Synthesis of functionalized pyroglutamic acids, part 2: The stereoselective condensation of multifunctional groups with chiral levulinic acids

Article abstract of DOI:10.1055/s-2008-1078211

A general procedure to access 3-hydroxy-4-oxopentanoic acid derivatives is described. A key feature is an aldol reaction with an enal as a masked pyruvic aldehyde. Chiral levulinic acid derivatives are provided as precursors for isocyanide-mediated condensation of multifunctional groups, which affords functionalized pyroglutamic acids. The stereoselectivity in the Ugi 4C-3C reaction with the chiral keto acids is examined. Georg Thieme Verlag Stuttgart.

Full text of DOI:10.1055/s-2008-1078211

LETTER
2249
Synthesis of Functionalized Pyroglutamic Acids, Part 2: The Stereoselective
Condensation of Multifunctional Groups with Chiral Levulinic Acids
S
ynthesis of Func
y
tionalized P
n
yroglutamic
t
A
cids:
h
Par
t
2
ia B. Gilley, Matthew J. Buller, Yoshihisa Kobayashi*
Department of Chemistry and Biochemistry, University of California, San Diego,
9500 Gilman Drive, Mail Code 0343, La Jolla, CA 92093-0343, USA
E-mail: ykoba@chem.ucsd.edu
Received 2 May 2008
densation of levulinic acid and a variety of aldehydes fol-
Abstract: A general procedure to access 3-hydroxy-4-oxopenta-
noic acid derivatives is described. A key feature is an aldol reaction
with an enal as a masked pyruvic aldehyde. Chiral levulinic acid de-
rivatives are provided as precursors for isocyanide-mediated con-
lowed by hydrogenation of the resulting enone or enoic
acid.4 Nevertheless, we could not find a general procedure
to obtain 3-hydroxy-4-oxopentanoic acid derivatives.
densation of multifunctional groups, which affords functionalized Therefore, the application of Ugi 4C-3C reaction with a
pyroglutamic acids. The stereoselectivity in the Ugi 4C-3C reaction
with the chiral keto acids is examined.
levulinic acid derivative was unknown to prepare a func-
tionalized pyroglutamic acid. The hydroxy group could be
Key words: pyroglutamic acid, Ugi reaction, levulinic acid, aldol
reaction, Amadori rearrangement
important as a directing group for stereocontrol and also
allow us to tune the stereoselectivity by protection, and
creation of diversity by post-Ugi modification of the un-
protected hydroxy group.
In the preceding Letter, we demonstrated a stereoselective
Ugi 4-center 3-component condensation (4C-3C) reaction
of 3-hydroxy-4-oxopentanoic acid (1) with a convertible
isocyanide to afford functionalized pyroglutamic acid de-
rivatives.1 Levulinic acid (4-oxopentanoic acid) deriva-
tives could be potentially useful precursors to prepare
both functionalized g-lactams and g-lactones by isocya-
nide-mediated multifunctional group condensation reac-
tions.2
In this Letter, we report the synthesis of a series of 3-hy-
droxy-4-oxopentanoic acid derivatives by a unified aldol
strategy with an enal as a masked pyruvic aldehyde, and
their stereoselectivity in the Ugi 4C-3C reaction.
We decided to introduce a methyl group at C2 or C3 posi-
tion of 3-hydroxy-4-oxopentanoic acid (see compounds
2–5 in Figure 1) as a representative substitute of a levulin-
ic acid derivative. While the synthesis of 1 has been re-
ported in the literature by others previously,5 our own
efficient synthesis of 1 is reported in the preceding paper
featuring aldol reaction with enal.1 Now, we explore the
validity of our strategy for the synthesis of keto acids 2–5.
O
Me
O
Me
OH
HO
HO
O
O
O
levulinic acid
1
hydrogenolysis
O
Me
OH
O
Me
OH
of
HO
BnO
O
Me
O
Me
OH
benzyl ester
HO
HO
O
O
R2
R2
O
O
OH
R1
R1
Me
Me
A
B
2
3
ozonolysis
O
Me
OH
O
Me
OH
HO
HO
R
R
O
Me
OH
Me
Me
O
stereoselective
aldol with enal
Me Me
X
X
4
5
O
O
R2
R2
Figure 1 Structures of levulinic acid derivatives 1–5
R1
R1
E
D
C
Levulinic acid and its ethyl ester are commercially avail-
able and 483 derivatives of the acid, bearing an alkyl sub-
stituent at C2 or C3, are known in the literature.3 Most of
the derivatives are conveniently synthesized by aldol con-
Scheme 1 Retrosynthetic analysis of b-hydroxy-g-keto acid
Scheme 1 shows the unified retrosynthetic analysis of 3-
hydroxy-4-oxopentanoic acid derivatives 1–5. Levulinic
acid A is an interesting and challenging compound due to
its structural character of four different oxidation states in
the five-carbon frame. Importantly, the last step to reach
acid A required the use of hydrogenolysis of benzyl ester
SYNLETT 2008, No. 15, pp 2249–2252
1
5
.0
9
.2
0
0
8
Advanced online publication: 28.08.2008
DOI: 10.1055/s-2008-1078211; Art ID: S04308ST
© Georg Thieme Verlag Stuttgart · New York

2250
C. B. Gilley et al.
LETTER
Ph
Ph
MgCl2 (0.2 equiv)
TMSCl (1.5 equiv)
Et3N (2.0 equiv)
Me
Me
OH
O
O
O
O
+
CHO
(1.2 equiv)
Me
O
N
O
N
EtOAc, r.t.
cat. TFA
MeOH
Me
7
Bn
Bn
88%
8
6
2. BnBr (1.1 equiv)
cat. TBAI
excess K2CO3
DMF, 92%
1. H2O2 (20 equiv)
LiOH (3 equiv)
THF– H2O (9:1)
Ph
1. O3, MeOH
–78 °C; DMS
87%
Me
OH
O
Me
OH
BnO
O
HO
O
2. H2, Pd/C
MeOH, 100%
Me
Me
9
2
Scheme 2 Asymmetric synthesis of 2
On the other hand, (2S,3S)-3-hydroxy-2-methyl-4-oxo-
pentanoic acid (3) was prepared by following a similar
procedure used for the synthesis of 2 starting from boron-
mediated stereoselective aldol reaction7 with oxazolidi-
none 6 and enal 7 to furnish adduct 10 (Scheme 3).8 Com-
pound 3 was obtained in 70% yield from 10 over four
steps including the conversion of 10 to benzyl ester (79%,
2 steps) followed by ozonolysis of the benzylidene (88%)
and hydrogenolysis of the benzyl ester (quant.).
Ph
Ph
Me
Me
OH
Bu2BOTf (1.5 equiv)
Et3N (1.6 equiv)
O
O
+
O
O
CHO
O
N
O
N
(1.2 equiv)
CH2Cl2, 0 °C;
aldehyde 7
–78 °C to 0 °C
Me
Me
7
Bn
10
Bn
6
43%
Scheme 3 Aldol reaction for asymmetric synthesis of 3
Next, compound 4, bearing two additional methyl groups
at the C2 position of compound 1, was prepared as follows
(Scheme 4). Aldol reaction of the lithium enolate of ben-
zyl isobutyrate 11 and enal 7 gave adduct 12 as a racemic
mixture. Again, ozonolysis of benzylidene 12 followed by
hydrogenolysis of benzyl ester 13 afforded ( )-3-hy-
droxy-2,2-dimethyl-4-oxopentanoic acid (4).9
B, due to facile elimination of the hydroxy group under
both acidic and basic conditions. Ester B could be directly
prepared by aldol reaction of pyruvic aldehyde, but we ex-
pected difficulty with the aldol reaction. Instead, we
planned the installation of the carbonyl group by ozonol-
ysis of olefin C. Therefore, the starting materials for the
aldol reaction were reduced to ester D and enal (or enone)
E as a synthetic equivalent of pyruvic aldehyde. Conve-
niently, commercially available a-methylcinnamaldehyde
was chosen as E (R = Ph, R2 = H), because cinnamalde-
hyde is much more stable than acrylaldehyde. One might
imagine a variety of cinnamaldehyde derivatives with a
different substituent at the a-position could be applicable
in the synthesis. The well-developed aldol reactions might
secure the stereoselective preparation of those levulinic
acid derivatives.
Lastly, compound 5, bearing an additional methyl group
at the C3 position of compound 1, was prepared as a race-
mic mixture by aldol reaction of the lithium enolate of
Ph
Ph
Me
LDA
(1.8 equiv)
Me
+
BnO
BnO
CHO
Me
THF
O
OH
O
7
–78 °C
Me
Me
96%
Diastereomers 2 and 3, bearing an additional methyl
group at C2 position of compound 1, were prepared in
enantiomerically pure form by use of a chiral auxiliary in
the aldol reaction with a-methylcinnamaldehyde (7).
Evans’ Mg-mediated stereoselective aldol reaction6 of ox-
azolidinone 6 with enal 7 cleanly gave adduct 8 as the sole
product (Scheme 2). The acyloxazolidinone 8 was con-
verted to benzyl ester 9 under standard conditions. Ozo-
nolysis of the benzylidene followed by hydrogenolysis of
the benzyl ester afforded (2R,3S)-3-hydroxy-2-methyl-4-
oxopentanoic acid (2).
Me
(1.8 equiv)
11
12
O3, MeOH
–78 °C; DMS
55%
BnO
H2
O
O
Me
OH
Me
HO
Pd/C
MeOH
90%
O
O
OH
Me
Me
Me
Me
13
4
Scheme 4 Racemic synthesis of ( )-4
Synlett 2008, No. 15, 2249–2252 © Thieme Stuttgart · New York

LETTER
Synthesis of Functionalized Pyroglutamic Acids: Part 2
2251
MeO
MeO
Table 1 The Stereoselectivity in Ugi 4C-3C Reaction with b-Hy-
droxy-g-keto Acids [Ar = (2,2-dimethoxyethyl)phenyl]
LHMDS
Me
O
OMe
OMe
(2.0 equiv)
Me
BnO
+
BnO
Me
THF
–78 °C
–
+
O
OH
C N
O
15
PMB
Me
O
Me
Me
R1
OMe
OMe
PMBNH2
O
Me
OH
N
(2.0 equiv)
14
16
N
18
HO
O
H
OMe
OMe
R3
99%
CSA (1.0 equiv)
(2 steps) THF–H2O (4:1)
OH
O
CF3CH2OH
60°C
R1
R2
R3
R2
H2
O
O
Me
Me
OH
Entry Keto acid
Yield
dr (anti/syn)
Product
(shown anti)
HO
Pd/C
BnO
MeOH
96%
O
O
OH
Me
Me
O
PMB
O
Me
OH
Me
5
17
HO
Ar
Ar
N
65%
2:1
N
1
O
1
O
H
Scheme 5 Racemic synthesis of ( )-5
OH
O
19
benzyl acetate 14 and 3,3-dimethoxybutan-2-one (15), in-
stead of enone E (R = Ph, R2 = Me) due to availability
(Scheme 5). This was followed by acid-catalyzed depro-
tection of the dimethyl ketal 16 and hydrogenolysis of
benzyl ester 17 to afford ( )-3-hydroxy-3-methyl-4-oxo-
pentanoic acid (5).
PMB
O
Me
OH
Me
HO
N
N
H
74%
2:1
2
O
O
OH
O
Me
Me
2
20
PMB
A series of chiral b-hydroxy-g-keto acids 2–5 was effi-
ciently prepared by using a unified strategy of aldol reac-
tion with enal as a masked pyruvic aldehyde. The
feasibility was proven even for their enantioselective syn-
thesis.
O
Me
OH
Me
HO
Ar
Ar
N
N
H
73%
3:2
O
3
O
OH
O
Me Me
Me
Me
4
21
Then, we investigated the effect of additional alkyl sub-
stituents at the 2- or 3-position of keto acid 1 on the ste-
reoselectivity in the Ugi 4C-3C reaction (Table 1). As
described in the preceding Letter, 3-hydroxy-4-oxopen-
tanoic acid (1) showed a modest selectivity (anti/syn =
2:1) in the Ugi reaction to give 19. Now, compound 2
bearing an additional methyl group at C2 of 1, compound
4 bearing two additional methyl groups at C2, and com-
pound 5 bearing an additional methyl group at C3 were
subjected to the Ugi 4C-3C reaction with PMBNH2 and
isocyanide 18 (Table 1).10 The three keto acids 2, 4, and 5
afforded the desired Ugi products 20–22 in good yields
(73–84%) with a modest anti selectivity (up to 2:1). The
relative stereochemistry of each Ugi product was deter-
mined by chemoselective formation of syn methyl ester
via b-lactone as described for compound 1 in the preced-
ing paper.1
PMB
O
Me
OH
Me
Me
HO
N
N
H
84%
2:1
O
4
O
Me
OH
5
22
–
+
PMB
HN
C N
O
Me
OH
Me
O
OMe
OMe
O
18
HO
PMBNH2
CF3CH2OH
60 °C
Me
Me
23
3
68%
CO2
PMB
PMB
HN
Amadori
rearrangement
N
Me
Me
O
Interestingly,
(2S,3S)-3-hydroxy-2-methyl-4-oxopen-
O
O
tanoic acid (3) exclusively furnished a-aminoketone 23
derived from an Amadori rearrangement under the same
conditions as were utilized in the Ugi reaction
(Scheme 6).11 Contrary to the reaction of 2, the formation
of a pyroglutamic acid amide by the Ugi reaction was not
detected. Amadori rearrangement of the resulting imine F
might afford g-amino-b-keto acid G as a putative interme-
diate. Then, decarboxylation of the b-keto acid gave a-
aminoketone 23. Although all compounds bearing a a-hy-
droxyketone moiety, like 1–4, have the risk of undergoing
an Amadori rearrangement under the conditions of the
tautomerization
HO
OH
HO
Me
Me
F
G
Scheme 6 Amadori rearrangement of compound 3
Ugi reaction, the side reaction was not observed in any
case except with keto acid 3.12
In conclusion, we have disclosed a general synthesis of le-
vulinic acid derivatives and their stereoselective outcome
in the Ugi reaction. The stereoselectivity induced by
chiral keto acids with an unprotected alcohol was modest.
Synlett 2008, No. 15, 2249–2252 © Thieme Stuttgart · New York

2252
C. B. Gilley et al.
LETTER
(11) The stereochemistry of compound 23 was not determined.
For a recent application of the Amadori rearrangement in
natural product synthesis, see: Guzi, T. J.; Macdonald, T. L.
Tetrahedron Lett. 1996, 37, 2939.
The relative stereochemistry of the major product was as-
signed as anti. To improve the stereoselectivity, protec-
tion of the alcohol with a sterically hindered silyl group
could be effective, as well as use of a bulkier isocyanide.
Also, further investigation into the synthesis of function-
alized g-lactones by a substrate-controlled Passerini 3C-
2C reaction with a variety of chiral g-keto acids will be re-
ported in due course.
(12) 1H NMR data of the selected compounds are shown below.
Compounds 1 and 19 are reported as compounds 10 and 11a,
respectively, in the preceding paper.1 Compound 2: 1H NMR
(400 MHz, CDCl3): d = 5.01 (br s, 1 H), 4.10 (br s, 1 H), 3.11
(d, J = 4.8 Hz, 1 H), 2.20 (br s, 4 H), 1.34 (d, J = 6.8 Hz, 3
H). Compound 3: 1H NMR (400 MHz, CDCl3): d = 7.02 (br
s, 1 H), 4.53 (br s, 1 H), 2.95 (br s, 1 H), 2.13 (br s, 4 H), 1.05
(br s, 3 H). Compound 4: 1H NMR (400 MHz, CDCl3): d =
4.76 (br s, 1 H), 4.05 (s, 1 H), 1.91 (s, 3 H), 1.27 (s, 3 H), 1.22
(s, 3 H). Compound 5: 1H NMR (300 MHz, CDCl3): d = 5.97
(br s, 1 H), 2.99 (d, J = 12.3 Hz, 1 H), 2.66 (d, J = 12.3 Hz,
1 H), 2.25 (s, 3 H), 1.32 (s, 3 H). Compound 8: 1H NMR (400
MHz, CDCl3): d = 7.24–7.35 (m, 10 H), 6.56 (s, 1 H), 4.69–
4.76 (m, 1 H), 4.16–4.37 (m, 4 H), 3.34 (dd, J = 3.2, 13.6 Hz,
1 H), 2.78 (dd, J = 9.6, 13.6 Hz, 1 H), 1.96 (s, 3 H), 1.18 (d,
J = 6.8 Hz, 3 H). Compound 9: 1H NMR (400 MHz, CDCl3):
d = 7.21–7.36 (m, 10 H), 6.52 (s, 1 H), 5.21 (d, J = 12.4 Hz,
1 H), 5.17 (d, J = 12.4 Hz, 1 H), 4.30 (d, J = 8.4 Hz, 1 H),
2.84 (quin, J = 7.6 Hz, 1 H), 1.86 (s, 3 H), 1.16 (d, J = 7.2
Hz, 3 H). Compound 16: 1H NMR (300 MHz, CDCl3): d =
7.28–7.36 (m, 5 H), 5.13 (s, 2 H), 3.67 (s, 1 H), 3.32 (s, 3 H),
3.24 (s, 3 H), 2.73 (d, J = 14.1 Hz, 1 H), 2.40 (d, J = 14.1 Hz,
1 H), 1.30 (br s, 6 H). Compound 17: 1H NMR (300 MHz,
CDCl3): d = 7.32–7.38 (m, 5 H), 5.14 (d, J = 12.3 Hz, 1 H),
5.09 (d, J = 12.3 Hz, 1 H), 3.05 (d, J = 16.5 Hz, 1 H), 2.69
(d, J = 16.5 Hz, 1 H), 2.29 (s, 3 H), 1.32 (s, 3 H). Compound
20 (major diastereomer, anti): 1H NMR (400 MHz, CDCl3):
d = 8.97 (s, 1 H), 7.65 (d, J = 8.0 Hz, 1 H), 7.13–7.26 (m, 5
H), 6.79–6.83 (m, 2 H), 5.15 (d, J = 15.6 Hz, 1 H), 4.46–4.49
(m, 2 H), 4.03–4.12 (m, 2 H), 3.77 (s, 3 H), 3.42 (s, 3 H), 3.37
(s, 3 H), 2.77–2.85 (m, 3 H), 1.42 (s, 3 H), 1.28–1.36 (m, 3
H). Compound 21 (major diastereomer, anti): 1H NMR (300
MHz, CDCl3): d = 9.14 (s, 1 H), 7.69–7.73 (m, 2 H), 7.11–
7.22 (m, 5 H), 6.77–6.84 (m, 2 H), 5.36 (d, J = 15.3 Hz, 1 H),
4.40–4.45 (m, 1 H), 4.00 (d, J = 15.3 Hz, 1 H), 3.77 (s, 3 H),
3.41 (s, 3 H), 3.36 (s, 3 H), 2.81–3.00 (m, 3 H), 1.46 (s, 3 H),
1.28 (s, 3 H), 1.24 (s, 3 H). Compound 22 (major
Acknowledgment
We greatly acknowledge the University of California, a Kaplan fel-
lowship (M.J.B.), and a GAANN fellowship (C.B.G.) for financial
support of this research.
References and Notes
(1) See the preceding paper: Buller, M. J.; Gilley, C. B.;
Nguyen, B.; Olshansky, L.; Fraga, B.; Kobayashi, Y. Synlett
2008, 2244.
(2) For g-lactam synthesis, see: (a) Short, K. M.; Mjalli, A. M.
M. Tetrahedron Lett. 1997, 38, 359. (b) Harriman, G. C. B.
Tetrahedron Lett. 1997, 38, 5591. (c) Hanusch-Kompa, C.;
Ugi, I. Tetrahedron Lett. 1998, 39, 2725. (d) Tye, H.;
Whittaker, M. Org. Biomol. Chem. 2004, 2, 813. (e) For g-
lactone synthesis, see: Passerini, M. Gazz. Chim. Ital. 1923,
53, 331. For recent reviews on multicomponent
condensation reactions, see: (f) Ramon, D. J.; Yus, M.
Angew. Chem. Int. Ed. 2005, 44, 1602. (g) Dömling, A.
Chem. Rev. 2006, 106, 17.
(3) Database search on the reported synthesis of levulinic acid
derivatives by MDL CrossFire Commander was conducted
on April 25, 2008.
(4) For a recent example in the literature, see: Mahajan, V. A.;
Borate, H. B.; Wakharkar, R. D. Tetrahedron 2006, 62,
1258.
(5) (a) For compound 1, see: Lueoend, R. M.; Walker, J.; Neier,
R. W. J. Org. Chem. 1992, 57, 5005. (b) For an ester of
compound 4, see: Kende, A. S.; Kawamura, K.; Orwat, M. J.
Tetrahedron Lett. 1989, 30, 5821.
(6) Evans, D. A.; Tedrow, J. S.; Shaw, J. T.; Downey, C. W.
J. Am. Chem. Soc. 2002, 124, 392.
diastereomer, anti): 1H NMR (400 MHz, CDCl3): d = 8.94 (s,
1 H), 7.49 (d, J = 8.0 Hz, 1 H), 7.09–7.28 (m, 5 H), 6.79–6.88
(m, 2 H), 4.76 (d, J = 15.2 Hz, 1 H), 4.41–4.45 (m, 1 H), 4.22
(d, J = 15.2 Hz, 1 H), 4.08–4.13 (m, 1 H), 3.74 (s, 3 H), 3.41
(s, 3 H), 3.39 (s, 3 H), 2.74 (m, 3 H), 2.92 (m, 1 H), 1.47 (s,
3 H), 1.41 (s, 3 H). Compound 23: 1H NMR (400 MHz,
CDCl3): d = 7.21 (d, J = 8.3 Hz, 2 H), 6.83 (d, J = 8.3 Hz, 2
H), 3.74–3.77 (m, 5 H), 3.60 (q, J = 8.8 Hz, 1 H), 3.11 (br s,
1 H), 2.53 (dd, J = 7.2, 17.6 Hz, 1 H), 2.37 (dd, J = 7.6, 17.6
Hz, 1 H), 1.21 (d, J = 6.8 Hz, 3 H), 1.04 (d, J = 7.2 Hz, 3 H).
(7) Abiko, A.; Liu, J.-F.; Masamune, S. J. Org. Chem. 1996, 61,
2590.
(8) The anti isomer was also isolated in 18% yield.
(9) Enantioselective synthesis of 4 would be achieved by
Mulzer’s procedure: Kögl, M.; Brecker, L.; Warrass, R.;
Mulzer, J. Angew. Chem. Int. Ed. 2007, 46, 9320.
(10) (a) Gilley, C. B.; Buller, M. J.; Kobayashi, Y. Org. Lett.
2007, 9, 3631. (b) Isaacson, J.; Loo, M.; Kobayashi, Y. Org.
Lett. 2008, 10, 1461. (c) Isaacson, J.; Gilley, C. B.;
Kobayshi, Y. J. Org. Chem. 2007, 72, 3913. (d) Vamos,
M.; Ozboya, K.; Kobayashi, Y. Synlett 2007, 1595.
(e) Kreye, O.; Westermann, B.; Wessjohann, L. A. Synlett
2007, 3188.
Synlett 2008, No. 15, 2249–2252 © Thieme Stuttgart · New York

Copyright of Synlett is the property of Georg Thieme Verlag Stuttgart and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.