close

Вход

Забыли?

вход по аккаунту

?

Cellular Profiling of Small-Molecule Bioactivities an Alternative Tool for Chemical Biology.

код для вставкиСкачать
Highlights
DOI: 10.1002/anie.200500721
Drug Research
Cellular Profiling of Small-Molecule Bioactivities: an
Alternative Tool for Chemical Biology**
Thorsten Berg*
Keywords:
fluorescent probes · inhibitors · phenotypic screening ·
proteins · signal transduction
O
ne of the main goals of chemical
biology is to identify cell-permeable
organic molecules that interfere with
protein function to provide the scientific
community with novel research tools for
the investigation of biological questions.[1] There are currently two major
approaches for the primary screening of
chemical libraries to identify such bioactive organic molecules. In the first
approach, inhibitors or activators of
purified proteins or reconstituted, narrowly defined signaling pathways are
sought in vitro. In the second approach,
cells or organisms are treated with small
molecules; those compounds which
cause a desired phenotype are selected.
The success of research projects that
begin with such a phenotypic screen is
dependent on the subsequent identification of the molecular target(s) of the
identified compounds that elicit the
phenotype, as a meaningful analysis of
the biological profiles of the hit compound can only be performed when the
cellular target is known. Most of the
current methods used to pinpoint the
molecular target(s) of compounds identified in phenotypic screens can be
[*] Dr. T. Berg
Max Planck Institute of Biochemistry
Department of Molecular Biology
Am Klopferspitz 18, 82152 Martinsried
(Germany)
Fax: (+ 49) 89-8578-2454
E-mail: berg@biochem.mpg.de
[**] I extend my thanks to Professor Dr. Axel
Ullrich for his support of my research. I
also thank Angela Hollis for critical reading of the manuscript.
Supporting information for this article
(complete citations) is available on the
WWW under http://www.angewandte.org
or from the author.
5008
classified into two subgroups: 1) affinity-based biochemical methods based on
small-molecule binding to target proteins and the subsequent identification
of the protein bound and 2) genetic
approaches that monitor cellular
changes in response to treatment with
the compound. As discussed at the end
of this Highlight, neither approach is
generally applicable; alternative approaches aimed at the identification of
molecular targets of bioactive compounds from phenotypic screens would
therefore be highly welcome.
The research groups of Altschuler,
Wu, and Mitchison recently reported a
method which can be used to profile the
effects of interfering agents on cells by
the analysis of multiple parameters
related to the expression, post-translational modification, and localization of
preselected proteins.[2] This approach
registers the biological effects of a compound with an unknown mode of action
on a defined set of biological markers
and then compares them with the effects
observed with a set of reference compounds with known activities.[3] If the
biological effects of a compound with
unknown target(s) resemble one of
more of the biological profiles of the
reference compounds, a potential mechanism of action for the new inhibitor can
be proposed. This could provide valuable information for decreasing the number of potential targets for subsequent
investigation.
To demonstrate the principle of this
method, cells grown in microtiter plates
were treated with 100 different chemical
substances at a range of concentrations.
90 of the 100 compounds had a known
mechanism of action, and thereby
served as references for the classifica-
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
tion of the remaining 10 test compounds.
Six of these 10 were simply blinded
references, three had an unknown biological mechanism, and one compound
was known to have multiple effects.
Treated cells were then stained with
fluorescently labeled antibodies raised
against a small number of proteins
involved in a range of cell signaling
pathways. In this way, a very broad
overview of the effects of compounds on
the cells was obtained. The wells were
photographed with an automated microscope, and special software was employed to convert the large amount of
data obtained into interpretable information. Changes in various parameters
related to the fluorescence of the antibodies in response to various compound
concentrations were analyzed; these
changes were subsequently expressed
as color codes, as shown in the example
case of the topoisomerase inhibitor
camptothecin (Figure 1 a).
For a meaningful comparison of the
biological effects of two compounds, it is
necessary to depict the compound profiles in a manner that is independent of
the absolute concentration at which an
effect is initially observed. Otherwise,
two compounds with identical biological
mechanisms yet different potencies
would be erroneously regarded as having different biological effects, as the
observation of any effect would occur at
different concentration thresholds.
Therefore, the authors devised a method
to analyze the similarity of the compound effects in a concentration-independent manner and used this information for the unsupervised clustering of
the compounds depicted in Figure 1 b.
The names of the compounds and their
mode of action as described in the
Angew. Chem. Int. Ed. 2005, 44, 5008 – 5011
Angewandte
Chemie
Figure 1. a) Cellular responses to different concentrations of the topoisomerase inhibitor camptothecin. The response of each parameter to a
given concentration of the drug is expressed as a colored rectangle. The vertical axis depicts 93 biological parameters. b) Clustering analysis of
drug effects; excerpted from Ref. [2] (Copyright 2004, AAAS).
literature are given in the left, and on
the right, the similarities between compound pairs are given: high similarities
are indicated by dark squares; lower
degrees of similarity are represented
with lighter squares. (The list of compounds as given on the left side of
Figure 1 b is omitted from top of the
right part of the figure.) Naturally, the
biological profiles of each compound
match perfectly in self-comparison,
which explains the diagonal line of black
squares.
Successful clustering analysis lists
compounds with similar modes of action
adjacent to each other, as observed by
adjacent black squares in the left part of
Figure 1 b, and leads to an accumulation
of dark squares around the diagonal line
in the right portion. Good clustering is
indeed observed for histone deacetylase
inhibitors, topoisomerase inhibitors, and
microtubule poisons. However, for other compound classes such as protein
synthesis inhibitors, clustering was not
ideal.
Another criterion for success was to
determine if the 10 test compounds
Angew. Chem. Int. Ed. 2005, 44, 5008 – 5011
mentioned above could be properly
classified. Five of the six blinded alternate titrations of known drugs showed
usable activity profiles and clustered
well with their unmasked counterparts,
or at least with drugs of similar mechanism, as indicated by the blue squares in
the left part of Figure 1 b. The single
compound with many cellular functions
clustered correctly with others known to
have one of its functions.[4] Of the three
remaining compounds with uncharacterized mechanisms, only one could be
grouped with a certain inhibitor class (it
would have been helpful to see the
independent experimental data supporting the correct classification of the
compound).
Whereas the success rate of the
current study is not optimal, it is the
principle of the approach which deserves credit. About 60 of the 100
compounds showed a meaningful response in the experiment, and it is
possible that the other 40 compounds
either acted on proteins or pathways not
represented by the relatively small set of
antibodies used in this study, or that
their biological targets were not expressed in the chosen cell line. A larger
set of reference compounds would also
improve the chances of pinpointing the
biological functions of novel compounds
identified in phenotypic screens. Therefore, the power of this method is likely
to be improved with the use of larger
sets of antibodies and reference compounds, and by extending the experiments to additional cell lines. The antibodies and reference compounds could
be adapted to the specific needs of the
researcher. A distinctly positive aspect
of the method is that it permits the
analysis of compound concentrations at
which various cellular parameters
change. Such information could be useful for the identification of an optimal
dose below the threshold of undesirable
(toxic) effects. One severe limitation of
the approach, however, is that the
hardware and software necessary for
these experiments are not currently in
place at most academic settings; acquisition of these resources may require an
industrial-level budget. The expansion
of the set of reference compounds, anti-
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5009
Highlights
body probes, and cell lines might require
even more investment in screening infrastructure and computing power.
The cellular targets of most chemical
genetic screens have been identified
through biochemical affinity-based
techniques.[5–8] Such approaches require
the attachment of a tag that facilitates
immobilization of the bioactive molecule. The resulting hybrid molecule is
then attached to a solid support by the
tag and then incubated with cell lysates.
Proteins
successfully
immobilized
through specific interactions with the
compound are subsequently analyzed by
protein microsequencing, MS-based
methods, or immunoblotting. Examples
of successful applications of this approach include the discovery of the
molecular targets of the natural product
FK506[9] and of several kinase inhibitors.[10–13] One general concern is that
compound-binding proteins identified
by this method may not be the physiologically relevant binding partners in
vivo. In this respect, irreversible inhibitors with reactive functional groups are
advantageous, in that intact cells, instead
of cell lysates, can be treated with the
tagged inhibitor. Therefore, it can be
assumed that only the physiologically
relevant proteins will bind to the inhibitor. Examples of the successful discovery of cellular targets of irreversible
inhibitors include the natural products
fumagillin,[14] parthenolide,[15] epoxomicin,[16] and lactacystin.[17]
In a variation of the approach referred to as “photoaffinity labeling”, a
reversible inhibitor is tagged with a
photolabile group, which inserts itself
into chemical bonds in close proximity
upon irradiation, thereby forming covalent bonds with its protein-binding partners.[18, 19] Despite the success of affinitybased purification approaches, they suffer from the following limitations: 1) the
approach requires a tight interaction
between the small molecule and its
protein target, 2) the protein target must
be abundant in the cell, and 3) the
addition of a tag without significant loss
of biological activity requires precise
knowledge of structure–activity relationships, and there is no guarantee that
a tagged inhibitor with sufficient biological potency can be identified and synthesized. Several modified approaches
have been suggested to overcome these
5010
www.angewandte.org
hurdles, for example: 1) the alignment
of proteins on solid supports (protein
arrays)[20] or their display on the surface
of phages[21] to artificially increase the
local concentration of potential target
proteins, 2) the yeast three-hybrid system,[22] and 3) the screening of smallmolecule libraries pre-equipped with a
linker that allows rapid immobilization
without the need for structural modification of the inhibitor.[23] Future research will assess the power of these
modified approaches.
Most genetic approaches for target
identification have been developed in
yeast, and make use of the availability of
complete sets of viable homozygous and
heterozygous gene-deletion mutants.
The most obvious approach is to compare the effect of an interfering agent on
the gene expression profile of a wildtype yeast strain with that of yeast
strains in which single genes have been
deleted.[24–27] Another experimental approach exploits the phenomenon that
heterozygous deletion mutants (yeast
strains in which only one of the two
alleles of the respective gene is deleted)
are more susceptible to inhibition by
compounds which act on the gene product encoded by the remaining undeleted
gene.[28–30] Genetic approaches with
yeast as a model system may be an
interesting alternative for cases in which
the affinity purification approach cannot
be applied; perhaps the main drawback
is that a homologue of the gene product
targeted in human cells (30 000–40 000
genes) may not be present in yeast
( 6000 genes).
The cellular profiling of small-molecule bioactivities highlighted herein represents a significant advantage over
most methods in current use, as it does
not require a tag and is not restricted to
yeast. However, it appears unlikely that
this method will permit the definition of
the target of an interfering agent down
to a single protein, as the information
provided is of a more general nature.
Therefore, this novel method may represent a useful supplement to the toolbox of chemical geneticists in cases in
which 1) the affinity purification approach fails, 2) the aim is to decrease
the mode of action of a compound to a
pathway or a group of proteins, or
3) cellular systems other than yeast are
necessary. It will be interesting to wit-
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ness the results obtained with this method in its future applications.
Published online: June 23, 2005
[1] T. U. Mayer, Trends Cell Biol. 2003, 13,
270 – 277.
[2] Z. E. Perlman, M. D. Slack, Y. Feng, T. J.
Mitchison, L. F. Wu, S. J. Altschuler,
Science 2004, 306, 1194 – 1198.
[3] T. J. Mitchison, ChemBioChem 2005, 6,
33 – 39.
[4] M. D. Vera, M. M. Joullie, Med. Res.
Rev. 2002, 22, 102 – 145.
[5] L. Burdine, T. Kodadek, Chem. Biol.
2004, 11, 593 – 597.
[6] S. Wang, T. B. Sim, Y.-S. Kim, Y.-T.
Chang, Curr. Opin. Chem. Biol. 2004, 8,
371 – 377.
[7] G. P. Tochtrop, R. W. King, Comb.
Chem. High Throughput Screening
2004, 7, 677 – 688.
[8] S. Ding, T. Y. H. Wu, A. Brinker, E. C.
Peters, W. Hur, N. S. Gray, P. G. Schultz,
Proc. Natl. Acad. Sci. USA 2003, 100,
7632 – 7637.
[9] M. W. Harding, A. Galat, D. E. Uehling,
S. L. Schreiber, Nature 1989, 341, 758 –
760.
[10] C. Kung, K. M. Shokat, ChemBioChem
2005, 6, 523 – 526.
[11] K. Godl et al., Proc. Natl. Acad. Sci.
USA 2003, 100, 15 434–15 439.
[12] G. R. Rosania, J. Merlie, Jr., N. Gray, Y.T. Chang, P. G. Schultz, R. Heald, Proc.
Natl. Acad. Sci. USA 1999, 96, 4797 –
4802.
[13] M. Knockaert, P. Lenormand, N. Gray, P.
Schultz, J. PouyssIgur, L. Meijer, Oncogene 2002, 21, 6413 – 6424.
[14] N. Sin, L. Meng, M. Q. W. Wang, J. J.
Wen, W. G. Bornmann, C. M. Crews,
Proc. Natl. Acad. Sci. USA 1997, 94,
6099 – 6103.
[15] B. H. B. Kwok, B. Koh, M. I. Ndubuisi,
M. Elofsson, C. M. Crews, Chem. Biol.
2001, 8, 759 – 766.
[16] L. Meng, R. Mohan, B. H. B. Kwok, M.
Elofsson, N. Sin, C. M. Crews, Proc.
Natl. Acad. Sci. USA 1999, 96, 10 403 –
10 408.
[17] G. Fenteany, R. F. Standaert, W. S. Lane,
S. Choi, E. J. Corey, S. L. Schreiber,
Science 1995, 268, 726 – 731.
[18] J. R. Colca, G. G. Harrigan, Comb.
Chem. High Throughput Screening
2004, 7, 699 – 704.
[19] G. DormKn, G. D. Prestwich, Trends
Biotechnol. 2000, 18, 64 – 77.
[20] J. Huang, H. Zhu, S. J. Haggarty, D. R.
Spring, H. Hwang, F. Jin, M. Snyder,
S. L. Schreiber, Proc. Natl. Acad. Sci.
USA 2004, 101, 16 594 – 16 599.
[21] P. P. Sche, K. M. McKenzie, J. D. White,
D. J. Austin, Chem. Biol. 1999, 6, 707 –
716.
Angew. Chem. Int. Ed. 2005, 44, 5008 – 5011
Angewandte
Chemie
[22] E. J. Licitra, J. O. Liu, Proc. Natl. Acad.
Sci. USA 1996, 93, 12 817 – 12 821.
[23] S. M. Khersonsky et al., J. Am. Chem.
Soc. 2003, 125, 11 804–11 805.
[24] M. J. Marton et al. , Nat. Med. 1998, 4,
1293 – 1301.
Angew. Chem. Int. Ed. 2005, 44, 5008 – 5011
[25] T. R. Hughes et al., Cell 2000, 102, 109 –
126.
[26] T.-F. Chan, J. Carvalho, L. Riles, X. F. S.
Zheng, Proc. Natl. Acad. Sci. USA 2000,
97, 13 227 – 13 232.
[27] A. B. Parsons et al., Nat. Biotechnol.
2004, 22, 62 – 69.
[28] P. Y. Lum et al., Cell 2004, 116, 121 –
137.
[29] K. Baetz et al., Proc. Natl. Acad. Sci.
USA 2004, 101, 4525 – 4530.
[30] G. Giaever et al., Proc. Natl. Acad. Sci.
USA 2004, 101, 793 – 798.
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5011
Документ
Категория
Без категории
Просмотров
0
Размер файла
435 Кб
Теги
chemical, bioactivities, profiling, molecules, small, tool, biologya, alternative, cellular
1/--страниц
Пожаловаться на содержимое документа