Identification of promiscuous small molecule activators in high-throughput enzyme activation screens

Article abstract of DOI:10.1021/jm701583b

It is recognized that high-throughput enzyme inhibition screens often return nonspecific inhibitors as hits . Recently, high-throughput screens for enzyme activators have led to the identification of several compounds with novel and potent biological activity. Here, we show that enzyme activation screens can also uncover compounds that activate multiple enzymes in a nonspecific fashion. Described herein are the general structural features of such compounds and methods to differentiate between specific and general enzyme activation.

Full text of DOI:10.1021/jm701583b

2346
J. Med. Chem. 2008, 51, 2346–2349
Letters
Identification of Promiscuous Small Molecule
Activators in High-Throughput Enzyme
Activation Screens
David R. Goode,† Ryan K. Totten,† James T. Heeres,‡ and
Paul J. Hergenrother*,†,‡
Departments of Chemistry and Biochemistry, Roger Adams
Laboratory, UniVersity of Illinois, 600 S. Mathews AVenue,
Urbana, Illinois 61801
ReceiVed December 17, 2007
Abstract: It is recognized that high-throughput enzyme inhibition
screens often return nonspecific inhibitors as “hits”. Recently, high-
throughput screens for enzyme activators have led to the identification
of several compounds with novel and potent biological activity. Here,
we show that enzyme activation screens can also uncover compounds
that activate multiple enzymes in a nonspecific fashion. Described herein
are the general structural features of such compounds and methods to
differentiate between specific and general enzyme activation.
Figure 1. Structures of hits and their potency in the procaspase-2
activation assay.
to determine if certain compounds induce a nonspecific activa-
tion effect, analogous to the effect of nonspecific inhibitors. A
brief mention of such an activation effect was noted during a
high-throughput screen for promiscuous enzyme inhibitors,
although the effect was attributed to low-volume conditions of
HTS formats.16 Here, we document compounds that show
promiscuous enzyme activation; as described below, these
compounds were initially discovered during a high-throughput
screen for procaspase-2 activation.
High-throughput screening (HTS) is now a major method for
the discovery of new small molecule leads.1–3 It is increasingly
recognized that in high-throughput enzyme inhibition assays
some compounds will inhibit the enzyme of interest in a
nonspecific fashion.4 This promiscuous inhibition appears to be
the result of compound aggregation into colloidal particles.5,6
The nonspecific inhibition effect is well documented and has
allowed for the early identification of potential aggregators via
the development of structural filters.7 Importantly, there now
exists a battery of experiments that can be used to distinguish
specific from nonspecific enzyme inhibition.8,9
Although high-throughput screens for the identification of
enzyme inhibitors comprise the majority of HTS efforts, there
is a growing interest in enzyme activation assays, where small
molecules that “turn on” enzymatic activity are sought. These
can be compounds that replace biological ligands that are known
to enhance enzymatic activity, or they can be compounds that
enhance enzyme/proenzyme activity for which there is no known
endogenous biological activator. For instance, the activation of
glucokinase to alter glucose homeostasis has been proposed to
treat type 2 diabetes,10 and multiple allosteric activators are
currently being studied.11 Similarly, small molecule activators
of soluble guanylate cyclase are also being pursued to control
cGMP signaling.12 HTS campaigns were used to discover novel
activators of RNase L13 and SIRT1.14 Finally, a small molecule
activator of procaspase-3 (called PAC-1) was recently identified
through HTS; PAC-1 enhances the activity of procaspase-3 in
vitro and induces apoptosis in cancer cells in cell culture and
in vivo.15
Compounds from an in-house library17 (∼22 000 compounds)
were screened for their ability to increase the catalytic activity
of procaspase-2. Procaspase-2 was incubated with compound
(∼20 µM) for 16 h at 37 °C in 384-well plates. At this point,
the known caspase-2 substrate N-acyl-Val-Asp-Val-Ala-Asp-
p-nitroanalide (Ac-VDVAD-pNA)18 was added and the absor-
bance at 405 nm was monitored for 2 h. Compounds were
considered hits if they induced an activity greater than 2 standard
deviations above the average of vehicle treated controls. This
screen produced 147 hits, a hit rate of 0.7%. Through secondary
evaluations in the same assay, it was determined that seven
compounds enhanced procaspase-2 enzyme activity in a dose-
dependent manner and with reasonable potency. Potency was
evaluated by the concentration required to induce half-maximal
activation; the seven compounds shown in Figure 1 have EC50
values for procaspase-2 activation in the low micromolar or mid-
micromolar range.
Several hits from this screen showed very unusual structural
features, with some having polyamine structures (1 and 2) and
others appearing surfactant-like (3 and 4). The remaining library
hits (5-7) showed more druglike characteristics, two of which
(6 and 7) had considerable structural similarity. Thus, further
chemical derivatization was focused on compounds such as 6
and 7, with the goal of identifying even more potent activators
of procaspase-2.
On the basis of these and other success stories, the use of
HTS for identification of compounds that enhance the activity
of specific enzymes will likely increase. As such enzyme
activation screens become more prevalent, it will be important
The synthesis of 6 commenced with the functionalization of
epichlorohydrin to generate epoxide 8, followed by amine
opening of the epoxide to provide 6 (Scheme 1). The efficient
synthesis of 6 inspired a parallel synthesis strategy whereby
* To whom correspondence should be addressed. Phone: (217) 333-0363.
Fax: (217) 244-8024. E-mail: hergenro@uiuc.edu.
†
Department of Chemistry.
‡
Department of Biochemistry.
10.1021/jm701583b CCC: $40.75
2008 American Chemical Society
Published on Web 03/27/2008

Letters
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 8 2347
Scheme 1. Synthesis of 6
surfactants and surfactant-like compounds showed activation of
chymotrypsin. The effect of different concentrations of CHAPS,
6, and 7 on chymotrypsin activity is shown in Figure 2A, and
the activity vs time progress curves for 7 are shown in Figure
2B. The activating effect of surfactants on enzymatic activity
has been previously documented and is thought to be due to
prevention of oligomerization or loosening of protein secondary
structure.20 This effect could also be due to the surfactant-like
molecules masking hydrophobic patches on the protein surface,
as has been seen previously for CHAPS.21
In order to investigate the mechanism of activation by the
HTS hits, the chymotrypsin assay was repeated with 0.1% (w/
v) CHAPS added to the assay buffer. In all cases, the addition
of CHAPS abrogated the activation conferred by the HTS hits.
Maximal enzymatic activity was obtained when CHAPS was
added to the buffer, regardless of the presence of the HTS hits;
this is seen when comparing the highest activation of a HTS
hit to the activity of enzyme alone in the CHAPS buffer (shown
for 7 in Figure 2B,C). This nonadditive effect of CHAPS with
the compounds points to their behaving in a similar manner,
likely as surfactants in the buffer.
adamantyl-linked epoxides 8 and 9 were crossed with 42
different amines to provide an 84-member library (Scheme 2).
The library members were synthesized in an average yield of
63% to give an average mass of 19.5 mg after purification by
SiO2 column chromatography. After this purification, all final
products appeared as a single spot upon analysis by thin-layer
chromatography (see Supporting Information Figure S1).
Interestingly, upon evaluation in the procaspase-2 activation
assay, none of the derivatives were found to be more potent
activators than the parent 6, although the vast majority of the
compounds retained the ability to enhance the activity of
procaspase-2.
Compounds found to enhance the activity of procaspase-2
were generally surfactant-like, containing a charged headgroup
and a hydrophobic tail. This observation and the inability to
identify more potent compounds from the parallel synthesis
library led us to suspect that the activation may be occurring
through a nonspecific mechanism. To investigate the general
effect of these compounds on enzyme activity, we investigated
the ability of the original seven hits in Figure 1 to activate
another protease, chymotrypsin. The activity of chymotrypsin
was monitored by cleavage of the N-succinyl-Ala-Ala-Pro-Phe-
p-nitroanalide (Suc-AAPF-pNA) substrate.19 Known surfactants
(CHAPS, Triton-X, Tween-20, and SDS) were also investigated
for their ability to enhance the activity of this enzyme. The
The generality of the nonspecific activation phenomenon was
further examined by assaying multiple enzymes with different
catalytic activities: ꢀ-lactamase, ꢀ-galactosidase, horse radish
peroxidase, alkaline phosphatase, trypsin, and caspase-3. Stan-
dardized assay conditions were used for most enzymes, namely,
50 mM phosphate buffer, pH 7.0, and known substrates (see
Supporting Information for assay details). CHAPS was also
included in the compound set as the known surfactant control.
As shown in Table 1, surfactant activation is not seen for every
enzyme, and some of the compounds identified herein actually
inhibit the enzymes as the compound concentrations approach
the solubility limit in the specified assay buffer (see Supporting
Information Figure S4). However, every enzyme that was
activated by CHAPS was also activated by some HTS hits,
supporting the notion of a nonspecific surfactant effect for these
Scheme 2. Parallel Synthesis of Derivatives of Compound 6

2348 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 8
Letters
tamase appear to be the most sensitive to this surfactant
activation (Table 1; see Supporting Information Figure S4 for
full activation curves for each compound and enzyme).
These results show the presence of promiscuous activators
in HTS libraries, which pose a danger as nuisance compounds
in future high-throughput enzyme activation screens. A search
of PubChem reveals that 1–3 and 5–7 have been tested in
multiple high-throughput screens, and several of these com-
pounds are members of commercial screening libraries. The
structural features of the promiscuous activators identified herein
are the same as those of surfactants: a polar or charged
headgroup and a large hydrophobic tail. However, these
structural features may be embedded within more druglike
compounds, as is the case for 6 and 7. In addition, these
compounds induce activation at concentrations relevant in high-
throughput screens (low micromolar), which is not always the
case for classical surfactants. While this activation effect is not
completely general across all enzyme classes, similar compounds
will likely appear as hits in other enzyme activator screens, and
thus, any small molecule enzyme activator should be examined
for this nonspecific effect.
The promiscuous activation may be due to some combination
of enzyme stabilization in solution, loosening of the protein
secondary structure, or alteration of adsorption of the enzymes
to surfaces that they make contact with during the course of
the experiments. To investigate the latter possibility, an experi-
ment examining the effect of preincubation time on chymo-
trypsin activity was performed. The results from this experiment
show that simple incubation of chymotrypsin in a 384-well plate
for 1 h causes the enzyme to lose significant activity (Supporting
Information Figure S6). This effect is abrogated, however, by
the addition of CHAPS or 7. Thus, it appears that CHAPS and
7 allow enzymes to retain their full activity, at least in part by
preventing their adsorption onto surfaces commonly used in
enzymatic assays.
Shoichet has recommended the inclusion of detergents in
primary HTS for enzyme inhibition to eliminate the identifica-
tion of compounds that inhibit enzymes through formation of
colloidal aggregates.8 As shown herein, when HTS for enzyme
activators is conducted, the inclusion of detergent (or BSA) in
the enzyme assay buffer should also reduce the number of “hits”
that are actually nonspecific activators. However, time is a
variable in these experiments as well. For example, we have
found that in the presence of CHAPS, 7 will not enhance the
activity of procaspase-2 when incubated for short (30 min) times.
When 7 is incubated with procaspase-2 for long times (16 h),
however, enhanced activity is observed, even in the presence
of CHAPS.
Figure 2. Activation of chymotrypsin by small molecules: (A) rep-
resentative dose response activation of chymotrypsin by 6, 7, and
CHAPS, where error bars represent standard deviations from the mean;
(B) progress curves showing activation of chymotrypsin by 7 in
phosphate assay buffer; (C) progress curves showing no activation of
chymotrypsin by 7 in phosphate assay buffer + 0.1% CHAPS.
Table 1. Activation of Enzymes by 1–7 and CHAPSa
enzymes
1
2
3
4
5
6
7
CHAPS
procaspase-2
chymotrypsin
ꢀ-lactamase
ꢀ-galactosidase
HR peroxidase
alkaline phosphatase
trypsin
+
-
-
-
-
+
-
+
+
+
-
-
-
-
-
+
+
+
+
-
-
-
+
-
+
+
+
-
-
-
-
-
+
-
+
-
-
-
+
-
+
+
+
-
-
-
+
+
+
+
+
-
-
-
-
+
+
+
+
-
-
-
+
+
Described herein is the unexpected discovery of general
enzyme activators in HTS assays. Also described are facile
methods for identification of such promiscuous activators,
namely, counter-screening hits against a readily available pro-
tease (such as chymotrypsin) and screening for elimination of
compound activation by addition of a detergent (such as
CHAPS) to the assay buffer. These experiments have allowed
us to authenticate that the compounds in Figure 1 are promiscu-
ous enzyme activators. Importantly, similar experiments have
confirmed that one reported enzyme activator, PAC-1, does not
act through this nonspecific mechanism (see Supporting Infor-
mation Figure S7). Analogous to promiscuous inhibitors,7 a
computational filter set could be developed to flag possible
promiscuous activators among HTS collections. Avoiding
promiscuous activators by these simple screens or filters will
save time and effort in future HTS endeavors.
caspase-3
a Minus sign denotes no effect or inhibition, and plus sign denotes
activation. See Supporting Information for full activation curves for each
compound and enzyme.
compounds. This data are consistent with a report by Shoichet
and co-workers, whereby multiple compounds in a large high-
throughput screen (70 000) showed enzyme activation that was
detergent-sensitive.16 Importantly, addition of BSA to the assay
buffer also prevents chymotrypsin activation by CHAPS and 7
(see Supporting Information Figure S5). Of the enzymes
surveyed thus far, trypsin, caspase-3, chymotrypsin, and ꢀ-lac-

Letters
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 8 2349
(11) Guertin, K. R.; Grimsby, J. Small molecule glucokinase activators as
glucose lowering agents: a new paradigm for diabetes therapy. Curr.
Med. Chem. 2006, 13, 1839–1843.
(12) Evgenov, O. V.; Pacher, P.; Schmidt, P. M.; Hasko, G.; Schmidt, H. H.;
et al. NO-independent stimulators and activators of soluble guanylate
cyclase: discovery and therapeutic potential. Nat. ReV. Drug DiscoVery
2006, 5, 755–768.
(13) Thakur, C. S.; Jha, B. K.; Dong, B.; Das Gupta, J.; Silverman, K. M.
Small-molecule activators of RNase L with broad-spectrum antiviral
activity. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 9585–9590.
(14) Milne, J. C.; Lambert, P. D.; Schenk, S.; Carney, D. P.; Smith, J. J.;
et al. Small molecule activators of SIRT1 as therapeutics for the
treatment of type 2 diabetes. Nature 2007, 450, 712–716.
(15) Putt, K. S.; Chen, G. W.; Pearson, J. M.; Sandhorst, J. S.; Hoagland,
M. S., et al. Small-molecule activation of procaspase-3 to caspase-3
as a personalized anticancer strategy. Nat. Chem. Biol. 2006, 2, 543–
550.
(16) Feng, B. Y.; Simeonov, A.; Jadhav, A.; Babaoglu, K.; Inglese, J., et
al. A high-throughput screen for aggregation-based inhibition in a large
compound library. J. Med. Chem. 2007, 50, 2385–2390.
(17) Hergenrother, P. J. Obtaining and screening compound collections: a
user’s guide and a call to chemists. Curr. Opin. Chem. Biol. 2006,
10, 213–218.
(18) Talanian, R. V.; Quinlan, C.; Trautz, S.; Hackett, M. C.; Mankovich,
J. A., et al. Substrate specificities of caspase family proteases. J. Biol.
Chem. 1997, 272, 9677–9682.
(19) DelMar, E. G.; Largman, C.; Brodrick, J. W.; Geokas, M. C. A
sensitive new substrate for chymotrypsin. Anal. Biochem. 1979, 99,
316–320.
(20) D’Auria, S.; Di Cesare, N.; Gryczynski, I.; Rossi, M.; Lakowicz, J. R.
On the effect of sodium dodecyl sulfate on the structure of beta-
galactosidase from Escherichia coli. A fluorescence study. J. Biochem.
(Tokyo) 2001, 130, 13–18.
(21) Gall, A. L.; Ruff, M.; Moras, D. The dual role of CHAPS in the
crystallization of stromelysin-3 catalytic domain. Acta Crystallogr.,
Sect. D: Biol. Crystallogr. 2003, 59, 603–606.
Acknowledgment. We thank the Department of Defense
(Grant W81XWH-06-1-0608) and the University of Illinois for
funding this work.
Supporting Information Available: Experimental details, in-
cluding biochemical assay conditions, compound synthesis, and
compound characterization. This material is available free of charge
via the Internet at http://pubs.acs.org.
References
(1) Pereira, D. A.; Williams, J. A. Origin and evolution of high throughput
screening. Br. J. Pharmacol. 2007, 152, 53–61.
(2) Carnero, A. High throughput screening in drug discovery. Clin. Transl.
Oncol. 2006, 8, 482–490.
(3) Posner, B. A. High-throughput screening-driven lead discovery:
meeting the challenges of finding new therapeutics. Curr. Opin. Drug
DiscoVery DeV. 2005, 8, 487–494.
(4) Shoichet, B. K. Screening in a spirit haunted world. Drug DiscoVery
Today 2006, 11, 607–615.
(5) McGovern, S. L.; Helfand, B. T.; Feng, B.; Shoichet, B. K. A specific
mechanism of nonspecific inhibition. J. Med. Chem. 2003, 46, 4265–
4272.
(6) McGovern, S. L.; Caselli, E.; Grigorieff, N.; Shoichet, B. K. A common
mechanism underlying promiscuous inhibitors from virtual and high-
throughput screening. J. Med. Chem. 2002, 45, 1712–1722.
(7) Seidler, J.; McGovern, S. L.; Doman, T. N.; Shoichet, B. K.
Identification and prediction of promiscuous aggregating inhibitors
among known drugs. J. Med. Chem. 2003, 46, 4477–4486.
(8) Feng, B. Y.; Shelat, A.; Doman, T. N.; Guy, R. K.; Shoichet, B. K.
High-throughput assays for promiscuous inhibitors. Nat. Chem. Biol.
2005, 1, 146–148.
(9) Feng, B. Y.; Shoichet, B. K. A detergent-based assay for the detection
of promiscuous inhibitors. Nat. Protoc. 2006, 1, 550–553.
(10) Grimsby, J.; Sarabu, R.; Corbett, W. L.; Haynes, N. E.; Bizzarro, F. T.;
et al. Allosteric activators of glucokinase: potential role in diabetes
therapy. Science 2003, 301, 370–373.
JM701583B