F. Vacondio et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5290–5294
5293
Figure 3. Time course of A549 cell uptake for compds 3, 5–6 at 37 °C (black diamonds) and 4 °C (white diamonds). Each point represents the mean SD of three independent
experiments.
chemically- or metabolically-driven reactive species.18 We had
previously monitored the formation of conjugates with cysteine
in the presence of A549 cell lysate by HPLC–MS for both the acryl-
amide 2 and the 3-dimethylaminopropanamide 4, which generates
acrylamide 2 by a retro-Michael reaction. In that test, both com-
pounds gave the same cysteine-conjugate at m/z = 490.05.9 For
compounds 5 and 7–8, no conjugate was detected up to 24 h of
incubation under the same experimental conditions. This result
rules out any bioconversion to electrophilic species able to react
with cysteine.
To further characterize the mechanism by which reversible
inhibitors 5 and 6 appear to accumulate in A549 cells, we exam-
ined the time-course of their uptake at the concentration of 1 lM
at 37 °C versus 4 °C (see Supplementary data for experimental de-
tails). The observed temperature dependency (see Fig. 3) indicated
that A549 cell uptake of 5 and6 was predominantly an active pro-
cess. In fact, the passive influx component at 4 °C never exceeded
17.9( 3.3)% or 25.2( 3.5)% of the total influx for 5 or6, respectively.
Moreover, cell uptake increased in a time-dependent manner
reaching a plateau for both compounds, so it was a saturable pro-
cess. On the contrary, cell uptake of compound 3 which does not
accumulate within cells, showed no temperature dependency, sug-
gesting the predominance of a passive diffusion mechanism. We
also evaluated if the high intracellular concentrations of 5–6 could
be explained by a high affinity of these basic compounds for intra-
cellular components. Equilibrium dialysis experiments on com-
pound-spiked A549 cell homogenates (see Supplementary data
for experimental details) returned a percentage of unbound drug
equal to 15.7( 2.2)% for 5, 7.7( 0.3) for6, if compared to a
32.2( 0.8)% for gefitinib (1). The high and long-lasting intracellular
accumulation of 5 and 6 can therefore also be related to their high
binding to intracellular components.
Intracellular concentrations of compounds 5–8 were then mea-
sured in A549 lung cancer cells by HPLC-ESI-MS/MS under condi-
tions reproducing those of the EGFR autophosphorylation assay,
that is, immediately after 1 h exposure to the inhibitor (1 lM) or
after a following period (8 h) of drug-free incubation. Results, re-
ported in Table 1, were compared with those obtained for com-
pounds 2–4.9 Intracellular concentrations of 2 were below the
limit of detection (LOD) of the bioanalytical method at both 1 h
and 8 h. This is probably due to the high reactivity of the acrylam-
ide, able to covalently interact with different cell components.
Compounds 3 and 4 were detected in cells after 1 h incubation,
while their concentrations dropped below LOD after 8 h. The disap-
pearance from the intracellular environment of compound 4 is
consistent with the intracellular conversion into the reactive acryl-
amide 2, previously shown by us.9 All the other EGFR inhibitors un-
der evaluation (5–8) showed high intracellular concentrations 1 h
after compound administration, well above their IC50 values on re-
combinant EGFR. Compound 6 was the most concentrated in A549
Compounds 7 and 8, having side chains in 6 without a basic
group, showed intracellular concentrations at 1 h of 72 and
240 lM, respectively. At 8 h, their intracellular concentrations
dropped of a 103-fold (7) and 67-fold (8) factor, even if their
IC50s under the same conditions showed only a limited decrease
(1.8-fold for 7 and 2.4-fold for 8). On the other hand, in the case
of the reversible inhibitor 1, comparable decrease in intracellular
concentrations led to a 12.5-fold drop in IC50 1 h versus 8 h. Mean-
while, for compound 5, a 2.4-fold decrease in IC50 was related to
the maintenance of higher intracellular concentrations. Com-
pounds 7 and 8 were also the least potent of the series in the kinase
assay on recombinant EGFR. Therefore, even if a rapid covalent
interaction with Cys797 of EGFR was ruled out by fluorescence-
cells at 1 h (324
lM) and concentrations in the micromolar range
were also maintained at 8 h (105
lM). Despite compound 6 had
been designed as a reversible inhibitor, it showed 90% inhibition
of EGFR autophosphorylation at 8 h. This can be fully explained
by its intracellular concentration at 8 h, which largely exceeds
the reported IC50 value. The other reversible inhibitor 5, where
homologation of the side chain was intended to avoid b-elimina-
tion and acrylamide generation, had an intracellular concentration
based and reactivity assays,
a reversible inhibition of EGFR
of 147
l
M at 1 h, which dropped to 7.2
l
M 8 h after. In spite of the
followed by a slower covalent inactivation in cells during the 8 h
cannot be excluded.
reduction in intracellular concentration, this still exceeded the re-
ported IC50 and is thus sufficient to explain the observed 96% inhi-
bition of the enzyme. Thus, reversible 5 and 6 demonstrated to be
highly accumulated in cells at 8 h. Compound 6, which is as potent
as the acrylamide 2 in inhibiting EGFR autophosphorylation both at
1 h and 8 h, shows with particular evidence the possibility of get-
ting long-lasting EGFR autophosphorylation inhibition in A549
cells by the intracellular accumulation of the inhibitor, even in
the absence of reactive groups able to covalently inhibit the
enzyme.
A similar behavior has also been reported for the reference
reversible inhibitor gefitinib (1), bearing a basic morpholine in
the side chain.9,10 In fact, the calculated intracellular concentration
of 1 at 8 h (2.3
inhibition of autophosphorylation at the same time (46% of inhibi-
tion, 1 M).
In conclusion, with our selected set of EGFR inhibitors we
proved that the long-lasting inhibition of EGFR autophosphoryla-
tion in cells, if taken as a single indicator, is not adequate to classify
irreversible covalent inhibitors from reversible inhibitors which
accumulate within cells. Cell-based autophosphorylation assay
need to be complemented by other experimental data to shed light
on the mechanism of action of newly synthesized EGFR inhibitors.
The combination of: (i) a fluorescence-based assay on EGFR kinase
for covalent bond formation, (ii) a reactivity assay in the presence
of a trapping agent for reactive species in cell lysate, and (iii)
HPLC–MS/MS dosing of intracellular concentrations at different
time points can be efficiently employed to this aim. Moreover,
the proposed approach could be exploited to improve the potency
and to better understand the mechanism of action of inhibitors tar-
geting non catalytic cysteines by covalent bond formation.
lM) was consistent with the observed percentage of
l
Vacondio, Federica
