Five-membered ring analogues of shikimic acid

Article abstract of DOI:10.1021/jo001121k

Shikimate and other intermediates of the shikimate-chorismate pathway are densely functionalized structures that seem to offer limited options for skeletal modification. We designed and synthesized cyclopentylidenes 1 and 2, as well as cyclopentenes 3 and 4, as novel ring-contracted analogues of shikimic acid. Enzymatic studies showed that analogues 1-3 are indeed processed by shikimate kinase to give phosphates 1-P, 2-P, and 3-P as five-membered ring analogues of shikimate-3-phosphate. In particular, analogue 1 is converted by the enzyme at a rate only 3.5-fold slower than that of the native substrate, while analogue 3 binds to shikimate kinase with an apparent K_m of 1.7 mM, compared to 0.14 mM for shikimate.

Full text of DOI:10.1021/jo001121k

1326
J . Org. Chem. 2001, 66, 1326-1333
F ive-Mem ber ed Rin g An a logu es of Sh ik im ic Acid
Ming An, Tab Toochinda, and Paul A. Bartlett*
†
Center for New Directions in Organic Synthesis, Department of Chemistry, University of California,
Berkeley, California 94720-1460
paul@fire.cchem.berkeley.edu
Received J uly 24, 2000
Shikimate and other intermediates of the shikimate-chorismate pathway are densely functionalized
structures that seem to offer limited options for skeletal modification. We designed and synthesized
cyclopentylidenes 1 and 2, as well as cyclopentenes 3 and 4, as novel ring-contracted analogues of
shikimic acid. Enzymatic studies showed that analogues 1-3 are indeed processed by shikimate
kinase to give phosphates 1-P , 2-P , and 3-P as five-membered ring analogues of shikimate-3-
phosphate. In particular, analogue 1 is converted by the enzyme at a rate only 3.5-fold slower than
that of the native substrate, while analogue 3 binds to shikimate kinase with an apparent Km of
1.7 mM, compared to 0.14 mM for shikimate.
In tr od u ction
The shikimate-chorismate pathway constitutes the
initial stage of the biosynthesis of aromatic compounds.1
The enzymes in this pathway are unique to plants and
some microorganisms, and thus they are important
targets for herbicide and antibiotic development.2 Shiki-
mate and other substrates of these enzymes are small,
densely functionalized molecules that pose unique syn-
thetic challenges for developing enzyme inhibitors. One
notable difficulty is the tendency of the native cyclohex-
ene-based structures to aromatize.3 Therefore, we became
interested in altered ring templates that are less prone
to aromatization yet are still accepted by the enzymes
in the shikimate-chorismate pathway. Such novel skel-
etons may potentially provide a basis for the design of
more synthetically accessible inhibitors. Furthermore,
equally intriguing to us is the prospect of mimicking the
cyclohexene-based intermediates with ring altered ana-
logues.
Berchtold and co-workers have studied substrate ana-
logues of chorismate in which the ring skeleton is altered
electronically, as in dihydropyran 5,4 or structurally, as
in cycloheptadiene 6.5 The ring-expanded analogue 6 is
not processed by chorismate mutase, anthranilate syn-
thase, or p-aminobenzoate synthase; however, it is a
competitive inhibitor of the reactions they catalyze.5
* To whom correspondence should be addressed.
† The Center for New Directions in Organic Synthesis is supported
by Bristol-Myers Squibb as a Sponsoring Member.
(1) Reviews: (a) Haslam, E. Shikimic Acid Metabolism and Me-
tabolites; Wiley: New York, 1993. (b) Weiss, U.; Edwards, J . M. The
Biosynthesis of Aromatic Compounds; Wiley-Interscience: New York,
1987; pp 260-270.
(2) (a) Sikorski, J . A.; Gruys, K. J . Acc. Chem. Res. 1997, 30, 2-8.
(b) Roberts, F.; Roberts, C. W.; J ohnson, J . J .; Kyle, D. E.; Krell, T.;
Coggins, J . R.; Coombs, G. H.; Milhous, W. K.; Tzipori, S.; Ferguson,
D. J . P.; Chakrabarti, D.; McLeod, R. Nature 1998, 393, 801-805.
(3) For a recent example, see: Kozlowski, M. C.; Tom, N. J .; Seto,
C. T.; Sefler, A. M.; Bartlett, P. A. J . Am. Chem. Soc. 1995, 117, 2128-
2140.
We proposed cyclopentylideneacetates 1 and 2, as well
as the cyclopentenyl structures 3 and 4, as ring-
contracted analogues of shikimic acid. Replacing the
cyclohexene template of shikimate with a five-membered
ring unit represents a drastic but calculated change in
structure. For example, the low energy conformers of
shikimate and the cyclopentylidene analogues 1 and 2
can be superimposed by their hydroxyl oxygens and
carboxyl carbon with rms differences of only 0.42 and 0.26
Å, respectively (Figure 1a).6 The electrostatic surfaces of
these three molecules also resemble each other closely
(4) Delany, J . J ., III.; Padykula, R. E.; Berchtold, G. A. J . Am. Chem.
Soc. 1992, 114, 1394-1397.
(5) Pawlak, J . L.; Berchtold, G. A. J . Org. Chem. 1988, 53, 4063-
4069.
10.1021/jo001121k CCC: $20.00 © 2001 American Chemical Society
Published on Web 01/23/2001

Analogues of Shikimic Acid
J . Org. Chem., Vol. 66, No. 4, 2001 1327
Sch em e 1
undesirable dihydro derivative 11 as the major product
(20-30% yield), with the desired cyclopentenylidene 8a
isolated in low yield (5-13%) as an approximate 1:1 E/ Z
mixture (Scheme 1).10 Similar results were obtained using
trimethyl phosphonoacetate instead of the phosphorane
ylide (not shown). Formation of the dihydro compound
11 presumably begins with 1,4-addition to the enone (7a
f 9a ), which is precedented for reaction of both phos-
phonate and phosphorane ylides.11 Intramolecular trans-
fer of the phosphorus moiety (9a f 10) then activates
the hydroxyl group for â-elimination (10 f 11).12
However, when phosphorus transfer is prevented by
protection of the hydroxyl group as the benzoate ester in
7b, treatment with the phosphorane ylide affords another
unexpected product: the fully conjugated 4-oxo-2-cyclo-
pentenylidene ester 13, as a 2:1 mixture of E/Z isomers
F igu r e 1. (a) Superpositions of shikimate (shown in gray and
red) and analogues 1 (blue, top stereopair) and 2 (green); (b)
comparison of electrostatic surfaces.
(Figure 1b).7 We designed the two cyclopentylidenes 1
and 2 to best mimic the low energy conformation of
shikimate while allowing cyclopentenes 3 and 4 to retain
some degree of flexibility. In this study, we describe the
syntheses of analogues 1-4 and their enzymatic evalu-
ations as alternative substrates for shikimate kinase.
Resu lts
Syn th eses of An a logu es 1-4. In the most straight-
forward fashion, we planned to build the cyclopentylide-
neacetate skeleton in the target structures 1 and 2 from
a direct Wittig-type olefination of the known prostaglan-
din intermediate 4-hydroxycyclopent-2-en-1-one 7a.8 How-
ever, treatment of cyclopentenone 7a with carbethoxy-
triphenylphosphorane9 under various conditions gave the
(10) The stereochemistry of the exocyclic double bond in 8a was
determined from 1H NMR analysis and NOE difference NMR experi-
ments. NOE’s were observed from Hd to Hc in the E isomer and from
Hd to Ha/Hb in the Z isomer. The chemical shift of the vinyl proton Hc
in the Z isomer was found to be 1.03 ppm downfield from that in the
E isomer as a consequence of the magnetic anisotropy of the cis ester
carbonyl group. A similar anisotropic effect was observed for the ring
methylene protons Ha and Hb in the E isomer.
(6) Structures were minimized with MacroModel 7.0’s MMFF94s
forcefield (Halgren, T. A. J . Comput. Chem. 1996, 17, 490.) and
superimposed using the program Cerius2 4.0.
(7) Charges were calculated with MacroModel 7.0’s MMFF94s
forcefield and displayed with the program GRASP (A. Nicholls and B.
Honig, Columbia University).
(8) Racemic enone 7a was prepared from 2-methylfuran in 39% yield
over two steps (see Experimental Section for details).
(9) The ylide was prepared from triphenyl phosphine and ethyl
bromoacetate in 84% yield over 2 steps, see: Denney, D. B.; Ross, S.
T. J . Org. Chem. 1962, 27, 998-1000.
(11) (a) Freeman, J . P. J . Org. Chem. 1966, 31, 538-541. (b)
Bergmann, E. D.; Sulomonovici, A. Tetrahedron 1971, 27, 2675-2678.
(c) Wadsworth, W. S., J r. Org. React. 1977, 25, 73.

1328 J . Org. Chem., Vol. 66, No. 4, 2001
An et al.
in 54% yield.13 In this instance, 1,4-addition (7b f 9b)
is followed by tandem â-eliminations (9b f 12 f 13).
Indeed, triphenylphosphine was isolated from reaction
of benzoate 7b, whereas triphenylphosphine oxide was
the byproduct from reaction of alcohol 7a .
Sch em e 2
Although our original Wittig strategy was based on
literature precedents,14 we had not anticipated the roles
that oxygen substituents at the 4-position would play.
Nevertheless, enone 13 proved to be a useful product,
affording the key intermediate 8a via Luche reduction.15
This alternative sequence proved to be advantageous in
several ways: it affords 8a in better overall yield than
the Wittig reactions and in a more favorable E/ Z ratio.
Furthermore, the unwanted Z isomer 13z can be isomer-
ized to the desired E isomer in the presence of a trace
amount of acid. Although this isomerization reaches
equilibrium at 1:1 E/ Z ratio, the E and Z isomers of
enone 13 are readily separated by column chromatogra-
phy, and nearly all of 13z can be converted to 13e by
recycling the unwanted isomer. As a result, 4-hydroxy-
cyclopentenone 7a can be converted to the stereochemi-
cally pure E isomer of cyclopentenylidene 8a in four steps
in >40% overall yield on a gram scale. In contrast, no
such isomerization is observed for the reduced product
8a itself, and the separation of the E/ Z isomers of
compound 8a proved to be difficult. Enone ester 13 can
also be prepared from a direct olefination of 4-cyclopen-
tene-1,3-dione using the phosphorane ylide, but only in
poor yield (7-10%).
Sch em e 3
Hydroxy ester 8a served as the key intermediate in
synthesis of all the analogues. En route to 1 and 3,
epoxidation of the endocyclic double bond with m-CPBA
gives the cis product 14 in 70-82% yield as a single
diastereomer (Scheme 2). Treatment of 14 with aqueous
sulfuric acid opens the epoxide at the allylic position to
afford triol ester 15 in 92% yield. Both cyclopentylidene
analogue 1 and cyclopentene analogue 3 can be obtained
from triol ester 15 in a single step under basic conditions.
When the saponification is carried out in water, analogue
(()-1 is isolated as the only product in 90% yield.
Alternatively, when a 2:1 water/methanol mixture is used
as the solvent, isomerization competes with hydrolysis
to give the cyclopentene analogue (()-3 as the major
product. The extent of cyclopentylidene to cyclopentene
isomerization seems to increase with the content of
methanol in the solvent mixture up to the 2:1 ratio: a
higher methanol content, and thus a higher proportion
of methoxide, leads to side reactions that lower the
(12) In fact, enolate â-addition to a 4-acyloxycyclopentenone followed
by â′-elimination of the carboxylate group is a known method to
generate 4-substituted cyclopentenones. (see: (a) Koksal, Y.; Oster-
thun, V.; Winterfeldt, E. Liebigs Ann. Chem. 1979, 1300-1308. (b)
Harre, M.; Raddatz, P.; Walenta, R.; Winterfeldt, E. Angew. Chem.,
Int. Ed. Engl. 1982, 21, 480-492. (c) West, F. G.; Gunawardena, G.
U. J . Org. Chem. 1993, 58, 5043-5044.) We are not aware of an
instance in which the incoming nucleophile provides the activation for
â′-elimination of a hydroxyl substituent, as occurs in the phosphorus
transfer steps (cf. 9a f 10).
combined yield of analogues 1 and 3, while lower metha-
nol content gives more of analogue 1.
To obtain the stereoisomeric analogues 2 and 4, allylic
alcohol 8a is first converted to the tert-butyldiphenylsilyl
ether 8b, then dihydroxylated selectively from the less
hindered side using a catalytic amount of OsO4 (Scheme
3). The desired diol 16a is isolated as the predominant
diastereomer in 60-63% yield. The diastereomeric excess
for this substrate-directed dihydroxylation is >92%, since
the all-cis diastereomer is isolated in only 1-3% yield.
This reaction must be carried out under the Sharpless
biphasic conditions if the desired regioselectivity (ring
double bond vs the exocyclic unsaturation) is to be
obtained and overoxidation avoided.16 Subsequently,
(13) Similar transformations from 4-acetoxycyclopentenone to 4-oxo-
2-cyclopentenylidene products have been reported with sulfoxide and
sulfone ylides, albeit with yields of 2% (see ref 12a) and 24% (see: Ito,
M.; Kodama, A.; Hiroshima, T.; Tsukida, K. J . Chem. Soc., Perkin
Trans. 1. 1986, 905-907.), respectively. To our knowledge, the reaction
of 7b f 13 is the first example involving a phosphorane ylide.
(14) For examples involving phosphorous ylides, see: (a) Nakayama,
M.; Ohira, S.; Shinke, S.; Matsushita, Y.; Matsuo, A.; Hayashi, S. Chem.
Lett. 1979, 1245-1246. (b) Tanaka, K.; Uchiyama, F.; Ikeda, T.;
Inubushi, Y. Chem. Pharm. Bull. 1983, 31, 8-1971. (c) Novak, L.;
Rohaly, J .; Galik, G.; Fekete, J .; Varjas, L.; Szantay, C. Liebigs Ann.
Chem. 1986, 509-524.
(16) Kolb, H. C.; Vannieuwenhze, M. S.; Sharpless, K. B. Chem. Rev.
1994, 94, 2483.
(15) Luche, J . L. J . Am. Chem. Soc. 1978, 100, 2226.

Analogues of Shikimic Acid
J . Org. Chem., Vol. 66, No. 4, 2001 1329
F igu r e 3. Conversion of shikimate and analogues 1-3 to
phosphorylated products.
give phosphates 1-P , 2-P , and 3-P , respectively, as ring
contracted analogues of S-3-P; second, the reactions for
racemates 1-3 are complete at 50% conversion, reflecting
clean enzymatic resolution of these alternative substrates
(Figure 3);18 and third, no phosphorylation is detected
for analogue 4 after 100 days of incubation with SK.
F igu r e 2. Enzymatic resolution of analogue 1, as monitored
by H NMR.
1
when diol 16a is stirred in 5 M ethanolic HCl, the silyl
group is removed smoothly to afford the triol ester 16b,
which is then hydrolyzed in aqueous K2CO3 to give the
cyclopentylidene analogue (()-2 in 90% yield over two
steps.
Isomerizing the exocyclic double bond in ester 16b into
the thermodynamically favored17 endocyclic position would
provide analogue 4 upon hydrolysis (cf. 15 f 3). However,
in contrast to the ease with which this isomerization can
be accomplished with the stereoisomer 15, all attempts
to do so with 16b in various alkaline aqueous alcohol
solutions were unsuccessful. In all cases, the cyclopen-
tylidene 2 was isolated as the only product. These results
were somewhat perplexing, since triol 16b differs from
triol 15 only in the configuration of the middle hydroxyl
group.
An alternative route was devised to force the cyclo-
pentylidene to cyclopentene isomerization (Scheme 3).
The cis diol in compound 16b is protected to give
acetonide 17, which is then deprotonated with excess
LDA to generate the alkoxide/ester-enolate dianion. On
acidification of the reaction mixture, R-protonation gives
the desired cyclopentenyl acetonide 18 in 47% yield, along
with enol ether 19 in 16% yield. Finally, heating a
mixture of acetonide 18 and sulfonic acid resin in water
removes all protecting groups to afford analogue (()-4
in 81% yield.
E n zym a t ic E va lu a t ion s of R a cem ic An a logu es
1-4. Shikimate kinase (SK) phosphorylates the 3-hy-
droxyl group of shikimate selectively to give shikimate-
3-phosphate (S-3-P) at the expenditure of one equivalent
of ATP.1 When racemic analogues 1-4 were treated with
type II SK (EC 2.7.1.71) from Escherichia coli K12,
monitoring of these enzymatic reactions by 1H NMR
immediately revealed three important results (for ex-
ample, see Figure 2): first, cyclopentylidenes 1 and 2,
as well as cyclopentene 3, are phosphorylated by SK to
To isolate the phosphorylated products 1-P , 2-P , and
3-P , the enzymatic reactions were carried out on 30-60
mg scale. After SK-catalyzed phosphorylation, reaction
mixtures were treated with apyrase to degrade all
residual ATP and byproduct ADP to AMP and inorganic
phosphate. Subsequent anion exchange chromatography
gave the phosphates as the triethylammonium salts,
which were further purified via reverse phase HPLC to
afford enantiomerically pure phosphates 1-P , 2-P , and
3-P in 79%, 78%, and 88%19 yields, respectively.
Stea d y-Sta te En zym e Kin etics Assa ys. Analogues
1-3 were further evaluated in steady-state kinetics
assays. E. coli SK was assayed in the forward direction
at 25 °C by coupling the release of ADP to the reactions
catalyzed by pyruvate kinase (ADP + phosphoenolpyru-
vate f ATP + pyruvate) and lactate dehydrogenase
(pyruvate + NADH f lactate + NAD+).20 This phospho-
(18) The data points in Figure 3 were obtained from SK-catalyzed
phosphorylations carried out with 0.1 mmol of the racemic analogues
1-3 and 0.05 mmol of (-)-shikimate at 50 mM [substrate], 50 mM
[ATP], and 16 mM [Mg2+], using 2 units of SK (pH 7.1-7.4).
(19) The yield for phosphate 3-P is that of the triethylammonium
salt after anion exchange.
(20) Kinetics assays were performed according to the reported
conditions except that the pH was changed to 7.8, see: Millar, G.;
Lewendon, A.; Hunter, M. G.; Coggins, J . R. Biochem. J . 1986, 237,
427.
(17) Allinger, N. L.; Dodziuk, H.; Rogers, D. W.; Naik, S. N.
Tetrahedron 1982, 38, 1593.

1330 J . Org. Chem., Vol. 66, No. 4, 2001
An et al.
Ta ble 1. Kin etic P a r a m eter s for Sh ik im a te a n d
An a logu es 1 a n d 3
weakly and turned over more slowly. These analogues
represent the first examples of alternative substrates for
the shikimate-chorismate pathway in which the ring
size has been altered. Moreover, their activity in the SK-
catalyzed phosphorylation reaction demonstrates the first
step in establishing an unnatural “biosynthetic” sequence
based on these five-membered ring analogues. The next
enzyme in the pathway, EPSP synthase, catalyzes the
reaction of S-3-P with phosphoenolpyruvate to give
5-enolpyruvylshikimate-3-phosphate (EPSP).1 It will be
interesting to see if the ring-contracted S-3-P analogues
1-P , 2-P , and 3-P can serve as alternative substrates of
EPSP synthase as well.
substrate
Km (mM)
kcat (s-1
)
kcat/Km (M-1s-1) × 103
shikimic acid 0.14 ( 0.02
8.23
2.41
0.238
58.0
1
3
11 ( 1
1.7 ( 0.2
0.214
0.144
rylation-dependent oxidation of NADH (A340 ) 6180 M-1
cm-1) is monitored by following the decrease in absor-
bance at 340 nm. The apparent kcat and Km values for
analogues 1 and 3 are compared with those of the native
substrate shikimate in Table 1.21 SK-catalyzed phospho-
rylation of cyclopentylidene 2 is too slow to be character-
ized kinetically even though, from the modeling point of
view, it appeared to be the most faithful mimic of the
low energy conformer of shikimate.
Exp er im en ta l Section
Gen er a l. Unless otherwise noted, chemicals were obtained
from commercial sources and were used without further
purification. Chromatography employed 60-mesh silica gel
from E. Merck & Co. and was performed according to the
method of Still, Kahn, and Mitra.25 Unless otherwise specified,
CDCl3 was the NMR solvent. NMR chemical shifts are
referenced as follows, depending on solvent. 1H: to internal
TMS or residual CHCl3 (7.27 ppm), H2O (4.80 ppm), or CD2-
HOD (3.31 ppm). 13C: CDCl3 (77.23 ppm) or MeOH-d4 (49.15
ppm). 31P: internal trimethyl phosphate (3.086 ppm).
At 2.5 mM ATP and pH 7.8, the apparent Km for
shikimate was found to be 140 µM, compared to the
reported value of 200 µM at 5 mM ATP.22 Under the same
conditions, cyclopentenyl analogue 3 binds to SK with
an apparent Km of 1.7 mM, versus 11 mM for cyclopen-
tylidene analogue 1. The kcat values for analogues 1 and
3 are 2.41 s-1 and 0.238 s-1, respectively, compared to
that of 8.23 s-1 for the native substrate. Hence, the
cyclopentenyl analogue 3 binds to SK with 6-fold higher
affinity than does the cyclopentylidene analogue 1, while
the latter is turned over 10 times faster, resulting in
similar kcat/Km values for both ring-contracted analogues.
(()-(E )-[2(R *),3(R *),4(R *)]-Tr ih yd r oxycyclop e n t yli-
d en ea cetic Acid (1). A mixture of 71 mg (0.35 mmol) of triol
ester 15, described below, and 20 mL of 0.5 M aq NaOH (10
mmol, 29 equiv) was stirred at room temperature for 1 h, then
acidified with 1 M aq HCl. After lyophilization to remove
water, the residue was chromatographed (7:1:1:1 EtOAc/
Discu ssion
1
MeOH/dH2O/acetone) to afford 55 mg (90%) of triol acid 1. H
NMR (MeOH-d4) δ 2.74-2.90 (m, 1), 2.97-3.13 (m, 1), 3.67
(dd, 1, J ) 4.4, 8.8), 4.08 (t, 1, J ) 4.3), 4.50 (td, 1, J ) 2.3,
8.9), 5.93-5.97 (m, 1). 13C NMR (MeOH-d4) δ 37.8, 70.4, 78.5,
79.2, 116.2, 164.4, 170.3. IR (neat) 3361 (vbr), 2926, 1695, 1405,
1332, 1217, 1119, 1038, 875 cm-1. HRMS (FAB+) MH+ Calcd.
for C7H11O5: 175.0606. Found: 175.0605.
We designed these five-membered ring analogues of
shikimate in the absence of any knowledge of the
structure of substrate-bound SK.23 Our design adopted
two divergent approaches: first, constraint of cyclopen-
tylidenes 1 and 2 to best mimic the low energy conforma-
tion of the native substrate; second, retention of flexibility
in cyclopentenyl analogues 3 and 4. In the event, we
obtained two alternative substrates of equal efficacy
(similar kcat/Km values for 1 and 3) but with considerably
different characteristics.
Cyclopentenyl analogue 3 binds to SK with the highest
affinity among the ring-contracted analogues. This result
is interesting considering the fact that the acetate side-
chain in cyclopentene 3 would have to adopt a high
energy, ring-eclipsed conformation to mimic the native
substrate. However, tighter ground-state binding is not
maintained at the transition state in this case, since kcat
for analogue 3 is relatively poor. It is the cyclopentylidene
1, which has a lower ground-state affinity, that forms
the most productive enzyme complex among all ana-
logues. In fact, at 50 mM ATP and saturating substrate
concentrations (50 mM), analogue 1 is turned over even
faster than shikimate itself (see Figure 3).24
(()-(E )-[2(R *),3(S*),4(R *)]-Tr ih yd r oxycyclop e n t yli-
d en ea cetic Acid (2). A solution of 45 mg (0.22 mmol) of triol
ester 16b (described below) and 304 mg (2.2 mmol, 9.9 equiv)
of K2CO3 in 5 mL of water was stirred at 60 °C for 25 h. The
reaction mixture was acidified with 5 mL of 1 M aq HCl and
lyophilized, and the residue was chromatographed (7:1:1:1
EtOAc/MeOH/dH2O/acetone) to give 37 mg (95%) of triol acid
1
(()-2. H NMR (MeOH-d4) δ 2.62-2.71 (m, 1), 3.17-3.27 (m,
1), 3.83-3.88 (m, 1), 4.06-4.11 (m, 1), 4.62-4.66 (m, 1), 5.95-
5.98 (m, 1). 13C NMR (MeOH-d4) δ 38.4, 74.3, 76.4, 77.9, 116.3,
165.9, 170.5. IR (neat) 3333 (vbr), 2929, 1699, 1408, 1218, 1112,
1046 cm-1. HRMS (FAB-) (M - H+)- Calcd. for C7H9O5:
173.0450. Found: 173.0454.
(()-[3(R*),4(R*),5(R*)]-Tr ih yd r oxy-1-cyclop en ten ea ce-
tic Acid (3). To a solution of 25 mg (0.12 mmol) of triol ester
15 (described below) in 1 mL MeOH were added 2 mL of 1 M
aq NaOH (2 mmol, 16 equiv). The reaction mixture was stirred
at room temperature for 10 min, acidified with 1 M aq HCl,
and lyophilized. The residue contained 14 mg (65%) of cyclo-
pentene acid 3 and 5 mg (23%) of cyclopentylidene acid 1,
which were separated via chromatography (7:1:1:1 EtOAc/
MeOH/dH2O/acetone; spontaneous acetonide formation on the
column was sometimes observed for compound 3). 1H NMR
(MeOH-d4) δ 3.08 (s, 2), 3.89 (t, 1, J ) 5.3), 4.43-4.47 (m, 1),
4.57-4.60 (m, 1), 5.73-5.76 (m, 1). 13C NMR (MeOH-d4) δ 38.2,
73.2, 80.2, 81.6, 129.5, 146.5, 179.4. IR (neat) 3324 (vbr), 2922,
1574, 1391, 1091 cm-1. HRMS (FAB-) (M - H+)- Calcd. for
C7H9O5: 173.0450. Found: 173.0444.
In addition to analogues 1 and 3, cyclopentylidene 2
is also phosphorylated by SK, although it is bound more
(21) All steady-state enzyme kinetics parameters were calculated
using the EnzymeKinetics (v1.11) program and the reported molecular
weight of 18 93720 for E. coli K12 aroL SK II.
(22) Feyter, R. D.; Pittard, J . J . Bacteriol. 1986, 165, 331.
(23) A crystal structure of SK from Erwinia chrysanthemi in complex
with shikimate was reported in 1998, see: Krell, T.; Coggins, J . R.;
Lapthorn, A. J . J . Mol. Biol. 1998, 278, 983-997. This SK is also type
II and shares 53% sequence analogy with E. coli SKII. However, there
was not enough electron density to allow an unambiguous positioning
of shikimate in the active site.
(()-[3(R*),4(S*),5(R*)]-Tr ih yd r oxy-1-cyclop en ten ea ce-
tic Acid (4). A mixture of 12 mg (0.05 mmol) of protected triol
ester 18 (described below), 200 mg of Bio-Rad AG 50W-X4
(24) It is known that SK activity is inhibited by high levels of
shikimate, see: Feyter, R. D. Methods Enzymol. 1987, 142, 355-361.
(25) Still, W. C.; Kahn, M.; Mitra, S. J . Org. Chem. 1978, 43, 2923-
2925.

Analogues of Shikimic Acid
J . Org. Chem., Vol. 66, No. 4, 2001 1331
resin, and 4 mL of water was stirred at 95 °C for 3 h. The
reaction mixture was filtered, and the aqueous solution was
lyophilized to give a colorless residue, which was chromato-
graphed (7:1:1:1 f 6:2:1.25:1 EtOAc/MeOH/dH2O/acetone) to
cally stirred mixture of 18.2 g (222 mmol) of 2-methylfuran,
47.0 g (443 mmol, 2.02 equiv) of Na2CO3, and 170 mL of MeOH
at -55 °C was added a solution of 35.4 g (222 mmol, 1.0 equiv)
of bromine in 12 mL of CH2Cl2 over a 1-h period. The reaction
mixture was allowed to warm to 10 °C over 3 h, then the
mixture was filtered through Celite. The filtrate was mixed
with 200 mL of brine and extracted with CH2Cl2 (one 200-mL
portion followed by five 100-mL portions). The combined
organic fractions were dried, filtered, and evaporated under
reduced pressure to give 25.9 g (82%) of 2,5-dihydro-2,5-
dimethoxy-2-methylfuran (mixture of stereoisomers) as a
colorless oil, which was used in the subsequent reaction
1
afford 7 mg (81%) of the triol acid (()-4. H NMR (MeOH-d4)
δ 3.08 (d, 1, J ) 15.8), 3.17 (d, 1, J ) 15.8), 3.86 (dd, 1, J )
3.8, 5.5), 4.52 (d, 1, J ) 5.5), 4.56 (s, 1), 5.62 (s, 1). 13C NMR
(MeOH-d4) δ 40.2, 77.0, 80.8, 81.8, 132.7, 143.9, 179.3. IR
(neat) 3297 (vbr), 2919, 1574, 1389, 1111 cm-1. HRMS (FAB-)
(M - H+)- Calcd. for C7H9O5: 173.0450. Found: 173.0449.
(+)-(E)-O-4-P h osp h o-[2(R),3(S),4(R)]-tr ih yd r oxycyclo-
p en tylid en ea cetic Acid (1-P ). To a solution of 52 mg of
racemic triol acid 1 (0.3 mmol, 2 equiv), 3.6 mg of MgCl2 (0.038
mmol, 0.25 equiv), and 151 mg of ATP‚Na2‚3H2O (0.25 mmol,
1.68 equiv) in deuterated triethanolamine‚HCl/NaOH buffer
(2.5 mL, 520 mM, pH ) 9.8), was added 4 units of SK. After
the reaction mixture was allowed to stand at room temperature
for 187 h with occasional mixing, 12 units of apyrase was
added. The resulting mixture was incubated at room temper-
ature for 112 h, diluted with 500 mL of water, titrated with 1
M aq NaOH to pH 8.8, then subjected to anion exchange
chromatography (DEAE Sephadex A-25, 0 f 1 M NEt3H+-
HCO3- buffer, pH ) 8.1). Fractions were lyophilized, and the
desired product was detected by 1H NMR. A total of 36 mg
(95%) of the desired cyclopentylidene phosphate 1-P anion was
obtained in various forms of triethylammonium salts and as
a mixture with AMP. The crude product was further purified
by RP-HPLC with a water (0.1% TFA) to 99:1 water (0.1%
TFA)/acetonitrile (0.1% TFA) gradient over 20 min at a flow
rate of 10 mL/min (UV detection at 210 nm, product Rt ) 6
min, AMP Rt ) 12 min). All fractions corresponding to the
product peak were combined and lyophilized to give 30 mg
(79%) of the desired phosphate 1-P in the acid form as a
colorless residue. 1H NMR (D2O) δ 3.02-3.22 (m, 2), 3.86 (ddd,
1, J ) 1.8, 4.3, 9.3), 4.51-4.68 (m, 2), 5.94-6.03 (m, 1). 13C
NMR (10:1 D2O/MeOH-d4) δ 36.6 (d, J ) 1.5), 75.2 (d, J )
5.3), 77.58 (d, J ) 5.3), 77.61, 115.8, 163.2, 170.9. 31P NMR
1
without further purification. H NMR major isomer δ 1.49 (s,
1
3), 3.17 (s, 3), 3.48 (s, 3), 5.47 (s, 1), 5.9-6.0 (m, 2). H NMR
minor isomer 1.55 (s, 3), 3.10 (s, 3), 3.40 (s, 3), 5.76 (s, 1), 5.9-
6.0 (m, 2). 13C NMR δ 25.7, 26.1, 49.3, 49.9, 54.2, 55.5, 106.7,
107.7, 111.5, 112.7, 129.7, 130.1, 134.0, 134.4.
For the second step, a modified version of the patented
procedure27 was kindly provided by Ingrid Choong (UC Ber-
keley): To a stirred, deoxygenated solution of 10.0 g (69.4
mmol) of 2,5-dihydro-2,5-dimethoxy-2-methylfuran and 50 mg
(0.45 mmol) of hydroquinone in 278 mL of water at 0 °C was
added 0.84 g (14 mmol, 0.20 equiv) of acetic acid. After 1 h, a
solution of 18.6 g (69.4 mmol, 1.0 equiv) of Na2HPO4‚7H2O in
60 mL of water was added, and the mixture was heated to 60
°C for 2 h. Nonpolar side-products were removed by washing
with 100 mL of hexanes, the aqueous layer was extracted with
1-butanol (seven 100-mL portions), and the combined 1-bu-
tanol fractions were concentrated under reduced pressure to
give a crude brown oil, which was purified by chromatography
(1:1 hexanes/EtOAc, Rf ) 0.09) to afford 3.18 g (47%) of
1
cyclopentenone 7 as a yellow oil. H NMR δ 2.27 (dd, 1, J )
2.0, 18.6), 2.76 (dd, 1, J ) 6.0, 18.5), 4.25 (d, 1, J ) 6.2), 5.02-
5.05 (m, 1), 6.21 (d, 1, J ) 5.7), 7.63 (dd, 1, J ) 2.3, 5.7). 13C
NMR δ 44.1, 70.0, 134.5, 164.4, 207.8. Spectra obtained were
identical to those reported.28
Eth yl (E)- a n d (Z)-4-Oxo-2-cyclop en ten ylid en ea ceta te
(13e,z). Rou te 1: F r om (()-4-Ben zoyloxy-2-cyclop en ten -
on e (7b). To a solution of 1.90 g (9.4 mmol) of benzoyloxycy-
clopentenone 7b29 in 40 mL of toluene was added 3.37 g (9.7
mmol, 1.03 equiv) of carbethoxytriphenylphosphorane, and the
reaction mixture was stirred at 60 °C for 8 h. After toluene
was removed under reduced pressure, the residue was chro-
matographed (9:1 f 1:1 hexanes/EtOAc) to give 569 mg (37%)
of cyclopentenone 13e and 260 mg (17%) of cyclopentenone 13z
as individual isomers, as well as 154 mg (6%, 1:1 E/ Z ratio)
of the direct Wittig product (benzoate of 8a ); traces of triph-
enylphosphine oxide were removed by additional chromatog-
raphy (19:1 CH2Cl2/acetone).
20
(D2O) δ 0.96. [R]D ) +59.6° (c ) 2.11, H2O). HRMS (FAB-)
(M - H+)- Calcd. for C7H10O8P: 253.0113. Found: 253.0121.
(+)-(E)-O-2-P h osp h o-[2(R),3(S),4(R)]-tr ih yd r oxycyclo-
p en tylid en ea cetic Acid (2-P ). As described above for isomer
1, 61 mg of racemic triol acid 2 (0.35 mmol) was phosphory-
lated and purified to give 34 mg (78%) of phosphate 2-P in
1
the acid form as a colorless residue. H NMR (D2O) δ 2.68 (d,
1, J ) 20.5), 3.17-3.32 (m, 1), 4.12 (d, 1, J ) 4.1), 4.21 (d, 1,
J ) 6.6), 5.04-5.16 (m, 1), 6.07 (d, 1, J ) 2.3). 13C NMR (10:1
D2O/MeOH-d4) δ 37.0, 73.4, 76.1 (d, J ) 2.7), 79.3 (d, J ) 5.2),
116.7, 162.4 (d, J ) 6.0), 170.8. 31P NMR (162 MHz, D2O) δ
20
0.90. [R]D ) +26.5° (c ) 1.02, dH2O). HRMS (FAB+) MH+
Less polar isomer 13z: 1H NMR δ 1.32 (t, 3, J ) 7.2), 3.04
(d, 2, J ) 1.2), 4.23 (q, 2, J ) 7.2), 5.89-5.94 (m, 1), 6.54 (dd,
1, J ) 1.7, 5,9), 8.86 (dd, 1, J ) 0.8, 5.9). 13C NMR δ 14.4,
39.9, 60.8, 115.9, 140.0, 151.5, 154.8, 165.4, 204.0. IR (neat)
Calcd. for C7H12O8P: 255.0270. Found: 255.0264.
(-)-O-3-P h osp h o-[3(R ),4(S),5(R )]-t r ih yd r oxy-1-cyclo-
p en ten yl Aceta te Tr isod iu m Sa lt (3-P ). As described above
for isomer 1, 30 mg of racemic triol acid 3 (0.172 mmol) was
phosphorylated and purified to give 19 mg of crude product
(88%) after anion exchange chromatography. This crude
product was dissolved in water, titrated with 1 M aq NaOH
to pH 12, and then purified by RP-HPLC using water as the
solvent (UV detection at 191 nm, product Rt ) 5.6 min, AMP
Rt ) 7 min). Fractions containing the product were combined
and titrated with 0.1 M aq NaOH to pH 9, then lyophilized to
give the desired phosphate 3-P as the trisodium salt. 1H NMR
(D2O) δ 3.01-3.14 (m, 2), 4.02 (t, 1, J ) 5.3), 4.63 (d, 1, J )
4.4), 4.88-4.98 (m, 1), 5.81 (s, 1). 13C NMR (10:1 D2O/MeOH-
3083, 2979, 2914, 1718, 1645, 1382, 1310, 1226, 1145 cm-1
.
MS (EI+) m/z 166 (M+). Anal. Calcd. for C9H10O3: C, 65.05;
H, 6.07. Found: C, 65.16; H, 6.17.
More polar isomer 13e: For elemental analysis, 13e was
further purified to a white powder via sublimation; mp 49.5-
50.0 °C. 1H NMR δ 1.32 (t, 3, J ) 7.2), 3.38 (d, 2, J ) 1.7),
4.23 (q, 2, J ) 7.2), 6.00-6.01 (m, 1), 6.55 (dd, 1, J ) 0.6, 5.6),
7.79 (d, 1, J ) 5.5). 13C NMR δ 14.4, 39.4, 60.8, 116.8, 139.3,
152.8, 158.0, 165.8, 205.2. IR (KBr) 3069, 2988, 1708, 1649,
1239, 906 cm-1. MS (EI+) m/z 166 (M+). Anal. Calcd. for
C9H10O3: C, 65.05; H, 6.07. Found: C, 65.05; H, 6.32.
Rou te 2: By Isom er iza tion of Eth yl (Z)-4-Oxo-2-cyclo-
p en ten ylid en ea ceta te (13z). To a solution of 40 mg (0.24
mmol) of cyclopentenone 13z in 30 mL of chloroform was added
20 µL of concentrated HCl. The reaction mixture was stirred
d4) δ 38.5, 76.8 (d, J ) 5.2), 79.3 (d, J ) 5.9), 81.3, 128.0 (d, J
) 2.7), 146.9, 180.2. 31P NMR (162 MHz, D2O) δ 5.17. [R]D
)
20
-84.8° (c ) 0.28, dH2O). HRMS (FAB-) (M-3Na++2H+)-
Calcd. for C7H10O8P: 253.0113. Found: 253.0104.
(()-4-Hyd r oxy-2-cyclop en ten on e (7a ). Enone 7 was pre-
pared from 2-methylfuran in two steps. For the first step, a
modified version of Clauson-Kaas’ procedure26 was kindly
provided by Lorin Thompson (UC Berkeley): To a mechani-
(27) Tanaka, T. J pn. Pat. 54-163510; J apan Kokai 56-86128.
(28) Paquette, L. A.; Earle, M. J .; Smith, G. F. Org. Synth. 1995,
73, 39.
(29) Alcohol 7a was treated with benzoyl chloride and pyridine in
CH2Cl2 to give the known benzoate 7b in 87% yield. (Dols, P. P. M. A.;
Klunder, A. J . H.; Zwanenburg, B. Tetrahedron 1994, 50, 8515-8538.)
(26) Clauson-Kaas, N. Kgl. Danske Vidensk. Selsk. Math.-Fys. Medd.
1947, 22, 6.

1332 J . Org. Chem., Vol. 66, No. 4, 2001
An et al.
at room temperature for 3 d and washed with 20 mL of water.
The aqueous portion was further extracted with CH2Cl2 (two
10-mL portions), and the combined organic fractions were dried
and concentrated under reduced pressure to give 40 mg (100%)
of 13e,z in a 1:1 ratio. The E/ Z isomers were separated by
chromatography (4:1 hexanes/EtOAc).
(()-Eth yl (E)-4(R*)-t-Bu tyldiph en ylsiloxy-[2(R*),3(R*)]-
d ih yd r oxycyclop en tylid en ea ceta te (16a ). To a solution of
1.176 g (2.897 mmol) of silyl ether 8b in 20 mL of tert-butyl
alcohol was added 1.57 g of 2.5 wt % OsO4 (39.3 mg, 0.155
mmol, 0.05 equiv) in tert-butyl alcohol, followed by 116 mg
(1.04 mmol, 0.36 equiv) of DABCO and a solution of 3.05 g
(9.27 mmol, 3.2 equiv) of K3Fe(CN)6 and 1.28 g (9.27 mmol,
3.2 equiv) of K2CO3 in 20 mL of water. The biphasic reaction
mixture was stirred at room temperature for 50 h, and then
3.9 g of Na2SO3 was added. The resulting mixture was stirred
at room temperature for 3 h, diluted with 100 mL of water,
and extracted with EtOAc (six 100-mL portions). All EtOAc
fractions were combined, dried, and concentrated under re-
duced pressure, and the residue was chromatographed (7:3
hexanes/EtOAc) to afford 805 mg (63%) of the desired triol 16a
and 35 mg (3%) of the all-cis triol. For 16a : 1H NMR δ 1.04
(s, 9), 1.26 (t, 3, J ) 7.1), 2.05-2.09 (m, 1), 2.43 (d, 1, J ) 8.7),
2.77-2.87 (m, 1), 3.09-3.21 (m, 1), 3.83-3.89 (m, 1), 4.14 (q,
2, J ) 7.1), 4.16-4.21 (m, 1), 4.79-4.87 (m, 1), 6.04 (q, 1, J )
2.4), 7.33-7.48 (m, 6), 7.57-7.69 (m, 4). 13C NMR δ 14.5, 19.3,
27.1, 38.2, 60.1, 74.9, 75.6, 76.7, 115.9, 128.0, 130.1, 133.5,
133.8, 135.9, 163.6, 166.7. IR (neat) 3434 (br), 2931, 2858, 1716,
1699, 1428, 1203, 1112, 822, 701 cm-1. MS (FAB+) m/z 441
(MH+). Anal. Calcd. for C25H32O5Si: C, 68.15; H, 7.32. Found:
C, 67.86; H, 7.55.
(()-Eth yl (E)-4-Hyd r oxy-2-cyclop en ten ylid en ea ceta te
(8a ). To a solution of 0.31 g (1.87 mmol) of enone 13e and 0.836
g (2.24 mmol, 1.2 equiv) of CeCl3‚7H2O in 50 mL of MeOH at
-3 °C was added slowly over 2 min 85 mg (2.25 mmol, 1.2
equiv) of NaBH4. The reaction mixture was stirred at room
temperature for 30 min, mixed with 10 mL of 0.5 M aq HCl (5
mmol, 2.7 equiv) and 25 mL of brine, and extracted with ether
(five 30-mL portions). The combined ether fractions were dried
and concentrated under reduced pressure, and the crude
product was purified via chromatography (7:3 hexanes/EtOAc)
to give 296 mg (94%) of allylic alcohol 8a as a viscous, colorless
1
oil. H NMR δ 1.29 (t, 3, J ) 7.1), 1.78-1.80 (m, 1), 2.78 (td,
1, J ) 2.1, 19.4), 3.45 (ddd, 1, J ) 2.2, 6.5, 19.3), 4.18 (q, 2, J
) 7.1), 4.99-5.08 (m, 1), 5.80 (t, 1, J ) 2.2), 6.39 (d, 1, J )
5.4), 6.52 (dd, 1, J ) 2.3, 5.4). 13C NMR δ 14.2, 39.9, 59.8, 75.2,
110.8, 135.7, 147.5, 162.9, 167.4. IR (neat) 3417 (br), 3060,
2981, 2935, 2903, 1698, 1633, 1580, 859 cm-1. HRMS (EI+)
M+ Calcd. for C9H12O3: 168.0786. Found: 168.0789.
(()-E t h yl (E)-4-t-Bu t yld ip h en ylsiloxy-2-cyclop en t e-
n ylid en ea ceta te (8b). To a solution of 410 mg (2.44 mmol)
of alcohol 8a in 3 mL of DMF were added 581 mg (8.54 mmol,
3.5 equiv) of imidazole and 2 g (7.31 mmol, 3 equiv) of tert-
butyldiphenylsilyl chloride. The reaction mixture was stirred
at room temperature for 2 h, then diluted with 50 mL of brine
and 50 mL of ether. The aqueous layer was extracted with
ether (one 50 mL-portion followed by two 25-mL portions), the
combined organic layers were washed with water, dried, and
concentrated under reduced pressure, and the residue was
chromatographed (9:1 hexanes/Et2O) to give 939 mg (95%) of
(()-Eth yl (E)-[2(R*),3(S*),4(R*)]-Tr ih yd r oxycyclop en -
tylid en ea ceta te (16b). A solution of 46 mg (0.104 mmol) of
silyl ether 16a in 2 mL of 5 M ethanolic HCl was stirred at
room temperature for 62 h. After ethanol was removed under
reduced pressure, the crude product was chromatographed
(EtOAc f 10:1 EtOAc/MeOH) to give 20 mg (95%) of triol 16b
1
as a colorless film. H NMR (MeOH-d4) δ 1.28 (t, 3, J ) 7.1),
2.60-2.73 (m, 1), 3.21 (tdd, 1, J ) 2.5, 6.2, 20.2), 3.84-3.88
(m, 1), 4.08 (d, 1, J ) 6.3), 4.15 (q, 2, J ) 7.1), 4.62-4.68 (m,
1), 5.94-6.00 (m, 1). 13C NMR (MeOH-d4) δ 14.8, 38.4, 61.1,
74.3, 76.5, 77.8, 115.5, 166.5, 168.5. IR (neat) 3370 (br), 2923,
1696, 1666, 1213, 1042 cm-1. MS (EI+) m/z 202 (M+). Anal.
Calcd. for C9H14O5: C, 53.46; H, 6.98. Found: C, 53.11; H, 7.29.
(()-Eth yl (E)-[2(R*),3(S*)]-O-Isop r op ylid en e-4(R*)-h y-
d r oxycyclop en t ylid en ea cet a t e (17). A mixture of 91 mg
(0.45 mmol) of triol ester 16b and 0.75 g of Bio-Rad AG 50W-
X4 resin in 5 mL of acetone was stirred at room temperature
for 20 h, then diluted with 10 mL of EtOAc and filtered. After
solvent was removed under reduced pressure, the colorless
residue was chromatographed (1:1 hexanes/EtOAc and 2:1
EtOAc/MeOH) to afford 78 mg of acetonide 17 (94% overall
yield) as a colorless oil along with 22 mg of recovered starting
1
8b as a colorless, viscous oil. H NMR δ 1.07 (s, 9), 1.26 (t, 3,
J ) 7.1), 2.90 (td, 1, J ) 2.2, 19.0), 3.34 (ddd, 1, J ) 2.0, 6.3,
19.0), 4.16 (q, 2, J ) 7.2), 4.95-5.00 (m, 1), 5.72 (t, 1, J ) 2.2),
6.23 (d, 1, J ) 5.7), 6.26 (dd, 1, J ) 1.9, 5.5) 7.30-7.50 (m, 6),
7.64-7.78 (m, 4). 13C NMR δ 14.5, 19.2, 25.4, 40.5, 59.8, 77.0,
110.8, 127.8, 127.9, 129.9, 133.7, 134.2, 135.3, 135.8, 135.9,
147.6, 162.8, 167.4. IR (neat) 3070, 2958, 2931, 2893, 2857,
1703, 1634, 1214, 1110, 1064, 702 cm-1. MS (EI+) m/z 406
(M+). Anal. Calcd. for C25H30O3Si: C, 73.85; H, 7.44. Found:
C, 73.89; H, 7.66.
(()-Eth yl (E)-[2(S*),3(R*)]-Ep oxy-4(R*)-h yd r oxycyclo-
p en tylid en ea ceta te (14). To a solution of 281 mg (1.67 mmol)
of hydroxycyclopentenylidene 8a in 50 mL of toluene at 0 °C
was added slowly 1.08 g (3.57-5.44 mmol, 2.1-3.2 equiv) of
57-86% pure m-CPBA. The mixture was allowed to warm to
room temperature over 1 h and then was stirred for 23 h. After
toluene was removed under reduced pressure, the residue was
chromatographed (1:1 hexanes/EtOAc, Rf ) 0.26) to give 217
mg (70%) of epoxide 14 as a colorless oil. 1H NMR δ 1.27 (t, 3,
J ) 7.2), 2.11 (ddd, 1, J ) 3.0, 6.6, 18.4), 2.70-3.25 (br s, 1),
3.56 (ddd, 1, J ) 1.8, 8.1, 18.3), 3.72 (d, 1, J ) 2.7), 3.83-3.87
(m, 1), 4.11-4.22 (m, 2), 4.32-4.40 (m, 1), 5.99-6.02 (m, 1).
13C NMR δ 14.5, 34.2, 59.3, 60.5, 60.7, 71.7, 117.7, 156.4, 165.7.
IR (neat) 3413 (br), 2983, 1712, 1668, 1222, 857 cm-1. MS (EI+)
m/z 184 (M+). Anal. Calcd. for C9H12O4: C, 58.69; H, 6.57.
Found: C, 58.38; H, 6.52.
(()-Eth yl (E)-[2(R*),3(R*),4(R*)]-Tr ih yd r oxycyclop en -
tylid en ea ceta te (15). A mixture of 125 mg of epoxide 14 (0.68
mmol) and 5 mL of 0.75 M aq H2SO4 (3.75 mmol, 5.5 equiv)
was stirred at room temperature for 20 min, then neutralized
to pH 7 with 1 M NaHCO3. After water was removed via
lyophilization, the residue was chromatographed (EtOAc) to
yield 126 mg (92%) of triol ester 15. 1H NMR (MeOH-d4) δ
1.27 (t, 3, J ) 7.2), 2.77-2.90 (m, 1), 2.98-3.12 (m, 1), 3.67
(dd, 1, J ) 4.4, 8.9), 4.06-4.12 (m, 1), 4.15 (q, 2, J ) 7.2), 4.51
(td, 1, J ) 2.4, 9.0), 5.95 (q, 1, J ) 2.4). 13C NMR (MeOH-d4)
δ 14.8, 37.8, 61.1, 70.5, 78.5, 79.4, 115.6, 164.9, 168.3. IR (neat)
3366 (br), 2925, 1710, 1455, 1207, 1113 cm-1. MS (FAB+) m/z
203 (MH+). Anal. Calcd. for C9H14O5: C, 53.46; H, 6.98.
Found: C, 53.37; H, 7.28.
1
material. H NMR δ 1.22-1.32 (m, 3), 1.34 (s, 3), 1.41 (s, 3),
2.18-2.32 (br s, 1), 2.90-3.06 (m, 1), 3.25 (d, 1, J ) 18.2), 4.17
(q, 2, J ) 7.1), 4.37 (d, 1, J ) 4.7), 4.47 (d, 1, J ) 5.3), 4.95 (d,
1, J ) 5.3), 6.09 (s, 1). 13C NMR δ 14.3, 24.8, 26.8, 37.1, 60.4,
73.9, 82.3, 84.3, 111.9, 119.5, 160.5, 166.7. IR (neat) 3466 (br),
2986, 2933, 1716, 1374, 1212, 1158, 1039, 861 cm-1. MS (EI+)
m/z 242 (M+). Anal. Calcd. for C12H18O5: C, 59.49; H, 7.49.
Found: C, 59.61; H, 7.84.
(()-Eth yl [4(S*),5(R*)]-O-Isopr opyliden e-3(R*)-h ydr oxy-
1-cyclop en ten ea ceta te (18). To a solution of 300 µL (2.14
mmol, 10.6 equiv) of dry diisopropylamine in 2 mL of dry THF
in a flame-dried flask under Ar at -70 °C was slowly added
600 µL of 1.6 M n-BuLi in hexanes (0.96 mmol, 4.7 equiv),
and the mixture was stirred for 10 min. A solution of 49 mg
(0.20 mmol) of acetonide 17 in 2 mL of dry THF was added
slowly over 10 min. The reaction mixture was allowed to warm
to -35 °C over 20 min, mixed with 5 mL of 1 M aq HCl, and
partitioned between 30 mL of brine and 25 mL of ether. The
aqueous portion was extracted with ether, and the combined
organic layer was dried and concentrated under reduced
pressure to give 45 mg of yellow residue, which was chromato-
graphed (1:1 hexanes/EtOAc) to afford 23 mg (47%) of the
desired isomerized acetonide 18 and 9 mg (16%) of the enol
ether 19. For 18: 1H NMR δ 1.28 (t, 3, J ) 7.1), 1.35 (s, 3),
1.36 (s, 3), 2.09-2.17 (m, 1), 3.13-3.30 (m, 2), 4.17 (q, 2, J )
7.1), 4.53 (d, 1, J ) 5.5), 4.70 (s, 1), 5.23 (d, 1, J ) 5.5), 5.74 (s,
1). 13C NMR δ 14.4, 26.3, 27.6, 34.3, 61.2, 80.0, 85.0, 86.3,
112.1, 130.9, 142.5, 170.7. IR (neat) 3414 (br), 2986, 2935, 1738,

Analogues of Shikimic Acid
J . Org. Chem., Vol. 66, No. 4, 2001 1333
1372, 1248, 1208, 1157, 1035, 872 cm-1. HRMS (FAB+) MLi+
Calcd. for C12H18O5Li: 249.1314. Found: 249.1319.
were initiated by the addition of SK. Observed activity was
corrected against control runs in which the particular sub-
strate was omitted.
SK Stea d y-Sta te En zym e Kin etics Assa ys for Sh ik i-
m a te, An a logu e 1, a n d An a logu e 3. Escherichia coli K12
aroL shikimate kinase II (EC 2.7.1.71) was kindly provided
by Dr. Kenneth Gruys at Monsanto. The enzyme concentration
was 7.6 mg/mL (280/205 UV absorption method)30 with a
specific activity of 26.1 units/mg (one unit of enzyme activity
is defined as the amount of enzyme catalyzing the conversion
of 1 µmol of substrate/min at 25 °C), estimated using Michae-
lis-Menten Vmax values projected by the EnzymeKinetics
(v1.11) program. The assay mixture comprised, in a final
volume of 1 mL at 25 °C, the following: 50 mM triethanolamine‚
HCl/KOH buffer, pH 7.8; 50 mM KCl; 5 mM MgCl2; 2.5 mM
ATP (di-Na+ salt); 1 mM phosphoenolpyruvate (mono-K+ salt);
0.1 mM NADH (di-Na+ salt); 1.75 units of pyruvate kinase;
and 2.5 units of lactate dehydrogenase. Assays with shikimate
as substrate were performed using 0.05 units of SK and
substrate concentrations from 0.05 to 0.4 mM. For analogues
1 (0.1-4 mM) and 3 (0.1-1 mM), 0.4 units and 0.8 units of
SK, respectively, were used for each assay. Assay reactions
Ack n ow led gm en t. We thank Dr. Kenneth Gruys
(Monsanto) for generous gifts of SK and Dr. Lorin
Thompson and Dr. Ingrid Choong (University of Cali-
fornia, Berkeley) for helpful discussions about the
preparation of 4-hydroxycyclopent-2-en-1-one, 7a . This
work was supported by the National Institutes of Health
(Grants GM-28965 and GM-30759) and a fellowship
from Glaxo Wellcome to M.A.
Su p p or tin g In for m a tion Ava ila ble: Detailed general
experimental conditions, a figure depicting the enzymatic
resolution of analogues 2 and 3, as monitored by 1H NMR,
and additional characterization material (1H, 13C, and 31P NMR
spectra). This material is available free of charge via the
Internet at http://pubs.acs.org.
(30) Scopes, R. K. Protein Purification: Principles and Practice, 3rd
ed.; Springer-Verlag: New York, 1994; pp 46-48.
J O001121K