Stereocontrol of the Horner-Wadsworth-Emmons Reaction: Application to the Synthesis of HIV-1 Protease Inhibitors

Article abstract of DOI:10.1071/CH01093

A systematic study on the Horner-Wadsworth-Emmons (HWE) reaction has shown that ethyl diphenylphosphonoacetate and methyl diphenylphosphonoacetate give a high excess of (Z)-alkenes. These reaction conditions were then used to prepare (Z)-ethyl-5-phenylpent-2-enoate, the corresponding (E)-isomer being prepared by standard Wittig chemistry. Reduction of each allylic ester, with diisobutylaluminium hydride (DIBAL), gave the allylic alcohols (15) and (19), respectively. Epoxidation of (15) and (19), under Sharpless conditions, gave separate samples of all four stereoisomers of 2,3-epoxy-5-phenylpentan-1-ol. Esterification of each isomer with Cbz-valine, under Mitsunobu conditions, provided (9)-(12) which were assayed against HIV protease. The cis series (9)-(10) proved to be significantly more potent than the trans (11)-12) and, within each of these series, the isomers derived from L-diisopropyltartrate [(9) and (11)] were the most active.

Full text of DOI:10.1071/CH01093

P u b l i s h i n g
AUSTRALIAN JOURNAL OF
CHEMISTRY
A N I N T E R N A T I O N A L J O U R N A L F O R C H E M I C A L S C I E N C E
publishing research papers from all fields of chemical
science, including synthesis, structure, new materials,
macromolecules, supramolecular chemistry, biological
chemistry, nanotechnology, surface chemistry, and
analytical techniques.
Volume 54, 2001
CSIRO 2001
©
All enquiries and manuscripts should be directed to:
Dr Alison Green
Australian Journal of Chemistry–
an International Journal for Chemical Science
CSIRO PUBLISHING
PO Box 1139 (150 Oxford St)
Collingwood, Vic. 3066, Australia
Telephone: +61 3 9662 7630
Fax: +61 3 9662 7611
E-mail: publishing.ajc@csiro.au
Published by CSIRO PUBLISHING
for CSIRO and the Australian Academy of Science
w w w. p u b l i s h . c s i r o . a u / j o u r n a l s / a j c

Aust. J. Chem. 2001, 54, 391–396
Stereocontrol of the Horner–Wadsworth–Emmons Reaction:
Application to the Synthesis of HIV-1 Protease Inhibitors
A
A
A,B
Derek C. Martyn, Deborah A. Hoult and Andrew D. Abell
A
Department of Chemistry, University of Canterbury, Private Bag 4800, Christchurch, New Zealand.
Author to whom correspondence should be addressed (e-mail: a.abell@chem.canterbury.ac.nz).
B
A
systematic study on the Horner–Wadsworth–Emmons (HWE) reaction has shown that ethyl
diphenylphosphonoacetate and methyl diphenylphosphonoacetate give a high excess of (Z)-alkenes. These reaction
conditions were then used to prepare (Z)-ethyl-5-phenylpent-2-enoate, the corresponding (E)-isomer being
prepared by standard Wittig chemistry. Reduction of each allylic ester, with diisobutylaluminium hydride (DIBAL),
gave the allylic alcohols (15) and (19), respectively. Epoxidation of (15) and (19), under Sharpless conditions, gave
separate samples of all four stereoisomers of 2,3-epoxy-5-phenylpentan-1-ol. Esterification of each isomer with
Cbz-valine, under Mitsunobu conditions, provided (9)–(12) which were assayed against HIV protease. The cis
series (9)–(10) proved to be significantly more potent than the trans (11)–(12) and, within each of these series, the
isomers derived from L-diisopropyltartrate [(9) and (11)] were the most active.
Manuscript received: 26 July 2001.
Final version: 17 October 2001.
C
H
01039
D
S
.
t
e
C
r
e
.
o
M
c
a
o
r
n
t
y
r
o
n
l
o
,
D
f
.
t
h
A
.
e
H
o
o
r
u
n
l
r
e
a
t
n
–
d
W
a
A
.
d
s
D
.
w
o
A
b
t
r
e
h
–
E
l
m
m
o
n
s
R
e
a
c
i
t
o
n
Introduction
Results and Discussion
The Horner–Wadsworth–Emmons (HWE) reaction of a
resonance-stabilized phosphonate carbanion (1) and an
aldehyde (2), or ketone, is a well established method for the
introduction of carbon–carbon double bonds (see
Scheme 1).[ The utility of this reaction is facilitated by the
fact that its stereochemical outcome can be controlled; for
example, the structure of the phosphonate, the nature of the
base, the structure of the carbonyl reagent and the reaction
temperature all influence the E-to-Z product ratios.[
However, most of the work published in this area is rather
fragmented so that there are comparatively few systematic
studies[ addressing these issues. In this paper we present a
detailed study of the effect of some of these reaction
parameters on the outcome of reactions of phosphonates (1)
with a range of aldehydes. The optimum conditions for the
formation of the (Z)-alkene were then used to synthesize
some epoxide-based inhibitors of HIV protease which were
assayed against the enzyme.
Model Studies
We initially investigated reactions of the phosphonates
[
(
(1a–c), see Scheme 1] with three types of aldehyde, (2a–b),
2c–f), (2g–j), where R is conjugated to the carbonyl, is
3
1,2]
unbranched aliphatic, and is α-branched aliphatic,
respectively. This was done in an attempt to ascertain the
effect of aldehyde structure on the stereochemical outcome
of HWE reactions (see Table 1). In this study, the aldehydes
were reacted at either room temperature or –78°C with the
anion derived from the phosphonate. The anion itself was
generated from a tetrahydrofuran (THF) solution of the
phosphonate at 0°C, on treatment with n-butyllithium. The
E-to-Z ratio of products was determined by integration of the
1–6]
6]
1
alkene protons in the H nuclear magnetic resonance (NMR)
spectrum of the crude reaction mixtures. The product
alkenes were then isolated and characterized.
The reactions of the conjugated aldehydes benzaldehyde
(
2a) and cinnamaldehyde (2b) with the anion derived from
1
2
3
3
2
1
−
trimethyl phosphonoacetate (1a) at –78°C gave a very strong
preference for formation of the (E)-isomer. For both
aldehydes the E-to-Z product ratio was greater than 19:1 (see
(
R O)2POCH2CO2R
+
R CHO
R CH=CHCO2R
+
(R O)2PO2
1
2
(
1a) R , R , = Me
1
2
(
1b) R , R , = Et
1
2
(
1c) R = Ph, R = Me
Table 1). This selectivity has been noted previously in related
1
2
(
1d) R = Me, R = Et
[7,8]
studies.
Similarly, reactions of (1a) with the unbranched
1
2
(
1e) R = Et, R = Me
1
i
2
aliphatic aldehydes (2c–f) gave a strong preference for the
(E)-alkene at both room temperature and –78°C (see Table
(
1f) R = Pr , R = Me
1
2
(
1g) R = 4-MeOC6H4, R = Me
1
2
(
1h) R = 3-MeOC6H4, R = Me
1
). However, it is interesting to note that the preference for
Scheme 1
the (E)-isomer, in these reactions, decreases somewhat with
©
CSIRO 2001
10.1071/CH01093
0004-9425/01/060391

3
92
D. C. Martyn, D. A. Hoult and A. D. Abell
Table 1. HWE reactions of phosphonates (1a–c) with aldehydes (2)
1
2
3
3
2
(
R O)2POCH2CO2R
R CHO
R CH=CHCO R
2
(1)
(2)
(3)
Phosphonates (E:Z)
(1b); R = R = Et
Room temp. –78°C
1
2
1
2
1
2
Aldehyde (2)
(1a); R = R = Me
(1c); R = Ph, R = Me
3
R
Room temp.
–78°C
–78°C
(
(
(
(
(
(
(
(
(
(
2a)
2b)
2c)
2d)
2e)
2f)
2g)
2h)
2i)
Ph
—
—
11:1
7:1
5.5:1
5.7:1
4.5:1
3.8:1
3.4:1
3.6:1
>19:1
>19:1
8:1
3.6:1
2:1
—
—
—
>19:1
>19:1
—
13:1
—
>19:1
—
—
9.3:1
9.4:1
—
1.5:1
—
—
—
—
1:1.3
1:7
—
1:3.3
1:5.7
—
PhCH=CH
Me
Et
Me(CH2)5
Me(CH2)9
Me CH
2:1
1:2.8
1:2.8
1:3
2
Me(CH ) CHMe
cyclohexyl
2
8
—
12:1
—
1.5:1
2j)
Et CH
1:5
1:3.8
2
1
2
reactions where the amount of the (Z)-alkene was greatest for
reactions involving (1c) followed by (1a) and (1b).
Table 2. Effect of R and R on the outcome of HWE reactions at
–78°
1
R2
3
3
2
The reactions of the anions derived from the phosphonates
(
R O)
2
POCH
2
CO
2
R CHO
(2)
R CH=CHCO R
2
1
2
(
1d–f) were studied to determine which of R or R was most
important to the stereochemical outcome of these reactions
see Table 2). From these results it is clear that the nature of
(1)
(3)
(
Phosphonate (1) E-to-Z ratio of alkenes (3) obtained on reaction of the indicated aldehyde (2)
1
2
(2a); R3 = Ph
3
3
R
R
(2e); R = Me(CH2)5
(2g); R = Me CH
the phosphonate has little effect on reactions with aromatic
2
aldehydes such as (2a), where an E-to-Z ratio of > 19:1 was
observedin all cases. However, the natureofR does appearto
(
(
(
(
(
1a) Me Me
1b) Et Et
1d) Me Et
>19:1
>19:1
>19:1
>19:1
>19:1
2:1
9.4:1
1.2:1
6.7:1
>19:1
1:2.8
1.5:1
1:2.9
2.4:1
10:1
1
be important for the stereochemical outcome in reactions of
the other aldehydes. For example, reaction of (2e) or (2g) with
1e) Et
Me
1f) iPr Me
2
phosphonates (1a) and (1e), both of which have R = Me but
1
R
= Me and Et, respectively, gave products with a
3
increasing chain length of R . By contrast, reactions of the
more sterically hindered α-branched aliphatic aldehydes
2g–j), with the anion of (1a) at –78°C, gave a predominance
significantly different E-to-Z ratio. The formation of
1
[10]
(Z)-alkene is favoured when R is small, i.e. Me. The
(
phosphonate (1f), which contains a large isopropyl group at
1
of the (Z)-alkene (Table 1), the extreme case being the
reaction of (2j) at –78°C where an E-to-Z ratio of 1:5 was
observed (cf. (2g) which gave an E-to-Z ratio of 1:2.8 under
the same conditions). A preference for the (E)-alkene
remained in the equivalent reactions at room temperature.
The anion derived from triethyl phosphonoacetate (1b)
was also reacted with the aldehydes (2d), (2e), (2g) and (2j) to
ascertain the generality of these observations (see Table 1).
Here, again, we found that the amount of (Z)-isomer formed
increased with the increasing bulk of the R group. However,
unlike the reactions of (1a), the (E)-isomer still predominated
in the reactions of aldehydes (2g) and (2j) at –78°C. Again we
noted, as have others, that a lower reaction temperature
favourstheformationofthe(Z)-alkene.
R , gave the poorest (Z)-selectivity in reactions with (2a), (2e)
and (2g). A similar trend is evident on comparing the
outcomes of the reactions of (1b) and (1d). The nature of R2
seems to be comparatively unimportant for the
stereochemical outcome of these reactions (cf. reactions of
(1a)–(1d) and (1b)–(1e) in Table 2).
In the last series of reactions we considered the effect of
the electronic properties of the phenyl group in phosphonates
of the type (1c). We had already observed that (1c) led to
elevated levels of the (Z)-alkene relative to the other
phosphonates (see Table 1). Two new phosphonates were
prepared and studied, one bearing a para-methoxy
substituent (1g) and the other a meta-methoxy substituent
(1h) (see Scheme 1).* Reaction of the phosphonate (1g) with
(2e) and (2g) at –78°C gave a significant increase in the
amount of (E)-alkene produced (an E-to-Z ratio of 1:4.2 and
1:2.1 for (2e) and (2g), respectively) relative to the
equivalent reactions of (1c) (an E-to-Z ratio of 1:7 and 1:3.3
for (2e) and (2g), respectively, see Table 1). By contrast, the
reaction of (1h) with the same aldehydes gave equivalent
results to reactions of (1c) (an E-to-Z ratio of 1:7.2 and 1:3.3
for (2e) and (2g), respectively). From these results it is
3
[6]
The results of the reactions of a selection of aldehydes
1
[9]
with a third phosphonate (1c), where R is aromatic, are
also summarized in Table 1. In this case all aldehydes gave a
predominance of the (Z)-alkene with the sterically most
demanding examples again giving the greatest excess, in
general. From the studies presented in Table 1 it appears that
1
2
the nature of R and R of the phosphonate (1) has a
significant effect on the stereochemical outcome of these
*
These phosphonates were chosen due to the availability of the starting materials used in their synthesis[9] and due to the differing electronic
properties of meta and para substitution.

Stereocontrol of the Horner–Wadsworth–Emmons Reaction
393
apparent that (E)-alkene formation is favoured by using
phosphonates that have an electron-rich aryl group at R .
−
(
PhO)2PO CHCO2Et
1
Ph
Ph
Ph
CO2Et
(14)
CHO
13)
4
9%
(
Synthesis of HIV-1 Protease Inhibitors
DIBAL
5
5%
i
A number of epoxide-based inhibitors of HIV protease are
Ti(OPr)4,
L-DIPT, TBHP
known. For example, 1,2-epoxy-3-(p-nitrophenoxy)propane
Ph
OH
OH
11]
(
EPNP) [(4),[ see Fig. 1] is a relatively weak irreversible
O
68%
(
16)
(15)
i
inhibitor (kinact 11 mM), where the nitrophenoxy group is
N-Cbz-L-Valine,
Ti(OPr)4,
6
6%
thought to reside in the S pocket† of the enzyme’s active
46%
1
PPh3, DEAD
D-DIPT, TBHP
site. Modifications to the structure of EPNP led to the
development of the tripeptidomimetic epoxide (5) (kinact
Ph
OH
(9)
O
2
0 µM), which spans the S –S ´ subsites of the enzymes
(17)
3
1
[
12]
active site. C -Symmetric epoxides such as compound (6)
are reversible HIV-1 protease inhibitors (K 75 nM), while
N-Cbz-L-Valine,
PPh3, DEAD
2
38%
[
13]
i
the non-peptidic epoxide (7) is an irreversible inhibitor of
[14]
(10)
HIV-1 protease (Kinact 65 µM).
Previous work in our
Scheme 2
laboratories led to the development of the peptidomimetic
[
15]
irreversible inhibitor (8) (Kinact 1.8 µM).
the synthesis and assay of a series of epoxide-based HIV-1
We now report
Ph3PCHCO2Et
94%
(13)
Ph
CO2Et
18)
protease inhibitors (9)–(12) (see Fig. 2). These compounds
(
O
2
N
70%
DIBAL
i
Ti(OPr)4,
L-DIPT, TBHP
O
−
Cbz = PhCH
2
OCO
O
Ph
OH
Ph
OH
O
31%
(4)
(19)
(20)
Ph
Ph
N
Ph
i
Ti(OPr)4,
N-Cbz-L-Valine,
PPh3, DEAD
18%
Cbz-Phe-Ala
45%
D-DIPT, TBHP
Cbz-Val
Val-Cbz
N
H
N
H
H
O
O
(5)
(6)
(11)
Ph
OH
O
(21)
O
2
O N
N
O
N-Cbz-L-Valine,
PPh3, DEAD
37%
O
O Val-Cbz
O
O
(8)
(12)
(7)
Scheme 3
Fig. 1. Structures of a number of epoxide-based inhibitors of HIV
protease.
were prepared to probe the effect of extending the peptide
sequence of (4) in the C-direction and also the effect of the
configuration of the epoxide on inhibitory activity. The
strategy for the synthesis of these compounds was based
upon key HWE and Wittig preparations of (14) and (18).
Subsequently, epoxidation under Sharpless conditions gave
the four isomeric epoxides which were elaborated into the
target compounds (see Schemes 2 and 3).‡
O
O
H
N
H
N
Ph
O
Ph
Ph
O
Ph
O
O
O
O
O
O
(9)
(10)
O
O
A HWE reaction of hydrocinnamaldehyde (13) with the
anion of ethyl diphenylphosphonoacetate,§ generated by
n-butyllithium in THF at –78°C, gave the (Z)-allylic ester
Ph
O
Ph
O
N
H
O
Ph
N
H
O
Ph
O
O
O
O
(
11)
(12)
(
(
14) in 49% yield, the crude mixture contained the (Z) and
E) isomers in a ratio of 8:1 [cf. results for (1c) in Table 1].
Fig. 2. Structures of the epoxide-based HIV-1 protease inhibitors
9)–(12) synthesized here.
(
Reduction of (14) with diisobutylaluminium hydride
†
Note the use of Schechter–Berger nomenclature (I. Schechter, A. Berger, Biochem. Biophys. Res. Commun. 1967, 27, 157). The residues on the
amino-terminal side of the peptide bond that is to be cleaved are denoted P –P , and those on the carboxy-terminus are denoted P ´–P ´. In turn,
1
n
1
n
the corresponding subsites on the enzyme are denoted S –S ´.
n
n
‡
A phenyl group was chosen in place of the nitrophenoxy group of (7) because the synthetic precursor is commercially available and the
nitrophenoxy group interferes with the HIV protease inhibitory assay.
Ethyl diphenylphosphonoacetate was used in place of methyl diphenylphosphonoacetate (1c) in these reactions because it was at hand and since
§
2
we had already demonstrated that the nature of R is comparatively unimportant to the stereochemical outcome.

3
94
D. C. Martyn, D. A. Hoult and A. D. Abell
Table 3. Inhibition of HIV Protease
Conclusions
In summary, an investigation of the HWE reaction has led to
the determination of experimental conditions to optimize the
formation of (E)- and (Z)-alkenes. We have shown, as have
Compound
Inhibition (%)A
(
(
8)B
9)
88C
100
26
16
9
1
others, that the nature of the aldehyde and R in the phospho-
(
(
(
10)
11)
12)
nate have a significant influence on the stereochemical out-
come. The optimum phosphonate for (Z)-alkene formation,
namely ethyl diphenylphosphonoacetate or methyl diphenyl-
phosphonoacetate, was then used to prepare a series of iso-
meric epoxide-based HIV protease inhibitors that reveal
some interesting structure–activity trends.
A
Assay conditions: after 1 min, pH 6.5,
7°C, [S] 50 µM, [I] 200 µM.
Reference 15.
3
B
C
[I] 20 µM.
(
5
DIBAL) gave the corresponding (Z)-allylic alcohol (15) in
5% yield. Separate samples of this were then epoxidized
under Sharpless conditions, using both L-diisopropyltartrate
L-DIPT) and D-DIPT, to give (16) and (17) respectively.
These were then treated with N-Cbz-L-valine and
diethylazodicarboxylate to give the
Experimental
NMR spectra were recorded on either a Varian Unity 300 or XL-300
spectrometer, and are reported in ppm relative to Me Si. Mass
spectrometry (MS) was performed on
spectrometer. Infrared (IR) spectra were recorded on a Shimadzu
FTIR-8201PC spectrophotometer. Optical rotations were measured on
4
(
a Kratos MS80RFA
(DEAD)/PPh3
compounds (9) and (10) in 46 and 38% yields, respectively.
The remaining stereoisomers (11) and (12) were prepared by
an analogous route using the (E)-alkene (18), which was
prepared using standard Wittig chemistry (see Scheme 3).
a JASCO J-20C recording spectropolarimeter and [α] values are given
D
–
1
2
–1
in units of 10 ° cm g , with the concentration (c) given in units of mg
–
3
cm . Microanalyses were performed at the Department of Chemistry,
University of Otago, Dunedin, New Zealand. Radial chromatography
was performed on a Harrison and Harrison Chromatotron using 1 and
Here, the key Wittig reaction gave a 14:1 ratio of (E)- to
2
mm plates. Column chromatography was performed using 230–400
16]
(
Z)-alkenes.[ The required (E)-alkene (18), isolated in 94%
mesh Merck Silica Gel 60 under positive air pressure. Light petroleum
refers to the fraction with a boiling point of 50–70°C. The starting
phosphonates (1) were either purchased or prepared using standard
yield, was then reduced and used in the subsequent Sharpless
chemistry as detailed in Scheme 3.
[9]
methods.
It is interesting to note that the 13C NMR spectra of the
isomeric pairs (9)–(10) and (11)–(12) show some diagnostic
features. The spectra of (11) and (12) are essentially identical
with the following exceptions: (i) one of the epoxide carbon
resonances was observed at δ 55.9 and 56.0 for (11) and (12),
General Procedure for Reactions of the Phosphonates (1)
A stirred solution of the phosphonate (1) (1.4 mmol) in dry THF
(2.5 mL) was cooled to 0°C under a nitrogen atmosphere and
n-butyllithium (1.4 mmol of a 1.6 M solution in hexane) was added
dropwise. After 10 min, the solution was brought to the desired
temperature, and the aldehyde (2) (1.22 mmol) was added dropwise.
The resultant mixture was stirred for 1.5 h and then quenched with
water (3 mL) and ether (5 mL). The organic phase was washed with
respectively; (ii) the epoxide-CH O resonance was observed
2
at δ 65.0 and 65.4 for (11) and (12) respectively; (iii) (11)
revealed seven aromatic resonances while (12) showed only
six; (iv) the carbobenzyloxy (Cbz) carbonyl resonance was
observed at δ 156.2 and 159.4 for (11) and (12), respectively.
The spectra of (9) and (10) reveal similar trends: (i) the
epoxide carbon resonance differed (δ 53.6 and 53.8 for (9)
saturated brine, dried over anhydrous MgSO , filtered, and the solvent
4
was removed by evaporation under reduced pressure. The residue was
1
analysed by H NMR spectroscopy and finally purified by column
chromatography on silica. The NMR spectra of the simple alkenes (2)
[
19]
were compared with literature data.
and (10), respectively); (ii) the epoxide-CH O resonance
2
differed (δ 63.3 and 63.4 for (9) and (10) respectively); (iii)
(
Z)-Ethyl-5-phenylpent-2-enoate (14)
(
9) revealed seven aromatic resonances while (10) showed
The general procedure was adopted using ethyl diphenylphosphonoac-
etate (1.74 g, 5.43 mmol), dry THF (15 mL), n-butyllithium (4.2 mL of
a 1.6 M solution in hexane, 6.7 mmol), a temperature of –78°C and the
aldehyde (13) (610 mg, 4.55 mmol). The residue was purified by col-
umn chromatography (90% light petroleum, 10% ethyl acetate) to give
only six; (iv) the carbobenzyloxy (Cbz) carbonyl resonance
differed (δ 156.1 and 155.7 for (9) and (10), respectively).
Some other minor differences were also observed for (9) and
(
10), see experimental section for details.
[16]
(14) (457 mg, 49%) and (18) (53 mg, 6%).
Assay Against HIV Protease
(Z)-5-Phenylpent-2-en-1-ol (15)
Compounds (9)–(12) were assayed against HIV protease as
The first thing to note from the
results presented in Table 3 is that, as expected, all
A solution of (14) (321 mg, 1.6 mmol) in dry CH Cl (20 mL) was cooled
2 2
previously detailed.[
15,17]
to –78°C under a nitrogen atmosphere, and diisobutylaluminium hydride
DIBAL, 6.1 mL of a 1 M solution in CH Cl , 6.1 mmol) was added
(
2
2
dropwise.Theresultant solutionwasstirredfor 3hat–78°C,and water (1
mL) was then added over 5 min, followed by 10% aqueous potassium
sodium tartrate (10 mL). The mixture was then warmed to room
temperature and the aqueous layer was separated and extracted three
times with CH2Cl2. The combined organic extracts were washed with
compounds assayed in the current study were less active than
(
8),[ where this compound contains a more optimum
15]
[11]
para-nitrophenoxy group as found in EPNP (4). However,
what is critical is that the cis series, (9)–(10), was
significantly more active
[18]
water, dried over anhydrous MgSO , and the solvent was removed by
than the trans, (11)–(12). In
4
evaporation under reduced pressure. The residue was purified by radial
chromatography on silica (70% light petroleum, 30% ethyl acetate) to
give (15) as a colourless oil (140 mg, 55%) (Found: 162.1045. C H O
addition, it is apparent that within each series, the two
isomers derived from L-DIPT [(9) and (11)] are the most
active.
1
1
14
1
requires 162.1045). H NMR (CDCl ) δ 7.17–7.32, 5H, m, ArH; 5.59,
3

Stereocontrol of the Horner–Wadsworth–Emmons Reaction
395
2
2
1
H, m, CH; 4.00, 2H, d, J 5.9 Hz, CH OH; 2.69, 2H, t, J 7.3 Hz, PhCH ;
m, PhCH CH ; 0.98, 3H, d, J 6.8 Hz, Me; 0.90, 3H, d, J 6.8 Hz, Me. 13C
2
2
2
2
1
3
.40,2H,q,J 7.3Hz, CH ;1.23, br s, 1H, OH. C NMR(CDCl ) δ141.5;
NMR (CDCl ) δ 171.9; 155.7; 140.6; 128.5; 128.4; 128.2; 128.1; 126.3;
2
3
3
28.6;128.3;126.0;131.5;129.3;58.3;35.6;29.2.
67.1;63.4;59.0;55.9;53.8;32.7;31.2;29.8;19.0;17.5.
(2S,3R)-2,3-Epoxy-5-phenylpentan-1-ol (16)
(2S,3S)-2,3-Epoxy-5-phenylpentan-1-ol (20)
A solution of titanium tetraisopropoxide (701 mg, 2.5 mmol) and 4Å
molecular sieves (250 mg) in dry CH Cl (25 mL) was cooled to –25°C
The preparation was performed as described by the procedure for (16)
i
using Ti(OPr )4 (429 mg, 1.5 mmol), 4 Å molecular sieves (150 mg),
2
2
[16,20]
under a nitrogen atmosphere. A solution of L-(+)-diisopropyl tartrate
L-DIPT, 691 mg, 3.0 mmol) in dry CH Cl2 (1 mL) was added
and L-diisopropyl tartrate (421 mg, 1.8 mmol) (19)
(245 mg,
(
1.5 mmol), and TBHP in CH2Cl2 (1 mL of a 3 M solution, 3 mmol) was
added. Purification of the residue by column chromatography (60%
light petroleum, 40% ethyl acetate) gave (20) as a colourless oil (84 mg,
2
dropwise, followed by (15) (395 mg, 2.5 mmol) in dry CH Cl (4 mL).
2
2
The resultant solution was stirred for 30 min, t-butylhydroperoxide
TBHP, 1 mL of a 4.8 M solution in CH Cl , 4.8 mmol) was added and
1
13
(
31%). The H NMR, C NMR and HRMS data of the product (20)
2
2
[
20]
the resultant solution was stored at –20°C for 20 h. The reaction was
quenched with 10% aqueous tartaric acid (25 mL), the organic phase
was separated, and the aqueous layer was extracted three times with
CH Cl . The combined organic fractions were dried over anhydrous
were identical to those published.
2S,3S)-N-Cbz-L-Valine 2,3-Epoxy-5-phenylpentyl Ester (11)
(
2
2
The preparation was performed as described for (9) using (20) (23 mg,
0.13 mmol). Purification of the residue by column chromatography
(80% light petroleum, 20% ethyl acetate) gave (12) as a colourless oil
(24 mg, 45%) (Found: C, 70.0; H, 7.2; N, 3.5. C H NO requires C,
MgSO , filtered, and the solvent was removed by evaporation under
4
reduced pressure. Excess peroxide was removed by azeotropic
distillation with CCl and the crude mixture was purified by radial
4
24 29
5
1
chromatography on silica (60% light petroleum, 40% ethyl acetate) to
70.0; H, 7.1; N, 3.4%). H NMR (CDCl ) δ 7.17–7.37, m, 10H, m, ArH;
3
give (16) as a colourless oil (303 mg, 68%) (Found: 160.0888. C H O
5.25, 1H, d, J 8.8 Hz, CHNH; 5.11, s, 2H, CO CH Ph; 4.31–4.41, 2H,
1
1
12
2
2
+
requires [M–H O] 160.0888). IR (thin film) 3422(br), 3026, 2928,
860, 1603 cm . [α] –1.9 (c, 10.9 in CHCl ). H NMR (CDCl ) δ
.19–7.34, 5H, m, ArH; 3.57, 2H, d, J 5.4 Hz, CH OH; 3.11, 2H, m,
xCHO; 2.88, 1H, m, PhCH ; 2.73, 1H, m, PhCH ; 2.01, 1H, m, CH ;
.81, 1H, m, CH ; 1.34, 1H, br s, OH. C NMR (CDCl ) δ 140.4;
m, CHCH O and CHNH; 3.96, 1H, dd, J 5.9, 12.2 Hz, CHCH O;
2
2
2
–
1
23
D
1
2
7
2
1
1
2.72–2.92, 4H, m, 2×–CHO–, PhCH CH ; 2.17, 1H, m, CH(Me) ;
2 2 2
3
3
2
1.88, 2H, m, PhCH CH ; 0.98, 3H, d, J 6.8 Hz, Me; 0.90, 3H, d, J 6.8
2 2
1
3
2
2
2
Hz, Me. C NMR (CDCl ) δ 171.8; 156.2; 140.8; 128.54; 128.5; 128.4;
128.2; 128.1; 126.2; 67.1; 65.0; 59.0; 55.9; 55.2; 33.2; 32.1; 31.3; 18.9;
17.5.
3
1
3
2
3
28.0; 127.9; 125.7; 60.0; 56.6; 56.1; 32.2; 29.2.
(2S,3R)-N-Cbz-L-Valine 2,3-Epoxy-5-phenylpentyl Ester (9)
(2R,3R)-2,3-Epoxy-5-phenylpentan-1-ol (21)
The preparation was performed as described for (16) using Ti(OPri)
To a solution of (16) (15 mg, 0.08 mmol) in dry THF (1.5 mL) was
added N-Cbz-L-valine (46 mg, 0.19 mmol) and triphenylphosphine (46
mg, 0.18 mmol). After stirring the mixture for 5 min, DEAD (31 mg,
4
(113 mg, 0.40 mmol), 4 Å molecular sieves (50 mg), D-diisopropyl
1
6,20
tartrate (112 mg, 0.48 mmol) (19)
(64 mg, 0.40 mmol), and TBHP
0.18 mmol) was added dropwise, and the resultant solution was stirred
in CH Cl (0.25 mL of a 3 M solution, 0.75 mmol). Purification of the
2
2
at room temperature for 3 h, after which CH Cl was added. The crude
residue by column chromatography (60% light petroleum, 40% ethyl
2
2
1
mixture was washed three times with saturated aqueous NaHCO , the
acetate) gave (21) as a colourless oil (13 mg, 18%). The H NMR and
3
1
3
organic layer was dried over anhydrous MgSO , filtered, and the solvent
C NMR spectra of the product (21) were identical to those
4
[21]
was removed by evaporation under reduced pressure. The residue was
purified by radial chromatography on silica (80% light petroleum, 20%
ethyl acetate) to give (9) as a colourless oil (16 mg, 46%) (Found:
published and reported for (20).
(2R,3R)-N-Cbz-L-Valine 2,3-Epoxy-5-phenylpentyl Ester (12)
4
1
7
11.2049. C H NO requires 411.2046). IR (CHCl ) 3436, 2968,
24 29 5 3
The preparation was performed as described for (9) using (21) (13 mg,
–
1
23
1
722, 1511 cm . [α] –6.1 (c, 16.0 in CHCl ). H NMR (CDCl ) δ
D
3
3
0
.07 mmol). Purification of the residue by column chromatography
.19–7.35, 10H, m, ArH; 5.25, 1H, d, J 8.8 Hz, CHNH; 5.11, 2H, br s,
(80% light petroleum, 20% ethyl acetate) gave (11) as a colourless oil
CO CH Ph; 4.36, 1H, dd, J 4.4, 8.8 Hz, CHNH; 4.19, 1H, dd, J 4.4,
2
2
(
7
5
11 mg, 37%) (Found: C, 69.9; H, 6.9; N, 3.6. C H NO requires C,
24
29
5
12.2 Hz, CHCH O; 4.05, 1H, dd, J 6.3, 12.2 Hz, CHCH O; 3.14, 1H,
1
2
2
0.0; H, 7.1; N, 3.4%). H NMR (CDCl ) δ 7.18–7.38, 10H, m, ArH;
3
m, CHCH O; 3.06, 1H, m, CH CH CH; 2.72–2.89, 2H, m, PhCH CH ;
2
2
2
2
2
.29, 1H, d, J 9.3 Hz, CHNH; 5.12, 2H, s, CO CH Ph; 4.34, 2H, m,
2
2
2
.18, 1H, m, CH(Me) ; 1.87, 2H, m, PhCH CH ; 0.98, 3H, d, J 6.9 Hz,
2 2 2
CHCH O and CHNH; 3.96, 1H, dd, J 6.3, 12.2 Hz, CHCH O;
1
3
2
2
Me; 0.90, 3H, d, J 6.9 Hz, Me. C NMR (CDCl ) δ 171.8; 156.1; 140.6;
3
2
1
.70–2.95, 4H, m, 2×–CHO–, PhCH CH ; 2.18, 1H, m, CH(Me) ;
2 2 2
1
3
28.52; 128.50; 128.4; 128.2; 128.1; 126.2; 67.0; 63.3; 59.0; 55.9; 53.6;
2.6; 31.2; 29.8; 18.9; 17.5.
.89, 2H, m, PhCH CH ; 0.98, 3H, d, J 6.8 Hz, Me; 0.90, 3H, d, J 6.8
2
2
13
Hz, Me. C NMR (CDCl ) δ 171.8; 159.4; 140.8; 128.5; 128.4; 128.2;
3
1
28.1; 126.2; 67.1; 65.4; 59.0; 56.0; 55.2; 33.2; 32.1; 31.2; 18.9; 17.5.
(2R,3S)-2,3-Epoxy-5-phenylpentan-1-ol (17)
The reaction was performed as described for (16) using Ti(OPri)
4
Acknowledgments
(113 mg, 0.40 mmol), D-(–)-diisopropyl tartrate (105 mg, 0.45 mmol),
We thank Drs D. Fairlie and D. Bergman at the Centre for
Drug Design and Development, University of Queensland,
Brisbane, Australia, for performing the HIV-1 protease
inhibitor assays. Partial funding of this research was through
a New Zealand PGSF grant and a New Zealand Royal
Society of Chemistry Marsden grant.
4
Å molecular sieves (50 mg) (15) (58 mg, 0.36 mmol) and
t-butylhyroperoxide (TBHP) in CH Cl (0.24 mL of a 3 M solution,
2
2
0.72 mmol). Workup and purification gave (17) as a colourless oil (42
1
mg, 66%), with identical H NMR data to (16).
(
2R,3S)-N-Cbz-L-Valine 2,3-Epoxy-5-phenylpentyl Ester (10)
The preparation was performed as described for (9) using (17) (16 mg,
.09 mmol). Purification of the residue by radial chromatography on
0
References
silica (75% light petroleum, 25% ethyl acetate) gave (10) as a colourless
oil (14 mg, 38%) (Found: 411.2051. C H NO requires 411.2046). H
NMR (CDCl ) δ 7.19–7.36, 10H, m, ArH; 5.28, 1H, d, J 8.8 Hz, CHNH;
5
dd, J 3.9, 11.7 Hz, CHCH O; 4.05, dd, 1H, J 6.8, 11.7 Hz, CHCH O;
3
1
[1] B. E. Maryanoff, A. B. Reitz, Chem. Rev. 1989, 89, 863.
[2] J. Boutagy, R. Thomas, Chem. Rev. 1974, 74, 87.
[3] W. C. Still, C. Gennari, Tetrahedron Lett. 1983, 24, 4405.
[4] D. Wadsworth, O. Schupp, E. Seus, J. Ford, J. Org. Chem. 1965,
30, 680.
24
29
5
3
.11, 2H, br s, CO CH Ph; 4.34, 1H, dd, J 4.9, 9.3 Hz, CHNH; 4.15, 1H,
2 2
2
2
.16, 1H, m, CHCH O; 3.07, 1H, m, CH CH CH; 2.87, 1H, m,
2
2
2
PhCH CH ; 2.74, 1H, m, PhCH CH ; 2.19, 1H, m, CH(Me) ; 1.87, 2H,
[5] E. Breuer, D. Bannet, Tetrahedron Lett. 1977, 13, 1141.
2
2
2
2
2

3
96
D. C. Martyn, D. A. Hoult and A. D. Abell
[
[
6] S. K. Thompson, C. H. Heathcock, J. Org. Chem. 1990, 55, 3386.
7] I. C. Popoff, J. L. Dever, G. R. Leader, J. Org. Chem. 1969, 34,
128.
8] G. Etemad-Moghadam, J. Seyden-Penne, Tetrahedron 1984, 40,
153.
9] K. Ando, Tetrahedron Lett. 1995, 36, 4105.
[17] D. A. Bergman, D. Alewood, P. F. Alewood, J. L. Andrews, R. I.
Brinkworth, D. R. Englebretsen, S. B. H. Kent, Lett. Pept. Sci.
1995, 2, 99.
1
[
[
[18] C. S. Lee, N. Choy, C. Park, H. Choi, Y. C. Son, S. Kim, J. H. Ok,
H. Yoon, S. C. Kim, Bioorg. Med. Chem. Lett. 1996, 6, 589.
[19] E. Jacobsen, L. Deng, Y. Furukawa, L. Martinez, Tetrahedron
1994, 50, 4323; C. Pouchet, The Aldrich Library of NMR Spectra
2nd Edn 1983 (Aldrich Chemical Company, Inc.: Wisconsin); J.
Rodriguez, B. Waegell, Synthesis 1988, 7, 534; M. Orfanopoulos,
I. Smonou, C. Foote, J. Am. Chem. Soc. 1990, 112, 3067; J.
Gajewski, R. Weber, R. Braun, M. Manion, B. Hymen, J. Am.
Chem. Soc. 1977, 99, 816; E. Breuer, D. Bannet, Tetrahedron
1978, 34, 997; K. Tamao, T. Kaikui, M. Kumada, Tetrahedron
Lett. 1979, 7, 619; S. Matsui, Bull. Chem. Soc. Jpn 1984, 57, 426;
D. McGreer, N. Chui, Can. J. Chem. 1996, 61, 947; B. Gung, M.
Yanik, J. Org. Chem. 1996, 61, 947.
5
[
[
10] H. Nagaoka, Y. Kishi, Tetrahedron 1981, 37, 3873.
11] T. D. Meek, B. D. Dayton, B. W. Metcalf, G. B. Dreyer, J. B.
Strickler, J. G. Gorniak, M. Rosenberg, M. L. Moore, V. W.
Magaard, C. Debouck, Proc. Natl Acad. Sci. U.S.A. 1989, 86,
1841.
[
12] S. K. Grant, M. L. Moore, S. A. Fakhoury, T. A. Tomaszek, T. D.
Meek, Bioorg. Med. Chem. Lett. 1992, 2, 1441.
13] C. Park, J. S. Koh, Y. C. Son, H. Choi, C. S. Lee, N. Choy, K. Y.
Moon, W. H. Jung, S. C. Kim, H. Yoon, Bioorg. Med. Chem. Lett.
[
1995, 5, 1843.
[
14] Z. Yu, P. Caldera, J. J. De Voss, P. R. Jones, A. L. Burlingame, I.
D. Kuntz, C. S. Craik, P. R. Ortiz de Montellano, J. Am. Chem.
Soc. 1996, 118, 5846.
[20] R. A. Barrow, T. Hemscheidt, J. Liang, S. Paik, R. E. Moore, M.
A. Tius, J. Am. Chem. Soc. 1995, 117, 2479.
[21] M. Alcon, M. Poch, A. Moyano, M. A. Pericas, A. Riera,
Tetrahedron: Asymmetry 1997, 8, 2967.
[15] A. D. Abell, D. A. Hoult, D. A. Bergman, D. P. Fairlie, Bioorg.
Med. Chem. Lett. 1997, 7, 2853.
[16] J. T. Pinhey, M. J. Stoermer, J. Chem. Soc., Perkin Trans. 1 1991,
2455.