Cell biology. Going global on ubiquitin

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