a switch from kinase to phosphatase activity, which results in
dephosphorylation of the downstream proteins LuxU and LuxO.
Dephosphorylated LuxO does not promote the production of
small regulatory RNAs, and this event leads to the LuxR-dependent
transcription of target genes [21, 24].
Most of the AI-2 QS inhibitors reported primarily target the
receptor protein LuxP. However, in most cases LuxP is still able to
bind to LuxQ and AI-2 to form the LuxPQ complex. Hence, the
LuxPQ complex is a more accurate target for AI-2 QS inhibitors
with respect to the LuxP protein alone [25].
Herein, we describe a high-throughput virtual screening [26]
to identify AI-2 QS inhibitors based on the crystal conformations of
LuxPQ in complex with AI-2 (PDB entry: 2HJ9) published by the
Bassler lab [24], by the use of the DOCK6 program and SPECS
chemical database. The MM32 strain ofV. harveyiis chosen for the
anti-QS assay due to the absence of the LuxN receptor, needed to
respond to AI-1, and of LuxS, needed to catalyze the biosynthesis
of DPD [25]. The top 1000 compounds are selected by virtual
screening. Then 42 hits that are chosen by using consensus scoring
can be evaluated for their ability to inhibit the AI-2 QS system. We
found that seven inhibitors, including compound 5
(IC 50 ¼3.060.18μM), compound 12 (IC 50 ¼40.114.7μM),
compound 19 (IC 50 ¼ 5.51 1.05 μM), compound 23
(IC 50 ¼3.420.63μM), compound 27 (IC 50 ¼41.00.99μM),
compound 31 (IC 50 ¼19.8 1.48μM), and compound 33
(IC 50 ¼14.81.77μM) (Fig.2), were able to interfere with the
AI-2-QS system ofV. harveyiin the micromolar range. In particu-
lar, compounds 5, 19, and 23, exhibited IC 50 values of 3.06, 5.51,
and 3.42μM, respectively, with low effect or no effect onV. harveyi
growth rate.Fig. 2Chemical structures of seven compounds as LuxPQ inhibitors
In silicoIdentification of AI-2 Inhibitors 355