Diastereoselective formation of a quaternary center in a pyroglutamate derivative. Formal synthesis of Monatin

Article abstract of DOI:10.1016/S0040-4039(01)01401-0

In this communication we describe a highly diastereoselective formation of a quaternary center present in the structure of Monatin, a potent sweetening agent isolated from natural sources. The synthesis of a derivative useful for biological studies on the interactions of this type of sweetening substance on taste receptors is described. The strategy relies on the oxidation of an enolate originating from a pyroglutamate derivative followed by a highly diastereoselective alkylation with an electrophile obtained from indole.

Full text of DOI:10.1016/S0040-4039(01)01401-0

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 6793–6796
Diastereoselective formation of a quaternary center in a
pyroglutamate derivative. Formal synthesis of Monatin
Davi de Jesus Oliveira and Fernando Coelho*
DQO, IQ/Unicamp, PO Box 13083-970, Campinas, Sa˜o Paulo, Brazil
Received 23 June 2001; revised 27 July 2001; accepted 30 July 2001
Abstract—In this communication we describe a highly diastereoselective formation of a quaternary center present in the structure
of Monatin, a potent sweetening agent isolated from natural sources. The synthesis of a derivative useful for biological studies on
the interactions of this type of sweetening substance on taste receptors is described. The strategy relies on the oxidation of an
enolate originating from a pyroglutamate derivative followed by a highly diastereoselective alkylation with an electrophile
obtained from indole. © 2001 Elsevier Science Ltd. All rights reserved.
Monatin [1, (2S,4S)-4-hydroxy-4-(indol-3-ylmethyl)-
glutamic acid] is an unusual aminoacid isolated in
South Africa from the roots of Schlerochiton illicifolius
by Vleggaar et al.1
Despite the important biological properties of this
molecule, very few attempts have been made to prepare
it. The syntheses of racemates and an analogue of
Monatin have been reported.2
This interesting non-proteinogenic aminoacid exhibits
an intense sweet taste (1200–1400 times sweeter than
glucose) and is an attractive target to synthesize as a
molecule having potent biological activity.
In a current research program on the relationship
between some functional groups and the sweet taste, it
was necessary to have a Monatin derivative in which
the two nitrogen atoms were protected (2, Scheme 1).
Scheme 1. Retrosynthetic analysis of 1 and 2.
Keywords: Monatin; pyroglutamate; N-Boc-Monatin.
* Corresponding author. Tel.: 55 19 37883085; fax: 55 19 3788-3023; e-mail: coelho@iqm.unicamp.br
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01401-0

6794
D. J. Oli6eira, F. Coelho / Tetrahedron Letters 42 (2001) 6793–6796
in the new stereogenic center was determined using
NOE experiments (Fig. 1).
Our interest was focused on determining the importance
of the interactions of the hydrogens bonded to nitrogen
with taste receptors and the perception of the sweet
taste.
First we irradiated the hydrogen at C5 in order to deter-
mine the spatial orientation of the hydrogens at C4.
Irradiation of the hydrogen at C3 showed an increment
of 2.02% in the b-oriented hydrogen at C4 and an incre-
ment of only 0.13% in the a-oriented hydrogen. These
results confirm that OCBz group is a-oriented.10
Recently the first asymmetric total synthesis of this sub-
stance has been described by Goodman et al.3 This
report encouraged us to disclose our results concerning
a highly diastereoselective formation of a quaternary
center in a pyroglutamate derivative and the synthesis of
N-Boc-N%-Boc Monatin (2), which could be then trans-
formed into Monatin (1) by the procedure of Goodman
et al.3 To achieve our objective, we utilized the strategy
outlined in Scheme 1.
To prepare the quaternary center of 1 and 2 we have
taken advantage of a highly diastereoselective method-
ology for the alkylation of the pyroglutamate derivative
5, recently described.11 Alkylation of the enolate gener-
ated from 5 with N-Boc-3-bromomethylindole (6) at
−78°C in the presence of DMPU or HMPA (10 equiv.)
furnished the intermediate 3 (Scheme 3).12 The bromoin-
dole employed in this step as alkylating agent was pre-
pared from indole in three steps and 76% overall yield
according the procedure described by Scho¨lkopf et al.13
Monatin (1) and its N-Boc derivative (2) contain two
stereogenic centers at C2 and C4. From our point of
view, the pivotal problem in the conception of an
approach to their syntheses relies upon the formation of
the quaternary center at C4. If we consider 1 and 2 as
direct derivatives of glutamic acid, the control of the
absolute configuration at C2 can be secured if we use
glutamic acid or a derivative as starting material in our
strategy. In the retrosynthetic analysis shown, the qua-
ternary center of Monatin (C4) was prepared with a
high degree of diastereoselectivity by the alkylation of
an enolate obtained from the derivative 4, with the elec-
trophile 5, prepared from commercial indole (Scheme
1).
Using chiral HPLC analysis14 it was possible to detect
only one diastereoisomer (Fig. 2).
The stereochemical assignments of 3 were established
by an NOE experiment in deuterobenzene, by the irra-
diation of the hydrogens at C4, C5 and C6 (Fig. 3).
The lactam derivative 4 could be prepared through the
oxidation of an enolate obtained from a pyroglutamate
derivative. Control of the stereochemistry of the asym-
metric center at C4 can be achieved in the alkylation
step and is basically due to the presence of the volumi-
nous silyl substituent at C5, which directs the elec-
trophile preferentially to the a face of the enolate.
Our synthesis begins with the preparation of the lactam
4, which was obtained from
using a standard sequence described in the literature.4,5
Thus, esterification of the -pyroglutamic acid with
L-pyroglutamic acid by
L
SOCl2/CH3OH followed by ester reduction with
NaBH4/CH3OH furnished the alcohol 7 (Scheme 2).
Protection of the hydroxyl group with TBSCl, followed
by nitrogen protection as a carbamate (t-butyloxycar-
bonyl, Boc) afforded the pyroglutamate derivative 9
([h]2D0=−61 (c 1.1, CHCl3) (lit.6 −61 (c 1.1, CHCl3), in
four steps and with an overall yield of 65%.
Scheme 2. Synthesis of the pyrogutamate derivative 5. (a)
SOCl2, MeOH, 24 h, rt, 85%; (b) NaBH4, i-PrOH, 20 h, rt,
86%; (c) imidazol, DMF, TBSCl, 0°C“rt, 24 h, 95%; (d)
(Boc)2O, DMAP, Et3N, 3 h, rt, 100%; (e) (i) LHDMS, −78°C,
THF, 30 min; (ii) (BnOCO)2O, −78°C, 30 min; (iii) AcOH,
−78“0°C, then H2O 0°C“rt, 50%.
Treatment of 9 with lithium bis(hexamethyldisilazide)
(LHDMS) at −78°C, followed by the in situ oxidation of
the enolate with dibenzyldicarbonate peroxide7 fur-
nished 5 in 50% yield [isolated product, purified by silica
gel column chromatography (hexane:ethyl acetate
50:50); 92% yield based on recovered starting material],
as the only detectable diastereoisomer. MoOPh8 and
MoOPD9 were also tried as reagents for this oxidation
step, however both failed to give lactam 5.
Although control of the stereochemistry in this step was
unnecessary, the spatial orientation of the substituents
Figure 1. NOE experiments with 5.

-D. J. Oli6eira, F. Coelho / Tetrahedron Letters 42 (2001) 6793–6796
6795
and the a-oriented hydrogen at C4 have a cis relation-
ship. In order to evaluate the accuracy of these results
we did another experiment. The methylene group at C6
(double duplet centered at l 3.30) was irradiated and an
effect was observed on the hydrogens at C4. We had no
increment in the signal of the b-oriented hydrogen,
however an increment of 1.19% was observed in the
double doublet centered at l 2.42, previously attributed
to the a-oriented hydrogen.
Figure 2. HPLC of 3.14
At this stage of the work we have developed a highly
diastereoselective methodology to generate a quater-
nary center in a pyroglutamate derivative. To follow
our planned synthetic strategy it was necessary to
remove the protective groups and cleave the lactam
ring. Treatment of lactam 3 with LiOH in a mixture of
THF:H2O15 gave 10 in 97% yield. Under these condi-
tions the CBz protective group was also removed
(Scheme 4).
This experiment was similar to the one on intermediate
5. First we determined the orientation of the hydrogen
at C4 by irradiation of the C5 hydrogen, then we
separately irradiated the a- and b-oriented hydrogens.
The b-oriented hydrogen at C4 showed no increment
(0%) in the absorption of the benzylic CH2 (C6), how-
ever the a-oriented hydrogen showed an increment of
2.81%. These results indicated that the indolic moiety
Scheme 3. Formation of the quaternary center. (a) i. LHDMS, THF, −78°C, 30 min; ii. DMPU (10 equiv.), 78°C, 30 min; iii. 6,
−78°C, 3 h, 75%.
Figure 3. NOESY experiments with 3.
Scheme 4. Synthesis of 2. (a) LiOH, THF:H2O (5:1), rt, 3 h, 97%; (b) Jones, 0°C, acetone, 30 min 65%.

6796
D. J. Oli6eira, F. Coelho / Tetrahedron Letters 42 (2001) 6793–6796
To remove the silyl group and transform the free alcohol
into the carboxylic acid, intermediate 10 was treated with
Jones reagent in acetone, at 0°C, according to a known
procedure,16 to furnish the aminoacid 2, in 65% yield
(Scheme 4). From our point of view, this aminoacid (2)17
should be readily transformed in Monatin, using known
methodologies, such as that described by Goodman et al.3
2. (a) Holzappfel, C. W.; Bischofberger, K.; Olivier, J.
Synth. Commun. 1994, 24, 3197–3211; (b) Abushanab,
E.; Arumugan, S. US Patent 5,994,559, November 30,
1999; (c) Holzapfel, C. W.; Olivier, J. Synth. Commun.
1993, 23, 2511–2526.
3. Nakamura, K.; Baker, T. J.; Goodman, M. Org. Lett.
2000, 2, 2967–2970.
4. Saijo, S.; Wada, M.; Himizu, J.-I.; Ishida, A. Chem.
Pharm. Bull. 1980, 28, 1449–1458.
5. Ackermann, J.; Mathes, M.; Tamm, C. Helv. Chim.
Acta 1990, 73, 122–132.
6. Ohfune, Y.; Tomita, M. J. Am. Chem. Soc. 1982, 104,
3511–3515.
7. Gore, M. P.; Vederas, J. C. J. Org. Chem. 1986, 51,
3700–3704.
In conclusion, we have developed a highly diastereoselec-
tive methodology to generate a quaternary center in a
pyroglutamate derivative. This strategy has culminated
with the preparation of our target compound 2. Since the
methodology to remove the N-Boc group is already
known,3 from our point of view, we have also made a
formal synthesis of Monatin (1).
8. Vedejs, E.; Engler, D. A.; Telschow, J. E. J. Org. Chem.
1978, 34, 188–196.
A simple modification in the alkylation agent used in the
formation of the quaternary center permits access to other
compounds of this class, this methodology seems to be
versatile. Biological tests using aminoacid 2 are ongoing
in our laboratory.
9. Anderson, J. C.; Smith, S. C. Synlett 1990, 107–
108.
10. [h]2D0 +76.5 (c, 0.01 g/ml, CHCl3), HRMS calcd for
479.2339; found 479.2335.
11. Oliveira, D. J.; Coelho, F. Synth. Commun. 2000, 30,
2533–2543.
Supplementary material
The result of the HPLC analysis, the 500 MHz 1H NMR
spectra and all spectra obtained in the NOE experiments
with compound 3 are available.
12. [h]2D0 +83.0 (c, 0.08 g/ml, CHCl3), HRMS calcd for
708.3442; found 708.3439.
13. Scho¨llkopf, U.; Lonsky, R.; Lehr, P. Liebigs Ann. Chem.
1985, 413–417.
14. Chrompack-CP-Chirasil-Dex CB column, ethyl ace-
tate:hexane as mobile phase.
Acknowledgements
15. (a) Oba, M.; Terauchi, A. M.; Kamo, H.; Nishiyama,
K. Tetrahedron Lett. 1998, 39, 1595–1598; (b) Yoda, H.;
Oguchi, T.; Takabe, K. Tetrahedron: Asymmetry 1996, 7,
213–2116.
We thank CNPq for fellowships to D.J.O. and F.C.
(301369/87-9) and Fapesp for a grant (c 96/5710-9).
16. Mulzer, J.; Meier, A.; Buschamann, J.; Luger, P. J. Org.
Chem. 1996, 61, 566–572.
References
17. All the spectral data are compatible for 2. [h]2D0 −9.8 (c,
0.01 g/ml, CH3OH); HRMS (M+) calcd for C24H32N2O9
492.2106; found 492.2103.
1. Vleggaar, R.; Ackerman, L. G.; Steyn, P. S. J. Chem.
Soc., Perkin. Trans. 1 1992, 22, 3095–3098.