Design, synthesis and structure-activity relationships of new phosphinate inhibitors of MurD

Article abstract of DOI:10.1016/j.bmcl.2005.09.086

A series of new phosphinate compounds were designed and synthesized as inhibitors of the d-glutamic acid-adding enzyme (MurD) involved in peptidoglycan biosynthesis. They were tested against the MurD enzyme from Escherichia coli, allowing initial structure-activity relationships to be deduced. Two compounds had IC₅₀ values near 100 μM and constitute a promising starting point for further development.

Full text of DOI:10.1016/j.bmcl.2005.09.086

Bioorganic & Medicinal Chemistry Letters 16 (2006) 343–348
Design, synthesis and structure–activity relationships of new
phosphinate inhibitors of MurD
a
b
a,
*
ˇ
Katja Strancar, Didier Blanot and Stanislav Gobec
aUniversity of Ljubljana, Faculty of Pharmacy, Asˇkercˇeva 7, 1000 Ljubljana, Slovenia
bEnveloppes Bacte´riennes et Antibiotiques, UMR 8619 CNRS, Universite´ de Paris-Sud, 91405 Orsay, France
Accepted 28 September 2005
Available online 3 November 2005
Abstract—A series of new phosphinate compounds were designed and synthesized as inhibitors of the D-glutamic acid-adding
enzyme (MurD) involved in peptidoglycan biosynthesis. They were tested against the MurD enzyme from Escherichia coli, allowing
initial structure–activity relationships to be deduced. Two compounds had IC50 values near 100 lM and constitute a promising
starting point for further development.
ꢀ 2005 Elsevier Ltd. All rights reserved.
The increasing emergence of pathogenic bacterial strains
with high resistance to antibiotic therapy constitutes a
serious public threat.1 This has created an urgent need
for the development of new antibacterial agents directed
towards novel targets. One of the best known and most
validated targets for antibacterial therapy is the machin-
ery for peptidoglycan biosynthesis.2,3 However, the early
biosynthetic steps have received relatively little attention
as potential drug targets, even though this part of the
biosynthetic pathway utilizes essential enzymes that
have no mammalian counterparts.
There have been several attempts to target MurD with
novel inhibitors.6–13 The most potent compounds were
designed as phosphinate transition state analogues.6,7,12
In all of them the phosphinodipeptide part, Ala-w(PO2-
CH2)-Glu, was conserved. They differ in the moiety
mimicking the UDP-MurNAc part, for which a QSAR
study has recently been made.14 On the other hand,
phosphinate transition state analogues for MurD with
modifications in the peptide part have not been
reported.
We were interested in the design, synthesis and biologi-
cal evaluation of simplified phosphinate inhibitors of
MurD. For this purpose, we selected compound 27 as
the starting point for an exploratory structure–activity
relationship (SAR) study (Fig. 1). This compound,
although lacking the UMP moiety, still possesses good
inhibitory potency (IC50 = 20 nM). As a first step we
sought a simple substitute for the phospho-sugar residue
as well as an appropriate replacement of the D-lactoyl
residue (ꢀlinkerꢁ). Following this we introduced modifi-
cations into the phosphinodipeptide part of the inhibi-
tors. In lieu of phosphinoalanine, certain other
phosphino amino acids were incorporated and, in addi-
tion, the glutamic acid mimetic, 2-methylenepentane-
1,5-dioic acid, was truncated (Fig. 1).
Peptidoglycan is an essential macromolecular component
of the cell wall of both Gram-positive and Gram-negative
bacteria. Its main function is to preserve cell integrity by
withstanding the internal osmotic pressure. The glycan
chains are composed of alternating units of N-acetylglu-
cosamine (GlcNAc) and N-acetylmuramic acid (Mur-
NAc). The carboxyl group of MurNAc residues is
substituted in most bacteria by a peptide unit, L-alanyl-
c-D-glutamyl-meso-diaminopimeloyl(or L-lysyl)-D-ala-
nine.3 The enzyme MurD (UDP-MurNAc-L-alanine:
D-glutamate ligase), which catalyses the addition of
D-glutamate to the cytoplasmic peptidoglycan precursor
UDP-MurNAc-L-Ala, is present in all bacteria and is
highly specific for D-glutamate.4 High specificity, ubiqui-
ty among bacteria and absence in mammals make MurD
a promising target for antibacterial therapy.5
The synthesis of phosphinate inhibitors 8a, 12a–g and
13a–j is presented in Scheme 1. The key intermediate,
the trimethyl ester of (L,D)-Ala-w(PO2-CH2)(L,D)-Glu
(9), was prepared according to a known procedure.6
Acetaldehyde, diphenylmethylamine hydrochloride and
Keywords: MurD; Inhibitors; Antibacterials.
*
Corresponding author. Tel.: +386 1 47 69 585; fax: +386 1 42 58
031; e-mail: gobecs@ffa.uni-lj.si
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.09.086

ˇ
K. Strancar et al. / Bioorg. Med. Chem. Lett. 16 (2006) 343–348
344
PO2H-UMP
O
PO3H2
O
OH
OH
O
O
substitute for
phospho-sugar residue
NHCOCH3
NHCOCH3
HO
HO
O
O
H3C
H3C
variations of linker
O
O
HN
HN
P
H2O3P
O
substitute for
phosphinoalanine
O
CH3
CH3
HO
OH
NH
HOOC
HOOC
tetrahedral mimic of
transition state
COOH
COOH
1
2
truncation of "glutamate" residue
Figure 1. Transition state of the reaction catalyzed by MurD (1) and the structure of the potent phosphinate inhibitor 2 (IC50 = 20 nM). Structural
features of the compounds synthesized are mentioned on the right-hand side.
Ph
Ph
HN
H2O,
reflux
1. conc. HCl,
reflux
Cbz-Cl,
NaOH/dioxane
O
P
O
P
O
P
O
P
Cbz
HN
OH
H
OH
H
OH
H
OH
H
Ph
NH+
3Cl
H2N
EDC/MeOH
+
H
Ph
2. propylene
oxide
CH3
CH3
CH3
H
O
3
4
5
CH3
COOMe
O
O
P
Cbz
HN
Cbz
O
P
COOMe
OMe
H
1. MeONa/MeOH
2. dimethyl
H2/Pd/C
MeOH
P
CH3
6
H2N
HN
OMe
CH3
OMe
CH3
2-methylenepentanedioate (7)
COOMe
COOMe
9
8
R1-COOH,
EDC, HOBt
DCM, base R2-SO2Cl
O
DMF, base
O
O
R2
R1
S
O
P
COOMe
O
P
COOMe
HN
HN
OMe
OMe
CH3
CH3
COOMe
COOMe
10a-g
11a-j
1 M LiOH,
dioxane
1 M LiOH,
dioxane
1 M LiOH,
dioxane
O
R2
O
O
R1
O
O
O
S
COOH
O
P
COOH
O
P
COOH
HN
P
HN
HN
OH
CH3
OH
OH
CH3
CH3
COOH
COOH
COOH
13a-j
8a
12a-g
Scheme 1. Synthesis of phosphinate inhibitors 8a, 12a–g and 13a–j.
hypophosphorous acid were condensed into diph-
enylmethylaminophosphonous acid 3, which was
converted in two steps into Cbz N-protected a-amin-
oalkylphosphonous acid 5.15 In our hands, the
preparation of the methyl ester of N-protected a-amin-
oalkylphosphonous acids (e.g., 6) initially caused much
trouble. From the literature it is known that a-amin-
oalkylphosphonous acids are prone to oxidation,16 but
we found that compound 6 decomposed back to the
starting phosphinic acid 5. For this reason, the formation
of the methyl ester was carried out using EDC and dry
MeOH in an argon atmosphere, and the product was
used immediately in the next reaction step. From inter-
mediates 8 and 9 we obtained carbamate 8a, amides
12a–g and sulfonamides 13a–j. Alkaline hydrolysis of tri-
methyl ester 8 afforded compound 8a. From the crucial
intermediate amine 9 the synthesis proceeded in two
directions: the amino group was substituted by different
moieties mimicking the MurNAc residue of lead com-
pound 2 via either an amide (10a–g) or a sulfonamide
linkage (11a–j). The resulting trimethyl esters 10a–g
and 11a–j were converted, by alkaline hydrolysis, into
the target compounds 12a–g and 13a–j, respectively.
These compounds were tested for their inhibitory activity
on MurD from Escherichia coli.17–19 Results are present-
ed as residual activity (RA) of the enzyme in the presence
of 1 mM inhibitor (Table 1). For the most active
compounds, IC50 values have also been determined.17
The results for compounds 8a and 12a–g are shown in
Table 1. Compound 8a had previously been prepared

methylene group was used. For compound 12e, the D
-al-
ˇ
K. Strancar et al. / Bioorg. Med. Chem. Lett. 16 (2006) 343–348
345
Table 1. MurD inhibitory activity of carbamate- and amide-substitut-
ed phosphinates 8a and 12a–g
anyl residue was introduced in order to mimic the D-lac-
toyl residue of inhibitor 2. However, both compounds
were found to be less effective inhibitors (Table 1).
R1
O
O
P
COOH
HN
OH
In spite of these results, compounds 12f and 12g, in
which the nitrobenzylsulfonyl moiety is linked to the
D-alanyl residue, were prepared. The latter, with a nitro
group at the para position, exhibited the strongest inhib-
itory activity (IC50 = 78 lM). Its meta analogue, 12f,
inhibited the enzyme about 2 to 3 times less strongly
(Table 1).
COOH
RA (%)a
Compound
R1
8a
41
2
O
The sulfonamides 13a–j consist of key phosphinodipep-
tide Ala-W(PO2-CH2)-Glu substituted by several func-
tionalized phenylsulfonyl and benzylsulfonyl moieties
(Table 2). The most active compounds in this series were
found to be 13i and 13h, with the m-nitrobenzylsulfonyl
12a
12b
12c
28 2 (432 28 lM)
OH
23
2
O
O
O
17 1 (95 15 lM)
Table 2. MurD inhibitory activity of sulfonamide-substituted
phosphinates 13a–j
O
R2
O
O
O
S
O
P
COOH
12d
12e
36
44
1
1
HN
OH
COOH
O
R2
RA (%)
O
Compound
N
H
R
13a
54
52
55
40
49
5
2
1
1
2
O
12f
19
8
3
S
H
N
O2N
R
O
13b
13c
13d
13e
O2N
O2N
O
S
12g
1 (78 19 lM)
H
N
R
O
a IC50 values are shown in parentheses.
and evaluated as an inhibitor of MurD from Streptococ-
cus pneumoniae with IC50 100 lM.12 From our RA value
we estimate that the IC50 of this compound for the en-
zyme from E. coli is almost one order of magnitude
greater. This relatively large difference could be ex-
plained by different assay conditions as well as by differ-
ences in the active sites of MurD from E. coli and S.
pneumoniae.
CF3
13f
50
37
2
2
CF3
Cl
13g
The introduction of the trans-cinnamoyl moiety in com-
pound 12a resulted in a higher inhibitory activity than in
compound 8a. This trend was also observed in com-
pound 12b, which possesses a 3-hydroxy group on the
trans-cinammoyl residue. Further improvement was
achieved with compound 12c, in which the trans-cinna-
moyl moiety was further substituted by the 3,4-methy-
lenedioxy ring. Since this 1,3-benzodioxolyl fragment
proved to be a promising pharmacophore, we prepared
two new compounds in which the same pharmacophore
was connected to the phosphinodipeptide by two differ-
ent linkers. For compound 12d, the shorter, flexible
F
NO2
13h
13i
13j
30
22
47
5
2
2
O2N
O2N

ˇ
K. Strancar et al. / Bioorg. Med. Chem. Lett. 16 (2006) 343–348
346
Table 3. MurD inhibitory activity of phosphinates 14–16 with
branched side chains and comparison with phosphinoalanine deriva-
tive 8a
In order to study the importance of the phosphinoala-
nine residue, we prepared analogues of 8a in which it
was replaced by phosphinovaline (14), phosphinoleucine
(15) and phosphinoisoleucine (16). These compounds
were synthesized by a procedure similar to that de-
scribed in Scheme 1, using an appropriate starting alde-
hyde, but with some improvements which will be
published elsewhere. All compounds inhibited MurD
less than the parent compound 8a (Table 3). Thus,
replacement of the phosphinoalanine side chain by bulk-
ier ones decreases the inhibitory activity, presumably
due to the small volume of the L-alanine-binding subsite.
O
P
COOH
H
O
N
OH
O
R3
COOH
Compound
R3
RA (%)
8a
41
66
2
1
CH3
14
H3C
CH3
CH3
In order to determine the importance of the c-carboxyl-
ate group of the phosphinodipeptide glutamate residue,
we prepared compounds 21, 22 and 23 as analogues of
12c, 12g and 8a, respectively, but lacking the glutamic
acid side chain. The synthesis of these truncated ana-
logues is similar to that presented in Scheme 1, using
methyl acrylate instead of dimethyl 2-methylenepen-
tanedioate 7 (Scheme 2). Compounds 21–23 did not
inhibit MurD (Table 4), confirming our previous
hypothesis9 that the presence of the glutamate residue,
or its appropriate mimetic, in this type of inhibitor is
essential for strong inhibition.
15
16
65
83
1
3
CH3
CH3
H3C
and o-nitrobenzylsulfonyl substituents, respectively.
Their positional isomer 13j inhibited the enzyme to a
lesser extent, as did the homologue of 13i (m-nitrophenyl
substituted derivative 13b). However, all of them were
less potent than compound 12g.
In the complex of MurD with its product UDP-
MurNAc-L-Ala-D-Glu (MurD Æ UMAG, pdb entry
4UAG),20 the a-carboxylic acid of D-glutamic acid is
H-bonded to Lys348 and Thr321, the c-carboxylate is
1. MeONa/MeOH
O
P
H2/Pd/C,
O
Cbz
HN
Cbz O
HN
OM e
H
2. methylacrylate
MeOH
H2N
P
P
COOMe
COOMe
OM e
OM e
CH3
CH3
CH3
18
4
17
O
O
COOH
O
O
O
1 M LiOH,
dioxane
O
O
O
O
EDC, HOBt
DMF, base
O
HN
P
HN
P
COOH
COOMe
OH
OMe
CH3
CH3
O2N
O2N
21
19
1 M LiOH,
dioxane
O
O
S
S
NH
HN
H
NH
O
N
O
O
COOH
S
O
O
O
O
H3C
H3C
O
P
O2N
P
HN
COOMe
COOH
OMe
OH
CH3
CH3
20
22
1 M LiOH,
dioxane
O
O
O
O
HN
O
O
P
OH
HN
P
COOMe
OMe
COOH
CH3
CH3
17
23
Scheme 2. Synthesis of the truncated phosphinates 21–23.

ˇ
K. Strancar et al. / Bioorg. Med. Chem. Lett. 16 (2006) 343–348
347
Table 4. MurD inhibitory activity of truncated compounds 21–23 and comparison with that of their non-truncated analogues
Compound
R4
RA (%)a
O2N
O
P
COOH
R4
O
S
H
N
N
OH
O
H
O
12g
22
8
1 (78 19 lM)
COOH
100
5
H
O
O
O
COOH
R4
H
N
P
OH
O
12c
21
17 1 (95 15 lM)
COOH
91
1
H
O
P
COOH
R4
H
O
N
OH
O
8a
40
92
2
1
COOH
23
H
a IC50 values are shown in parentheses.
held in place by hydrogen bonds with Ser415 and Phe
422, and the amide bond between lactoyl moiety and
L-Ala interacts with the side chain of Asn138.20 The ab-
sence of inhibition by truncated analogues 21–23 is con-
sistent with the loss of binding energy of the two
hydrogen bonds provided by the c-carboxylate group.
We can thus expect that the best phosphinate inhibitors
found here (12c, 12f and 12g) bind the enzyme in a sim-
ilar manner as UMAG. It is possible that they utilize
their substituted aromatic rings to interact also with
Leu15, Thr16 or Gln162, which are the residues that
bind the phospho-sugar part of UMAG.
the Ministry of Education, Science and Sport of the
Republic of Slovenia, the Centre National de la Recher-
che Scientifique, the Proteus programme of integrated
actions and Lek Pharmaceuticals d.d.
References and notes
1. Courvalin, P.; Davies, J. Curr. Opin. Microbiol. 2003, 6,
425.
2. Labischinski, H.; Maidhof, H. In Bacterial Cell Wall;
Ghuysen, J.-M., Hakenbeck, R., Eds.; Elsevier Science
B.V.: Amsterdam, 1994; p 23.
3. van Heijenoort, J. Nat. Prod. Rep. 2001, 18, 503.
4. Pratviel-Sosa, F.; Acher, F.; Trigalo, F.; Blanot, D.;
Azerad, R.; van Heijenoort, J. FEMS Microbiol. Lett.
1994, 115, 223.
5. El Zoeiby, A.; Sanschagrin, F.; Levesque, R. C. Mol.
Microbiol. 2003, 47, 1.
6. Tanner, M. E.; Vaganay, S.; van Heijenoort, J.; Blanot, D.
J. Org. Chem. 1996, 61, 1756.
7. Gegnas, L. D.; Waddell, S. T.; Chabin, R. M.; Reddy, S.;
Wong, K. K. Bioorg. Med. Chem. Lett. 1998, 8, 1643.
8. Auger, G.; van Heijenoort, J.; Blanot, D.; Deprun, C. J.
Prakt. Chem. 1995, 337, 351.
In the present paper, we reported the synthesis and
activity of a series of new phosphinate inhibitors of
MurD from E. coli. A common feature of all the active
compounds is the N-substituted phosphinodipeptide
Ala-w(PO2-CH2)-Glu, both alanine and glutamate resi-
dues being essential in this series. Compounds 12c and
12g had IC50 values near 100 lM. Since they consist of
mixtures of four diastereoisomers, the actual IC50 for
the one most closely mimicking the L-Ala-D-Glu stereo-
chemistry is presumably lower than that observed. These
compounds constitute promising starting points for
further structural modifications.
9. Gobec, S.; Urleb, U.; Auger, G.; Blanot, D. Pharmazie
2001, 56, 295.
10. Horton, J. R.; Bostock, J. M.; Chopra, I.; Hesse, L.;
Phillips, S. E. V.; Adams, D. J.; Johnson, A. P.; Fishwick,
C. W. G. Bioorg. Med. Chem. Lett. 2003, 13, 1557.
11. Li, Z.; Francisco, G. D.; Hu, W.; Labthavikul, P.;
Petersen, P. J.; Severin, A.; Singh, G.; Yang, Y.;
Rasmussen, B. A.; Lin, Y.-I.; Skotnicki, J. S.; Mansour,
T. S. Bioorg. Med. Chem. Lett. 2003, 13, 2591.
Acknowledgments
We thank Prof. Roger Pain for critical reading of the
manuscript. This work was financially supported by

ˇ
K. Strancar et al. / Bioorg. Med. Chem. Lett. 16 (2006) 343–348
348
12. Snyder, N. J.; Tebbe, M. J.; Victor, F.; Blaszczak, L. C.;
Halligan, N. G.; Sigmund, S.; Thompson, R. C.; Wilkie, S.
C.; Stack, D. R.; Lee, C.; Birch, G. B.; Wu, C. E.; Smith,
M. C. Abstracts of the 39th Interscience Conference on
Antimicrobial Agents and Chemotherapy, September 26–
29, San Francisco, USA, 1999; p 330.
13. Victor, F.; Tebbe, M. J.; Birch, G. B.; Smith, M. C.;
Letourneau, D. L.; Wu, C. E. Abstracts of the 39th
Interscience Conference on Antimicrobial Agents and Che-
motherapy, September 26–29, San Francisco, USA, 1999;
p 330.
14. Kotnik, M.; Oblak, M.; Humljan, J.; Gobec, S.; Urleb, U.;
ˇ
Solmajer, T. QSAR Comb. Sci. 2004, 23, 399.
15. Baylis, E. K.; Campbell, C. D.; Dingwall, J. G. J. Chem.
Soc., Perkin. Trans. 1 1984, 2845.
16. Bartlett, P. A.; Hanson, J. E.; Morgan, B. P.; Ellsworth, B.
A. In Houben-Weyl: Synthesis of Peptides and Peptidom-
imetics, 4th ed.; Goodman, M., Ed.; Houben-Weyl Meth-
ods in Organic Chemistry; Thieme: Stuttgart, 2003; Vol.
E22c, p 492.
a Nucleosil 5C18 column (150 · 4.6 mm) as stationary
phase, and isocratic elution at a flow rate of 0.6 ml/min
with 50 mM ammonium formate, pH 4.7. Compounds
were detected and quantified with an HPLC radioactivity
monitor LB 506 C-1 (Berthold France, Thoiry, France)
using Quickszint Flow 2 scintillator (Zinsser Analytic,
Maidenhead, UK) at 0.6 ml/min. Residual activity was
calculated with respect to
a similar assay without
inhibitor. Values are expressed as means standard
deviations of duplicate determinations. IC50 values were
determined from a range of inhibitor concentrations;
values standard deviations at 95% of confidence were
calculated from the fitted regression equations using the
logit-log plot.
18. Auger, G.; Martin, L.; Bertrand, J.; Ferrari, P.; Fanchon,
´
E.; Vaganay, S.; Petillot, Y.; van Heijenoort, J.; Blanot,
D.; Dideberg, O. Protein Expr. Purif. 1998, 13, 23.
19. All the compounds gave satisfactory spectroscopic data
and elemental analyses. Representative results for com-
pound 12c are given: mp 82–84 ꢁC; IR (KBr): m 3427, 1717,
17. Enzymatic assays were performed as described8 with
slight modifications. The compounds were tested for their
ability to inhibit the addition of D-[14C]Glu to UDP-
MurNAc-L-Ala in a mixture (final volume: 50 ll) con-
taining 0.1 M Tris–HCl, pH 8.6, 5 mM MgCl2, 25 lM
UDP-MurNAc-L-Ala, 25 lM D-[14C]Glu (50,000 cpm),
5% (v/v) DMSO, and purified MurD18 (diluted with
20 mM potassium phosphate, pH 7.0, 1 mM dithiothre-
itol and 1 mg/ml BSA) and 1 mM tested compound (all
compounds were soluble in the assay mixture containing
5% DMSO). The mixture was incubated for 30 min at
37 ꢁC, and the reaction was stopped by adding 10 ll
glacial acetic acid. The mixture was lyophilized and taken
up in the HPLC elution buffer. The radioactive substrate
and product were separated by reverse-phase HPLC with
1653, 1449, 1254, 1036, 970, 807 cmÀ1 1H NMR
;
(300 MHz, DMSO-d6): d 1.14–1.28 (m, 3H, –CH3), 1.53–
2.33 (m, 6H, –CH2–CH(COOH)–CH2–CH2–COOH),
2.57–2.70 (m, 1H, –CH2–CH(COOH)–CH2), 4.07–4.23
3
(m, 1H, –PCH), 6.06 (s, 2H, –O–CH2–O–), 6.59 (d, JH–
H (E) = 15.1 Hz, 1H, –HC@CH–CONH), 6.95 (d,
Jortho = 7.9 Hz, 1H, ArH-5), 7.07 (d, Jortho = 7.9 Hz, 1H,
3
ArH-6), 7.11 (s, 1H, ArH-2), 7.36 (d, JH–H (E) = 15.4 Hz,
1H, Ar-HC@CH–). 31P NMR (DMSO-d6): d 45.65 (1P),
45.27 (1P). MS (FAB): m/z 428 (M+H)+; Anal. Calcd for
C18H22NO9P · H2O: C, 48.54; H, 5.43; N, 3.14. Found: C,
48.75; H, 5.42; N, 2.99.
20. Bertrand, J. A.; Auger, J.; Martin, L.; Fanchon, E.;
Blanot, D.; Le Beller, D.; van Heijenoort, J.; Dideberg, O.
J. Mol. Biol. 1999, 289, 579.