 (s, 1H), 3.97 (m, 1H), 3.54 (s, 2H), 2.01–1.69 (m, 5H),
3
1
.50–1.30 (m, 2H), 0.94 (s, 3H, CH ), 0.81 (s, 3H, CH ). Anal. calcd for C H N O S: C, 63.69; H,
3
3
19 22
2
3
6.14; N, 7.82. Found: C, 63.57; H, 6.17; N, 7.68.
3
.3. Reduction by LiAlH , N-Boc protection and oxidation: (−)-(S)-(N-tertbutyloxycarbonyl)-3-amino-
4
2
2
2-phenylpropanoic acid, (Boc-β homophenylglycine, Boc-β -HPhg) 4
To a commercial solution of LiAlH (8.9 mmoles, 8.9 mL) in THF was added AlCl (8.8 mmoles, 1.17
4
3
g) and then dropwise at 0°C a solution of 3a (8 mmoles, 2.86 g) in 2-methyltetrahydrofuran (90 mL).
After four hours stirring at room temperature and acidification (1 N HCl), extraction was performed
with ethylacetate (3×). The organic layers containing recovered sultam were washed with 1 N HCl, the
aqueous layer being reextracted (3×) with ethyl acetate.
The combined aqueous layers (100 mL) were brought to pH 8–9 with 10 N NaOH. After addition of
dioxane (50 mL) and cooling at 0°C the resulting amino alcohol was reacted with ditertbutyldicarbonate
(16.5 mmoles, 3.6 g) first at 0°C (1 hour) and then for 16 hours at room temperature. The resulting white
powder obtained after concentration in vacuo was suspended with a saturated solution of Rochelle salt
(
potassium sodium tartarate tetrahydrate) which was extracted with ethyl acetate (3×). The organic layers
were dried (MgSO ) and concentrated in vacuo.
4
The resulting oil dissolved in dimethylformamide (30 mL) was directly reacted, in the dark, with
pyridium dichromate (3.5 equiv., 38 mmoles, 14.4 g). After 14 hours stirring, saturated aqueous NH Cl
4
(
200 mL) was added. The reaction mixture was extracted with diethyl ether (3×). The organic layers
were washed with 1 N NaOH (3×) and then the aqueous layers was acidified to pH 2 with 1 N HCl in
the presence of CH Cl . After extraction with CH Cl (2×), the combined dichloromethane extractions
2
2
2
2
were washed (NH Cl, ×3), dried (MgSO ) and concentrated in vacuo, yielding a slightly yellow powder
4
4
which was purified by flash chromatography (eluant, CHCl :MeOH:AcOH=95:5:0.2) leading to 4 as a
2
9
20
white powder (516 mg, 20% yield); m.p. 135–136°C (lit. 144–146°C); [α]
−86 (c 1.25, CHCl ),
D
3
9
20
2
+
1
(
7
lit. [α]
+88 (c 1.25, CHCl ) for (R)-Boc-β -HPhg); MH : 266; H-NMR (400 MHz, CDCl ) δ
D
3
3
.38–7.28 (m, 5H), 7.07 (broad s, 0.5H cis–trans isomerism NH), 5.0 (broad s, 0.5H, cis–trans isomerism
NH), 3.94–3.8 (m, 1H), 3.6–3.5 (m, 2H), 1.5 and 1.44 (two d, cis–trans isomerism Boc). Anal. calcd for
C H O N: C, 63.40; H, 7.17; N, 5.28. Found; C, 63.30; H, 7.08; N, 5.31.
1
4
19
4
Acknowledgements
This work was supported by a grant from ISOCHEM (Vert le Petit, France).
References
1
2
. (a) S. M. Hecht, Acc. Chem. Res. 1986, 19, 383; (b) J. P. Michael, G. Pattenden, Angew. Chem., Int. Ed. Engl. 1993, 32, 1.
. (a) H. Umezawa, T. Aoyagi, H. Suda, M. Hamda, T. Takeuchi, J. Antibiot. 1976, 29, 97; (b) J. Lynas, B. Walker, Biochem.
J. 1997, 325, 609.

R. Ponsinet et al. / Tetrahedron: Asymmetry 9 (1998) 865–871
871
3
. (a) M. A. Ondetti, S. L. Engel, J. Med. Chem. 1975, 18, 761; (b) O. W. Griffith, Ann. Rev. Biochem. 1986, 55, 855; (c) M.
Rodriguez, P. Fulcrand, J. Laur, A. Aumelas, J. P. Bali, J. Martinez, J. Med. Chem. 1989, 32, 522; (d) B. Bozu, F. Fulop, G.
K. Toth, G. Toth, M. Szucs, Neuropeptides 1997, 31, 367.
4. (a) V. J. Hruby, In Peptides: Chemistry, Structure and Biology, Eds: R. S. Hodges, J. A. Smith; ESCOM: Leiden, 1994, pp.
3; (b) K. H. Hsieh, T. R. Lanhann, R. C. Speth, J. Med. Chem. 1989, 32, 898; (c) H. I. Mosberg, H. B. Kroona, J. Med.
Chem. 1992, 35, 2384; (d) H. Josien, S. Lavielle, A. Brunissen, M. Saffroy, Y. Torrens, J. C. Beaujouan, J. Glowinski, G.
Chassaing, J. Med. Chem. 1994, 37, 1586.
5. (a) D. Seebach, M. Overhand, F. N. M. Kühnle, B. Martinoni, L. Oberer, U. Hommel, H. Widmer, Helv. Chim. Acta 1996,
79, 913; (b) D. Seebach, P. E. Ciceri, M. Overhand, B. Jaun, D. Rigo, L. Oberer, U. Hommel, R. Amstutz, H. Widmer, Helv.
Chim. Acta 1996, 79, 2043; (c) T. Hintermann, D. Seebach, Synlett 1997, 437.
6
. (a) G. P. Dado, S. H. Gellman, J. Am. Chem. Soc. 1994, 116, 1054; (b) D. H. Appella, L. A. Christianson, I. Karle, D. R.
Powell, S. H. Gellman, J. Am. Chem. Soc. 1996, 118, 13071; (c) D. H. Appella, L. A. Christianson, D. A. Klein, D. R.
Powell, X. Huang, J. A. Barchi Jr., S. H. Gellman, Nature 1997, 387, 381.
7. (a) D. C. Cole, Tetrahedron 1994, 50, 9517; (b) E. Juaristi, D. Quintana, J. Escalante, Aldrichimica Acta 1994, 27, 3; (c) G.
Cardillo, C. Tomasini, Chem. Soc. Rev. 1996, 117.
8. (a) E. Juaristi, D. Quintana, B. Lamatsch, D. Seebach, J. Org. Chem. 1991, 56, 2553; (b) K. Burgess, L. T. Liu, B. Pal, J.
Org. Chem. 1993, 58, 4758; (c) E. Juaristi, D. Quintana, M. Balderas, E. Garcia-Perez, Tetrahedron: Asymmetry 1996, 7,
2233.
9. J. F. Bower, J. M. J. Williams, Synlett 1996, 685.
10. Reviews: (a) W. Oppolzer, Pure Appl. Chem. 1990, 62, 1241; (b) W. Oppolzer, Tetrahedron 1987, 43, 1969; (c) B. H. Kim,
D. P. Curran, Tetrahedron 49, 293.
1
1
1
1
1
1. W. Oppolzer, H. Bienayme, A. Genevois-Borella, J. Am. Chem. Soc. 1991, 113, 9660.
2. W. Oppolzer, C. Chapuis, G. Bernardinelli, Helv. Chim. Acta 1984, 67, 1397.
3. M. Ayoub, G. Chassaing, A. Loffet, S. Lavielle, Tetrahedron Lett. 1995, 36, 4069.
4. M. Ayoub, PhD Dissertation, Université P. et M. Curie, Paris, France, 1996.
5. (a) W. Oppolzer, R. Moretti, S. Thomi, Tetrahedron Lett. 1989, 30, 6009; (b) H. Josien, A. Martin, G. Chassaing,
Tetrahedron Lett. 1991, 32, 6547; (c) H. Josien, G. Chassaing, Tetrahedron: Asymmetry 1992, 3, 1351.