Synthesis and SAR studies of novel benzodiazepinedione-based inhibitors of Clostridium difficile (C. difficile) toxin B (TcdB)

Article abstract of DOI:10.1016/j.bmcl.2018.10.047

Synthesis and structure-activity relationships (SAR) of a novel series of benzodiazepinedione-based inhibitors of Clostridium difficile toxin B (TcdB) are described. Compounds demonstrating low nanomolar affinity for TcdB, and which possess improved stability in mouse plasma vs. earlier compounds from this series, have been identified. Optimized compound 11d demonstrates a good pharmacokinetic (PK) profile in mouse and hamster and is efficacious in a hamster survival model of Clostridium difficile infection.

Full text of DOI:10.1016/j.bmcl.2018.10.047

Accepted Manuscript
Synthesis and SAR studies of novel benzodiazepinedione-based inhibitors of
Clostridium difficile (C. difficile) toxin B (TcdB)
Jeffrey J. Letourneau, Ilana L. Stroke, David W. Hilbert, Andrew G. Cole, Laurie
J. Sturzenbecker, Brett A. Marinelli, Jorge G. Quintero, Joan Sabalski, Yanfang
Li, Linh Ma, Igor Pechik, Philip D. Stein, Maria L.Webb
PII:
S0960-894X(18)30849-7
https://doi.org/10.1016/j.bmcl.2018.10.047
BMCL 26106
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
21 September 2018
23 October 2018
28 October 2018
Please cite this article as: Letourneau, J.J., Stroke, I.L., Hilbert, D.W., Cole, A.G., Sturzenbecker, L.J., Marinelli,
B.A., Quintero, J.G., Sabalski, J., Li, Y., Ma, L., Pechik, I., Stein, P.D., L.Webb, M., Synthesis and SAR studies of
novel benzodiazepinedione-based inhibitors of Clostridium difficile (C. difficile) toxin B (TcdB), Bioorganic &
Medicinal Chemistry Letters (2018), doi: https://doi.org/10.1016/j.bmcl.2018.10.047
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com
Synthesis and SAR studies of novel benzodiazepinedione-based inhibitors of
Clostridium difficile (C. difficile) toxin B (TcdB)
Jeffrey J. Letourneaua,*, Ilana L. Strokea, David W. Hilbertb, Andrew G. Colea,c, Laurie J. Sturzenbeckera,
Brett A. Marinellia, Jorge G. Quinteroa,c, Joan Sabalskia, Yanfang Lia, Linh Maa,d, Igor Pechika, Philip D.
Steina,e, Maria L.Webba,f
aVenenum Biodesign, a member of Genesis Biotechnology Group, 8 Black Forest Road, Hamilton, NJ 08691, USA
bFemeris Women’s Health Research Center, a member of Genesis Biotechnology Group, 2000 Waterview Drive, Hamilton, NJ 08691, USA
c Current address is Arbutus Biopharma Inc., 701 Veterans Circle, Warminster, PA 18974, USA
d Current address is Bristol-Myers Squibb, Route 206 and Province Line Road, Princeton, NJ 08543, USA
e Deceased
f Current address is Merck Research Labs, 2015 Galloping Hill Rd, Kenilworth, NJ 07033, USA
ARTICLE INFO
ABSTRACT
Article history:
Received
Revised
Accepted
Available online
Synthesis and structure-activity relationships (SAR) of a novel series of benzodiazepinedione-
based inhibitors of Clostridium difficile toxin B (TcdB) are described. Compounds
demonstrating low nanomolar affinity for TcdB, and which possess improved stability in mouse
plasma vs. earlier compounds from this series, have been identified. Optimized compound 11d
demonstrates a good pharmacokinetic (PK) profile in mouse and hamster and is efficacious in a
hamster survival model of Clostridium difficile infection.
Keywords:
2009 Elsevier Ltd. All rights reserved.
Clostridium difficile
Toxin
TcdB
Inhibitor
Benzodiazepinedione
Clostridium difficile infection (CDI) is the most common cause of infectious hospital-acquired diarrhea, with significant morbidity,
mortality and associated healthcare costs.1,2 The major risk factor for CDI is antimicrobial therapy, which disrupts the normal gut
microbiota and allows C. difficile to flourish. Treatment of CDI with antimicrobials is generally effective in the short-term, but recurrent
infections are frequent and problematic,3 indicating that improved treatment options are necessary. Symptoms of disease are largely due
to two paralogous toxins, TcdA and TcdB, which are glucosyltransferases that hydrolyze UDP-glucose and transfer the glucose moiety
to host Rho GTPases, resulting in irreversible inhibition.
To date, several experimental approaches specifically targeting the bacterial toxins have been explored, including development of
toxin binding antibodies.4 Bezlotoxumab, a human monoclonal antibody which binds TcdB and neutralizes its effects, was recently
approved by the FDA and is indicated to reduce recurrence of CDI in adults who are receiving antibacterial drug treatment for CDI and
are at a high risk of recurrence.5 Another emerging approach towards possible treatment of CDI involves direct inhibition of the
glucosyltransferase activity of C. difficile toxins.6 We recently reported details of our hit-to-lead optimization effort around a novel
series of benzodiazepinedione-based inhibitors of TcdB.7
This effort resulted in highly potent inhibitors 1 (IC50 = 1 nM) and 2 (IC50 = 3 nM) which were demonstrated to inhibit induction of
apoptosis of mammalian cells (CHO cells) by either TcdB or TcdA (Figure 1). However, these compounds were unstable in mouse
plasma leading to suboptimal mouse PK profiles, thus precluding further advancement into rodent models of CDI. As such, we initiated
a lead optimization program, utilizing compounds 1 and 2 as our starting point, aiming to identify potent inhibitors with improved
stability in mouse plasma and hence improved PK profiles suitable for advancement into rodent models of CDI.
Testing of compound 2 in a mouse plasma stability assay revealed a rapid disappearance of the parent compound, with amide 2 being
undetectable after a 2-hour incubation in mouse plasma.7 In addition, the formation of the potential hydrolysis product, carboxylic acid
3, was assayed for in the mouse plasma stability testing and it was detected after a 2-hour incubation of amide 2 in mouse plasma. We
reasoned that the likely cause of instability in mouse plasma for compound 2, and related congeners, was enzyme-mediated hydrolysis
(e.g., by amidases) of the N-heteroaryl amide moiety (Figure 2).8

H
O
N
Cl
N
(R)
O
N
N
plasma
enzyme-
mediated
R
N
O
X
H
H
H
O
O
N
N
Cl
Cl
N
N
N
1
; X = CH, R = Me
TcdB IC50 = 1 nM
; X = N, R = OMe
TcdB IC50 = 3 nM
(R)
O
O
N
N
N
hydrolysis
of amide
O
OH
N
O
2
O
N
H
2
3
H
O
N
H
Cl
O
R
N
Cl
N
O
N
N
O
O
N
N
O
N
R1
H
R
O
A
B
O H
H
ketone
N
sterically block
hydrolysis
R =
*
N
*
etc.
N
Figure 1. TcdB inhibitors
Described herein are efforts to identify potent inhibitors of TcdB with improved stability in mouse plasma, for possible advancement
into rodent models of CDI. These efforts focused on structural changes in the region of the amide linkage that were intended to reduce
hydrolysis. Two distinct strategies were employed to achieve this: 1) introduction of steric shields (i.e., R1 and/or R2 substituents) onto
the phenyl ring and next to the amide moiety in order to block the labile amide linkage (A in Figure 2); and, 2) replacement of the labile
amide linkage with a ketone (B in Figure 2).
The compounds appearing in Tables 1 and 2 were synthesized via the general routes outlined in Schemes 1 and 2.9 As illustrated in
Scheme 1, the synthesis of amides 11 bearing substituents (R1 and/or R2) on the phenyl ring of the amide started with conversion of the
appropriate 4-bromobenzoic acid to the acid chloride followed by reaction with potassium tert-butoxide to provide bromo-t-butylester
5. Metal halogen exchange with n-BuLi at low temperature, and reaction of the resultant aryllithium intermediate with DMF provided
the aldehyde 6. Reductive amination of 6 with (R)-methyl 2-amino-3-(pyridin-2- yl)propanoate hydrochloride provided secondary
amine 7. Acylation with acid chloride 810 gave azidobenzamide intermediate 9. Conversion of 9 to the benzodiazepinedione-based
carboxylic acid 10 was achieved via an aza-Wittig reaction11 followed by acidic hydrolysis of the resultant iminoether intermediate and
finally deprotection of the tert-butyl ester with TFA in DCM. Finally, arylamides 11 were prepared by reaction of carboxylic acid 10
with aryl amines (ArNH2) using the coupling reagent HATU and heating in a microwave reactor. As shown in Scheme 2, the synthesis
of heteroaryl ketones 13 started with conversion of carboxylic acid 3 to the corresponding Weinreb amide 12. The Weinreb amide was
then reacted with either lithiated-heteroaryls or heteroaryl Grignard reagents to provide the heteroaryl ketones 13.
The compounds in Tables 1 and 2 were evaluated for their ability to inhibit the UDP-glucose hydrolysis activity of TcdB.12
Figure 2. Plasma enzyme-mediated hydrolysis of amide 2 to give carboxylic acid 3. Strategies to improve plasma stability: A - introduce steric shields (R1 and/or
R2 around amide linkage; B - replace labile amide linkage with ketone.
We prepared a number of N-heteroaryl amide-based analogs containing phenyl substituents (R1 and/or R2) intended to block amide
hydrolysis (Table 1). To determine whether or not substitution of the phenyl ring would be tolerated from a potency perspective we first
prepared analogs of carboxylic acid 3 (IC50 = 0.18 µM) containing either a single chloro substituent (compound 10a) or two chloro
substituents (compound 10b). Compound 10a was only 2 to 3-fold less potent than compound 3, but compound 10b was more than an
order of magnitude less potent than 10a. Having previously established that replacement of the carboxylic acid moiety in 3 with certain
N-heteroaryl amides significantly increases affinity for the target,7 we then prepared a number of analogs containing a single substituent
in the ortho-position with relation to the labile amide functionality. Substitution with chloro methyl, or methoxy, in combination with a
variety of differently substituted N-pyridin-2-yl (R2 = Cl; 11a-11e, R2 = Me; 11j-11n, R2 = OMe; 11q-11t), N-pyrimidin-2-yl (R2 = Cl;
11f, 11g, R2 = Me; 11o, 11p) or N-pyrazin-2-yl (R2 = Cl; 11h, 11i) amides provided analogs with good binding affinity, with IC50
values ranging from 3 to 41 nM. In order to ascertain if this strategy could retard amide hydrolysis, we tested the stability of 11b in
mouse plasma. Indeed, 100% of parent compound remained after 2 hours following incubation with mouse plasma. In comparison, no
parental 2 remained after 2 hours following incubation with mouse plasma.
R1
O
R1
O
R1
O
R1
O
O
N
O
H
a,b
c
d
O
O
OH
R2
N
O
R2
R2
Br
R2
Br
O
4
6
5
7
O
H
H
O
N
O
O
N
O
O
N
N
Cl
R1
R1
i
N
Cl
f,g,h
R1
e
7
+
Cl
O
N3
N
O
O
O
N
N
Cl
N3
OH
HetAr
N
O
O
R2
R2
H
R2
8
10
11
Cl
9
Scheme 1. Reagents and conditions: a) oxalyl chloride, DMF (cat), DCM; b) KOt-Bu, THF, 0 °C to 23 °C, 16 h; c) n-BuLi, THF, -100 °C, 30 min; DMF, -100
°C, THF, 45 min; d) (R)-methyl 2-amino-3-(pyridin-2-yl)propanoate hydrochloride, Et3N, MeOH, 30 min; NaCNBH3, HOAc, MeOH, 16 h; e) DIPEA, DCM,
DMAP (cat); f) P(Bu)3, toluene, 70 °C; g) TFA/H2O/THF (1:1:10); h) 20% TFA/DCM; i) HetArNH2, HATU, DIPEA, NMP, 4Å molecular sieves, microwave @

100 °C
H
H
O
O
H
N
N
O
Cl
Cl
N
N
a
b or c
N
Cl
N
O
O
O
N
N
N
O
OH
N
O
O
HetAr
O
3
12
13
Scheme 2. Reagents and conditions: a) CDI, DCM; N,O-dimethylhydroxyl amine hydrochloride, DIPEA; b) XMg-HetAr (X = Cl or Br), THF, 0 °C or 23 °C; c)
Li-HetAr, THF, -78 °C.
Table 1. SAR – N-heteroaryl amides containing steric shields
H
H
O
O
N
N
Cl
Cl
N
N
R1
R2
R1
R2
(R)
(R)
N
O
N
O
O
N
N
O
R3
OH
X Y
H
3, 10a-b
11a-11t
a
CHO cell
TcdB,
CHO cell
TcdA,
HLM/MLM,
Compd
R1
R2
R3
X
Y
TcdB, IC50a (µM)
Value
s
means
of at
least
% rem. @ 15 minb
IC50a (μM)
ND
IC50a (μM)
ND
are
3
H
H
H
-
-
-
-
-
-
-
0.18 (±0.08)
0.48 (±0.25)
ND
ND
10a
Cl
ND
ND
two
10b
11a
11b
11c
11d
11e
Cl
H
H
H
H
H
Cl
Cl
Cl
Cl
Cl
Cl
-
-
7.4 (±1.3)
ND
ND
ND
experi
ments
unless
other
wise
indicat
ed,
and
standa
rd
deviati
on is
given
in
parent
heses;
[n=1]
indicat
es
Me
OMe
F
CF3
CN
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
0.005 (±0.001)
0.005 (±0.002)
0.004 (±0.000)
0.018 (±0.014)
0.008 (±0.002)
0.024 (±0.009)
0.061 (±0.070)
0.13 (±0.023)
0.25 (±0.14)
0.052 (±0.027)
0.58 (±0.11)
0.76 (±0.072)
1.8 (±0.36)
6.3 (±2.6)
ND
57/57
100/71
88/69
94/92
90/84
11f
11g
11h
11i
H
H
H
H
Cl
Cl
Cl
Cl
Me
N
CH
CH
N
0.004 (±0.001)
0.004 (±0.001)
0.009 (±0.004)
0.021 (±0.002)
0.13 (±0.052)
0.053 (±0.045)
0.22 (±0.45)
0.49 (±0.19)
1.5 [n=1]
79/33
73/38
81/79
100/68
OMe
Me
N
0.92 (±0.45)
2.1 (±0.46)
3.4 [n=1]
CH
CH
OMe
N
11j
11k
11l
11m
11n
H
H
H
H
H
Me
Me
Me
Me
Me
Me
OMe
F
CF3
CN
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
0.007 (±0.002)
0.010 (±0.013)
0.008 (±0.007)
0.020 (±0.017)
0.003 (±0.001)
0.061 (±0.004)
0.071 (±0.021)
0.062 (±0.015)
0.17 (±0.11)
0.86 (±0.11)
0.48 (±0.15)
0.86 (±0.14)
2.4 (±1.2)
ND
61/31
68/51
85/60
100/99
90/70
0.078 (±0.072)
11o
11p
11q
11r
11s
11t
H
H
H
H
H
H
Me
Me
Me
N
CH
CH
CH
CH
CH
CH
0.004 (±0.001)
0.004 (±0.000)
0.009 (±0.002)
0.005 (±0.002)
0.014 (±0.007)
0.041 (±0.022)
0.13 (±0.063)
0.029 (±0.016)
1.9 (±2.5)
3.0 (±0.51)
0.69 (±0.14)
11(±6.6)
74/44
82/39
62/49
78/71
94/67
100/83
value
is
from a
single
experi
ment;
ND
not
deter
OMe
Me
N
OMe
OMe
OMe
OMe
CH
CH
CH
CH
OMe
F
0.13 (±0.082)
1.1 (±0.56)
1.1 (±0.21)
17 (±9.3)
=
mined.
b
CF3
0.66 (±0.26)
31 (±7.4)
Comp
ound assayed in human liver microsomes (HLM) or mouse liver microsomes (MLM), percentage compound remaining unchanged after 15 min was determined.
Next, we investigated replacement of the N-heteroaryl amide moiety with various heteroaryl ketones (Table 2). As described
elsewhere,7 we initially designed N-heteroaryl amide-based inhibitors based on a hypothesis that the N-heteroaryl amide moiety could
bind to the Mn+2 cofactor of TcdB via a bidentate chelate interaction. This hypothesis has been subsequently confirmed by x-ray
crystallography (Pechik (2018), manuscript in preparation). Therefore the heteroaryl ketone-based analogs that were prepared each
possessed strategic placement of a heteroatom in the heteroaryl moiety to allow possible bidentate coordination of the manganese
cofactor in the enzyme (Figure 3). First we prepared the 2-pyridyl ketone 12a, envisioning that it could coordinate the Mn+2 cofactor in
a bidentate fashion akin to B as illustrated in Figure 3. This analog possessed moderate potency against TcdB (IC50 = 45 nM).
Methylation of the pyridyl ring was tolerated at C(5) (12c) and at C(4) (12d) resulting in analogs with affinity at TcdB comparable to
that of 12a.

Table 2. SAR – Heteroaryl ketones
H
O
N
Cl
N
(R)
N
O
O
HetAr
12a-12j
aValue
are
means
of at
least
HLM/MLM,
CHO cell TcdB, IC50a (μM) CHO cell TcdA, IC50a (μM)
Compd
HetAr
TcdB, IC50a (µM)
0.045 (±0.014)
% rem. @ 15 minb
s
N
N
N
N
*
12a
4.2 (±2.0)
ND
37 (±1.0)
ND
64/79
25/3
two
*
experi
ments
unless
other
wise
indicat
ed,
and
12b
12c
0.34 (±0.06)
*
*
0.038 (±0.011)
5.9 [n=1]
23 [n=1]
57/68
12d
12e
12f
0.074 (±0.073)
0.19 (±0.14)
0.23 (±0.08)
ND
ND
26/48
49/33
44/16
standa
rd
deviati
on is
given
in
parent
heses;
[n=1]
indicat
es
value
is
from a
single
experi
ment;
*
*
N
> 100 [n=1]
> 100 [n=1]
> 100 [n=1]
> 100 [n=1]
N
N
N
N
*
*
12g
12h
0.006 (±0.004)
0.002 (±0.001)
0.33 (±0.14)
1.8 (±0.69)
5.3 (±0.81)
18 (±2.3)
67/76
87/93
N
N
ND
not
=
N
H
deter
mined
N
*
*
.
N
12i
12j
0.008 (±0.006)
0.061 (±0.015)
0.66 (±0.35)
1.6 (±0.34)
8.1 (±2.8)
ND
44/27
15/44
b
Comp
ound
assaye
N
N
N
d
in
human
liver
microsomes (HLM) or mouse liver microsomes (MLM); percentage of compound remaining unchanged after 15 min was determined.
However methylation of the pyridyl ring at C(6) (12b) led to an 8-fold drop in potency vs. 12a. Replacement of the 2-pyridyl ring with
1-methylimidazol-2-yl (12e) or 1-methylimidazol-4-yl (12f) led to a five-fold loss of potency in both instances. Next we investigated
introduction of bicyclic heteroaromatics (12g-j) which we envisioned could coordinate the Mn+2 cofactor in a bidentate fashion
analogous to C as shown in Figure 3. Azaindoles 12g and 12h and imidazopyridine 12i all possessed excellent binding affinity at TcdB;
N-heteroaryl amides
Heteroaryl ketones
Mn+2
N
Mn+2
Mn+2
O
O
N
O
N
N
H
N
A
B
C
of these,
Figure 3. Proposed bidentate coordination of the Mn+2 cofactor in TcdB by a representative N-heteroaryl amide (A) or by representative heteroaryl ketones (B
and C).
12h was the most potent (IC50 = 2nM). The triazolopyridine 12j had decreased affinity (IC50 = 61 nM).
Most of the compounds in Tables 1 and 2 underwent further evaluation in a whole cell assay; in which their ability to inhibit
induction of apoptosis in cultured mammalian cells (CHO) by full-length toxin (TcdB and/or TcdA) was determined.13 A number of
amide-based inhibitors (Table 1) with good binding affinity at TcdB demonstrated submicromolar activity in the cell-based assay in the
presence of TcdB. Activity, albeit more modest, was also demonstrated in the cell-based assay in the presence of TcdA for this class of
inhibitors. Methoxy substitution of the phenyl ring (R2) generally had an adverse effect on activity in the whole cell assays vs. chloro or
methyl substitution (e.g., compare 11q to 11a and 11j). Only two of the more potent ketone-based inhibitors (12g and 12i)

demonstrated submicromolar activity in the cell-based assay in the presence of TcdB. In vitro microsomal stability, in the presence of
human liver microsomes (HLM) or mouse liver microsomes (MLM), was also assessed for both amide and ketone-based inhibitors to
Table 3. Pharmacokinetic parameters for 11d in CD-1 mice after i.v. and p.o. dosing
Route (dose)a
Cl (mL/min/kg)
T1/2 (h)
i.v. (1 mg/kg)
28
2.6
6.2
592
312
p.o. (10 mg/kg)
Vz (L/kg)
AUC0-inf (ng/mL h)
Cmax (ng/mL)
Tmax (h)
9423
1533
0.25
159
F (%)
aVehicle for i.v. dosing is 40% PEG400/5% ethanol/ 55% water (v/v/v), and for p.o. dosing is 20% PEG400/80% (0.5% (w/v) CMC containing 0.25% (v/v)
Tween 80) (v/v).
identify possible candidates for progression into mouse pharmacokinetic (PK) experiments.
The main focus of our lead optimization program was to identify compounds for progression into mouse and hamster models of CDI.
Therefore the pharmacokinetic profile of various analogs demonstrating both submicromolar activity in the whole cell assay in the
presence of TcdB, and reasonable stability in the mouse microsome stability assay (i.e., MLM stability > 50% remaining at 15 min),
was determined after oral (p.o.) administration in CD-1 mice. Compound 11d emerged as having particularly high exposure after oral
dosing. Therefore full PK (i.e., both intravenous (i.v.) and oral (p.o.) dosing) was subsequently obtained for this compound (Table 3).
Compound 11d showed moderate clearance (Cl) and high apparent volume of distribution during the terminal phase (Vz) leading to a
moderate elimination half-life (T1/2) of 2.6 hours. After oral dosing (10 mg/kg), 11d exhibited rapid absorption (Tmax = 0.25 h) and
excellent oral bioavailability (F > 100%). Hamster PK was also determined after p.o. administration of compound 11d (Table 4) since a
hamster survival model is widely considered the gold standard amongst animal models for development of potential CDI therapeutics.14
In hamsters, excellent exposure was observed with low apparent clearance (Cl/F) and an elimination half-life of 10 hours. Additionally,
rapid absorption (Tmax = 0.25 h) was observed, as it was in mice.
Table 4. Pharmacokinetic parameters for 11d in hamsters after p.o. dosing
Route (dose)a
Cl/F (mL/min/kg)
T1/2 (h)
p.o. (10 mg/kg)
7.8
10
AUCinf (ng/mL h)
Cmax (ng/mL)
Tmax (h)
21477
2747
0.25
aVehicle for i.v. dosing is 40% PEG400/ 5% ethanol/ 55% water (v/v/v), and for p.o. dosing is 20% PEG400/80% (0.5% (w/v) CMC containing 0.25% (v/v)
Tween 80) (v/v).
Given its inhibitory activity at TcdB, activity in the whole cell assay, and favorable PK profiles in both mouse and hamster,
compound 11d was selected for further profiling in vivo. Compound 11d was demonstrated to be efficacious in both mouse and hamster
models of CDI, for which the details have been previously disclosed.15 In a hamster survival model, oral administration of 11d (50
mg/kg b.i.d.) for 7 days post-infection significantly increased survival with a median survival time of 12.3 days (Table 5).16 In contrast,
the median survival time in the vehicle group was only 2.3 days (with nine of ten animals expiring within three days). In comparison,
oral administration of the positive control vancomycin (5 mg/kg b.i.d.) for 5 days post-infection significantly increased survival with a
median survival time of 12.0 days.
Table 5. Summary of results from in vivo profiling of 11d in a hamster survival model of CDI15
Compound
Dosing
Duration
of dosing
7 days
5 days
-
Median
survival time
12.3
12.0
2.3
p-value b
Regimena
50 mpk, b.i.d.
5 mpk, b.i.d.
b.i.d.
11dc
0.002
0.006
Vancomycind
Vehiclec
a Oral (p.o.) administration for all three groups; mpk = mg/kg; b.i.d. = twice daily.
b p-value is for comparison between treatment group (11d or vancomycin) and vehicle group; data were analyzed by log-rank (Mantel-Cox) test.
c Dosing vehicle is 20% PEG400/80% (0.5% (w/v) CMC containing 0.25% (v/v) Tween 80) (v/v).
d Dosing vehicle is DI water
In conclusion, a series of TcdB inhibitors, exemplified by lead compounds 1 and 2, was further optimized with respect to plasma
stability. In general, compounds showed good affinity for TcdB. Additionally, a number of analogs demonstrated good activity in a
whole cell assay in the presence of either TcdB or TcdA and also possessed good stability in the presence of human or mouse liver
microsomes. Compound 11d exhibited a promising pharmacokinetic profile in both mouse and hamster and was profiled in vivo in
mouse and hamster models of CDI. In a hamster survival model, oral administration of 11d significantly extended life span vs. vehicle
control. As such, compound 11d and congeners represent an intriguing starting point for further development of TcdB inhibitors as a
new therapeutic approach to CDI infection.
Acknowledgments
Thank you to Albert Uveges for providing in vitro microsomal stability data. Thank you to Yan Xu and Invivotek for performing
mouse and hamster PK experiments and in vivo efficacy experiments including the hamster survival model. And thank you to Lining
Cai and Biotranex for analysis of mouse and hamster plasma samples for the generation of PK data and for performing mouse plasma
stability studies.
References and notes
1.
2.
Lessa, F. C., Mu, Y.; Bamberg, W. M. et al. N. Eng. J. Med. 2015; 273:825-834.
Zhang, S., Palazuelos-Munoz, S., Balsells, E. M., Nair, H., Chit, A., Kyaw, M. H. BMC Infectious Diseases 2016;16:447.

3.
Ananthrakrishnan, A. N., Binion, D. G. Expert Rev. Gastroent. Hepatol. 2010;4:589-600.
4.
For excellent overviews of current therapies and the pipeline of potential experimental therapies see: (a) Jarrad, A. M., Karoli, T., Blaskovich, M. A.
T., Lyras, D., Cooper, M. A. J. Med. Chem. 2015;58:5164-5185, and (b) Hopkins, R. J., Wilson, R. B. Gastroenterol. Rep. 2018;6:21-28.
(a) Merck & Co. Zinplava (bezlotoxumab) injection for intravenous use prescribing information. Whitehouse Station, NJ. 2016 Oct. (b) Wilcox M. H.,
Gerding D. N., Poxton I. R., et al. N. Engl. J. Med. 2017;376:305-317.
(a) Puri, A. W., Lupardus, P. J., Deu, E., Albrow, V. E., Garcia, K. C., Bogyo, M., Shen, A. Chem. Biol. 2010;17:1201-1211. (b) Abdeen, S. J., Swett,
R. J., Feig, A. L. ACS Chem. Biol. 2010;5:1097-1103. (c) Tam, J.; Beilhartz, G. L., Auger, A., Gupta, P., Therien, A. G., Melnyk, R. A. Chem. Biol.
2015;22:175-185.
5.
6.
7.
8.
9.
Letourneau, J. J., Stroke, I. L., Hilbert, D. W., et al. Bioorg. Med. Chem. Lett. 2018;28:756-761.
Uetrecht, J. and Trager, W. Drug Metabolism: Chemical and Enzymatic Aspects. 1st ed. New York: Taylor & Francis Ltd. 2007.
For synthetic procedures and characterization data (1H NMR and/or LC/MS) regarding the preparation of these and related structures see: Letourneau,
J. J., Cole, A. G., Marinelli, B. A., Quintero, J. G. Novel Clostridium Difficile Toxin Inhibitors. PCT Int. Appl. WO 2017/214359 A1.
10. Acid chloride 8 was prepared via reaction of 2-azido-4-chlorobenzoic acid with oxalyl chloride. For synthesis of 2-azidobenzoic acids see: Lamara,
K., Redhouse, A. D., Smalley, R. K., Thompson, J. R. Tetrahedron 1994;50:5515-5526.
11. Sugimori, T., Okawa, T., Eguchi, S., Kakehi, A.; Yashima, E.; Okamoto, Y. Tetrahedron 1998;54:7997-8008.
12. Fluorescence polarization assay for C. difficile TcdB glucosyltransferase domain UDP-glucose hydrolysis activity: See reference #9 in Letourneau, J.
J., Stroke, I. L., Hilbert, D. W., et al. Bioorg. Med. Chem. Lett. 2018;28:756-761. for a brief description of assay conditions.
13. Cell-based assay for C. difficile full-length toxin activity: Caspase 3/7 activation: See reference #15 in Letourneau, J. J., Stroke, I. L., Hilbert, D. W.,
et al. Bioorg. Med. Chem. Lett. 2018;28:756-761. for a brief description of assay conditions.
14. Best, E. L., Freeman, J., Wilcox, M. H. Gut Microbes, 2012;3:145-167.
15. Stroke, I. L., Letourneau, J. J., Miller, T. E., Xu, Y., Pechik, I., Savoly, D. R., Ma, L., Sturzenbecker, L, J., Sabalski, J., Stein, P. D., Webb, M. L.,
Hilbert, D. W., Antimicrob. Agents Chemother. 2018;62:e00107-18.
16. Sambol, S. P., Tang, J. K., Merrigan, M. M., Johnson, S., Gerding, D. N., J. Infect. Dis., 2001;183:1760-1766.
Highlights



Highly potent C. diff. toxin B (TcdB) inhibitors identified
Improved plasma stability enabled compound progression into in vivo models of CDI
Compound 11d efficacious in hamster survival model of CDI upon oral dosing
Rapid disappearance
of parent compound in
mouse plasma via amide
hydrolysis
H
H
O
O
N
N
Cl
Cl
N
N
O
O
N
N
N
N
O
CF3
N
H
N
H
O
O
N
Cl
2
11d
Retards hydrolysis of
amide in mouse plasma
TcdB IC50 = 3 nM
TcdB IC50 = 18 nM
0% Parent compund remaining
after 2 h incubation in mouse plasma
Improved stability in mouse plasma
and favorable PK profile
Efficacious in hamster survival model of CDI