Design, synthesis, and evaluation of novel L-phenylglycine derivatives as potential PPARγ lead compounds

Article abstract of DOI:10.1016/j.bmc.2018.07.005

In accordance with the structural characteristics of thiazolidinedione drugs and highly bioactive tyrosine derivatives, we tentatively designed the L-phenylglycine derivatives TM1 and TM2 based on basic principles of drug design and then synthesized them.

Full text of DOI:10.1016/j.bmc.2018.07.005

Accepted Manuscript
Design, synthesis, and evaluation of novel L-phenylglycine derivatives as po-
tential PPARγ lead compounds
Liu Jinyu, Su Xiaoyan, Li Huachong, Fan Li, Li Yuanyuan, Tang Xuemei, Yan
Jufang, Chen xin, Chen Feifei, Liu Jie, Yang Dacheng
PII:
S0968-0896(17)32494-X
https://doi.org/10.1016/j.bmc.2018.07.005
BMC 14442
DOI:
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
29 December 2017
13 June 2018
4 July 2018
Please cite this article as: Jinyu, L., Xiaoyan, S., Huachong, L., Li, F., Yuanyuan, L., Xuemei, T., Jufang, Y., xin,
C., Feifei, C., Jie, L., Dacheng, Y., Design, synthesis, and evaluation of novel L-phenylglycine derivatives as
potential PPARγ lead compounds, Bioorganic & Medicinal Chemistry (2018), doi: https://doi.org/10.1016/j.bmc.
2018.07.005
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.

Design, synthesis, and evaluation of novel L-phenylglycine derivatives
as potential PPARγ lead compounds
Liu Jinyu a,†, Su Xiaoyan a,†, Li Huachong a, Fan Li a,*, Li Yuanyuan a,Tang Xuemei b, Yan Jufang c, Chen
xin c, Chen Feifei a, Liu Jie a, Yang Dacheng a,*
a Institute of Bioorganic and Medicinal Chemistry, Key Laboratory of Applied Chemistry of Chongqing Municipality, School of Chemistry and
Chemical Engineering, Southwest University, Chongqing 400715, PR China.
b School of Life Science, Southwest University, Chongqing 400715, China
c Drug Screening Center, Chengdu Di Ao Pharmaceutical Group Co., Ltd., Chengdu 610041, China
† Liu Jinyu and Su Xiaoyan have identical contribution to this work
Abstract: In accordance with the structural characteristics of thiazolidinedione drugs and highly bioactive
tyrosine derivatives, we tentatively designed the L-phenylglycine derivatives TM1 and TM2 based on
basic principles of drug design and then synthesized them. The in vitro screening of peroxisome
proliferator-activated receptor gamma (PPARγ) activated activity, α-glucosidase inhibitory and dipeptidyl
peptidase-4 inhibitory activities showed that the novel molecule M5 had efficient PPAR response element
(PPRE) activated activity (PPRE relative activity 105.04% at 10 μg·mL-1 compared with the positive
control pioglitazone, with 100% activity). Therefore, M5 was selected as the hit compound from which the
TM3 and TM4 series of compounds were further designed and synthesized. Based on the PPRE relative
activities of TM3 and TM4, we discovered another new molecule, TM4h, which had the strongest PPRE
relative activity (120.42% at 10 μg·mL-1). In addition, the concentration-dependent activity of the highly
active compounds was determined by assaying their half-maximal effective concentration (EC50) values.
The molecular physical parameter calculation and the molecular toxicity prediction were used to
theoretically evaluate the lead-likeness and safety of the active compounds. In conclusion, we identified a
potential PPARγ lead molecule and developed a tangible strategy for antidiabetic drug development.
Keywords: Diabetes mellitus; L-Phenylglycine; Tyrosine derivatives; Thiazolidinediones; Peroxisome
proliferator-activated receptor; Dipeptidyl peptidase-4; α-Glucosidase
1

1. Introduction
Diabetes mellitus (DM) is an endocrine and metabolic disease, which has become the third most
common non-infectious chronic disease after cardiovascular diseases and tumors. The Global Report on
Diabetes published by World Health Organization in 2016 reported that there were approximately 422
million adults with DM worldwide, which will reach 592 million by 2035, and 90–95% are expected to be
patients with type 2 diabetes mellitus (T2DM) [1]. T2DM (formerly called non-insulin-dependent or
adult-onset diabetes) results from the body’s ineffective use of insulin [1], characterized especially by
insulin resistance [1]. T2DM can cause hypertension, diabetic retinopathy, and other chronic complications
that lead to an increase in mortality [2]. Drugs currently used for the treatment of T2DM include insulin
sensitizers (e.g., pioglitazone), insulin secretagogues (e.g., sulfonylureas), aldose reductase inhibitors (e.g.,
acarbose), dipeptidyl peptidase-4 (DPP-4) inhibitors (e.g., sitagliptin), sodium-glucose cotransporter
(SGLT2) inhibitors (e.g., canagliflozin), and glucagon-like peptide-1 (GLP-1) receptor agonists (e.g.,
liraglutide) [1]. Although these drugs have clearly made outstanding contributions to the treatment of
T2DM, the side effects caused by long-term use or high-dose administration has encouraged the continued
search for novel T2DM drugs, which has attracted the attention of drug development researchers.
Peroxisome proliferator-activated receptor gamma (PPARγ) is mainly responsible for the regulation
of the transcription of insulin-responsive genes, which control the production, transmission, and utilization
of glucose [3]. PPARγ is also the key decisive factor controlling the development of adipocytes [4]. Of note,
anti-diabetic drugs inhibit obesity-linked phosphorylation of PPARγ by cyclin-dependent kinase 5 (CDK5)
[5]. PPARγ agonists are insulin sensitizers, and clinical drugs in this class include thiazolidine-2,4-dione
(TZD) small molecule compounds such as rosiglitazone and pioglitazone [6]. Although TZDs have resulted
in a significant breakthrough in the treatment of T2DM, their long-term use and high-dose administration
may also lead to weight gain, edema, liver injury, and other side effects [7], which limited their clinical
application. Thus, the development of high-efficacy and low-toxicity non-TZD PPARγ agonists has become
a goal of researchers in this field. The study of non-TZD PPARγ agonists has revealed that several natural
amino acids including phenylalanine, leucine, and histidine exert blood glucose-reducing effects [8] and
numerous phenylalanine derivatives [9,10] (e.g., nateglinide) and tyrosine derivatives [11] (e.g., farglitazar,
GW409544, and GW1929) have excellent PPARγ activated activity. Farglitazar, a classical tyrosine
derivative, is also a strong PPARγ agonist [12]. Unfortunately, farglitazar was not successful in phase III
2

trials because of its adverse events such as angioneurotic edema [13]. Nonetheless, the studies mentioned
above provided insights into the development of non-TZD PPARγ agonists with higher potency and lower
toxicity than the existing agents.
A structural analysis of some TZDs (rosiglitazone, pioglitazone, and troglitazone) and tyrosine
derivatives (GW409544, farglitazar, and GW1929) [14,15] (Scheme 1) revealed that PPARγ agonists
typically include a hydrophilic group attached to the aromatic ring, a linker, and a hydrophobic group
[14,16]. Unlike TZDs, tyrosine derivatives have a lipophilic group derived from the α-amino group, which
is believed to improve the PPARγ affinity of farglitazar compared with that of TZDs [15]. The above
information indicates that the target molecule should contain the required pharmacophore for PPARγ
agonists as well as the previously mentioned lipophilic groups and, accordingly, we tentatively designed
molecule A (Scheme 1).
Scheme 1. Designed molecules inspired by thiazolidine-2,4 -diones (TZDs) and tyrosine derivatives.
Molecule A is a structural derivative of tyrosine. The truncation of the CH2 of A converted it into a
phenylglycine derivative, which has been used widespreadly, for instance, the chlorophenylglycine
derivatives are used in the synthesis of penicillin and cephalosporin derivatives [17] and the D-p-hydroxy
-phenylglycine is a side chain of semi-synthetic β-lactam antibiotics [18]. However, there are few reports
on the integration of a phenylglycine unit into antidiabetic drug molecules. Therefore, we hypothesized that
3

an antidiabetic molecule with high potency and low toxicity would be obtained by changing the tyrosine
residue to a phenylglycine residue. Thus, we further designed molecule B from A (Scheme 2).
It is well known that the ligand-binding domain (LBD) of PPAR is a large, Y-shaped site surrounded
by 13 α-helices and a small four-stranded β-sheet [13]. Ligands usually adopt a U-shaped conformation
when binding to the LBD of PPARγ [19]. Most PPARγ agonists contain a flexible alkyl ether linkage to
enhance their binding to receptors [20]. Khanna et al. shifted alkyl ether bond from para-site to meta-site
[21] and Suh et al. replaced the flexible linker at the meta-site with a rigid linker [18], which resulted in
strong PPARγ activated activity. Thus, we not only changed the location of the alkyl ether group of the
PPARγ agonist but also performed rigidity-flexibility exchange. Consequently, we changed the flexible
alkyl ether bond of B in the para-site into an alkyl amide in the meta-site to obtain molecule C (Scheme 2).
The mechanism of action of TZD drugs indicates that either the aliphatic group or the heterocyclic
aryl could be selected as the hydrophobic part of target molecule because of the large binding cavity of the
receptor [22]. We used fragment-based drug design and, therefore, drug molecular fragments or drug-like
molecular fragments were given prior consideration as hydrophobic R1 groups of the target molecules.
Therefore, the selected fragment R1 comprised of sulfamethoxazole (antibacterial drug), benzimidazole (the
fragments of omeprazole and rabeprazole) and phthalimide (derivatives thereof have anti-inflammatory
[23], antibacterial [24], and antitumor [25] activities). We also chose benzyloxycarbonyl (Cbz), an ordinary
and readily removable group, as the lipophilic group of the target compounds. In order to get more
information about structure-activity relationships (SARs), the acidic COOH of the target molecules was
converted into neutral COOCH3. Based on the above requirements, we designed and synthesized the target
molecules TM1 and TM2 (Scheme 2) with L-phenylglycine as the basic skeleton, and then assessed their
PPARγ activated activity.
4

Scheme 2. Design of target molecules TM1–TM4.
Preliminary activity screening showed that both TM1 and TM2 exhibited poor PPARγ activated
activities, whereas the intermediate M5 (Scheme 2) exhibited better PPARγ activated activity (the PPRE
relative activity was 105.04% at 10 μg·mL-1). Therefore, we designed and synthesized TM3 and TM4 by
altering both the carboxyl-terminal and chlorine atom of M5. The biological activity screening of TM3 and
TM4 revealed that compound TM4h had the highest activity, which not only confirmed the correctness of
our molecule design, but also provided a new direction and foundation for subsequent research.
2. Results and discussion
2.1 Synthesis
2.1.1 Synthesis of TM1 and TM2
TM1 and TM2 can be structurally divided into an L-3-aminophenylglycine moiety and a hydrophobic
moiety linked to the aromatic amine group; thus, we designed four synthetic routes (Scheme 3).
Scheme 3. Synthetic routes of designed target compounds TM1 and TM2
5

Route A consists of six steps: nitration of phenylglycine, protection of the α-amino group, reduction of
the nitro group, acetylation of the amino group at the benzene ring, substitution of the chlorine atom, and
methyl esterification of the carboxyl group. Route A had the shortest steps and, thus, was our initial
preference. However, we found that although the crude TM2 product could be successfully obtained after
the step substituting the chlorine atom, purification of the product was difficult owing to the numerous side
reactions and its poor solubility. In addition, it was difficult to convert some TM2 compounds to their TM1
, owing to their poor solubility. The reaction for route B was similar to that of route A, except
for an extra step (hydrolysis of the methyl ester) and the different order of reactions. Route B is a typical
step-by-step synthesis method and all the intermediates and target molecules can be obtained theoretically.
However, in this route, the raw material and product of the nitro reduction are often covered in zinc powder,
which results in a long reduction time, incomplete reduction, low extraction and separation efficiency, and
extremely low yield. Route C is similar to route B, which contains the same unit reaction, but the formation
of methyl ester is performed prior to the reduction of nitro. Route C overcame the problem associated with
a reduction of the nitro in route B, but two amino groups still exist in the molecule. This indicates that
benzylamine would also be protected in the step protecting α-amino with Cbz, making selective protection
difficult. Although route D still uses L-phenylglycine as the raw material, Cbz protection of α-amino group
and methyl esterification of carboxyl group are carried out before nitro group reduction. This process not
only addresses the reduction issues but also selectively protects amino group easily and provides essential
intermediates and target molecules. Therefore, route D is a feasible synthetic route in theory.
We previously reported the nitration of phenylglycine and the subsequent modification of the
α-amino group [26~28] using experimental methods well-developed by our laboratory and the overall yield
of both steps was over 80%. The combination system of Alcohol (ROH)/thionyl chloride (SOCl2) is usually
used as the reagent for esterification of amino acids in our laboratory [29] to deliver an alkyl ester with high
purity and a satisfactory yield. We also adopted SOCl2/methanol (MeOH) as the reaction reagent in the
synthesis of M2 to M3 and found that this simple and convenient process generated the corresponding
alkyl ester at a very high yield (over 85%). The significant step in the synthesis of the target products TM1
and TM2 is the reduction of nitro group (the synthesis of M4), which requires both the successful reduction
of -NO2 and prevention of side reactions on the -CO2CH3 and -Cbz group. We attempted reduction
reactions using 80% hydrazine hydrate, Zn/ammonium chloride (NH4Cl)/acetic acid (HOAc), Zn/HOAc,
Zn/NH4Cl, and Zn/CaCl2 and finally discovered that the Zn/NH4Cl/HOAc reduction was the best. However,
we identified an impurity close to the target molecule after repeated TLC analyses, which indicated that it
would be very difficult to obtain a pure product using column chromatography alone. Additionally,
prolonged exposure of the reduction product to air would oxidize it and, so, it is not efficient to purify M4
just using column chromatography. We attempted the next chloroacetylation reaction with crude M4 and,
fortunately, found that the reaction proceeded smoothly, the purification was simple, and the total yield of
both steps was approximately 50–70%. Thus, the large-scale synthesis of M5 became practical, which
provided a solid foundation for the synthesis of the target molecules TM1 and TM2.
6

Scheme 4. Synthesis of TM1 and TM2: (a) HNO3, H2SO4, 85%; (b) benzyloxycarbonyloxysuccinimide, 10% K2CO3,
acetone, 92%; (c) MeOH, SOCl2, 83%; (d) Zn, NH4Cl, MeOH; (e) 2-chloroacetyl chloride, K2CO3, acetone, 80%; (f) Y1-H,
K2CO3, DMF, 38% for TM1a, 75% for TM1b, 66% for TM1c, 73% for TM1d, and 85% for TM1e; (g) LiOH, THF; 10%
HCl, 86% for TM2a, 83% for TM2b, 87% for TM2c, 95% or TM2d, and 90% for TM2e.
The synthesis of TM1 through the reaction of M5 with amine or thiophenol (Y1-H) is a typical
nucleophilic substitution, and the related experimental results are presented in the supporting information.
The yields of TM1b–TM1e were above 60%, whereas that of TM1a was only 38.7%. This was attributed
to the nucleophilicity of the nucleophile. Specifically, since the acidity of the N-H of phthalimide is greater
than that of benzimidazole, the N-H of phthalimide is more likely to convert into a negative nitrogen ion
under alkaline conditions. Therefore, TM1b exhibited a stronger reactivity and higher yield than TM1c did.
N-Boc-piperazine and 2-mercaptobenzimidazole are both strong nucleophiles; however, 2-mercapto-
benzimidazole is a planar molecule and the nucleophilic ability of the sulfur atom is stronger than that of
the nitrogen atom and, therefore, the yield of TM1e was higher than that of TM1d. In addition, the NH2 on
the benzene ring of sulfamethoxazole was conjugated with the benzene ring bearing a strong electron-
withdrawing group sulfonamide, which resulted in the poor nucleophilicity of NH2 group and, thus, a low
reaction rate. Furthermore, the relatively large molecular weight and poor solubility of TM1a made column
chromatography difficult and, thus, TM1a was obtained with the lowest yield.
TM2 can conveniently be prepared by the hydrolyzing the corresponding TM1. Methyl ester is
usually hydrolyzed under acidic or alkaline conditions. The -Cbz would likely be removed under strong
acidic conditions, and the chiral molecules would probably undergo racemization under intense conditions
and, therefore, TM1 must be hydrolyzed under mild condition. By exploring various experimental
conditions, we found that THF was used as the solvent, pH adjusted to 10–11 by LiOH, temperature was
controlled between 16–25oC, and the reaction was completed within 1 hour without racemization. The
7

experimental results are given in the supporting information.
2.1.2 Synthesis of TM3 and TM4
The TM3 and TM4 series were the target molecules we further designed using M5 as the hit
molecule. TM3 is a series of compounds derived from M5 by the substitution of a chlorine atom and,
therefore, they can be synthesized by conventional amidation and the substitution of heterocyclic amine.
The former process requires the reaction of M4 with the corresponding acyl chloride R3CH2COCl and the
latter involves a nucleophilic substitution reaction between heterocyclic amines and M5. TM4 is an
alternative derivative of the carboxyl methyl ester of M5, which is similar to M5 structurally and can be
prepared using two synthetic routes. 1) Starting from M2, TM4b–TM4h were assembled by the
esterification or amidation of a carboxyl group (converted to M6), followed by direct chloroacetylation of
the amino group after the reduction of nitro group. 2) TM4a was generated through a two-step reaction
between nitro reduction of M2 (affording M7) and chloroacetylation of amino group. The synthetic route
is shown in Scheme 5.
Scheme 5. Synthesis of TM3 and TM4: (h) Y2-acyl chloride, K2CO3, acetone, 65% for TM3e, 53% for TM3f, 51% for
TM3g, 65% for TM3h, 62% for TM3i, and 52% for TM3j; (i) Y2-H, K2CO3, DMF, 66% for TM3a, 55% for TM3b, 53%
for TM3c, and 32% for TM3d; (j) H-XY3, SOCl2 or dicyclohexylcarbodiimide, 85% for M6b, 88% for M6c, 88% for M6d,
92% for M6e, 88% for M6f, 67% for M6g, and 86% for M6h; (k) Zn, NH4Cl, MeOH; (l) 2-chloroacetyl chloride, K2CO3,
acetone, 73% for TM4a, 65% for TM4b, 70% for TM4c, 62% for TM4d, 27% for TM4e, 46% for TM4f, 56% for TM4g,
and 38% for TM4h; (m) Zn, CaCl2, 85%; (n) K2CO3, DCM.
The results of the TM3 preparation are shown in the supporting information. TM3a–TM3d were
synthesized by the nucleophilic substitution reaction between M5 and azoles with a yield of 32–66%. The
yield of TM3d was only 32.7%, which was attributed to the steric hindrance of the ortho-methyl group.
8

The azoles used to prepare TM3b and TM3c contained resonance structures (Scheme 6), which include a
plurality of reaction sites and, therefore, the purification of these products is not easy. In fact, we only
separated the main target products while other isomers generated at other sites were not isolated.
TM3e–TM3i were obtained by acylation of M4 with different acyl chlorides, and the total yield of the
two-step reaction was over 51%.
Scheme 6. Resonant structures of azoles used to prepare TM3b and TM3c.
The key synthetic step of the target products TM4b–TM4h is the synthesis of M6. As a derivative of
M2, M6 is prepared by esterification or amidation of the carboxyl group of M2. There are numerous
methods for esterifying carboxyl groups including the carboxylic acid coupling method aided by dicyclo-
hexylcarbodiimide (DCC) or 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDCI) [30]. We used the
SOCl2/ROH method previously used in our laboratory to produce M6b–M6d with high purity and
excellent yield (85–92%) [29,31,32]. The general synthetic methods of amides include the reaction of
acyl halides, mix acid anhydrides, acyl azides as well as active esters with amines. Although numerous
methods are available for activating carboxylic function, we usually used a condensation reagent (such as
carbodiimide and onium-type regents) to convert chiral amino acids into corresponding amides [30]. While
the purification process is tedious, the yield is high, and racemization rarely occurs [30]. In this experiment,
DCC was used as activation agent, and hydroxybenzotriazole (HOBt) was used as additive. The molar ratio
of M2, DCC, HOBt, and base components was 1:1.2:1.2:1.2, and the yields of M6e–M6h were 67–92%.
As shown in the supporting information, the yield of M6e was the highest (92.0%), which may be related to
the small volume of methylamine while that of M6g was only 67.5%, mainly because there were two
reaction sites in hydroxylamine hydrochloride.
The synthesis of TM4b–TM4h from M6 required two steps, consisting of reduction of nitro group
and acylation of the resulting amino group. Given the properties of amino acids, the reaction conditions
were set as mild as possible to obtain high purity products without racemization, and the reaction yield was
not the main priority. The characterization results of TM4 derived from M6 are presented in the supporting
information. The yields of TM4e and TM4h were relatively low, which was thought to be the result of
product loss during the repeated recrystallization steps.
9

The key step of the synthesis of TM4a is the production of M7 from M2. Owing to the amphoteric
properties of amino acids, we initially selected Zn/NH4Cl/HOAc as the reducing reagent for M2. However,
due to the high water solubility of M7, this lead to a low yield which prevented the effective separation of
the product from the reaction mixture after adding excess water in the subsequent process. To improve the
yield of M7, we systematically studid possible strategies and finally established the Zn/CaCl2 reduction,
and thus obtaining excellent yield with high purity (see supporting information). When we adopted the
optimal procedure, the acylation of M7 proceeded smoothly, and the preparative results of TM4a are
shown in the supporting information.
2.2 SAR
2.2.1 SAR of TM1 and TM2
PPARs are a nuclear hormone receptor superfamily, comprised of three subtypes namely PPARα,
PPARγ, and PPARδ [33,34]. Presently, the fibrate-type lipid-lowering drugs (such as fenofibrate) and
certain insulin sensitizers sold on the market belong to PPARα and PPARγ agonists, respectively. PPARδ is
involved in lipid transport and metabolism, but currently, there is no listed drug with PPARδ activity [35].
In the presence of ligands, the PPAR heterodimer, which contains the retinoid X receptor (RXR), regulates
the transcription of target genes by binding to the PPRE in the promoter region of the target gene.
Stimulation by endogenous ligands such as saturated or unsaturated fatty acids and arachidonic acid
derivatives or exogenous blood lipid-reducing antidiabetic drugs initiate the PPAR response by acting as a
molecular sensor and trans-activating the target genes. Therefore, PPRE is an effective tool for the
evaluation of potential PPAR agonists. In this study, the antidiabetic activity of target molecules was
assessed by assaying PPRE relative activity.
The in vitro PPRE relative activity of the target molecules (10 µg·mL-1) was evaluated using
pioglitazone (0.78 μg·mL−1, 100%) as the positive control. As shown in Table 1, most molecules of the
TM1 and TM2 series possessed weak PPRE relative activity at tested concentrations. However, TM2a,
TM2b, and TM2e exhibited potent PPRE relative activity of 40.63 %, 39.93%, and 32.91%, respectively.
The activity intensities indicated that the PPRE relative activity of the TM2 series of compounds, which
have a carboxyl group at the hydrophilic end, were more active than those of the corresponding TM1 series
compounds, which have a carboxylic acid methyl ester at the hydrophilic end. This may be related to the
water solubility of these compounds and the presence of carboxyl groups.
10

Table 1
PPRE relative activity of TM1, TM2, and intermediates
PPRE/%
Relative Activity*
Concentration
µg·mL-1/µmol·L-1
Compound
-Y1
-R
Activity
105.04
M5
-Cl
-CH3
-CH3
10/25.59
10/16.45
100.00
TM1a
18.85
11.92
3.93
27.91
TM1b
TM1c
TM1d
-CH3
-CH3
-CH3
10/19.94
10/21.16
10/18.50
14.56
4.52
-8.27
-10.89
TM1e
M8
-CH3
-H
10/19.82
10/26.54
7.66
-9.41
-Cl
53.49
49.10
TM2a
TM2b
TM2c
TM2d
-H
-H
-H
-H
-H
10/16.84
10/20.51
10/21.81
10/18.99
40.63
39.93
3.85
58.77
47.43
4.30
21.43
27.49
39.37
TM2e
10/20.38
0.78/2.19
32.91
100
/
Pioglitazone
*The data in relative activity is the PPRE activity at the equal molar concentration (µmol·L-1) converted by (PPRE/Conc.)
×24.36, and the value of PPRE is an average value of four measurements
The PPRE relative activity of TM2 was stronger than that of TM1, which was consistent with our
design idea. However, unexpectedly, the activity of TM2 was not as potent as that of the intermediate, M5
(methyl ester group), which exhibited higher potency than that of M8 (carboxyl group). Therefore, we
suggested that M5 may be a novel PPARγ hit compound.
To predict the selectivity of the compounds, we further evaluated the inhibitory effects of TM1,
TM2, and selected intermediates on α-glucosidase and DPP-4. α-Glucosidase retards the absorption of
glucose in the intestinal tract, thereby delaying the peak postprandial blood glucose levels to control
11

excessive elevation. Currently marketed α-glucosidase inhibitors include voglibose, acarbose, and miglitol
[36]. DPP-4 is a transmembrane serine protease that belongs to the proline oligopeptide family [37]. DPP-4
inhibitors, as the name suggests, inhibit the activity of DPP-4, leading to increased levels of glucagon-like
peptide (GLP)-1 and gastric inhibitory peptide (GIP) [38]. These actions regulate the blood glucose and
other indicators to achieve the goal of diabetes treatment. KR-62436, a DPP-4 inhibitor, was approved for
clinical use by the US Food and Drug Administration (FDA) in 2007 [39,40]. As shown in Table 2,
compared with the reference drugs, voglibose and KR-62436, the α-glucosidase and DPP-4 inhibitory
activities of TM1, TM2, and their intermediates were extremely weak, showing that both TM1 and TM2
displayed good target specificity.
Table 2
α-Glucosidase and DPP-4 inhibitory activity of TM1, TM2, and selected intermediates*
α-Glucosidase
Inhibition
/%
DPP-4
Inhibition
/%
α-Glucosidase
Inhibition
/%
DPP-4
Inhibition
/%
Concentration
/μmol·L-1
Concentration
/μmol·L-1
Compd.
Compd.
25.59
16.45
19.94
21.16
18.50
19.82
3.74
26.54
16.84
20.51
21.81
18.99
20.38
0.81
17.66
-17.33
M5
TM1a
TM1b
TM1c
M8
TM2a
14 .63
11.85
17.53
2.91
16.12
5.87
8.89
17.83
0.99
11.10
-33.95
-0.74
5.26
TM2b
TM2c
2.17
-0.31
TM1d
TM1e
TM2d
TM2e
15.46
-6.12
13.69
9.26
8.99
/
12.58
/
-7.54
73.11
103.95
Voglibose
KR-62436
*The value of measured activities is an average value of four measurements.
2.2.2 SAR of TM3 and TM4
The proposed design of the TM3 series of molecules included replacing the chlorine atoms in the M5
molecule with different groups to investigate the effect of different groups on the biological activity.
Therefore, azoles and simple groups with similar or slightly larger size as chlorine atom became preferred
substituents in molecular design. It is reported that diazole derivatives have anti-inflammatory and
antibacterial effects, and can regulate plant growth [41]. Tetrazole and its derivatives can be used to
synthesize pesticides, medicines, photographic materials, and other substances [42]. We attempted to
design and synthesize 10 derivatives of TM3 by selecting diazole, tetrazoles, chlorine-free alkyl groups,
12

chloroalkyl and methoxy groups, as well as hydrogen atom to replace the chlorine atom in M5 molecule.
The PPRE relative activity of TM3 was measured with pioglitazone as the positive control, and the results
are presented in Table 3.
Table 3
PPAR response element (PPRE) relative activity of TM3.
Concentration
µg·mL-1/µmol·L-1
10/25.59
PPRE/%
Relative Activity*
Compound
-Y2
Activity
105.04
M5
-Cl
100.00
10/21.25
10/22.81
19.53
14.45
22.39
TM3a
15.43
14.94
1.49
TM3b
TM3c
TM3d
10/23.56
10/22.20
14.45
1.36
-H
-CH3
10/28.06
10/27.00
10/26.01
10/24.70
10/23.87
24.93
18.28
12.01
-3.55
11.58
21.64
16.49
11.25
-3.50
TM3e
TM3f
TM3g
TM3h
TM3i
-C2H5
-CH2Cl
-CH2CH2Cl
11.81
-16.07
-OCH3
/
10/25.88
0.78/2.19
-17.07
100
TM3j
Pioglitazone
*The data in relative activity is the PPRE activity at the equal molar concentration (µmol·L-1) converted by (PPRE/Conc.)
×24.36, and the value of PPRE is an average value of four measurements
Table 3 shows that all compounds containing azole unit in TM3 series exhibited weak PPRE relative
activity (< 30%) at 10 μg·mL-1 in the following decreasing order of magnitude: TM3a > TM3b > TM3c >
TM3d. The substitution of a chlorine atom with a hydrogen atom, methyl, and ethyl produced the
following compounds with decreasing magnitude of activity: TM3e > TM3f > TM3g, which was possibly
related to the alkyl volume. In contrast, the compounds with chloroalkyl group (TM3h and TM3i) or a
methoxy group (TM3j), showed PPRE relative activity in the following decreasing order: TM3i > TM3h >
13

TM3j, which shows that the stronger the electron-withdrawing effect of the substituent, the weaker the
PPRE relative activity of the corresponding molecule. However, because the changes described above did
not result in a more active compound than M5, the Y2 part of the molecule requires further investigation.
Based on the PPARγ activated activities of the TM3 series, we decided to retain the chloroacetyl unit
of M5 and vary the methyl carboxylate moiety, which led to the design of the TM4 series of molecules.
Initially, a soluble potassium carboxylate (TM4a) with good solubility was designed, and then the
homologs of the methyl ester, that is, the ethyl (TM4b), propyl (TM4c), and butyl (TM4d) esters were also
designed. Furthermore, adopting the principle of drug design by converting the ester group into an amido,
we further designed the target molecules TM4e–TM4h, corresponding to the amides of methylamine,
ethylamine, hydroxylamine, and -N(Me)C2H4OH. After these eight target molecules were synthesized ,
their PPRE relative activities in vitro were assayed. The results presented in Table 4 showed that the PPRE
relative activities of TM4b–TM4g were extremely weak while that of TM4a was 53.49%, but still weaker
than that of M5. Fortunately, the PPRE relative activity of TM4h was 120.42%, not only stronger than that
of pioglitazone, but also higher than the lead molecule M5. The activity of the TM4 series showed that
carboxyl terminus of phenylglycine derivatives could be either ester or amido group, and the other part of
these molecules should contain a hydrophobic moiety, which is consistent with the structural characteristics
of TZD and tyrosine derivatives.
Table 4
PPRE relative activity of TM4.
Concentration
Relative PPRE/%
Activation*
Compound
-XY3
µg·mL-1/µmol·L-1
Activation
53.49
-O-K+
-OEt
10/24.70
10/25.32
10/24.46
10/21.89
10/26.33
10/25.39
10/26.19
TM4a
TM4b
TM4c
TM4d
TM4e
TM4f
TM4g
52.75
-23.89
-17.07
-5.58
-24.83
-17.14
-5.01
-OPr
-OBu
-NHMe
-NHEt
-NHOH
-18.80
-20.13
-5.33
-17.39
-19.31
-4.96
14

-N(Me)C2H4OH
/
10/23.59
0.78/2.19
TM4h
120.42
124.35
Pioglitazone
100
*The data in relative activity is the PPRE activity at the equal molar concentration (µmol·L-1) converted by (PPRE/Conc.)
×24.36, and the value of PPRE is an average value of four measurements
Based on the preliminary screening, we further evaluated the half-maximal effective concentration
(EC50) of M5 and TM4h which were the most active among the designed molecules (Table 5).
Table 5
EC50 values of most active compounds.
Compound
EC50(μmol·L-1)
Max Activated activity (%)
Max Conc. (μmol·L-1)
22.39
8.97
0.24
178.74
208.04
142.47
25.59
23.05
2.19
M5
TM4h
Pioglitazone
2.3 Lead-likeness and toxicity
2.3.1 Lead-likeness and toxicity prediction of TM1 and TM2 series
To predict the lead-likeness of the TM1 and TM2 series of compounds, we calculated their
physicochemical parameters using the PLEXUS software provided by Eli Lilly and Co., and the results
were given in Table 6. It is well known that the physicochemical properties of molecules are one of the
factors that must be considered in selecting lead compounds. The cLogP is an important parameter
describing the hydrophobic properties of compounds; the larger the cLogP, the better the hydrophobicity of
the compound [43]. The topological polar surface area (tPSA) is the sum of the surfaces of polar atoms in a
molecule. Presently, it is generally considered that a tPSA value ≤ 140 is best [44]. Table 6 shows that the
cLogP values of the TM1 series were generally much higher than those of the TM2 series were, but the
TM2 series exhibited stronger PPRE relative activity. This observation indicates that the PPAR target
protein maybe requires a better spatial matching rather than stronger lipophilicity. Table 6 also shows that
the tPSA values of TM2 series compounds are higher than those of TM1 series, indicating that the higher
the tPSA value of a compound, the stronger its interaction with the PPAR target protein. Therefore, the
corresponding PPARγ activated activity of the compound with greater tPSA value is better than that of
compounds with lower values.
Lipinski’s rule of five (RO5) is an indicator of the oral bioavailability of compounds, involving
molecular weight, cLogP, and the number of hydrogen bond donors (H-bond donors) or acceptors (H-bond
15

acceptors) [45]. The molecular weights of our compounds are all approximately 500, which essentially
meet the requirements of RO5. The TM2 molecules, having carboxyl groups as their hydrophilic end
groups, have better water solubility, lower cLogP, and generally better activity than the TM1 molecules.
Both TM1 and TM2 series of molecules have H-bond acceptors and donors, accordingly, TM2a, with the
largest number of H-bond acceptors (8) and donors (5), has the strongest activity (40.63%, 16.84 µmol·L-1)
among TM1 and TM2 series. This observation indicates that there is a probable relationship between the
H-bond acceptors, H-bond donors, and PPARγ activated activity. It is worth noting that the M5
intermediate has the lowest number of H-bond acceptors (3) and donors (2), but its activity (105.04 %,
25.59 µmol·L-1) is stronger than that of TM2a. Similarly, the number of H-bond acceptors and donors of
pioglitazone, are all less than those of TM2a, but the activity of pioglitazone is stronger than that of TM2a
and M5. Therefore, we thought that the number of H-bonds alone is not the only factor that affected the
activity of these molecules.
Table 6
Physical properties of TM1 and TM2.
H-bond
H-bond
donors
Fraction of
sp3 carbons
Rotatable
bonds
cLogP
tPSA
RO5
Leadlikeness
Compound
acceptors
3.53
2.47
67.76
93.73
1
1
4
3
1
2
0.32
0.21
7
9
1
1
Pioglitazone
M5
1.80
2.00
2.08
3.00
2.97
2.23
1.50
1.70
1.78
0.43
2.67
173.52
131.11
109.33
126.51
118.12
104.73
184.52
142.11
120.33
137.51
129.12
0
0
1
0
0
1
0
1
1
0
1
7
5
4
5
4
4
8
6
5
6
5
4
2
2
2
3
3
5
3
3
3
4
0.17
0.15
0.15
0.43
0.15
0.17
0.14
0.12
0.12
0.41
0.12
13
10
10
12
11
8
0
0
0
0
0
1
0
0
0
0
0
TM1a
TM1b
TM1c
TM1d
TM1e
M8
12
9
TM2a
TM2b
TM2c
TM2d
TM2e
9
11
10
Lipinski’s RO5 (drug-likeness) requires mass ≤ 500, cLogP (pH 7.4) ≤ 5, H-bond donors ≤ 5, and H-bond acceptors ≤
10. Lead-likeness requires mass, 200–350 (optimization might add approximately 100–200); cLogP < 1.0–3.0 (optimization
might increase this by 1–2 log units), single charge present (secondary or tertiary amine preferred); and non-substrate
16

peptides are suitable.
After Lipinski proposed the RO5, other scientists developed more drug-likeness parameters. For example, Fsp3 is
the ratio of the sp3 hybrid carbon number to total carbon number, which describes the unsaturation of the compound
[46]. Studies have shown that a high Fsp3 value indicates that a molecule has more rotatable bonds, which means that
its molecular flexibility is better and space adaptability is stronger than that of compounds with low values. A rotatable
bond is a single bond other than the single bonds in the ring and C-N bonds [47]. The number of rotatable bonds
determines the rigidity and flexibility of a compound. For TM1 and TM2, the number of rotatable bonds is between 8
and 13, more than pioglitazone, indicating that the flexibility of these molecules is better than that of pioglitazone. The
Fsp3 values of TM1d and TM2d were closest to that of pioglitazone, but their PPRE activities are very weak. In
contrast, the Fsp3 value of M5 differs considerably from that of pioglitazone, but its PPRE relative activity is the
highest. These data revealed that the molecular activity was not necessarily related to the Fsp3 value, but the
appropriate Fsp3 value may help increase molecular flexibility and thus help molecule adapt to protein cavity.
Diabetes is a chronic disease that requires long-term drug administration and, therefore, antidiabetic drugs must
have little or no toxicity. To predict the safety of TM1 and TM2, we calculated the toxicity of these compounds using
the ADMET Predictor 7.0 software from Simulations Plus, Inc. (USA) (Table 7).
Table 7
Toxicity data of TM1 and TM2.
hERG
pIC50
TOX-RAT
(mg/kg)
TOX-BRM-Rat
(mg/kg/day)
TOX-BRM-Mouse
(mg/kg/day)
TOX-SGPT
0
TOX-MUT-Risk
0
TOX-Risk
1
Compound
5.16
629.57
4.81
61.68
Pioglitazone
M5
4.13
3.14
1.69
1.89
1.89
1.00
15.05
8.84
5.95
5.57
6.85
2.98
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
2
2
2
2
3
2
1
0
2
1
0
1
5.08
5.15
5.26
5.31
5.27
5.23
5.10
5.16
5.28
5.25
5.16
5.12
672.24
518.80
687.22
844.75
278.95
441.57
829.66
1133.07
966.63
1253.13
591.56
391.72
724.74
681.47
829.91
324.03
686.26
163.59
979.71
984.36
1146.36
519.32
999.17
265.37
TM1a
TM1b
TM1c
TM1d
TM1e
M8
TM2a
TM2b
TM2c
TM2d
TM2e
The human ether-à-go-go-related gene (hERG) is used to quantitatively predict the cardiac toxicity of a
compound, and a-log half-maximal inhibitory concentration (pIC50) ≤ 6 mol·L-1 is stipulated to be normal. Table 7
shows that the overall performance of the predicted compounds in the order of cardiac toxicity was pioglitazone ≤ TM2
< TM1, the molecules with the lowest predicted cardiac toxicity were M5, M8 and TM2e. Totally, the pIC50 values
of all predicted molecules were < 6 mol·L-1, indicating that they may have no cardiotoxicity.
TOX-RAT predicts the acute toxicity of compounds and is expressed as the median lethal dose (LD50) in rat
models. The LD50 predictive value of most drug molecules should be ≥ 300 mg/kg. According to Table 7, the predicted
acute toxicities of the TM2 series of molecules, which are more hydrophilic, were lower than that of the TM1 series.
17

Furthermore, the predicted acute toxicities of the intermediates M5 and M8 were lower than that of pioglitazone.
Although the predicted acute toxicities of TM1d and TM2d (contain Boc piperazine groups), TM1e and TM2e
(contain 2-mercapto benzimidazole groups) were stronger than that of pioglitazone, except for TM1d, the LD50 of all
the molecules were higher than 300 mg/kg and, therefore, they probably lacked acute toxicity.
TOX-BRM-Rat and TOX-BRM-Mouse predict the carcinogenicity of compounds. For most of the drug
molecules, the median predictive value of the carcinogenic dose (TD50) should be ≥ 4 and ≥ 25 mg·kg-1·day-1 (rat and
mouse models, respectively). In the prediction of carcinogenicity in rat, the order of magnitude of carcinogenicity of
the compounds was M8 < TM2 < pioglitazone < M5 < TM1, and TM2a < TM2d < TM2b < TM2c < TM2e (8.84,
6.85, 5.95, 5.57, 2.98 mg·kg-1·day-1, respectively), where TM1a and TM2a had the lowest predicted carcinogenicity
among that series of compounds. Unfortunately, the TM1 series and TM2e had carcinogenic risk. The predicted results
of carcinogenicity in mice revealed the following order TM2 < M8 < M5 < TM1 < pioglitazone, where TM2b <
TM2d < TM2a < TM2c < TM2e (1146.36, 999.17, 984.36, 519.32, 265.37 mg·kg-1·day-1, respectively)，TM1b <
TM1d < TM1a < TM1c < TM1e (829.91, 686.26, 681.47, 324.03, 163.59 mg·kg-1·day-1, respectively). TM1b and
TM2b exhibited the lowest predicted carcinogenicity in their respective series, and the order was also consistent in
both series. The predicted carcinogenicity of these compounds generally followed the order TM2 < TM1 in both the rat
or mouse models.
2.3.2 Lead-likeness and toxicity prediction of TM3 and TM4 series
Similarly, we also carried out the physicochemical parameter calculation and toxicity prediction of the TM3 and
TM4 series (Fig.1). Although TM3 and TM4 are parallel structures based on M5, due to the different transformation of
functional groups, the correlation between the two series of data is very poor. First, the cLogP values of the TM3 and
TM4 series (0.3–3 and -0.7–4.5, respectively) appeared to meet the RO5 requirements, but the calculated values of
most molecules were lower than that of pioglitazone. The tPSA values of the TM3 and TM4 series of compounds were
in the range of 90–140 and met the requirements for a lead compound, but were higher than that of pioglitazone. The
number of rotatable bonds of compound in these two series was between 8 and 12, higher than that of pioglitazone,
which indicated that these molecules were more flexible than pioglitazone. The PPARγ activated activity of TM4a was
only inferior to that of TM4h, but its Fsp3 value was the lowest in the TM4 series. There were five compounds with a
similar Fsp3 value to that of pioglitazone, but only TM4h showed excellent activity.
3

Fig.1. Comparisons of physicochemical properties and toxicities between TM3 and TM4 (physicochemical properties in the first row:
cLogP, tPSA, Fsp3, and rotatable bonds; predicted toxicity in the second row: hERG, acute toxicity, carcinogenicity in rat and mouse).
Toxicity is an important consideration during the development of new diabetes drugs. The predicted results
shown in Fig.1 that the pIC50 values of all molecules in the TM3 and TM4 series are less than 6 mol·L-1, which
indicates that they may not have cardiac toxicity.
The predicted results in Fig.1 revealed that all the other molecules in the TM3 and TM4 series may have lower
acute toxicity levels than that of pioglitazone, except for TM3b and TM3c. The predicted LD50 of all molecules in the
TM3 and TM4 series was > 300 mg/kg, showing that these compounds may possess no acute toxicity.
The carcinogenicity of the TM3 and TM4 molecules was predicted using rat (TOX-BRM-Rat) and mouse
(TOX-BRM-Mouse) models and the predicted carcinogenic risk was in the order of TM4 < TM3. All the molecules in
the TM3 and TM4 series had a lower predicted carcinogenic risk in mice than pioglitazone did, except for TM3a and
TM3d. Therefore, both the TM3 and TM4 series showed no predicted carcinogenic risk to mice.
3. Conclusion
We used L-phenylglycine as the starting raw material to prepare novel compounds of the TM1 and TM2 series
by nitration, α-amino protection, carboxyl protection, nitro reduction, acylation of the amino group on the benzene ring,
and reaction with different heterocyclic amines or thiophenol. The assays of PPAR activated activitiy, α-glucosidase
inhibitory activity and DPP-4 inhibitory activity identified one novel compound (M5) with high PPRE relative activity.
We adapted M5 as a hit compound and subsequently designed and synthesized the TM3 and TM4 series of molecules
by replacing the chlorine atom or methyl ester of M5 with other moieties. The results of the PPRE relative activity
assay revealed that TM4h had the highest PPRE relative activity. The calculation of physicochemical properties and
prediction of toxicities suggested that TM4h is worth further investigation. Finally, this study established the
structure-activity relationship of the target molecules, which has provided ideas and directions for further research
studies.
4. Experimental section
4.1 Reagents and instrumentation
1
Melting points were determined by an electrothermal X-6 apparatus. H NMR and 13C NMR spectra were
recorded on a Bruker AV 300 spectrometer using TMS as an internal standard. The chemical shifts were reported in
parts per million (ppm), the coupling constants (J) are expressed in hertz (Hz) and signals were described as singlet (s),
doublet (d), triplet (t), quartet (q), as well as multiplet (m). ESI-MS analyses were obtained on an Agilent 1946B
instrument. High resolution mass spectra (HRMS) analyses were conducted on a Daltonics Data Analysis 3.2 (Bruker)
using ESI (electrospray ionization) techniques. The purity of the compounds was examined by thin-layer
4

chromatography (TLC) on silica gel plate using petroleum ether and ethyl acetate. All chemical reagents and solvents
were commercially available, and were used without further purification [16].
4.2 Chemistry
4.2.1 Preparation of M1 and M2
M1 and M2 have been reported by our laboratory [26, 27].
4.2.2 Preparation of M3
In a round bottom bottle，SOCl2 (2.2 mmol) was added dropwise into methanol (10 mL) at ice-water temperature.
After stirring 0.5 h, M2 (2.1 mmol) was added into the mixture solution. The resultant mixture was refluxed at 70℃and
monitored by TLC. The crude product was obtained through concentrating under reduced pressure. Added a small
amount of methanol to dissolve the residue, dropped cooled petroleum ether to precipitate the target compound. After a
large amount of white solid appeared, filtrated, washed with water three times, and dried in vacuo, the pure M3 was
obtained.
M3: Benzyl (S)-(methoxycarbonyl)(3-nitrophenyl)methylcarbamate
1
Yield 83%, m.p.: 93~94℃. H NMR (DMSO-d6, 300 MHz) δ: 3.76(s, 3H, -COOCH3)，5.04~5.15(m, 2H,
ArCH2O), 5.49(d, J=6.4 Hz, 1H, CH), 6.11(d, J=5.36 Hz, 1H, -CONH-), 7.26~7.35(m, 4H, ArH), 7.56(t, 1H, J1=7.9 Hz,
J2=7.8 Hz, ArH), 7.75 (d, J=7.4 Hz, 1H, ArH), 8.18~8.26(m, 2H, ArH); 13C NMR (DMSO-d6, 75 MHz) δ: 53.3, 57.2,
67.4, 122.0, 123.5, 128.2, 128.3, 128.5, 129.8, 133.3, 135.7, 139.0, 148.5, 155.2, 169.9.
4.2.3 Preparation of M4
The mixture of M3 (1.1 mmol), Zn powder (11.8 mmol), ammonium chloride (6.4 mmol) and ethyl acetate
(EtOAc, 10 mL) in a 50 mL round bottom flask was stirred at 75 ℃ and monitored by TLC until completion. After hot
filtration, the EtOAc was employed to wash the residue on the filter paper to ensure that products obtained were all
dissolved in the filtrate. The combined filtrate was washed with water (3×6 mL), and dried over anhydrous Na2SO4.
The organic solvent was then removed under reduced pressure, and thus the yellow crude product M4 was obtained.
Without further purification, the crude product was used directly in the synthesis of M5.
4.2.4 Preparation of M5
M4 (12.1 mmol), dried acetone (10 mL), anhydrous K2CO3 (36.4 mmol) were added into the 50 mL round bottom
flask. While the resultant mixture was stirred in an ice bath, the acetone solution (10 mL) of chloroacetyl chloride (12.7
mmol) was dripped slowly from the constant pressure funnel. After 5 min, removed the ice bath, stirred the reaction
solution at 10℃and monitored the reaction process by TLC. Upon completion, filtrated off potassium carbonate,
distilled under reduced pressure, added EtOAc (25 mL) and saturated brine (10 mL), extracted with EtOAc (3×12 mL),
dried over anhydrous Na2SO4, concentrated to remove the solvent, and the crude product M5 was obtained. The crude
products were recrystallized from choroform/petroleum ether to give a white solid, that is the pure product M5.
M5: Methyl (S)-2-(benzyloxycarbonylamino)-2-(3-(2-chloroacetamido)phenyl)acetate
Yield 80%, m.p.: 170~171℃, [] 2D0 +22.0(c 1.0 mg/mL, Acetone). H NMR (DMSO-d6, 300 MHz) δ: 3.64(s,
1
5

3H, -COOCH3)，4.25(s, 2H, COCH2Cl), 5.06(s, 2H, ArCH2O), 5.25(d, J=7.5 Hz, 1H, CH), 7.12(d, J=7.6 Hz, 1H, ArH),
7.30~7.36(m, 6H, ArH), 7.60(d, J=12.9 Hz, 2H, ArH), 8.32(d, J=7.4 Hz, 1H, -CONH-), 10.40(s, 1H, -CONH-). 13C
NMR (DMSO-d6, 75 MHz) δ: 43.6, 52.3, 58.0, 65.8, 118.8, 119.3, 123.3, 127.8, 127.9, 128.4, 129.2, 136.9, 137.0,
138.7, 155.9, 164.8, 171.1; HR ESI-MS calcd for C19H19ClN2O5 [M-H]- 389.0910，found 389.0915.
4.2.5 General preparation of TM1
In a 50 mL flask, Y1-H (1.1 mmol), dimethyl formamide (2.0 mL) and fine anhydrous potassium carbonate (1.3
mmol) were added. After stirring vigorously 0.5 hour, added the compound M5 (0.9 mmol). Stirred evenly at room
temperature, and monitored the reaction process by TLC. After completion, added saturated brine (15 mL) to the
reaction mixture, adjusted pH to 5 ~ 6 with 2 mol/L hydrochloric acid in an ice bath, extracted by EtOAc (3 × 10 ml),
combined the organic layer, washed with saturated brine (15 mL), dried over anhydrous Na2SO4, the crude product was
obtained by drying in vacuo, and the pure TM1 was obtained furtherly by silica gel column chromatography or
recrystallization.
TM1a
：
Methyl
(S)-2-(((benzyloxy)carbonyl)amino)-2-(3-(2-((4-(N-(5-methylisoxazol-3-yl)sulfamoyl)
phenyl)amino)acetamido)phenyl)acetate
Yield 38%, m.p.: 195~197℃, [] 2D0 +19.0(c 1.0 mg/mL, Acetone). 1H NMR (DMSO-d6, 300 MHz) δ: 2.3(s, 3H,
CH3)，3.63(s, 3H, -COOCH3)，4.49(s, 2H, COCH2), 5.06(s, 2H, ArCH2O), 5.22(d, J=7.1 Hz, 1H, CH), 6.25(S, 1H, NH),
6.40(s, 1H, -C=CH-), 6.62(d, J=7.9 Hz, 2H, ArH), 7.08(d, J=7.4 Hz, 1H, ArH), 7.30~7.36(m, 6H, ArH), 7.51~7.56(m,
4H, ArH), 7.94(s, 1H, -CONH-) 8.21(d, J=7.2 Hz, 1H, -CONH-), 10.24(s, 1H, -SO2NH-); 13C NMR (DMSO-d6, 75
MHz) δ: 12.1, 50.2, 52.3, 59.8, 65.7, 96.7, 112.8, 118.4, 118.8, 122.0, 122.8, 128.4, 129.1, 129.4, 136.8, 139.0, 154.0,
155.9, 159.5, 162.3, 165.2, 170.4, 171.1; HR ESI-MS calcd for C29H29N5O8S [M+Na]+630.1624，found 630.1625.
TM1b: Methyl (S)-2-(benzyloxycarbonylamino)-2-(3-(2-(1,3-dioxoisoindolin-2-yl)acetamido)phenyl) acetate
Yield 75%, m.p.: 98~100℃, [] 2D0 +46.0(c 1.0 mg/mL, Acetone); []D20 +25.0(c 1.0 mg/mL, EtOAc). 1H NMR
(DMSO-d6, 300 MHz) δ: 3.64(s, 3H, -COOCH3), 4.01(s, 2H, COCH2)，5.07(s, 2H, ArCH2O), 5.25(d, J=7.3 Hz, 1H,
CH), 7.15(d, J=7.3 Hz, 1H, ArH), 7.32~7.95(m, 12H, ArH), 8.33(d, J=7.4 Hz, 1H, -CONH-), 9.89(s, 1H, -CONH-). 13C
NMR (DMSO-d6, 75 MHz) δ: 43.4, 52.3, 58.1, 65.7, 118.9, 119.2, 122.7, 127.7, 128.4, 129.0, 129.4, 130.0, 131.7,
136.8, 138.2, 138.9, 155.9, 168.0, 169.1, 171.1; HR ESI-MS calcd for C27H23N3O7[M-H]- 500.1463，found 500.1455.
TM1c: Methyl (S)-2-(3-(2-(1H-benzo[d]imidazol-1-yl)acetamido)phenyl)-2-(benzyloxycarbonylamino) acetate
Yield 66%, m.p.: 114~115℃, []D20 +17.0(c 1.0 mg/mL, EtOAc). 1H NMR (DMSO-d6, 300 MHz) δ: 3.62(s, 3H,
-COOCH3)，5.05(s, 2H, ArCH2O), 5.20(s, 2H, COCH2), 5.24(d, J=7.1 Hz, 1H, CH), 7.11(d, J=7.4 Hz, 1H, ArH),
7.19~7.35(m, 7H, ArH), 7.52~7.69(m, 5H, ArH), 8.24(s, 1H, -N=CH-), 8.31(d, J=6.9 Hz, 1H, -CONH-), 10.57(s, 1H,
-CONH-); 13C NMR (DMSO-d6, 75 MHz) δ: 30.7, 52.3, 58.0, 65.7, 110.3, 119.3, 122.4, 123.1, 127.8, 128.4, 129.2,
134.4, 136.8, 137.1,138.8, 143.2, 145.0, 155.9, 165.6, 171.0. HR ESI-MS calcd for C26H26N4O5[M-H]- 473.1825，
found 473.1824.
TM1d: Tert-butyl (S)-4-(2-(3-(1-(benzyloxycarbonylamino)-2-methoxy-2-oxoethyl)phenylamino)-2- oxoethyl)
6

piperazine-1-carboxylate
Yield 73%, m.p.: 101~102℃, [] 2D0 +37.0(c 1.0 mg/mL, EtOAc). H NMR (CDCl3, 300 MHz) δ: 1.48(s, 9H,
C(CH3)3), 2.57(s, 4H, CH2 and CH2), 3.15(s, 2H, COCH2)，3.52(s, 4H, CH2 and CH2), 3.73(s, 3H, -COOCH3),
5.04~5.15(m, 2H, ArCH2O), 5.35(d, J=6.7 Hz, 1H, CH), 5.90(d, J=6.3 Hz, 1H, -CONH-), 7.12(d, J=7.1 Hz, 1H, ArH),
7.26(s, 1H, ArH), 7.35(s, 5H, ArH), 7.46(s, 1H, ArH), 7.69(d, J=7.4 Hz, 1H, ArH), 9.05(s, 1H, -CONH-); 13C NMR
(CDCl3, 75 MHz) δ: 28.3, 43.4, 52.9, 53.2, 57.8, 62.1, 67.1, 80.0, 118.1, 119.8, 122.9, 128.2, 128.5, 129.7, 136.0, 137.6,
137.9, 154.6, 155.3, 168.0, 171.0; HR ESI-MS calcd for C28H36N4O7[M-H]- 539.2511，found 539.2511.
TM1e: Methyl (S)-2-(3-(2-(1H-benzo[d]imidazol-2-ylthio)acetamido)phenyl)-2-(benzyloxycarbonyl
-amino)acetate
1
Yield 85%, m.p.: 95~96℃, [] 2D0 +17.0(c 1.0 mg/mL, EtOAc). H NMR (CDCl3, 300 MHz) δ: 3.65(s, 3H,
1
-COOCH3), 3.92(s, 2H, COCH2)，5.01~5.13(m, 2H, ArCH2O), 5.33(d, J=6.9 Hz, 1H, CH), 5.94(d, J=6.9 Hz, 1H,
-CONH-), 7.07(d, J=7.4 Hz, 1H, ArH), 7.22~7.31(m, 9H, ArH), 7.51(d, J=6.0 Hz, 3H, ArH, NH), 7.63(s, 1H, -CONH-);
13C NMR (CDCl3, 75 MHz) δ: 36.1, 52.9, 57.8, 67.2, 118.3, 119.8, 122.7, 128.0, 128.5, 129.6, 135.9, 137.2, 138.9,
150.5, 155.5, 168.2, 171.0. HR ESI-MS calcd for C26H24N4O5S[M-H]- 503.1395，found 503.1401.
4.2.6 General preparation of TM2
TM1 (0.4 mmol), THF (2.0 mL) were added into the 50 mL flask. While stirring and cooling in an ice bath,
added 1 mol/L LiOH solution to adjust pH to 10~11. Stirred the mixture at room temperature, and monitored by TLC
until completion. After adding saturated brine (10 mL), adjusted pH to 5~6 with 2 mol/L hydrochloric acid. Extracted
with EtOAc (3×8 mL), combined the organic layer, dried over anhydrous Na2SO4, then the filtrate was concentrated
under reduced pressure to give pure TM2.
TM2a
：
(S)-2-(((benzyloxy)carbonyl)amino)-2-(3-(2-((4-(N-(5-methylisoxazol-3-yl)sulfamoyl)phenyl)
amino)acetamido)phenyl)acetic acid
Yield 86%, m.p.: 235~236℃, [] 2D0 +19.7(c 1.0 mg/mL, DMF). H NMR (DMSO-d6, 300 MHz) δ: 2.3(s, 3H,
CH3), 4.49(s, 2H, COCH2), 5.06(s, 2H, ArCH2O), 5.22(d, J=7.1 Hz, 1H, CH), 6.25(S, 1H, NH), 6.40(s, 1H, CH), 6.62(d,
J=7.9, Hz 2H, ArH), 7.08(d, J=7.4 Hz, 1H, ArH), 7.30~7.36(m, 6H, ArH), 7.51~7.56(m, 4H, ArH), 7.94(s, 1H,
-CONH-) 8.21(d, J=7.2 Hz, 1H, -CONH-), 10.24(s, 1H, -SO2NH-); 13C NMR (DMSO-d6, 75 MHz) δ: 12.4, 50.2, 59.8,
65.7, 96.7, 112.8, 118.4, 118.8, 122.0, 122.8, 128.4, 129.1, 129.4, 136.8, 139.0, 154.0, 155.9, 159.5, 162.3, 165.2, 170.4,
171.1; HR ESI-MS calcd for C28H27N5O8S [M+Na]+616.1467，found 616.1467.
1
TM2b: (S)-2-(((benzyloxy)carbonyl)amino)-2-(3-(2-(1,3-dioxoisoindolin-2-yl)acetamido)phenyl)acetic acid
Yield 83%, m.p.: 177~178℃, [] 2D0 +29.7(c 1.0 mg/mL, DMF). 1H NMR (DMSO-d6, 300 MHz) δ: 4.01(s, 2H,
COCH2)，5.07(s, 2H, ArCH2O), 5.25(d, J=7.3 Hz, 1H, CH), 7.15(d, J=7.3 Hz, 1H, ArH), 7.32~7.95(m, 12H, ArH),
8.33(d, J=7.4 Hz, 1H, -CONH-), 9.89(s, 1H, -CONH-). 13C NMR (DMSO-d6, 75 MHz) δ: 43.4, 58.1, 65.7, 118.9,
119.2, 122.7, 127.7, 128.4, 129.0, 129.4, 130.0, 131.7, 136.8, 138.2, 138.9, 155.9, 168.0, 169.1, 171.1; HR ESI-MS
7

calcd for C26H21N3O7[M-H]- 486.1301，found 486.1302.
TM2c: (S)-2-(3-(2-(1H-benzo[d]imidazol-1-yl)acetamido)phenyl)-2-(((benzyloxy)carbonyl)amino)
acetic acid
Yield 87%, m.p.: 169~171℃, [] 2D0 +19.9(c 1.0 mg/mL, DMF). 1H NMR (DMSO-d6, 300 MHz) δ: 3.99(s, 2H,
COCH2), 5.05(s, 2H, ArCH2O), 5.13(d, 2H, COCH2), 5.24(d, J=7.1 Hz, 1H, CH) 7.11(d, J=7.8 Hz, 1H, ArH),
7.31~7.85(m, 12H, ArH), 8.10(d, J= 7.8 Hz, 1H, -OCONH-), 8.76(d, J=5.7 Hz, 1H, -N=CH-), 9.86(s, 1H, -CONH-).
13C NMR (DMSO-d6, 75 MHz) δ: 30.7, 58.0, 65.7, 110.3, 119.3, 122.4, 123.1, 127.8, 128.4, 129.2, 134.4, 136.8,
137.1,138.8, 143.2, 145.0, 155.9, 165.6, 171.0. HR ESI-MS calcd for C25H24N4O5[M-H]- 459.1668，found 459.1669.
TM2d:
(S)-2-(((benzyloxy)carbonyl)amino)-2-(3-(2-(4-(tert-butoxycarbonyl)piperazin-1-yl)acetamido)
phenyl)acetic acid
Yield 95%, m.p.: 151~153℃, [] 2D0 +45.0(c 1.0 mg/mL, DMF). 1H NMR (DMSO-d6, 300 MHz) δ: 1.42(s, 9H,
-C(CH3)), 2.04~2.09(m, 4H, 2CH2), 3.02~3.23(m, 4H, 2CH2), 4.02(s, 2H, -COCH2), 5.05(s, 2H, ArCH2O), 5.13(d, J=
7.8 Hz, 1H, CH), 7.16(d, J= 7.2 Hz, 1H, ArH), 7.31~7.36(m, 6H, ArH), 7.59~7.64(m, 2H, ArH), 8.14(d, J= 7.8 Hz, 1H,
13
-OCONH-), 10.48(s, 1H, -CONH-); C NMR (DMSO-d6, 75 MHz) δ: 28.4, 31.1, 52.0, 57.8, 58.5, 66.0, 80.2, 119.4,
119.5, 123.6, 128.1, 128.2, 128.8, 129.4, 137.2, 138.3, 138.5, 153.9, 156.2, 172.3; HR ESI-MS calcd for C27H34N4O7
[M+Na]+ 549.2320, found 549.2327.
TM2e: (S)-2-(3-(2-((1H-benzo[d]imidazol-2-yl)thio)acetamido)phenyl)-2-(((benzyloxy)carbonyl) amin o)acetic
acid
Yield 90%, m.p.: 147~149℃, [] 2D0 +24.0(c 1.0 mg/mL, DMF). H NMR (DMSO-d6, 300 MHz) δ: 4.60(s, 2H,
1
COCH2), 5.04(s, 2H, ArCH2O), 5.09~5.18(m, 2H, CH and -NHC=N-), 7.12(d, 1H, J= 7.5 Hz, ArH), 7.27~7.35(m, 8H,
ArH), 7.42~7.65(m, 4H, ArH), 8.13(d, 1H, J= 7.8 Hz, -OCONH-), 10.46(s, 1H, -CONH-); 13C NMR (CDCl3, 75 MHz)
δ: 36.1, 57.8, 67.2, 118.3, 119.8, 122.7, 128.0, 128.5, 129.6, 135.9, 137.2, 138.9, 150.5, 155.5, 168.2, 171.0. HR
ESI-MS calcd for C25H22N4O5S [M-H]- 489.1233，found 489.1238.
4.2.7 General preparation of TM3
The synthetic conditions of TM3a~TM3d from M5 are the same as TM1.
The preparative procedure of TM3e~TM3j from M4. M4 (0.314 g, 1.0 mmol), acetone (4 ml) and anhydrous
K2CO3 (0.414 g, 3.0 mmol) were added to the 50 mL round bottom flask, cooled in an ice bath, and the RCH2COCl
(2.0 mmol) was added dropwise. Then the reaction mixture was stirred at 20~30℃ and detected by TLC. Upon
completion, the saturated brine was added, the pH of the solution was adjusted to 5 - 6 with 2 mol/L HCl in ice bath.
After the acetone was removed, the residue was extracted by EtOAc (3×10 mL). The combined organic phase was
washed with saturated brine (3×10 mL), dried over anhydrous Na2SO4, then the crude product was obtained by vacuum
distillation, and the purified TM3e - TM3j was obtained by silica gel column chromatography.
TM3a:
Methyl
(S)-2-(benzyloxycarbonylamino)-2-(3-(2-(5-(methylthio)-1H-tetrazol-1-yl)acetamido)
phenyl)acetate
8

Yield 66%, m.p.: 182~183℃, [] 2D0 +11.0(c=1.0 mg /mL, EtOAc). 1H NMR (CDCl3, 300 MHz) δ: 3.74 (s, 3H,
-COOCH3), 3.98(s, 3H, SCH3), 4.06(s, 2H, ArCH2O), 5.15(s, 2H, CH2CO), 5.35(d, J=7.01 Hz, 1H, -CONH-), 5.90(s,
1H, CH), 7.13(d, J=7.3 Hz, 2H, ArH), 7.28~7.36(m, 5H, ArH), 7.51(d, J=7.8 Hz, 1H, ArH), 7.60(s, 1H, ArH), 9.45(s,
1H, -CONH-); 13C NMR (DMSO-d6, 75 MHz) δ: 31.1, 43.9, 52.7, 58.4, 66.1, 119.2, 119.6, 123,6, 128.2, 128.3, 128.8,
129.6, 137.2, 137.4, 138.9, 156.2, 165.2, 171.4; HRMS calcd for C21H22N6O5S[M-H]- 469.1300, found 469.1297.
TM3b: Methyl (S)-2-(benzyloxycarbonylamino)-2-(3-(2-(5-methyl-1H-tetrazol-1-yl)acetamido)
phenyl) acetate
Yield 55%, m.p.: 156~157℃, [] 2D0 +24.0(c 1.0 mg/mL, EtOAc). 1H NMR (DMSO-d6, 300 MHz) δ: 2.57(s, 3H,
CH3), 3.76(s, 3H, -COOCH3)，5.14(s, 4H, ArCH2O and COCH2), 5.31(d, J=6.3 Hz, 1H, CH), 6.18 (d, J=6.3 Hz, 1H,
-CONH-), 7.12(d, J=7.2 Hz, 1H, ArH), 7.20~7.51(m, 7H, ArH), 7.62(s, 1H, ArH), 8.84(s, 1H, -CONH-); 13C NMR
(DMSO-d6, 75 MHz) δ: 8.8, 49.8, 53.0, 57.9, 67.2, 118.7, 120.1, 123,4, 127.9, 128.2, 128.5, 129.7, 136.0, 137.6, 137.7,
155.6, 162.7, 170.8; HRMS calcd for C21H22N6O5 [M+ Na]+ 461.1544, found 461.1546.
TM3c: Methyl (S)-2-(3-(2-(1H-tetrazol-1-yl)acetamido)phenyl)-2-(((benzyloxy)carbonyl)amino)
acetate
Yield 53%, m.p.: 130~131℃, [] 2D0 +33.0(c 1.0 mg/mL, Acetone). 1H NMR (DMSO-d6, 300 MHz) δ: 3.63(s, 3H,
-COOCH3)，5.06(s, 2H, ArCH2O), 5.24(d, J=7.5 Hz, 1H, CH), 5.49(s, 2H, COCH2), 7.13(d, J=7.5 Hz, 1H, ArH),
7.31~7.36(m, 6H, ArH), 7.57(d, J=9.3 Hz, 2H, ArH), 8.33(d, J=7.5 Hz, 1H, -CONH-), 9.43(s, 1H, -N=CH-), 10.65(s,
1H, -CONH-); 13C NMR (DMSO-d6, 75 MHz) δ: 50.5, 52.8, 58.3, 60.2, 66.2, 119.1, 119.5, 123.7, 128.1, 128.2, 128.3,
128.7, 129.7, 137.2, 137.5, 138.8, 145.7, 156.3, 164.0, 171.4; HRMS calcd for C20H20N6O5[M-H]- 423.1422, found
423.1426.
TM3d:
Methyl
(S)-2-(((benzyloxy)carbonyl)amino)-2-(3-(2-(3,5-dimethyl-1H-pyrazol-1-yl)acetamido)
phenyl)acetate
Yield 32%, m.p.: 135~136℃, []D20 +17.0(c 1.0 mg/mL, EtOAc). 1H NMR (DMSO-d6, 300 MHz) δ: 3.62(s, 3H,
-COOCH3)，5.05(s, 2H, ArCH2O), 5.20(s, 2H, COCH2), 5.24(d, J=7.1 Hz, 1H, CH), 7.11(d, J=7.4 Hz, 2H, ArH),
7.19~7.35(m, 7H, ArH), 8.24(s, 1H, CH), 8.31(d, J=6.9 Hz, 1H, -CONH-), 10.57(s, 1H, -CONH-); 13C NMR
(DMSO-d6, 75 MHz) δ: 30.7, 52.3, 58.0, 65.7, 110.3, 119.3, 122.4, 123.1, 127.8, 128.4, 129.2, 134.4, 136.8,
137.1,138.8, 143.2, 145.0, 155.9, 165.6, 171.0. HRMS calcd for C24H26N4O5 [M-H]- 449.1825，found 449.1826.
TM3e: Methyl (S)-2-(3-acetamidophenyl)-2-(benzyloxy)carbonyl)amino)acetate
Yield 65%, m.p.: 127~128℃.1H NMR (DMSO-d6, 300 MHz) δ: 3.03(s, 3H, -COCH3), 3.63(s, 3H, -COOCH3),
5.06(s, 2H, ArCH2O), 5.20(d, J=7.5Hz, 1H, CH), 7.05(d, J=7.8Hz, 1H, ArH), 7.25~7.36(m, 6H, ArH), 7.57 (t, J=8.1Hz,
2H, ArH), 8.29(d, J=7.5Hz, 1H, -CONH-), 10.01(s, 1H, -CONH-); 13C NMR (DMSO-d6, 75 MHz) δ: 24.4, 52.7, 58.5,
66.1, 118.9, 119.3, 122.7, 128.2, 128.3, 128.8, 129.4, 137.1, 137.2, 140.0, 156.3, 168.8, 171.5. HRMS calcd for
C19H20N2O5[M+Na]+ 379.1264, found 379.1271.
9

TM3f: Methyl (S)-2-(((benzyloxy)carbonyl)amino)-2-(3-propionamidophenyl)acetate
20
Yield 53%, m.p.: 136~137℃, [] D =+26.0°(c=1.0 mg/mL, EtOAc). 1H NMR (CDCl3, 300 MHz) δ: 1.25(t, J = 4.2
Hz, 3H, CH2CH3), 2.35-2.43(m, 2H, CH2CH3), 3.79(s, 3H, CH3), 5.09(s, 2H, ArCH2O), 5.33(d, J = 4.2 Hz, 1H, CH),
7.29(s, 1H, ArH) 7.35(s, 5H, ArH), 7.46-7.52 (m, 2H, ArH), 7.85(d, J = 7.2 Hz, 1H, ArH). 13C NMR (CDCl3, 75 MHz)
δ: 29.7, 52.9, 57.9, 67.1, 118.4, 120.0, 122.6, 127.8, 128.2, 128.5, 129.6, 135.9, 137.3, 138.7, 155.4, 171.1, 172.3.
HRMS calcd for C20H22N2O5[M+Na] +393.1421, found 393.1427.
TM3g: Methyl (S)-2-(((benzyloxy)carbonyl)amino)-2-(3-butyramidophenyl)acetate
20
1
Yield 51%, m.p.: 140~141℃, [] D =+33.0°(c=1.0 mg/mL, EtOAc). H NMR (DMSO-d6, 300 MHz) δ: 0.90(t, J
= 7.2 Hz, 3H, CH2CH3), 1.55-1.63(m, 2H, CH2CH3), 2.26(t, J = 7.2 Hz, 2H, -CH2CH2CO-), 3.62(s, 3H, OCH3), 5.05(s,
2H, ArCH2O), 5.19(d, J=7.2 Hz, 1H, CH), 7.04(d, J = 7.5 Hz, 1H, ArH) 7.35(m, 6H, ArH), 7.57(m, 2H, ArH), 8.26(d, J
= 7.5 Hz, 1H, CONH), 9.94(s, 1H, CONH). 13C NMR (DMSO-d6, 75 MHz) δ: 14.1, 19.0, 38.7, 52.7, 58.5, 66.1, 119.3,
122.7, 128.2, 128.3, 128.8, 129.4, 137.1, 137.2, 139.9, 156.4, 171.5, 171.7, 118.9. HRMS calcd for C21H24N2O5
[M+Na]+407.1577, found 407.1574.
TM3h: Methyl (S)- 2-(((benzyloxy)carbonyl)amino)-2-(3-(3-chlorobutanamido)-phenyl)acetate
20
Yield 65%, m.p.: 152~154℃, [] D =+22.0°(c=1.0 mg/mL, EtOAc). 1H NMR (CDCl3, 300 MHz) δ: (CDCl3, 300
MHz) δ: 2.78(t, J = 6.3 Hz, 2H, -COCH2), 3.71(s, 3H, -OCH3), 3.86(t, J = 6.3 Hz, 2H, -CH2Cl), 5.09(d, J = 4.2Hz, 2H,
ArCH2O), 5.33(d, J=6.9 Hz, 1H, CH), 7.10(d, J = 7.8 Hz, 1H, ArH), 7.29-7.33(m, 5H, ArH), 7.48(s,1H, ArH), 7.55(d, J
= 7.5 Hz, 1H, ArH), 7.70(s, 1H, ArH). 13C NMR (CDCl3, 75 MHz) δ: 39.8, 40.2, 52.9, 57.9, 67.2, 118.7, 120.2, 122.9,
128.1, 128.2, 128.5, 129.6, 135.9, 137.3, 138.4, 155.5, 168.2, 171.1. HRMS calcd for C20H21ClN2O5 [M+Na]+
427.1035, found 427.1031
TM3i: Methyl (S)-2-(((benzyloxy)carbonyl)amino)-2-(3-(4-chloropropanamido)phenyl)acetate
20
D
Yield 62%, m.p.: 158~160℃, []
=+59.0°(c=1.0 mg/mL, EtOAc). 1H NMR δ: 2.17-2.24(m, 2H,
-CH2CH2CH2-), 2.52(t, J = 7.8 Hz, 2H, -COCH2), 2.67(t, J = 5.4 Hz, 2H, -CH2Cl), 3.72(s, 3H, OCH3), 5.08(d, J=3.3 Hz,
2H, ArCH2O), 5.35(d, J = 9.6 Hz, 1H, -CH), 7.09(s, 1H, ArH), 7.29-7.34(m, 6H, ArH), 7.49(s, 1H, ArH), 7.58(m, 1H,
ArH). 13C NMR (CDCl3, 75 MHz) δ: 26.8, 33.2, 42.8, 52.5, 57.6, 66.8, 118.7, 120.3, 122.8, 128.1, 128.2, 128.5, 129.6,
136.0, 137.3, 138.6, 156.2, 171.5, 175.8. HRMS calcd for C21H23ClN2O5 [M+Na]+ 441.1182, found 441.1184.
TM3j: Methyl (S)-2-(((benzyloxy)carbonyl)amino)-2-(3-(2-methoxyacetamido)phenyl)acetate
Yield 52%, m.p.: 116~117℃.1H NMR (DMSO-d6, 300 MHz) δ: 3.37(s, 3H, -OCH3), 3.64(s, 3H, -COOCH3),
3.99(s, 2H, COCH2O), 5.06(s, 2H, ArCH2O), 5.23(d, J=7.5Hz, 1H, CH), 7.09(d, J=7.5Hz, 1H, ArH), 7.27~7.36(m, 5H,
ArH), 7.64 (t, J1=7.5, 3H, ArH), 8.28(d, J=7.8, 1H, -CONH-), 9.82(s, 1H, -CONH-). 13C NMR (CDCl3, 75 MHz) δ:
52.7, 59.1, 66.1, 72.1, 119.7, 119.9, 120.3, 123.3, 128.1, 128.2, 128.3, 128.6, 128.8, 129.3, 131.5, 137.2, 139.0, 156.3,
168.6, 171.5. HRMS calcd for C20H22N2O6[M+Na]+ 409.1365, found 409.1366.
4.2.8 Preparation of M6
10

Preparation of M6b~M6d: SOCl2 (5.3 mmol) was added dropwise into alcohols (15 mL) in an ice salt bath. After
stirring for 0.5 h, M2 (5.0 mmol) was added into the above solution. The resulting mixture was refluxed at 70℃ and
monitored by TLC until completion. The crude product was obtained through concentrating under reduced pressure,
followed by recrystallization with alcohol and petroleum ether to give white solid M6b~M6d.
Preparation of M6e~ M6h: To a solution of M2 (0.5 mmol) and HOBt (0.6 mmol) in dried THF (2 mL) which
was cooled to 0℃, dicyclohexylcarbodiimide (DCC, 0.6 mmol) and DIPEA (0.25 mmol) were added. The mixture was
stirred in an ice-water bath for 0.5 h and amines (0.6 mmol) was added. The resulting reaction mixture was stirred at
room temperature and monitored by TLC until completion. Filtered by suction and washed the filter cake with cold
EtOAc (3×3 mL).The combined filtrate was concentrated under reduced pressure, the obtained residue was dissolved
by EtOAc (20 mL), and then washed with 0.5N NaHCO3 (10 mL×3), 10% citric acid (10 mL×3) and saturated brine
(10 mL) sequentially. The organic solution was dried over Na2SO4, concentrated in vacuo, the residue was purified
through silica gel column chromatography to give M6e~ M6h.
M6b: Ethyl (S)-2-(benzyloxycarbonylamino)-2-(3-nitrophenyl)acetate
Yield 85%, m.p.: 109~110℃.1H NMR (DMSO-d6, 300 MHz) δ: 1.12(t, J=6.8 Hz, 3H, CH3), 4.10~4.12(m, 2H,
CH2), 5.07(s, 2H, ArCH2O), 5.53(m, 1H, CH), 7.35(s, 5H, ArH), 7.67(t, J=7.9 Hz, 1H, ArH), 7.89(d, J=7.3 Hz, 1H,
ArH,), 8.20(d, J=8.2 Hz, 1H, -CONH-), 8.32(s, 1H, ArH), 8.50(d, J=7.6 Hz, 1H, ArH); 13C NMR (DMSO-d6, 75 MHz)
δ: 14.3, 57.6, 61.8, 66.3, 123.0, 123.5, 128.2, 128.3, 128.8, 130.5, 135.2, 137.1, 139.2, 148.2, 156.3, 170.3.
M6c: Propyl (S)-2-(benzyloxy)carbony)lamino)-2-(3-nitrophenyl)acetate
Yield 88%, m.p.: 106~107℃.1H NMR (DMSO-d6, 300 MHz) δ: 0.77(t, J=6.9 Hz, 3H, CH3), 1.48~1.53(m, 2H,
CH2), 4.04(s, 2H, OCH2), 5.09(s, 2H, ArCH2O), 5.57(d, J=7.5Hz, 1H, CH), 7.36(s, 5H, ArH), 7.66~7.71(m, 1H, ArH),
7.91(d, J=7.2 Hz, 1H, ArH), 8.21(d, J=8.1 Hz, 1H, -CONH-), 8.35(s, 1H, ArH), 8.52(d, J=7.5 Hz, 1H, ArH); 13C NMR
(DMSO-d6, 75 MHz) δ: 10.4, 21.8, 57.6, 66.3, 67.1, 123.0, 123.5, 128.2, 128.3, 128.8, 130.4, 135.2, 137.1, 139.3,
148.1, 156.3, 170.3.
M6d: Butyl (S)-2-(benzyloxycarbonylamino)-2-(3-nitrophenyl)acetate
Yield 88%, m.p.: 132~133℃.1H NMR (DMSO-d6, 300 MHz) δ: 0.79(t, J1=7.2 Hz, 3H, CH3), 1.16~1.23 (m, 2H,
CH2), 1.46~1.50(m, 2H, CH2), 4.05~4.09(m, 2H, OCH2), 5.06(s, 2H, ArCH2O), 5.56(d, J=8.1 Hz, 1H, CH), 7.37(s, 5H,
ArH), 7.66~7.71(m, 1H, ArH), 7.91(d, J=7.5 Hz, 1H, ArH), 8.21(d, J=8.4 Hz, 1H, -CONH-), 8.35(s, 1H, ArH), 8.52(d,
13
J=8.1 Hz, 1H, ArH); C NMR (DMSO-d6, 75 MHz) δ: 13.8, 18.8, 30.4, 57.6, 65.4, 66.3, 123.0, 123.5, 128.2, 128.3,
128.8, 130.4, 135.2, 137.1, 139.3, 148.1, 156.3, 170.3.
M6e: Benzyl (S)-(2-((methylamino)oxy)-1-(3-nitrophenyl)-2-oxoethyl)carbamate
Yield 90%, m.p.: 113~114℃.1H NMR (DMSO-d6, 300 MHz) δ: 2.59(d, J=4.2 Hz, 3H, CH3), 5.04(s, 2H,
ArCH2O), 5.38(d, J= 7.8 Hz, 1H, CH), 7.36(s, 5H, ArH), 7.62~7.70(m, 2H, ArH), 7.87(d, J=7.5 Hz, 1H, -CONH-),
13
8.15~8.25(m, J= 8.4 Hz, 2H, ArH), 8.33(d, 1H, -CONH-). C NMR (DMSO-d6, 75 MHz) δ: 33.8, 57.9, 66.1, 122.4,
11

123.8, 128.1, 128.2, 128.7, 128.9, 130.3, 134.4, 137.2, 137.3, 148.1, 156.1, 169.7.
M6f: Benzyl (S)-(2-((ethylamino)oxy)-1-(3-nitrophenyl)-2-oxoethyl)carbamate
Yield 88%, m.p.: 116~117℃.1H NMR (DMSO-d6, 300 MHz) δ: 0.99(t, J=7.2 Hz, 3H, CH3), 3.03~3.13(m, 2H,
CH2), 5.04(s, 2H, ArCH2O), 5.38(d, J=8.4 Hz, 1H, CH), 7.31~7.37(m, 5H, ArH), 7.62~7.68(m, 1H, ArH), 7.87(t, J=8.1
Hz, 1H, -CONH-), 8.15(d, J=8.6 Hz, 1H, ArH), 8.24(d, J=8.7 Hz, 1H, ArH), 8.33(s, 1H, ArH), 8.40(d, J=8.3 Hz, 1H,
-CONH-). 13C NMR (DMSO-d6, 75 MHz) δ: 12.1, 33.8, 57.9, 66.1, 122.4, 123.8, 128.1, 128.2, 128.7, 128.9, 130.3,
134.4, 137.2, 137.3, 148.1, 156.1, 169.7.
M6g: Benzyl (S)-(2-((hydroxyamino)oxy)-1-(3-nitrophenyl)-2-oxoethyl)carbamate
Yield 67%, m.p.: 125~126℃. 1H NMR (DMSO-d6, 300 MHz) δ: 5.04(s, 2H, CH2), 5.30(d, J=8.4 Hz, H, CH),
7.36(s, 5H, ArH), 7.65~7.71(m, J=7.8 Hz, H, ArH), 7.88(d, J=7.8 Hz, H, NH), 8.16~8.25(dd, J=8.4, 9.6 Hz, H, ArH),
8.38(t, J=8.7 Hz, 2H, ArH), 9.13(s, H, NH), 11.08(s, H, OH). 13C NMR (DMSO-d6, 75 MHz) δ: 55.8, 66.2, 122.2,
123.2, 124.0, 128.1, 128.3, 128.8, 130.4, 134.3, 137.2, 141.1, 148.1, 156.1.
M6h: Benzyl (S)-(2-(((2-hydroxyethyl)(methyl)amino)oxy)-1-(3-nitrophenyl)-2-oxoethyl)carbamate
20
D
1
Yield 86%, m.p.: 133~135℃, [] +39.8(c=1.0 mg/mL, EtOAc); H NMR (DMSO-d6, 300 MHz) δ:
2.87~2.99(m, 4H, N(CH3)CH2CH2OH and N(CH3)CH2CH2OH), 3.51~3.52(m, 3H, CH3), 5.04(s, 2H, ArCH2O),
5.79~5.96(dd, J=7.8, 7.8 Hz, 1H, CH), 7.35(s, 4H, ArH), 7.67 (d, J=6.0 Hz, 1H, ArH), 7.85(d, J=7.2 Hz, 1H, ArH),
13
8.10(d, J=8.1 Hz, 1H, -CONH-), 8.16~8.23(m, 2H, ArH), 8.27 (s, 1H, ArH). C NMR (DMSO-d6, 75 MHz) δ: 36.4,
54.5, 55.2, 58.8, 66.2, 123.0, 123.9, 128.1, 128.3, 128.8, 129.6, 135.1, 137.2, 155.9, 169.0, 169.6.
4.2.9 Preparation of TM4b~TM4h
The mixture of M6 (1.0 mmol), Zn powder (11.0 mmol), NH4Cl (6.0 mmol) and MeOH (15 mL) was stirred at
75~85 ℃ and monitored by TLC until completion. After hot filtration and washing residue with MeOH (2×2 mL), the
filtrate was concentrated under reduced pressure to obtain the crude product.
The mixture of the crude product (0.9 mmol), dried acetone (5 mL) and K2CO3 (2.7 mmol) were added into the
50 mL round bottom flask. While the resultant mixture was stirred in an ice bath, the acetone solution (3 mL) of
chloroacetyl chloride (0.97 mmol) was dripped slowly from the constant pressure funnel. After 5 min, removed the ice
bath, stirred the reaction solution at 15℃-18℃and monitored the reaction process by TLC until completion. Filtered by
suction and washed the residue with acetone. The filtrate was concentrated under reduced pressure to obtain crude
product, and followed the purification by silica gel column chromatography to obtain the pure TM4b~TM4h.
TM4b: Ethyl (S)-2-(benzyloxycarbonylamino)-2-(3-(2-chloroacetamido)phenyl)acetate
Yield 65%, m.p.: 83~84℃.1H NMR (DMSO-d6, 300 MHz) δ: 1.13(t, J=6.9 Hz, 3H, CH3), 4.06~4.13(m, 2H,
CH2), 4.25(s, 2H, COCH2Cl), 5.07(s, 2H, ArCH2O), 5.21(d, J=7.5 Hz, 1H, CH), 7.12(s, 1H, ArH), 7.30~ 7.36(m, 6H,
ArH), 7.59(d, J=6.6 Hz, 2H, ArH), 8.28(d, J=7.5 Hz, 1H, -CONH-), 10.38(s, 1H, -CONH-); 13C NMR (DMSO-d6, 75
MHz) δ: 14.4, 44.0, 58.5, 61.4, 66.1, 119.2, 119.6, 123.6, 128.2, 128.3, 128.8, 129.5, 137.3, 137.4, 139.1, 156.1, 165.1,
170.9; HRMS calcd for C20H21ClN2O5[M-H]- 403.1066, found 403.1072.
12

TM4c: Propyl (S)-(benzyloxycarbonylamino)-2-(3-(2-chloroacetamido)phenyl)acetate
Yield 70%, m.p.: 95~96℃.1H NMR (DMSO-d6, 300 MHz) δ: 0.74~0.79(m, 3H, CH3), 1.49~1.56(m, 2H, CH2),
4.02(t, J=6.0 Hz, 2H, OCH2), 4.25(s, 2H, CH2Cl), 5.07(s, 2H, ArCH2O), 5.22(d, J=7.2 Hz, 1H, CH), 7.14(d, J=7.5 Hz,
1H, ArH), 7.31~7.36(m, 6H, ArH), 7.58(d, J=14.1 Hz, 1H, ArH), 8.29(d, J=7.5 Hz, 1H, -CONH-), 10.39(s, 1H,
-CONH-); 13C NMR (DMSO-d6, 75 MHz) δ: 10.4, 21.8, 44.0, 58.3, 66.1, 66.8, 119.6, 128.1, 128.3, 128.7, 129.5, 137.2,
137.4, 139.0, 156.4, 165.1, 170.9. HRMS calcd for C21H23ClN2O5 [M-H]- 417.1223，found 417.1221.
TM4d: Butyl (S)-2-(benzyloxycarbonylamino)-2-(3-(2-chloroacetamido)phenyl)acetate
Yield 62%, m.p.: 114~115℃.1H NMR (DMSO-d6, 300 MHz) δ: 0.77~0.82(t, J=7.5 Hz, 3H, CH3) , 1.16~ 1.26(m,
2H, CH3CH2), 1.44~1.53(m, 2H, CH3CH2CH2), 4.04~4.08(m, 2H, OCH2), 4.25(s, 2H, CH2Cl), 5.11(s, 2H, ArCH2O),
5.21(d, J=4.2 Hz, 1H, CH), 7.12(d, J=7.8 Hz, 1H, ArH), 7.30~7.36(m, 5H, ArH), 7.59(s, 1H, ArH), 8.28 (d, J=7.5 Hz,
1H, -CONH-), 10.38(s, 1H, -CONH-); 13C NMR (DMSO-d6, 75 MHz) δ: 13.9, 18.2, 30.4, 44.0, 58.6, 65.0, 66.1, 119.2,
119.5, 123.6, 128.2, 128.3, 129.5, 137.3, 139.1, 156.3, 165.1, 171.0. HRMS calcd for C22H25ClN2O5[M-H]- 431.1379，
found 431.1376.
TM4e: Benzyl (S)-(1-(3-(2-chloroacetamido)phenyl)-2-( methyl amino)-2-oxoethyl)carbamate
Yield 27%, m.p.: 101~102℃.1H NMR (DMSO-d6, 300 MHz) δ: 2.58(d, J=4.2 Hz, 3H, CH3), 4.24(s, 2H, CH2Cl),
5.03(s, 2H, ArCH2O), 5.15(d, J=7.8 Hz, 1H, CH), 7.12(d, J=7.8 Hz, 1H, ArH), 7.26~7.35(m, 6H, ArH), 7.54~7.58(m,
2H, ArH), 7.85(d, J=8.1 Hz, 1H, ArH), 8.18(d, J=4.5 Hz, 1H, -CONH-), 10.35(s, 1H, -CONH-); 13C NMR (DMSO-d6,
75 MHz) δ: 26.1, 44.0, 58.4, 65.8, 118.8, 119.1, 119.7, 123.3, 128.2, 128.8, 129.2, 134.5, 138.8, 139.9, 156.0, 165.1,
170.1. HRMS calcd for C19H20ClN3O5[M-H]- 404.1013，found 404.1016.
TM4f: Benzyl (S)-(1-(3-(2-chloroacetamido)phenyl)-2-(ethylamino)-2-oxoethyl)carbamate
Yield 46%, m.p.: 179~181℃.1H NMR (DMSO-d6, 300 MHz) δ: 1.00(t, J=7.2 Hz, 3H, CH3), 3.02~3.10(m, 2H,
-CH2-), 4.26(s, 2H, CH2Cl), 5.03(s, 2H, ArCH2O), 5.15(d, J=7.8 Hz, 1H, CH), 7.12(d, J=7.8 Hz, 1H, ArH),
7.26~7.37(m, 6H, ArH), 7.55~7.58(m, 2H, ArH), 7.81(d, J=8.7 Hz, 1H, ArH), 8.26(d, J=8.4 Hz, 1H, -CONH-), 10.26(s,
1H, -CONH-,); 13C NMR (DMSO-d6, 75 MHz) δ:14.9, 34.0, 44.0, 58.8, 66.0, 118.7, 119.1, 123.1, 128.1, 128.2, 129.2,
137.3, 140.0, 156.2, 165.1, 169.6. HRMS calcd for C20H22ClN3O5[M-H]- 418.1170，found 418.1173.
TM4g: Benzyl (S)-(1-(3-(2-chloroacetamido)phenyl)-2-((hydroxyamino)oxy)-2-oxoethyl)carbamate
Yield 56%, m.p.: 103~105℃.1H NMR (DMSO-d6, 300 MHz) δ: 4.24(s, 2H, CH2Cl), 5.04(s, 2H, ArCH2O),
5.14(d, J=8.1, 1H, CH), 7.15(d, J=7.5, 1H, ArH), 7.21(s, 1H, ArH), 7.27~7.35(m, 5H, ArH), 7.56~7.60(m, 2H, Ar-H),
7.74(d, J=7.8, 1H, -CONH-), 10.35(s, 1H, -CONH-). 13C NMR (DMSO-d6, 75 MHz) δ: 42.6, 55.8, 66.2, 122.2, 123.2,
124.0, 128.1, 128.3, 128.8, 130.4, 134.3, 137.2, 141.1, 148.1, 156.1, 165.4. HRMS calcd for C18H18ClN3O6[M-H]-
406.0806，found 406.0810.
TM4h: Benzyl (S)-(1-(3-(2-chloroacetamido)phenyl)-2-(((2-hydroxyethyl)(methyl)amino)oxy)-2-oxo
ethyl)carbamate
13

1
Yield 38%, m.p.:103~104℃. H NMR (DMSO-d6, 300 MHz) δ:2.85~2.91(m, 4H, N(CH3)CH2 CH2OH and
N(CH3)CH2CH2OH), 3.45~3.48(m, 3H, CH3), 4.24(s, 2H, COCH2Cl), 5.02(s, 2H, ArCH2O), 5.52~5.67(dd, J=7.8, 7.8
Hz, 1H, CH), 7.09(d, J=7.5, 1H, ArH), 7.29~7.34(m, 6H, ArH), 7.51 (d, J=7.2 Hz, 1H, ArH), 7.60~7.65(m, 1H, ArH),
7.69(s, J=7.5 Hz, 1H, -CONH-), 10.37(s, 1H, -CONH-); 13C NMR (DMSO-d6, 75 MHz) δ: 34.4, 36.2, 44.0, 49.1, 50.2,
58.8, 66.0, 119.0, 124.0, 128.1, 128.2, 128.8, 129.5, 137.4, 139.1, 139.2, 156.1, 165.1, 170.0; HRMS calcd for
C21H24ClN3O6 [M-H]- 448.1275，found 448.1278.
4.2.10 Preparation of M7 and TM4a
M2 (6.3 mmol), Zn powder (69.2 mmol), CaCl2 (5.0 mmol), and the water solution of 95% EtOH (16 mL,
volume ratio 4:1) were added into a 50 mL round bottom flask. Stirred and refluxed continuously at 80 ℃. The reaction
progress was monitored by TLC. Upon completion, the solid was collected by filtration. The obtained solid was washed
with EtOH (3 × 10 mL), the combined filtrate was concentrated under reduced pressure to obtain crude product.
Subsequently, the crude product was washed thoroughly with water (3 x 5 mL), and then dried in vacuo overnight to
afford the pure M7.
Synthesis of TM4a: M7 (5.1 mmol), DCM (15 mL), anhydrous K2CO3 (36.4 mmol) was added into a 50 mL
round bottom flask. While the mixture was stirred in an ice bath, the solution of chloroacetyl chloride (10.1 mmol) in
DCM (5 mL) was dripped slowly from the constant pressure funnel. Then the mixture was stirred at ambient
temperature and monitored by TLC. On completion of the reaction, cool saturated brine (20 mL) was added, the solid
was subsequently collected by filtration. The filter cake obtained was dried overnight to afford the crude product,
further recrystallization from mixed solvent of acetone and chloroform (6 mL, volume ratio 1:1) affords analytically
pure TM4a.
M7: (S)-2-(((benzyloxy)carbonyl)amino)-2-(3-nitrophenyl)acetic acid
Yield 71%, m.p.: 77~78℃, [] 2D0 +81.36(c 2.2 mg/mL, DMF). H NMR (DMSO-d6, 300 MHz) δ:4.68(s, 1H,
1
CH), 4.90(s, 2H, NH2), 5.03(s, 2H, CH2), 6.38(d, J1=7.5 Hz, 2H ArH), 6.43~7.34(m, 7H, ArH); 13C NMR (DMSO-d6,
75 MHz) δ: 57.9, 65.5, 113.6, 115.0, 121.5, 127.7, 127.8, 128.4, 129.0, 137.0, 137.5, 151.9, 155.9, 172.2.
TM4a: Potassium (S)-2-(((benzyloxy)carbonyl)amino)-2-(3-(2-chloroacetamido)phenyl)acetate
1
Yield 57%, m.p.: 79~80℃. H NMR (DMSO-d6, 300 MHz) δ:4.25(s, 1H, CH2), 5.08(d, J=7.5 Hz, 2H, CH2),
5.25(s, 1H, CH), 7.12(d, J=7.5 Hz, 1H, ArH), 7.33(t, J=8.4 Hz, 6H, ArH), 7.63(d, J=8.6 Hz, 2H, ArH), 8.32(d, J=8.4
Hz, 1H, NH), 10.40(s, 1H, NH); 13C NMR (DMSO-d6, 75 MHz) δ: 43.6, 58.0, 65.8, 118.8, 119.3, 123.3, 127.8, 127.9,
128.4, 129.2, 136.9, 137.0, 138.7, 155.9, 164.8, 171.1. HRMS calcd for C18H16ClKN2O5 [M-H]- 413.0307，found
413.0310.
4.3．Biology
4.3.1 The determination of α-glucosidase inhibition [16]
α-Glucosidase inhibitory activity was determined in a 100 μL reaction mixer containing optimal
rat-α-glucosidase(extracted from the rat small intestine of rat), 67 mmol L-1 pH 6.8 sodium phosphate buffer and
14