Substituted benzothiophene and benzofuran derivatives as a novel class of bone morphogenetic Protein-2 upregulators: Synthesis, anti-osteoporosis efficacies in ovariectomized rats and a zebrafish model, and ADME properties

Article abstract of DOI:10.1016/j.ejmech.2020.112465

The bone morphogenetic protein (BMP) pathway is a promising new target for the design of therapeutic agents for the treatment of low bone mass. This study optimized the structure of the anti-osteoporosis compound 38 by balancing its lipophilicity and impr

Full text of DOI:10.1016/j.ejmech.2020.112465

Journal Pre-proof
Substituted benzothiophene and benzofuran derivatives as a novel class of bone
morphogenetic Protein-2 upregulators: Synthesis, anti-osteoporosis efficacies in
ovariectomized rats and a zebrafish model, and ADME properties
Si-tu Xue, Lei Zhang, Zhuo-song Xie, Jie Jin, Hui-Fang Guo, Hong Yi, Zong-ying Liu,
Zhuo-rong Li
PII:
S0223-5234(20)30436-0
DOI:
https://doi.org/10.1016/j.ejmech.2020.112465
EJMECH 112465
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 9 March 2020
Revised Date: 28 April 2020
Accepted Date: 12 May 2020
Please cite this article as: S.-t. Xue, L. Zhang, Z.-s. Xie, J. Jin, H.-F. Guo, H. Yi, Z.-y. Liu, Z.-r. Li,
Substituted benzothiophene and benzofuran derivatives as a novel class of bone morphogenetic
Protein-2 upregulators: Synthesis, anti-osteoporosis efficacies in ovariectomized rats and a zebrafish
model, and ADME properties, European Journal of Medicinal Chemistry (2020), doi: https://
doi.org/10.1016/j.ejmech.2020.112465.
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.
©
2020 Published by Elsevier Masson SAS.

Substituted Benzothiophene and Benzofuran Derivatives as a Novel Class of
Bone Morphogenetic Protein-2 Upregulators: Synthesis, Anti-osteoporosis
Efficacies in Ovariectomized Rats and a Zebrafish Model, and ADME
Properties
†
†
Si-tu Xue , Lei Zhang , Zhuo-song Xie, Jie Jin, Hui-Fang Guo, Hong Yi, Zong-ying Liu*,
*
Zhuo-rong Li
Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences & Peking Union
Medical College, Beijing 100050, China
ABSTRACT
The bone morphogenetic protein (BMP) pathway is a promising new target for the design of
therapeutic agents for the treatment of low bone mass. This study optimized the structure of the
anti-osteoporosis compound 38 by balancing its lipophilicity and improving its stability. Twenty
derivatives which were not reported in the literature were designed and synthesized. The
ovariectomized rat model of osteoporosis was selected to evaluate the therapeutic effects.
Compound 125 showed better therapeutic efficacy than that of 38. We verified the
anti-osteoporosis activity and BMP-2 protein upregulation after treatment with 125 in a zebrafish
1

osteoporosis model. We found that 125 improved the ADME properties, therapeutic efficacy,
and pharmacokinetics of the drug. Overall, we evaluated the anti-osteoporosis effects of the
compounds of this type, preliminarily determined the target patient population, verified the
mechanism of action, clarified the level of toxicity, and provided preliminary ADME data. We
believe that these compounds can both correct bone loss that is already occurring in patients and
have broad clinical applicability.
KEYWORDS
Bone morphogenetic protein-2, Osteoporosis, Osteogenesis, Benzothiophene, Benzofuran
2

1
INTRODUCTION
Osteoporosis is a systemic disease in which the rigidity and mechanical stability of bones
decrease. Bone health in an adult involves the continual removal of old or damaged bone tissue
by osteoclasts (OCs) through a process called resorption, which is balanced by the equivalent
(
1,2)
deposition of new bone by osteoblasts (OBs).
As the skeleton ages, the action of OCs
outpaces that of OBs, leading to a shift toward excess resorption and development of the chronic
degenerative condition osteoporosis; this reduces bone mineral density (BMD) and increases the
(
3,4)
risk of bone fractures.
Osteoporosis has classically been divided into primary
(postmenopausal women and people over 70 years old), secondary (caused at least in part by
(
4,5)
other diseases or therapies), and idiopathic forms.
The most common forms of primary
osteoporosis occur as a result of menopause and aging. Secondary osteoporosis can result from,
for example, the chronic use of glucocorticoids and other drugs, systemic diseases, endocrine
(
6–8)
diseases, and malignant neoplasms.
Currently, it is estimated that over 200 million people
worldwide suffer from this disease, and osteoporosis is the cause of more than 8.9 million
(
9,10)
fractures annually.
men, who account for an estimated one in five Americans with osteoporosis or low BMD. In
015, direct medical costs totaled $637.5 million for fatal fall injuries and $31.3 billion for
nonfatal fall injuries. During that same year, hospitalizations for falls cost an average of $30,550
Although osteoporosis typically occurs in women, it is also diagnosed in
(
11)
2
(
12)
per admission, totaling $17.8 billion. By 2025, the cost of fractures in the United States is
3

expected to exceed $25 billion per year for the treatment of more than three million fractures. In
addition to being the major cause of fractures in elderly individuals, osteoporosis is highly
associated with the occurrence of becoming bedridden, which can lead to serious
(
13)
complications.
Understanding how osteoporosis develops and how to treat it has important
socioeconomic and medical consequences. Efforts have primarily concentrated on the
development of drugs that block bone resorption by decreasing the differentiation and activity of
OCs or on the development of drugs that promote bone formation by increasing the
differentiation and activity of OBs. The first category encompasses the antiresorptive agents,
which are the most commonly prescribed osteoporosis medications, and include
nitrogen-containing bisphosphonates, denosumab, raloxifene, hormone replacement therapy, and
(
14)
calcitonin. By inhibiting OC bone resorption, these therapies are able to improve the quality of
life of patients with osteoporosis. However, they cannot restore the skeletal integrity in most
(
15–25)
patients with established osteoporosis, and they have various side effects.
Recently,
attention has turned to the second category mentioned above, the bone anabolic agents. Such
agents have the ability to increase bone mass and strength, potentially reversing the structural
damage. To date, only one bone anabolic drug is available on the market. The discovery of more
novel, cost-effective bone anabolic agents is therefore a priority to treat those suffering from this
debilitating condition. Such agents will provide an additional option for osteoporosis patients and
4

represent a major advancement in the treatment of osteoporosis.
Anabolic therapies that stimulate bone formation are ideal therapies for patients with
osteoporosis. Despite the promise of various proteins (e.g., bone morphogenetic protein-2
(BMP-2)) as anabolic agents to increase bone formation, many are not optimal for therapeutic
applications due to their side effects outside the skeletal system and their long protein sequences,
which result in high manufacturing costs, limited supplies, and susceptibility to enzymatic
(
26-27)
degradation.
Thus, there is a need for inexpensive, novel agents that promote bone
formation to effectively treat bone loss. The selective augmentation of BMP-2 in OBs by small
molecules that are orally effective may be considered a advancement in this field.
Our research team has established a screening model for the BMP-2 protein. In this
screening, we found that 1-(benzo[b]thiophen-2-yl)ethanone (1) could upregulate the expression
(
28)
of BMP-2. Therefore, derivatives of 1 were synthesized, and their ability to upregulate BMP-2
was evaluated in vitro. Based on these results, SAR studies was performed to guide the next-step
synthesis studies. Three cycles of structural optimization were performed based on the
(
29–32)
precursor.
The SAR data revealed that the C-3 methyl group in the benzothiophene or
(
29)
benzofuran moiety did not enhance BMP-2 expression, the methyl carboxylate-substituted
(
30–32)
C-5 phenyl ring may be beneficial to enhance BMP-2 expression,
the C-6 position is more
favorable for substitution than the other positions, and the 2-carboxylate group is very important
(
32)
for the potent effects of the 1-(benzo-[b]thiophen/furan-2-yl) thanone analogs. (Figure 1) A
5

few derivatives of precursor 1 with higher activity were identified. Derivative 8a showed the
(
30)
highest activity in the senescence-accelerated mouse prone 6 (SAMP-6) model, while 38
showed the highest activity in a glucocorticoid-induced osteoporosis model, with similar
therapeutic effects to those of raloxifene. The safety of 38 was also systematically evaluated in
rat models, and the results indicated that 38 was safe, exhibiting low toxicity and making it
(
32)
worthy of further study.
Preliminary pharmacokinetic experiments suggested that there was high lipophilicity and
poor metabolic stability of both 38 and 8a. The prototype drugs were undetectable within 2
minutes in liver microsomes and within 5 minutes in plasma. To improve their therapeutic
effects in vivo, the ester chain of 38 and 8a was either extended or substituted with a long-chain
amide, and the side-chain proton donor was salified to improve the metabolic stability and
hydrophilicity. 20 derivatives which were not reported in the literature were designed and
synthesized in this work. An ovariectomized (OVX) rat model of osteoporosis was selected to
evaluate the therapeutic effects of the synthetic compounds toward estrogen deficiency-induced
osteoporosis. A study was also conducted using a zebrafish osteoporosis model, and the early
ADME parameters of the compounds were evaluated.
2
RESULTS AND DISCUSSION
2
.1 Chemistry
6

Although drugs with strong lipophilicity possess better membrane permeability, their
high lipophilicity also results in high metabolic clearance rates and strong first-pass effects,
leading to poor stability. In addition, drugs possessing low solubility have limited absorption into
the blood and circulatory system due to their low solubility in the aqueous phase, and most
(
33)
would be excreted in the feces. Therefore, we had to improve the absorption and stability of
the oral medications by balancing their lipophilicity and hydrophilicity to improve their
pharmaceutical efficacy.
To improve the metabolic stability and hydrophilicity of compounds 38 and 8a, we
designed 12 substituted benzofuran analogs and 8 substituted benzothiophene analogs. The
synthesis of these 20 analogs is shown in Scheme 1. The desired compounds (114–125, 132–139)
with different C-2 substituents as well as diverse substituents on the phenyl ring were
synthesized using 38, 8a, and 38a as the starting materials. Intermediates 112, 128, and 129 were
obtained through the reaction of 38, 8a, and 38a with dioxane, respectively, in 0.1 N sodium
hydroxide at 40 °C. The reactions of 112, 128, and 129 with thionyl chloride in the presence of
dimethyl formamide generated the corresponding targets 113, 130, and 131, respectively.
Compounds 113, 130 and 131 were then reacted with a substituted alcohol or amine in the
presence of triethylamine in anhydrous toluene to afford the desired products (114–118, 123, 124,
1
32–136, and 139). The salinization of 115, 123, 134, and 135 by methanesulfonic acid in ethyl
alcohol produced the desired compounds 119, 125, 137, and 138, respectively. Title compounds
7

1
20 and 121 were synthesized by first replacing the phosphate ethyl ester group with a
trimethylsilyl group, followed by alkaline hydrolysis to remove the siloxy group of 117 and 118,
respectively, in the presence of bromotrimethylsilane. The desired product 122 was obtained
through the reaction of 120 with sodium hydroxide.
2
.2 In vitro evaluation of hydrophilicity
The solubilities of the derivatives of precursors 38 and 8a were evaluated by precipitation
after dilution in PBS solution (PBS containing 0.05% Tween-20, pH=7.4). A total of 16
compounds were dissolved in DMSO with 2-fold serial dilutions, and these DMSO serial
solutions were further diluted 10-fold in the PBS solution. The samples precipitated to varying
degrees after PBS dilution (marked as “+” for precipitation). The results showed that the
hydrophilicity of the derivatives was greatly improved compared with that of the precursors 38
and 8a. As shown in Table 1, derivative 120 did not precipitate even at the highest tested
concentration and showed the best hydrophilicity; derivatives 114, 115, 116, 121, 123, 124, 133,
and 134 also showed better hydrophilicity than 38 and 8a.
8

Table 1. Precipitation of the derivatives in the PBS dilution*.
Concentration (nM)
Compound
10000 5000 2500 1250 625 312.5 156.25
1
14
15
16
++
++
++
-
-
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
1
1
-
-
1
1
1
1
1
1
1
1
1
1
20
21
23
24
32
33
34
35
36
39
/
-
/
/
/
/
/
/
/
/
/
/
++
++
++
++
++
++
++
++
++
++
++
++
+
-
/
/
/
/
++
++
++
++
++
++
++
++
++
++
-
+
+
/
+
-
/
-
/
/
/
/
/
/
/
+
+
-
/
-
/
/
/
/
+
++
+
++
+
++
++
++
++
++
++
+
-
++
-
-
/
+
/
/
/
-
/
+
8
a
8
3
3
8a
++
+
*"/" No solids are precipitated and the solution is clear; "+"Solids are precipitated sporadically and some of
them are visible in the solution; "++" Large number of solids are precipitated and the solution is turbid.
We then selected compounds for in vivo experiments. Compound 120, with the best
hydrophilicity in the PBS assay, was selected as the first choice. Then, to compare the SAR and
to investigate the influences of different terminal substituents on the activity, 115, with the same
carbon backbone as 120, was chosen from the group of 114, 115, and 116. To compare the
influences of ester and amide derivatives on the in vivo activity, compound 123, the amide
derivative with the highest hydrophilicity, was selected. In addition, to compare the influences of
sulfur and oxygen atoms on the in vivo activity, 135, the sulfur-containing derivative of precursor
9

3
8, was selected for the in vivo experiments. For the derivatives of 8a, compound 134, with a
cyclic aliphatic hydrocarbon side chain, was selected among the compounds with similar
hydrophilicity for the in vivo experiments. The polarity and water solubility of the 5 chosen
derivatives could be improved after salification. Therefore, 115, 123, 134, and 135 were salified
to their corresponding methanesulfonic acid derivatives 119, 125, 137, and 138, while 120 was
salified to its disodium salt derivative 122 (Figure 2).
2
.3 Activity screening assays in OVX-induced osteoporosis rats and preliminary SAR studies
(
29–32)
In previous studies,
in senile osteoporosis mouse models (SAMP-6) and drug-induced osteoporosis models
glucocorticoid-induced osteoporosis in rats). Compounds 38 and 8a, with excellent activity,
the bioactivity of precursor 1 and its derivatives were evaluated
(
were identified as the precursors for the current study. In the current study, the OVX rat model of
osteoporosis was selected to evaluate the therapeutic effects of the derivatives on estrogen
deficiency-induced osteoporosis and summarize the overall treatment effect of such structures on
the three types of osteoporosis.
First, the preliminary acute toxicity in Kunming mice was assessed. Compounds 119, 122,
1
25, 137, and 138 were all given to the mice orally at a dosages of 500 mg/kg/day for 3 days,
and the animals were closely monitored during the first 3 days and the following 14 days. No
animals died during the 17 days of observation, indicating that the LD values of all compounds
5
0
10

were greater than 500 mg/kg po.
2
.3.1 Influences of the elected compounds on the morphology of tibia tissue in OVX-induced
osteoporosis rats
Figure 3A shows stained slices of trabecular bone tissue from all groups of OVX-induced
osteoporosis rats. The number of trabeculae increased in the groups treated with 1, 38, 119, 125,
and raloxifene compared with that in the untreated model group, with increased thickness and
narrowed gaps between the trabeculae in the treatment groups, suggesting positive therapeutic
effects in OVX-induced osteoporosis rats.
2
.3.2 Effects of the tested compounds on trabecular bone volume (TBV%) of tibia and BMD of
L4-6 in OVX-induced osteoporosis rats
As shown in Figure 3B–D, the TBV% and BMD of the untreated model rats significantly
decreased compared with that of the blank control (MC) group and the sham group, while no
significant difference was observed between the MC group and the sham group. Therefore, the
influence of the surgery alone on the results was excluded, and the establishment of the animal
model was successful. Compared with the mice in the untreated model group, the TBV% and
BMD of the seven test groups and the raloxifene group significantly increased, indicating
positive therapeutic effects. The TBV% and BMD of the mice treated with 125 and 119 were
11

higher than those of in precursor 38 group. The optimal derivative 125 was less effective than the
positive control raloxifene.
In summary, derivatives 119 and 125 had similar or stronger activity than precursor 38,
suggesting that improved hydrophilicity could improve the therapeutic effects. Due to limited
resources, only 7 compound groups were evaluated, and the preliminary SAR comparisons
among these compounds suggested that the derivatives of 38 presented higher bioactivity than
the derivatives of 8a and 38a. Therefore, precursor 38 was a better structural parent compound
for further modification. The activity of compound 125 was higher than that of 119, indicating
that the activity of the amide side chains may be better than the activity of the ester side chains.
Therefore, in future studies, more derivatives with amide side chains will be considered when
designing experiments.
2
.4 Evaluation of the activity and mechanism of action of anti-osteoporotic compound 125 in a
zebrafish model
During preliminary studies, we found that anti-osteoporosis activity experiments in
mammals required an extended experimental period, complex technologies, and high research
costs, which became a technical bottleneck in the research and development of new
anti-osteoporosis drugs. Zebrafish belong to the group of “lower vertebrates” and feature a small
size and intact bones. The key genes involved in the regulation of bone formation (e.g.,
12

runt-related transcription factor 2 (Runx2) and osteonectin) are highly homologous with their
(
34)
mammalian counterparts. To overcome the problems listed above, our institute established a
zebrafish-based osteoporosis model, which dramatically shortened the experimental period and
enhanced research efficiency compared with that of the mammalian models. Moreover, the new
model overcomes the shortcomings of experiments that use osteoblast and osteoclast cell models
outside the body, such as the strict experimental conditions, the single section where the
experiments are applicable and that the experiment results could hardly reflect the internal
comprehensive effects. The dexamethasone zebrafish model was used to evaluate lead compound
(
35)
1
, and we found that the result was highly consistent with that of the rat model. Therefore, we
adopted the zebrafish model to assess the anti-osteoporotic activity of compound 125.
Concurrently, we used this model to study the anti-osteoporosis mechanism of action.
2
.4.1 Activity evaluation using the dexamethasone zebrafish osteoporosis model
The experimental results from the dexamethasone zebrafish osteoporosis model are
shown in Figure 4A. The stained area in the dexamethasone-treated zebrafish was apparently
smaller than that in the untreated fish, indicating that dexamethasone can significantly lower the
degree of bone mineralization in the zebrafish and that the model was established successfully.
The results show that compared with the model group, the stained area of the bone in the groups
treated with 1, 5, and 25 μM 125 was significantly increased. Image-Pro Plus 6.0 software was
13

used to analyze the branchial bones on the left flank of the embryos and calculate the total
stained area of the branchial cleithrum, branchial compass, branchial abopercular, and
branchiostegal rays as well as the total value of the integrated optical density (IOD). With the
control group of the untreated fish (MC) considered to have a value of 1.00, the results of all
groups, presented as fractions, are shown in Figure 4B and C. The findings reveal that treatment
with compound 125 at all concentrations could improve the stained bone area of the zebrafish
and the optical density of the stained bone compared with that of the model group, suggesting
that 125 could alleviate the bone loss induced by dexamethasone. This improvement was
dose-dependent. Treatment with 25 μM 125 could offset most of the bone mineralization damage
caused by dexamethasone, bringing the degree of bone mineralization close to normal.
2
.4.2 Influence of 125 on BMP-2 expression in the zebrafish osteoporosis model
Protein extracts of the untreated control (MC) zebrafish, dexamethasone-treated (model)
zebrafish, and zebrafish treated with 1 μM 125 for 24 h were prepared. Western blots were used
to show the relative expression levels of BMP-2. The results revealed that the relative expression
levels of the BMP-2 protein in the model group was dramatically lower than that of the MC
group, indicating that dexamethasone treatment can lower the expression level of BMP-2 (Figure
4
D and E). In contrast, the expression level of BMP-2 in the group treated with 125 was
14

significantly higher than that of the model group and approached the level in the MC group. This
indicates that 125 can ameliorate the expression of the BMP-2 protein.
2
.4.3 Effect of 125 on Runx2 transcription in vitro
Runx2 is a key transcription factor in osteoblast differentiation. Compounds that
effectively activate Runx2-driven transcription can induce osteoblast differentiation. After
confirming that 125 could upregulate the BMP-2 protein, we further investigated whether 125
could upregulate Runx2 transcription. MC3TC cells were transiently transfected with the
plasmid p6OSE2-luc and treated with the tested compounds for 48 h. Then, the activity of the
firefly luciferase reporter was measured to investigate the effects of the compounds on Runx2
transcription. Compound 125 was tested in this assay, with lovastatin as the positive control. The
results show that the upregulation induced by 10 μM lovastatin was as high as 255.6% of the
untreated cell level, but 100 μM lovastatin was apparently cytotoxic. Compound 125 had the
highest efficacy at 100 μM, upregulating luciferase activity to 267.2%, which was even higher
than the upregulation obtained with lovastatin at the optimal dosage (Figure 4F). These results
indicate that 125 could effectively activate Runx2 transcription and could have inductive effects
on osteoblast differentiation.
15

Considering the results of the above three experiments, by enhancing the expression of
the BMP-2 protein, 125 may effectively activate Runx2-driven transcription, which could further
induce osteoblast differentiation and improve the bone mass of the zebrafish.
2
.5 Study on the early ADME properties of active compounds
After determining the anti-osteoporosis activity of compounds of this type through in vivo
tests, we initiated ADME research to examine the pharmacological performance of these
compounds as drugs.
2
.5.1 Unidirectional permeability of the test compounds in Caco-2 cells
As a tool for the rapid screening of drug absorption capacity, the Caco-2 test was deemed
key in the early screening of oral medications. The apparent permeability coefficient (Papp)
(
36)
obtained from Caco-2 cells correlates well with human intestinal absorption. After preparation
of the Caco-2 cells, we measured the value of the apical to basolateral transepithelial electrical
resistance (TEER AP-BL) and lucifer yellow (LY) leakage AP-BL to evaluate monolayer
integrity, during which we also verified the accuracy of the model by introducing metoprolol,
erythromycin, and cimetidine as control compounds. As shown in Table 2, the results of the Papp
(AP-BL) measurements of 1, 38, and 125 indicate relatively good intestinal absorption of 1 and
16

1
3
25 but poor absorption of the lead compound 38. The Papp value of 125 was higher than that of
8.
Table 2. Permeability results of the test compounds in the Caco-2 cell line.
Papp (AP-BL)
TEER AP-BL
LY Leakage
AP-BL (%)
0.154
Compound
-6
2
(×10 , cm/s)
(Ω × cm )
Metoprolol
Erythromycin
Cimetidine
1
32.6
0.0324
0.679
326
396
318
351
373
341
0.137
0.134
0.179
0.152
22.1
3
8
< 0.00247
9.34
*
125
0.158
*
Receiver sample AB was below the limit of quantification (BLOQ).
2
.5.2 Metabolic stability of the test compounds in pooled human liver microsomes
We then used human liver microsomes for the metabolic stability study to evaluate the
metabolism and clearance rate of the compounds in vitro. We verified the accuracy of the model
with phenacetin, verapamil, metoprolol, imipramine, benzydamine, and diclofenac. As shown in
Table 3, the evaluation of 1, 38, and 125 revealed short half-lives for the in vitro metabolism of 1
and 38, as well as large intrinsic clearance rates and poor metabolic stability. In contrast, the
half-life for the in vitro metabolism of 125 was significantly longer than that of lead compound
3
8, with a smaller intrinsic clearance rate and better metabolic stability (Table 3).
Comprehensively considering the two tests, the ADME properties of 125 were significantly
17

better than those of the lead compound, which was consistent with the original goal of the
current round of derivative design.
Table 3 Metabolic stability of the test compounds in pooled human liver microsomes.
Compound
Phenacetin
Verapamil
Metoprolol
Imipramine
Benzydamine
Diclofenac
1
In vitro t1/2 (min)
CL (µL/min/mg protein)
int
26.5
3.40
159.0
67.0
43.1
3.52
1.25
26.2
204
4.35
10.3
16.1
197.0
>300
*
*
3
8
NC
NC
1
25
103
6.73
*
In vitro t1/2 and CLint values could not be calculated. The test compound was only detected at the 0.5 minute
time point, which indicates a fast metabolism. CL , intrinsic clearance
int
2
.5.3 CYP inhibition of the test compounds
We proceeded with the ADME research of 125 to identify whether it could inhibit the
hepatic drug-metabolizing enzyme system and thereby affect the efficacy and safety of other
drugs. The most important hepatic microsomal enzymes are the CYP450s. We selected the five
most common CYP450 enzymes and verified the model with specific inhibitors targeting each of
these enzymes. As shown in Table 4, 125 had slight inhibitory effects on CYP2D6 but no
inhibitory effects on the other four enzymes; therefore, 125 should have little impact on the
metabolism of other drugs.
18

Table 4. Summary of CYP inhibition results.
CYP*
1A2
2D6
3A4
2C9
2C19
Dextromethorphan
Midazolam
(1)
Substrate (μM)
Phenacetin (10)
Diclofenac (10)
Omeprazole (0.5)
(
5)
Inhibitor
Compound
Control
Naphthoflavone
IC50 (µM)
0.0117
Quinidine
IC50 (µM)
0.0356
Ketoconazole
IC50 (µM)
0.0217
Sulfaphenazole
IC50 (µM)
0.893
Tranylcypromine
IC50 (µM)
2.07
1
25
>10
7.84
>10
>10
>10
*
Human liver microsome concentration in the incubation was 0.2 mg/mL, incubation time = 20 minutes
2
.5.4 Pharmacokinetic evaluation of the 125 metabolites in Sprague-Dawley rats after iv and po
administration
In the human body, 125 would be hydrolyzed into the acid-form 112 to continue its
activity. A simple assessment of the pharmacokinetic parameters of 125 would not accurately
represent its in vivo pharmacokinetic properties and therapeutic effects. As a result, in the
pharmacokinetic rat model, we used iv and intragastric administration of the metabolite of 125
and measured its pharmacokinetic parameters. As shown in Table 5, the metabolite of 125
possessed good pharmacokinetic parameters and high oral bioavailability.
19

*
Table 5. PK parameters and plasma concentration of the metabolite of 125 .
C
0
(ng/mL)
T1/2 (h)
AUC (ng
17.2 ± 3.46
2.52 ± 0.89
T
1/2 (h)
4.53 ± 0.28
0.50.± 0.00
T
max (h)
iv
mg/kg AUC0-∞ (ng
MRT (h)
·
h/mL) 59.5 ± 24.7
po
Cmax (ng/mL)
54.2 ± 16.2
0
-t
2
·
h/mL) 69.0 ± 21.6 10 mg/kg AUC (ng
·
h/mL) 278.0 ± 28.0
0-t
2.85 ± 0.31
AUC0-∞ (ng
·
h/mL) 286.0 ± 30.0
Bioavailability (%) 83.1%
*
means ± SD, n=3, except for Bioavailability
2
.6 BALB/C mice 14-Day toxicology study
This experiment aimed to evaluate the in vivo toxic responses to 38 and 125. BALB/C
mice were divided into a control group, the 38 (100 mg/kg) group, the 38 (50 mg/kg) group, the
25 (100 mg/kg) group and the 125 (50 mg/kg) group to observe the status and growth of the
1
BALB/C mice and evaluate the influence of the compounds on the 14 days following intragastric
administration. Indicators of observation included hair, behavioral changes and neurological
signs (fremitus, convulsion, spasm, gait disturbance, etc.). After 14 days, all mice were weighed
and then dissected to observe the pathological changes in their organs; routine blood tests and
blood biochemistry analyses were also performed to study the relationship between the dose and
the toxic responses in mice and the extent of damage to the main target organs. The experiment
20

also explored the tolerant dose range in laboratory animals and the safe dose range that was free
of any toxic response.
According to the results of the toxicity study, the mice in the 4 medication administration
groups and the control group gained weight well, with no significant difference among the
groups. In the 4 medication administration groups, the hair of the animals was glossy and no
significant behavioral changes or neurotoxic symptoms were found. Some of the liver biopsies
showed a minor degree of hepatocyte nucleus mitosis, megakaryocyte and inflammatory
infiltration, and hepatocyte necrotic foci, etc., with no significant difference among the different
groups. Some renal biopsies indicated necrosis in individual renal tubular epithelial cells to a
minor degree instead of necrosis on the whole, but no significant difference among the groups
was found. During the treatment course, no tissue damage was observed in the livers or kidneys
of mice treated with 38 and 125 (Figure 5A). Additionally, the serum ALT, aspartate
transaminase (AST), urea, and Cre levels of the mice treated with 38 and 125 were still in the
normal range (Figure 5B-E).
3
CONCLUSIONS
In this study, the ester chains of precursors 38 and 8a were extended or substituted with a
long-chain amide, and the side-chain proton donor was salified to improve metabolic stability
and hydrophilicity. We desired to improve the pharmacokinetic properties of 38 and 8a, which
21

we had previously established as having excellent anti-osteoporosis activities. A total of 20
derivatives were designed and synthesized, none of which had been previously reported in the
literature, and all showed improved hydrophilicity compared with that of the precursor.
Five compounds with the highest hydrophilicity were selected for evaluation in an
OVX-induced osteoporosis rat model. The TBV% and BMD of the groups treated with 119, 125,
3
8, and 1 were higher than those of the untreated model group but lower than those of the
raloxifene group. Compound 125 demonstrated better properties than those of its precursor 38.
This indicated that the goal of improving the in vivo therapeutic efficacy of the precursor
compound by improving the pharmacokinetic properties was achieved. Preliminary SAR
comparisons among these compounds suggested that precursor 38 was a better structural parent
compound than 8a and that amide side chains could produce better activity than lipid side chains.
Our institute established a zebrafish model of hormone-induced osteoporosis. With this
model, the therapeutic effects of test drugs can be evaluated within 2 weeks using a simplified
operation; thus, in vivo activity screening in batches is possible. All of these factors provide a
solid foundation for the more rapid and accurate identification of novel anti-osteoporosis drugs.
This study verified the anti-osteoporosis activity of 125 with the zebrafish osteoporosis model,
and we further quantitatively evaluated BMP-2 protein expression in zebrafish. The results
showed that 125 could effectively enhance BMP-2 protein expression. Runx2 is a key
transcription factor in OB differentiation. We found that 125 could effectively activate
22

Runx2-driven transcription. These results confirmed the mechanism of how these compounds
enhance OB activity and protect against osteoporosis. In addition, this model could be used for
future in vitro verification of compounds that upregulate BMP-2.
After verifying the anti-osteoporosis activity of 125, we evaluated its early ADME
properties. In both the Caco-2 cell absorption and liver microsome stability tests, compound 125
presented better parameters than those of compounds 38 and 1. These results are in agreement
with the original goal of the current round of derivative design, which was to significantly
enhance the absorptive capacity and stability of 125 compared with that of the lead compound 38.
We also learned from the CYP450 enzyme inhibition test that 125 should have little impact on
the metabolism of other drugs, which reflects its safety. The metabolite of 125 showed good
pharmacokinetic parameters and high oral bioavailability. Moreover, the high likelihood of oral
administration of 125 is of great importance to osteoporosis patients in need of chronic drug
administration.
Based on BMP-2 screening, our group first discovered that lead compound 1, which had
a novel structure, possessed anti-osteoporosis activity, and we completed three rounds of
structural optimization of 1 over the years. During this process, our group synthesized more than
1
50 new compounds. Moreover, our group created multiple evaluation models related to
osteoporosis at both the cellular and animal level and thereby noticeably improved the level of
the research platform. Representative compound structures and major experimental results are
23

shown in Table 6.
24

Table 6. Major experimental results on these compounds of our group.
Compound
Experimental content
1
8a
38
125
Positive control
Lovastatin
Upregulated activity of
Upregulated
Upregulated
64.2%
Upregulated
89.4%
a,
BMP-2 expression in vitro
Not Done
Upregulated
3
5.6%
b
2
0.9%
Effects on bone
Lower than the
positive control
Slightly lower than
the positive control
histomorphometry in
Not Done
Not Done
Not Done
Lovastatin
a
SAMP6 mice
Effects on bone
histomorphometry in
glucocorticoid-induced
Higher than the
positive control
Lower than the
positive control
Higher than the
positive control
Alfacalcidol
Activity
b
osteoporosis rats
Effects on bone
histomorphometry in
Lower than the
positive control,
higher than 1
and 38
Offset most of
the bone
Lower than the
positive control
Lower than the
positive control
Not Done
Not Done
Raloxifene
Not Done
OVX-induced osteoporosis
c
rats
Effects on bone
histomorphometry in
Not Done
Not Done
zebrafish-based osteoporosis
mineralization
25

c
model
damage
5
000 mg/kg/day
orally, 3 days
500 mg/kg/day
orally, 3 days
No animals died
500 mg/kg/day
orally, 3 days
No animals died No animals died
500 mg/kg/day
orally, 3 days
Preliminary acute toxicity in
Not Done
Not Done
a, b, c
Kunming mice
No animals died
50 mg/kg/day
1
Pathological sections of the
hearts, livers, spleens, lungs,
orally, 3 months
No pathological
changes
30 mg/kg/day
orally, 3 months
Not Done
Not Done
Not Done
Not Done
Not Done
Not Done
a
and kidneys of rats
Toxicity
Pathological sections of the
hearts, livers, spleens, lungs,
30 mg/kg/day
orally, 3 months
No pathological
changes
Not Done
and kidneys of SAMP6 mice No pathological
a
changes
Blood hematology and
Alfacalcidol
30 mg/kg/day
orally, 3 months
No effects
3
0 mg/kg/day
30 mg/kg/day
orally, 3 months
No effects
30 mg/kg/day
orally, 3 months
No effects
biochemistry of rats with
glucocorticoid-induced
orally, 3 months
No effects
b
osteoporosis
1
00 mg/kg/day
100 mg/kg/day
orally, 14 days
No effects
BALB/C mice 14-Day
Not Done
Not Done
orally, 14 days
Not Done
c
toxicology tests
No effects
Immunohistochemical
analysis of BMP-2
Increased BMP-2
expression
Increased BMP-2
expression
Not Done
Not Done
Not Done
Lovastatin
expression in the femora of
a
SAMP6 mice
Mechanism
The influence on the BMP-2
expression of the
Enhanced the
expression of the
BMP-2 protein
Not Done
Not Done
Not Done
zebrafish-based osteoporosis
c
model
26

Effectively
activated
transcription
Effect on Runx2
transcription in vitro
Not Done
Not Done
Not Done
Lovastatin
c
a
b
c
(30)
Our group published in The Journal of Medicinal Chemistry in 2010.
(32)
Our group published in The European Journal of Medicinal Chemistry in 2015.
Content of this article
27

In terms of activity, through two rounds of SAR studies, the anti-osteoporosis activity of
each round of derivatives was better than that of the previous round. Among these derivatives, 38
showed therapeutic effects in the glucocorticoid rat model that were better than those of the
positive control drug, leading to almost normalized values. During the third round of structural
modification, both the therapeutic effects and pharmacological performance of the compounds
were enhanced by improving their pharmacokinetic properties. Three large-scale animal tests
were adopted, each representing a different type of osteoporosis for a more thorough evaluation
of the therapeutic effects of these compounds. The results indicated that compounds of this type
worked well on senile osteoporosis (SAMP-6 mouse model) and drug-induced osteoporosis
(
glucocorticoid rat model) but did not work well on osteoporosis resulting from menopause
OVX rat model).
(
We assessed the safety of these compounds by evaluating the overall state of the animals
and the visceral and blood indexes; the results indicated low toxicity. A preclinical study on the
toxicity of 125 will be started shortly.
Regarding the mechanism of action of the compounds, using cellular, zebrafish, and
mouse models, we verified that these compounds exhibited their anti-osteoporosis effects via
osteogenesis and upregulation of BMP-2. As a result, these compounds could correct bone loss
that is already occurring in patients, which is difficult for the existing first-tier drugs currently in
clinical practice. In addition, these new compounds could be used in combination with existing
28