Design, synthesis, functional and structural characterization of an inhibitor of N-acetylneuraminate-9-phosphate phosphatase: Observation of extensive dynamics in an enzyme/inhibitor complex

Article abstract of DOI:10.1016/j.bmcl.2013.05.052

The design, synthesis and characterization of a phosphonate inhibitor of N-acetylneuraminate-9-phosphate phosphatase (HDHD4) is described. Compound 3, where the substrate C-9 oxygen was replaced with a nonlabile CH₂ group, inhibits HDHD4 with a binding affinity (IC₅₀ 11 μM) in the range of the native substrate Neu5Ac-9-P (compound 1, K_m 47 μM). Combined SAR, modeling and NMR studies are consistent with the phosphonate group in inhibitor 3 forming a stable complex with native Mg²⁺. In addition to this key interaction, the C-1 carboxylate of the sugar interacts with a cluster of basic residues, K141, R104 and R72. Comparative NMR studies of compounds 3 and 1 with Ca²⁺ and Mg²⁺ are indicative of a highly dynamic process in the active site for the HDHD4/Mg²⁺/3 complex. Possible explanations for this observation are discussed.

Full text of DOI:10.1016/j.bmcl.2013.05.052

Bioorganic & Medicinal Chemistry Letters 23 (2013) 4107–4111
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis, functional and structural characterization
of an inhibitor of N-acetylneuraminate-9-phosphate phosphatase:
Observation of extensive dynamics in an enzyme/inhibitor complex
⇑
⇑
Soong-Hoon Kim , Keith L. Constantine , Gerald J. Duke, Valentina Goldfarb, John T. Hunt,
Stephen Johnson, Kevin Kish, Herbert E. Klei, Patricia A. McDonnell, William J. Metzler, Luciano Mueller,
Michael A. Poss, Craig R. Fairchild, Rajeev S. Bhide
Bristol-Myers Squibb Research and Development, 311 Pennington-Rocky Hill Road, Pennington, NJ 08534, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The design, synthesis and characterization of a phosphonate inhibitor of N-acetylneuraminate-9-phos-
phate phosphatase (HDHD4) is described. Compound 3, where the substrate C-9 oxygen was replaced
Received 14 February 2013
Revised 10 May 2013
Accepted 14 May 2013
Available online 23 May 2013
with a nonlabile CH2 group, inhibits HDHD4 with a binding affinity (IC50 11
lM) in the range of the native
substrate Neu5Ac-9-P (compound 1, Km 47 M). Combined SAR, modeling and NMR studies are consis-
l
tent with the phosphonate group in inhibitor 3 forming a stable complex with native Mg2+. In addition
to this key interaction, the C-1 carboxylate of the sugar interacts with a cluster of basic residues, K141,
R104 and R72. Comparative NMR studies of compounds 3 and 1 with Ca2+ and Mg2+ are indicative of a
highly dynamic process in the active site for the HDHD4/Mg2+/3 complex. Possible explanations for this
observation are discussed.
Keywords:
N-Acetylnuraminate-9-phosphate
phosphatase
Inhibitor
Comparative NMR studies
Enzyme/inhibitor complex
Ó 2013 Elsevier Ltd. All rights reserved.
Post-translational modification of proteins with carbohydrates
generates divergent epitopes that are critical in multiple roles,
including protein trafficking and cell signaling.1 Defects in the bio-
synthesis of glycosylated proteins have been implicated in several
disease states, including inflammation and tumor progression.1–3
The prevalent ‘capping’ carbohydrate in the branches of oligosac-
charides is N-acetylneuraminate (Neu5Ac), a sialic acid derivative.
N-Acetylnuraminate-9-phosphate phosphatase (E.C. 3.1.3.29) is
one of the enzymes involved in the biosynthesis of Neu5Ac in
eukaryotes and higher species.4 This enzyme is the human ortho-
log of a rat protein initially annotated as ‘haloacid dehalogenase-
like hydrolase domain containing 40 (HDHD4).4 It cleaves the C-9
phosphate group from its substrate N-acetylnuraminate-9-phos-
phate 1 (Fig. 1). HDHD4 is a member of the haloacid dehydroge-
nase family (HAD) of enzymes. The HAD family of enzyme
catalyzes a variety of related reactions, including the cleavage of
the oxygen–phosphorous bond during phosphate hydrolysis.5,6
The key step in the hydrolysis involves the transfer of substrate
phosphate to an active-site aspartic acid mediated by an active-site
Mg2+ ion. The active-site aspartic acid residue coordinates the Mg2+
ion in a monodentate manner, leaving the other carboxylate oxy-
gen free to perform a nucleophilic attack on the substrate phospho-
rous. Subsequent hydrolysis of the phosphorylated aspartic acid
residue releases the phosphate group, and regenerates the aspartic
acid. The general mechanism of phosphate transfer in the HAD
class of enzymes has received a great deal of attention, including
studies of substrate selectivity,6 studies of the nature of phosphate
transfer (dissociative, concerted or addition/elimination5––a sub-
ject of continuing debate), studies of protein conformational
changes that occur during catalysis,7,8 and detailed ab initio calcu-
lations of magnesium ion–phosphate group interactions.9
The design, synthesis and characterization of reversible inhibi-
tor 3 of HDHD4 and closely related analogs are described with
O
P
HO
HO
OH
9
O
9
OH
Neu5Ac-9-Pase
OH
CO2H
OH
OH
CO2H
2
O
O
HO
HO
6
AcHN
AcHN
OH
2
Neu5Ac
OH
1
Neu5Ac-9-P
⇑
Corresponding authors.
E-mail address: soonghoon.kim@bms.com (S.-H. Kim).
Figure 1. Hydrolysis of C-9 phosphate group by HDHD4.
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.05.052

4108
S.-H. Kim et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4107–4111
O
P
tection of silyl and anomeric methoxy groups with acidic Dowex
50 W resin to give phosphonate 3.
O
P
HO
HO
HO
HO
O
Scheme 2 shows the synthesis of analogs 12 and 13. These ana-
logs address the role of C2 anomeric hydroxyl and methoxy groups.
Compound 10 was treated with Dowex 50 W acidic resin to re-
move the silyl groups to give ester 12, which was hydrolyzed with
1 N NaOH in methanol to give acid 13.
9
OH
OH
9
OH
CO2H
OH
CO2H
O
O
HO
HO
AcHN
AcHN
OH
OH
Synthesis of compound 16, where the phosphonate group in 3 is
replaced by a carboxylic acid, is shown in Scheme 3. Olefination of
aldehyde 8 gave the olefin 14, which was reduced with Pd/C to give
compound 15. Hydrolysis followed by deprotection gave com-
pound 16.
1
3
Figure 2. Phosphonate inhibitor 3. The labile phosphorous–oxygen bond (marked
with an arrow) has been replaced with stable phosphorous–carbon bond.
As illustrated in Table 1, phosphonate 3 showed inhibition15
that is close to the Km of the substrate (Km of 1 is 47 l
M4,15), sug-
the aim of providing tool compounds for a comprehensive struc-
tural characterization of enzyme inhibitor complexes by NMR
and X-ray crystallography (Fig. 2).
gesting that similar interactions are involved in substrate binding.
These comparable activities also suggest that the internal C-9
phosphate oxygen in native substrate Neu5Ac 1 may not be in-
volved in an additional binding interaction with the enzyme.
Replacing the phosphonate group with a carboxylic acid in com-
pound 16 significantly reduced inhibition (>10-fold loss).
Capping the C-2 anomeric oxygen with a methyl group in com-
pound 12 led to fivefold reduction in potency. Replacing both the
hydroxyl and carboxyl group with methoxy and methyl ester
groups respectively in analog 13 resulted in 10-fold loss in activity.
Sodium orthovanate 4, which is a well-known inhibitor of phos-
phate hydrolysis,16 showed inhibition similar to phosphonate 3.
The pKa of inhibitors were determined to assess the extent of
ionization at physiological pH.17 The Mg2+ interaction with the
phosphonate oxygen is a metal–ligand interaction where the
charge and coordination number determine the bond enthalpy.
As shown in Table 2, both phosphate and phosphonate groups
are likely to exist as di-anion at pH 7. The calculated literature9 val-
ues for Mg2+ mono-dentate ligand interaction in the gas phase
Traditionally, inorganic complexes such as sodium orthovana-
date 4 (Na3VO4, Table 1) have been employed to study the inhibi-
tion of phosphatase activity, but they provide little information on
the roles of the organic features of the substrate to overall binding
and selectivity, and are not suitable starting points for structure-
based inhibitor design. The inhibitor 3 replaces the phosphate
group in Neu5Ac-9-P 1 with a phosphonate group,10 while main-
taining all the features of the native substrate, including the sugar
portion. Unlike the phosphate group in substrate 1,4,11 the carbon–
phosphorous bond cannot be hydrolyzed by HDHD4. Initial bio-
chemical characterization suggests that inhibitor
HDHD4 with an affinity similar to native substrate Neu5Ac-9-P 1
in a reversible manner (Table 1).
3 binds to
Additional analogs (Table 1) were prepared to understand the
role of C-9 phosphate and C-1 carboxylate groups in substrate 1.
The SAR of these closely related analogs, together with modeling
and NMR studies helped define the critical features of substrate/
HDHD4 interaction. Moreover, comparative NMR studies of com-
pounds 3 and 1 with Ca2+ and Mg2+ indicate a previously unob-
served active site dynamic equilibrium process in the HDHD4/
Mg2+/3 complex.
Scheme 1 shows the synthesis of phosphonate 3. Esterification
of N-acetylneuraminic acid 212 was followed by protection of alco-
hols as silyl ethers. Selective deprotection of the least hindered
silyl ether from compound 6 was accomplished with acetic acid
in THF/water to give C-9 hydroxyl compound 7. Swern oxidation
gave aldehyde 8 which was used immediately for the next step.
Olefination13 in toluene for 3 days at 100 °C gave a modest yield
(30%) of olefin 9 as a single isomer. Reduction14 with PtO2 reduced
the olefin and removed the bis-phenoxy esters in a single step to
give compound 10. Methyl ester hydrolysis was followed by depro-
show the following trend in bond enthalpy: HPO24ꢀ > H2PO1ꢀ
ꢁ
4
MeCO12ꢀ, where the overall charge imparts greater ligand binding
to Mg2+. Taken together, the >10-fold difference in inhibition
between phosphonate 3 and carboxylate 16 is likely due to differ-
ence in bond enthalpy between a di-anion phosphonate and mono-
anion carboxylate.
Comparative NMR data (Supporting information) were consis-
tent with the small difference in the chemical structures between
inhibitor 3 and native Neu5Ac-9-P 1. Their solution NMR chemical
shifts and coupling constants revealed virtually no difference in su-
gar ring conformation. In both compounds, the rigid chair confor-
mation of the sugar was maintained. The acyclic portion defined
by C-6 to C-9 carbon was not well resolved in either analogs, pre-
sumably because the linear chain can adopt several conformations.
Thus, the close IC50 of compound 3 to the Km value of native sub-
strate 1 reflects the similarity in key conformational and structural
features (phosphonate/Mg2+ interaction and Neu5Ac) that are crit-
ical for binding to HDHD4; compound 3 seemed like an ideal tool
compound for more extensive structural characterization of the
HDHD4 enzyme inhibitor complex.
In order to further verify and characterize interactions of
HDHD4 with 1 and/or 3, experimental structural information was
needed. We have determined an X-ray structure of HDHD4 in com-
plex with Mg2+ and vanadate.18 An X-ray structure of HDHD4 in
complex with Na+ has also been reported.19 However, X-ray struc-
tures of HDHD4 in complex with 1 and/or 3 are not yet available.
NMR assignments have been obtained for apo-HDHD4, and for
HDHD4 in complex with Ca2+ and native substrate 1. Ca2+ occupies
the same site as Mg2+ in the HAD phosphatases but, unlike Mg2+, it
does not support phosphatase activity. The HAD phosphatases all
contain a catalytic aspartate residue (Asp12 in HDHD4) that has
one carboxylate oxygen coordinated to Mg2+. The other carboxyl-
ate oxygen is available to perform a nucleophilic attack on the
Table 1
Inhibition of Neu5Ac-P inhibitors
O
P
O
HO
HO
OH
OH
OR1
HO
OH
O
CO2R2
6
6
O
CO2H
HO
2
HO
2
16
AcHN
AcHN
OH
OH
Compounds
R1
R2
IC50 (l
M)*
4
3
12
13
16
Na3VO4
H
Me
Me
––
––
H
6.5
11.0
49.8
93.7
H
Me
––
>100
*
Km and Vmax for HDHD4/substrate 1 were determined to be 47
min/mg protein, respectively.
lM and 143 lmol/

S.-H. Kim et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4107–4111
4109
OTBS
OMe
OH
OH
TBSO
HO
HO
OH
CO2H
HO
OMe
CO2Me
a
b
O
CO2Me
O
O
TBSO
AcHN
HO
AcHN
AcHN
OTBS
OH
OH
6
5
2
c
O
PhO
PhO
P
OTBS
OHC
TBSO
OTBS
OMe
CO2Me
OTBS
OMe
CO2Me
HO
OMe
CO2Me
e
d
h
O
O
O
TBSO
TBSO
AcHN
AcHN
AcHN
OTBS
OTBS
OTBS
f
7
8
9
O
P
O
P
O
P
HO
HO
HO
HO
HO
HO
OTBS
OMe
OTBS
g
OH
OMe
CO2H
OH
CO2H
O
CO2Me
O
O
TBSO
AcHN
TBSO
AcHN
HO
AcHN
OTBS
10
OTBS
11
OH
3
Scheme 1. (a) Dowex 50 W, 65 °C, 48 h, 60%; (b) TBS-OTf, 2,6-lutidine, RT, 16 h, 82%; (c) AcOH/water/THF (3:1:1), RT, 6 h, 56% (d) Swern oxidation, 85%; (e) (PhO)2POC = PPh3,
toluene, 100 °C, 3 days, 30%; (f) PtO2, H2 (ballon), ethanol, RT, 2 h, 94%; (g) 1 N NaOH/methanol (1:1), RT, 2 h, 80%; (h) Dowex 50 W, 80 °C, 1 h, 67%.
O
P
O
P
O
P
HO
HO
HO
HO
HO
HO
a
OTBS
OMe
b
OH
OH
OMe
CO2H
OMe
CO2Me
O
CO2Me
O
O
TBSO
AcHN
HO
AcHN
HO
AcHN
OTBS
OH
OH
13
12
10
Scheme 2. (a) Dowex 50 W, 80 °C, 1 h, 76%; (b) 1NaOH/MeOH, RT, 2 h, 80%.
HO2C
HO2C
MeO2C
OHC
OTBS
O
OH
OTBS
c
OTBS
OMe
CO2Me
OMe
CO2Me
a
b
OH
CO2H
OMe
CO2H
O
O
TBSO
O
HO
AcHN
TBSO
AcHN
TBSO
AcHN
AcHN
OTBS
OH
OTBS
OTBS
14
16
8
15
Scheme 3. (a) Ph3P = CHCO2Me, toluene, 70 °C, 16 h, 37%; (b) Pd/C, H2 (ballon), EtOAc, 3 h, 95%; (c) 1 N NaOH/MeOH, RT, 12 h; Dowex 50 W, water, 80 °C, 2 h. 25% yield for
two steps.
phosphorous atom of the substrate.5 When Mg2+ is replaced by
Ca2+ it coordinates both carboxylate oxygens of the catalytic aspar-
tate, preventing nucleophilic attack on the substrate.20,21
Using the available data (NMR, X-ray, SAR), the binding site was
modeled22 with high confidence. As shown in Figure 3, the key
interaction is the coordination of the phosphonate oxygen atom
with Mg2+. The Mg2+–oxygen bond length is 2.0 Å, in-line with
published data for Mg2+–oxygen bond lengths in related enzyme
complexes.7,23 The phosphonate oxygens may also be involved in
additional interactions with T131, and K164. The sugar C-1 carbox-
ylate makes key contacts with a cluster of basic residues: K141,
R104 and R72.24 The binding model is consistent with SAR where
removal of the phosphonate or C-1 carboxylate groups leads to loss
of inhibition (Table 1).
In the presence of a large excess of 1, the HDHD4/Ca2+ chemical
shift changes were not accompanied by additional NMR line broad-
ening of HDHD4 resonances. These results indicate that 1 does not
display multiple binding modes.25 More specifically, 1 does not
undergo significant binding mode transitions on the
ls–ms time-
scale when bound to HDHD4/Ca2+. Such bound-state motions on
this timescale are expected to cause readily detectable NMR line
broadening.25

4110
S.-H. Kim et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4107–4111
Table 2
pKa of HDHD4 inhibitors
O
O
P
OH
HO
HO
OH
HO
CO2H
OH
OH
CO2H
HO
OH
X
OH
CO2H
2
O
6
O
6
HO
O
HO
AcHN
2
HO
AcHN
AcHN
OH
OH
OH
2
16
1
(X=O)
3 (X=CH2)
Compounds
Calc pKa phosphorous
Calc pKa carboxylic acid
Experimental pKa
1
3
2
1.8, 6.8
2.3, 7.9
––
2.3
2.3
2.3
6.3, 3.6
7.6, 3.5
3.2
16
––
2.3, 4.5
NT
Figure 4. Portions of overlain 1H–15N HSQC spectra. Blue peaks are from HDHD4/
Ca2+/3, and red peaks are from HDHD4/Mg2+/3. Peaks labeled in the former
spectrum correspond to broadened peaks in the latter spectrum.
Figure 3. Model of inhibitor 3 bound to HDHD4 binding site.
Two additional ternary complexes were examined by NMR:
HDHD4/Ca2+/3 and HDHD4/Mg2+/3. Like the HDHD4/Ca2+/1 com-
plex, the HDHD4/Ca2+/3 complex induced chemical shift changes
in and around the active site without additional line-broadening
under saturating conditions; NMR assignments were readily ob-
tained for the HDHD4/Ca2+/3 complex by comparison to the data
obtained for the HDHD4/Ca2+/1 complex. In contrast to the com-
plexes with Ca2+, extreme line broadening is observed for a number
of backbone amide peaks in the 2D 1H–15N HSQC spectrum in the
presence of a large excesses of 3 and Mg2+ (Fig. 4). Figure 4 shows
the line broadening perturbations that have been identified with
high confidence mapped on to an X-ray/NMR-based model of the
HDHD4/Mg2+/1 ground state complex. The line broadening is espe-
cially pronounced in the catalytic core of the enzyme, with many
resonances broadened beyond detection (see Fig. 4). Broadening
extends throughout the central b-sheet (Fig. 5).
ing observed in the HDHD4/Mg2+/3 complex reflects some highly
dynamic process occurring on the
ls–ms timescale in the active
site. Line-broadening has been reported for residues in direct con-
tact with the phosphates of ATP in complex with human sulfire-
doxin due to bound state conformational exchange.26 We observe
line-broadening across the core domain of HDHD4; simple confor-
mational exchange of 3 is unlikely to be the cause of the effects we
observe. Another possible explanation (discussed in more detail in
the Supporting information) for the line-broadening observed is a
reversible nucleophilic attack of the Asp12 carboxylate oxygen
on the phosphorous atom of 3, resulting in a ground state—phos-
phorane intermediate state equilibrium (Supporting information).
In this mechanism, the phophorous atom undergoes a change in
coordination from four to five when it accepts the Asp12 carboxyl-
ate oxygen into its coordination sphere. Of the three alternative
mechanisms for the first step of phosphoryl transfer by HAD super-
family members,5 the addition–elimination pathway appears to be
the only one consistent with our observations. The concerted path-
way does not have an intermediate. The intermediate expected for
the dissociative pathway has the P–C bond broken; this cleavage is
unlikely, since HDHD4 is not a phosphonatase.5
Possible causes of active-site line-broadening include the pres-
ence of paramagnetic impurities, exchange between ligand-bound
and ligand-free forms of the protein, and dynamic processes that
occur in the ligand-bound state. We have ruled out paramagnetic
impurities and exchange between ligand-bound and ligand-free
forms of the enzyme (Supporting information). The line-broaden-

S.-H. Kim et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4107–4111
4111
References and notes
1. Dove, A. Nat. Biotechnol. 2001, 19, 913.
2. Grewal, P. K.; Hewitt, J. E. Hum. Mol. Genet. 2003, 12, 259.
3. Brooks, S. A.; Hall, D.; Buley, I. Br. J. Cancer 2001, 858, 1014.
4. Maliekal, P.; Vertommen, D.; Delpierre, G.; Schaftingen, E. R. Glycobiology 2006,
16, 165.
5. Allen, K. N.; Dunaway-Marino, D. TIBS 2004, 29, 496.
6. Kuznetsova, E.; Proudfoot, M.; Gonzalez, C. F.; Brown, G.; Omelchenko, M. V.;
Borozan, I.; Carmel, L.; Wolf, Y. I.; Mori, H.; Savchenko, A. V.; Arrowsmith, C.
H.; Koonin, E. V.; Edwards, A. M.; Yakunin, A. F. J. Biol. Chem. 2006, 281,
36149.
7. Lahiri, S. D.; Zhang, G.; Dunaway-Marino, D.; Allen, K. N. Science 2003, 299,
2067.
8. Lahiri, S. D.; Zhang, G.; Dunaway-Mariano, D.; Allen, K. N. Bioorg. Chem. 2006,
34, 394.
9. Mayaan, R.; Range, K.; YorkLahiri, D. M. J. Biol. Inorg. Chem. 2004, 9, 807.
10. Phosphonates have been validated as inhibitors of phosphate hydrolysis. For
recent review as inhibitors of tyrosine phosphatase, see: Combs, A. P. J. Med.
Chem. 2010, 53, 2333.
11. Preparation of N-acetylnuraminate-9-phosphate 1 has been described in Ref. 4
A modified procedure to prepare 1 was developed and will be subject of future
publication.
12. Yu, R. K.; Ledeen, R. J. Biol. Chem. 1959, 244, 1306.
13. Ogino, Y.; Nishiyama, S.; Yamamura, S.; Izawa, T.; Kato, K. Bioorg. Med. Chem.
Lett. 1994, 4, 1991.
14. Mlynarski, J.; Banaszek, A. Carbohydr. Res. 1996, 295, 69.
15. HDHD4 phosphatase activity was determined using
(malachite green) assay. The reaction mixture contained 50
1, 25 mM HEPES pH 7.5, 25 mM KCl, 1 mM MgCl2, 1 mM DTT, 50
a
Biomol Green™
M Neu5Ac-9-P
g/ml BSA
l
l
and 1 ng of recombinant HDHD4. After a 30 min incubation time at 30° with
the enzyme, released phosphate was quantified by addition of four volumes
of the Biomol Green™ reagent (Biomol; Plymouth Meeting, PA). After
incubation for 30 min at room temperature, absorbance at 620 nm was
Figure 5. Line-broadening effects mapped on to a model of HDHD4/Mg2+/1. Ribbon
representations of residues 7–25, 55–71, and 100–242 are shown. Residues
displaying broadened peaks are orange, the remaining residues are grey.
read on
a Titertek Multiskan MCC/340 microplate reader. Under these
conditions Km and Vmax were determined to be 47 lM and 143 lmol/min/
mg protein, respectively.
16. Swarup, G.; Cohen, S.; Garbers, D. L. Biochem. Biophys. Res. Commun. 1982, 107,
1104.
17. The pKa values were obtained using a Sirius GLpKa instrument. The pKa was
measured using the potentiometric titration method.
18. McDonnell, P. A.; Constantine, Keith L.; Klei, H. E.; Johnson, Stephen R.;
Goldfarb, V.; Kish, K.; Kim, S. -H. Three-dimensional structure of HDHD4
complexed with magnesium and a phosphate mimetic. Bristol-Myers Squibb,
assignee. PCT Int. Appl. (2007), WO 2,007,112,377 A2 20,071,004. PDB IDs for
the HDHD4 apo structure determined by SeMet MAD phasing is 4KNV and for
HDHD4/Mg+2 & VO42ꢀ complex is 4KNW.
In summary, inhibitor 3 was prepared in eight steps from N-
acetylneuramic acid 2. The inhibitor 3 has binding affinity similar
to Neu5Ac-9-P 1. The SAR suggests that the key interaction is be-
tween the active site Mg2+ and the phosphonate/phosphate group
of the substrate. pKa measurement shows that, like the substrate
phosphate, the phosphonate exists as a di-anion ligand, which con-
tributes significantly to binding enthalpy. Combined NMR and
modeling studies confirm that the conformation of Neu5Ac-9-P 1
and inhibitor 3 in solution are essentially identical, and that the
binding modes of 1 and 3 to HDHD4 are highly similar as well, with
either Mg2+ or Ca2+ in the active site. Thus, inhibitor 3 is a useful
tool compound for further structural studies and inhibitor design.
Our studies also reveal a highly unusual dynamic process in the en-
zyme active site when 3 and Mg2+ are bound. The results are con-
sistent with the first half-reaction of the addition-elimination
mechanism5—and equilibrium between the ground state phospho-
nate and a phosphorane intermediate state can account for our
observations. The fundamental mechanism of phosphate transfer
in the HAD enzymes is the subject of much debate.5,7,27–31 Further
characterization of the HDHD4/Mg2+/3 complex and related sys-
tems should enhance our understanding of this fundamental
process.
19. PDB ID (revised 13-Jul-2011): 2W4M, E. Ugochukwu, R. Cocking, W.W. Yue,
A.C.W. Pike, A. Roos, J.R.C. Muniz, F. Von Delft, C. Bountra, C.H. Arrowsmith, J.
Weigelt, A. Edwards, U. Oppermann. Structural genomics consortium. The
crystal structure of human N-acetylneuramic acid phosphatase, NANP.
20. Peeraer, Y.; Rabijns, A.; Collett, J.-F.; Van Schaftingen, E.; Ranter, C. D. Eur. J.
Biochem. 2004, 271, 3421.
21. Ca2+ had an IC50 of 101
lM under assay conditions described in Ref. 15.
22. The X-ray structure of HDHD4 in complex with Mg2+ and vanadate was utilized
for as a starting point for molecular modeling/docking. As discussed in the
Supporting information, this structure revealed an ‘open cap’ domain
orientation. Based on NMR data (Supporting information), the cap domain
undergoes a slight twist in orientation relative to the core domain upon
binding of 1. Details on the modeling of the complexes are described in the
Supporting information.
23. A distance of 2.2 Å has been reported for magnesium(II)–oxygen bond in
related b-glucophosphomutase. See Ref. 7.
24. Similar interaction between a carboxylate and basic amino acids residues have
been detailed with sialic acid and influenza neuraminidase. Varghese, J. N.;
McKimm-Bresckhim, L.; Caldwell, J. B.; Kortt, A.; Colman, P. M. Proteins: Struct.
Funct. Genet. 1992, 14, 327.
Acknowledgments
25. Reibarkh, M.; Malia, T. J.; Wagner, G. J. Am. Chem. Soc. 2006, 128, 2160.
26. Lee, D. Y.; Park, S. J.; Jeong, W.; Sung, H. J.; Oho, T.; Wu, X.; Rhee, S. D.;
Gruschus, J. M. Biochemistry 2006, 45, 15301.
27. Knowles, J. Science 2003, 299, 2002.
28. Webster, C. E. J. Am. Chem. Soc. 2004, 126, 6840.
29. Tremblay, L. W.; Zhang, G.; Dai, J.; Dunaway-marino, D.; Allen, K. N. J. Am.
Chem. Soc. 2005, 127, 5298.
30. Baxter, N. J.; Olguin, L. F.; Golicnik, M.; Feng, G.; Hounslow, A. M.; Bermel, W.;
Blackburn, G. M.; Hollfelder, F.; Waltho, J. P.; Williams, N. H. Proc. Natl. Acad.
Sci. U.S.A. 2006, 103, 14732.
31. Baxter, N. J.; Bowler, M. W.; Alizadeh, T.; Cliff, M. J.; Hounslow, A. M.; Wu, B.;
Berkowitz, D. B.; Williams, N. H.; Blackburn, G. M.; Waltho, J. P. Proc. Natl. Acad.
Sci. U.S.A. 2010, 173, 4555.
We thank Dr. Steven Sheriff for assistance in depositing the
coordinates of the HDHD4 structures in the PDB, and we thank
Dr. Janet Caceres-Cortez for providing the solution NMR data for 3.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2013.
05.052. These data include MOL files and InChiKeys of the most
important compounds described in this article.