Achieving improved permeability by hydrogen bond donor modulation in a series of MGAT2 inhibitors

Article abstract of DOI:10.1039/c3md00156c

Monoacylglycerolacetyltransferase-2 (MGAT2) is a potential target for the treatment of type II diabetes. We report here the optimisation of a series of MGAT2 inhibitors with regard to their potency and permeability. Improvements in permeability, as measured by increased flux in a Caco-2 assay, were achieved through substitution at the 9-position of the core. We propose that reduction of the NH hydrogen-bond donor strength was primarily responsible for these effects. The Royal Society of Chemistry.

Full text of DOI:10.1039/c3md00156c

MedChemComm
View Article Online
View Journal | View Issue
CONCISE ARTICLE
Achieving improved permeability by hydrogen bond
donor modulation in a series of MGAT2 inhibitors†
Cite this: Med. Chem. Commun., 2013,
4, 1305
a
James S. Scott, David J. Berry,b Hayley S. Brown,a Linda Buckett,a David S. Clarke,a
*
Kristin Goldberg,a Julian A. Hudson,a Andrew G. Leach,a Philip A. MacFaul,a
Piotr Rauboa and Graeme Robba
Monoacylglycerolacetyltransferase-2 (MGAT2) is a potential target for the treatment of type II diabetes. We
report here the optimisation of a series of MGAT2 inhibitors with regard to their potency and permeability.
Improvements in permeability, as measured by increased flux in a Caco-2 assay, were achieved through
substitution at the 9-position of the core. We propose that reduction of the NH hydrogen-bond donor
strength was primarily responsible for these effects.
Received 4th June 2013
Accepted 29th July 2013
DOI: 10.1039/c3md00156c
www.rsc.org/medchemcomm
Monoacylglycerolacetyltransferase-2 (MGAT2) is a member of the examined.5 This had resulted in improvements in permeability, as
family of enzymes that catalyses the synthesis of diacylglycerol assessed in a Caco-2 assay (apical pH¼ 6.5; basolateral pH ¼ 7.4).8
from monoacylglycerol and fatty acyl CoA.1 It is located in the This was consistent with a reduction in hydrogen bond donor
endoplasmic reticulum and is highly expressed in the small count, but had led to an unacceptable loss in potency for the three
intestine.2 Preclinical studies have shown that MGAT2 decient molecular matched pairs (Fig. 1). Notably, methylation of R4
mice show resistance to obesity, hyperglycaemia, hyper- appeared to have the least impact on Caco-2 permeability, with
cholesterolaemia and hepatic steatosis induced by a high fat moresignicantchangesbeingobservedatR1andR7.Inattempts
diet.3 Increased energy expenditure, even in the absence of a high to achieve potent yet permeable compounds, we sought to retain
fat diet, is at least partly responsible for these protective effects.4 the three donors but to modulate their strength by electronic and
Inhibition of MGAT2 has therefore attracted interest as a steric effects through substitution of the core aryl ring.9–12
potential mechanism for treatment of the metabolic syndrome.
Results from aryl ring substitution of the core are shown in
We have recently reported a novel series of MGAT2 inhibitors5 Table 1. In the case of uoro substitution (2–4), the 6-uoro
and herein report further optimization, primarily focused on resulted in a large drop in potency whereas substitution in the
increasing the membrane permeability of the compounds.
8- or 9-positions resulted in less than a two-fold change. In
Membrane permeability is a key consideration in drug terms of permeability, the 6- and 8-substitutions resulted in no
discovery, particularly in the in vivo absorption of oral drugs and change, however, 9-substitution, proximal to the amide NH, did
the ability of molecules to access intracellular targets. In result in increased permeability. In the case of methoxy
general, within a chemical series, the permeability decreases substituents (5–7), substitution at the 6-position was not toler-
with increasing polarity although other factors, such as ated with 8-substitution also resulting in a drop in potency. By
molecular weight, are important.6 In recent years, the strategy of contrast, 9-substitution resulted in an increase in both potency
keeping lipophilicity of drug candidates low has been high- and permeability. Methyl substitution (8–10) showed similar
lighted as a way of increasing the likelihood of achieving soluble SAR in terms of potency, with 6- and 8-substitution resulting in
compounds that are metabolically stable and avoid potential lower potencies relative to the un-substituted core. As before,
toxicological risks.7 One of the consequences of working in low 9-substitution resulted in increased potency and an improve-
lipophilicity space however, is that permeability is more likely to ment in terms of Ligand Lipophilicity Efficiency (LLE)13 of +1.3
become an issue.
relative to the un-substituted core. In terms of permeability, for
Systematic methylation of the three hydrogen bond donors all three methyl isomers, the data was very similar to the un-
present in our series of MGAT2 inhibitors had been previously substituted core with no improvement observed, even at the 9-
position. The pyridyl analogues (11, 12) were also examined. In
the case of the 8-pyridyl compound the potency was lowered,
however, this was accompanied by an expected drop in lip-
aAstraZeneca, Mereside, Alderley Park, Maccleseld, Cheshire, SK10 4TG, UK. E-mail:
jamie.scott@astrazeneca.com; Fax: +44 (0)1625 516667; Tel: +44 (0)1625 232567
ophilicity with the value for LLE remaining the same as the
bDurham University, School of Medicine, Pharmacy and Health, Wolfson Research
phenyl matched pair. Here the permeability was measured to be
Institute, Queen's Campus, Stockton on Tees, TS17 6BH, UK
lower as may be expected based on a lower lipophilicity
† Electronic supplementary information (ESI) available. See DOI:
compound. The 9-pyridyl compound lost a greater degree of
10.1039/c3md00156c
This journal is ª The Royal Society of Chemistry 2013
Med. Chem. Commun., 2013, 4, 1305–1311 | 1305

View Article Online
MedChemComm
Concise Article
Fig. 1 Effect of methylation on MGAT2 potency and permeability.
Table 1 Human MGAT2 potencies, lipophilicity and permeability data for core
Table 2 Human MGAT2 potencies, lipophilicity and permeability data for R9
substitution
substitution
MGAT2
Cpd Substitution IC50
LLE (pIC50 Caco-2 A–B Papp
R9
MGAT2
LLE (pIC50 Caco-2 A–B Papp
(ꢁ10ꢀ6 cm sꢀ1
)
Cpd substitution IC50
Log D7.4 ꢀ log D)
(ꢁ10ꢀ6 cm sꢀ1
)
a
a
Log D7.4 ꢀ log D)
1
2
3
4
5
6
7
8
H
6-F
8-F
9-F
6-OMe
8-OMe
9-OMe
6-Me
8-Me
9-Me
8-azaN
9-azaN
0.911
40.4
1.49
1.95
>100
7.61
0.109
13.6
13.4
0.052
10.5
>66
2.6
2.4
2.3
2.3
1.9
2.8
2.6
—
3.0
2.6
1.6
1.8
3.4
2.0
3.5
3.4
—
2.3
4.4
—
1.9
4.7
3.4
—
6
4
5
14
7
1
12
9
7
7
1
13
14
15
16
H
0.058
0.016
0.405
0.030
1.3
1.4
—
5.9
6.4
—
5
10
3
Me
CN
CF3
1.9
5.6
46
a
IC50 value from the human MGAT2 Rapidre LCMSꢀ assay.
9
substitution of the bis-amide ring (Table 3). Here the potency
enhancements seen with 9-substitution were less marked with
all substituents examined (OMe, Et, Cl) showing similar values
of LLE to the un-substituted compound 17. In terms of
permeability, an enhancement was seen for all 9-substituted
10
11
12
10
a
IC50 value from the human MGAT2 Rapidre LCMSꢀ assay.
Table 3 Human MGAT2 potencies, lipophilicity and permeability data for R9
potency than would be expected from lipophilicity alone.
Interestingly however, it showed a surprising increase in
permeability relative to the isolipohilic 8-pyridyl isomer and
retained similar permeability when compared to the more
lipophilic aryl core. Our conclusion from these data was that
modulation of the 9-position offered an opportunity to increase
permeability (F, OMe, aza-N) and potency (Me, OMe) with
methoxy offering an opportunity to do both simultaneously.
Exploration of the R9 substituent (Table 2) was carried out
contemporaneously in a 2,4-uoro aniline series which had
been previously shown5 to be considerably more potent and less
lipophilic than the 4-triuoromethyl aniline (contrast matched
pair 1 and 13). Methyl substitution resulted in an increase in
potency although this was less dramatic than seen previously
(DLLE for 10 and 1 ¼ +1.3; DLLE for 14 and 13 ¼ +0.5). Cyano
substitution 15 resulted in lower potency and permeability.
Triuoromethyl 16 retained potency and provided a dramatic
enhancement in permeability.
substitution
R9
MGAT2
LLE (pIC50 Caco-2 A–B Papp
a
Cpd substitution IC50
Log D7.4 ꢀ log D)
(ꢁ10ꢀ6 cm sꢀ1
)
17b
18
19
20
H
OMe
Et
0.010
0.010
0.004
0.009
0.007
0.062
1.7
1.8
2.1
1.9
1.7
1.8
6.2
6.2
6.3
6.1
6.5
5.4
8
22
24
35
9
Cl
21ac Me
21bc Me
14
a
IC50 value from the human MGAT2 Rapidre LCMSꢀ assay.
b
c
The effect of 9-substitution was also examined in a series
with racemic methyl, ethyl (as opposed to methyl, methyl)
Compound is the enantiomerically pure (S)-isomer. Compounds
are enantiomerically pure but the absolute stereochemistry is unknown.
1306 | Med. Chem. Commun., 2013, 4, 1305–1311
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Concise Article
MedChemComm
compounds, notably so for the chloro compound 20. In the case Comparison of the 2-uoro-4-triuoromethyl sulphonamide
of the methyl substituted compounds, these were synthesised in pair 22 and 23 with and without a 9-chloro substituent shows an
enantiomerically pure form based on the separation of an approximate three-fold increase in permeability for the chloro
intermediate by chiral HPLC. Isomer 21a (absolute stereo- substituted compound 23. In the case of the 2-methoxy-4-tri-
chemistry undetermined) was approximately nine-fold more uoromethyl sulphonamide 24 an increase in lipophilicity was
potent than enantiomeric 21b resulting in a high value for LLE observed together with a further increase in permeability. This
(6.5). In common with the data presented in Tables 1 and 2, compound had the highest Caco-2 value (Papp 58 ꢁ 10ꢀ6 cm sꢀ1
)
methyl substitution offered little to no enhancement of of any member in the series together with high potency (IC50
permeability.
7 nM). Interestingly, even low lipophilicity compounds con-
With the chloro substituent offering the largest improve- taining the 2-cyano-4-methyl sulphonamide (25) retained
ments in terms of permeability,14 this was examined in reasonable Caco-2 values (>10 ꢁ 10ꢀ6 cm sꢀ1) despite low
conjunction with other sulphonamide substituents (Table 4).15
Table 5 Log ka and Log kb values for 9-substituted compounds
Table 4 Human MGAT2 potencies, lipophilicity and permeability data for
combinations
LLE
Caco-2 A–B
ꢀ
Papp (ꢁ10ꢀ6
R9
R aryl
MGAT2 Log (pIC50
cm sꢀ1
)
a
Cpd substitution substitution IC50
D7.4 log D)
22b
23
24
H
Cl
Cl
2F,4CF3
2F,4CF3
2OMe,4CF3 0.007
2CN,4Me
2CN,4Me
0.012
0.012
1.9 6.0
2.2 5.7
3.0 5.2
0.4 7.2
0.6 6.3
9
29
58
14
11
Entry
9-Substituent
Log kb
A
Log ka
B
Log kb
C
Log ka D
1
2
3
4
5
6
H
Me
Et
Cl
CF3
CN
2.12/2.12
2.22/2.28
2.25/2.32
1.88/1.89
1.78.1.74
1.45/1.54
0.35
0.31
0.30
0.57
0.68
0.84
1.37/1.38
1.44/1.43
1.48/1.46
1.22/1.31
1.26/1.15
1.08/0.92
0.98
0.88
0.86
0.71
0.83
1.06
25ac Cl
25bc Cl
0.023
0.117
a
IC50 value from the human MGAT2 Rapidre LCMSꢀ assay.
Compound is the enantiomerically pure (S)-isomer. Compounds
b
c
are enantiomerically pure but the absolute stereochemistry is unknown.
Fig. 2 Plots of permeability against log D7.4 (A) and MGAT2 potency against permeability (B), grouped by 9-substituent and shaped by 9-substituent.
This journal is ª The Royal Society of Chemistry 2013
Med. Chem. Commun., 2013, 4, 1305–1311 | 1307

View Article Online
MedChemComm
Concise Article
pattern. A dashed line showing the value of Caco-2 A–B Papp
¼
10 ꢁ 10ꢀ6 cm sꢀ1 has been added to aid visual comparison. For
the compounds with no substitution at the 9-position, only 2/41
(5%) has a Papp value >10 ꢁ 10ꢀ6 cm sꢀ1. For the 9-methyl set,
12/47 (26%) have a Papp value >10 ꢁ 10ꢀ6 cm sꢀ1 and when only
compounds with a log D7.4 > 1.5 are considered, this proportion
increases to 12/25 (48%). All of the 15 compounds with a
methoxy, ethyl or chloro substituent show Papp value >10 ꢁ 10ꢀ6
cm sꢀ1, even for those chloro substituted compounds with low
lipophilicity (log D7.4 0.5). Fig. 2B shows MGAT2 potency plotted
against permeability with a target area (pIC50 > 8; Papp > 10 ꢁ
10ꢀ6 cm sꢀ1 shaded in grey). For the un-substituted series only
one compound of the set of 54 satisfy this criteria. For the
9-methyl compounds, 4/44 (10%) are both potent and perme-
able. When the 9-OMe/Et/Cl set are considered, 8/15 (53%)
achieve the target, highlighting the importance of 9-substitu-
tion in achieving potent and permeable compounds in this
series.
Fig. 3 Bivariate plot of Caco-2 Papp (log scale) and log ka (D) showing a corre-
lation between the two. Open, blue points are from Table 2; closed, green points
are from Table 3. R2 ¼ 0.77; RMSE ¼ 0.46; probability of such a relationship
happening by chance is 0.0018.
In an attempt to understand the inuence of the 9-substit-
uent on the electronics of the core, the strength of hydrogen-
bond donors (log ka) and hydrogen-bond acceptors (log kb)
labelled A–D were calculated using quantum mechanics
(Gaussian 09 using B3LYP functional and 6-31G* basis set)16.
The labelling and values for the 9-methyl substituted core is
shown in Table 5. Each of the carbonyls has two values corre-
sponding to the orbital lobes on the oxygen lone pairs. These
log D7.4 values (ꢂ0.5). As in the case of compound 21, a clear
difference in potency between the separated enantiomers was
observed.
A graphical overview of the full dataset for this series is
shown in Fig. 2, of which the compounds described in this
paper are a subset. Fig. 2A shows permeability plotted against
measured log D7.4 and trellised according to the 9-substitution
Fig. 4 (a) Strong H-bonding interactions in 26. These bonds support a chain structure down the b-axis. (b) 26 viewed down the b-axis (c) 27 viewed down the b-axis
(d) 28 viewed down the b-axis.
1308 | Med. Chem. Commun., 2013, 4, 1305–1311
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Concise Article
MedChemComm
calculations account for both through-bond and through-space have been put on a log-scale so as to be normally distributed).
electronic effects of the 9-substituent. The relationship with the strength of the adjacent hydrogen-
The calculated hydrogen-bond donor and acceptor strengths bond donor (D) clearly shows a strong correlation with log Caco-2
for a range of 9-substituents corresponding to the compounds Papp as shown in Fig. 3. The probability (P) of this correlation
from Tables 2 and 3 are shown in Table 5. These compounds arising by chance is 0.0018, whilst for all other calculated
have xed aryl substitution (2,4-diuoro) and differ only in the hydrogen-bond strengths P $ 0.5. The correlation of perme-
substitution of the bis-amide ring (where ethyl or methyl ability with log ka (D) is also better than for log D7.4 (P ¼ 0.024).
substitution result in equivalent permeability). Simple bivariate This result helps rationalize the observed permeability: Higher
statistics were used to elucidate any relationships between Caco- permeability is associated with lower hydrogen-bond donor
2 Papp and calculated hydrogen-bond strength (Caco-2 Papp values strength (Cl, CF3), and low permeability is associated with higher
hydrogen-bond donor strength (H, CN).
In order to understand any effects on the solid state of these
compounds, small molecule X-ray crystal structures were
obtained of several members of this series. Although there are
differences in symmetry and packing density across the un-
substituted compounds (26–28), similar packing motifs around
a strong H-bonded chain between the amides of the cores with a
sulphonamide backbone are consistently observed in all three
structures (Fig. 4). In the 9-chloro analogue (23), similar amide
interactions between the cores are retained, however, it is
noteworthy that the sulphonamide H-bonding motifs are
completely absent in this structure (Fig. 5).
Compounds were synthesized using a modied version of
the previously described route5 whereby cyclisations were
carried out on the amino ester intermediates of type 29 (Scheme
1). This circumvented difficulties associated with isolation of
the previously employed amino acids. The amino esters were
synthesized from amide coupling of 2-amino benzoic acids 30,
regioselective ring opening of isatoic anhydrides 31 or amide
coupling of 2-nitro benzoic acids 32 followed by reduction of the
nitro group 33. Cyclisation to form the core 34 was carried out
using acidic conditions and the various substitutions (R6, R8,
R9) were generally well tolerated with the exception of (R6 ¼ F)
which proceeded in low yield (8%). Acetic acid was initially used
in the cyclisation, however, it was noted that acetylation of the
aniline was a signicant by-product in some cases. This
problem was overcome by switching to the more sterically
hindered pivalic acid resulting in cleaner reaction proles and
higher isolated yields. Treatment with chlorosulphonic acid
Fig. 5 (a) Strong H-bonding interactions in 23. These bonds support a chain struc-
ture down the b-axis with weaker interactions supporting layering down the a-axis.
generated the corresponding sulphonyl chlorides 35 with
This is viewed along an intersection of the b/c-axis. (b) 23 packing down a-axis.
Scheme 1 General synthetic approaches. Reagents: (a) H2NCR3(Me)COOMe, HATU, (iPr)2NEt, DMF, 76–100%; (b) H2NCR3(Me)COOMe, (iPr)2NEt, DMF, 60 ꢃC, 55–
70%; (c) H2, Pd/C, EtOH, 93–100%; (d) MeCOOH or tBuCOOH, mW, 180–200 ꢃC, 8–99%; (e) ClSO3H, 65–90 ꢃC, 32–100%; (f) ArNH2, pyridine, CH2Cl2, 25–100 ꢃC, 13–
88%; (g) Pd2(dba)3, XPhos, Zn, Zn(CN)2, DMA, mW, 120 ꢃC, 2 h, 82%.
This journal is ª The Royal Society of Chemistry 2013
Med. Chem. Commun., 2013, 4, 1305–1311 | 1309

View Article Online
MedChemComm
Concise Article
Scheme 2 Synthesis of aza-analogues 11 & 12. Reagents: (a) H2NC(Me2)COOMe, HATU, (iPr)2NEt, DMF, 20 ꢃC, 45 min, 90%; (b) TFA, CH2Cl2, 20 ꢃC, 3 h, 100%; (c)
tBuCOOH, mW, 200 ꢃC, 3 h, 35%; (d) BnSH, tBuOK, DMF, 80 ꢃC, 3 h, 94%; (e) (i) NCS, AcOH, H2O, 20 ꢃC, 10 min; (ii) 4-CF3C6H4NH2, pyridine, 20 ꢃC, 3 h, 22%; (f) ClSO3H, 65
ꢃC, 18 h; (g) 4-CF3C6H4NH2, pyridine, 20 ꢃC, 3 days, 9% over 2 steps; (h) H2NC(Me2)COOMe, HATU, (iPr)2NEt, DMF, 20 ꢃC, 18 h, 29%; (i) tBuCOOH, 200 ꢃC, 28 h, 56%.
Table 6 Physical properties of compound 20
excellent regioselectivity for the 7-position regardless of core
substitution (R6, R8, R9) in all cases (presumably directed by
Aqueous
solubilitya (mM) (% free)
Mouse PPBb Rat PPBb Human
hERG
the aniline amide N1). In the case of R9 ¼ CN, this compound
was made from the fully elaborated sulphonamide with R9 ¼ Br
and subjected to palladium catalysed cyanation conditions
giving compound 15 in good yield.
PPBb (% free) IC50 (mM)
c
(% free)
150
3.7
3.0
1.7
>100
a
Solubility of compounds in aqueous phosphate buffer at pH 7.4 aer
24 hours at 25 ꢃC (mM). b PPB was assessed by equilibrium dialysis in the
appropriate species plasma at 37 ꢃC. Free and bound concentrations
The enantiomerically pure compounds 17 and 22 (R9 ¼ H)
were obtained by chiral HPLC separation of the cyclised core
prior to sulphonylation. Treatment with chlorosulphonic acid
and sulphonamide coupling gave the desired products with the
absolute conguration being identied as (3S) based on X-ray
crystallography. Compounds 21a and 21b (R9 ¼ H) were made
in a similar fashion although the absolute conguration was
not determined. In the case of compounds 25a and 25b these
were made by chiral separation of racemic 25 and the absolute
conguration was not determined.
c
were quantied by LC-UVMS. Inhibition of hERG channel IC50 (mM)
in a electrophysiology (IonWorksꢁ) assay.
Table 7 Pharmacokinetic parameters for selected compound 20a
Clp (mL
Vdss
IV half- Oral half- Bioavailability
Species minꢀ1 kgꢀ1
)
(L kgꢀ1
)
life (h) life (h)
(%)
For the aza-analogues it was found that the sulphonylations
would not occur on the fully elaborated core (presumably due to
the electron deciency of the pyridyl ring) and so these were
made according to the procedures outlined in Scheme 2. For the
8-aza compound 11, the core was constructed from the 2-
chloropyridine 36 via amide coupling, Boc deprotection to
aniline 38 and cyclisation using pivalic acid to the 8-aza core 39.
The chloro substituent was then displaced with a benzylthiol to
give 40 and then oxidized to a sulphonyl chloride and quenched
with 4-triuoromethyl aniline to generate sulphonamide 11. For
the 9-aza compound 12, an alternative approach was utilized
whereby the 2-anilinopyridine 41 was regioselectively converted
to the sulphonyl chloride 42 and then coupled in low yield to
form the sulphonamide 43. Amide coupling to form anili-
noester 44 and cyclisation under pivalic acid conditions gave
sulphonamide 12.
Mouse 29
0.5
0.4
0.2
1.3
1.2
2.4
74
69
Rat
4.1
a
Dosed PO at 3 mg kgꢀ1 (rat) and 50 mg kgꢀ1 (mouse) in suspension
(HPMC/Tween) and IV (1 mg kgꢀ1) in 5% DMSO: 95% hydroxypropyl
b cyclodextrin. Data is mean of two animals in a single experiment.
sulphonamide NH was 5.9. The aqueous solubility as measured
on crystalline material was 150 mM.
The oral pharmacokinetic prole of compound 20 was
examined in rat at low dose (3 mg kgꢀ1) and mouse at high dose
(50 mg kgꢀ1) with the results are summarised in Table 7. From
IV data the compound showed moderate clearance in mouse
and low clearance in rat together with a low volume of distri-
bution in both species. Good bioavailabilities (>60%) were
observed in both species.
As a representative example of this series, compound 20 was
proled further (Table 6). The compound showed no signicant
inhibition against four isoforms of the cytochrome P450
Conclusion
enzymes (CYP1A2, CYP2C9, CYP2D6 and CYP3A4 all IC50
>
In summary we have reported the optimization of a series of
30 mM) and only weak activity against CYP2C19 (IC50 27 mM). No MGAT2 inhibitors to give potent compounds such as 20 that
inhibition of the hERG ion channel (IC50 > 100 mM) was have favourable physical and pharmacokinetic properties. In
observed. Plasma protein binding showed reasonable free levels particular the permeability, as measured by increased ux in a
(1.7–3.7% free across species) in line with the low lipophilicity Caco-2 assay, has been improved through substitution at the
of the compound (log D7.4 1.9). The measured pKa of the 9-position of the core. This was achieved without recourse to
1310 | Med. Chem. Commun., 2013, 4, 1305–1311
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Concise Article
MedChemComm
increasing the lipophilicity of the compounds but rather
through modulation of the hydrogen bond donor strength of an
amide NH.
M. Rooseboom and A.-L. Ungell, Eur. J. Pharm. Sci., 2008,
35, 383.
9 Increasing permeability through alteration of hydrogen
bond strength through addition of proximal groups has
been described in the literature. For a recent review see
P. V. Desai, T. J. Raub and M.-J. Blanco, Bioorg. Med. Chem.
Lett., 2012, 22, 6540.
Acknowledgements
Scott Boyd, Ben Chappell, Paul Kemmitt and Paul Schoeld are
thanked for synthetic contributions. Michael Tonge is 10 C. D. Cox, M. J. Breslin, D. B. Whitman, P. J. Coleman,
acknowledged for generating enzyme inhibition data. Per
Svensson and Anne Ertan are thanked for crystallography
contributions. We also thank Nicholas Newcombe and Andrew
Turnbull for project leadership.
R. M. Garbaccio, M. E. Fraley, M. M. Zrada, C. A. Buser,
E. S. Walsh, K. Hamilton, R. B. Lobell, W. Tao,
M. T. Abrams, V. J. South, H. E. Huber, N. E. Kohl and
G. D. Hartman, Bioorg. Med. Chem. Lett., 2007, 17, 2697.
11 J. F. Levesque, K. Bleasby, A. Chefson, A. Chen, D. Dube,
Y. Ducharme, P. A. Fournier, S. Gagne, M. Gallant,
E. Grimm, M. Hafey, Y. Han, R. Houle, P. Lacombe,
S. Laliberte, D. MacDonald, B. Mackay, R. Papp and
R. Tschirret-Guth, Bioorg. Med. Chem. Lett., 2011, 21, 5547.
12 S. B. Ra, B. R. Hearn, P. Vedantham, M. P. Jacobson and
A. R. Renslo, J. Med. Chem., 2012, 55, 3163.
References
1 C.-L. E. Yen and R. V. Farese Jr, J. Biol. Chem., 2003, 278,
18532.
2 R. M. Bell and R. A. Coleman, Annu. Rev. Biochem., 1980, 49,
459.
13 P. D. Leeson and B. Springthorpe, Nat. Rev. Drug Discovery,
2007, 6, 881.
3 C.-L. E. Yen, M.-L. Cheong, C. Grueter, P. Zhou, J. Moriwaki,
J. S. Wong, B. Hubbard, S. Mamour and R. V. Farese Jr, Nat.
Med., 2009, 15, 442.
4 D. W. Nelson, Y. Gao, N. M. Spencer, T. Banh and
C.-L. E. Yen, J. Lipid Res., 2011, 52, 1723.
14 Other groups such as 9-F (Table 1) and 9-CF3 (Table 2)
showed permeability increases in other sets of matched
pairs but were viewed on balance as being less optimal
than the Cl. In the case of the F, the permeability rise was
less pronounced and the CF3 added both lipophilicity and
molecular weight. The exact matched pairs with 9-Cl were
not made as part of this program of work.
¨
5 J. G. Barlind, L. K. Buckett, S. G. Crosby, O. Davidsson,
¨
H. Emtena, A. Ertan, U. Jurva, M. Lemurell, K. Nilsson,
˚
G. O'Mahony, A. U. Petersson, A. Redzic, F. Wagberg and
Z.-Q. Yuan, Bioorg. Med. Chem. Lett., 2013, 23, 2721.
6 M. J. Waring, Bioorg. Med. Chem. Lett., 2009, 19, 2844.
7 M. J. Waring, Expert Opin. Drug Discovery, 2010, 5, 235.
8 R. Hayeshia, C. Hilgendorf, P. Artursson, P. Augustijns,
B. Brodin, P. Dehertogh, K. Fisher, L. Fossati,
E. Hovenkamp, T. Korjamo, C. Masungi, N. Maubon,
15 Structure–activity-relationship
information
from
sulphonamide library work revealed that 2,4-disubstitution
was a preferred motif in this series and that electron
withdrawing substituents (particularly in the 2-position)
were favourable for potency.
¨
¨
16 (a) P. Kenny, J. Chem. Inf. Model., 2009, 49, 1234; (b)
P. W. Kenny, J. Chem. Soc., Perkin Trans. 2, 1994, 199.
¨
R. Mols, A. Mullertz, J. Monkkonen, C. O'Driscoll,
H. M. Oppers-Tiemissen, E. G. B. Ragnarsson,
This journal is ª The Royal Society of Chemistry 2013
Med. Chem. Commun., 2013, 4, 1305–1311 | 1311