1262
Table 1. In vitro PTP1B and TCPTP inhibitory activities of
9-24
IC50/¯Ma
Compound
Foldb
PTP1B
TCPTP
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
13.4 « 1.1
31.2 « 1.4
20.4 « 3.0
9.3 « 0.8
16.8 « 2.9
25.8 « 1.5
19.4 « 1.2
14.1 « 1.4
1.5 « 0.1
7.7 « 0.2
6.6 « 1.5
4.6 « 0.8
4.5 « 0.6
19.4 « 1.1
4.1 « 0.03
7.0 « 0.9
n.d.
>100
35.7 « 2.8
33.6 « 2.3
37.8 « 1.0
>100
47.6 « 0.8
44.8 « 0.8
15.5 « 0.1
17.9 « 0.06
20.0 « 1.2
12.6 « 0.5
11.7 « 2.1
28.6 « 1.0
5.2 « 0.3
10.0 « 1.3
®
>3.2
1.8
3.6
2.3
>3.9
2.5
3.2
10.3
2.3
3.0
2.7
2.6
1.5
1.3
1.4
aValues are mean of three experiments (n.d.: not detected).
bThis is the inhibition ratio of TCPTP/PTP1B.
Scheme 2. Synthesis of monomeric glycosides via click
chemistry, reagents and conditions: (a) VcNa (8 equiv),
CuSO4¢5H2O (6 equiv) in CH2Cl2/H2O, 6 h; (b) VcNa (8
equiv), CuSO4¢5H2O (6 equiv) in CH2Cl2/H2O/DMF, 24 h.
dimeric glucosyl 21 and mannosyl 23 showed best inhibitory
activities (IC50 = 4.5 and 4.1 ¯mol L¹1, respectively) and the
galactosyl dimer 22 was almost 5-fold less active (IC50
=
required for dimers (VcNa from 3 equiv to 8 equiv and
CuSO4¢5H2O from 2 equiv to 6 equiv).
19.4 ¯mol L¹1). This suggests the conformational change on the
C-2 position of sugar moiety rendered little impact toward
PTP1B activity whereas the C-4 epimerization was crucial for
inhibition with the equatorial bond being privileged. However,
the 1-distributed glucosyl dimer 24 (IC50 = 7.0 ¯mol L¹1) was
less potent compared to the side chain modified 21 and 23,
which is different from the structure-activity relationship (SAR)
pattern of related monomers (16 vs. 13 and 15) described above.
Furthermore, the IC50 value of dimeric galactoside contain-
ing amide bonds (compound 18, 7.7 ¯mol L¹1) lowered remark-
ably in comparison with its ester counterpart (compound 22,
19.4 ¯mol L¹1) while the same tendency was also observed for
the glucosyl derivatives (17 vs. 21) but not for mannosyl
derivatives (19 vs. 23). Finally, the best hit (amide 6-triazolyl
dimeric glucoside 17, IC50 = 1.5 ¯mol L¹1) was discovered
which exhibited almost 10-fold and 3-13-fold enhanced
inhibitory potency, respectively, compared to its corresponding
monomer 9 (IC50 = 13.4 ¯mol L¹1) and other dimeric glyco-
sides.
Our next attempt was to assess the specificity of the
acquired glycoside inhibitors. As noted, achievement of selec-
tive PTP1B inhibitors over other homologous PTPases (espe-
cially TCPTP with 77% sequence identity) is a challenging
task.3 Evidently, all assayed products 9-24 exhibited at least
several fold selectivity for PTP1B over TCPTP, SHP-1 (>3.3-
fold), SHP-2 (>4.9-fold), and LAR (>4.9-fold).
Monomeric glucosides 12, 13, and 16 and mannosides 11
and 15 exhibited satisfactory specificity over SHP-1, SHP-2,
and LAR (Table S1)13 though moderate selectivity over TCPTP
(1.8-3.6-fold, Table 1) was gained. Interestingly, the monomeric
galactosides 10 and 14 displayed culminant selectivity over
TCPTP as well as SHP-1, SHP-2, and LAR among the monomer
The mixed solvent system CH2Cl2/H2O for the cyclo-
addition of di-2-propynylamide benzene 7 with the azido-
glycosides 1-4 rendered unsatisfactory result due probably to the
poor solubility of this propargyl material. We then tended to add
DMF (CH2Cl2/H2O/DMF = 5:2:1, V/V/V) into the system for
simultaneously enhancing the solubility of the starting material
and the formed products (Scheme 2). To our delight, under such
improved condition with vigorous stirring for 24 h, amide
triazolyl glycosyl dimers were afforded smoothly in moderate
yields (71% for 17, 56% for 18, and 63% for 19) with the 1-
modified triazole 20 being an exception that was yielded almost
quantitatively (97%).
Next, all synthesized triazolyl-glycosides were assayed
through PTP1B and a panel of homologous PTPases including
TCPTP (77% sequence identity with PTP1B), SHP-1, SHP-2, and
LAR by previously developed methods.10,11 The data are listed
in Table 1 (for PTP1B and TCPTP) and Table S1 (for SHP and
LAR).13 The monomers were first observed to display micromole-
ranged PTP1B inhibitory activity (Table 1, compounds 9-16).
¹1
Obviously, the 1-distributed products 12 (IC50 = 9.3 ¯mol L
)
and 16 (IC50 = 14.1 ¯mol L¹1) were more active toward PTP1B
than the 6-distributed ones (compounds 9-11, IC50 = 13.4-
31.2 ¯mol L¹1; compounds 13-15, IC50 = 16.8-25.8 ¯mol L¹1),
indicating that the conjugation pattern of anomeric carbon-
modified glycosides was structurally advantageous comparing to
that of the side-chain (C-6) modified structures.
Generally, the dimers (Table 1, compounds 17-24, except
22) showed at least 2-fold enhanced activity compared to their
corresponding monomers which demonstrated our design of
bidentate inhibitors desirable. For the ester series 21-24, the
Chem. Lett. 2010, 39, 1261-1263
© 2010 The Chemical Society of Japan