3-Phenylpropanoic acid-based phosphotyrosine (pTyr) mimetics: Hit evolution to a novel orally active protein tyrosine phosphatase 1B (PTP1B) inhibitor

Article abstract of DOI:10.1002/cmdc.201400007

Protein tyrosine phosphatase 1B (PTP1B) is a promising therapeutic target for type 2 diabetes. Herein, we report the evolution of a previously identified 3-phenylpropanoic acid-based PTP1B inhibitor to an orally active lead compound. A series of 3-phenylp

Full text of DOI:10.1002/cmdc.201400007

CHEMMEDCHEM
COMMUNICATIONS
DOI: 10.1002/cmdc.201400007
3-Phenylpropanoic Acid-Based Phosphotyrosine (pTyr)
Mimetics: Hit Evolution to a Novel Orally Active Protein
Tyrosine Phosphatase 1B (PTP1B) Inhibitor
Yan-Bo Tang,[a] Jun-Zheng Liu,[a] Shu-En Zhang,[a] Xin Du,[a] Feilin Nie,[a] Jin-Ying Tian,[b]
Fei Ye,[b] Kai Huang,[a] Jin-Ping Hu,[a] Yan Li,[a] and Zhiyan Xiao*[a]
Protein tyrosine phosphatase 1B (PTP1B) is a promising thera-
peutic target for type 2 diabetes. Herein, we report the evolu-
tion of a previously identified 3-phenylpropanoic acid-based
PTP1B inhibitor to an orally active lead compound. A series of
3-phenylpropanoic acid-based PTP1B inhibitors were synthe-
sized, and three of them, 3-(4-(9H-carbazol-9-yl)phenyl)-5-(3,5-
di-tert-butyl-4-methoxyphenyl)-5-oxopentanoic acid (9), 3-(4-
(9H-carbazol-9-yl)phenyl)-5-(4’-bromo-[1,1’-biphenyl]-4-yl)-5-ox-
opentanoic acid (10) and 3-(4-(9H-carbazol-9-yl)-2-fluorophen-
yl)-5-(4-cyclohexylphenyl)-5-oxopentanoic acid (16), showed
IC50 values at sub-micromolar level. Further in vivo evaluation
indicated the sodium salt of 9 not only exhibited significant in-
sulin-sensitizing and hypoglycemia effects, but also decreased
the serum levels of triglyceride and total cholesterol in high-
fat-diet-induced insulin resistance model mice. Preliminary
in vivo pharmacokinetic studies on the sodium salt of 9 re-
vealed its pharmacokinetic profile after oral administration in
rats. These results provide proof-of-concept for the dual effects
of PTP1B inhibitors on both glucose and lipid metabolisms.
PTP1B inhibitors, namely ertiprotafib,[5] trodusquemine,[6] and
JTT-551,[7] have hitherto reached clinical trials.
Ertiprotafib is a phosphotyrosine (pTyr) mimetic that binds
to the catalytic site of PTP1B. It is a mono-carboxylic acid pTyr
mimetic characterized by the extensive presence of hydropho-
bic fragments, which likely increases cellular permeability. In
addition, the benzyl group in ertiprotafib is assumed to inter-
act with a less conserved subpocket near the active site, which
could be a key structural feature to improve affinity as well as
selectivity. Although the clinical trial on ertiprotafib was termi-
nated in phase II, ertiprotafib still offers a good chemical tem-
plate for developing new PTP1B inhibitors with improved se-
lectivity and oral availability.[5] In our previous study, a series of
novel PTP1B inhibitors with various pTyr mimetic moieties
have been obtained via pharmacophore-oriented scaffold hop-
ping from ertiprotafib.[8] Hit structure I (Scheme 1) is a novel
PTP1B inhibitor recently identified by us that bears a 3-phenyl-
propanoic acid moiety as a pTyr mimetic and exhibits an IC50
value of 10.2 mm against human recombinant PTP1B.[8] Here,
Protein tyrosine phosphatases (PTPs) are key regulators in the
reversible tyrosine phosphorylation of a variety of functional
proteins.[1] Among the PTPs, protein tyrosine phosphatase 1B
(PTP1B) has emerged as an attractive target for the treatment
of diabetes and obesity, owing to its important role in both in-
sulin and leptin signal transduction processes.[2–4] However,
conserved and cationic nature of the catalytic site of PTP1B
makes it difficult to achieve selectivity and in vivo activity, and
presents formidable challenges in developing PTP1B inhibitors
into therapeutic drugs. As a result, only three small-molecule
[a] Dr. Y.-B. Tang,+ J.-Z. Liu,+ Dr. S.-E. Zhang, X. Du, F. Nie, Dr. K. Huang,
Dr. J.-P. Hu, Prof. Y. Li, Prof. Z. Xiao
Beijing Key Laboratory of Active Substance Discovery & Druggability Evalu-
ation, Institute of Materia Medica, Chinese Academy of Medical Sciences &
Peking Union Medical College
1 Xiannongtan Street, Beijing 100050 (China)
E-mail: xiaoz@imm.ac.cn
[b] J.-Y. Tian, Prof. F. Ye
Beijing Key Laboratory of New Drug Mechanisms & Pharmacological Evalu-
ation Study, Institute of Materia Medica, Chinese Academy of Medical Scien-
ces & Peking Union Medical College
1 Xiannongtan Street, Beijing 100050 (China)
[+] These authors contribute equally.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201400007.
Scheme 1. Hit structure (I) and general formula (II) of target compounds.
ChemMedChem 2014, 9, 918 – 921 918
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

CHEMMEDCHEM
COMMUNICATIONS
www.chemmedchem.org
10 mm. Most compounds showed significant inhibition against
PTP1B (Table 1), and twelve compounds were further tested to
establish their IC50 values against PTP1B. Among them, ten
compounds were more potent than hit compound I in PTP1B
inhibition, and three compounds were even superior to refer-
ence compound 20, which is one of the most potent PTP1B in-
hibitors in the compound class of ertiprotafib analogues
(Table 1).
A quick glance at the results shown in Table 1 suggested
that bulky substituents on phenyl ring A could be favourable
for PTP1B inhibition (5, 7–12, 14, 16–17 vs I). However, replac-
ing phenyl ring A with 2-methylthiophene led to a substantial
loss of activity (15 vs I). The introduction of a fluorine atom on
phenyl ring B significantly improved the PTP1B inhibitory activ-
ity (7 vs 16; 13 vs 17). Although the limited number of target
compounds failed to provide conclusive structure–activity rela-
tionships, the discovery of novel potent PTP1B inhibitors
makes it possible to further investigate this prototype class for
in vivo activities.
Compound 9 was selected for further in vivo antidiabetic
evaluation. To improve its water solubility and facilitate oral ad-
ministration, compound 9 was converted into sodium salt 21
by treating with 4m sodium hydroxide in 1,4-dioxane. Com-
pound 21 showed in vitro PTP1B inhibitory activity (IC50 =
0.40 mm) comparable to that of compound 9. To generate an
insulin resistance model (IRM), obesity was induced in mice
using a high-fat diet (HFD), and in vivo evaluation of com-
Scheme 2. Synthesis of target compounds. Reagents and conditions: a) Cu,
K2CO3, PhNO2, N2, reflux, 6 h, 42–80%; b) NaOH, C2H5OH, o/n, RT, 51–94%;
c) diethyl malonate, KOH, EtOH, ultrasonication, 3 h; or 1. dimethyl malonate,
KOH, EtOH, CH2Cl2, reflux, 6 h, 65–85%; d) KOH, H2O, acetone, o/n, RT; 2. 5%
HCl, 90–97%; e) 6m HCl, AcOH, reflux, 6–10 h, 35–99%.
we report the evolution of this hit structure
to a novel lead compound with potent
PTP1B inhibitory activity as well as significant
Table 1. Inhibitory activity of target compounds against human recombinant protein tyrosine
phosphatase 1B (PTP1B).[a]
in vivo effects on both insulin resistance and
lipid abnormalities.
Compd Ra
R
Inhibition [%] IC50
Compd Ra
R
Inhibition [%] IC50
Target compounds with the general formu-
la of II (Scheme 1) were designed and syn-
thesized. The phenyl ring A in I was substi-
tuted with various functional groups and
even replaced with heteroaromatic rings to
examine the effects of this molecular area on
PTP1B inhibition. Fluoro substitution was also
introduced to the phenyl ring B to explore its
possibility to potentiate PTP1B inhibition.
The synthesis of target compounds 1–17
started from commercially available carbazole
18 and 4-bromobenzaldehydes (19). As
shown in Scheme 2, all target compounds
were generally prepared through successive
Ullmann reaction (step a), aldol condensation
(step b), Michael addition (step c), hydrolysis
(step d) and decarboxylation (step e).
100 mm 10 mm [mm]
100 mm 10 mm [mm]
1
H 104.3 ND
18.5
ND
11
H ND
97.6 1.28
99.6 3.79
2
3
H 58.0
H 89.3
NA
NA
12
13
H ND
H ND
ND
65.3 ND
96.2 1.51
4
5
6
H 107.4 ND
15.5
14
15
16
H ND
H ND
H 94.9
H 91.7
68.4 8.60
56.0 ND
NA
ND
F
F
ND
ND
100.5 0.34
97.4 2.06
7
8
H ND
H ND
94.0 1.26
96.0 1.03
17
I
H 106.1 ND
10.2
9
H ND
H ND
99.0 0.43
96.8 0.42
20
– 100.0 99.5 0.87
All target compounds were evaluated
against human recombinant PTP1B with pro-
tocols described previously.[9] Hit compound
I[8] and compound 20,[10] which is an ana-
logue of ertiprotafib, were used as reference
compounds. Initial screening was performed
to measure the percentage inhibition against
PTP1B at a concentration of 100 mm and/or
10
[a] For full structure, see Scheme 1. ND: not determined; NA: not active—the percentage inhibi-
tion was less than 50% at the tested concentration. Hit I and compound 20 were used as refer-
ence compounds. The IC50 value was interpolated from dose–response data with at least eight
different concentrations. Each test was performed in triplicate. Data represent the average of
triplicate measurements.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 918 – 921 919

CHEMMEDCHEM
COMMUNICATIONS
www.chemmedchem.org
Figure 1. Compound 21 improved insulin resistance in insulin resistant model (IRM) mice (n=8). a) Changes in blood glucose determined by an intraperito-
neal glucose tolerance test (IPGTT); b) AUC values in IPGTT results; c) Changes in blood glucose determined by an insulin tolerance test (ITT); d) AUC values
in ITT results. ##, p<0.01 vs control mice (con); **, p<0.01 vs IRM mice.
pound 21 was performed as previously described.[9] Com-
pound 21 was orally administrated at a dose of 50 mgkgÀ1 for
8–14 days, and rosiglitazone was given orally at 10 mgkgÀ1 as
a positive control. In the intraperitoneal glucose tolerance test
(IPGTT), impaired glucose tolerance in IRM mice was ameliorat-
ed by treatment with 21 as indicated by both changes in
blood glucose levels (Figure 1a) and the area under glucose–
time curve (AUC) (Figure 1b) after glucose loading. Similarly,
insulin intolerance was also improved by treatment with 21 in
the insulin tolerance test (ITT) (Figure 1c,d). These results were
similar to those observed in the rosiglitazone-treated group.
Compound 21 was also tested for its effects on fasting
blood glucose levels and body weight in IRM mice. The blood
glucose levels in fasting mice were significantly lower in both
rosiglitazone- and 21-treated groups than in the IRM group
(Figure 2a). Although the food intake by mice in all groups
had no obvious variation, the body weight of IRM mice in the
21-treated group showed a decreasing trend, and in contrast,
body weight gain was observed in both IRM and rosiglitazone-
treated groups (Figure 2b), indicating that compound 21 has
a unique pharmacological profile different from that of rosigli-
tazone.
only modulate glucose metabolism as expected, but also affect
lipid abnormality, which might be an advantage of this com-
pound class as compared with other available antidiabetic
agents.
Compound 21 was then further tested for its effects on the
levels of serum triglyceride (TG) and total cholesterol (TC)
(Figure 3). Compared to the IRM group, the levels of serum TG
and TC were significantly decreased in the 21-treated group.
Such results indicate that the PTP1B inhibitor 21 could not
Figure 2. Effects of compound 21 on fasting blood glucose and body
weight of insulin resistant model (IRM) mice (n=8). a) Effects on fasting
blood glucose; b) Changes in body weight. #, p<0.05; ###, p<0.001 vs
control mice (con); **, p<0.01; ***, p<0.001 vs IRM mice.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 918 – 921 920

CHEMMEDCHEM
COMMUNICATIONS
www.chemmedchem.org
In conclusion, a series of 3-phenylpropanoic acid-based
phosphotyrosine (pTyr) mimetics were synthesized and evalu-
ated as protein tyrosine phosphatase 1B (PTP1B) inhibitors.
Compound 21 was identified as a novel orally active PTP1B in-
hibitor, which not only exhibited significant insulin-sensitizing
and hypoglycemia effects, but also decreased the serum levels
of triglyceride and total cholesterol in IRM mice. The results
could be regarded as proof-of-concept in insulin-resistant ani-
mals for dual effects of PTP1B inhibitors on both glucose and
lipid metabolisms. Further structural optimization of the lead
compounds is undergoing.
Acknowledgements
This research work was financially supported by the National
Natural Science Foundation of China (20972192), the Beijing Nat-
ural Science Foundation (7102117 to Z.X.), and the Chinese Na-
tional S & T Major Project (2012ZX09103-101-063 to F.Y.).
Figure 3. Effects of compound 21 on the metabolic profiles induced by
high-fat diet in mice (n=8). a) Effects on serum triglyceride (TG); b) Effects
on total cholesterol (TC). ###, p<0.001 vs control mice (con); *, p<0.05; ***,
p<0.001 vs IRM mice.
Keywords: 3-phenylpropanoic acids · inhibitors · medicinal
chemistry · protein tyrosine phosphatase 1B · type 2 diabetes
[1] A. J. Barr, Future Med. Chem. 2010, 2, 1563–1576.
[2] S. Zhang, Z.-Y. Zhang, Drug Discovery Today 2007, 12, 373–381.
[3] A. P. Combs, J. Med. Chem. 2010, 53, 2333–2344.
[4] S. Thareja, S. Aggarwal, T. R. Bhardwaj, M. Kumar, Med. Res. Rev. 2012,
32, 459–517.
Compound 21 was further subjected to pharmacokinetic
study in rats by following the protocols described previously.[11]
After oral administration at the dose of 50 mgkgÀ1, compound
21 could be readily detected in blood and sustained at a sub-
stantial plasma level within 24 hours after administrated. The
major pharmacokinetic parameters of compound 21 were
listed in Table 2.
[5] Z.-Y. Zhang, S.-Y. Lee, Expert Opin. Invest. Drugs 2003, 12, 223–233.
[6] K. A. Lantz, S. G. Emeigh Hart, S. L. Planey, M. F. Roitman, I. A. Ruiz-
White, H. R. Wolfe, M. P. McLane, Obesity 2010, 18, 1516–1523.
[7] S. Fukuda, T. Ohta, S. Sakata, H. Morinaga, M. Ito, Y. Nakagawa, M.
Tanaka, M. Matsushita, Diabetes Obes. Metab. 2010, 12, 299–306.
[8] J.-Z. Liu, S.-E. Zhang, F. Nie, Y. Yang, Y.-B. Tang, W. Yin, J.-Y. Tian, F. Ye, Z.
Xiao, Bioorg. Med. Chem. Lett. 2013, 23, 6217–6222.
[9] para-Nitrophenyl phosphates (pNPPs) were used as the substrates for
the measurement of PTP1B activity. All of the animal procedures were
approved by the Animal Care and Use Committee in Beijing, China. All
the measurements were performed by following the experimental pro-
tocols described in: Y.-M. Ma, R.-Y. Tao, Q. Liu, J. Li, J.-Y. Tian, X.-L.
Zhang, Z.-Y. Xiao, F. Ye, Mol. Cell. Biochem. 2011, 357, 65–72.
[10] J. Wrobel, J. Sredy, C. Moxham, A. Dietrich, Z. Li, D. R. Sawicki, L. Seestal-
ler, L. Wu, A. Katz, D. Sullivan, C. Tio, Z.-Y. Zhang, J. Med. Chem. 1999,
42, 3199–3202.
[11] Male Sprague Dawley rats (210Æ21 g, n=3) were used for the pharma-
cokinetic study in vivo. All experimental procedures were in compliance
with the guidelines of China for animal care. Compound 21
(50 mgkgÀ1) was orally administrated, and plasma samples were collect-
ed at 0.08, 0.25, 0.5, 1, 2, 4, 6, 8, 12 and 24 h after dosing. For detailed
experimental protocols, please refer to: F. You, J. Hu, X. Li, Y. Li, Cancer
Chemother. Pharmacol. 2013, 71, 1453–1462.
Table 2. Phamrcokinetic parameters of 21 (50 mgkgÀ1) determined in
rats after oral administration (n=3).[a]
Parameter [unit]
Value
AUC(0–24)
MRT(0–24)
t1/2(z)
CLz/F
Cmax
[mgLÀ1 hÀ1
]
37239.79Æ4113.14
6.65Æ0.60
[h]
[h]
3.45Æ0.37
[LhÀ1 kgÀ1
]
1.34Æ0.14
[mgLÀ1
]
5014.01Æ1499.01
[a] Abbreviations: area under the concentration–time curve (AUC); mean
residence time (MRT); half-life (t1/2); plasma clearance (CLz/F); maximum
concentration (Cmax).
Received: January 2, 2014
Published online on March 18, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 918 – 921 921