Tetrahydroquinoline and tetrahydroisoquinoline derivatives as potential selective PDE4B inhibitors

Article abstract of DOI:10.1016/j.bmcl.2018.04.068

Tetrahydroquinoline and tetrahydroisoquinoline derivatives containing 2-phenyl-5-furan moiety were designed and synthesized as phosphodiesterase type 4 (PDE4) inhibitors. The bioassay results showed that title compounds showed good inhibitory activity aga

Full text of DOI:10.1016/j.bmcl.2018.04.068

Accepted Manuscript
Tetrahydroquinoline and tetrahydroisoquinoline derivatives as potential selec-
tive PDE4B inhibitors
Ya-Sheng Li, Xing-Yu Liu, Dong-Sheng Zhao, Yi-Xian Liao, Lian-Hui Zhang,
Feng-Zhi Zhang, Gao-Peng Song, Zi-Ning Cui
PII:
S0960-894X(18)30386-X
https://doi.org/10.1016/j.bmcl.2018.04.068
BMCL 25817
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
12 March 2018
24 April 2018
29 April 2018
Please cite this article as: Li, Y-S., Liu, X-Y., Zhao, D-S., Liao, Y-X., Zhang, L-H., Zhang, F-Z., Song, G-P., Cui,
Z-N., Tetrahydroquinoline and tetrahydroisoquinoline derivatives as potential selective PDE4B inhibitors,
Bioorganic & Medicinal Chemistry Letters (2018), doi: https://doi.org/10.1016/j.bmcl.2018.04.068
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.

Tetrahydroquinoline and tetrahydroisoquinoline derivatives as potential
selective PDE4B inhibitors
Ya-Sheng Li a, c, #, Xing-Yu Liu a, #, Dong-Sheng Zhao b, #, Yi-Xian Liao a, Lian-Hui Zhang a,
Feng-Zhi Zhang c, Gao-Peng Song d,*, Zi-Ning Cui a,*
a
State Key Laboratory for Conservation and Utilization of Subtropical Agro-bioresources,
Integrative Microbiology Research Centre, Guangdong Province Key Laboratory of Microbial
Signals and Disease Control, South China Agricultural University, Guangzhou 510642, China
b Department of Pharmacy, Quanzhou Medical College, Quanzhou 362100, China
c College of Pharmaceutical Science & Green Pharmaceutical Collaborative Innovation Center of
Yangtze River Delta Region, Zhejiang University of Technology, Hangzhou, 310014, China
d College of Materials and Energy, South China Agricultural University, Guangzhou 510642, China
#The authors contributed equally to this paper.
*Corresponding author. Tel.: +86-20-85288229; Fax: +86-20-85288229
E-mail address: ziningcui@scau.edu.cn (Z. C.), vinsin1021@126.com (G. S.)
1

Abstract:
Tetrahydroquinoline
and
tetrahydroisoquinoline
derivatives
containing
2-phenyl-5-furan moiety were designed and synthesized as phosphodiesterase type 4 (PDE4)
inhibitors. The bioassay results showed that title compounds showed good inhibitory activity
against PDE4B and blockade of LPS (lipopolysaccharide) induced TNF-α release, which also
exhibited considerable in vivo activity in animal models of asthma/COPD (chronic obstructive
pulmonary disease) and sepsis induced by LPS. The bioactivity of compounds containing
tetrahydroquinoline (series 4) was higher than that of tetrahydroisoquinoline derivatives (series 3).
Compound 4m with 4-methoxybenzene moiety exhibited the best potential selective activity
against PDE4B. The primary structure–activity relationship study and docking results showed that
the tetrahydroquinoline moiety of compound 4m played a key role to form hydrogen bonds and
π-π stacking interaction with PDE4B protein while the rest part of the molecule extended into the
catalytic domain to block the access of cAMP and formed the foundation for inhibition of PDE4B.
Based on LPS induced sepsis model for the measurement of TNF-α inhibition in Swiss Albino
mice and neutrophilia inhibition for asthma and COPD in Sprague Dawley rats with the potential
molecules, compound 4m would be great promise as a hit inhibitor in the future study.
Keywords: synthesis; tetrahydroquinoline; tetrahydroisoquinoline; PDE4B inhibitor; molecular
simulation
2

Asthma and chronic obstructive pulmonary diseases (COPD) known as the kind of chronic
obstructive lung disease have been a major public health burden worldwide.1 Including Asthma and
COPD, kinds of diseases and biological responds are mediated by intracellular concentration of the
secondary signal messengers, such as cyclic adenosine monophosphate (cAMP) and cyclic
guanosine monophosphate (cGMP), which are able to regulate the function of airway smooth muscle,
inflammatory cells, and immune cells.2,3 As one of the 11-membered Phosphodiesterases (PDEs),
PDE4 belongs to the superfamily of 3’,5’-cyclic-nucleotide phosphodiesterase and specifically
catalyze the hydrolysis of cAMP to its inactive 5’-monophosphate metabolite, thus blocking the
associated signal transduction systems. PDE4 is predominantly expressed in inflammatory and
immune cells like eosinophils, lymphocytes, macrophages, and neutrophils.4,5 Hence PDE4
inhibitors can increase the concentration of cAMP in cell, which elicits a variety of functional
responses in the inflammatory cells such as reduction of tumour necrosis factor-α (TNF-α)
production.6-8 Therefore the development of PDE4 inhibitors as anti-inflammatory drugs for new
therapies in clinic has made along standing research effort.
Since the first-generation PDE4 inhibitors rolipram and piclamilast (Fig. 1) have been extensively
studied as anti-inflammatory drugs, many research were reported in this field.9 Structurally, rolipram
and piclamilast belong to diether derivatives of catechol class, some of which have been evaluated in
clinic for the treatment of patients with depression, asthma, COPD, dermatitis,
neuroinflammation10,11 and rheumatoid arthritis.12 Further structural optimization demonstrated that
the 8-methoxyquinoline-5-carboxamides moiety (such as SCH365351, Fig. 1)13 had excellent PDE4
inhibitory activity. Molecular simulation studies suggested that the quinoline moiety of SCH365351
could bind to the adenosine recognition site. Base on structural optimization, we changed the amide
linker of the SCH365351 into furan ring and oxazole ring to design and synthesize series of
3

derivatives containing 5-phenyl-2-furan14 and 2-phenyl-4-oxazole15 residue in our previous work
(Fig. 1), which exhibited considerable inhibitory activity against PDE4B, the important subtype of
PDE4. Primary SAR results showed that a five-membered heterocycles including oxazole and furan
rings could be a key pharmacophore of PDE4 inhibitors. In addition, we reported that some
7-(cyclopentyloxy)-6-methoxy 1,2,3,4-tetrahydroisoquinoline derivatives as “like-rolipram” drugs
(Fig. 1) showed strong inhibitory activity against PDE4B.11
Figure 1. The designed strategy for the title compounds
Based on the above results, five membered heterocyclics and 1,2,3,4-tetrahydroisoquinoline
moiety received the special pharmaceutical effect to inhibit PDE4. However, as isomeride of
1,2,3,4-tetrahydroisoquinoline, 1,2,3,4-tetrahydroquinoline has been drawn less attention to being as
the PDE4 inhibitors. Because of the different spatial configurations, 1,2,3,4-tetrahydroquinoline was
expected to have a good binding effects with the domain of PDE4. In the present work, two series of
4

1,2,3,4-tetrahydroisoquinoline
and
1,2,3,4-tetrahydroquinoline
derivatives
containing
5-phenyl-2-furan moiety were designed rationally and synthesized, of which the inhibitory activity
against PDE4B and blockade of LPS-induced TNF-α release were evaluated.
The synthetic route of title compounds 3 and 4 was shown in Scheme 1. The key intermediates 2
(the different carboxylic acids 2 ) were synthesized from substituted aniline by Meerwein arylation
reaction according to the reported procedure.16,17 With the different carboxylic acids 2 and
tetrahydroisoquinoline, their coupling reaction was then performed in the presence of EDCI and
HOBt to provide the different title compounds 3 (3a-3o), respectively. And a mixture of
5-substituted phenyl-2-furancarboxylic acid 2 and thionyl chloride was refluxed in anhydrous
toluene for 3.5 h to afford the different 5-phenyl-2-furancarbonyl chloride, which was added into
tetrahydroquinoline in anhydrous dichloromethane to react and obtain the title compounds 4 (4a-4o),
respectively.
Reagents and conditions: (a) NaNO2, hydrochloric acid, 0 - 5 ºC, 3h; (b) furoic acid, CuCl2 (cat.), acetone - H2O,
r.t., 5 h (two steps); (c) HOBt, EDCI, tetrahydroisoquinoline, anhydrous dichloromethane, 0 ºC – r.t., 3 h; (d)
SOCl2, anhydrous toluene, reflux; (e) tetrahydroquinoline , anhydrous dichloromethane, reflux, 4 h. R = a: H; b:
2-Cl; c: 3-Cl; d: 4-Cl; e: 2-F; f: 3-F; g: 4-F; h: 2-NO2; i: 3-NO2; j: 4-NO2; k: 4-CH3; l: 4-Br; m: 4-OCH3; n:
2,4-di-F; o: 2,6-di-F.
5

Scheme 1. The synthetic route of the title compounds 3a-3o and 4a-4o.
As shown in Table 1, rolipram as the positive control (see Supplementary data for the method), in
vitro data for the inhibition of PDE4 and blockade of LPS-induced TNF-α release were listed. The
data revealed that the inhibitory activity of title compounds containing tetrahydroquinoline was
better than that of the compounds containing tetrahydroisoquinoline. For instance, the IC50 values of
compounds 4b and 4j (containing tetrahydroquinoline) against PDE4B were 41.6 µM and 12.1 µM,
which were better than that of compounds 3b and 3j (containing tetrahydroisoquinoline) against
PDE4B were 51.6 µM and 15.8 µM.
Table 1. Inhibition (IC50, µM) of enzymatic potency (PDE4B and PDE4D) and TNF-α release from
human blood mononuclear cells stimulated with lipopolysaccharide a
Compd.
R
PDE4B
IC50 (µM)
79.6 4.4
51.6 1.9
21.1 1.0
4.1 0.5
46.1 1.5
25.7 1.3
7.4 0.6
66.3 3.4
39.9 3.0
15.8 1.4
9.4 0.8
58.4 1.9
2.6 0.5
PDE4D
Selectivity (D/B)
TNF-α
IC50 (µM)
188.5 7.3
121.2 4.0
57.1 1.6
28.6 1.1
95.4 3.2
78.1 2.4
27.2 1.4
151.2 4.9
72.5 3.7
46.3 2.3
44.1 1.8
156.2 6.6
13.6 0.7
17.1 0.7
91.9 2.9
141.5 5.7
101.4 4.2
62.5 2.1
27.2 1.2
92.8 3.7
64.3 3.1
22.1 1.0
128.1 5.3
62.3 3.0
H
2-Cl
3-Cl
4-Cl
2-F
NTb
NTb
--
--
--
6.14
--
--
6.66
--
--
5.39
6.04
--
3.92
6.03
4.21
---
3a
3b
3c
3d
3e
3f
NTb
25.2 1.2
NTb
3-F
4-F
NTb
49.3 1.8
NTb
3g
3h
3i
2-NO2
3-NO2
4-NO2
4-CH3
4-Br
4-OCH3
2,4-di-F
2,6-di-F
H
2-Cl
3-Cl
4-Cl
2-F
NTb
88.3 3.4
56.8 2.2
NTb
3j
3k
3l
10.2 0.4
18.1 1.0
89.2 1.7
NTb
3m
3n
3o
4a
4b
4c
4d
4e
4f
3.0 0.3
17.1 1.4
70.9 3.3
41.6 1.9
18.9 1.1
3.5 0.4
50.3 1.5
22.2 1.1
6.1 0.4
NTb
--
96.8 3.7
21.3 1.1
NTb
5.12
6.09
--
--
6.60
--
3-F
4-F
2-NO2
3-NO2
NTb
40.3 1.5
NTb
4g
4h
4i
66.9 2.9
30.9 2.1
NTb
--
6

4-NO2
4-CH3
4-Br
4-OCH3
2,4-di-F
2,6-di-F
12.1 1.2
9.4 0.3
46.8 1.6
1.5 0.2
2.6 0.4
15.2 0.9
2.0 0.3
77.2 2.8
50.1 2.5
NTb
10.3 0.5
15.6 0.9
81.3 2.2
2.8 0.7
6.38
5.33
--
6.86
6.00
5.34
1.40
46.8 2.3
35.2 1.4
131.2 5.2
11.8 0.6
16.2 0.7
89.3 2.8
18.9 0.9
4j
4k
4l
4m
4n
4o
rolipram
a Results are the average of at least three assays. b NT, not tested.
From Table 1, it can be seen that the substituted group at the benzene had decisive influence on
the inhibitory activity and selectivity against PDE4B since the compounds (3a and 4a, R = H)
without any substituted group displayed the poorest inhibitory activities among all the title
compounds. While compound 4m (R = 4-OCH3) exhibited the best inhibitory against PDE4B and
TNF-α with the IC50 value of 1.5 µM and 11.8 µM, which shown more competitive than rolipram
(2.0 µM and 18.9 µM). Accordingly, among compounds 3a-3o, compound 3m (R = 4-OCH3)
showed strong inhibitory activity (IC50 value of 2.6 µM and 13.6 µM, respectively) against PDE4B
and TNF-α. Herein, this result was well consistent with the conclusion of our recent studies14,15
,
methoxyl group at para-position of benzene ring was favorable to the inhibitory activity against
PDE4B. The locations of substituted groups at the benzene ring had important effects on the
bioactivity. Activity with respect to substitution at the benzene ring follows the obvious trend: 4- >
3- > 2-. It was noted that introduction of 2,4-di-F group in the benzene ring (3n and 4n) was also
helpful to enhance inhibitory activity (IC50 value of 3.0 µM and 2.6 µM) against PDE4B. The
bioactivity against PDE4D was tested when the potential inhibitors display IC50 < 20 µM against
PDE4B. It was interesting to note that all the title compounds with strong inhibition against PDE4B
showed better selectivity towards PDE4D comparable to rolipram. Most of the title compounds
exhibited the selectivity (D/B) ratio around 5 to 7 folds in Table 1. Notably, compound 4m displayed
about 6.86 times selectivity towards the PDE4D subtype better than that of rolipram, demonstrating
that compound 4m is exploitable as a potential lead compound for the design of selective PDE4B
7

inhibitors.
Gratifyingly, four active compounds from selected in vivo showed significant TNF-α inhibition
(LPS induced sepsis model for the measurement in Swiss Albino mice) and neutrophilia inhibition
for asthma and COPD (in Sprague Dawley rats). The details were listed in Table 2. The results
showed that compound 4m demonstrated better inhibitory activity against TNF-α release (46.8%)
and LPS induced neutrophilia inhibition (40.1%) than that of rolipram (42.9% and 34.2%), and
compounds 3m and 4n also revealed similar bioactivity (TNF-α: 39.2% and 39.6%; LPS: 32.7% and
34.1%) compared to rolipram.
Table 2. LPS induced TNF-α in SA mice and neutrophil influx in BALF of SD rats
Swiss Albino mice (n = 6)
Sprague Dawley rats (n = 6)
Does
(mg/kg, po)
TNF-α
Inhibition (%)
Does
(mg/kg, po)
LPS induced
neutrophilia (%
inhibition)
Compd.
R
2.4-di-F
4-OCH3
2.4-di-F
4-OCH3
10
10
10
10
10
30.8
39.2
39.6
46.8
42.9
10
10
10
10
10
29.5
32.7
34.1
40.1
34.2
3n
3m
4n
4m
rolipram
According to the results of the bioassay, 3m and 4m showed the best activity for further discovery.
Docking studies of compounds 3m and 4m at PDE4B (PDB ID: 1XMY) was conducted using
Surflex-Dock in Sybyl 8.0 (see the Supplementary data for the method),16,25 and the docking contour
maps were shown in Figure 2. The docking orientation demonstrated that the furan moiety as the
pivotal pharmacophore formed integral hydrogen bonds with the conserved glutamine residue
(Gln443) (Fig. 2) and the heterocyclic ring was evidently positioned between the phenylalanine
(Phe446) and isoleucine (Ile410) (Fig. 2A and C), which formed the cavity accommodating the
8

hydrophobic moiety of compounds 3m and 4m. Obviously, oxygen atom of amide(-CO-N-) in
compounds 3m and 4m forming steady hydrogen bond with hydrogen atom of amide(-CO-NHH) in
Gln443, and the tetrahydroisoquinoline (3m) and the tetrahydroquinoline (4m) formed stronger π-π
stacking interaction with benzene ring of the phenylalanine (Phe446). The remainder of the molecule
was displayed to extend into the catalytic domainin close to both the Zn2+ and Mg2+ cations (Fig. 2F
and H), which played important roles in the catalytic mechanism of cAMP hydrolysis. The
para-methoxy group formed coordinate bond with the Zn2+ (3m: 3.92 Å and 4m: 2.02 Å, Fig. 2E
and 2G) and Mg2+ (3m: 3.60 Å and 4m: 3.34 Å, Fig. 2E and 2G) cations.
A
B
D
F
C
E
9

G
H
Figure 2. Model of PDE4 and docking of compounds 3m and 4m. (A, B) The entire PDE4B
bound to 3m. (C, D) The entire PDE4B bound to 4m. (E, F) The catalytic domain bound to 3m. (G,
H) The catalytic domain bound to 4m.
The catalytic domain bound to 3m and 4m overlaid with rolipram were shown in Figure 3. It
illustrated in Figure 3 that the oxygen atom of carbonyl group and the nitrogen atom of
tetrahydroquinoline or tetrahydroisoquinoline interacted with Gln443, which played the same role as
the oxygen atoms of methoxy group and cyclopentyloxy group in rolipram respectively. Such
orientation and interactions would block the access of cAMP to the catalytic domain and formed the
foundation for inhibition of PDE4. That could be the reason why the activity of most title
compounds containing tetrahydroquinoline (series 4) was better than that of compounds containing
tetrahydroisoquinoline (series 3).
A
B
10

C
D
Figure 3. Model of PDE4B and docking of compounds 3m and 4m overlaid with rolipram (green).
(A, B) The catalytic domain bound to 3m. (C, D) The catalytic domain bound to 4m.
In summary, 1,2,3,4-tetrahydroisoquinoline and 1,2,3,4-tetrahydroquinoline derivatives containing
2-phenyl-5-furan moiety as new phosphodiesterase type 4 inhibitors were designed, synthesized and
evaluated. The bioactivity exhibited that 1,2,3,4-tetrahydroquinoline moiety was an excellent
pharmacophore of PDE4 inhibitors, and 1,2,3,4-tetrahydroquinoline derivatives (series 4) exhibited
more competitive biological activity than that of the 1,2,3,4-tetrahydroisoquinoline derivatives
(series 3), both in vivo and in vitro. Among the title compounds, compound 4m showed the strongest
inhibitory activity against PDE4B and the highest selectivity toward PDE4D. The primary SAR
showed that substituents at para-position of benzene ring favored to the bioactivity. Docking results
suggested that the formation of the hydrogen bonds, π-π stacking interaction, hydrophobic
interaction and metal coordination in the ligand-receptor complex were vital for the binding affinity.
We hope that such efforts were contributed to the further development of novel PDE4 inhibitors with
better activity and selectively.
Acknowledgements
We acknowledge the financial supports from the National Key Research and Development
Program of China (2017YFD0200504), the National Key Project for Basic Research (973 Program,
2015CB150600), the Special Fund for the Research and Construction of Public Service Ability
11

(2016A020217014).
References and notes
1. Raghavan, N. ; Webb, K. ; Amornputtisathaporn, N. ; O’ Donnell, D. E. Curr. Opin. Pharmacol.
2011, 11, 204.
2. Jeon, Y. H.; Heo, Y. S.; Kim, C. M.; Hyun, Y. L.; Lee, T. G.; Ro, S.; Cho, J. M. Cell. Mol. Life
Sci. 2005, 62, 1198.
3. Manallack, D. T.; Hughes, R. A.; Thompson, P. E. J. Med. Chem. 2005, 48, 3449.
4. Beavo, J. A.; Reifsnyder, D. H. Trends Pharmacol. Sci. 1990, 11, 150.
5. Tasken, K.; Aandahl, E. M. Physiol. Rev. 2004, 84, 137.
6. Kodimuthali, A.; Jabaris, S. L.; Pal, M. J.Med.Chem. 2008, 51, 5471.
7. Huang, Z.; Mancici, J. A. Curr. Med. Chem. 2006, 13, 3253.
8. Houslay, M.; Schafer, P.; Zhang, K. Y. J. Drug Discovery Today. 2005, 10, 1503.
9. Beavo, J. A.; Reifsnyder, D. H. Trends Pharmacol. Sci. 1990, 11, 150.
10. Norman, P. Expert Opin. Ther. Pat. 1998, 8, 771.
11. Song, G. P.; Zhao, D. S.; Hu, D. K.; Li, Y. S.; Jin, H. W.; Cui, Z. N. Bioorg. Med. Chem. Lett.
2015, 25, 4610.
12. Kuang, R.; Shue, H. J.; Blythin, D. J.; Shih, N. Y.; Gu, D.; Chen, X.; Schwerdt, J.; Lin, L.; Ting,
P. C.; Zhu, X.; Aslanian, R.; Xiao, L.; Prelusky, D.; Wu, P.; Zhang, J.; Zhang, X.; Celly, C. S.;
Minnicozzi, M.; Billah, M.; Wang, P. Bioorg. Med. Chem. Lett. 2007, 17, 5150.
13. Kuang, R.; Shue, H. J.; Xiao, L.; Blythin, D. J.; Shih, N. Y.; Chen, X.; Gu, D.; Schwerdt, J.;
Ling, L.; Ting, P. C.; Cao, J.; Aslanian, R.; Piwinski, J. J.; Prelusky, D.; Wu, P.; Zhang, J.; Zhang,
X.; Celly, C. S.; Billah, M.; Wang, P. Bioorg. Med. Chem. Lett. 2012, 22, 2594.
14. Li, Y. S.; Tian, H.; Zhao, D. S.; Hu, D. K.; Liu, X. Y.; Jin, H. W.; Song, G. P.; Cui, Z. N. Bioorg.
12

Med. Chem. Lett. 2016, 26, 3632.
15. Li, Y. S.; Hu, D. K.; Zhao, D. S.; Liu, X. Y.; Jin, H. W.; Song, G. P.; Cui, Z. N.; Zhang, L. H.
Bioorg. Med. Chem. 2017, 25, 1852.
16. Cui, Z. N.; Li, X.; Tian, F.; Yan, X. J. Int. J. Mol. Sci. 2014, 15, 8941.
17. Cui, Z. N.; Li, Y.; Ling, Y.; Huang, J.; Cui, J. R.; Wang, R.; Yang, X. L. Eur. J. Med. Chem. 2010,
45, 5576.
18. Zhou, Z. Z.; Ge, B. C.; Chen, Y. F.; Shi, X. D.; Yang, X. M.; Xu, J. P. Bioorg. Med. Chem. 2015,
23, 7332.
13

Highlights
ꢀ
ꢀ
5-Phenyl-2-furan derivatives containing tetrahydro(iso)quinoline were synthesized.
The title compounds as selective PDE4B inhibitors exhibited strong activity both in vitro and in
vivo
ꢀ
SAR and molecular simulation studies were conducted.
14

Graphical Abstract
Tetrahydroquinoline and tetrahydroisoquinoline derivatives as potential
selective PDE4B inhibitors
Ya-Sheng Li a,c,#, Xing-Yu Liu a, #, Dong-Sheng Zhao b, #, Yi-Xian Liao a,
Lian-Hui Zhang a, Feng-Zhi Zhang c, Gao-Peng Song d,*, Zi-Ning Cui a,*
15