Elucidation of a Screening Artifact
COMMUNICATION
electrostatic interactions between the negatively charged
carboxylate groups and the positively charged protein sur-
face. In this context it is of interest that some other polyan-
ions including polysulfated heparins have already been re-
To find out whether these inhibitory profiles can also be
observed for RNA polymerases from other species, RNAP
from bacteriophage T7 was employed. P1 showed a strong
À1
inhibitory effect (IC =0.06 mgmL ) whereas compound 1
5
0
[15–17]
À1
ported to inhibit RNAP.
was inactive at 200 mm (43.3 mgmL ). In contrast to these
findings, P1 showed no or only moderate activity against
seven other enzymes (Table S2 in the Supporting Informa-
tion). In a further experiment, P1 did not inhibit growth of
The electrostatic interaction of P1 with RNAP was dem-
onstrated by gel electrophoresis. In the presence of RNAP
the mobility of P1 was reduced (Figure S7 in the Supporting
Information). Considering the averaged molecular weight of
around 40 kDa, the IC50 value of P1 against E. coli RNAP
can be calculated to be approximately 15 nm. Since the con-
centration of enzyme in the assay is 36 nm, it can be con-
cluded that P1 binds to the protein at a molar ratio of about
À1
E. coli and Pseudomonas aeruginosa (at 20 mgmL ). The
lack of in vivo activity was expected as P1 is too hydrophilic
for passive diffusion and too large to permeate the porins
and enter the bacterial cell.
Nevertheless, it is feasible that the knowledge gained in
this study could be exploited for an antibacterial therapy.
Biological in vitro effects of polyanions including inhibition
1
:1 and is a very potent inhibitor. Under identical experi-
mental conditions a 70-fold excess of heparin is required to
obtain the same inhibitory effect.
[15–17]
of RNAP have indeed already been described.
Howev-
On closer examination it is obvious that the bound poly-
mer should sterically block the loading of the DNA tem-
plate to RNAP and thus inhibit transcription at its earliest
step. To corroborate this hypothesis, the influence of P1 on
DNA binding to the RNAP main channel was investigated.
We expected that P1 should be less effective if the DNA
template already occupied the main channel. Indeed, our re-
sults revealed a significantly reduced RNAP inhibition when
the DNA template was added prior to P1 compared to a
vice versa chronology of addition. For rifampicin, which af-
fects the transcription process by binding to an allosteric site
er, the polymers were devoid of antibacterial activity. It
seems viable that in vivo activity could be achieved for our
polymers by strongly reducing their size, thus enabling hy-
drophilic substances to penetrate the porins. On the other
hand polyanions of such a molecular weight should not be
capable of passing the membranes of mammalian cells, and
therefore they could be potential therapeutics for local in-
fections of the skin and lung.
In summary, we could demonstrate that the activity of the
hit compound 1 found in an experimental screening ap-
proach can be attributed to a highly active polymeric impur-
ity formed by decomposition of 1. Remarkably, the closely
[19]
close to the active site within the DNA binding channel,
a
[7]
similar trend was observed (Figure 5). As a negative control,
CBR703, was used as it is described to bind to a surface-ex-
posed groove distant from the main channel, which should
not influence the binding of the DNA template to its chan-
related compound 4 is also described as a RNAP inhibitor.
In our assay “synthesized” 4 was inactive, whereas the poly-
merized compound strongly inhibited the bacterial enzyme.
Furthermore, 2 was also reported to be active against polio-
[
20]
[21]
nel. As expected, CBR703 inhibited RNAP independently
of whether the DNA channel was already occupied by the
DNA template or not. In summary, this experiment supports
our hypothesized mechanism of P1 RNAP inhibition.
virus RNAP,
which seems doubtful with respect to our
findings. It is worth mentioning that compounds 1–4 are re-
[
22]
[23]
ported to inhibit HIV-1 gp41, anthrax lethal factor and
Eph receptors.
[24]
Remarkably around 500 related deriva-
tives of N-phenyl-2,5-dimethyl
pyrrole are reported in numer-
ous patents and publications as
biologically
pounds;
active
nevertheless, they
com-
[
25]
do not have a clean slate and
have been propagated as pan
assay interference compounds
[26]
(
PAINS). However, the mo-
lecular mechanism of the false
positive results remained “mys-
terious” until this investigation.
In conclusion, researchers
should be cautious with com-
pounds that tend to polymer-
ize, as even traces of polymeric
impurities might cause tremen-
dous effects in biological
assays.
Figure 5. E. coli RNAP transcription assay with different chronology of addition of DNA template and inhibi-
tor. Solid line: normal test procedure: inhibitor was added before the DNA template. Dashed line: reversed
test procedure: DNA template was added before the inhibitor. Concentration-dependent inhibition of: a) P1,
b) rifampicin, c) CBR703; d) structures of rifampicin and CBR703.
Chem. Eur. J. 2013, 00, 0 – 0
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
3
&
ÞÞ
These are not the final page numbers!