Metallo-β-lactamase inhibitors by bioisosteric replacement: Preparation, activity and binding

European Journal of Medicinal Chemistry 135 (2017) 159e1 73

Contents lists available at ScienceDirect

European Journal of Medicinal Chemistry

journal homepage: h ttp://www.elsevier.com/locate/ejmech

Research paper

Metallo-

b-lactamase inhibitors by bioisosteric replacement:

Preparation, activity and binding

,

Susann Skagseth ^{a 1}, Sundus Akhter ^{b 1}, Marianne H. Paulsen ^{b 2}, Zeeshan Muhammad ^b,

Silje Lauksund ^c, Ørjan Samuelsen ^{c d}, Hanna-Kirsti S. Leiros ^{a **}, Annette Bayer ^b

,

, *

^aThe Norwegian Structural Biology Centre (NorStruct), Department of Chemistry, Faculty of Science and Technology, UiT The Arctic University of Norway, N-

9037 Tromsø, Norway

^bDepartment of Chemistry, Faculty of Science and Technology, UiT The Arctic University of Norway, N-9037 Tromsø, Norway

^cNorwegian National Advisory Unit on Detection of Antimicrobial Resistance, Department of Microbiology and Infection Control, University Hospital of

North Norway, N-9038 Tromsø, Norway

^dDepartment of Pharmacy, UiT The Arctic University of Norway, N-9037 Tromsø, Norway

a r t i c l e i n f o

a b s t r a c t

Article history:

Bacterial resistance is compromising the use of

b

-lactam antibiotics including carbapenems. The main

-lactam ring mediated by serine- or

-lactamases (MBLs). Although several inhibitors of MBLs have been reported, none has been

Received 12 December 2016

Received in revised form

24 March 2017

resistance mechanism against

metallo-

b-lactams is hydrolysis of the b

b

developed into a clinically useful inhibitor. Mercaptocarboxylic acids are among the most prominent

scaffolds reported as MBL inhibitors. In this study, the carboxylate group of mercaptocarboxylic acids was

replaced with bioisosteric groups like phosphonate esters, phosphonic acids and NH-tetrazoles. The

inﬂuence of the replacement on the bioactivity and inhibitor binding was evaluated. A series of bio-

isosteres of previously reported inhibitors was synthesized and evaluated against the MBLs VIM-2, NDM-

1 and GIM-1. The most active inhibitors combined a mercapto group and a phosphonate ester or acid,

with two/three carbon chains connecting a phenyl group. Surprisingly, also compounds containing

thioacetate groups instead of thiols showed low IC₅₀values. High-resolution crystal structures of three

inhibitors in complex with VIM-2 revealed hydrophobic interactions for the diethyl groups in the

phosphonate ester (inhibitor 2b), the mercapto bridging the two active site zinc ions, and tight stacking

of the benzene ring to the inhibitor between Phe62, Tyr67, Arg228 and His263. The inhibitors show

reduced enzyme activity in Escherichia coli cells harboring MBL. The obtained results will be useful for

further structural guided design of MBL inhibitors.

Accepted 11 April 2017

Available online 14 April 2017

Keywords:

Crystal structure

Inhibition properties

Carboxylate bioisosters

Thiols

Metallo-b-lactamase inhibitors

1. Introduction

serine dependent

b-lactamases (SBLs; Amber class A, C, and D) and

metallo- -lactamases (MBLs; Amber class B) [5,6]. The MBLs are

b

Antibiotics are essential for modern medicine and as antibiotics

are now becoming increasingly ineffective, effective treatment of

an ever-increasing range of infections is threatened [1,2]. As

classiﬁed into three subclasses (B1, B2 and B3), with the majority of

nosocomial MBLs belonging to the B1 subclass. This includes vari-

ants of the Verona integron-encoded metallo-b-lactamase (VIM),

defence mechanism against

b

-lactam antibiotics, bacteria are able

the New Delhi metallo- -lactamase (NDM) [7], and the German

b

to produce

biotic [3,4].

b

-lactamases that hydrolyze and inactivates the anti-

-Lactamases are grouped into two super families, the

Imipenemase (GIM). The active site of the B1 MBLs is generally well

preserved despite large variation in the second shell interacting

residues. The active site holds two cationic zinc ions (Zn1, Zn2), as a

metal co-factor of enzyme activity, which are coordinated by H116,

H118 and H196 (Zn1) and D120, C221 and H263 (Zn2) [8,9]. Here

and throughout the paper, the standard numbering scheme for

* Corresponding author.

** Corresponding author.

E-mail addresses: hanna-kirsti.leiros@uit.no (H.-K.S. Leiros), annette.bayer@uit.

no (A. Bayer).

These authors have contributed equally to this work.

Current address: Department of Pharmacy, UiT The Arctic University of Norway,

N-9037 Tromsø, Norway.

class B b-lactamases is used [10]. In general, the sequence identity

among the class B1 MBLs is low. For the enzymes used in this study,

the sequence identity is 32% between VIM-2 and NDM-1, 28% for

VIM-2 versus GIM-1 and 24% between NDM-1 and GIM-1 (Fig. S1 in

1

2

http://dx.doi.org/10.1016/j.ejmech.2017.04.035

160

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

SI).

were prepared as racemic mixtures. Phosphonate ester 2aec and

The MBLs hydrolyze a broad spectrum of

b

-lactam antibiotics

phosphonic acid 3aec analogues with differing chain lengths

(n ¼ 2e4) were prepared according to the synthetic strategy pre-

sented in Scheme 2. Triethyl phosphonoacetate 5 was alkylated

using potassium tert-butoxide (KOtBu) as base to afford the mono-

alkylated acetates 6aec in moderate yields (45e73%). Chemo-

selective reduction of the ester in presence of the phosphonate was

obtained with lithium borohydrid to provide the corresponding

alcohols 7aec. Subsequent mesylation followed by substitution

with potassium thioacetate gave the thioacetates 8aec. Several

methods for deprotection of the thioacetates were evaluated

[36e38]. Best results were obtained by treatment with sodium

methylthiolate (NaSMe) providing the free thiols 2aec in good

yields (74e77%) and purity (>95% by HPLC). The phosphonates

ester analogues 2, 7 and 8 were puriﬁed by normal-phase ﬂash

column chromatography to >95% purity as determined by HPLC

analysis. The diethyl phosphonates 2, 7 and 8 were transferred to

the corresponding phosphonic acid derivatives 3, 9 and 10,

respectively, by treatment with trimethylsilyl bromide (TMSBr)

followed by a MeOH quench. Puriﬁcation of the phosphonic acid

analogues 3, 9 and 10 could be achieved by washing with EtOAc/

pentane to provide the target compounds with moderate to good

yields and purity >95% as determined by HPLC analysis.

[3,8,11,12] including the last resort antibiotics; the carbapenems. In

addition, genes encoding certain MBLs are associated with mobile

genetic elements and can spread across different types of Gram-

negative bacteria [7,11]. Inhibitors of SBLs are available clinically

[13], including the broad spectrum inhibitor avibactam [14]. Un-

fortunately, all SBL inhibitors lack activity against MBLs [15].

Although the search for MBL inhibitors started from mid 1990s, and

a wide variety of scaffolds has been reported [16e18], clinically

useful MBL inhibitors remain an elusive goal.

In this study we decided to explore the use of thiol-based

compounds as one of our strategies to gain increased knowledge

on MBL inhibitors. Thiols have the potential to coordinate the zinc

ions in the MBL active site, due to the thiophilic nature of zinc, and

thereby preventing b-lactam hydrolysis. Some mercaptocarboxylic

acids have been evaluated successfully against several MBLs

[19e32], making this class of compounds an interesting starting

point for further investigations. Our focus was to elucidate the ef-

fect of replacement of the carboxylate group in the promising

mercaptocarboxylic acid scaffold with bioisosters on the bioactivity

and enzyme-inhibitor binding. Bioisosteric replacement is

a

concept in medicinal chemistry to rationally improve the activity or

physicochemical properties of biologically active compounds [33].

A structural element of a biological active compound is replaced

with a substitute that maintain some of the properties of the parent

structure, e.g. inhibitor binding to a biological target, while others

are changed, e.g. lipophilicity or steric size. Typical bioisosters of

carboxylic acids include among others phosphonates, NH-tetra-

zoles and sulphonamides [34].

The target structures 2e4 (Scheme 1) were envisioned by sub-

stitution of the carboxylate group of the known mercaptocarboxylic

acid inhibitor 1 [25,35] with bioisosteric groups like phosphonate

esters, phosphonic acids or NH-tetrazoles, respectively. The carbon

chain of the alkylphenyl substituent was varied (n ¼ 2e4; Scheme

1) with the goal to evaluate the effect of chain length on hydro-

phobicity and bioactivity. The compounds and relevant in-

termediates from the synthetic pathway were simultaneous

evaluated against the three MBLs VIM-2, NDM-1 and GIM-1 with

the goal to ﬁnd MBL inhibitors able to act on several different MBL

subfamilies. For each target we have determined the IC₅₀value

through enzyme inhibition assays and the effect of the inhibitors in

whole cell E. coli assays with VIM-2, GIM-1 or NDM-1. Some in-

hibitors were also tested in a synergy assay with two or three

clinical isolates from Pseudomonas aeruginosa, Klebsiella pneumo-

niae or Escherichia coli. The enzyme-inhibitor complexes of the best

inhibitors were further evaluated by X-ray analysis and the impact

from second shell interacting residues on ligand binding were

examined.

The NH-tetrazole analogues 4, 15 and 16 were prepared as

shown in Scheme 3 a-Substituted acrylonitriles 13a and 13b with

varying chain lengths (n ¼ 1, 3) were prepared by a procedure

based on the work of Baraldi et al. [39]. Microwave promoted re-

action of the acrylonitriles with trimethylsilyl azide with dibu-

tyltinoxide as catalyst (20 mol%) [40] gave the corresponding NH-

tetrazoles 14a and 14b (74e78% yield), which proved to be excellent

Michael acceptors. The addition of potassium thioacetate resulted

in the tetrazolyl thioacetates 15a and 15b in good yields (>95%),

while the addition of cyclohexanethiol gave 16 (93% yield).

Deprotection of the thioacetates by treatment with sodium meth-

ylthiolate (NaSMe) provided the free thiol 4 in moderate yield (74%)

and purity >96% as determined by HPLC analysis.

2.2. Characterization of inhibitor properties against VIM-2, GIM-1

and NDM-1

The inhibitory activities of the mercaptocarboxylic acid 1c [25]

and bioisosters 2aec, 3aec and 4, as well as several in-

termediates from the synthesis (7e10 aec, 15a, 15b and 16) were

determined by the half maximal inhibitory concentration (IC₅₀

)

values (Table 1). For VIM-2 and GIM-1, the IC₅₀values were

measured using nitroceﬁn as a reporter substrate, while IC₅₀values

for NDM-1 were measured with imipenem as reporter substrate.

Nitroceﬁn is hydrolyzed by NDM-1 with a high catalytic efﬁciency

and unsuitable as a reporter substrate for NDM-1 [41]. The inﬂu-

ence of dimethyl sufoxide (DMSO) on the enzyme activity was

investigated in the assay, and a concentration of 2.5% DMSO was

tolerated without inﬂuencing the enzyme activity (data not

shown). The IC₅₀values were determined by measuring the initial

rate of the reactions with inhibitors at different concentration in a

2-fold dilution series, and were ﬁtted to a dose-response curve (IC₅₀

2. Results and discussion

2.1. Chemistry

For the purpose of this work, all compounds (Schemes 2 and 3)

Scheme 1. Design of new MBL inhibitors through bioisosteric substitution of the carboxylate group.

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

161

Scheme 2. Synthesis of phosphonate and phosphonic acid containing thiol-based inhibitors. Reagents and conditions: a: n ¼ 2; b: n ¼ 3; c: n ¼ 4; (a) RdBr, KOtBu, DMF, 0 ^ꢀC, 6a:

62%, 6b: 73%, 6c: 45%; (b) LiBH₄, THF, MW 80 ^ꢀC for 10 min, 7a: 60%, 7b: 95%, 7c: 56%; (c) MsCl, Et₃N, DMAP, CH₂Cl₂, rt; (d) KSAc, DMF, rt.; 8a: 54%, 8b: 34%, 8c: 71% over two steps;

(e) NaSMe, MeOH, ꢁ20 ^ꢀC, 2a: 30%, 2b: 74%, 2c: 77%; (f) TMSBr, CH₂Cl₂, then MeOH, rt., 3a: 96%, 3b: 90%, 3c: 98%, 9a: 76%, 9b: 61%, 9c: 55%, 10a: 91%, 10b: 71%, 10c: 65%.

Scheme 3. . Synthesis of NH-tetrazole containing thiol based inhibitors. Reagents and conditions: a: n ¼ 1; b: n ¼ 3; (a) RdBr, K₂CO₃, acetone, reﬂux; (b) 5% aq. NaOH, 13a: 51%, 13b:

21%; (c) TMSN₃, n-Bu₂SnO (20 mol%), 1,4-dioxane, MW 150 ^ꢀC for 50 min, 14a: 78%, 14b: 74%; (d) HSAc, DMF, 60 ^ꢀC, 15a: 95%, 15b: 98%; (e) NaSMe, MeOH, ꢁ20 ^ꢀC, 4: 74%; (f)

cyclohexylthiol, DMF, 60 ^ꢀC, 16: 93%.

curves are given in Fig. S2eS4 in SI).

In order to validate our assay, the IC₅₀of the previously reported

VIM-2 inhibitor 1c [25] was measured against VIM-2. We obtained

from 1.8 to 4.7

respectively. With GIM-1 the inhibitors 10aec showed slightly

higher IC₅₀values of 12e26 M. The same trend was seen with the

mercaptophosphonic acids 3a and c, where the IC₅₀values for VIM-

2 and NDM-1 were in the same range (7.8 and 8.6 M with VIM-2

and 5.9 and 8.5 M with NDM-1, respectively), while the values for

GIM-1 were 2e4 fold higher (23 and 16 M, respectively). A

possible explanation for this might be that GIM-1 has a narrower

active site compared to other MBLs due to the aromatic side chains

Trp228 and Tyr233 [42]. However, this does not explain the low

IC₅₀values for inhibitors 2a and 2b.

Substituting the carboxylic acid of compound 1 with a NH-tet-

razole (4) had a deteriorating effect on the inhibitory activity, while

substitution with phosphonate esters and phosphonic acids groups

led to similar (2 and 10) or improved (3) activity. Comparing the

inactive alcohols 7 and 9 with the corresponding active thiols and

thioacetates 2, 3 and 10 illustrates the importance of the sulfur

mM, and from 2.5 to 6.6 mM for VIM-2 and NDM-1,

m

an IC₅₀of 2.9

of 1.1 M (Table 1). The difference in IC₅₀values is most likely due to

different assay buffers and protein constructs. The new compounds

showed IC₅₀values ranging from 0.38 to 133 M with VIM-2, 0.18 to

>5000 M with GIM-1 and 1.8e144 M with NDM-1 (Table 1 and

S1 in SI). The synthetic intermediates 7e9 aec, 15a and 15b did not

have inhibitory activity (IC₅₀> 10 M, see Table S1 in SI) against any

mM, which is in the same range as the reported value

m

of the MBLs. Compounds 2aec, 3aec, 4, 10aec and 16 showed ac-

tivity at varying levels. In general, the inhibitors showed better IC₅₀

values towards VIM-2 compared to GIM-1 and NDM-1, with the

exception of inhibitor 2a, which had a ~ﬁve times lower IC₅₀for

GIM-1 (0.18

Inhibitors 2a and 2b showed IC₅₀values ranging from 0.18 to

M for all three MBLs. The IC₅₀s were within the same range for

VIM-2 and GIM-1, and slightly increased for NDM-1 (1.8e2.2 M).

mM) compared to VIM-2 (0.89 mM).

2.2

m

atom on the b-carbon for the inhibitor activity. However, activity

Inhibitor 2c precipitated from the buffer solution and was not

evaluated further. The results indicate that for the mercapto-

phosphonates 2, lengthening the side chain from two (2a) to three

methylene groups (2b), did not have a great effect on the activity.

While the mercaptophosponate esters 2 showed highest activity

against GIM-1, the mercapto- and thioacetate phosphonic acids, 3

and 10, were more active towards VIM-2 and NDM-1 than GIM-1.

The IC₅₀values for thioacetate phosphonic acids 10aec ranged

seem to depend on a subtle combination of the sulfur atom and a

carboxylic acid bioisoster as revealed when comparing the mer-

capto and thioacetate substituted phosphonate esters and phos-

phonic acids 2, 3, 8 and 10. The mercapto substituted phosphonate

esters 3 were the overall most active inhibitors, while the thio-

acetate substituted phosphonate esters 8 showed the lowest ac-

tivity of all four. Surprisingly, the same trend was not observed for

the pair of phosphonic acids 2 and 10. Here, the thiol 2 and the

162

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

Table 1

Evaluation of compound 1c and bioisosteres thereof as inhibitors of VIM-2, GIM-1 or NDM-1 measured as inhibition concentrations (IC₅₀) against puriﬁed enzyme and percent

inhibition in E. coli SNO3 bacterial whole cell experiments.

Compound

VIM-2^a

GIM-1^a

NDM-1^b

IC₅₀M)

(

m

% inhib

IC₅₀

(m

M)

% inhib

IC₅₀

56

(m

M)

% inhib

1c (n ¼ 4)

2.9 (1.1)^c

e

2a (n ¼ 2)

2b (n ¼ 3)

2c (n ¼ 4)

0.89

0.38

p^d

78

79

p^d

0.18

0.31

p^d

64

48

p^d

2.2

1.8

p^d

33

62

p^d

3a (n ¼ 2)

3b (n ¼ 3)

3c (n ¼ 4)

7.8

33.2

8.6

92

95

94

23

28

16

37

46

31

5.9

7.1

8.5

42

45

42

4 (n ¼ 3)

28

67

68

25

12

16

8a (n ¼ 2)

8b (n ¼ 3)

8c (n ¼ 4)

133

34

20

17

14

12

18

26

13

25

21

25

nh^e

i^f

10a (n ¼ 2)

10b (n ¼ 3)

10c (n ¼ 4)

2.3

4.7

1.8

94

95

97

12

26

20

40

37

50

2.9

6.6

2.5

39

38

37

16 (n ¼ 3)

5

73

36

8

10

i^f

a

Reporter substrate used was nitroceﬁn (NCF).

Reporter substrate used was imipenem (IPM).

Values in parentheses as reported by Jin et al. [25].

p ¼ precipitated in buffer and was not tested.

nh ¼ no hydrolysis.

b

c

d

e

f

i ¼ inactive.

thioacetate 10 inhibited the MBLs at the same level, which may

indicate that the interaction of the phosphonic acid with the

enzyme is more important for the activity than the interaction with

the thiol. This interpretation would be in agreement with the

previously reported observation that a mercaptophosphonic acid

coordinated the MBL CphA by the two oxygen atoms of the phos-

phonic acid and not by the sulfur atom of the thiol [43].

bacterial cells, two different assays were used. Initially, the in-

hibitors were tested in a whole cell assay using E. coli SNO3 cells

with bla_VIM-2, bla_GIM-1or bla_NDM-1in pET26b(þ) (Table 1). The

inhibitory activity were measured as percent inhibition calculated

by measuring the slope of the initial hydrolysis of the reporter

substrates for up to 60 min, and the slope of a control without in-

hibitor present, according to equation (1).

The most active inhibitors showed activity in the low micro-

molar to high nanomolar range for the racemic mixtures. In two

recent studies, pure stereoisomers of the mercaptocarboxylic acids

captopril [19] and bisthiazolidine [21] were found to display

10e100 times differences in inhibitory activity between different

stereoisomers. This will most likely be the case for the compounds

prepared in this work, and for future work this can be pursued. A

large variation in inhibitory activity towards three B1 MBLs was

observed, which indicates that these inhibitors do not fulﬁll the

requirements for a broad-spectrum MBL inhibitor.

The percent inhibition for the most active compounds is shown

in Table 1. High percent inhibition (>70%) was observed for several

compounds, which indicates that the compounds pass the bacterial

cell membrane of E. coli. The highest percent inhibition was

observed for inhibitors 3aec and 10 aec in E. coli with VIM-2, with

92e95% and 94e97% inhibition, respectively. These results are in

agreement with the low IC₅₀values (1.8e8.6 and 33

mM). The

inhibitory activities of 3aec and 10aec were less pronounced with

GIM-1 and NDM-1 resulting in an inhibition of around 30e50%,

although the IC₅₀values against NDM-1 were in the same range as

against VIM-2.

The inhibitors 2a and 2b, showing lowest IC₅₀values, showed

78% and 79% inhibition with VIM-2, respectively. While the percent

inhibition of inhibitor 2a and 2b were 64% and 48% for GIM-1 and

2.3. Evaluation of inhibitors in bacterial cell assays

To investigate if the inhibitors were active against MBLs in

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

163

Table 2

33% and 62% for NDM-1, respectively. The phosphonate esters 2

were found to be more hydrophobic at pH 7.1 (clogD ¼ 3.4e3.8),

which is the pH of the buffer medium, compared to the phosphonic

acids 3 and 10 (clogD ¼ ꢁ0.9e0.1), which probably results in poorer

passive permeability [44] of the former explaining the lower % in-

hibition observed for these compounds. For NDM-1, inhibitor 2b

was the best inhibitor in terms of the lowest IC₅₀and highest effect

in the cell-based screening assay. Modiﬁcation of 2 to alter the

physicochemical properties may by a direction for further inhibitor

improvement.

Additionally, the inhibitors were tested for synergistic activity

with meropenem against MBL-producing clinical strains, and effect

from the inhibitors alone on the bacterial growth. The results are

given in Table S2 in the SI. Disappointingly, synergy testing of the

inhibitors with meropenem using bacterial strains from

P. aeruginosa or K. pneumoniae did not show a lowering of the MIC.

However, promising results were observed when 3b and 10b were

tested against an E. coli strain with VIM-29. In these experiments,

the MIC was lowered from 8 to 32 mg/L with only meropenem to

1 mg/L in presence of inhibitors (Table S2 in SI). Since VIM-2 has

90% sequence identity to VIM-29 it is likely to believe that VIM-2

could also be inhibited in a similar assay. In addition, the tested

X-ray data collection and crystallographic reﬁnement statistics for VIM-2 complexes

of 2b, 10b and 10c. Values in parenthesis are for the highest resolution shell.

VIM-2_2b

VIM-2_10b

VIM-2_10c

PDB entry

X-ray source

Data collection statistics

Space group

5MM9

ID29, ESRF

5NHZ

ID29, ESRF

5NI0

ID29, ESRF

C2

P2₁2₁2₁

Unit cell (Å)

a

b

c

101.11

79.07

67.47

130.3

24.94e1.55

(1.61e1.55)

0.97 625

45.78

90.87

124.08

45.72

91.07

122.92

b

(^ꢀ)

Resolution (Å)

24.75e1.85

(1.92e1.85)

0.97 625

24.77e1.67

(1.73e1.67)

0.97 625

Wavelength (Å)

No. unique reﬂections

Multiplicity

Completeness (%)

Mean (<I>/<s_I>)

R-merge^a

54 949 (5447) 45 000 (4421) 56 573 (5666)

2.0 (1.9)

6.5 (6.7)

6.4 (6.3)

93.58 (93.51)

8.16 (1.92)

0.064 (0.417)

0.995 (0.679)

24.94e1.55

(1.61e1.55)

0.97 625

99.93 (99.98)

14.60 (2.30)

0.073 (0.791)

0.999 (0.785)

24.75e1.85

(1.92e1.85)

0.97 625

94.12 (92.38)

10.80 (1.61)

0.102 (0.925)

0.997 (0.634)

24.77e1.67

(1.73e1.67)

0.97 625

CC

1/2

Resolution (Å)

Wavelength (Å)

No. unique reﬂections

Multiplicity

54 949 (5447) 45 000(4421)

2.0 (1.9)

56 573 (5666)

6.4 (6.3)

compounds were not toxic at high concentrations (>500 mM) to the

6.5 (6.7)

Completeness (%)

Mean (<I>/<s_I>)

R-merge^a

93.58 (93.51)

8.16 (1.92)

0.064 (0.417)

0.995 (0.679)

14.84

99.93 (99.98)

14.60 (2.30)

0.073 (0.791)

0.999 (0.785)

27.24

94.12 (92.38)

10.80 (1.61)

0.102 (0.925)

0.997 (0.634)

20.49

three clinical cell lines included in this study (Table S2 in SI).

2.4. Crystal structure complexes of VIM-2 with 2b, 10b and 10c

CC

1/2

Wilson B-factor (Å²)

Reﬁnement statistics

Resolution (Å)

To better understand the effects of bioisosteric substitution and

to address the question of the coordinating properties of the

phosponic acid group compared to the thiol, we attempted to

obtain complex structures of the new inhibitors. Initial attempts to

soak native crystals were found unsuccessful, whereas our new

protocol for a DMSO free co-crystallization method [45] was suc-

cessful for inhibitors 2b, 10b and 10c. Several more inhibitors were

included in the experiment without providing satisfactory com-

plexes. All the new VIM-2 complex structures have two protein

24.94e1.55

0.1575

0.2006

0.011

25e1.85

0.1662

0.2039

0.013

1.32

97.0

0.47

3.00

25e1.67

0.1778

0.2140

0.011

1.32

97.0

0.95

5.62

R-factor (all reﬂections)

R-free^b

RMSD bond lengths (Å)

RMSD bond angles (^ꢀ)

Ramachandran favored (%) 98.0

Ramachandran outliers (%) 0.46

Clashscore

Average B-factor (Å²)

All atoms

1.31

5.50

20.8

36.5

27.4

molecules in the asymmetric unit, a ab/ba fold with mobile loops

Chain A/B

Solvent

Ligand A/B

Mean occupancy

Zn1 A/B

20.6/20.6

34.0

40.3/32.8

32.6/37.2

47.9

47.5/58.5

25.3/26.18

37.0

63.04/64.24

adjacent to the active site, and there are two zinc ions in the active

site, Zn1 bound to His116, His118 and His196 and Zn2 coordinated

by Asp120, Cys221 and His263. For all structures, the interactions in

the best-deﬁned protein chain will be described, and only dis-

crepancies for other chain will be mentioned.

0.9/0.9

0.75/1.0

1.0/0.9

1.0/1.0

0.6/0.7

0.7/0.8

1.0/1.0

0.7/0.5

0.9/0.9

Zn2 A/B

Ligand A/B

The VIM-2_2b complex was the best-deﬁned complex to 1.55 Å

(Table 2) and this complex was used to investigate the two different

enantiomers. Herein, the (S)-form of the inhibitor 2b did not ﬁt into

the observed polder omit map (Fig. 1A), whereas for the (R)-form

(Fig. 1B), the substituents at the stereogenic center are all deﬁned

within the polder omit map. Therefore, we anticipate that VIM-2 is

selectively binding the (R)-enantiomer of 2b, since the CH₂-SH

group, the CH₂-CH₂chain and the phosphonate ester are deﬁned

inside both the polder omit and the ﬁnal 2Fo-Fc map (Fig. 1B, C).

The three-dimensional arrangement of substituents on the ster-

eogenic center of (R)-2b (Fig. 2) is in agreement with the reported

complex of VIM-2_1b [35] selectively bound to (S)-1b. Note that the

difference in naming of the stereoisomers ((S)-1b versus (R)-2b)

relates to differences in priority according to the Khan-Ingold-

In the VIM-2_2b crystal structure there are hydrogen bonds

from the OAE atom of the P¼O double bond in compound 2b to the

Asn233 sidechain (d(OAE_ligand… ND2_N233) ¼ 2.84 Å) and from the

SH group to both zinc ions (d(S_ligand… Zn1/Zn2) ¼ 2.30/2.30 Å).

Further, the benzene ring of compound 2b is sandwiched between

Arg228 and Tyr67, making T-shaped

p-p stacking interactions with

Tyr67 and cation- interactions with the positively charged Arg288

p

in a parallel manner. The benzene ring is also in a T-shaped stacking

conformation with His263. Additionally, the ethyl groups of the

phosphonate ester make hydrophobic interactions with Phe61,

Tyr67, Trp87, and CA and CB carbon atoms of Asp119 (Fig. 3A). These

two ethyl groups are absent in the corresponding phosphonic acids

Prelog rules (for 1b: CH₂SH

>

CO₂H while for 2b:

3, which display higher IC₅₀values (7.8e33.2

mM) towards VIM-2

P(O)(OH)₂> CH₂SH) not to inversion of the stereocenter. For both

VIM-2_10b and 10c structures, the (R)-conﬁguration ﬁts best in the

polder omit and 2Fo-Fc maps (Fig. 1DeI), but less convincing than

for the highest resolved VIM-2_2b complex (Fig. 1AeC). This in-

dicates that for the structurally related compounds of this study the

three-dimensional arrangement corresponding to the (R)-conﬁgu-

ration probably leads to inhibitors with improved binding and in-

hibition properties.

than the phosphonate esters 2 (0.38e1.7 M) (Table 1), indicating

that the phosphonic acid group is not involved in corresponding

favorable interactions.

Comparison of the VIM-2_2b structure with the previously re-

ported complex of VIM-2_1b [35] (Fig. 2D) showed that the sulfur

atom of thiol 2b is bridging the zinc(II) atoms of VIM-2 resembling

the earlier reported complex of VIM-2_1b and numerous other

m

164

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

Fig. 1. Observed electron density maps for the new inhibitors with polder omit maps depicted at 2.5s (green), 2Fo-Fc maps at 1.0 s (blue) and Fourier difference maps (Fo-Fc) at 4.0

s

(green) and ꢁ4.0

s

(red). The polder omit maps for VIM-2 with the inhibitors 2b, 10b and 10c in (S)-form (left panels), (R)-form (middle panels) and the ﬁnal 2Fo-Fc map (right

panels) are displayed. The zinc ions (orange) and adjacent water (red) are shown, and hydrogen bonds are shown as red dashed lines. (For interpretation of the references to colour

in this ﬁgure legend, the reader is referred to the web version of this article.)

thiol-B1 MBL complex structures [19,23,28,35,46e49]. Also, the

propylphenyl chains of both 1b and 2b occupy the same region in

VIM-2. Still there is a T-shaped stacking for Phe61 and phenyl in our

Tyr67, Trp87, and CA and CB carbon atoms in the Asp119 (Fig. 3A).

The resolution of the VIM-2_10b structure is 1.85 Å, and this

inhibitor was reﬁned in the (R)-form since this ﬁtted best in the

observed electron density maps (Fig. 1D, E). In the ﬁnal electron

density maps the orientation of the thioacetate group is described

adequately but the width of the electron density is indicating

ﬂexibility and there is positive election in the Fo-Fc map. This

additional density could be ascribed to the hydroxyl ion between

Zn1 and Zn2, but this could not be modeled and is thus left out in

new 2b complex, and a face-to-face more parallel p-p stacking for

1b (PDB ID: 2YZ3) [35]. Both the carboxyl group of 1b and the

phosphonate P¼O of 2b are forming hydrogen bonds with the ND2

atom in the side chain of Asn233. The slightly different orientations

of the inhibitors 1b and 2b may be attributed to the hydrophobic

interactions of the two ethyl groups of 2b with residues Phe61,

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

165

Fig. 2. A-C) Stereochemical view of the enantiomeric form of 1b, 2b, 10b and 10c coordinated and bound to VIM-2. D) Overlay of VIM-2_2b (cyan) and previously reported VIM-

2_1b [35] (grey). (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the web version of this article.)

the ﬁnal structure. Further, there is additional unmodeled density

in elongation of the methyl group of the thioacetate of 10b. The

complex of VIM-2_10c is very similar to VIM-2_10b. Still, for the

VIM-2_10c structure, a hydroxyl ion bridging Zn1 and Zn2 in both

chain A and chain B could be modeled, but the methylene chain is

not well deﬁned in the electron density and displays high B-factors

(Table 2, Fig. 1H and I).

2YNW [50]), it becomes clear that the R2 binding sites accommo-

dating the phenyl group of the inhibitors are very different. For

GIM-1, both Trp228 and Tyr64 conﬁne the R2 site making it smaller

and more hydrophobic. Our reported IC₅₀values for compounds 2

and 10 are not signiﬁcantly different for the longest (n ¼ 4)

compared to the shortest (n ¼ 2) inhibitors in a series (Table 1). This

indicates that similar favorable T-shaped stacking from Trp228 and

From the VIM-2_10b and 10c complex structures, it is clear that

the sulfur atom of the thioacetate is binding the zinc atoms

differently when compared to typical thiol complexes e.g. the VIM-

2_2b complex. In both the VIM-2_10b and 10c structures, the thi-

oacetate interacts with Zn2 and the ND2 atom of Asn233 through

the sulfur atom, with His263 and through a water molecule to the

guanidinium group of Arg228 (Fig. 3B and C, respectively). In VIM-

2_10b the S-atom is close to both zinc ions (d(S_ligand… Zn1/

Zn2) ¼ 2.96/3.22 Å) (Fig. 3B), while for VIM-2_10c, the S to zinc

distances are longer (d(S_ligand… Zn1/Zn2) ¼ 3.29/4.02 Å; Fig. 3C). In

both structures, the phosphonic acid is adjacent to Trp87 and

interacting with the side chain of Asn233. These ﬁndings differ

from the reported CphAemercaptophosphonic acid complex, in

which the Zn was coordinated by two oxygen atoms of the phos-

phonic acid and not by the sulfur atom of the thiol [43]. However,

CphA belongs to the B2 MBL subclass containing one zinc ion in the

active site compared to two zinc ions for B1 MBLs, which may

explain the observed differences. In addition, the benzene ring of

compound 10b is more parallel to Tyr67 compared to VIM-2_2b

structures, and both Arg228 and the main chain of Gly232 are

stacking at the opposite side, and His263 in a T-shaped orientation.

Furthermore, the position of residue Phe61 differs signiﬁcantly

between complexes of 2b and 10b/c. In the complexes of 10b and

10c, the residue Phe61 is moved and closed down on the inhibitors

interacting with the phosphonic acid group, whereas in the 2b

structure Phe61 is in a slight more open conformation interacting

with the ethyl groups (Fig. 4A).

cation-p interactions from Arg224 to the phenyl groups of the in-

hibitors are possible for all compounds within a series. Several

conformations are described for Trp228 in the native GIM-1 crystal

structures [50], which all can contribute to the described inhibitor

interactions. Furthermore, residue 233 is a tyrosine in GIM-1 and

the polar interactions observed in VIM-2_2b, 10b and 10c from

Asn233 to the P¼O group cannot be formed. Still the low IC₅₀of for

2a of 0.18 mM and 64% inhibition towards GIM-1 (Table 1) tells that

it is possible to inhibit GIM-1 in a cell based assay.

The new synthetized inhibitors are less efﬁcient towards E. coli

SNO3 cell with NDM-1, but 2a, 2b and 10a-c show low IC₅₀values

with pure enzyme (1.8e6.6 mM; Table 1). As with VIM-2 (and in

contrast to GIM-1), NDM-1 has an Asn233 residue, which can bind

the inhibitors through the P¼O group as for VIM-2. However, hy-

drophobic stabilization of the phenyl rings of the inhibitors is

limited for NDM-1 as the R2 binding site consists of Ala228 and

Lys224, where the lysine residue is shorter than Arg228 (VIM-2)

and Arg224 (GIM-1). Further, Phe63 in NDM-1 is too far away to

reach down to the phenyl groups of the modeled 2b, 10b and 10c

inhibitor complexes (Fig. 4E). In the reported complex structure of

NDM-1 with hydrolyzed ampicillin (PDB ID: 3Q6X; Fig. 4F), the

substrate interacts with Ile64 and Val73 (in the R2 site), and the

“R1” part of the ampicillin with Met60 and Gln119. The NDM-1

binding site is more hydrophobic and also different from both

VIM-2 and GIM-1 with few conserved second shell residues, thus

the design of an inhibitor hitting all three MBL targets is

challenging.

2.5. Modeling of complexes with GIM-1 and NDM-1

3. Conclusion

The sequence identity between the three MBLs included in this

study is 32% (VIM-2 versus NDM-1), 28% (VIM-2 versus GIM-1) or

24% (NDM-1 versus GIM-1) (Fig. S1). Low sequence identity is also

found between the second shell sphere residues that constitute the

active site together with the six conserved zinc binding residues

(116, 118, 196, 120, 221, 263). The impact from these second shell

interacting residues on ligand binding with GIM-1 and NDM-1 was

examined by modeling the inhibitor binding based on the VIM-2

complexes obtained.

Our bioisosteric approach on designing new metallo-

mase inhibitors was found successful with low dose rate mea-

surement for the 2 series of inhibitors (IC₅₀¼ 0.18e2.2 M) towards

all three enzymes, VIM-2, GIM-1 and NDM-1. For the cell-based

assay with a -lactamase-negative E. coli SNO3 cells inducing

b-lacta-

m

b

expression of one MBL, the best inhibitors were the 3 and 10 series.

The three new crystal structures of VIM-2 with 2b, 10b or 10c,

reveal energetic favorable cation-p interactions from Arg228, and

stacking from Tyr67 and Phe61 with the phenyl ring of the in-

hibitors. Asn233 in VIM-2 is also contributing with polar in-

teractions to the P¼O group (2b) and sulphur atom (10b, 10c).

When superimposing the new VIM-2 complexes with inhibitors

2b, 10c and 10b onto the GIM-1 structure (Fig. 4C, D) (PDB ID:

166

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

structure is therefore a valuable starting point for a structure-

guided design of an inhibitor targeting both VIM-2 and NDM-1

simultaneously.

4. Experimental

4.1. Chemistry

All reagents and solvents were purchased from commercial

sources and used as supplied unless otherwise stated. Compounds

1c and 17 were prepared according to the literature [36]. Reactions

were monitored by thin-layer chromatography (TLC) with Merck

pre-coated silica gel plates (60 F₂₅₄). Visualization was accom-

plished with either UV light or by immersion in potassium per-

manganate or phosphomolybdic acid (PMA) followed by light

heating with a heating gun. Puriﬁcation of reactions was carried out

by ﬂash column chromatography using silica gel from Merck (Silica

gel 60, 0.040e0.063 mm). Analytical HPLC was carried out on a

Purity analysis was carried out on Waters Acquity UPLC^®BEH C18

(1.7

m

m, 2.1 ꢂ 100 mm) column on a Waters Acquity I-class UPLC

with Photodiode Array Detector. NMR spectra were obtained on a

400 MHz Bruker Avance III HD equipped with a 5 mm SmartProbe

BB/1H (BB ¼ 19F, 31P-15 N). Data are represented as follows:

chemical shift, multiplicity (s ¼ singlet, d ¼ doublet, t ¼ triplet,

q ¼ quartet, p ¼ pentet, m ¼ multiplet), coupling constant (J, Hz)

and integration. Chemical shifts (d) are reported in ppm relative to

the residual solvent peak (CDCl₃: d_H7.26 and d_C77.16; Methanol-d₄:

d_H3.31 and d_C49.00). ³¹P NMR spectra were recorded using an

insertion NMR tube ﬁlled with PPh₃

(

d

¼ ꢁ5.4 ppm) solution in

C₆D₆as a reference. Positive ion electrospray ionization mass

spectrometry was conducted on a Thermo electron LTQ Orbitrap XL

spectrometer. IR spectra were recorded with Model Varian 7000e

FT-IR spectrometer.

4.1.1. Preparation and spectroscopic data of 2-(Mercaptomethyl)-6-

phenylhexanoic acid (1c)

The title compound 1c was prepared as reported in the litera-

ture [36]. The ﬁnal deacetylation was performed according to the

general procedure (see chapter 4.1.5). The preceding thioacetate

(1.67 mmol, 468 mg, 1 equiv) in methanol (6 mL) was treated with

1 M NaSMe in MeOH (3.4 mmol, 3.4 mL, 2 equiv). Reaction time was

1.5 h. Column chromatography (6% MeOH:CH₂Cl₂) gave the product

1c as a white solid (1.23 mmol, 290 mg, 74%) of high purity (>95% as

determined by HPLC). ¹H NMR (400 MHz, Chloroform-d)

d 10.85 (s,

1H), 7.30e7.16 (m, 4H), 2.79e2.74 (m, 1H), 2.68e2.60 (m, 2H),

1.75e1.66 (m, 4H), 1.53 (t, J ¼ 8.5 Hz, 1H), 1.44 (q, J ¼ 7.8 Hz, 2H). ¹³

C

NMR (101 MHz, Chloroform-d)

d 180.7, 142.2, 128.4, 128.3, 125.8,

125.7, 49.0, 35.6, 31.2, 31.2, 30.9, 26.6, 25.6. HRMS (ESI): Calcd. for

C

₁₃H₁₇O₂S [MþH]^þ237.0954; found 237.0944. The NMR and MS

data were in accordance with the literature [36].

Fig. 3. Interactions found in the VIM-2_2b (A), VIM-2_10b (B) and the VIM-2_10c (C)

complexes. For the two latter (10b and 10c), residue Phe61 is situated above the paper

plane and left out of the ﬁgure for simpliﬁcation.

4.1.2. General procedure for the alkylation of triethyl

phosphonoacetate (6aec)

A solution of ethyl-2-(diethoxyphosphoryl) acetate 5 in DMF

(25 mL/mol) was stirred at 0 ^ꢀC. The base was added and the so-

lution was continued to stir for 20 min before the alkyl halide was

added drop wise. The reaction mixture was allowed to warm to

room temperature and then heated at 60 ^ꢀC for 18 h. After cooling

the reaction mixture was acidiﬁed with 10% citric acid and

extracted with diethyl ether (3 ꢂ 30 mL). The combined organic

phases were washed with H₂O (3 ꢂ 30 mL), brine (1 ꢂ 30 mL) and

dried over Na₂SO₄. The product 6aec was puriﬁed with column

chromatography using 2:1 diethyl ether/pentane as eluent.

Disappointingly, no synergistic effect was seen in P. aeruginosa nor

in K. pneumoniae bacterial strains, but some effects on a clinical

isolate from E. coli with VIM-29 were observed for 3b and 10b

showing reduced MIC from 8e32 mg/L with only meropenem to 1

mg/L for meropenem in combination with inhibitor. A broad in-

hibitor hitting VIM-2, GIM-1 and NDM-1 was not found, but one

promising hit is the thioacetate phosphonic acid 10b, which seems

to target both VIM-2 and NDM-1. The reported VIM-2_10b

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

167

Fig. 4. Ribbon diagrams (left panels) and calculated electrostatic surfaces (right panels) of VIM-2 (A, B), GIM-1 (C, D) and NDM-1 (E, F). In panel F, the structure of ampicillin

(orange) in complex with NDM-1 (PDB ID: 4RLO) is given. Residues discussed or being highly charged are labeled. (For interpretation of the references to colour in this ﬁgure legend,

the reader is referred to the web version of this article.)

4.1.2.1. Ethyl

2-(diethoxyphosphoryl)-4-phenylbutanoate

(6a).

bromo-2-phenylethane (0.028 mol, 5.18 g, 1 equiv) gave 6a

Potassium tert-butoxide (0.028 mol, 3.15 g, 1 equiv), ethyl 2-

(0.018 mol, 5.76 g, 62%) as a pale yellow oil. ¹H NMR (400 MHz,

(diethoxyphosphoryl) acetate (0.028 mol, 6.30 g, 1 equiv) and 1-

Chloroform-d) d 7.28e7.26 (m, 2H), 7.21e7.16 (m, 3H), 4.23e4.20

168

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

(m, 2H), 4.14e4.07 (m, 4H), 2.96 (ddd, J ¼ 22.9, 11.0, 3.6 Hz, 1H),

2.79e2.69 (m, 1H), 2.64e2.56 (m, 1H), 2.35e2.26 (m), 2.18e2.12 (m,

4.1.3.2. Diethyl 1-hydroxy-5-phenylpentane-2-ylphosphonate (7b).

Ethyl 2-(diethoxyphosphoryl)-5-phenylpentanoate 6b (4.7 mmol,

1.60 g, 1 equiv) and LiBH₄(0.012 mol, 255 mg, 2.5 equiv) gave 7b

(4.5 mmol, 1.34 g, 95%) as a colorless oil. ¹H NMR (400 MHz,

1H), 1.30e1.28 (m, 9H). ¹³C NMR (101 MHz, Chloroform-d)

d 169.5

(d, J ¼ 5.0 Hz), 140.9, 128.9 (d, J ¼ 10.7 Hz), 126.7, 63.2 (dd, J ¼ 8.4,

6.7 Hz), 61.9, 46.2, 44.9, 34.9 (d, J ¼ 15.4 Hz), 29.1 (d, J ¼ 4.4 Hz), 16.9

Chloroform-d) d 7.29e7.25 (m, 2H), 7.19e7.16 (m, 3H), 4.15e4.08 (m,

(dd, J ¼ 6.1, 2.3 Hz), 14.7.³¹P NMR (162 MHz, Chloroform-d):

d

23.0.

4H), 3.82e3.77 (m, 2H), 3.46 (s, 1H), 2.62 (t, J ¼ 7.4 Hz, 2H),

HRMS (ESI): Calcd. for

329.1519.

C

₁₆H₂₆O₅P [MþH]^þ329.1512; found

1.81e1.71 (m, 1H), 1.68e1.66 (m, 2H), 1.66e1.50 (m, 1H), 1.31 (m,

6H). ¹³C NMR (101 MHz, Chloroform-d)

d

141.8, 128.3 (d, J ¼ 5.7 Hz),

125.7, 61.8 (dd, J ¼ 13.1, 6.8 Hz), 60.5 (d, J ¼ 4.4 Hz), 39.9, 38.5, 35.6,

29.2 (d, J ¼ 10.4 Hz), 25.1 (d, J ¼ 3.7 Hz), 16.4 (d, J ¼ 5.5 Hz). ³¹P NMR

4.1.2.2. Ethyl 2-(diethoxyphosphoryl)-5-phenylpentanoate (6b).

Potassium tert-butoxide (0.064 mol, 7.18 g, 0.8 equiv), ethyl 2-

(diethoxyphosphoryl) acetate (0.08 mol, 20.01 g, 1 equiv) and 1-

bromo-3-phenylpropane (0.064 mol, 12.71 g, 0.8 equiv) gave 6b

(0.046 mol, 15.89 g, 73%) as a pale yellow oil. ¹H NMR (400 MHz,

(162 MHz, Chloroform-d):

d 33.0. HRMS (ESI) Calcd. for C₁₅H₂₆O₄P

[MþH]^þ301.1563; found 301.1566.

4.1.3.3. Diethyl 1-hydroxy-6-phenylhexane-2-ylphosphonate (7c).

Ethyl 2-(diethoxyphosphoryl)-6-phenylhexanoate 6c (4.21 mmol,

1.5 g, 1 equiv) and LiBH₄(10.53 mmol, 229 mg, 2.5 equiv) gave 7c

(2.34 mmol, 737 mg, 56%) as a colorless oil. ¹H NMR (400 MHz,

Chloroform-d) d 7.21e7.17 (m, 2H), 7.11e7.08 (m, 3H), 4.08e4.02 (m,

4H), 3.78e3.65 (m, 2H), 2.56e2.52 (m, 2H), 1.91e1.87 (m, 1H),

1.63e1.53 (m, 3H), 1.46e1.44 (m, 2H), 1.35e1.34 (m, 1H), 1.25 (t,

Chloroform-d)

d 7.29e7.24 (m, 2H), 7.19e7.14 (m, 3H), 4.28e4.08

(m, 6H), 2.96 (ddd, J ¼ 22.7, 10.9, 3.8 Hz, 1H), 2.63 (t, J ¼ 7.8 Hz, 2H),

2.07e2.02 (m, 1H), 1.92e1.89 (m, 1H), 1.72e1.65 (m, 2H), 1.34e1.26

(m, 9H). ¹³C NMR (101 MHz, Chloroform-d)

d

169.1 (d, J ¼ 4.8 Hz),

141.8 (d, J ¼ 29.8 Hz), 128.3e128.2 (m), 125.8 (d, J ¼ 29.8 Hz), 62.6

(dd, J ¼ 9.1, 6.6 Hz), 61.3, 46.3, 44.9, 35.3, 30.1 (d, J ¼ 14.9 Hz), 26.6

(d, J ¼ 4.8 Hz), 16.3 (dd, J ¼ 5.9, 2.7 Hz), 14.1. ³¹P NMR (162 MHz,

J ¼ 7.1 Hz, 6H). ¹³C NMR (101 MHz, Chloroform-d)

d 142.4,

23.1. HRMS (ESI): Calcd. for C₁₇H₂₇O₅NaP [MþNa]^þ

128.5e128.4 (m), 125.9, 62.1 (dd, J ¼ 13.6, 6.9 Hz), 60.9 (d,

J ¼ 5.7 Hz), 40.0, 38.7, 35.8, 31.4, 27.3 (d, J ¼ 11.0 Hz), 25.3 (d,

Chloroform-d):

365.1488; found 365.1484.

d

J ¼ 3.8 Hz), 16.6. ³¹P NMR (162 MHz, Chloroform-d):

d 33.3. HRMS

(ESI): Calcd. for C₁₆H₂₈O₄P [MþH]^þ315.1720; found 315.1723.

4.1.2.3. Ethyl

2-(diethoxyphosphoryl)-6-phenylhexanoate

(6c).

Potassium tert-butoxide (0.014 mol, 1.55 g, 0.7 equiv), ethyl-2-

(diethoxyphosphoryl) acetate (0.019 mol, 4.25 g, 1 equiv) and 1-

bromo-4-phenylbutane (0.013 mmol, 2.83 g, 0.7 equiv) gave 6c

(0.009 mol, 3.37 g, 73%) as a yellow oil. ¹H NMR (400 MHz, Chlo-

4.1.4. General procedure for thioacetylation (8aec)

Compound 7aec (1 equiv) was dissolved in CH₂Cl₂(5 mL/mmol)

and added triethylamine (1.05 equiv) and a catalytic amount of 4-

dimethylaminopyridine (DMAP). The mixture was stirred for

5 min at RT, before methanesulfonyl chloride (1.05 equiv) was

slowly added. The reaction mixture was stirred for 3e24 h at RT.

The solvent was evaporated, and the remaining crude was

quenched with 100 mL aq. NH₄Cl and extracted with diethyl ether,

dried over Na₂SO₄, ﬁltered, and evaporated. To the crude was added

an excess (7e10 equiv) of 2 M potassium thioacetate in DMF and

stirred overnight at RT. The brown suspension was quenched with

aq. NH₄Cl (100 mL) and the product was extracted with diethyl

ether (3 ꢂ 30 mL). The combined organic layers were washed with

H₂O (3 ꢂ 50 mL), dried over Na₂SO₄and concentrate to yield an

orange oil. The crude was puriﬁed with column chromatography

using 5e15% MeOH:CH₂Cl₂as eluent.

roform-d)

d 7.33e7.28 (m, 2H), 7.22e7.15 (m, 3H), 4.27e4.12 (m,

6H), 2.96 (ddd, J ¼ 22.5, 11.0, 3.8 Hz, 1H), 2.65e2.58 (m, 2H),

1.72e1.68 (m, 1H), 1.68e1.65 (m, 1H), 1.38e1.36 (m, 2H) 1.3e1.25

(m, 11H). ¹³C NMR (101 MHz, Chloroform-d)

d

169.3 (d, J ¼ 4.8 Hz),

142.2, 128.3 (d, J ¼ 8.7 Hz), 125.7, 62.7(t, J ¼ 6.6, 14.6 Hz), 61.3, 46.5,

45.2, 35.5, 30.9, 28.1 (d, J ¼ 15.0 Hz), 26.9 (d, J ¼ 4.9 Hz), 16.4 (dd,

J ¼ 6.1, 3.2 Hz), 14.2. ³¹P NMR (162 MHz, Chloroform-d):

d 23.3.

HRMS (ESI): Calcd. for

357.1822.

C

₁₈H₃₀O₅P [MþH]^þ357.1825; found

4.1.3. General procedure for the reduction of alkylated triethyl

phosphonoacetate (7aec)

LiBH₄was dissolved in THF (2 mL/mmol) at 0 ^ꢀC and slowly

added to the ester 6aec in THF at 0 ^ꢀC. The suspension was stirred

at room temperature for 30 min and irradiated at 80 ^ꢀC for 10 min.

The mixture was cooled to RT and slowly diluted with MeOH and

stirred for 30 min until excess of the LiBH₄was quenched. The

mixture was acidiﬁed with 10% citric acid, saturated with NaCl and

the product was extracted with diethyl ether. The organics were

dried over Na₂SO₄, ﬁltered and concentrated in vacuum. The

product 3aec was puriﬁed with column chromatography with 5%

MeOH:CH₂Cl₂.

4.1.4.1. S-2-(diethoxyphosphoryl)-4-phenylbutyl ethanethioate (8a).

Diethyl 1-hydroxy-4-phenylbutan-2-ylphosphonate 7a (2.8 mmol,

825 mg, 1 equiv), triethylamine (2.9 mmol, 355 mg, 1.05 equiv),

methanesulfonyl chloride (2.9 mmol, 355 mg, 1.05 equiv), DMAP

(catalytic amount) and potassium thioacetate (20.3 mmol, 2.31 g, 7

equiv) gave 8a (1.5 mmol, 520 mg, 54%) as an orange oil. ¹H NMR

(400 MHz, Chloroform-d)

d 7.33e7.30 (m, 2H), 7.25e7.22 (m, 3H),

4.23e4.13 (m, 4H), 3.44e3.39 (m, 1H), 3.08e3.05 (m, 1H),

2.85e2.81 (m, 2H), 2.37 (s, 3H), 2.10e2.05 (m, 2H), 1.93e1.91 (m,

1H), 1.41e1.36 (m, 6H). ¹³C NMR (101 MHz, Chloroform-d)

d 195.3,

4.1.3.1. Diethyl 1-hydroxy-4-phenylbutan-2-ylphosphonate (7a).

Ethyl 2-(diethoxyphosphoryl)-4-phenylbutanoate 6a (1.6 mmol,

0.53 g, 1 equiv) and LiBH₄(4.8 mmol, 0.12 g, 3 equiv) gave 7a

(1.0 mmol, 0.28 g, 60%) as a colorless oil. ¹H NMR (400 MHz,

141.4, 128.5 (d, J ¼ 12.7 Hz), 126.0, 61.9 (dd, J ¼ 6.8, 2.5 Hz), 36.4,

35.1, 33.5 (d, J ¼ 7.5 Hz), 30.6, 29.6 (d, J ¼ 3.3 Hz), 28.3 (d, J ¼ 2.0 Hz),

16.5 (d, J ¼ 5.9 Hz). ³¹P NMR (162 MHz, Chloroform-d):

d 30.3.

HRMS (ESI): Calcd. for C₁₆H₂₆O₄PS [MþH]^þ345.1285; found

Chloroform-d)

d

7.31e7.26 (m, 2H), 7.21e7.18 (m, 3H), 4.14e4.07 (m,

345.1289.

4H), 3.89e3.79 (m, 2H), 3.15e3.12 (m, 1H), 2.81 (ddd, J ¼ 14.5, 9.2,

5.7 Hz, 1H), 2.69 (ddd, J ¼ 13.8, 9.4, 6.6 Hz, 1H), 2.01e1.98 (m, 2H),

1.97e1.83 (m, 1H), 1.35e1.29 (m, 6H). ¹³C NMR (101 MHz, Chloro-

4.1.4.2. S-2-(diethoxyphosphoryl)-5-phenylpentyl

(8b). Diethyl

ethanethioate

1-hydroxy-5-phenylpentane-2-ylphosphonate

7b

form-d)

d

141.1,128.4 (d, J ¼ 5.7 Hz),126.1, 61.9 (t, J ¼ 6.7 Hz), 60.7 (d,

(0.93 mmol, 289 mg, 1 equiv), triethylamine (0.98 mmol, 111 mg,

1.05 equiv), methanesulfonyl chloride (0.98 mmol, 111 mg, 1.05

equiv), DMAP (catalytic amount) and potassium thioacetate

(9.3 mmol, 1.06 g, 10 equiv) gave 8b (0.32 mmol, 114 mg, 34%) as an

J ¼ 5.5 Hz), 39.1, 37.7, 33.5 (d, J ¼ 11.1 Hz), 27.0 (d, J ¼ 3.5 Hz), 16.5 (t,

J ¼ 5.5 Hz). ³¹P NMR (162 MHz, Chloroform-d):

d 33.3. HRMS (ESI)

Calcd. for C₁₄H₂₄O₄P [MþH]^þ287.1407; found 287.1408.

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

169

orange oil. ¹H NMR (400 MHz, Chloroform-d)

d

7.22e7.18 (m, 2H),

(dd, J ¼ 7.3, 3.1 Hz), 40.4, 38.9, 35.9, 28.9 (d, J ¼ 8.2 Hz), 26.7 (d,

J ¼ 3.6 Hz), 23.7 (d, J ¼ 2.7 Hz), 16.6 (d, J ¼ 6.2 Hz). ³¹P NMR

7.12e7.08 (m, 3H), 4.07e4.00 (m, 4H), 3.26 (ddd, J ¼ 15.2, 13.9,

5.1 Hz, 1H), 2.92 (ddd, J ¼ 13.9, 12.0, 8.2 Hz, 1H), 2.54 (t, J ¼ 7.3 Hz,

2H), 2.25 (s, 3H),1.99e1.86 (m,1H),1.76e1.70 (m, 3H),1.63e1.43 (m,

(162 MHz, Chloroform-d) d 31.9. HRMS (ESI): Calcd. for C₁₅H₂₆O₃PS

[MþH]^þ317.1335; found 317.1337.

1H), 1.27e1.23 (m, 6H). ¹³C NMR (101 MHz, Chloroform-d)

d 195.1,

141.8, 128.3 (dd, J ¼ 16.7, 1.9 Hz), 125.7, 61.8 (t, J ¼ 6.9 Hz), 37.0,

35.7e35.4 (m), 30.4 (d, J ¼ 2.5 Hz), 28.9 (dd, J ¼ 7.3, 2.3 Hz), 28.3 (d,

J ¼ 1.9 Hz), 27.4 (d, J ¼ 3.7 Hz), 16.4 (d, J ¼ 5.8 Hz). ³¹P NMR

4.1.5.3. Diethyl 1-mercapto-6-phenylhexan-2-ylphosphonate (2c).

Thioacetate 8c (1.07 mmol, 400 mg, 1 equiv) in methanol and 1 M

NaSMe in MeOH (1.1 mmol, 1.1 mL, 1 equiv) gave 2c (0.82 mmol,

270 mg, 77%) as a colorless oil after puriﬁcation with column

chromatography (5e20% MeOH: CH₂Cl₂). ¹H NMR (400 MHz,

(162 MHz, Chloroform-d): d 30.5. HRMS (ESI): Calcd. for C₁₇H₂₈O₄PS

[MþH]^þ359.1440; found 359.1449.

Chloroform-d)

d

7.28e7.23 (m, 2H), 7.17e7.15 (d, J ¼ 6.8 Hz, 3H),

4.1.4.3. S-2-(diethoxyphosphoryl)-6-phenylhexyl ethanethioate (8c).

Diethyl 1-hydroxy-6-phenylhexane-2-ylphosphonate 7c (13 mmol,

4.20 g, 1 equiv), triethylamine (0.014 mmol, 1.42 g, 1.05 equiv),

methanesulfonyl chloride (1.34 mmol, 152 mg, 1.05 equiv)), DMAP

(catalytic amount) and potassium thioacetate (0.14 mmol, 15.98 g,

10 equiv) gave 8c (9.4 mmol, 3.502 g, 71%) as an orange oil. ¹H NMR

4.12e4.05 (m, 4H), 2.94e2.83 (m, 1H), 2.70e2.60 (m, 3H), 1.98e1.80

(m, 1H),1.76e1.66 (m, 5H), 1.60e1.43 (m, 2H),1.30 (t, J ¼ 7.0 Hz, 6H).

¹³C NMR (101 MHz, Chloroform-d)

d

142.5, 128.5 (d, J ¼ 11.5 Hz),

125.8, 61.8 (d, J ¼ 6.9 Hz), 40.3, 38.9, 35.7 (d, J ¼ 1.9 Hz), 31.4,

26.9e26.8 (m), 23.6 (d, J ¼ 2.3 Hz), 16.6 (d, J ¼ 5.9 Hz). ³¹P NMR

(162 MHz, Chloroform-d) d 31.5. HRMS (ESI): Calcd. for C₁₆H₂₈O₃PS

(400 MHz, Chloroform-d)

d

7.34e7.30 (m, 2H), 7.24e7.21 (m, 3H),

[MþH]^þ331.1486; found 331.1491.

4.18e4.15 (m, 4H), 3.38 (ddd, J ¼ 15.2, 13.8, 5.0 Hz, 1H), 3.03 (ddd,

J ¼ 13.9, 11.9, 8.2 Hz, 1H), 2.67 (t, J ¼ 7.6 Hz, 2H), 2.38 (s, 3H),

2.06e1.97 (m, 1H), 1.83e1.77 (m, 1H), 1.69e1.62 (m, 3H), 1.62e1.56

(m, 2H), 1.40e1.36 (m, 6H). ¹³C NMR (101 MHz, Chloroform-d)

4.1.5.4. 5-Phenyl-2-(1H-tetrazol-5-yl)pentane-1-thiol

(4).

Thioacetate 15b (0.1 mmol, 30 mg,1 equiv) in methanol was treated

with 1 M NaSMe in MeOH (0.2 mmol, 0.2 mL, 2 equiv). The reaction

time was 1.5 h at RT. Column chromatography (3% MeOH: CH₂Cl₂)

provided 4 as white solid (0.073 mmol, 18.1 mg, 73%) with high

purity (>95% as determined by HPLC). ¹H NMR (400 MHz, Meth-

d

195.4, 142.5, 128.4 (d, J ¼ 13.2 Hz), 125.7, 61.9 (dd, J ¼ 6.9, 5.0 Hz),

37.2, 35.7 (d, J ¼ 8.9 Hz), 31.5, 30.6, 28.5 (d, J ¼ 1.8 Hz), 27.9 (d,

J ¼ 3.5 Hz), 27.1 (d, J ¼ 7.4 Hz), 16.5 (d, J ¼ 6.0 Hz). ³¹P NMR

(162 MHz, Chloroform-d):

d

31.1. HRMS (ESI): calcd. For C₁₈H₂₆O₄PS

anol-d₄) d 7.19e7.04 (m, 5H), 3.45e3.43 (m, 1H), 3.04e2.95 (m, 2H),

[MþH]^þ373.1602; found 373.1289.

2.56 (q, J ¼ 7.6 Hz, 2H), 1.82 (ddd, J ¼ 35.0, 8.0, 4.1 Hz, 2H), 1.75e1.72

(m, 2H), 1.49e1.36 (m, 2H). ¹³C NMR (101 MHz, Methanol-d₄)

4.1.5. Deacetylation of thioacetates (preparation of 1c, 2aec and 4)

Except for 2a, the following general procedure was applied. To a

stirred solution of thioacetate (8 or 15) (1 equiv) in methanol (6 mL)

at ꢁ20 ^ꢀC was added sodiumthiomethoxide, NaSMe (1 M solution

in MeOH, 1 equiv). The reaction mixture was stirred for 30 min

at ꢁ20 ^ꢀC, before aqu. HCl (0.1 M) was added. The reaction mixture

was extracted with DCM (3 ꢂ 30 mL), the combined organic layers

were washed with brine, dried over MgSO₄, ﬁltered and concen-

trate. The resulting crude was puriﬁed by column chromatography.

d 158.3, 141.5, 127.9, 125.5, 41.1, 35.3, 34.8, 32.7, 28.5. HRMS (ESI):

Calcd. for C₁₂H₁₇N₄S [MþH]^þ249.1170; found 249.1168.

4.1.6. General procedure for the hydrolysis of phosphonic esters

(preparation of 3, 9 and 10)

Bromotrimethylsilane (3 equiv) was added drop wise to 2, 7 or 8

(1 equiv) in 10 mL dry CH₂Cl₂under argon. The mixture was stirred

overnight at room temperature. The solvent and excess TMSBr were

removed under reduced pressure, MeOH (0.2 mL/mmol of sub-

strate) was added and the reaction mixture was stirred for 1 h at RT.

The reaction mixture was evaporated to dryness. The crude 3, 9 or

10 was obtained as viscous oil. Crude product was dissolved in

minimal volume of EtOAc and cold pentane was added resulting in

the separation of compound as oily or crystalline material. The

pentane layer was discarded and the remaining compound was

dried under vacuum.

4.1.5.1. Diethyl 1-mercapto-4-phenylbutan-2-ylphosphonate (2a).

According to the reported procedure [38], a 0.1 M NaOMe/MeOH

solution (3 mmol, 3 mL, 2 equiv) was added to thioacetate 8a

(1.5 mmol, 0.52 g, 1 equiv) at 0 ^ꢀC. The resulting mixture was stirred

for 15 min at rt under H₂and concentrated. Aq. NH₄Cl (30 mL) and

0.1 M HCl (10 mL) were added to the crude and the product was

extracted with diethyl ether. The combined organics were dried

and concentrated. Repeated column chromatography (5e20%

MeOH: CH₂Cl₂and 3:1 EtOAc:Et₂O) gave 2a (0.46 mmol, 139 mg,

30%) as a slightly orange oil. ¹H NMR (400 MHz, Methanol-d₄)

4.1.6.1. 1-Hydroxy-4-phenylbutan-2-ylphosphonic

acid

(9a).

Diethyl 1-hydroxy-4-phenylbutyl-2-ylphosphonate 7a (0.31 mmol,

95 mg, 1 equiv) and TMSBr (140 mg) gave crude 9a (77 mg) as a

colorless oil. A small batch (30 mg) was dissolved in 2e3 drops of

EtOAc and washed with cold pentane, afforded the pure product 9a

d

7.33e7.29 (m, 2H), 7.26e7.20 (m, 3H), 4.18e4.10 (m, 4H),

3.04e2.92 (m, 1H), 2.86e2.72 (m, 4H), 2.15e2.01 (m, 3H), 1.71 (t,

J ¼ 8.3 Hz, 1H), 1.37e1.33 (m, 6H). ¹³C NMR (101 MHz, Methanol-d₄)

(0.086 mmol, 20 mg, 76%). ¹H NMR (400 MHz, Chloroform-d)

d 8.65

d

142.9, 129.7, 127.3, 63.7 (dd, J ¼ 7.1, 1.9 Hz), 40.5, 39.1, 34.4 (d,

(s, 2H), 7.15e7.08 (m, 5H), 3.84e3.80 (m, 2H), 2.69e2.58 (m, 2H),

J ¼ 8.2 Hz), 30.2 (d, J ¼ 3.2 Hz), 23.9 (d, J ¼ 2.6 Hz), 16.9 (d,

2.02e1.80 (m, 2H), 1.78e1.73 (m, 1H). ¹³C NMR (101 MHz, Meth-

J ¼ 5.8 Hz). ³¹P NMR (162 MHz, Choroform-d)

d

32.9. HRMS (ESI):

anol-d₄)

d

142.0, 128.0 (d, J ¼ 13.0 Hz), 125.5, 60.3, 40.5, 39.2, 33.6

Calcd. for C₁₆H₂₈O₃PS [MþH]^þ331.1491; found 331.1494.

(d, J ¼ 8.3 Hz), 28.1 (d, J ¼ 3.2 Hz), 22.8. ³¹P NMR (162 MHz,

Methanol-d₄)

231.0781; found: 231.0779.

d

26.1. HRMS (ESI): Calcd. for C₁₀H₁₆O₄P [MþH]^þ

4.1.5.2. Diethyl 1-mercapto-5-phenylpentan-2-ylphosphonate (2b).

Thioacetate 8b (0.56 mmol, 0.20 g, 1 equiv) in methanol and 1 M

NaSMe in MeOH (0.6 mmol, 0.6 mL, 1 equiv) gave 2b (0.41 mmol,

130 mg, 74%) as white oil after puriﬁcation with column chroma-

tography (5e20% MeOH: CH₂Cl₂). ¹H NMR (400 MHz, Chloroform-

4.1.6.2. 1-Hydroxy-5-phenylpentane-2-ylphosphonic

acid

(9b).

Diethyl 1-hydroxy-5-phenylpentan-2-ylphosphonate 7b (0.36 mmol,

0.1 g, 1 equiv) and TMSBr (1.06 mmol, 0.16 g, 3 equiv) gave crude 9b

(110 mg) as a colourless oil. A small batch (64 mg) was dissolved in

2e3 drops of EtOAc and washed with cold pentane, afforded the

pure product 9b (0.16 mmol, 39 mg, 61%). ¹H NMR (400 MHz,

d)

d 7.29e7.25 (m, 2H), 7.19e7.15 (m, 3H), 4.12e4.04 (m, 4H),

2.96e2.85 (m, 1H), 2.69e2.62 (m, 3H), 1.98e1.93 (m, 1H), 1.80e1.76

(m, 4H), 1.65 (t, J ¼ 8.4 Hz, 1H), 1.32e1.28 (m, 6H). ¹³C NMR

(101 MHz, Chloroform-d)

d

142.0, 128.5 (d, J ¼ 8.8 Hz), 125.9, 61.9

Methanol-d₄)

d

7.29e7.15 (m, 5H), 3.92 (ddd, J ¼ 14.3, 11.0, 4.9 Hz,

170

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

1H), 3.73e3.64 (m, 1H), 2.68e2.63 (m, 2H), 1.98e1.69 (m, 5H). ¹³

C

4.1.6.7. 1-Sulfanyl-4-phenylbutan-2-ylphosphonic

acid

(3a).

NMR (101 MHz, Methanol-d₄)

d

142.2, 127.9 (d, J ¼ 15.9 Hz), 125.3,

Diethyl 1-sulfanyl-4-phenylbutan-2-ylphosphonate 2a (0.14 mmol,

43 mg, 1 equiv) and TMSBr (0.43 mmol, 65 mg, 3 equiv) gave crude

3a (38 mg) as an orange oil. The crude was dissolved in 2e3 drops

of EtOAc and washed with cold pentane, afforded the pure product

3a (0.13 mmol, 33 mg, 96%).¹H NMR (400 MHz, Chloroform-d)

60.4, 40.9, 39.5, 35.7, 29.6 (d, J ¼ 8.3 Hz), 25.8 (d, J ¼ 3.4 Hz). ³¹P

NMR (162 MHz, Methanol-d₄)

d 31.1. HRMS (ESI): Calcd. for

C

₁₁H₁₈O₄P [MþH]^þ245.0937; found 245.0936.

d

9.05 (s, 2H), 7.28e7.19 (m, 5H), 3.05e2.94 (m, 1H), 2.78e2.74 (m,

4.1.6.3. 1-Hydroxy-5-phenylpentane-2-ylphosphonic

acid

(9c).

3H), 2.14e2.07 (m, 3H), 1.56 (t, J ¼ 8.4 Hz, 1H). ¹³C NMR (101 MHz,

Diethyl 1-hydroxy-5-phenylpentan-2-ylphosphonate 7c (0.36 mmol,

100 mg, 1 equiv) and TMSBr (0.95 mmol, 150 mg, 3 equiv) gave

crude 9c (110 mg) as colorless oil. A small batch (38 mg) was dis-

solved in 2e3 drops of EtOAc and washed with cold pentane,

afforded the pure product 9c (21 mg, 0.081 mmol, 55%). ¹H NMR

Methanol-d₄)

d

141.8, 128.1 (d, J ¼ 12.3 Hz), 125.6, 40.4, 39.0, 33.1 (d,

J ¼ 7.5 Hz), 28.9, 22.9. ³¹P NMR (162 MHz, Methanol-d₄)

d 30.43.

HRMS (ESI): Calcd. for C₁₀H₁₆O₃PS [MþH]^þ247.0552; found

247.0552.

(400 MHz, Chloroform-d)

d 8.22 (s, 2H), 7.27e7.20 (m, 2H),

4.1.6.8. 1-Sulfanyl-5-phenylpentan-2-ylphosphonic

acid

(3b).

7.15e7.10 (m, 3H), 3.91e3.74 (m, 2H), 2.57e2.53 (m, 2H), 1.94e1.73

(m, 1H), 1.69e1.56 (m, 1H), 1.55e1.38 (m, 5H). ¹³C NMR (101 MHz,

Diethyl 1-sulfanyl-5-phenylpentan-2-ylphosphonate 2b (0.05 mmol,

15 mg, 1 equiv) and TMSBr (0.14 mmol, 22 mg, 3 equiv) gave crude

3b as an orange oil. The crude was dissolved in 2e3 drops of EtOAc

and washed with cold pentane, afforded the pure product 3b

(0.04 mmol, 11 mg, 90%). ¹H NMR (400 MHz, Methanol-d₄)

Methanol-d₄)

d

142.4, 127.9 (d, J ¼ 13.5 Hz), 125.2, 60.3, 40.9, 39.5,

35.3, 31.5, 27.2 (d, J ¼ 8.5 Hz), 25.7 (d, J ¼ 3.4 Hz). ³¹P NMR

(162 MHz, Methanol-d₄) d 32.0. HRMS (ESI): Calcd. for C₁₂H₁₉O₄PNa

[MþH]^þ281.0916; found 281.0913.

d

7.27e7.18 (m, 4H), 7.16e7.12 (m, 1H), 2.97 (ddd, J ¼ 17.8, 13.7,

3.8 Hz, 1H), 2.66e2.60 (m, 3H), 1.91e1.78 (m, 5H). ¹³C NMR

(101 MHz, Methanol-d₄)

142.9, 128.8 (d, J ¼ 16.9 Hz), 126.1 (d,

J ¼ 3.3 Hz), 41.8, 40.4, 36.3, 29.9(d, J ¼ 7.4 Hz), 27.5 (d, J ¼ 2.5 Hz),

4.1.6.4. 1-Acetylthio-4-phenylbutan-2-ylphosphonic acid, (10a).

S-2-(diethoxyphosphoryl)-4-phenylbutyl

(0.77 mmol, 268 mg, 1 equiv) and TMSBr (2.31 mmol, 353 mg, 3

equiv) gave crude 10a (232 mg) as an orange oil. The crude was

dissolved in 2e3 drops of EtOAc and washed with cold pentane,

afforded the pure product 10a (0.69 mmol, 201 mg, 91%). ¹H NMR

d

ethanethioate

8a

23.8. ³¹P NMR (162 MHz, Methanol-d₄)

d 30.12. HRMS (ESI): Calcd.

for C₁₁H₁₈O₃PS [MþH]^þ261.0709; found 261.0712.

4.1.6.9. 1-Sulfanyl-6-phenylhexan-2-ylphosphonic

acid,

(3c).

(400 MHz, Methanol-d₄)

d 7.26e7.22 (m, 2H), 7.19e7.14 (m, 3H),

Diethyl 1-sufanyl-6-phenylhexan-2-ylphosphonate 2c (0.30 mmol,

100 mg, 1 equiv) and TMSBr (0.91 mmol, 144 mg, 3 equiv) gave

crude 3c as an orange oil. The crude was dissolved in 2e3 drops of

EtOAc and washed with cold pentane, afforded the pure product 3c

3.44e3.36 (m, 1H), 3.04 (ddd, J ¼ 13.8, 10.4, 8.4 Hz, 1H), 2.79e2.76

(m, 2H), 2.30 (s, 3H), 2.07e1.81 (m, 3H). ¹³C NMR (101 MHz,

Methanol-d₄)

d

195.5, 141.8 (d, J ¼ 17.5 Hz), 128.1e128.0 (m), 125.5,

40.4, 39.0, 37.2, 35.8, 33.1 (t, J ¼ 8.0 Hz), 29.7 (d, J ¼ 2.9 Hz), 28.9 (dd,

(0.29 mmol, 81 mg, 98%). ¹H NMR (400 MHz, Chloroform-d)

d 8.59

J ¼ 10.7, 2.5 Hz), 28.0, 22.9. ³¹P NMR (162 MHz, Methanol-d₄)

d 30.4.

(s, 2H), 7.29e7.25 (m, 2H), 7.19e7.16 (m, 3H), 2.98 (ddd, J ¼ 20.0,

13.9, 4.6 Hz, 1H), 2.75e2.63 (m, 1H), 2.62 (dd, J ¼ 8.8, 6.4 Hz, 2H),

2.11e2.02 (m, 1H), 1.83 (ddd, J ¼ 16.5, 9.4, 6.2 Hz, 2H), 1.67e1.63 (m,

2H), 1.49e1.47 (m, 2H), 1.34 (t, J ¼ 7.1 Hz, 1H). ¹³C NMR (101 MHz,

HRMS (ESI): Calcd. for C₁₂H₁₈O₄PS [MþH]^þ289.0658; found

289.0660.

4.1.6.5. 1-(Acetylthio)-5-phenylpentan-2-ylphosphonic acid, (10b).

S-2-(diethoxyphosphoryl)-5-phenylpentan

(0.53 mmol, 190 mg, 1 equiv) and TMSBr (1.58 mmol, 242 mg, 3

equiv) gave crude 10b (207 g) as an orange oil. The crude was

dissolved in 2e3 drops of EtOAc and washed with cold pentane,

afforded the pure product 10b (0.37 mmol, 113 mg, 71%). ¹H NMR

Methanol-d₄)

d

142.5 (d, J ¼ 3.2 Hz), 128.1e127.9 (m), 125.2, 41.3 (d,

ethanethioate

8b

J ¼ 134.8 Hz), 35.3, 31.5e31.4 (m), 26.9e26.9 (m), 23.3. ³¹P NMR

(162 MHz, Methanol-d₄) d 30.5. HRMS (ESI): Calcd. for

C₁₂H₁₉O₃NaPS [MþH]^þ297.0690; found 297.0685.

4.1.7. General procedure for the synthesis of acrylonitriles (13a and

13b)

(400 MHz, Methanol-d₄)

3.00 (ddd, J ¼ 14.0, 10.9, 8.5 Hz, 1H), 2.58 (t, J ¼ 7.3 Hz, 2H), 2.27 (s,

3H), 1.94e1.56 (m, 6H). ¹³C NMR (101 MHz, Methanol-d₄)

195.4,

d 7.25e7.13 (m, 5H), 3.39e3.30 (m, 1H),

Compound 13a and 13b were prepared by a modiﬁed procedure

based on the work of Baraldi et al. [39] K₂CO₃(3 equiv) and 4-

cyano-3-oxotetrahydrothiophene 12 (1 equiv) were stirred in dry

acetone (20 mL). To this suspension alkyl halide (1 equiv) was

added and the mixture was reﬂuxed for 5 h. The residue was dis-

solved in ether and stirred vigorously with 5% NaOH aq. (10 mL).

The solution was kept at RT until complete fragmentation of the

intermediate adduct was observed by TLC. The organic phase was

separated and dried with sodium sulfate. The crude oil was further

puriﬁed by column chromatography using 20% diethyl ether in

pentane as eluent.

d

141.9, 128.0 (d, J ¼ 18.7 Hz), 125.4, 37.8, 36.4, 35.4, 29.0 (d,

J ¼ 7.4 Hz), 28.0, 27.5 (d, J ¼ 3.1 Hz). ³¹P NMR (162 MHz, Methanol-

d₄)

d

30.8. HRMS (ESI): Calcd. for C₁₃H₂₀O₄PS [MþH]^þ303.0814;

found 303.0818.

4.1.6.6. 1-(Acetylthio)-6-phenylhexan-2-ylphosphonic acid, (10c).

S-2-(diethoxyphosphoryl)-6-phenylhexyl

(0.478 mmol, 0.178 g, 1 equiv) and TMSBr (1.434 mmol, 0.219 g, 3

equiv) gave crude 10c (113 mg) as an orange oil. The crude was

dissolved in 2e3 drops of EtOAc and washed with cold pentane,

afforded the pure product 10c (0.31 mmol, 98 mg, 65%). ¹H NMR

ethanethioate,

8c

4.1.7.1. 2-Benzylacrylonitrile (13a). K₂CO₃(11.7 mmol, 1.62 g, 3

equiv), compound 4-cyano-3-oxotetrahydrothiophene (3.9 mmol,

0.5 g, 1 equiv), and benzyl bromide (3.9 mmol, 0.673 g, 1 equiv),

gave 13a as an oil (1.99 mmol, 284 mg, 51%). ¹H NMR (400 MHz,

(400 MHz, Methanol-d₄)

d 7.26e7.20 (m, 3H), 7.19e7.12 (m, 2H),

3.39e3.33 (m, 1H), 3.04 (ddd, J ¼ 13.8, 10.8, 8.6 Hz, 1H), 2.65e2.58

(m, 3H), 2.30 (s, 3H), 1.96e1.89 (m, 1H), 1.82e1.72 (m, 4H),

1.64e1.59 (m, 1H). ¹³C NMR (101 MHz, Methanol-d₄)

d

195.5, 141.9,

Chloroform-d)

d

7.40 (t, J ¼ 8.0 Hz, 2H), 7.33 (d, J ¼ 7.3 Hz, 1H), 7.26

128.1e127.8 (m), 125.3, 40.9, 39.6, 37.8, 36.4, 35.5 (d, J ¼ 11.7 Hz),

(d, J ¼ 7.4 Hz, 2H), 5.95 (s, 1H), 5.73 (s, 1H), 3.59 (s, 2H). ¹³C NMR

29.2e28.8 (m), 28.1, 27.5 (d, J ¼ 3.1 Hz), 26.7 (d, J ¼ 3.0 Hz), 23.0. ³¹

P

(101 MHz, Chloroform-d) d 135.6, 131.1, 129.2, 128.9, 128.9, 128.6,

NMR (162 MHz, Methanol-d4)

d

31.5. HRMS (ESI): Calcd. for

128.6, 128.4, 127.4, 122.7, 118.5, 40.8. HRMS (ESI): Calcd. for C₁₀H₁₀N

C

₁₄H₂₀O₄PS[MþH]^þ315.0820; found 315.0817.

[MþH]^þ144.0808; found 144.0808.

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

171

4.1.7.2. 2-Methylene-5-phenylpentanenitrile

(47.1 mmol, 6.51 g, 3 equiv), 4-cyano-3-oxotetrahydrothiophene

(15.7 mmol, 2.00 g, equiv), 1-bromo-3-phenylpropane

(15.7 mmol, 3.13 g, 1 equiv), gave 13b as an oil (3.29 mmol,

564 mg, 21%). ¹H NMR (400 MHz, Chloroform-d)

7.34 (dd, J ¼ 8.1,

(13b). K₂CO₃

(12.1 mmol, 0.925 g, 26 equiv) gave 15b (0.456 mmol, 132 mg, 98%)

as a colorless oil. ¹H NMR (400 MHz, Methanol-d₄)

7.22e7.11 (m,

d

1

5H), 4.93 (s, 1H), 3.38e3.30 (m, 2H), 3.16 (dd, J ¼ 15.2, 10.3 Hz, 1H),

2.58 (t, J ¼ 7.5 Hz, 2H), 2.24 (s, 3H), 1.91e1.77 (m, 2H), 1.61e1.45 (m,

d

2H). ¹³C NMR (101 MHz, Methanol-d₄)

d 196.2, 159.6, 142.9, 129.4,

6.7 Hz, 2H), 7.24e7.19 (m, 3H), 5.87 (s, 1H), 5.71 (d, J ¼ 1.6 Hz, 1H),

126.9, 37.2, 36.2, 33.9, 33.3, 30.4, 29.9. HRMS (ESI): Calcd. for

2.68 (t, J ¼ 7.6 Hz, 2H), 2.29e2.27 (m, 2H), 1.96e1.90 (m, 2H),. ¹³

C

₁₄H₁₉ON₄S [MþH]^þ291.1284; found 291.1274.

NMR (101 MHz, Chloroform-d)

d 141.1, 130.5, 128.5, 128.5, 128.4,

126.1, 122.9, 118.6, 34.7, 33.9, 29.1. HRMS (ESI): Calcd. for C₁₂H₁₄N₄

4.1.9.3. ⁰5-(1-(cyclohexylthio)-5-phenylpentan-2-yl)-1H-tetrazole

(16). 14a (0.23 mmol, 0.050 g, 1 equiv), and cyclohexyl thiol

(0.933 mmol, 0.108 g, 4 equiv) gave 16 (0.216 mmol, 71.50 mg, 93%).

[MþH]^þ172.1114; found 172.1121.

4.1.8. General procedure for the preparation of tetrazoles (14a and

14b)

¹H NMR (400 MHz, Methanol-d₄)

d

7.23e7.09 (m, 5H), 2.95e2.82

(m, 1H), 2.58e2.55 (m, 2H), 1.86e1.68 (m, 6H), 1.59e1.18 (m, 10H).

¹³C NMR (101 MHz, Methanol-d₄)

158.94, 141.6, 127.9, 127.9, 125.5,

Dibutyltin oxide (0.2 equiv) and trimethylsilyl azide (2 equiv)

were added to a solution of 13 (1 equiv) in anhydrous 1,4-dioxane

(2 mL/mmol). The reaction mixture was subjected to microwave

irradiation in a tightly sealed vessel for 50 min at 150 ^ꢀC, then

cooled to room temperature. The solvent was removed under

reduced pressure. The residue was dissolved in diethyl ether (10 mL

and extracted with 2 M aq. NaOH (3 ꢂ 10 mL). The aqueous layer

was acidiﬁed with 4 M aq. HCl to pH 1 and extracted with ethyl

acetate (4 ꢂ 10 mL). The organic extract was washed with brine

(10 mL), dried over MgSO4, and evaporated under reduced pressure

to give the target tetrazoles, which were recrystallized from EtOAc.

d

43.6, 36.7, 34.9, 33.6, 33.4, 33.3, 32.8, 28.6, 25.6, 25.5. HRMS (ESI):

Calcd. for C₁₈H₂₇N₄S [MþH]^þ331.1950; found 331.1951.

4.2. Biological activity

4.2.1. Gene constructs of VIM-2, NDM-1 and GIM-1

In this study two types of gene constructs were used. The ﬁrst

included the native leader sequence to allow the proteins to be

transported to the periplasm, for the three enzymes VIM-2 from

Pseudomonas aeruginosa strain 301-5473 (GenBank no Q9K2N0

[51]), GIM-1 from P. aeruginosa (GenBank no Q704V1 [50,52]) and

NDM-1 (GenBank no. E9NWK5, e.g. refs. [53,54]), where the latter

bla_NDM-1gene is reported from several organisms. Cloning of

bla_NDM-1or bla_GIM-1genes into the Escherichia coli pET26b(þ) vector

(Novagen) was performed using the Primers listed in Table S3, and

by restriction cutting as described for VIM-26 [55]. Cloning of bla-

4.1.8.1. 5-(3-Phenylprop-1-en-2-yl)-1H-tetrazole

(14a).

Dibutyltin oxide (0.07 mmol, 0.017 g, 0.2 equiv), trimethylsilyl azide

(1.4 mmol, 0.161 g, 2 equiv) and 13a (0.100 g, 0.7 mmol, 1 equiv)

gave 14a (0.55 mmol, 102 mg, 78%). ¹H NMR (400 MHz, Methanol-

d₄) d 7.35e7.18 (m, 5H), 6.04 (s, 1H), 5.47 (s, 1H), 4.92 (s, 1H), 3.92 (s,

2H). ¹³C NMR (101 MHz, Methanol-d₄)

d

137.5, 133.2, 128.7, 128.1,

into pET26b(þ)are described previously [56]. The obtained

VIM-2

126.3, 120.3, 38.9. HRMS (ESI): Calcd. for C₁₀H₁₀N₄[MþH]^þ

E. coli pET-26b(þ) MBL constructs were further used in the in the

187.0981; found 187.0978.

whole cell-based inhibitor assays.

In the second gene constructs used for the recombinant gene

expression, the native leader sequence was removed and replaced

with a hexa-His tag and a TEV cleavage site as reported earlier for

VIM-2 (residues V27-E268 [51]) and GIM-1 (residues Q19-D250

[50]) both in pDEST14. NDM-1 used a codon optimized synthetic

gene (Life Technologies, Thermo Fisher Scientiﬁc), with a TEV

cleavage site with sequence ENLYFQG and residues G36-R280 in

NDM-1 transformed in pDONR™221, and further sub cloned into

pDEST17 with carries an N-terminal hexa His-tag, yielding

4.1.8.2. 5-(5-Phenylpent-1-en-2-yl)-1H-tetrazole

(14b).

Dibutyltin oxide (0.58 mmol, 145.39 mg, 0.2 equiv), trimethylsilyl

azide (5.84 mmol, 0.672 g, 2 equiv) and 13b (2.92 mmol, 1 equiv)

gave 14b (2.16 mmol, 462 mg, 74%). ¹H NMR (400 MHz, Chloroform-

d)

d

7.14e7.00 (m, 5H), 5.87 (s, 1H), 5.47 (t, J ¼ 1.4 Hz, 1H), 4.84e4.83

(m, 1H), 2.58 (dd, J ¼ 15.7, 7.8 Hz, 4H), 1.78 (m, 2H). ¹³C NMR

(101 MHz, Chloroform-d)

d 157.3, 143.1, 134.7, 129.4, 129.4, 126.9,

120.6, 36.2, 34.0, 30.9. HRMS (ESI): Calcd. for C₁₂H₁₅N₄[MþH]^þ

215.1292; found 215.1291.

pDest17-NDM-1. Herein the residue numbering is the class B

lactamase numbering scheme will be applied [10].

b-

4.1.9. General procedure for the Michael addition to vinylic

tetrazoles (preparation of 15 and 16)

4.2.2. Recombinant protein expression and puriﬁcation of VIM-2,

Vinylic tetrazole 14 (1 equiv) and thioacetic acid (26 equiv) or

thiol (4 equiv) were dissolved in anhydrous DMF. The reaction

mixture was subjected to heating at 60 ^ꢀC in a tightly sealed vessel

for 22 h with stirring. The solvent was removed under reduced

pressure and the crude was subjected to column chromatography

using 5% MeOH in CH₂Cl₂.

GIM-1 and NDM-1

The three proteins were expressed and puriﬁed following this

protocol. pDest17-NDM-1 was transformed into E. coli BL21 Star

(DE3) pLysS (Invitrogen), and pDEST14 plasmids with VIM-2 or

GIM-1 were transformed into in-house modiﬁed E. coli BL21 Star

(DE3) pLysS (Invitrogen) cells with the pRARE plasmid (Novagen) in

order to allow expression of genes encoding tRNAs for rare codons

4.1.9.1. S-(3-phenyl-2-(1H-tetrazol-5-yl)propyl) ethanethioate (15a).

14a (0.32 mmol, 0.060 g, 1 equiv), and thioacetic acid (8.6 mmol,

0.662 g, 26 equiv) gave 15a (0.30 mmol, 79 mg, 95%) as a colorless

[57]. Precultures grown in Terriﬁc Broth (TB) media with 100

ampicillin (Sigma-Aldrich) and 34 g/mL chloramphenicol (Sigma-

Aldrich). The precultures were inoculated to 2 L Luria-Bertani (LB)

media containing 100 g/mL ampicillin and 34 g/mL chloram-

phenicol and grown at 37 ^ꢀC to reach an optical density (OD₆₀₀nm)

of 0.5e1.0 before induced expression with 0.4 mM isopropyl -D-1-

mg/mL

m

oil. ¹H NMR (400 MHz, Chloroform-d)

d

7.20e7.16 (m, 3H),

m

7.08e7.06 (m, 2H), 3.86 (q, J ¼ 10.5, 7.7 Hz, 1H), 3.44e3.40 (m, 1H),

3.33 (d, J ¼ 14.0 Hz,1H), 3.23 (d, J ¼ 7.6 Hz, 2H), 2.27 (s, 3H). ¹³C NMR

b

(101 MHz, Chloroform-d)

d

195.9, 157.9, 137.0, 128.9, 128.9, 128.7,

thigalactopyranoside (IPTG; Sigma-Aldrich). The induced cultures

were grown overnight at 20 ^ꢀC before collecting the cells by

centrifugation (8900 ꢂ g, 30 min, 4 ^ꢀC). Buffer A containing 50 mM

127.0, 39.7, 38.0, 31.8, 30.6. HRMS (ESI): Calcd. for C₁₂H₁₅ON₄S

[MþH]^þ263.0964; found 263.0961.

HEPES pH 7.2, 100 mM ZnCl₂and 150 mM NaCl was used to resus-

4.1.9.2. S-(5-phenyl-2-(1H-tetrazol-5-yl)pentyl) ethanethioate (15b).

14b (0.466 mmol, 0.100 mg, 1 equiv), and thioacetic acid

pend the cell pellets, following sonication and collecting the su-

pernatants by centrifugation (3000 g, 40 min,

4

^ꢀC). The

172

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

recombinant proteins were afﬁnity puriﬁed using a 1 mL or 5 mL

His-Trap HP column (GE Healthcare) in buffer A washed with 5%

buffer B (50 mM HEPES pH 7.2, 100 mM ZnCl₂, 150 mM NaCl and 1 M

of 1.6 mM in the assay) was added. Nitroceﬁn hydrolysis was

measured at OD₄₈₂every minute for 3 h with shaking (47 s) in

between reads using a Spectramax M2^espectrophotometer (Mo-

imidazole), before eluted in a gradient of 5e100% buffer B. Peak

fractions were investigated using 4e20% sodium dodecyl sulfate

polyacrylamide gel electrophoresis (SDS-PAGE; Bio-Rad) [58]. The

peaks containing MBL protein was added in-house-made His-tag-

ged TEV protease in a 1:100 mg ratio of TEV:protein and dialyzed at

lecular Devices). EDTA (250 mM concentration) was used as positive

control and wells containing no inhibitor as negative controls. The

percent inhibition was calculated according to equation (1).

!

Slope

ꢁ Slope

ðNo inhibitorÞ

ðInhibitorÞ

% inhibition ¼

x 100 %

4

^ꢀC overnight using 10-kDa cutoff Snakeskin (Pierce) in buffer C

(50 mM HEPES pH 7.2, 150 mM NaCl, 1 mM EDTA and 1 mM

Slope

b

-

(1)

mercaptoethanol). To remove uncleaved protein and TEV protease a

second His-Trap puriﬁcation was performed. SDS-PAGE analysis

was used to estimate a purity of ~95% of the fractions containing

protein, which then were pooled and dialyzed in buffer A overnight.

The synergistic effect of the inhibitors with meropenem was

tested against selected clinical bacterial strains containing MBLs.

The bacterial strains were plated on lactose agar plates with

100 mg/L ampicillin and lactose agar and incubated overnight at

37 ^ꢀC. The inhibitors were diluted to a ﬁnal concentration of 125

mM

4.2.3. Co-crystallization of VIM-2 with 2b, 10b and 10c

Co-crystallization of VIM-2 was successful with inhibitors 2b,

10b and 10c, when applying our new protocol for DMSO free co-

in Mueller Hinton (MH) broth. In order to monitor the effect of the

DMSO in the assay, a DMSO control was included with a concen-

tration of 5%. Meropenem was diluted in MH broth in a 2-fold

crystallization [45]. MRC 96-wells were pre-coated with 3

hibitor (100 mM) dissolved in DMSO, and DMSO evaporated after

leaving the plates in the fume hood for 24 h 60 L reservoir solution

ml in-

dilution series with ﬁnal concentrations of 256 mg/mL e 0.0625 mg/

m

mL. The microtiter plates were inoculated with a 0.5 McFarland

suspension of the bacterial strain in 0.85% NaCl, which were diluted

in MH broth. A quality check of bacterial suspension in 0.85% NaCl

in a 1:100 ratio was incubated on MH agar plates overnight at 37 ^ꢀC.

The ﬁnal CFU/mL inoculum was calculated and compared to a

standard. The microtiter plates were incubated for 20 h at 37 ^ꢀC.

The minimum inhibitory concentrations (MIC) were read visually of

the plates the next day.

consisting of 22e27% PEG 3350 and 0.2 M magnesium formate was

added to each well. The crystallization plates were then shaken for

24 h to allow the inhibitors to dissolve, followed by manual mixing

of hanging drops in a 1:1 ratio of VIM-2 (9.4 mg/mL) and reservoir.

The crystals were cryoproceeded in 25% PEG 3350, 0.2 M MgCl₂,15%

ethylene glycol and 50 mM HEPRS pH 7.2, before they were ﬂash

cooled in liquid nitrogen and X-ray data collected at ESRF, ID29.

Coordinate and structure factor ﬁles for the VIM-2_2b (PDB

accession number 5MM9), VIM-2_10b (PDB accession number

5NHZ) and VIM-2_10c (PDB accession number 5NI0) were depos-

ited in the Protein Data Bank (PDB).

Acknowledgments

This work has been funded by the Research Council of Norway

(FRIMEDBIO 2011; 213808) and this is highly acknowledged. The

provision of beam time at ID29, ESRF is highly valued. We

acknowledge Trine Joseﬁne O. Carlsen for puriﬁcation and crystal-

lization of proteins and Kinga Leszczak for contributions to the

synthetic work.

4.2.4. Dose rate inhibition studies for IC₅₀determination

The half-maximal inhibitory concentration (IC₅₀) against the

VIM-2, NDM-1 and GIM-1 enzymes were determined by using

sixteen different concentration of inhibitor compounds ranging

from 0

(pH 7.2), 100

GIM-1) and 2.5e0 mM inhibitor were incubated in a 96 well plate

at 25 ^ꢀC for 5 min. In addition, the enzyme buffer contained 400

g/

mL Bovine Serum Albumin (BSA) to prevent protein unfolding and

loss of activity due to low concentrations [59,60]. 100 M of the

mM to 250

mM. A 100 ml solution with 50 mM HEPES buffer

mM ZnCl₂), puriﬁed enzyme (10 nM VIM-2, NDM-1 or

Appendix A. Supplementary data

m

Supplementary data related to this article can be found at http://

dx.doi.org/10.1016/j.ejmech.2017.04.035.

m

reporter substrate nitroceﬁn (VIM-2, GIM-1) or imipenem (NDM-1)

was added to the enzyme-inhibitor solution and the increase in

absorbance at 482 nm (nitroceﬁn) or 260 nm (imipenem) was

recorded on a Spectramax M2e spectrophotometer (Molecular

Devices). Each data point was performed in triplicates and the

initial velocity for each inhibitor concentration was analyzed by a

log [inhibitor] vs. response curve ﬁtting to calculate IC₅₀in

GraphPad Prism 5.0 software.

References

[1] World Health Organization (WHO), Antimicrobial Resistance: Global Report

on Surveillance, 2014.

[2] World Health Organization (WHO), The Evolving Threat of Antimicrobial

Resistance : Options for Action, 2012.

[3] A.M. Queenan, K. Bush, Clin. Microbiol. Rev. 20 (2007) 440e458.

[4] J.D. Pitout, Drugs 70 (2010) 313e333.

[5] R.P. Ambler, Philos. Trans. R. Soc. B 289 (1980) 321e331.

[6] K. Bush, G.A. Jacoby, Antimicrob. Agents Chemother. 54 (2010) 969e976.

[7] T.R. Walsh, Int. J. Antimicrob. Agents 36 (2010) S8eS14.

[8] T. Palzkill, Ann. N. Y. Acad. Sci. 1277 (2013) 91e104.

[9] J. Jeon, J. Lee, J. Lee, K. Park, A. Karim, C.-R. Lee, B. Jeong, S. Lee, Int. J. Mol. Sci.

16 (2015) 9654e9692.

[10] G. Garau, I. García-Saez, C. Bebrone, C. Anne, P. Mercuri, M. Galleni, J.M. Frere,

O. Dideberg, Antimicrob. Agents Chemother. 48 (2004) 2347e2349.

[11] T.R. Walsh, M.A. Toleman, L. Poirel, P. Nordmann, Clin. Microbiol. Rev. 18

(2005) 306e325.

[12] G. Patel, R. Bonomo, Front. Microbiol. 4 (2013) 1e17.

[13] J.D. Buynak, Biochem. Pharmacol. 71 (2006) 930e940.

[14] J.L. Liscio, M.V. Mahoney, E.B. Hirsch, Int. J. Antimicrob. Agents 46 (2015)

266e271.

[15] S.M. Drawz, K.M. Papp-Wallace, R.A. Bonomo, Antimicrob. Agents Chemother.

58 (2014) 1835e1846.

4.2.5. Cell-based screening assay of the inhibition potential

The inhibitory activity of the inhibitors was investigated in a

cell-based assay using a b-lactamase-negative E. coli SNO3 (ampA1

ꢀ

ꢁ

ampC8 pyrB recA rpsL) (24) transformed with pET26b(þ) containing

bla_VIM-2, bla_GIM-1or bla_NDM-1. The screen was conducted in 96-well

plates (Corning) in duplicates. One

250 M), 50 l overnight culture (adjusted to an OD₆₀₀of 1 in LB

broth) of E. coli SNO3 containing one of the MBLs and 50 l LB

media with ﬁnal concentration 0.8 mM Isopropyl ß- -1-

ml inhibitor (ﬁnal concentration

m

D

thiogalactopyranoside (Sigma, IPTG) was added to each well. The

ꢀ

[16] M.M. Gonzalez, A.J. Vila, in, Springer Berlin Heidelberg, Berlin, Heidelberg,

plate was incubated at 37 ^ꢀC for 20 min with shaking to induce the

2016, pp. 1e34.

expression of the MBL. Subsequently, 50

ml nitroceﬁn (diluted in

[17] W. Jing-Fang, C. Kuo-Chen, Curr. Top. Med. Chem. 13 (2013) 1242e1253.

[18] W. Fast, L.D. Sutton, Biochim. Biophys. Acta (2013) 1648e1659.

50 mM HEPES pH 7.2 and 100 M ZnCl₂to give a ﬁnal concentration

m

S. Skagseth et al. / European Journal of Medicinal Chemistry 135 (2017) 159e173

173

[19] J. Brem, S.S. van Berkel, D. Zollman, S.Y. Lee, O. Gileadi, P.J. McHugh,

T.R. Walsh, M.A. McDonough, C.J. Scho ﬁ eld, Antimicrob. Agents Chemother. 60

(2016) 142e150.

[41] T. Christopeit, H.K. Leiros, Bioorg. Med. Chem. Lett. 26 (2016) 1973e1977.

[42] S. Skagseth, T.J. Carlsen, G.E. Bjerga, J. Spencer, Ø. Samuelsen, H.K.S. Leiros,

Antimicrob. Agents Chemother. 60 (2015) 990e1002.

[20] F.-M. Klingler, T.A. Wichelhaus, D. Frank, J. Cuesta-Bernal, J. El-Delik,

[43] P. Lassaux, M. Hamel, M. Gulea, H. Delbrück, P.S. Mercuri, L. Horsfall,

€

ꢁ

H.F. Müller, H. Sjuts, S. Gottig, A. Koenigs, K.M. Pos, D. Pogoryelov, E. Proschak,

D. Dehareng, M. Kupper, J.M. Frere, K. Hoffmann, M. Galleni, C. Bebrone,

J. Med. Chem. 58 (2015) 3626e3630.

[21] M.M. Gonzalez, M. Kosmopoulou, M.F. Mojica, V. Castillo, P. Hinchliffe,

J. Med. Chem. 53 (2010) 4862e4876.

ꢀ

[44] H. Nikaido, Microbiol. Mol. Biol. Rev. 67 (2003) 593e656.

[45] T. Christopeit, K.-W. Yang, S.-K. Yang, H.-K.S. Leiros, Acta Cryst. F72 (2016)

813e819.

I. Pettinati, J. Brem, C.J. Scho ﬁ eld, G. Mahler, R.A. Bonomo, L.I. Llarrull,

J. Spencer, A.J. Vila, ACS Infect. Dis. 1 (2015) 544e554.

[22] N. Li, Y. Xu, Q. Xia, C. Bai, T. Wang, L. Wang, D. He, N. Xie, L. Li, J. Wang, H.-

G. Zhou, F. Xu, C. Yang, Q. Zhang, Z. Yin, Y. Guo, Y. Chen, Bioorg. Med. Chem.

Lett. 24 (2014) 386e389.

[46] H. Kurosaki, Y. Yamaguchi, H. Yasuzawa, W.C. Jin, Y. Yamagata, Y. Arakawa,

ChemMedChem 1 (2006) 969e972.

ꢀ

[47] P. Hinchliffe, M.M. Gonzalez, M.F. Mojica, J.M. Gonzalez, V. Castillo, C. Saiz,

M. Kosmopoulou, C.L. Tooke, L.I. Llarrull, G. Mahler, R.A. Bonomo, A.J. Vila,

J. Spencer, Proc. Natl. Acad. Sci. U. S. A. 113 (2016) E3745eE3754.

[48] M.F. Mojica, S.G. Mahler, C.R. Bethel, M.A. Taracila, M. Kosmopoulou,

K.M. Papp-Wallace, L.I. Llarrull, B.M. Wilson, S.H. Marshall, C.J. Wallace,

M.V. Villegas, M.E. Harris, A.J. Vila, J. Spencer, R.A. Bonomo, Biochemistry-US

54 (2015) 3183e3196.

ꢀ

[23] B.M.R. Lienard, G. Garau, L. Horsfall, A.I. Karsisiotis, C. Damblon, P. Lassaux,

ꢁ

C. Papamicael, G.C.K. Roberts, M. Galleni, O. Dideberg, J.M. Frere, C.J. Scho ﬁ eld,

Org. Biomol. Chem. 6 (2008) 2282e2294.

[24] Q. Sun, A. Law, M.W. Crowder, H.M. Geysen, Bioorg. Med. Chem. Lett. 16

(2006) 5169e5175.

[25] W.C. Jin, Y. Arakawa, H. Yasuzawa, T. Taki, R. Hashiguchi, K. Mitsutani,

A. Shoga, Y. Yamaguchi, H. Kurosaki, N. Shibata, M. Ohta, M. Goto, Biol. Pharm.

Bull. 27 (2004) 851e856.

[49] D.T. King, L.J. Worrall, R. Gruninger, N.C.J. Strynadka, J. Am. Chem. Soc. 134

(2012) 11362e11365.

[26] U. Heinz, R. Bauer, S. Wommer, W. Meyer-Klaucke, C. Papamichaels,

J. Bateson, H.W. Adolph, J. Biol. Chem. 278 (2003) 20659e20666.

[27] C. Mollard, C. Moali, C. Papamicael, C. Damblon, S. Vessilier, G. Amicosante,

[50] P.S. Borra, Ø. Samuelsen, J. Spencer, T.R. Walsh, M.S. Lorentzen, H.-K.S. Leiros,

Antimicrob. Agents Chemother. 57 (2013) 848e854.

[51] T. Christopeit, T.J. Carlsen, R. Helland, H.-K.S. Leiros, J. Med. Chem. (2015)

8671e8682.

ꢁ

C.J. Scho ﬁ eld, M. Galleni, J.-M. Frere, G.C.K. Roberts, J. Biol. Chem. 276 (2001)

45015e45023.

[52] M. Castanheira, M.A. Toleman, R.N. Jones, F.J. Schmidt, T.R. Walsh, Antimicrob.

Agents Chemother. 48 (2004) 4654e4661.

[53] D. Yong, M.A. Toleman, C.G. Giske, H.S. Cho, K. Sundman, K. Lee, T.R. Walsh,

Antimicrob. Agents Chemother. 53 (2009) 5046e5054.

[28] N.O. Concha, C.A. Janson, P. Rowling, S. Pearson, C.A. Cheever, B.P. Clarke,

ꢁ

C. Lewis, M. Galleni, J.-M. Frere, D.J. Payne, J.H. Bateson, S.S. Abdel-Meguid,

Biochemistry-US 39 (2000) 4288e4298.

[29] M.I. Page, A.P. Laws, Chem. Commun. (1998) 1609e1617.

[30] S. Bounaga, A.P. Laws, M. Galleni, M.I. Page, Biochem. J. 331 (1998) 703e711.

[31] M. Goto, T. Takahashi, F. Yamashita, A. Koreeda, H. Mori, M. Ohta, Y. Arakawa,

Biol. Pharm. Bull. 20 (1997) 1136e1140.

[32] J. Brem, S.S. van Berkel, W. Aik, A.M. Rydzik, M.B. Avison, I. Pettinati, K.-

D. Umland, A. Kawamura, J. Spencer, T.D.W. Claridge, M.A. McDonough,

C.J. Scho ﬁ eld, Nat. Chem. 6 (2014) 1084e1090.

[33] G.A. Patani, E.J. LaVoie, Chem. Rev. 96 (1996) 3147e3176.

[34] C. Ballatore, D.M. Huryn, A.B. Smith, ChemMedChem 8 (2013) 385e395.

[35] Y. Yamaguchi, W. Jin, K. Matsunaga, S. Ikemizu, Y. Yamagata, J.I. Wachino,

N. Shibata, Y. Arakawa, H. Kurosaki, J. Med. Chem. 50 (2007) 6647e6653.

[36] J.D. Park, D.H. Kim, J. Med. Chem. 45 (2002) 911e918.

[37] R.V. Bikbulatov, F. Yan, B.L. Roth, J.K. Zjawiony, Bioorg. Med. Chem. Lett. 17

(2007) 2229e2232.

[38] L. Zervas, I. Photaki, N. Ghelis, J. Am, Chem. Soc. 85 (1963) 1337e1341.

[39] P.G. Baraldi, G.P. Pollini, V. Zanirato, A. Barco, S. Benetti, Synthesis (1985)

969e970.

[54] K.K. Kumarasamy, M.A. Toleman, T.R. Walsh, J. Bagaria, F. Butt,

R. Balakrishnan, U. Chaudhary, M. Doumith, C.G. Giske, S. Irfan, P. Krishnan,

A.V. Kumar, S. Maharjan, S. Mushtaq, T. Noorie, D.L. Paterson, A. Pearson,

C. Perry, R. Pike, B. Rao, U. Ray, J.B. Sarma, M. Sharma, E. Sheridan,

M.A. Thirunarayan, J. Turton, S. Upadhyay, M. Warner, W. Welfare,

D.M. Livermore, N. Woodford, Lancet Infect. Dis. 10 (2010) 597e602.

[55] H.-K.S. Leiros, K.S.W. Edvardsen, G.E.K. Bjerga, Ø. Samuelsen, FEBS J. 282

(2015) 1031e1042.

[56] Ø. Samuelsen, M. Castanheira, T.R. Walsh, J. Spencer, Antimicrob. Agents

Chemother. 52 (2008) 2905e2908.

[57] E.A. Rogulin, T.A. Perevyazova, L.A. Zheleznaya, N.I. Matvienko, Biochemistry-

Moscow þ 69 (2004) 1123e1127.

[58] U.K. Laemmli, Nature 227 (1970) 680e685.

[59] N. Laraki, N. Franceschini, G.M. Rossolini, P. Santucci, C. Meunier, E. de Pauw,

ꢁ

G. Amicosante, J.M. Frere, M. Galleni, Antimicrob. Agents Chemother. 43

(1999) 902e906.

[60] S. Siemann, D.P. Evanoff, L. Marrone, A.J. Clarke, T. Viswanatha,

G.I. Dmitrienko, Antimicrob. Agents Chemother. 46 (2002) 2450e2457.

[40] S.M. Lukyanov, I.V. Bliznets, S.V. Shorshnev, G.G. Aleksandrov, A.E. Stepanov,

A.A. Vasil'ev, Tetrahedron 62 (2006) 1849e1863.

Article Doi

Metallo-β-lactamase inhibitors by bioisosteric replacement: Preparation, activity and binding

DOI: 10.1016/j.ejmech.2017.04.035

Source and publish data:

Authors:

Article abstract of DOI:10.1016/j.ejmech.2017.04.035

Full text of DOI:10.1016/j.ejmech.2017.04.035

Products guided by the article

R&D Labs maybe for 104239-20-5

Relevant to this article

Hot Product