Analogs of 1-phosphonooxy-2,2-dihydroxy-3-oxo-5-(methylthio)pentane, an acyclic intermediate in the methionine salvage pathway: A new preparation and characterization of activity with E1 enolase/phosphatase from Klebsiella oxytoca

Article abstract of DOI:10.1016/j.bmc.2004.05.002

The methionine salvage pathway allows the in vivo recovery of the methylthio moiety of methionine upon the formation of methylthioadenosine (MTA) from S-adenosylmethionine (SAM). The Fe(II)-containing form of acireductone dioxygenase (ARD) catalyzes the p

Full text of DOI:10.1016/j.bmc.2004.05.002

Bioorganic & Medicinal Chemistry 12 (2004) 3847–3855
Analogs of 1-phosphonooxy-2,2-dihydroxy-3-oxo-5-(methylthio)-
pentane, an acyclic intermediate in the methionine salvage pathway:
a new preparation and characterization of activity with
E1 enolase/phosphatase from Klebsiella oxytoca
Yalin Zhang,a Melissa H. Heinsen,a Milka Kostic,a Gina M. Pagani,c Thomas V. Riera,b
a,
Iva Perovic,a Lizbeth Hedstrom,b Barry B. Snidera andThomas C. Pochapsky
*
aDepartment of Chemistry, Brandeis University, 415 South Street, MS 015, Waltham, MA 02454-9110 USA
bDepartment of Biochemistry, Brandeis University, 415 South Street, MS 015, Waltham, MA 02454-9110 USA
cBiophysics Program, Brandeis University, 415 South Street, MS 015, Waltham, MA 02454-9110 USA
Received8 March 2004; revised3 May 2004; accepted4 May 2004
Available online 2 June 2004
Abstract—The methionine salvage pathway allows the in vivo recovery of the methylthio moiety of methionine upon the formation
of methylthioadenosine (MTA) from S-adenosylmethionine (SAM). The Fe(II)-containing form of acireductone dioxygenase
(ARD) catalyzes the penultimate step in the pathway in Klebsiella oxytoca, the oxidative cleavage of the acireductone 1,2-dihydroxy-
3-oxo-5-(methylthio)pent-1-ene (2) by dioxygen to give formate and 2-oxo-4-(methylthio)butyrate (3). The Ni(II)-boundform (Ni–
ARD) catalyzes an off-pathway shunt, forming 3-(methylthio)propionate (4), carbon monoxide, and formate. Acireductone 2 is
formed by the action of another enzyme, E1 enolase/phosphatase, on precursor 1-phosphonooxy-2,2-dihydroxy-3-oxo-5-methyl-
thiopentane (1). Simple syntheses of several analogs of 1 are described, and their activity as substrates for E1 enolase/phosphatase
characterized. A new bacterial overexpression system and purification procedure for E1, a member of the haloacid dehalogenase
(HAD) superfamily, is described, and further characterization of the enzyme presented.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
tane (1) to the keto-acidprecursor of methionine, 2-
oxo-4-(methylthio)butyrate (3). Two enzymes have been
identified in the bacterium Klebsiella oxytoca (American
Type Culture Collection (ATCC) strain 8724) that are
involvedin these transformations, E1 enolase/phospha-
tase1;2 and acireductone dioxygenase (ARD).3–6 E1 cat-
alyzes the dephosphorylation and enolization of 1 to
acireductone 2 (Scheme 1). Previous publications have
dealt with mechanistic questions regarding this enzyme,
and these investigations were aided by the use of desthio
analogs 1a–c of E1 substrate 1.1;7;8 These compounds are
also substrates for E1 (Scheme 2).
S-Adenosylmethionine (SAM, or AdoMet) is a critical
source of activatedmethyl groups in all organisms, and
also serves as a source of activatedC 3 aminopropyl units
for polyamine biosynthesis during the cell cycle. The
methionine salvage pathway is a ubiquitous biochemi-
cal pathway in prokaryotes andeukaryotes that re-
incorporates the methylthio group of the proximate
metabolite of SAM, methylthioadenosine (MTA) into
methionine. Our laboratories are engagedin structural
andfunctional studies of enzymes in the methionine
salvage pathway, particularly those enzymes that cata-
lyze the final steps of the pathway, the conversion of
1-phosphonooxy-2,2-dihydroxy-3-oxo-5-methylthiopen-
However, the reportedsyntheses of these compounds
were laborious, gave low overall yields and required the
use of diazomethane.7 We now report simple syntheses
of 1a–c, andprovide a comparative analysis of their
ability to act as substrates of E1 enolase/phosphatase.
We also provide further characterization of the E1
enzyme as obtainedin active form upon overexpression
in Escherichia coli.
Keywords: Acireductone dioxygenase; Enolase; Phosphatase; Phos-
phoramidite.
* Corresponding author. Tel.: +1-781-736-2559; fax: +1-781-736-2516;
e-mail: pochapsk@brandeis.edu
URL: http://www.chem.brandeis.edu/pochapsky
0968-0896/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.05.002

3848
Y. Zhang et al. / Bioorg. Med. Chem. 12 (2004) 3847–3855
a) (i-Pr) N-P(OBn)
O
2
2
E1 enolase/
phosphatase
O
O
PyTFA, CH Cl
2
2
S
OP(OH)2
b) 30% H O
2
2
O-
S
HO OH
O
R
OH
R
OH
OP(OBn)
2
1
2
5a, R = n-propyl
6a, R = n-propyl (93%)
6b, R = cyclopropyl (86%)
6c, R = phenyl (90%)
6d, R = benzyl (70%)
5b, R = cyclopropyl
5c, R = phenyl
Fe-ARD
O2
S
Ni-ARD
O2
5d, R = benzyl
O
O
.
NaIO , RuO H O
O-
+ HCOOH
4
2
2
O-
S
in CCl /CH CN/H O
4
3
2
4
O
TFA/PhSH/H O
O
2
O
O
+ CO + HCOOH
O
OP(OH)
3
R
OP(OBn)
2
R
2
Scheme 1. Metabolic disposition of 1 in the methionine salvage
pathway of K. oxytoca.
HO OH
O
1a, R= n-propyl (79%)
1b, R= cyclopropyl (70%)
1c, R= phenyl (77%)
7a, R= n-propyl (75%)
7b, R= cyclopropyl (71%)
7c, R= phenyl (89%)
7d, R= benzyl (0%)
O
O
O
R
O-
Scheme 3. Syntheses of 1a–c. Percent yields are noted for each com-
pound.
R
OP(OH)2
E1 enolase/
phosphatase
OH
HO OH
2a, R = n-propyl
2b, R = cyclopropyl
2c, R = phenyl
1a, R = n-propyl
1b, R = cyclopropyl
1c, R = phenyl
zole, which must be replacedwith CH Cl2 prior to the
2
oxidation step.
Scheme 2. Desthio analogs of 1 that are substrates for E1 enolase/
phosphatase.
Dibenzyl 2,3-dioxo-1-alkyl phosphates 7a–c are pre-
paredfrom compounsd
6a–c by catalytic ruthenium
oxide oxidation using sodium periodate as the stoichio-
metric oxidant.12 Oxidations are performed in a biphasic
CCl4/H2O mixture, with acetonitrile added to accelerate
the reaction.13 Although the benzyl-substitutedalkynyl
phosphate 6d is readily prepared, it cannot be success-
fully oxidized to the diketone. It is likely that the enol
form of the product diketone is stabilized by the phenyl
ring, andthat this is susceptible to further oxidation
under the reaction conditions. We also note that the
oxidation must be closely monitored by TLC to prevent
over-oxidation because the benzyl groups of the phos-
phate esters are slowly oxidized to benzaldehyde.
2. Results and discussion
2.1. Synthesis of E1 substrate analogs
The general methodfor synthesis of E1 substrate ana-
logs 1a–c is shown in Scheme 3. The starting materials
for all three analogs are the 2-alkyn-1-ols 5. Compounds
5a and 5c are commercially available. Alkynols 5b and
5d are preparedby deprotonation of the appropriate
terminal alkynes with n-butyllithium followedby
hydroxymethylation with paraformaldehyde at low
temperature. Reaction of 5a–d with dibenzyl N,N-
diisopropylphosphoramide in the presence of pyridi-
nium trifluoroacetate under nitrogen for 25 min provides
the intermediate phosphoramidite, which is oxidized
with hydrogen peroxide to generate the dibenzyl phos-
phate ester 6a–d. This procedure is an adaptation
of chemistry commonly usedfor polynucleotide syn-
thesis.9
Compounds 7a–c are deprotected by stirring with a
mixture of trifluoroacetic acid(TFA), thiophenol, and
water for 25 min.14 Sufficient water andmethylene
chloride are added to form an organic and an aqueous
layer; 1a–c partition into the aqueous layer.
13C NMR spectra of E1 substrate analogs 1a–c indicate
that, as isolated under acidic conditions, the 2-carbonyl
is hydrated to form the geminal dihydroxy compound,
as indicated by 31P coupling to the hydrated carbon at d
94–96. Such hydration has been observed for biacetyl in
aqueous solution, andis reversible upon transfer of bi-
acetyl into nonpolar solvents.15;16 The three substrate
analogs 1a–c are stable indefinitely as hydrates if stored
as the free phosphoric acidin frozen 0.1 M solutions in
H2O. Under slightly basic conditions (pH 7.5 in potas-
sium phosphate buffer), NMR spectra indicate that
compound 1a enolizes almost completely to give enol-
phosphate 8a (see Scheme 4). Equilibration occurs
within 48 h, with a steady loss of the 1-CH2 signal of the
1H-Tetrazole is commonly usedas a catalyst for the
phosphoramidite chemistry employed for the synthesis
of 6a–d.10 Presently, solidtetrazole is a restrictedcom-
pound. However, pyridinium trifluoroacetate (PyTFA)
is an effective catalyst for phosphoramidite chemistry.11
For these reactions, PyTFA proves to be a better cata-
lyst than 1H-tetrazole. PyTFA accelerates the reaction
considerably (25 min to completion, vs >12 h with 1H-
tetrazole in THF) andworkup is considerably simpli-
fied. The use of PyTFA with CH2Cl2 as the solvent is
compatible with hydrogen peroxide oxidation of the
phosphite, as opposedto the THF usedwith 1 H-tetra-

Y. Zhang et al. / Bioorg. Med. Chem. 12 (2004) 3847–3855
3849
E1 enolase/phosphatase, pH 7.5
chromatographic mobilities identical to the enzyme de-
2
< 1 min, 25 oC.
O
3
scribedby Balakrishnan et al., andshows activity with
the substrates described here equivalent to the enzyme
1
O
or no enzyme, pH 7.5, 24 hrs, @ 25 oC (1a only),
or no enzyme, pH 7.5, 20 min, @ 100 oC (1a only).
4
2
R
OP(OH2)
originally isolatedfrom K. oxytoca.1
HO OH
1
O
O
1
4
2
R
OP(OH)2
3
2.3. Activity of E1 enolase/phosphatase toward com-
pounds 1a–c
OH
8 (λ
=278 nm)
max
E1
Enzymatic activity assays of E1 enolase/phosphatase
with substrates 1a–c were performedanaerobically using
20 mM HEPES (pH 7.4) containing 0.5 mM MgCl2.
Typically, substrate concentration for the assay is
ꢀ200 lM andthe reaction is initiatedby the injection of
sufficient enzyme to reach a final enzyme concentration
of 75 nM. The reactions can be monitoredoptically, as
products 2a and 2b have a characteristic optical
absorption maximum at ꢀ307 nm. For the phenyl-
O
1
4
2
O-
3
OH
(1, 2, 8)a, R= n-propyl
(1, 2, 8)b, R= cyclopropyl
(1, 2, 8)c, R= phenyl
2 (λ
=305 nm)
max
Scheme 4. Proposedsequence of E1 enzymatic activity with substrate
1a.
substitutedanalog
2c the maximum is shiftedto
hydrate at d 3.98 andthe appearance of a new signal at d
ꢀ318 nm. In the absence of any enol, there is no sig-
nificant UV absorbance for any of the substrates under
the conditions of the assay with the exception of 1c,
which exhibits the expectedphenyl absorbance maxima
near 226 and248 nm. We note that phosphate buffers
are not suitable for this assay because we findevidence
for significant inhibition of the dephosphorylation
reaction by phosphate, which differs from the previous
report (Fig. 1).1 Primary amine buffers such as Tris
should be avoided in studies of enzyme activity in this
system, as the total amount of product formed is lower
than is observedwith phosphate or HEPES buffer sys-
tems, possibly because primary amines are capable of
reacting with the intermediate and/or product that is
releasedfrom the enzyme (vide infra).
1
7.47. This signal shows an 8 Hz H–31P coupling, con-
firming that it is the 1-CH enol proton of 8a. Under the
same conditions, the hydrated carbonyl 13C signal of 1a
at d 95.5 is completely lost, as is the 1-C signal at d 68.5.
Two new signals are observedfor 8a at 136.5 (3JPC
9 Hz) and139.7 ( JPC ¼ 6 Hz) corresponding to the 2-C
and1-C enol carbons, respectively. The 3-C carbonyl
signal at d 212.0 in 1a shifts upfieldto d 200.9 in 8a.
¼
2
In the course of the enolization experiment in D2O, the
4-CH2 proton signals of 1a (andtherefore 8a) are rap-
idly lost (i.e., within several hours) due to exchange with
solvent, andthe 4- CH2 signal at d 38 is severely
2
broadened due to H coupling. The 1-CH signal of 8a is
also attenuatedby 1H/2H exchange, with about half of
the expectedintensity still observedat d 7.47 after 48 h.
However, the 1-C signal of 8a at d 139.7 is still clearly
observedafter 48 h, indicating that 1H/2H exchange at
this position is considerably slower than at 4-C. Storage
of 1a as a 0.1 M solution at neutral pH for several
months, even frozen, also results in considerable iso-
merization to 8a, andseveral minutes of heating in a
boiling water bath at neutral pH results in formation of
8a as well.7
The results of the assays are shown in Figure 2. Product
formation is monitoredat 320 nm rather than 307 nm, as
the intermediate enol-phosphate has an absorption at
kmax ꢀ278 nm for all three substrates, andtotal absor-
bance at 307 nm contains contributions from the inter-
mediate. Extinction coefficients for the acireductone
products of substrates 1a–c are similar, e320 ¼ 8000 MÀ1
cmÀ1. The kinetics of the reaction are complicatedby the
presence of the intermediate enol–phosphate 8, which is
releasedfrom the enzyme active site andbuilsd up
transiently in the course of the reaction. It is not yet
clear whether the two-step reaction (keto-gem-diol-
phosphate 1 to enol-phosphate 8 to acireductone 2)
can occur without release of the intermediate enol–
phosphate 8 or if release of intermediate 8 is obligate
(Scheme 4). A pH profile of enzyme activity as measured
by the rate of formation of acireductone 2 shows a
maximum near pH 7.8. At pH 5.0, the 307 bandis much
weaker than the 278 bandat all times, suggesting that
phosphatase activity is suppressedto a greater extent
than enolase activity at lower pH. A more complete
study of the mechanism of E1 enolase/phosphatase
activity is underway.
2.2. Characterization of bacterially overexpressed E1
enolase/phosphatase
E1 enolase/phosphatase was first identified as a com-
ponent of the methionine salvage pathway in K. oxytoca
ATCC 8724 (formerly K. pneumoniae),1 andthe gene
corresponding to E1 was cloned by Balakrishnan and
co-workers into pACYC184 to yielda construct pDiox-
2, which gives constitutive overexpression of E1 in E.
coli strain MM294.2 For our purposes, including NMR
characterization of E1, we wishedto obtain controllable
overexpression of E1 andfor that reason chose to use
the reliable pET vector expression system. We placedthe
E1 gene into a pET3a expression vector, andthe
resulting construct, pTP01, gives goodoverexpression of
the desired enzyme in E. coli BL21(DE3)(pLysS). The
over-expressedenzyme exhibits electrophoretic and
Analogs 1a–c are all substrates for E1, suggesting that
the nature of the R group is of little importance in
determining substrate binding and reactivity with E1.
However, the similarity in rates of product formation

3850
Y. Zhang et al. / Bioorg. Med. Chem. 12 (2004) 3847–3855
1.4
1.2
1
0.8
0.6
0.4
0.2
0
BES
BTP
HEPES
KPi
MES
Tris
MOPS
PIPES
0
50
100
150
200
250
time (s)
Figure 1. Formation of 2a from 1a catalyzedby E1 enolase/phosphatase as a function of buffer type. Reaction at 25 ꢁC was monitoredby UV
absorbance at 305 nm. Concentration of all buffers was 50 mM, pH 7.5, with 0.25 mM MgCl2. Note that the reaction in potassium phosphate buffer
(KPi) is inhibitedsignificantly, andtotal product formedin Tris buffer is less than in other buffers. Enzyme concentration was 120 nM andsubstrate
concentration was 150 lM for all runs. Buffer names are standard abbreviations found in the Sigma-Aldrich catalogue.
1.2
A
1
0.8
0.6
1a
1b
1c
0.4
0.2
0
0
1
2
3
4
5
6
7
8
9
10
Time (min)
Figure 2. Activity of E1 enolase/phosphatase towardsubstrates 1a–c. All assays were performedat 25 ꢁC in 50 mM HEPES buffer containing 0.5 mM
CaCl2. For each assay, substrate was dissolved in buffer in an anaerobic cuvette and degassed by gentle purging with argon gas for 30 min. 2 lL of a
37 lM solution of E1 (80:20 glycerol/water) was added by syringe and mixed rapidly to give a 75 nM final concentration of enzyme. Dead time of the
assay was ꢀ10 s. Concentrations of the individual substrates were 184 lM 1a, 219 lM 1b, and152 lM 1c.
andabsence of lag time for the any of the substrates
should not be over-interpreted, since the dead time of
the current assay is ꢀ10 s. Stopped-flow measurements
are underway to characterize the kinetics of E1 activity
with these substrates more thoroughly.
catalyze reactions initiatedby nucleophilic attack on the
substrate phosphate group by the side chain carboxylate
of a strictly conservedaspartate residue, followedby
hydrolysis of the resulting anhydride intermediate.
Many members of HAD superfamily require divalent
metals (usually Mg2þ) for their activity, as does E1.1
Several crystal structures of HAD superfamily members
have been solved.18;20–24 The consensus HAD structure
contains two domains, a highly conserved core domain
that assumes is a=b hydrolase fold (also known as the
Rossmann fold) and a more diverse cap domain, with
the active site locatedbetween the two odmains. A
homology model of E1 is shown in Figure 3, a model
that is strongly supportedby the results of NMR
structural analysis currently underway (Milka Kostic,
unpublishedresults).
2.4. Homology modeling of E1 enolase/phosphatase
Sequence homology identifies E1 as a member of the
haloaciddehalogenase (HAD) superfamily of enzymes.
The HAD superfamily contains L-2-haloaciddehalo-
genases, epoxide hydrolases, numerous phosphatases
andphosphotransferases, andsuperfamily members
have been identified in both prokaryotic and eukaryotic
organisms.17–19 In general phosphatases in this family

Y. Zhang et al. / Bioorg. Med. Chem. 12 (2004) 3847–3855
3851
tions of different motifs organizations are given in Fig-
ure 4.
E1 is the only enzyme isolatedto date for which both
enolase andphosphatase activities have been estab-
lished. BLAST searching identifies a significant number
of putative or predicted enolase/phosphatases from
various organisms, ranging from prokaryotes to mam-
mals. Sequence alignment revealedthe presence of all
three HAD superfamily conservedmotifs. It has also
revealedtwo conservedregions in the long linker do-
main, from which diversity of activity has been shown to
arise in other members of HAD superfamily. We are
currently attempting to establish the specific residues
responsible for substrate specificity andenolase/phos-
phatase activity of E1.
3. Experimental
3.1. General
NMR spectra were recorded on a Varian Unity INOVA
400 MHz spectrometer operating at 399.75 MHz (1H)
and100.51 MHz ( 13C). Chemical shifts are reportedin d
with the solvent resonance as the internal standard
Figure 3. Model of E1 enolase/phosphatase based on sequence
homology with b-phosphoglucomutase (PDB entry 1LVH).22 Model-
ing was accomplishedusing the nest subroutine of the program
Jackal.28 The positions of proposedactive site residues are shown and
labeled. These correspond to the strongly conserved residues shown in
Figure 4, andare color-codedto correspondto Figure 4. Figure was
1
(CDCl3: d 7.27 for H, d 77.0 for 13C). Spectra obtained
in D2O were referencedto the residual HDO line at
d
4.8. IR spectra were recorded on a Perkin–Elmer FTIR
spectrometer, andall absorption maxima are recorded
in cmÀ1. High resolution mass spectra were obtainedat
the University of California Riverside Mass Spectro-
metry facility. FAB spectra were obtainedon a
VG-ZAB instrument with running in negative ion mode
using a glycerol matrix. Desorption chemical ionization
(DCI) spectra were obtainedon a VG 7070 instrument
in positive ion mode using ammonia as proton source.
Purifiedyielsd are reportedwhere appropriate in
Scheme 3.
29
generatedusing MOLMOL.
Three conservedsequence motifs are foundin the HAD
superfamily. The first, motif I, is locatedclose to N-
terminus, andcontains the consensus sequence
hhhhDxDx[T/V][L/V]h, where h is a hydrophobic resi-
due (A, C, F, G, I, L, M, V, W or Y) and x can be any
residue (2). The first aspartate residue in this motif plays
a critical functional role in HAD enzyme superfamily,
acting as the nucleophile in enzyme catalyzedreac-
tions.25 In E1, motif I starts with Ala 4, with the strongly
conservedAsp 8 as a likely nucleophile involvedin
dephosphorylation and second conserved aspartate
conservatively replacedby a glutamate at position 10.
Conservedmotif II has the consensus sequence
hhhhhh[S/T]xx. The conservedserine/threonine (Ser 125
in E1) has been implicated in hydrogen bonding to the
substrate.18 Finally, conservedmotif III is locatednear
3.1.1. 3-Cyclopropyl-2-propyn-1-ol (5b). Under N2,
1.6 M n-butyllithium in hexane (9.72 mL, 15.5 mmol)
was added dropwise to a solution of cyclopropylacetyl-
ene (1.17 mL, 12.9 mmol) in anhydrous ethyl ether
(10 mL) at 0 ꢁC. The solution was stirredfor 30 min after
the addition was completed. To this solution, parafor-
maldehyde [(CH2O)n] (0.47 g, 15.5 mmol) was added.
The mixture was stirredfor 1 h at 0 ꢁC, andallowedto
warm to room temperature for another 1.5 h. 20% HCl
(5 mL) was added dropwise to the solution in an ice
bath. SaturatedNaHCO solution andH O were used
the C-terminus andis edfinedas K-(x)
18–30-(G/S)(D/
S)xxx(D/N)hhhh.18;20 In known structures, this motif
forms part of the active site, including residues respon-
sible for divalent metal ligation. The strictly conserved
lysine (Lys 159 in E1) forms a salt bridge to the nucleo-
philic aspartate of motif I.20–22 Lys 159 in E1 is separated
by a 22 amino acidlinker from conservedresiudes
Ser 183 andAsp 184 (the conservedG/S andD/S,
respectively). The conservedD/N residue of motif III is
conservatively replacedby Glu 188 in E1. Details of
primary sequence alignment andschematic representa-
3
2
to wash the organic layer. The aqueous layer was ex-
tractedwith diethyl ether, andthe organic layers were
combined, dried over Na2SO4, andconcentrated. Dis-
tillation (53–54 ꢁC at 2 Torr) gave compound 5b (0.77 g,
62%). The spectral data matched literature values:26
1
NMR (CDCl3) H 0.65–0.73 (m, 2H, cyclopropyl (cp)
CH2), 0.75–0.81 (m, 2H, cyclopropyl (cp) CH2), 1.22–
1.30 (m, 1H, cp CH), 1.55 (t (broad), 1H, J ¼ 3:6 Hz,

3852
Y. Zhang et al. / Bioorg. Med. Chem. 12 (2004) 3847–3855
Motif I
4AVLFDLDGVIT 108IKIALASAS 145KPAPD..165SIGLED----SQAGIQAI
7LILFDFDSTLV 93YVVAVVSGG 144KGEIL..162TVAVGD----GANDISMF
Motif II
Motif III
PGM
PSP
mSEH 5VAAFDLDGVLA 117FTTCIVTNN 160KPEPQ..180VVFLDD----FGSNLKPA
E-1
PAH
4AIVTDIEGTTS 119IDLYVYSSG 159KREAQ..179ILFLSD----IHQELDAA
8AVIFDWAGTTV 116IKIGSTTGY 156RPYPW..177MIKVGD----TVSDMKEG
EYA 489VFVWDLDETLI 664VNVLVTSTQ 699KIGHE..719YVVIGD----GNEEETAA
YKRX 7FIICDFDGTIT 93IPFYVISGG 151KPSVI..165IIMIGD----SVIDVEAA
HAD 6GIAFDLYGTLF 112LKLAILSNG 151KPDNR..172ILFVSS----NAWDATGA
Group A
T6P 121ALFLDYDGTLS 157FPTAIISGR 301KGKAV..321PIYVGD----DRTDEDAF
SP
9MIVSDLDHTMV
44SLLVFSTGR 174KGQAL..195TLACGD----SGNDAELF
41IGVVGGSDF 189KRYCL..204IYFFGDKTMPGGNDHEIF
35CYVVFASGR 191KGKAL..209IVVFGD----NENDLFMF
PMM
8LCLFDVDGTLT
P-TM 4VFVFDLDGTLL
Group B
DEM 167VAGFDLDGTLI 211YKLVIFTNQ 260 KPVTG..278PISIGD—-SIFVGDHAGR
HPIGP 5YLFIDRDGTLI
48YKLVMITNQ 105 KPKVK..125SYVIGD----RATDIQLA
Group C
Figure 4. Primary sequence alignment of several different members of HAD superfamily. Three conserved motifs of the HAD superfamily are color-
coded (purple, domain I, blue, domain II, and green, domain III). The three HAD main domain sequence distribution patterns, groups A, B and C
are listed separately (see text). The number of the first residue within a motif indicates the position of motifs within the sequence. Strictly conserved
residues are highlighted in red and those that express some variation are highlighted in green. Sequences are as follows: PGM, b-phosphoglucomutase
(Lactoccocus lactis); PSP, phosphoserine phosphatase (Methanococcus jannaschi); mSEH––N-terminal of mammalian soluble epoxide–hydrolase
(Mus musculus); E1, enolase/phosphatase (K. oxytoca); PAH, phosphonacetaldehyde hydrolase (Bacillus cereus); EYA, C-terminal of transcription
factor (Eya) (Drosophila melanogaster); YKRX, YkrX gene product (B. subtilis); HAD, L-2-haloaciddehalogenase ( Pseudomonas sp. YL); T6P,
trehalose-6-phosphate phosphatase (Arabidopsis thaliana); SP, sucrose phosphatase (A. thaliana); PMM, phosphomannomutase (Homo sapiens); P-
TM, phosphatase (Thermotoga maritima); DEM, DNA 50-kinase-30-phosphatase (H. sapiens); HPIGP, histidinol phosphatase/imidazole–glycerol
phosphate phosphatase (Escherichia coli).
OH), 4.22 (d, 2H, J ¼ 3:6 Hz, 1-CH2); 13C-0.6 (cp CH),
8.2 (2C, cp-CH2), 51.4 (1-CH2), 73.6, 89.7.
dropwise to the solution with cooling in an ice bath, and
the solution was stirredat room temperature for 10 min.
The mixture was washedwith water, andthe aqueous
layer extractedwith CH 2Cl2. The CH2Cl2 fractions were
combined, dried over Na2SO4, concentrated, and puri-
fiedby flash chromatography on silica gel (hexane/ethyl
acetate 2:1) to give 6a (3.40 g, 93% yield) as an oil: NMR
3.1.2. 4-Phenyl-2-butyn-1-ol (5d). Compoundwas
preparedin the same manner as 5b, starting with
3-phenyl-1-propyne. Product 5d (65% yield) was purified
by distillation (59–60 ꢁC, 2 Torr). Spectral data matched
literature values:27 NMR (CDCl3) 1H 3.64 (t, 2H,
J ¼ 2:0 Hz, 4-CH2), 4.32 (t, 2H, J ¼ 2:0 Hz, 1-CH2),
7.33 (m, 5H, ph CH); 13C 26.1 (4-CH2), 52.4 (1-CH2),
81.4, 85.0, 127.7 (ph CH), 128.9 (2C, ph–CH), 129.5
(2C, ph CH), 137.4 (ph C).
1
(CDCl3) H, 0.94 (t, 3H, J ¼ 7:6 Hz, 6-CH3), 1.48 (tq,
2H, J ¼ 7:6, 7.2 Hz, 5-CH2), 2.02 (tt, 2H, J ¼ 7:2,
3
2.4 Hz, 4-CH2), 4.64 (td, 2H, 2.4 Hz, JPH ¼ 9:6 Hz,
3
1-CH2), 5.07 (d, 4H, JPH ¼ 7:6 Hz, benzyl (bnz)–CH2),
7.35 (m, 10H, phenyl (ph)–CH); 13C 13.4 (6-CH3), 20.7,
2
21.7, 56.1 (2JPC ¼ 6 Hz, 1-CH2), 69.2 (2C, JPC ¼ 6 Hz,
bnz–CH2), 74.1 (3JPC ¼ 8 Hz, 2-yne-C), 88.8 (3-yne-C),
127.8 (2C), 128.4 (4C), 128.5 (4C), 135.7 (2C,
3JPC ¼ 7 Hz); IR (film) 2963, 2237, 1497, 1281,
1017 cmÀ1; Rf (silica gel TLC, 2:1 hexane/EtOAc) 0.40;
HRMS (DCI/NH3, pos. ion mode) MHþ calcd. for
C20H24O4P, 359.141223, obs. 359.141743.
3.1.3. 1-(Dibenzylphosphonooxy)-2-hexyne (6a). The
preparation of compound 6a is typical. A solution of
1.2 equiv of pyridinium trifluoroacetate (2.3 g,
12.2 mmol), 1.0 equiv (1.0 g, 10.2 mmol) of 2-hexyn-1-ol
and1.2 equiv dibenzyl N,N-diisopropylphosphoramide
(4.2 g, 12.2 mmol) in 10 mL of anhydrous CH2Cl2 was
stirredunder N
monitoredby TLC on silica gel using a 2% KMnO
solution spray for visualization. When all the starting
alcohol was consumed, 3.5 mL of 30% H2O2 was added
at 25 ꢁC for 25 min. The reaction was
3.1.4. 1-(Dibenzylphosphonooxy)-3-cyclopropyl-2-prop-
yne (6b). NMR (CDCl3) 1H 0.64–0.70 (m, 2H,
cp–CH2), 0.73–0.80 (m, 2H, cp–CH2), 1.18–1.28 (m, 1H,
cp–CH), 4.62 (dd, 2H, JPH ¼ 10:4 Hz, J ¼ 2 Hz,
2
4
3

Y. Zhang et al. / Bioorg. Med. Chem. 12 (2004) 3847–3855
3853
1-CH2), 7.30–7.38 (m, 10H); 13C-0.6 (cp–CH), 8.2
(cp–CH2), 56.1 (2JPC ¼ 5 Hz, 1-CH2), 69.3 (2JPC ¼ 5 Hz,
bnz–CH2), 72.8 (3JPC ¼ 6 Hz, 2-yne-C), 92.1 (3-yne-C),
127.9 (2C), 128.4 (4C), 128.5 (4C), 135.8 (2C,
3JPC ¼ 8 Hz); Rf (silica gel TLC plate, 2:1 hexane/
EtOAc) 0.31; IR (film) 2954, 2235, 1497, 1282,
1018 cmÀ1; HRMS (DCI/NH3, pos. ion mode) MHþ
calcd. for C20H22O4P, 357.125573, obs. 357.124391.
3.1.8.
oxopropane (7b). NMR (CDCl3) H 1.14–1.20 (m, 4H,
cp–CH2), 2.71–2.78 (m, 1H, cp–CH), 5.05 (d, 2H,
1-(Dibenzylphosphonooxy)-3-cyclopropyl-2,3-di-
1
3JPH ¼ 11:2 Hz, 1-CH2), 5.11 (dd, 2H, 2J ¼ 11 Hz,
2
3JPH ¼ 8 Hz, bzl–CH2), 5.15 (dd, 2H,
J ¼ 11 Hz,
3JPH ¼ 8 Hz, bzl–CH2), 7.32–7.38 (m, 10 H, ph–CH);
13C 14.2 (cp–CH2), 15.2 (cp–CH), 68.3 (2JPC ¼ 5 Hz, 1-
2
CH2), 69.7 (2C, JPC ¼ 5 Hz, bzl–CH2,), 128.0 (2C),
2
128.6 (4C), 128.6 (4C), 135.5 (2C, JPC ¼ 7 Hz), 190.1
(3JPC ¼ 10 Hz, 2-CO), 197.5 (3-CO); IR 2958, 1743,
1698, 1498, 1265, 1012 cmÀ1; HRMS (DCI/NH3, pos.
ion mode), calcd. for C20H22O6P, MHþ 389.115402, obs.
389.115100.
3.1.5.
1-(Dibenzylphosphonooxy)-3-phenyl-2-propyne
1
3
(6c). NMR (CDCl3) H 4.85 (d, 2H, JPH ¼ 10 Hz, 1-
CH2), 5.12 (d, 4H, 3JPH ¼ 8 Hz, bzl–CH2), 7.25–7.40 (m,
15H); 13C 55.9 (2JPC ¼ 5 Hz, 1-CH2), 69.3 (2JPC ¼ 5 Hz,
bnz–CH2), 82.7 (3JPC ¼ 7 Hz, 2-yne-C), 87.5 (3-yne-C),
121.7, 127.8 (2C), 128.2 (2C), 128.4 (2C), 128.4 (4C),
3.1.9. 1-(Dibenzylphosphonooxy)-3-phenyl-2,3-dioxopro-
pane (7c). NMR (CDCl3) 1H 5.10 (dd, 2H, 2J ¼ 11:6 Hz,
3
128.8 (4C), 131.7, 135.5 (2C, JPC ¼ 7 Hz); Rf (silica gel
3
3JPH ¼ 8 Hz, bzl–CH2), 5.12 (d, 2H, JPH ¼ 11:2 Hz, 1-
TLC plate, 2:1 hexane/EtOAc) 0.36; IR (film): 2955,
2234, 1491, 1280, 1019 cmÀ1; HRMS (DCI/NH3, pos.
ion mode: MHþ calcd. for C23H22O4P, 393.125573, obs.
393.127433.
3
CH2), 5.14 (dd, 2H, 2J ¼ 11:6 Hz, JPH ¼ 8 Hz, bzl–
CH2), 7.32–7.37 (m, 10 H, bzl–CH), 7.50 (dd, 2H, J ¼ 9,
8 Hz, m-ph–CH), 7.67 (tt, 1H, J ¼ 9 Hz, 1.2 Hz, p-ph–
CH), 8.04 (dd, 2 H, J ¼ 8:0 Hz, 1.2 Hz, o-ph–CH); 13C
2
68.9 (2JPC ¼ 5:3 Hz, 1-CH2), 69.8 (2C, JPC ¼ 5:3 Hz,
bzl–CH2), 128.1 (4C), 128.6 (4C), 128.7 (2C), 128.9 (2C),
3.1.6. 1-(Dibenzylphosphonooxy)-4-phenyl-2-butyne (6d).
NMR (CDCl3) 1H 3.59 (t, 2H, J ¼ 2:4 Hz, 4-CH2), 4.68
2
130.4 (2C), 131.6, 135.1, 135.4 (2C, JPC ¼ 7 Hz), 189.5
(3-CO), 194.1 (3JPC ¼ 10 Hz, 2-CO); IR (film) 2957,
1732, 1686, 1497, 1216, 1022 cmÀ1; HRMS (DCI/NH3,
pos. ion mode), calcd. for C23H23O6P, MHþ 425.115402,
obs. 425.115695.
3
(dt, 2H, JPH ¼ 10:4 Hz, 2.4 Hz, 1-CH2), 5.05 (d, 4H,
3JPH ¼ 8:0 Hz, bzl–CH2), 7.25–7.33 (m, 15H, ph–CH);
13C 26.1 (4-CH2), 56.9 (2JPC ¼ 6 Hz, 1-CH2,), 70.3
(2JPC ¼ 5 Hz, bzl–CH2), 84.7 (3JPC ¼ 7 Hz, 2-yne-C),
87.3 (3-yne-C), 127.8, 128.9 (4C), 128.9 (4C), 128.9 (2C),
3
129.5 (2C), 129.6 (2C), 136.5 (2C, JPC ¼ 8 Hz), 136.8;
3.1.10. 1-Phosphonooxy-2,2-dihydroxy-3-oxohexane (1a).
1.33 g (3.4 mmol) of 7a was added to a mixture of tri-
IR (film) 2958, 2234, 1493, 1281, 1017 cmÀ1
.
fluoroacetic
acid/thiophenol/H2O
(65 mL/1.7 mL/
1.7 mL) at 25 ꢁC. The mixture was stirredfor 25 min and
3.1.7. 1-(Dibenzylphosphonooxy)-2,3-dioxohexane (7a).
4.1 equiv of NaIO4 were added to a solution of 1 equiv
of 6a in CCl4/CH3CN/H2O (15 mL/15 mL/22.5 mL) to
form two clear phases. 0.022 equiv of RuO2. H2O were
added, and the solution was stirred vigorously for
20 min at room temperature. The reaction was moni-
toredby TLC andconsideredcomplete when no starting
material remained, with product showing a clear yellow
spot on TLC under ultraviolet light. Additional water
(10 mL) was added, and the aqueous layer was extracted
with CH2Cl2. The organic layers were combined, dried
over anhydrous Na2SO4, andconcentrate.d Flash
chromatography on silica gel deactivated with 10% H2O
with gradient elution from pure hexane to 2:1 hexane
EtOAc gave pure 7a as bright yellow oil (2.8 g, 75%
concentrated. H2O andCH Cl2 were added to the resi-
2
due. The two phases were separated and the organic
phase was extractedwith water five times. The combined
aqueous phases containing compound 1a were extracted
with CH2Cl2, filtered, passed over decolorizing carbon,
andconcentratedby rotary evaporation, yielding 0.57 g
(79% yield) of 1a. Purity was establishedby 1H NMR to
be >95%. Sufficient distilled water was then added to
prepare a 0.1 M solution of 1a, which was divided into
0.5 mL samples that were storedat )20 ꢁC: NMR (D2O)
1H 0.86 (t, 3H, J ¼ 7:6 Hz, 6-CH3), 1.55 (tq, 2H,
J ¼ 7:2, 7.6 Hz, 2-CH2), 2.73 (t, 2H, J ¼ 7:2 Hz, 4-CH2),
3
3.98 (d, 2H, JPH ¼ 4:8 Hz, 1-CH2), 13C 12.9 (6-CH3),
16.4 (5-CH2), 38.6 (4-CH2), 68.0 (2JPC ¼ 6 Hz, 1-CH2),
94.9 (3JPC ¼ 13 Hz, 2-C(OH)2), 212.0 (3-CO); IR (film)
3415, 2971,1719, 1654, 1460, 1022 cmÀ1; HRMS (FAB,
neg. ion mode), (M)H)À1 calcd. for anhydrous
C6H10O6P (diketone), 209.021502, obs. 209.022200.
1
yield) after removal of solvent: NMR (CDCl3) H 0.94
(t, 3H, J ¼ 7:2 Hz, 6-CH3), 1.61 (qt, 2H, J ¼ 7:2, 7.6 Hz,
5-CH2), 2.74 (t, 2H, J ¼ 7:6 Hz, 4-CH2), 5.03 (d, 2H,
3JPH ¼ 11:2 Hz, 1-CH2), 5.10 (dd, 2H, 2J ¼ 12 Hz,
3JPH ¼ 8 Hz, bzl–CH2) 5.14 (dd, 2H, 2J ¼ 12 Hz, 3JPH
¼
8 Hz, bzl–CH2), 7.34–7.37 (m, 10H, bzl–CH); 13C 13.5
3.1.11.
1-Phosphonooxy-2,2-dihydroxy-3-oxo-3-cyclo-
(6-CH3), 16.1 (5-CH2), 38.0 (4-CH2), 68.2 (2JPC ¼ 6 Hz,
propylpropane (1b). Synthesis was accomplishedas for
1a, except that activatedcharcoal was replacedby fil-
tration through reverse-phase C18-modified 40 l silica
using 7:3 acetonitrile/H2O. Purity was establishedby 1H
2
1-CH2), 69.1 (2C, JPC ¼ 6 Hz, bzl–CH2), 128.0 (2C),
2
128.6 (4C), 128.6 (4C), 135.5 (2C, JPC ¼ 7:6 Hz), 190.7
(3JPC ¼ 6 Hz, 2-CO), 198.3 (3-CO); IR (film) 2986, 1743,
1715, 1265, 1045 cmÀ1; HRMS (DCI/NH3, pos. ion
mode) calcd. for C20H24O6P, MHþ 391.131052, obs.
391.130697.
1
NMR to be >95%. NMR (D2O) H 1.12–1.15 (m, 2H,
cp–CH2), 1.18–1.22 (m, 2H, cp–CH2), 2.51–2.57 (m, 1H,
3
cp–CH), 4.10 (d, 2H, JPH ¼ 4:8 Hz, 1-CH2); 13C 14.4

3854
Y. Zhang et al. / Bioorg. Med. Chem. 12 (2004) 3847–3855
(cp–CH2), 17.6 (cp–CH), 68.9 (2JPC ¼ 6 Hz, 1-CH2),
96.3 (3JPC ¼ 10 Hz, 2-C(OH)2), 212.0 (3-CO); IR (neat)
3412, 2947, 17044, 1668, 1453, 1247 cmÀ1; HRMS (FAB,
neg. ion mode), (M)H)À1 cald. for anhydrous C6H8O6P
(diketone), 207.005852, obs. 207.005200.
protease activity. The lysate was clearedby centrifuga-
tion (10,000 ·g, 15 min) and the pellets discarded. An
ammonium sulfate precipitation was performed, with
the 30–70% ammonium sulfate precipitate containing
the majority of the E1. After centrifugation (10,000 ·g,
30 min) the supernatant was discarded and the 30–70%
ammonium sulfate precipitate was dissolved in a mini-
mum amount of 20 mM Tris buffer pH 7.5 anddialyzed
overnight against 20 mM Tris buffer pH 7.5 containing
0.5 mM MgSO4 (Buffer A). The resulting dialysate was
clearedby centrifugation (14,000 · g, 5 min), andpassed
through a 5 l syringe filter onto a MonoQ 10/10
(Pharmacia) column mountedon an AKFA FPLC
system. The following chromatographic program was
usedto elute the E1 with a flow rate of 4 mL/min: 2
column volumes (CV) of buffer A following sample
injection (1 CV ¼ 7.854 mL), 10 CV gradient from 0–
15% buffer B (Buffer A+1 M NaCl), 50 CV gradient
from 15–30% buffer B. E1 elutes between 20 CV
(160 mL)and24 CV (190 mL) ( ꢀ17% buffer B). These
fractions were pooled, and were mixed 1:1 with 2.4 M
(NH4)2SO4 in buffer A. After centrifugation andfiltra-
tion, the soluble fraction was appliedto a 16/10 Phenyl-
superose column (Pharmacia) equilibratedwith buffer C
(Buffer A + 1.2 M (NH4)2SO4). The E1 was elutedfrom
the phenyl-superose column using the following gradient
3.1.12. 1-Phosphonooxy-2,2-dihydroxy-3-oxo-3-phenyl-
propane (1c). Purity was establishedby 1H NMR to be
3
>95%. NMR (D2O) 1H 4.13 (d, 2H, JPH ¼ 5:6 Hz,
1-CH2), 7.4 (t, 2H, J ¼ 8:0 Hz, m-ph–CH), 7.58 (t, 1H,
J ¼ 8 Hz, p-ph–CH), 8.07 (d, 2H, J ¼ 7:6 Hz,
o-ph–CH); 13C 70.0 (3JPC ¼ 5 Hz, 1-CH2), 96.6
(3JPC ¼ 11 Hz, 2-C(OH)2), 131.1, 134.3, 135.2, 200.2 (3-
CO); IR (film) 3416, 2942,1711, 1659, 1454, 1185 cmÀ1
;
HRMS (FAB, neg. ion mode), (M)H)À1 cald. for
anhydrous C6H10O6P (diketone), 243.005852, obs.
243.00530.
3.1.13. 1-Phosphonooxy-2-hydroxy-3-oxohex-1-ene (8a).
10 mg of 1a was added to 0.4 mL of buffered (0.05 M
potassium phosphate, pH 7.5) D2O in a standard 5 mm
NMR tube. The pH was not correctedfor isotope ef-
fects. Formation of the enol 8a was followedby NMR,
with enolization complete within 48 h: (D2O) 1H 0.92 (t,
3H, J ¼ 7:6 Hz, 6-CH3), 1.48–1.60 (m, 2H, 2-CH2),
2.49–2.59 (m, mostly exchangedwith D 2O, 4-CH2), 7.47
(flow rate of 2 mL/min): 2 CV of buffer
C
(1 CV ¼ 20.106 mL) following injection, followedby a
gradient (6 CV) from 0–40% buffer A, and a second
gradient (20 CV) from 40% to 100% buffer A. The E1
elutedafter 17 CV (340 mL) ( ꢀ67% buffer A). Fractions
of pure E1, as judged by Coomassie-stained SDS–
PAGE, were pooled and concentrated, then dialyzed
against 50 mM HEPES buffer (pH 7.5) containing
0.5 mM MgCl2. The enzyme was then either reconcen-
tratedto ꢀ1 mM andstoredin liquidnitrogen or mixed
with glycerol (80/20 v/v glycerol/protein stock solution)
andstoredin )20 ꢁC freezer for enzyme assays (2 mg/mL
enzyme). Activity of the enzyme was comparedto that
of the enzyme isolatedfrom K. pneumoniae andfoundto
be equivalent, as measuredby the rate of formation of
acireductone in the absence of oxygen. Overexpression
typically gives a yieldof purifiedenzyme of ꢀ50 mg/L of
growth medium.
3
(d, JPH ¼ 8 Hz, 1-CH); 13C 13.5 (6-CH3), 19.7 (5-CH2),
38.6 (4-CH2), 136.5 (3JPC ¼ 9 Hz, 2-C(OH)), 139.7
(2JPC ¼ 6 Hz, 1-CH), 200.9 (3-CO).
3.1.14. Subcloning, overexpression, purification, and fur-
ther characterization of E1 enolase/phosphatase from K.
oxytoca. The E1 coding region was PCR amplifie0d
from the pDiox2 plasmidwith the forwardprimer 5
-
CACTCTGGAGAACATATGATCCGCGCT-30 and
reverse primer 50-GCGCGCGGATCCTTATGCTGG-
GATCTGCTCCGGATGAATA-30. The PCR product
was double-digested with BamHI and NdeI (New Eng-
landBiolabs), andsubclonedinto pET3a plasmid
(Novagen) to generate a new construct namedpTP01.
The subclonedvector sequences were confirmedby
standard methods. Single colonies of BL21(DE3)pLysS
E. coli transformedwith pTP01 were selectedfrom
LB-agar plates supplementedwith appropriate antibio-
tics andplacedin antibiotic-supplementedLB media.
Protein expression was induced by the addition of iso-
The calculatedextinction coefficient of E1 enolase/
phosphatase basedon the sequence of 21.62 mM À1 cmÀ1
at 280 nm, corresponding to an absorbance of 0.844 at
1 mg/mL. Mass spectra of the purifiedenzyme were
obtainedon an AppliedBiosystems Voyager 2102
MALDI/TOF spectrometer operating in positive ion
mode. The calculated molecular mass of E1 is
25482.6 amu, andMALDI/TOF mass spectral analysis
yields a molecular mass of 25469+/)3 m=z.
propyl b-D-thioglucopuranoside to 0.5 mM upon the
cell density reaching 0.6 at 600 nm. Cells were harvested
by centrifugation, andprotein expression was confirmed
by denaturing gel electrophoresis. If overexpression was
observed, cell-free extracts were assayed for enolase/
phosphatase activity using the procedures described
below.
Molecular modeling of E1 enolase/phosphatase was
performedusing the nest subroutine of Jackal, a molec-
ular modeling program available from Columbia Uni-
versity.28 The structure was modeled using a single
molecule from the deposited coordinates of b-phospho-
glucomutase (1LVH).22 The nest subroutine allows the
modeled structure to be refined by energy minimization
within loop andregular secondary structure regions.
3.1.15. Enzyme purification. A modification of the pro-
cedure of Myers et al. was used to purify E1 enolase/
phosphatase.1 Cells overexpressing E1 were harvested
by centrifugation andlysedby sonication in 20 mM Tris
buffer pH 8.0 with 0.1 % tosyl chloride added to inhibit

Y. Zhang et al. / Bioorg. Med. Chem. 12 (2004) 3847–3855
3855
6. Pochapsky, T. C.; Pochapsky, S. S.; Ju, T. T.; Mo, H. P.;
Al-Mjeni, F.; Maroney, M. J. Nat. Struct. Biol. 2002, 9,
966–972.
7. Dai, Y.; Pochapsky, T. C.; Abeles, R. H. Biochemistry
2001, 40, 6379–6387.
8. Al-Mjeni, F.; Ju, T.; Pochapsky, T. C.; Maroney, M. J.
Biochemistry 2002, 41, 6761–6769.
9. Beaucage, S. L.; Caruthers, M. H. Tetrahedron Lett. 1981,
22, 1859–1862.
10. Wakamiya, T.; Saruta, K.; Yasuoka, J.; Kusumoto, S.
Bull. Chem. Soc. Jpn. 1995, 68, 2699–2703.
11. Sanghvi, Y. S.; Guo, Z. Q.; Pfundheller, H. M.; Converso,
A. Org. Process Res. Dev. 2000, 4, 175–181.
12. Zibuck, R.; Seebach, D. Helv. Chim. Acta 1988, 71, 237–
240.
13. Schuda, P. F.; Cichowicz, M. B.; Heimann, M. R.
Tetrahedron Lett. 1983, 24, 3829–3830.
14. Kitas, E. A.; Knorr, R.; Trzeciak, A.; Bannwarth, W.
Helv. Chim. Acta 1991, 74, 1314–1328.
3.1.16. Enzyme assays. Enzyme assays were performed
in 50 mM HEPES buffer (pH 7.4) containing 0.5 mM
MgCl2, using anaerobic cuvettes with 1 cm path length.
1 mL of buffer containing ꢀ200 lM substrate was placed
in the cuvette. The solution was then degassed for at
least 15 min by bubbling a gentle stream of argon gas
introduced through the septum with a syringe needle,
with a smaller needle for exhaust. A drop of water was
placedon the top of the septum to ensure the seal, and
the needles removed. 2 lL of a 40 lM solution of E1
enolase/phosphatase in 80/20 glycerol/water was intro-
duced by syringe, the solution mixed rapidly, and the
course of the reaction monitoredby UV–visible spec-
trometry on a HP diode-array spectrophotometer
equippedwith temperature control. Temperature was
maintainedat 25 ꢁC. Basedon the mass of dehydrated
purifiedsamples of 1a, andassuming 100% conversion
of substrate to acireductone product, an extinction
coefficient was calculatedat 320 nm for the acireductone
15. Miyata, K.; Nakashima, K.; Koyanagi, M. Bull. Chem.
Soc. Jpn. 1989, 62, 367–371.
derivative of 1a of 8000 MÀ1 cmÀ1
.
16. Lee, Y. J.; Burr, J. G. Chem. Phys. Lett. 1976, 43, 146–148.
17. Collet, J. F.; Stroobant, V.; Pirard, M.; Delpierre, G.; Van
Schaftingen, E. J. Biol. Chem. 1998, 273, 14107–14112.
18. Wang, W. R.; Kim, R.; Jancarik, J.; Yokota, H.; Kim,
S. H. Structure 2001, 9, 65–71.
Acknowledgements
19. Selengut, J. D.; Levine, R. L. Biochemistry 2000, 39, 8315–
8324.
This work was supportedby a grant from the US PHS
(GM067786, TCP). G.M.P acknowledges previous
support from PHS training grant GM007596. The
authors thank Dr. R. Rogers Yokum (OmniGene, Inc.
Cambridge, MA) for providing us with the original
pDiox-2 plasmid. We thank Rebecca Myers, Brandeis
University Core Facility for DNA sequencing data. We
thank Prof. Karen N. Allen (Boston University School
of Medicine) for helpful discussions.
20. Morais, M. C.; Zhang, W. H.; Baker, A. S.; Zhang, G. F.;
Dunaway-Mariano, D.; Allen, K. N. Biochemistry 2000,
39, 10385–10396.
21. Hisano, T.; Hata, Y.; Fujii, T.; Liu, J. Q.; Kurihara, T.;
Esaki, N.; Soda, K. J. Biol. Chem. 1996, 271, 20322–
20330.
22. Lahiri, S. D.; Zhang, G. F.; Dunaway-Mariano, D.; Allen,
K. N. Science 2003, 299, 2067–2071.
23. Shin, D. H.; Roberts, A.; Jancarik, J.; Yokota, H.; Kim,
R.; Wemmer, D. E.; Kim, S.-H. Protein Sci. 2003, 12,
1464.
24. Argiriadi, M. A.; Morisseau, C.; Hammock, B. D.;
Christianson, D. W. Proc. Natl. Acad. Sci. U.S.A. 1999,
96, 10637–10642.
References and notes
25. Selengut, J. D. Biochemistry 2001, 40, 12704–12711.
26. Morgan, T. K.; Wohl, R. A.; Lumma, W. C.; Wan, C. N.;
Davey, D. D.; Gomez, R. P.; Marisca, A. J.; Briggs, M.;
Sullivan, M. E.; Wong, S. S. J. Med. Chem. 1986, 29,
1398–1405.
27. Claesson, A.; Sahlberg, C. Tetrahedron 1982, 38, 363–368.
28. Xiang, J. Z. ‘Jackal: A protein structure modeling pack-
age’, URL: http://trantor.bioc.columbia.edu/programs/
jackal.
1. Myers, R. W.; Wray, J. W.; Fish, S.; Abeles, R. H. J. Biol.
Chem. 1993, 268, 24785–24791.
2. Balakrishnan, R.; Frohlich, M.; Rahaim, P. T.; Backman,
K.; Yocum, R. R. J. Biol. Chem. 1993, 268, 24792–24795.
3. Wray, J. W.; Abeles, R. H. J. Biol. Chem. 1993, 268,
21466–21469.
4. Wray, J. W.; Abeles, R. H. J. Biol. Chem. 1995, 270, 3147–
3153.
29. Koradi, R.; Billeter, M.; Wuthrich, K. J. Mol. Graphics
1996, 14, 51.
5. Dai, Y.; Wensink, P. C.; Abeles, R. H. J. Biol. Chem.
1999, 274, 1193–1195.