Identification of lipid-like salicylic acid-based derivatives as potent and membrane-permeable PTP1B inhibitors

Article abstract of DOI:10.1016/j.bioorg.2019.103296

Developing protein tyrosine phosphatase-1B (PTP1B) inhibitors is an important strategy to treat type 2 diabetes mellitus (T2DM). Most existing ionic PTP1B inhibitors aren't of clinical useful due to their low cell-permeability, however. Herein, we introduced a series of lipid-like acid-based (salicylic acid) modules to prepare PTP1B inhibitors, and demonstrated a marked improvement of cell-permeability while maintaining excellent PTP1B inhibitory activity (e.g. compound B12D, IC₅₀ = 0.37 μM against PTP1B and P_app = 1.5 × 10^?6 cm/s). We believe that this strategy can be widely utilized to modify potent lead compounds with low cell-permeability.

Full text of DOI:10.1016/j.bioorg.2019.103296

Journal Pre-proofs
Identification of lipid-like salicylic acid-based derivatives as potent and mem-
brane-permeable PTP1B inhibitors
Liang Li, Mojdeh S. Tavallaie, Fangzhou Xie, Yu Xia, Yaoyao Liang, Faqin
Jiang, Lei Fu
PII:
S0045-2068(19)30837-5
DOI:
https://doi.org/10.1016/j.bioorg.2019.103296
YBIOO 103296
Reference:
Bioorganic Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
23 May 2019
11 September 2019
16 September 2019
Please cite this article as: L. Li, M.S. Tavallaie, F. Xie, Y. Xia, Y. Liang, F. Jiang, L. Fu, Identification of lipid-like
salicylic acid-based derivatives as potent and membrane-permeable PTP1B inhibitors, Bioorganic Chemistry
(2019), doi: https://doi.org/10.1016/j.bioorg.2019.103296
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will
undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing
this version to give early visibility of the article. 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.
© 2019 Published by Elsevier Inc.

Identification of lipid-like salicylic acid-based derivatives as
potent and membrane-permeable PTP1B inhibitors
Liang Li a, Mojdeh S. Tavallaie a, Fangzhou Xie a, Yu Xia b, Yaoyao Liang a, Faqin Jiang aꢀLei Fu a,
*
a Shanghai Key Laboratory for Molecular Engineering of Chiral Drugs, School of Pharmacy, Shanghai Jiao Tong
University (SJTU), 800 Dongchuan Road, Shanghai 200240, China
b Viva Biotech (Shanghai) Limited, Shanghai 201203, China
* To whom correspondence should be addressed. E. mail: leifu@sjtu.edu.cn (L. Fu)
Abstract
Developing protein tyrosine phosphatase-1B (PTP1B) inhibitors is an important strategy
to treat type 2 diabetes mellitus (T2DM). Most existing ionic PTP1B inhibitors aren’t of clinical
useful due to their low cell-permeability, however. Herein, we introduced a series of lipid-like
acid-based (salicylic acid) modules to prepare PTP1B inhibitors, and demonstrated a marked
improvement of cell-permeability while maintaining excellent PTP1B inhibitory activity (e.g.
compound B12D, IC50 = 0.37 μM against PTP1B and Papp = 1.5 × 10-6 cm/s). We believe that
this strategy can be widely utilized to modify potent lead compounds with low cell-permeability.
Keywords: T2DM; PTP1B inhibitors; Membrane permeability; Salicylic acid-based
derivatives
1. Introduction
Protein tyrosine phosphatase 1B (PTP1B) plays a crucial role in attenuating insulin
sensitivity by dephosphorylating the phosphor-tyrosine residues on the proteins of insulin
signal pathway1. The PTP1B knockout mice demonstrated increased insulin sensitivity and
obesity resistance compared to the wild-type2. Additionally, PTP1B inhibition was reported as
a potential target for preventing lung metastasis3 and treating Rett syndrome4. Therefore, there
is a considerable significance in developing PTP1B inhibitors for associated diseases such as
T2DM, obesity, breast cancer, and Rett syndrome 2, 5-8
.
Over the past few decades, numerous structure-based PTP1B inhibitors have been
developed9. Nevertheless, phosphorylated tyrosine (pTyr) is a charged substrate; thus, typical
PTP1B inhibitors (Fig. 1) that fit well into the protein’s active pocket shall carry a charged
pharmacophore to achieve adequate PTP1B inhibitory activity. However, the ionic

pharmacophore makes these compounds challenging to cross the cell membrane14, and
unsuitable for oral administration.
,
, ,
11 12 13
Fig. 1. Representative anionic PTP1B inhibitors10
.
To solve the membrane permeability issue, lately, many efforts have been made for the
identification of new PTP1B inhibitors by applying an uncharged pharmacophore (Fig. 2).
Nevertheless, these efforts often led to reduced PTP1B inhibitory activity.

Fig. 2. Representative nonionic PTP1B inhibitors15, 16, 17, 18
.
In this study, we demonstrated a novel strategy to produce a group of mono- or bis-anionic
amphipathic PTP1B inhibitors by designing and synthesizing a series of lipid-like salicylic
acid-based derivatives. The corresponding similarity to the amphiphilic lipids makes these
compounds capable of crossing the cell membrane adequately despite the anionic
pharmacophore. The acid groups were used to bind to the catalytic site of the protein, and the
alkane tails with various lengths were applied to generate further hydrophobic interactions.
Thus, we could reveal a rational strategy to provide potent and permeable PTP1B inhibitors.
These novel inhibitors could simulate the structure of amphiphilic lipids and demonstrate
excellent PTP1B inhibitory activity with sufficient membrane permeability. This study suggests
a new scenario for the optimization of potent lead PTP1B inhibitors with low cell permeability.
2. Results.
2. 1. The lead: the acid-based derivatives with an amphipathic core.
Focusing on the cell membranes nature, the compound A16S19 (Fig. 3), which is originally
used as a liposome substance, attracted our attention for its cell membrane crossing abilities.
We found out that it also possessed an inhibitory, however weak, activity against PTP1B (IC50

=174 μMˈPapp = 1.7 × 10-6 cm/s) which collectively could make A16S a good option as a lead
core in our designing strategy. Docking simulation studies depicted that the salicylic acid group
in A16S could bind to the catalytic site (A site20, Fig. 3) of PTP1B, while the hexadecyl group
formed multiple hydrophobic interactions with PTP1B. Despite the poor inhibitory activity, the
structural similarity to the cell membrane lipids inspired a path through identification and
designing of a new group of lipid-like PTP1B inhibitors.
A
a
Fig. 3. Molecular docking analysis of A16S bound to PTP1B (A) binding conformation of A16S with PTP1B
(different sites of PTP1B are marked with different colors respectively: red - A site - catalytic site, orange - B site-
secondary binding site, green - C site - negatively charged flat region, blue - D site - small pocket next to A site20).
(a) interactions with PTP1B. H-bonds are marked in red dashes (PDB: 2CNE).
Utilizing the structure of A16S, we designed and synthesized a series of A16S homologs
(Table 1) to investigate the structure-activity-relationships (SAR). These homologs which
contained a single salicylic acid group and a hydrophobic tail, all exhibited weak inhibitory
activity and even varying the hydrophobic tail could not exert much difference on the activity.
Subsequently, the compound A16D, designed with two groups of salicylic acids on the core,
demonstrated a graver inhibitory activity (Table 2). This finding led us to further delve into the
SARs of bis-salicylic acid-based derivatives for the identification of more potent and
membranophilic PTP1B inhibitors.

Table 1. PTP1B inhibitory activities of A16S homologs. a
Mono-salicylic acid-based scaffold
Cmpd
A6S
R’
IC50 (μM)
>200
Cmpd
B0S
R’
IC50 (μM)
>200
A12S
A16S
A18S
>200
174
152
2.9
B10S
C1S
C2S
55
>200
155
Sodium vanadate
a Experiments are conducted in triplicates.
Briefly, synthetic strategies for both mono- or bis-salicylic acid-based PTP1B inhibitors
were illustrated (Scheme 1). We initiated the synthesis by utilizing 2-hydroxy-4-methyl
benzoic acid and proceeded through a series of synthetic sequences, including protection,
brominating, alkylation, and deprotection. First, 2-hydroxy-4-methyl benzoic acid was
converted to compound 2 by using acetone as a protecting group; then compound 2 was
bromized by NBS to yield compound 3. Compound 3 reacted with the corresponding amines
and then deprotected by 1 N NaOH/ACN to yield mono and bis-salicylic acid-based inhibitors
of PTP1B (5 and 7 series). The ratio of compound 3 to amine is 1:1 and 2.5:1 for mono and bis-
salicylic acid-based products, respectively.

Scheme 1. The synthesis of mono- and bis-salicylic acid-based derivatives.
Reagents and conditions: (i) acetone, TFA, TFAA, r.t., overnightˈ58%; (ii) NBS, AIBN, reflux,
2 h, 39%; (iii) K2CO3, ACN, reflux, 2 h; (iv) NaOH, H2O, ACN, reflux, 2 h, 25%-38%; (v)
K2CO3, ACN, reflux, 48 h; (vi) NaOH/H2O, ACN, reflux, 2 h, 17%-39%.
2. 2. Length of the tail: essential benefit.
Following the discovery of A16D, various bis-salicylic acid-based derivatives were
designed and synthesized (Table 2). A short hydrophobic tail such as hexyl could not manifest
enough PTP1B inhibitory activity (A6D). Then, mounting a longer hydrophobic tail such as
dodecyl or hexadecyl demonstrated a vast improvement in the inhibitory activity by 2- and 7-
time respectively (A12D, A16D). However, the tail elongation was limit-dependent; for
example, A18D with an octadecyl tail (the most extended hydrophobic tail) demonstrated lower
inhibitory activity than A12D or A16D. We found that the replacement of four linear carbon
atoms with a benzene ring could rectify the PTP1B inhibitory activity (A12D/B8D). Following,
increasing the alkane chain length on the benzene ring from C2 to C6 doubled the inhibitory
activity (B2D/B6D), and C6 to C8 caused a 14-time leap in activity (B6D/B8D). Although, the
further increase of length slightly could enhance the inhibitory activity (B8D/B12D).

Interestingly, the elongation of the hydrophobic tail, which was initially intended to give the
molecule a lipid-like structure, could benefit the PTP1B inhibitory activity gravely.
Table 2. PTP1B inhibitory activities of bis-salicylic acid-based inhibitors. a
Bis-salicylic acid-based scaffold
Cmpd
A6D
R’
IC50 (μM˅ Cmpd
R’
IC50 (μM˅
>10
4.0
B2D
B4D
>10
A12D
A16D
A18D
4.6
6.1
1.5
5.4
B6D
B8D
0.45
C1D
C2D
>10
B10D
B12D
0.47
0.37
0.87
a Experiments are conducted in triplicates
2. 3. Membrane permeability and Selectivity over TCPTP.
Next, the permeability of bis-salicylic acid-based derivatives was assessed. We evaluated
the membrane permeability by parallel artificial membrane permeability assay (PAMPA21)
using the compounds with IC50 values ؄5 μM against PTP1B (A16D, A18D, C2D, B4D, B6D,
B8D, B10D and B12D). Permeability of Atenolol and Propranolol as the choice references in
PAMPA are also given. Clog P values explained the hydrophobicity of these compounds. As
shown in Table 3, tested compounds with two acid groups exhibited acceptable membrane
permeability (Papp = 0.6-1.5 × 10-6 cm/s). The lipid-like structure could highly enhance the

hydrophobicity and ability of cell-membrane crossing. Permeability of A18D was not detected,
and by the speculation, it might be stuck in the membrane due to excessive lipophilicity.
Simultaneously, the selectivity of these salicylic acid-based derivatives over TCPTP was
assessed. It is suggested that the inhibition of TCPTP, the homologous of PTP1B, caused
undesired side effects such as impaired glucose tolerance and attenuated glucose-stimulated
insulin secretion22. As shown in Table 3, these compounds showed a mild selectivity for
TCPTP. Overall, 3- to 10-fold PTP1B selectivity was achieved by A16D, A18D, C2D, B4D,
B6D, B8D, B10D, and 16-fold by B12D over TCPTP.
Table 3. The effects of various hydrophobic tail of salicylic acid-based inhibitors on
permeability, Clog P, PTP1B potency, and selectivity over TCPTP.
Permeability
PTP1B
TCPTP
Cmpd
Clog P a
Selectivity c
(10-6 cm/s) b IC50 (μM˅
IC50 (μM)
11.8
12.8
5.61
6.45
7.28
8.11
0.9
NDd
0.6
1.5
5.4
15.3
25.2
14.8
24.2
3.2
10.2
4.7
3.1
4.0
7.1
8.3
A16D
A18D
B4D
4.7
0.8
6.1
B6D
1.0
0.45
0.47
B8D
1.2
3.9
B10D
8.95
9.82
-0.11
1.5
NDd
1.2
0.37
0.87
-
5.8
6.7
-
15.7
7.7
-
B12D
C2D
Atenolol
Propranolol
2.76
18.1
-
-
-
a Predicted by ChemBioDraw Ultra 14.0
b Values detected by Parallel Artificial Membrane Permeability Assay
c TCPTP IC50/PTP1B IC50
d Not detected
2. 4. Docking studies analyses.
The bis-salicylic acid-based compounds demonstrated highly potent PTP1B inhibitory
activity in comparison with mono-salicylic acid-based ligands, which could implicate the
dynamic interactions with external sites of PTP1B. Thus, initially, molecular docking results of
the most potent bis-acid ligand called B12D was scrutinized (Fig. 4 A, a). As depicted, one of

the salicylic acid groups bound tightly to the catalytic site (A site) by forming three H-bonds
with residues Ala218 and Arg222 while another salicylic acid group formed one H-bond with
Lys117 and bound to D site20. Five benzene rings, including three on the molecule’s core and
two on the side chain of Tyr47 and Phe183, were spaced in the close range and could form
multiple π-π stacking interactions. Furthermore, the dodecyl chain of B12D formed other
hydrophobic interactions with Ala28, Val50, Ile220, and Met259.
A
B
b
a
Fig. 4. Molecular Docking comparisons of compounds B12D (bis-acid) and B10S (mono-acid) bound to PTP1B.
(A) Binding conformation of B12D with PTP1B. (B) Binding conformation of compound B10S with PTP1B
(different sites of PTP1B are marked with different colors respectively: red - A site - catalytic site, orange - B site-
secondary binding site, green - C site – negatively charged flat region, blue - D site – small pocket next to A site).
(a) Demonstration of interactions between compound B12D and PTP1B20. (b) Demonstration of interactions
between compound B10S and PTP1B. H-bonds are marked in red dashes.
To make a comparison with mono-acid-based compounds which demonstrated weaker
potency, simultaneously, we investigated the docking results of the most potent ligand among
our monoacid compounds: B10S (Fig. 4 B, b). The salicylic acid group bound to the catalytic
site (A site) by forming three H-bonds with residues Ala218 and Arg222, while Tyr47 and
Phe183 formed π-π stacking interactions. Decyl of B10S formed hydrophobic reactions with
mentioned above amino acids of PTP1B.

As seen, both B10S and B12D formed H-bonds with residues Ala218 and Arg222 while
Tyr47 and Phe183 were engaged in π-π stacking interactions, and they also formed multiple
hydrophobic interactions with PTP1B. The H-bond formed by the second salicylic acid with
Lys117 on the D site of PTP1B in bis-acid ligand might be quite critical for the dramatic
difference between inhibitory activities of mono- and bis-acid ligands (IC50 = 55, 0.37 μM
respectively against PTP1B). Besides, the second salicylic acid also could involve in few other
π-π stacking interactions with PTP1B. The superimposition of these two ligands
emphasized the fit conformation of bis-acid ligands lucidly (Fig. 5).
Fig. 5. Superimposition of compounds B12D (bis-acid) and B10S (mono-acid). H-bonds are marked in red dashes.
3. Discussion
Despite various efforts, such as the molecule charge manipulations and hydrophobicity
enhancements for the cell-membrane permeability improvement of PTP1B inhibitors23, the
identification of the potent and bioavailable inhibitors and specifically transforming those into
therapeutic applications is still a great challenge. In this research, we introduced a new scenario
by simulating the structure of the amphiphilic lipids and created a series of lipid-like anionic
molecules. By this strategy, we were able to rectify the potency and bioavailability of PTP1B
inhibitors extensively. These lipid-like salicylic acid-based derivatives were inspired by a
liposome substance (A16S) and could exhibit a notable PTP1B inhibitory activity and improved
bioavailability (e. g., B12D, IC50 = 0.37 μM against PTP1B andꢁPapp = 1.5 × 10-6 cm/s).
While our study has solely focused on the discovery of salicylic acid-based inhibitors of
PTP1B, the designing strategy created and brought to the surface by this study can also be
applied to other phosphatase inhibitors containing anionic moieties. This strategy supports
precise cell-membrane penetration without significant decrease of phosphatase inhibitory
activity. These compounds and the strategy behind confidently offer a potential path for the
future potent and membranophilic PTP1B inhibitors targeting T2DM.
Acknowledgments

We gratefully acknowledge Dr. Bin Hao, Dr. Xiaohui Wei, and Dr. Pei Cao of Shanghai Jiao
Tong University (China) for assistance in PAMPA evaluations.
Reference
1. Ali, M. I., et al., Deletion of Protein Tyrosine Phosphatase 1b Improves
Peripheral Insulin Resistance and Vascular Function in Obese, Leptin-Resistant Mice
via Reduced Oxidant Tone. Circul. Res. 2009, 105 (10), 1013-U172.
2. Agouni, A., et al., Liver-specific deletion of protein tyrosine phosphatase (PTP)
1B improves obesity- and pharmacologically induced endoplasmic reticulum stress.
Biochem. J 2011, 438, 369-378.
3. Julien, S. G., et al., Protein tyrosine phosphatase 1B deficiency or inhibition
delays ErbB2-induced mammary tumorigenesis and protects from lung metastasis.
Nat. Genet. 2007, 39 (3), 338-346.
4. Krishnan, N., et al., PTP1B inhibition suggests a therapeutic strategy for Rett
syndrome. J. Clin. Invest 2015, 125 (8), 3163-77.
5. Yip, S.-C.; Saha, S.; Chernoff, J., PTP1B: a double agent in metabolism and
oncogenesis. Trends Biochem. Sci. 2010, 35 (8), 442-449.
6. Verma, M.; Gupta, S. J.; Chaudhary, A.; Garg, V. K., Protein tyrosine
phosphatase 1B inhibitors as antidiabetic agents - A brief review. Bioorg. Chem.
2017, 70, 267-283.
7. Krishnan, N., et al., Targeting the disordered C terminus of PTP1B with an
allosteric inhibitor. Nat. Chem. Biol. 2014, 10, 558.
8. Tonks, N.; Krishnan, N. Method of treating Rett syndrome involves
administering effective amount of therapeutic agent which is ((2-bromo-4-(2-
carbamoyl-2-methanesulfonylaminoethyl)phenyl)difluorometh yl)phosphonic acid or
its derivative or analog to human subject. WO2016049110-A1.
9. Qian, S., et al., Recent advances in the development of protein tyrosine
phosphatase 1B inhibitors for Type 2 diabetes. Future Med. Chem. 2016, 8 (11),
1239-1258.
10. Skorey, K.; Waddleton, D.; Therien, M.; Leriche, T., Enzyme occupancy
measurement of intracellular protein tyrosine phosphatase 1B using photoaffinity
probes. Anal. Biochem. 2006, 349 (1), 49-61.
11. Van Zandt, M. C. Heteroaryl carboxylic acid derivatives as PTB-1B inhibitors
and their preparation, pharmaceutical compositions and use in the treatment of
metabolic disorders. WO2008033934A1, 2008.
12. Gray, J. L. N-Thiazolylalkyl amino acid derivatives as human protein tyrosine
phosphatase inhibitors and their preparation, and use in the treatment of angiogenesis-
related diseases. WO2008002569A2, 2008.
13. Hu, X., In silico modeling of protein tyrosine phosphatase 1B inhibitors with
cellular activity. Bioorg. Med. Chem. Lett. 2006, 16 (24), 6321-6327.
14. Mullard, A., Phosphatases start shedding their stigma of undruggability. Nat. Rev.
Drug Discov. 2018, 17, 847.

15. Wang, Y., et al., Novel chromenedione derivatives displaying inhibition of
protein tyrosine phosphatase 1B (PTP1B) from Flemingia philippinensis. Bioorg.
Med. Chem. Lett. 2016, 26 (2), 318-321.
16. Maheshwari, N., et al., Synthesis and biological evaluation of some N-(3-(1H-
tetrazol-5-yl) phenyl) acetamide derivatives as novel non-carboxylic PTP1B inhibitors
designed through bioisosteric modulation. Bioorg. Chem. 2018, 80, 145-150.
17. Meng, G., et al., Design and synthesis of new potent PTP1B inhibitors with the
skeleton of 2-substituted imino-3-substituted-5-heteroarylidene-1,3-thiazolidine-4-
one: Part I. Eur. J. Med. Chem. 2016, 122, 756-769.
18. Li, X.-Q., et al., Design, synthesis and biological evaluation of uncharged
catechol derivatives as selective inhibitors of PTP1B. Eur. J. Med. Chem. 2017, 136,
348-359.
19. Ashley, J. D., et al., Liposomal bortezomib nanoparticles via boronic ester
prodrug formulation for improved therapeutic efficacy in vivo. J. Med. Chem. 2014,
57 (12), 5282-92.
20. Ala, P. J., et al., Structural insights into the design of nonpeptidic
isothiazolidinone-containing inhibitors of protein-tyrosine phosphatase 1B. J. Biol.
Chem. 2006, 281 (49), 38013-21.
21. Liew, K. F., et al., Assessment of the Blood-Brain Barrier Permeability of
Potential Neuroprotective Aurones in Parallel Artificial Membrane Permeability
Assay and Porcine Brain Endothelial Cell Models. J. Pharm. Sci. 2017, 106 (2), 502-
510.
22. Xi, Y. N., et al., Pancreatic T cell protein-tyrosine phosphatase deficiency affects
beta cell function in mice. Diabetologia 2015, 58 (1), 122-131.
23. Maheshwari, N.; Karthikeyan, C.; Trivedi, P.; Moorthy, N., Recent Advances in
Protein Tyrosine Phosphatase 1B Targeted Drug Discovery for Type II Diabetes and
Obesity. Curr. Drug Targets 2018, 19 (5), 551-575.

1. A new series of memebranophilic PTP1B inhibitors was discovered.
2. These inhibitors showed moderate permeability despite their acid-based core.
3. A novel strategy for the optimization of PTP1B inhibitors with better permeability
was revealed.