G. Min et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3769–3774
3771
90
80
70
60
50
40
30
pounds with an amide (CG-11) or a pyridine ring (CG-15 and 16)
only moderately inhibited PRL-3 activity and DLD-1 (PRL-3) cell
migration.
CG-707
The recovery of phosphorylation was examined in an attempt to
confirm the inhibition of PRL-3 phosphatase activity by CG-707
and BR-1. Two PRL-3 substrates were used: ezrin and cytokeratin
8. Ezrin, a direct target of PRL-3, interacts with radixin and moesin,
constituting the ERM complex, which connects the actin cytoskel-
eton with cell surface receptor. Phosphorylation at amino acid
Thr567 activates ezrin and controls the structural transitions of
this protein, which plays a pivotal role in tumor progression and
metastasis, therefore, we detected the residue phosphorylation of
the Thr567 residue.20,21 Cytokeratin 8 (KRT8) is reported to colo-
calize and interact with PRL-3 at lamellipodias, participating in cel-
BR-1
0.1
0.5
1
2
5
10
lular movement.22 Treatment with 20
lM CG-707 or BR-1 in DLD-1
Concentration (μM)
(PRL-3) cells increased the level of phospho-ezrin more than three-
fold when compared with DMSO-treated cells. CG-707 and BR-1
also demonstrated the ability to rescue phospho-cytokeratin 8
which is a PRL-3 mediated effector protein (Fig. 4). These data sug-
gests that CG-707 and BR-1 specifically inhibited the phosphatase
activity of PRL-3.
Figure 2. Inhibition of PRL-3 phosphatase activity by CG-707 and BR-1 was
observed using a DiFMUP assay.
number of cells in the lower chamber after 20 h. As shown in Ta-
ble 1, most of the synthetic derivatives strongly inhibited the
The expression of PRL-3 has been shown to increase cell migra-
tory and invasive ability. We reported that DLD-1 colon cancer
cells overexpressing PRL-3 exhibited increased migration and inva-
sion rates when compared with DLD-1 cells.22 To observe the rela-
tionship between PRL-3 expression level and the anti-migratory
activity of CG-707 and BR-1, the two compounds were treated to
cell lines other than DLD-1 (PRL-3). The numbers of cells seeded
were varied depending on the size and mobility of each cell line.
The migration assay was stopped when the cells in a control well
had moved and penetrated the membrane to approximately 70%
confluence. As shown in Figure 5A, CG-707 and BR-1 inhibited
the migration of DLD-1, A375P, MIA PaCa-2, and AsPC-1 cells
which represent a colon cancer cell line, a melanoma cell line,
and 2 pancreatic cancer cell lines, respectively.17 However, the
compounds did not affect the migration of DU 145 or MDA-MB-
231, which have relatively low level of PRL-3 expression and are
derived from prostate cancer and breast cancer, respectively. CG-
707 and BR-1 strongly blocked the migration of A375P and MIA
PaCa-2 cells which have higher endogenous PRL-3 expression lev-
els than the cell lines such as DU 145 and MDA-MB-231, that were
not affected by the compounds. These results showed that the anti-
migratory effects of the compounds correlated with the PRL-3
expression level of each cell lines (Fig. 5B).
migration of DLD-1 (PRL-3) cells at a 10 lg/ml concentration. The
thiazolidinedione derivative (CG-18) displayed moderate anti-
migratory activity but this compound causes cell cytotoxicity.
Given the potent anti-migration efficacy demonstrated by most
of the rhodanine derivatives, we decided to study the dose-re-
sponse of the most potent derivatives to obtain their IC50 values
for suppressing the transwell migration of the DLD-1 (PRL-3) cells.
As shown in Figure 3A, CG-707 and BR-1 inhibited the migration of
DLD-1 (PRL-3) cells in a dose-dependent manner with IC50 values
of 5 and 7
lM, respectively. After the cells were treated with
20 M of the compounds, more than 90% of cell migration was
l
blocked and cell invasion was blocked by over 60% at the same
concentration (Fig. 3B).
We next performed proliferation assays on DLD-1 (PRL-3) can-
cer cells treated with the compounds to rule out any indirect ef-
fects on cell migration and invasion due to cytotoxicity. DLD-1
(PRL-3) cells were allowed to grow for 24 h in 96-well plates in
the presence or absence of the synthetic compounds at a concen-
tration of 10 lg/ml. The cell toxicity data for all derivatives are
listed in Table 1, along with the anti-migration data for compari-
son. The 14 compounds all inhibited cell migration by over 60%
(62–98%) without significant cell toxicity when the cells were trea-
ted with the derivatives at a dose of 10
l
g/ml, the exception was
Our primary screen revealed a class of rhodanine molecules as
potent PRL inhibitors. The structure and activities of selected com-
pounds, CG-707 and BR-1, are shown in Figures 1 and 2. These
CG-6, which significantly inhibited cell growth by approximately
60%. A few other anti-migratory derivatives greatly inhibited
DLD-1 (PRL-3) cell proliferation. For example, CG-17 and CG-18
blocked cell migration (approximately 50%) and also inhibited
the cell proliferation (approximately 63%). For these derivatives,
cytotoxicity may have partially contributed to their overall anti-
migratory effect. However, the potent PRL-3 inhibitors, CG-707
and BR-1, did not inhibit the proliferation of DLD-1 (PRL-3) cells,
compounds exhibited similar IC50 values, 0.8
lM for CG-707 and
1.1 M for BR-1 in PRL-3 phosphatase assays, respectively. To
l
examine the selectivity of these compounds, we further examined
the inhibitory activities of the rhodanines by performing a phos-
phatase assay with 9 other phosphatases that represent different
classes of phosphatases, including tyrosine and dual-specific phos-
phatases. As shown in Table 2, using DiFMUP as the substrate for
all of the protein phosphatase (PTPase) assays, the compounds dis-
played excellent selectivity for the PRL-3 phosphatase, but showed
minimal effects on other phosphatases.
even at concentration as high as 50 lM (Fig. 3C). These results indi-
cate that these compounds significantly inhibit cell migration and
invasion without cytotoxicity and may be good lead molecules for
designing a new PRL-3 inhibitor.
Despite of having very limited existing structure–activity rela-
tionships (SAR), the compounds with benzyl group showed more
promising activity against PRL-3 compared with the other com-
pounds, and a dramatic loss of inhibitory potency was seen when
the rhodanine group was substituted by a thiazolidinedione (CG-
18). The replacement of the oxybenzyl group with a sulfonylbenzyl
(CG-10) provided more promising starting points for the design of
a new PRL-3 inhibition, because CG-10 strongly inhibited both PRL-
3 activity and the migration of DLD-1 (PRL-3) cells. However, com-
EMT, epithelial-to-mesenchymal transition, is a critical step in
the metastatic progression and invasion of cancer cells during
which epithelial cells gain fibroblastic properties and lose cell adhe-
sive ability. A decreased level of the epithelial cell marker protein,
E-cadherin, resulted in a loss of cell-cell adhesion junctions and
was followed by an increased invasiveness of tumor cells. It has
been reported that PRL-3 is associated with EMT, where the overex-
pression of PRL-3 promotes the mesenchymal transition.23 To
observe the modulation of EMT marker proteins, such as E-cadherin