Synthesis and biological evaluation of xanthine derivatives on dipeptidyl peptidase 4

Article abstract of DOI:10.1248/cpb.c12-01046

A series of xanthine derivatives in which a methylene was inserted at position 8 of xanthine scaffold was synthesized and evaluated as inhibitors of dipeptidyl peptidase 4 (DPP-4) for the treatment of type 2 diabetes. As the results of structure-activity relationship (SAR) study of the series, the compounds with 4-methylquinazoline- 2-yl-methyl group at N-1 position and 2-aminoethylaminomethyl group gave better activities. Compounds H4 and H9 showed good DPP-4 inhibition and more than 100-fold selectivity over DPP-7 and DPP-8.

Full text of DOI:10.1248/cpb.c12-01046

April 2013
Note
Chem. Pharm. Bull. 61(4) 477–482 (2013)
477
Synthesis and Biological Evaluation of Xanthine Derivatives on
Dipeptidyl Peptidase 4
Kuaile Lin, Zhengyan Cai, Fei Wang, Wei Zhang, and Weicheng Zhou*
State Key Lab of New Drug & Pharmaceutical Process, Shanghai Key Lab of Anti-Infectives, Shanghai Institute of
Pharmaceutical Industry, State Institute of Pharmaceutical Industry; Shanghai 200437, China.
Received December 6, 2012; accepted January 12, 2013; advance publication released online January 28, 2013
A series of xanthine derivatives in which a methylene was inserted at position 8 of xanthine scaffold was
synthesized and evaluated as inhibitors of dipeptidyl peptidase 4 (DPP-4) for the treatment of type 2 diabe-
tes. As the results of structure–activity relationship (SAR) study of the series, the compounds with 4-methyl-
quinazoline-2-yl-methyl group at N-1 position and 2-aminoethylaminomethyl group gave better activities.
Compounds H4 and H9 showed good DPP-4 inhibition and more than 100-fold selectivity over DPP-7 and
DPP-8.
Key words dipeptidyl peptidase; dipeptidyl peptidase 4 inhibitor; xanthine
Diabetes is establishing itself as a serious epidemic of the xanthine scaffold, has been proved as a highly potent and
21st century. Data from global studies demonstrates that the selective DPP-4 inhibitor.5) The X-ray crystal structure of
number of people suffering from the disease in 2011 has linagliptin in the complex with human DPP-4 showed that the
reached a staggering 366 million, including 4.6 million deaths, aminopiperidine at C-8 of the xanthine scaffold occupied the
and the health care spending on diabetes has reached 464 S2 subsite. Its primary amine formed a network of charge-
billion dollars.1) Type 2 diabetes (T2D), formerly noninsulin- reinforced hydrogen bonds to Glu205, Glu206 and Tyr662, the
dependent diabetes, accounts for at least 90% of all cases of residues that constitute the recognition site for the amino ter-
the disease.1) Inhibition of dipeptidyl peptidase 4 (DPP-4) has minus of DPP-4. The butynyl substituent at N-7 occupied the
proven to be an effective treatment for improving glycemic hydrophobic S1 pocket of the enzyme. The xanthine moiety
control in type 2 diabetic patients.2) DPP-4 cleaves a wide laid on top of Tyr547, forming aromatic π-stacking interac-
range of peptides to modulate their biological activity. One of tions with the phenol of Tyr547. The quinazoline substituent
these peptides is glucagon-like peptide 1 (GLP-1), which plays at N-1 interacted with Trp629 by π-stacking interactions.5) So
an important role in the regulation of blood glucose level.3) far, most of existing reports on the structural modifications of
GLP-1 is released after food ingestion and stimulates insulin xanthine derivatives were focused on the substituents at N-1,
biosynthesis and secretion. GLP-1 also inhibits glucagon re- N-7, and the xanthine scaffold.5–10) In this paper, a series of
lease, delays gastric emptying, and induces pancreatic β-cell new xanthine derivatives in which a methylene was inserted
proliferation.4)
at position 8 in xanthine scaffold were designed, illustrated
In recent years, a number of DPP-4 inhibitors, such as as formula H (Table 1). And molecular docking between the
sitagliptin (1), vildagliptin (2), saxagliptin (3), alogliptin (4) target compounds and DPP-4 was studied with the molecu-
and linagliptin (5) (Fig. 1) have been launched and showed lar modeling software Discovery Studio 2.5 (Accelrys, Inc.).
great efficacy in patients with T2D. Linagliptin (5), with the As shown in Fig. 2, compound H6 bound to DPP-4 in very
Fig. 1. Structures of Launched DPP-4 Inhibitors
The authors declare no conflict of interest.
© 2013 The Pharmaceutical Society of Japan
*To whom correspondence should be addressed. e-mail: profzhouwc@yahoo.com.cn

478
Vol. 61, No. 4
Table 1. The DPP-4 Inhibitory Activity of H1–H45
Compound
Inhibition ratio (%)
IC50 (µm)
R1
R7
RN
H1
H2
94.31
21.2
1.57
—
H3
1.09
—
H4
94.43
69.55
77.86
32.23
33.56
97.04
38.43
49.57
53.31
43.49
95.64
68.59
79.03
0.84
7.17
5.10
—
H5
H6
H7
H8
–NH2
H9
0.72
—
H10
H11
H12
H13
H14
H15
H16
11.47
7.81
—
1.54
6.69
3.50

April 2013
479
Table 1. The DPP-4 Inhibitory Activity of H1–H45
Compound
Inhibition ratio (%)
IC50 (µm)
R1
R7
RN
H17
78.89
6.44
H18
H19
H20
H21
H22
H23
H24
79.03
38.77
58.89
59.69
29.81
38.58
47.42
5.34
—
14.63
—
—
—
18.09
H25
H26
43.62
41.03
—
—
H27
H28
H29
H30
H31
–NH2
2.79
28.08
2.72
—
—
—
—
—
0.87
5.98

480
Vol. 61, No. 4
Table 1. The DPP-4 Inhibitory Activity of H1–H45
Compound
Inhibition ratio (%)
IC50 (µm)
R1
R7
RN
H32
2.91
1.27
2.23
2.27
7.85
5.31
3.32
21.42
9.08
31.16
16.06
2.44
2.57
6.84
—
—
—
—
—
—
—
—
—
—
—
—
—
H33
H34
H35
H36
H37
H38
H39
–NH2
H40
H41
H42
H43
H44
H45
–CH3
–CH3
–CH3
–CH3
–CH3
–CH3
–NH2
Sitagliptin
Linagliptin
—
—
0.035
0.001

April 2013
481
Fig. 2. Docking Modes of Compounds H6 (Blue) and H9 (Magenta) Overlaid with the Bound Conformation of Linagliptin (Cyan, PDB Entry 2RGU)
Color image was converted into gray scale.
Reagents and conditions: (a) Na, EtOH, 70%; (b) NaNO2, AcOH, H2O, 75%; (c) Na2S2O4, ammonium hydroxide, 96%; (d) i) glycolic acid, H2O, ii) NaOH, H2O, 79%;
(e) R7X (X=Cl or Br), DIEA, DMSO, 53–82%; (f) R1X (X=Cl or Br), K2CO3, DMF, 30–89%; (g) MeSO2Cl, Et3N, CH2Cl2, 30–95%; (h) RNH, CH2Cl2 or i) RNH, CH2Cl2,
Et3N; ii) CF3COOH, CH2Cl2, 32–97%.
Chart 1
similar mode as linagliptin. So 45 novel xanthine derivatives
Biological Activity DPP-4 was extracted from confluent
were synthesized and their inhibition on DPP-4 was evaluated. Sf9 cell. The activity was measured as described,13,14) using the
substrate Gly-Pro-p-nitroanilide, which can be decomposed to
Gly-Pro and p-nitroaniline by DPP-4. Compounds H1–H45
Results and Discussion
Chemistry The synthesis of new xanthine derivatives were dissolved in 1% DMSO aqueous solution and were tested
was shown in Chart 1. Condensation of methylurea and ethyl at a fixed level of 10µg/mL. The compounds with the inhibi-
cyanoacetate followed by nitrosation gave compound B.11) tion rate over 50% went through into the second round of
Reduction of nitroso group with sodium hydrosulfite yielded screening where the inhibitory concentration 50% (IC50) was
compound C.12) The xanthine ring was formed by the conden- determined and the result was included in Table 1. Among
sation of compound C with glycolic acid. The alkylation at the compounds in the first round testing, 14 compounds were
N-7 position was carried out by treatment with R7X (X=Cl or found to possess moderate or significant activity. In the sec-
Br) and diisopropylethylamine in dimethyl sulfoxide (DMSO). ond screening, compounds H1, H4, H9 and H14 showed the
The compounds E were further alkylated at N-1 position best inhibitory activity. In order to confirm the results as well
with R1X (X=Cl or Br) and potassium carbonate in N,N- as to study the selectivity among DPP-4, DPP-7 and DPP-8,
dimethylformamide (DMF). The activation of the hydroxy compounds H4, H9 and H14 were sent to the National Cen-
with methanesulfonyl chloride gave methylsulfonate G. The ter for Drug Screening for the selective inhibition testing.
target products were achieved through nucleophilic substitu- The data were given in Table 2. As far as structure–activity
tion by amines. All the structures of target compounds were relationship (SAR) was concerned, variations at N-1 position
1
confirmed by H-NMR and MS.
showed that the 4-methyl-quinazolin-2-yl-methyl (H1–H13)

482
Vol. 61, No. 4
Table 2. The DPP-4, DPP-7 and DPP-8 Inhibitory Activity of H4, H9 and H14
DPP-4
DPP-7
Inhibition ratio at
DPP-8
Inhibition ratio at
Compound
Inhibition ratio at
IC50 (µm)
IC50 (µm)
IC50 (µm)
20µg/mL
250µg/mL
250µg/mL
H4
H9
H14
92.54%
92.29%
86.92%
3.08
2.66
5.34
14.43%
37.36%
87.61%
>530
>491
58.9
63.21%
68.26%
97.30%
314
223
63.8
and α-naphtylmethyl (H14–H24) substitutions exhibited can be found via article’s webpage.
better inhibitory effect for DPP-4 than 6,7,8-trifluoro-4-(4-
fluorophenylthio)quinoline-3-yl-methyl, methyl, 2-cyanobenzyl
Acknowledgements This work was supported by 973
or 6-fluoro-7-chloro-4-methoxyquinoline-3-yl-methyl substitu- program of the Ministry of Science and Technology of China
tion; and it seemed that compounds with 4-methyl-quinazo- (2010CB735601 and 2012CB724501). We thank the National
line-2-yl-methyl substitution gave better selectivity than those Center for Drug Screening for the selective inhibition testing
with α-naphtylmethyl group (Table 2). At N-7 position, both on DPP-4 over DPP-7 and DPP-8.
2-cyanobenzyl (H8–H13) and but-2-ynyl substitution (H1–
H7) performed similar effects on the activity. For RN com-
pounds, 2-aminoethylamino group was better for the inhibi-
tory activity than its homologues (H1>H2 and H3, H9>H8) in
all the series, indicating that the preferable distance between
two nitrogen atoms might be equivalent to two CH2. For the
compounds with the ring substitutions, some active com-
pounds were found in piperazinyl, homopiperazinyl and 3(R)-
aminopiperidinyl substitution in which the distance between
References
1) International Diabetes Federation. “Diabetes Estimates Excel
Tables”: ‹http://www.idf.org/diabetesatlas/diabetes-estimates-tables›,
cited 22 May, 2012.
2) Mclntosh C. H. S., Front. Biosci., 13, 1753–1773 (2008).
3) Kieffer T. J., McIntosh C. H., Pederson R. A., Endocrinology, 136,
3585–3596 (1995).
4) Farilla L., Hui H., Bertolotto C., Kang E., Bulotta A., Di Mario U.,
Perfetti R., Endocrinology, 143, 4397–4408 (2002).
two nitrogen also was two CH2. In most cases, flexible chain
(2-aminoethylamino) seemed better than the rigid ring (pi-
perazinyl, homopiperazinyl and 3(R)-aminopiperidinyl).
In beginning of the work, molecular docking strategy was
used to investigate the binding mode of linagliptin and guide
the procedure of molecular modification. Based on these
preliminary docking results, comparable biological activities
were expected. However, validated by bioassay, this series of
compounds bear far lower binding affinities than linagliptin
against DPP-4, for example, compound H6 which was very
similar to linagliptin did not give potent activity, indicating
the limit of molecular modeling strategy. In another hand,
when compound H9, the most active one in this series, was
taken into the molecular docking (Fig. 2), it was found that
H9 bound to DPP-4 in a pose more matched with the lina-
gliptin scaffold. In addition, 2-cyano benzyl bound to the con-
served hydrophobic pocket might contribute a more favorable
binding mode.
5) Eckhardt M., Langkopf E., Mark M., Tadayyon M., Thomas L., Nar
H., Pfrengle W., Guth B., Lotz R., Sieger P., Fuchs H., Himmels-
bach F., J. Med. Chem., 50, 6450–6453 (2007).
6) Kurukulasuriya R., Rohde J. J., Szczepankiewicz B. G., Basha F.,
Lai C., Jae H. S., Winn M., Stewart K. D., Longenecker K. L., Lub-
ben T. W., Ballaron S. J., Sham H. L., von Geldern T. W., Bioorg.
Med. Chem. Lett., 16, 6226–6230 (2006).
7) Nishio Y., Kimura H., Sawada N., Sugaru E., Horiguchi M., Ono
M., Furuta Y., Sakai M., Masui Y., Otani M., Hashizuka T., Honda
Y., Deguchi J., Nakagawa T., Nakahira H., Bioorg. Med. Chem., 19,
5490–5499 (2011).
8) Eckhardt M., Hauel N., Himmelsbach F., Langkopf E., Nar H.,
Mark M., Tadayyon M., Thomas L., Guth B., Lotz R., Bioorg. Med.
Chem. Lett., 18, 3158–3162 (2008).
9) Nishio Y., Kimura H., Tosaki S., Sugaru E., Sakai M., Horiguchi
M., Masui Y., Ono M., Nakagawa T., Nakahira H., Bioorg. Med.
Chem. Lett., 20, 7246–7249 (2010).
10) Sutton J. M., Clark D. E., Dunsdon S. J., Fenton G., Fillmore A.,
Harris N. V., Higgs C., Hurley C. A., Krintel S. L., MacKenzie R.
E., Duttaroy A., Gangl E., Maniara W., Sedrani R., Namoto K.,
Ostermann N., Gerhartz B., Sirockin F., Trappe J., Hassiepen U.,
Baeschlin D. K., Bioorg. Med. Chem. Lett., 22, 1464–1468 (2012).
11) Elzein E., Kalla R. V., Li X. F., Perry T., Gimbel A., Zeng D. W.,
Lustig D., Leung K., Zablocki J., J. Med. Chem., 51, 2267–2278
(2008).
Conclusion
In summary, a novel class of DPP-4 inhibitors, derived
from a xanthine scaffold, has been identified. The compounds
with 4-methyl-quinazoline-2-yl-methyl group at N-1 position
and 2-aminoethylaminomethyl group at RN showed better
activities. Compounds H4 and H9 showed good DPP-4 inhibi-
tion and more than 100-fold selectivity over DPP-7 and DPP-8,
which were worthy to the further modification and research.
12) Balssa F., Bonnaire Y., J. Labelled Compd. Radiopharm., 50, 33–41
(2007).
13) Nagatsu T., Hino M., Fuyamada H., Hayakawa T., Sakakibara S.,
Nakagawa Y., Takemoto T., Anal. Biochem., 74, 466–476 (1976).
14) Pascual I., Lopéz A., Gómez H., Chappé M., Saroyán A., González
Y., Cisneros M., Charli J. L., Chávez M. A., Enzyme Microb. Tech-
nol., 40, 414–419 (2007).
Supporting Information Full experimental details, physi-
cal and spectral data of the compounds A–H. This material