716
G. Manfroni et al. / Bioorg. Med. Chem. Lett. 19 (2009) 714–717
peptide is free in solution, but its emission dramatically decreases
when in close proximity to the quencher dabcyl molecule, that is,
when the QD complex is formed. Interference on Tat/TAR complex
formation in the presence of fixed concentration of quinolones is
clearly seen by the increase in fluorescence signal respect to what
observed in the peptide/nucleic acid titration. These measures
allow us to determine the inhibition constants for each compound,
reported in Table 1.25
in infected cells exhibited by the 2-phenylquinolones compounds
might be related to an unfavorable pharmacokinetic profile.
These new derivatives in fact are highly hydrophilic and dissolve
easily in aqueous buffers, as expected from highly charged com-
pounds. We hence measured the uptake of selected 2-phenylqu-
inolones by cultured cell lines to verify cell penetration for
compounds 2b, 2c, and 5b exhibiting in vitro activity, using 3a
and 16 as negative and 1 as positive controls. The amount of
compound found in the cell lysate after extensive washing (%
uptake) was measured at different times of incubation and did
not improve after 3 h, neither using different cells densities. At
the best experimental conditions, the percentages of quinolones
found in the cell lysate are shown in Table 1.26
It is clearly evident by these data how our positive control 1
shows a significant ability to be taken up by the cells, in particular
by the Jurkat cells employed for viral infections, its uptake percent-
age improving significantly, up to 40% by preincubating Jurkat cells
with the cationic detergent Lipofectamine (not shown). The
amount of all 2-phenylquinolones found in uninfected cells is on
the opposite very low, especially in Jurkat cells, at all conditions.
Compound 16, bearing no charged substituents at the phenyl ring
and no in vitro activity, is on the opposite the one with the highest
cell permeation profile. This data point out how the requirements
for activity in Tat/TAR system are also those unfavoring cellular
permeation. The remarkable uptake of 1 clearly allows activity in
infected cells, while the charged nature of 2-phenylquinolones
derivatives is detrimental to cell penetration. The unfavorable
pharmacokinetic parameters therefore likely account for the lack
of their activity and toxicity in cell lines, despite the good profile
of activity in Tat/TAR inhibition assays.
In conclusion, we report here the design, synthesis, and bio-
logical activity of novel quinolones characterized by the lack of
the usual C-3 carboxyl moiety and by the presence at the C-2
position of a phenyl group functionalized with different proton-
able side chains. These new small molecules are able to effi-
ciently interfere with Tat/TAR complex in vitro depending on
precise structural features. A clear important conclusion of our
structure–activity analysis is the requirement of only one basic
side chain at the para position of the phenyl ring for competition
of Tat/TAR binding, differently from the series of bi-aryl-hetero-
cycles17 chosen as model for the design of our phenylquinolones.
In that case it was clearly demonstrated how TAR binding is
achieved in the presence of two essential cationic substituents
at the phenyl ring.17
Our finding has also favorable implications, since it impacts
well on pharmacokinetic considerations. Although 5b, 2b, and 2c
in vitro data are similar or better than those found for a classical
quinolone as 1, it is evident that the cell parameters are profoundly
different, reflecting the ionic nature of 2-phenylquinolone deriva-
tives, devised from the pharmacophoric model of RNA-binding
compounds.10 However, the 2-phenylquinolones bearing a single
basic side chain have reduced molecular mass than the inactive
analogs, and are also those exhibiting reduced overall charge. The
analysis of in vitro data coupled with uptake studies therefore
gives us a clear indication on the requirements for antiviral activity
by the new 2-phenylquinolones. Namely, we will proceed explor-
ing the activity of novel compounds exhibiting only one side chain
but whose lipophilicity may be improved by variations at the quin-
olone nucleus as well as by modifications of the substituents at the
2-phenyl ring.
Compounds 2b, 2c, and 5b show inhibition constants compa-
rable or lower of the positive control (1) while all other deriva-
tives are less or totally inactive. Removal of all cationic side
chains is totally detrimental to activity, as demonstrated by
the absence of inhibition exhibited by 16. Notably, 2b and 2c
have only one basic side chain, while the corresponding phen-
ylquinolones 3b and 3c bearing two cationic side chains are less
active or inactive in these experimental conditions. The lower or
null activity of the two-side-chain compounds is also evidenced
by the Ki values of 3a and 4a. The low activity shown by 2a
points to the key importance of a nitrogen containing heterocy-
cle as protonable head in the mono-side-chain derivatives. Hav-
ing established that cationic substituents in meta of the phenyl
ring are detrimental to Tat/TAR inhibition activity, we tested
whether removal of any substituent at this position affected
activity. Very interestingly, compound 5b, the desamino analog
of the mono-side-chain 2b, is the most active derivative of all
2-phenylquinolones, indicating that this position could be free.
While steric and electronic contribution may be taken in ac-
count, it is tempting to speculate that in 2b an intramolecular
H-bond between the amino and the oxygen of the ether bridge
forces the orientation of the cationic side chain in a less favored
position for nucleic acid/peptide recognition.
The antiviral activity of the compounds able to interfere with
Tat/TAR interaction in FQA assays was evaluated using Jurkat
cells infected with laboratory-adapted HIV-1 strains. The virolog-
ical parameters considered were the reverse transcriptase (RT)
activity using specific template/primer. At 11, 18, and 21 days
post-infections of Jurkat cells, the RT titer of the drug treated
cells (1–100 lM) was only marginally lower compared to
the untreated control at all concentrations tested (see Sup-
plementary data). Furthermore, the cytotoxicity was evaluated
through the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium
bromide (MTT) assay: none of the compounds showed any cyto-
toxicity up to 100
the exception of 5b, showing a residual cytotoxicity in Jurkat
cells after 72 h of incubation (EC50 = 94.7 14.4 M). We there-
lM in CEM as well as in Jurkat cell lines, with
l
fore wondered if the lack of cytotoxicity and antiviral activity
Table 1
Biological evaluation of 2-phenylquinolone derivatives
Compound
Kia
(l
M)
% Uptakeb (Jurkat cells)
% Uptakeb (CEM cells)
2a
2b
2c
3a
3b
3c
4a
5b
16
1
3.85 0.86
1.95 0.50
1.58 0.29
3.03 0.51
2.60 0.43
NAc
NDd
0.94
0.82
0.67
NDd
NDd
NDd
3.27
25
NDd
1.80
1.10
1.20
NDd
NDd
NDd
NDd
NDd
7.50
NAc
1.02 0.12
NAc
2.18 0.34
23.2
a
Ki from FQA experiments performed with 10 aa FAM–Tat peptide in Tris 10 mM,
Acknowledgments
pH 7.5, Mg(ClO4)2 1 mM, NaCl 20 mM (TNMg), 0.01% Triton X-100.
b
Uptake experiments performed after 3 h of incubation at 37 °C with the indi-
This investigation was supported in part by the MIUR (PRIN
2006, Roma, Italy). We are grateful to Roberto Bianconi for excel-
lent technical assistance.
cated cell lines.
c
NA, not active.
d
ND, not determined.
Manfroni, Giuseppe
