PERSPECTIVES
the regenerative effect of PTEN deficiency.
Genetic ablation of a negative mTOR regulator, tuberous sclerosis complex 1 (TSC1), partially but not completely mimicked the effects
of PTEN deficiency on axon regeneration,
indicating that other PTEN-regulated pathways such as glycogen synthesis kinase–3
(GSK-3) could be involved in controlling
axon growth.
It is uncertain whether the intrinsic regenerative mechanism observed by Park et al.
is vigorous enough to overcome a hostile
extrinsic environment. For example, an optic
nerve crush produces much less inflammation and glial scarring than a contusion
injury that damages the spinal cord. It will be
important to determine the extent to which
the findings of Park et al. generalize to other
neuronal populations such as corticospinal
axons and to understand why mTOR activity
is reduced during development and after
axon injury.
Will the studies by Atwal et al. and Park
et al. lead to advances in treating human
spinal cord injuries? Prior work with
myelin-inhibitory proteins has initiated
exploratory clinical trials. The identification of LILRB2 as a human receptor for
myelin-inhibitory proteins should stimulate new thinking in this area. However,
work with primates, which more adequately model human injuries than rodents, is
just beginning (19). It is unclear whether
therapeutic approaches centered around
PTEN inhibition could be developed, and
whether PTEN inhibitors can mimic the
positive effects of deleting the PTEN gene
on axon regeneration. Nevertheless, the
idea of enhancing protein synthesis to promote long-distance axon growth after
injury is an appealing possibility.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
J. K. Atwal et al., Science 322,967 (2008).
K. K. S. Park et al., Science 322, 963 (2008).
P. M. Richardson et al., Nature 284, 264 (1980).
P. Caroni, M. E. Schwab, Neuron 1, 85 (1988).
J. Silver, J. H. Miller, Nat. Rev. Neurosci. 5, 146
(2004).
J. L. Goldberg et al., Science 296, 1860 (2002).
G. Yiu, Z. He, Nat. Rev. Neurosci. 7, 617 (2006).
A. E. Fournier et al., Nature 409, 341 (2001).
K. C. Wang et al., Nature 420, 74 (2002).
S. T. Wong et al., Nat. Neurosci. 5, 1302 (2002).
S. Mi et al., Nat. Neurosci. 7, 221 (2004).
Z. Shao et al., Neuron 45, 353 (2005).
J. B. Park et al., Neuron 45, 345 (2005).
B. Zheng et al., Trends Neurosci. 29, 640 (2006).
J. Syken et al., Science 313, 1795 (2006).
A. W. McGee et al., Science 309, 2222 (2005).
D. Fischer et al., J. Neurosci. 24, 8726 (2004).
Y. Yin et al., Nat. Neurosci. 9, 843 (2006).
S. Rossignol et al., J. Neurosci. 27, 11782 (2007).
10.1126/science.1166152
CELL BIOLOGY
A new technique that profiles protein stability
provides a powerful platform in which highthroughput screening can be performed in real
time with single-cell resolution.
Going Global on Ubiquitin
Caroline Grabbe1 and Ivan Dikic1,2,3
ur vision of protein control has for
many years been viewed from a transcriptional or activity-based perspective. More recently, protein stability and regulated degradation have emerged as equally
important issues to address. On pages 918 and
923 of this issue, Elledge and colleagues
describe a new technology to analyze global
protein stability (GPS). The studies introduce
the new approach (1) and illustrate how it can
be applied to identify substrates of a specific
enzyme (ubiquitin ligase) (2).
In addition to protein degradation that
occurs in the lysosomal compartment of cells,
degradation by a cellular shredding machine
known as the proteasome is another major
route to eliminate proteins. Targeting to the
proteasome is preceded by the addition of
ubiquitin chains to the selected substrates, an
event catalyzed by the sequential action of
activating, conjugating, and ligating enzymes
(E1, E2, and E3, respectively). The specificity of substrate recognition is mediated
mainly by divergent use of the estimated
∼617 ubiquitin ligases encoded by the human
O
1Institute
of Biochemistry II and Cluster of Excellence
Macromolecular Complexes, Goethe University, Frankfurt
am Main, Theodor-Stern-Kai 7, D-60590 Frankfurt (Main),
Germany. 2Mediterranean Institute for Life Sciences,
21000 Split, Croatia. 3Department of Immunology, School
of Medicine, University of Split, Soltanska 2, 21000 Split,
Croatia. E-mail:
[email protected]
872
genome (3). E3 ligases fall into different
classes, based on structural composition and
mechanism of action. In some cases, a group
of proteins comes together to form multisubunit ubiquitin ligases, as is the case for the
SCF (Skp1-cullin-F-box) complex where
Skp1, Cul1, Rbx1, and an F-box protein form
the core of the ligase (4).
Despite extensive efforts to map substrate targeting by individual ligases, the
methods used have been laborious and the
results far from complete. Most studies have
used direct substrate-ligase interaction as a
basis for substrate identification and have
thus been biased toward strongly interacting
targets (5). The GPS approach offers a new
mode to navigate the ubiquitin-proteasome
system and identify substrates for a given E3
ligase or, in general, to investigate how a
chemical or physical stimulus affects the
stability of a given protein. The system evaluates protein abundance at a global level in
living cells, with accuracy comparable to
conventional time-consuming experiments
that analyze only a few proteins at a time.
Eight thousand distinct complementary
DNAs were used to generate a library of
cultured human cells in which each cell
expresses a common stable red fluorescent
protein (DsRed) together with a variable
fusion protein composed of enhanced green
fluorescent protein (EGFP) and a unique
7 NOVEMBER 2008
VOL 322
SCIENCE
Published by AAAS
open reading frame (ORF), produced from
the same transcript.
The turnover of the EGFP-ORF fusion
proteins can thus be monitored by flow
cytometry (which counts, examines, and sorts
whole cells) as a ratio of red to green fluorescence in each cell (see the figure). Cells in
which the fluorescence ratio changes in
response to a gene perturbation or stimulus
can be sorted by the degree of change, and the
identity of the ORFs they express can be easily identified by a polymerase chain reaction–based microarray approach. During the
development of GPS profiling, a comparative
analysis of all ORFs tested was used to assign
each ORF a protein stability index value that
roughly categorizes each corresponding protein as having a short, medium, long, or extralong life span. An impressive power of the
GPS method is the capacity of single-cell
resolution, which is in contrast to other established methods that frequently generate population-averaged readouts. In addition, measurements can take place in live cells, in real
time, and can be integrated with systems for
automation to enable high-throughput studies.
The GPS approach is a major advance in
the quest to gain a comprehensive understanding of protein turnover in cells and will be a
valuable complement to the biophysical methods that have emerged in the past 5 years to
analyze substrate ubiquitination (6). Large-
www.sciencemag.org
PERSPECTIVES
Conditions
”en route“
Start
Finish
(FACS, Microarray)
Potential destinations
(what can be discovered)
Ribosome
mRNA 5' Cap
Protein
DsRed
IRES
EGFP X
DsRed
AAAAA
•Changes in gene
expression
Determine the ratio of red/green
fluorescence
•Compounds, mutations,
infections, and stress
responses that affect
protein stability
•Changes in external
stimulation
EGFP X
•E3 ligase-substrate pairs
Decreased protein stability
•Ubiquitination sites
and linkages
Red fluorescent protein
Increased protein stability
Green fluorescent protein fused to protein X
Navigating the world of ubiquitin and protein degradation. In GPS profiling,
the ratio between a constant factor (DsRed) and a variable factor (EGFP-X) is
measured by a combinatorial approach of flow cytometry (FACS analysis) and
microarray technology. This serves as a readout of protein abundance and stabil-
scale analysis of ubiquitinated proteins has
relied mainly on a combination of affinity
purification and mass spectrometry analysis
of proteins in yeast (7), human cell lines (8),
and transgenic mice (9). To more specifically
analyze ubiquitinated targets downstream of
individual E3 ligases, Ota et al. used SILAC
(stable isotope labeling with amino acids in
cell culture) to quantify the overall change in
protein ubiquitination after alterating E3 ligase activity (10). Substrates of the E3 ligase
Rsp5 were recently identified in a highthroughput assay in which all proteins expressed in yeast were spotted onto a nitrocellulose chip and directly tested for ubiquitination by Rsp5 in vitro (11).
The GPS technology moves the field
closer toward global in vivo mapping of ligase-substrate pairs. Illustrating the feasibility
of the system, Yen and Elledge used GPS profiling to identify substrates of the SCF ubiquitin ligase complex (2). Broadly outlined, in the
background of the described GPS library,
SCF function was abrogated by the expression
of a dominant negative Cul1, whereupon the
perturbed cells were analyzed by GPS profiling. An impressive number of 359 targets
were postulated as putative SCF substrates,
among which 66 were tested and 31 verified.
In a majority of cases, verification of a specific protein was accomplished by analyzing
individual samples by flow cytometry [fluorescence-activated cell sorting (FACS) analysis] as well as by biochemical detection in cell
extracts with antibodies (immunoblotting). In
the future, combining a GPS-based ORF
library with methods that perturb gene expression on a global scale, such as genome-wide
ity. By challenging the system with internal modifications or external stimuli, the
responses can be monitored on a global scale. Many questions concerning the
process of protein degradation, including E3 ligase specificity, routes to destruction, and importance of ubiquitin linkage, can be addressed.
RNA interference, leaves few limitations to
the amount of knowledge that may be
acquired with this innovative method.
A feature that is both a strength and a limitation is that the GPS technique does not monitor
E3 substrates directly but rather the outcome of
E3 activity. Thus, there is no discrimination
between the direct and indirect effects of an E3
ligase. However, this could be resolved by following the kinetics of protein degradation. GPS
profiling is also biased toward identifying ubiquitinated substrates that are destined for either
degradation or stabilization. This excludes proteins that are functionally affected by the modification with regard to their enzymatic activity,
localization in the cell, and ability to form complexes with other cellular constituents. Nevertheless, together with conventional proteomic
approaches, the GPS system will provide a
powerful means to distinguish the consequences of different types of ubiquitination, sorting
the proteolytic events from those of regulatory
nature (12).
In addition to assigning ligase-substrate
pairs, GPS profiling has the potential to elucidate the essence of differential ubiquitin chain
linkages, in particular how linkage affects the
efficiency of proteasomal targeting. It may
also help to identify degradation signals
(degrons) encoded by amino acid sequences,
to associate specific lysine residues that are
modified by ubiquitin to functionality, and to
explain why some proteins are directly routed
to the proteasome whereas others require
shuttle factors to ensure proper targeting.
There is also the potential to transfer GPS
profiling into model organisms such as
Drosophila melanogaster and Caenorhabditis
www.sciencemag.org
SCIENCE
VOL 322
Published by AAAS
elegans, creating ORF libraries that will enable substrate screening in vivo.
Given the importance of the ubiquitinproteasome system for cellular functions, dysfunctions of the involved players clearly have
the capacity to cause disease. Indeed, defective ubiquitination and protein degradation is
implicated in the etiology of cancer and neurodegenerative disorders, among others (13,
14). In this context, the GPS method could be
used to screen for compounds that counteract
such deficiencies. Another interesting aspect
would be to investigate how protein degradation is altered when cells are exposed to infectious agents or various stress situations.
Overall, the GPS approach will likely become
an important component of the integrated
approaches needed to systematically map the
mechanisms of regulated protein degradation.
References
1. H.-C. S. Yen, Q. Xu, D. M. Chou, Z. Zhao, S. J. Elledge,
Science 322, 918 (2008).
2. H. -C. S. Yen, S. J. Elledge, Science 322, 923 (2008).
3. W. Li et al., PLoS ONE 3, e1487 (2008).
4. T. Ravid, M. Hochstrasser, Nat. Rev. Mol. Cell Biol. 9, 679
(2008).
5. A. Peschiaroli et al., Mol. Cell 23, 319 (2006).
6. J. Peng, BMB Rep. 41, 177 (2008).
7. J. Peng et al., Nat. Biotechnol. 21, 921 (2003).
8. M. Matsumoto et al., Proteomics 5, 4145 (2005).
9. H. B. Jeon et al., Biochem. Biophys. Res. Commun. 357,
731 (2007).
10. K. Ota, K. Kito, S. Okada, T. Ito, Genes Cells 13, 1075
(2008).
11. R. Gupta et al., Mol. Syst. Biol. 3, 116 (2007).
12. F. Ikeda, I. Dikic, EMBO Rep. 9, 536 (2008).
13. G. Nalepa, M. Rolfe, J. W. Harper, Nat. Rev. Drug Discov.
5, 596 (2006).
14. D. Hoeller, C. M. Hecker, I. Dikic, Nat. Rev. Cancer 6, 776
(2006).
7 NOVEMBER 2008
10.1126/science.1166845
873