Development of a method for screening short-lived proteins using green fluorescent protein
© Jiang et al.; licensee BioMed Central Ltd. 2004
Received: 5 April 2004
Accepted: 6 August 2004
Published: 28 September 2004
We have developed a screening technology for the identification of short-lived proteins. A green fluorescent protein (GFP)-fusion cDNA library was generated for monitoring degradation kinetics. Cells expressing a subset of the GFP-cDNA expression library were screened to recover those in which the fluorescence signal diminished rapidly when protein synthesis was inhibited. Thirty clones that met the screening criteria were characterized individually. Twenty-three (73%) proved to have a half-life of 4 hours or less.
Cellular proteins differ widely in their lability, ranging from those that are completely stable to those with half-lives measured in minutes. Proteins with a short half-life are among the most critical to the cell. Regulated degradation of specific proteins contributes to the control of signal transduction pathways, cell-cycle control, transcription, apoptosis, antigen processing, biological clock control, differentiation and surface receptor desensitization [1, 2]. Rapid turnover makes it possible for the cellular level of a protein to change promptly when synthesis is increased or reduced . Furthermore, degradation rate is itself subject to regulation. For instance, inflammatory stimuli cause the rapid degradation of IκBα, the inhibitor of NFκB, resulting in the activation of that transcription factor [4–6].
Analysis of labile proteins has been time-consuming and labor-intensive. The most definitive form of analysis requires pulse-chase labeling cells and immunoprecipitation extracts. In vitro assay of degradation is simpler than in vivo analysis, but an in vitro assay system may not fully mimic the degradation of proteins in the cells. Genome-wide functional screening and systemic characterization of cellular short-lived proteins has received little attention . GFP, the green fluorescent protein from the jellyfish Aequorea victoria, has been widely used to monitor gene expression and protein localization . Recently, we demonstrated that fusion of GFP to the degradation domain of ornithine decarboxylase , a labile protein, can destabilize GFP  and that the degradation of an IκB-GFP fusion protein can be monitored by GFP fluorescence . These studies demonstrate that introducing GFP as a fusion within the context of a rapidly degraded protein does not alter the degradation properties of the parent molecule, and that the GFP moiety of the fusion protein is degraded along with the rest of the protein. GFP fluorescence, which provides a sensitive, rapid, precise and non-destructive assay of protein abundance, can therefore be used to monitor protein degradation . Furthermore, fluorescence associated with single cells can be analyzed using fluorescence-activated cell sorting (FACS), a technology easily adapted to high-throughput screening .
We developed a GFP-based, genome-wide screening method for short-lived proteins. We made a GFP fusion expression library of human cDNAs and introduced the library into mammalian cells. Transfected cells were FACS-fractionated into subpopulations of uniform fluorescence. Individual subpopulations were treated with cycloheximide (CHX) to inhibit protein synthesis and re-sorted after 2 hours of treatment. Sorting was gated to recover cells with a fluorescent signal that was diminished compared to the population mode. Repeated application of this process resulted in a high yield of clones that encode labile fusion proteins.
We selected R4 for further screening in this study. We collected 106 cells from the shifted population, the left shoulder of the population observed in the CHX-treated but not in the untreated R4 cells (Figure 3). Plasmid DNAs were recovered from the sorted cells and were propagated in Escherichia coli, resulting in a total of 400 clones. The individual clones were stored in 15% glycerol LB medium in a 96-well format.
To perform second-round selection, we grouped the 400 clones into 12 pools, each composed of approximately 33 clones. The individual pools of clones were cultured and used for plasmid preparation. We transfected these 12 groups of plasmid DNA into 293T cells and again subjected them to FACS analysis and gating as before. The EGFP-C1 vector was used as a control. Because enhanced green fluorescent protein (EGFP) is a stable protein, its fluorescence intensity would not be changed by treatment with CHX. We found that eight of the 12 groups showed a decrease of the fluorescence intensity peak by 30-50% (compared to untreated cells) after 2 hours of CHX treatment. In four out of 12 groups, no change in fluorescence intensity was detected.
The estimated half-lives of 22 labile proteins
Estimated half-life (h)
Similar to SH3-containing protein
A-kinase anchoring protein
Splicing factor SRp30c
Adaptor-related protein complex
ATP synthase, H+ transporting
H19, imprinted maternally expressed
Guanine-nucleotide-binding regulatory protein
Heat shock 70 kDa protein 1A
Cervical cancer 1 proto-oncogene protein p40
Gag-Pro-Pol precursor protein gene.
Insulin-like growth factors II
Serine (or cysteine) proteinase inhibitor
Thyroid hormone receptor interactors 13
The abundance of a given cellular protein is determined by the balance between its rate of synthesis and degradation. The two are of equal importance in their effect on the steady-state level. Furthermore, degradation determines the rate at which a new steady state is reached when protein synthesis changes . Despite its importance, degradation, the 'missing dimension' in proteomics , has received far less comprehensive attention than synthesis. This deficiency has arisen because developing the tools for a proteome-wide study of protein turnover is technically challenging. Proteins that are labile tend to be present at low abundance, and methods for characterizing turnover time are laborious.
We have developed an efficient and rather specific screen by combining GFP fluorescence, as a high-throughput measure of protein abundance, with pharmacologic shutoff of protein synthesis. Of 30 clones that were recovered from the screen (Figure 1) and individually examined by CHX treatment and FACS analysis, 22 (73%) are associated with proteins with a half-life of less than 4 hours. Given the relative rarity of rapidly degraded proteins in the proteome , this result demonstrates the specificity of the screening method. We have so far analyzed a restricted subset of the clones that were recovered in our screening procedure - 30 clones present in one of eight positive pools (among 12) from the R4 population. A second population, R3, appears to be equally rich in clones responsive to CHX. Extrapolation from this small sample implies that perhaps 300-400 (that is, 22 × 8 × 2) clones within the GFP-cDNA library may be found to be associated with proteins that are labile according to our secondary screening criterion. In contrast to the results with the less bright R3 and R4 cell populations, the failure to detect a CHX-sensitive subpopulation among the brighter R5-R6 cells is consistent with the expectation that labile proteins tend to be of lower abundance than more stable proteins.
For some of the proteins uncovered in this survey, rapid turnover can be rationalized as intrinsic to their cellular function. SRp30c factor (accession number U87279) is responsible for pre-mRNA splicing. Alterative splicing is a commonly used mechanism to create protein isoforms. It has been proposed that organisms regulate alternative splice site selection by changing the concentration and activity of splicing regulatory proteins such as SRp30c in response to external stimuli . The finding that SRp30c is a short-lived protein is consistent with its postulated regulatory function.
The G proteins are a ubiquitous family of proteins that transduce information across the plasma membrane, coupling receptors to various effectors [16, 17]. About 80% of all known hormones, neurotransmitters and neuromodulators are estimated to exert their cellular regulation through G proteins. The G protein (accession number M69013) shown here to short-lived is a G protein α subunit that transduces signals via a pertussis toxin-insensitive mechanism . Like other pertussis toxin-insensitive G proteins such as the Ga12 class, it causes the activation of several cytoplasmic protein tyrosine kinases: Src, Pyk2 (proline-rich tyrosine kinase 2) and Fak (focal adhesion kinase) . However, it is not known how this G protein is regulated. Its rapid turnover suggests a testable mechanism of its regulatory activation. Cervical cancer 1 proto-oncogene protein p40 (accession number AF195651), is a third protein shown here to turn over rapidly, but its function is unknown. Further studies of its turnover may provide important information on its function and regulation.
In mammalian cells, proteasomes have the predominant role in the degradation of short lived proteins, whereas lysosomal degradation appears to be quantitatively less important . Determining the mechanism that cells use to degrade the proteins uncovered by the method described here will require the use of specific inhibitors . Before degradation, most short-lived proteins are covalently coupled to multiple copies of the 76-amino-acid protein ubiquitin , a reaction catalyzed by a series of enzymes . These ubiquitinated proteins are recognized by the 26S proteasome and degraded within its hollow interior . This system of regulated degradation is central to such processes as cell-cycle progression, gene transcription and antigen processing. A few proteins have been found to be exceptions [25, 26]; like ODC, they do not require ubiquitin modification for degradation by the proteasome. In most cases it is not clear how short-lived proteins are selected to be modified and degraded. Some rapidly degraded proteins have been shown to contain an identifiable 'degradation domain'. Removal of this degradation domain makes such proteins stable, and appending this domain to a stable protein reduces its stability. Such a degradation domain has been identified in a number of short-lived proteins, including the carboxy terminus of mouse ODC [6, 27] and the destruction box of cyclins . In some cases, the signal is a primary sequence - like the PEST sequence [29, 30]. However, the identifiable structural features of such degradation domains are not sufficiently uniform to provide a reliable guide to identifying labile proteins. The method we have described does not use ubiquitin conjugation as a search criterion. This approach thus has the potential to discover labile proteins regardless of whether ubiquitin modification plays a role in their turnover. Once a large and representative sample of short-lived proteins is identified, a search for structural motifs among these proteins may facilitate the discovery of those motifs which correlate to protein degradation.
In this study we have developed an innovative technology to identify labile proteins using GFP-fusion expression libraries. Using this technology we have discovered short-lived proteins in a high-throughput format. This technology will greatly facilitate the discovery and study of short-lived proteins and their cellular regulation.
Materials and methods
Construction of GFP-cDNA expression libraries
Messenger RNAs from brain, liver, and the HeLa cell line (Clontech) were used as templates for cDNA synthesis, using a cDNA synthesis kit from Stratagene according to the manufacturer's recommendation, with some modifications. First-strand cDNA was synthesized using an oligo(dT) primer-linker containing an XhoI restriction site and with StrataScript reverse transcriptase. Synthesis was performed in the presence of 5-methyl dCTP, resulting in hemimethylated cDNA, which prevents endogenous cutting within the cDNA during cloning. Second-strand cDNA was synthesized using E. coli DNA polymerase and RNase H. Adaptors containing EcoRI cohesive ends were introduced into the double-stranded cDNA, which were then digested with XhoI. The cDNAs contained two different sticky ends: 5' EcoRI and 3' XhoI. The cDNAs were separated on a 1% SeaPlaque GTG agarose gel in order to collect those larger than 800 bp. After extracting cDNAs from the agarose gel with AgarACE-agarose-digesting enzyme followed by ethanol precipitation, the cDNAs were directionally cloned into EGFP-C1/2/3 expression vectors with three open reading frames (ORFs) (Clontech). The vectors were modified within the multiple cloning sites in order to be compatible with the cDNA orientation. By this means, cDNA ORFs were aligned to the carboxy terminus of EGFP. The host cell used for plasmid transfection and expression, 293T, expresses the SV40 large T antigen. Therefore, the cDNA EGFP-C1/2/3 vector containing the SV40 origin of replication can replicate independently from chromosome DNA in the host cells, which facilitates the recovery of plasmid DNAs from the host cells.
Transfection of the libraries into 293T cells
293T cells were cultured at 37°C in DMEM (Invitrogen) supplemented with 10% FBS, 1% nonessential amino acids and 100 U/ml penicillin, 0.1 mg/ml streptomycin. One day before transfection, cells were seeded in 10-cm plate in 10 ml growth medium without antibiotics. Transfection was performed using Lipofectamine 2000 reagent according to the manufacturer's instructions. Samples (25 μg) of a cDNA library were diluted in 1.5 ml Opti-MEM (Invitrogen). Lipofectamine 2000 was diluted in 1.5 ml Opti-MEM and mixed with diluted DNA. After 20 min incubation, the DNA-Lipofectamine 2000 complex was added to the cells. The cells were incubated for 16 h before analysis.
FACS analysis of GFP-expressing cells
Cells were harvested by trypsinization, washed, and resuspended in DMEM. Cytometric analysis and sorting were performed using a hybrid cell sorter combining a Becton Dickinson FACStarPLUS optical bench with Cytomation Moflo electronics (Stanford Beckman Center shared facility). Green fluorescence was measured using a 525/50 band pass filter. Gates were set to exclude cellular debris and the fluorescence intensity of events within the gated regions was quantified. Fluorescence-activated cell sorting was performed with a lower forward scatter threshold to detect transfected cells while ensuring that debris and electronic noise were not captured as legitimate events. Transfection efficiency was so high that normal voltages for detecting GFP were reduced. For fractionation, the cell population was gated on the basis of the fluorescence intensity. Cells were sorted at a rate of 8,000 events/sec. 106 cells were collected in 12 × 75 mm glass tubes containing 200 μl serum to enhance the cell survival rate. For short-lived protein screening, sorted cells were recultured in a 12-well plate and treated with or without 100 μg/ml CHX for 2 h. The cells then were collected and subjected to FACS analysis and sorting. The cells showing a decrease in fluorescence intensity with CHX treatment were collected for further analysis.
Plasmid DNA was extracted from sorted cells using a Qiagen mini-plasmid preparation kit. Plasmid DNAs were eluted in water and transformed into electro-competent DH10B E. coli (Invitrogen). Bacterial colonies were transferred to 96-well plates containing LB with 50 μg/ml kanamycin and 30% glycerol. After overnight growth at 37°C, the colonies are stored at -80°C. Plasmid DNAs were prepared from individual clones, sequenced and BLAST searches performed against the NCBI database.
Construction of Myc-tagged full-length coding sequences of genes
To obtain full-length coding sequence of the genes, we amplified them with a human full-length cDNA kit (Panomics) according to the manufacturer's instructions. The full-length coding sequences of cDNAs were then cloned into the pCMV-Myc vector (Clontech) for expression in 293T cells.
Western blot analysis of protein degradation
The plasmid DNAs of individual clones were prepared and transfected into 293T cells. The transfected cells, with or without CHX treatment, were collected in PBS and cell lysates were prepared by sonication. Proteins were resolved by SDS-polyacrylamide gel electrophoresis and transferred to a membrane. Fusion proteins were detected using a polyclonal antibody against GFP (Clontech), a monoclonal antibody against the Myc epitope (Sigma), a polyclonal antibody against G protein (Santa Cruz) or an antibody against Hsp70 (Santa Cruz). Bands were visualized with SuperSignal West Pico kit (Pierce).
Additional data files
Additional data file 1 contains the original data used to perform this analysis and is available with the online version of this paper.
We thank the staff of the FACS service center at the Beckman Center, Stanford University, for their technical support, and Robert Lam and Shanmei Li at Panomics for their help. This work was supported by NIH SBIR grant R43 GM64036 to X.L. and NIH grant RO1 45335 to P.C.
- Glickman MH, Ciechanover A: The ubiquitin-proteasome proteolytic pathway: destruction for the sake of construction. Physiol Rev. 2002, 82: 373-428.PubMedView ArticleGoogle Scholar
- Ciechanover A, Orian A, Schwartz AL: Ubiquitin-mediated proteolysis: biological regulation via destruction. BioEssays. 2000, 22: 442-451. 10.1002/(SICI)1521-1878(200005)22:5<442::AID-BIES6>3.0.CO;2-Q.PubMedView ArticleGoogle Scholar
- Schimke RT: Control of enzyme levels in mammalian tissues. Adv Enzymol. 1973, 37: 135-187.PubMedGoogle Scholar
- Sun SC, Ganchi PA, Ballard DW, Greene WC: NF-kappa B controls expression of inhibitor I kappa B alpha: evidence for an inducible autoregulatory pathway. Science. 1993, 259: 1912-1915.PubMedView ArticleGoogle Scholar
- Henkel T, Machleidt T, Alkalay I, Kronke M., Ben-Neriah Y, Baeuuerle PA: Rapid proteolysis of I kappa B-alpha is necessary for activation of transcription factor NF-kappa B. Nature. 1993, 365: 182-185. 10.1038/365182a0.PubMedView ArticleGoogle Scholar
- Beg AA, Finco TS, Nantermet PV, Baldwin AS: Tumor necrosis factor and interleukin-1 lead to phosphorylation and loss of I kappaB alpha: a mechanism for NF-kappa B activation. Mol Cell Biol. 1993, 13: 3301-3310.PubMedPubMed CentralView ArticleGoogle Scholar
- Pratt JM, Petty J, Riba-Garcia I, Robertson DHL, Gaskell SJ, Oliver SG, Beynon RJ: Dynamics of protein turnover, a missing dimension in proteomics. Mol Cell Proteomics. 2002, 1: 579-591. 10.1074/mcp.M200046-MCP200.PubMedView ArticleGoogle Scholar
- Tsien RY: The green fluorescent protein. Annu Rev Biochem. 1998, 67: 509-544. 10.1146/annurev.biochem.67.1.509.PubMedView ArticleGoogle Scholar
- Zhang M, Pickart CM, Coffino P: Determinants of proteasome recognition of ornithine decarboxylase, a ubiquitin-independent substrate. EMBO J. 2003, 22: 1488-1496. 10.1093/emboj/cdg158.PubMedPubMed CentralView ArticleGoogle Scholar
- Li X, Zhao X, Fang Y, Jiang X, Duong T, Fan C, Huang CC, Kain SR: Generation of destabilized green fluorescent protein as a transcription reporter. J Biol Chem. 1998, 273: 34970-349755. 10.1074/jbc.273.52.34970.PubMedView ArticleGoogle Scholar
- Li X, Fang Y, Zhao X, Jiang X, Duong T, Kain SR: Characterization of NFκB activation by detection of green fluorescent protein-tagged IκB degradation in living cells. J Biol Chem. 1999, 274: 21244-21250. 10.1074/jbc.274.30.21244.PubMedView ArticleGoogle Scholar
- Huh WK, Falvo JV, Gerke LC, Carroll AS, Howson RW, Weissman JS, O'Shea EK: Global analysis of protein localization in budding yeast. Nature. 2003, 425: 686-691. 10.1038/nature02026.PubMedView ArticleGoogle Scholar
- Galbraith DW, Anderson MT, Herzenberg LA: The lack of susceptibility of the brighter R5 and R6 subpopulations. In Green Fluorescent Proteins. 1999, San Diego: Academic Press, 315-341.Google Scholar
- Futcher B, Latter GI, Monardo P, McLaughlin CS, Garrels JI: A sampling of the yeast proteome. Mol Cell Biol. 1999, 19: 7357-7368.PubMedPubMed CentralView ArticleGoogle Scholar
- Stamm S: Signals and their transduction pathways regulating alternative splicing: a new dimension of the human genome. Hum Mol Genet. 2002, 11: 2409-2416. 10.1093/hmg/11.20.2409.PubMedView ArticleGoogle Scholar
- Spiegel AM, Shenker A, Weinstein LS: Receptor-effector coupling by G proteins: implications for normal and abnormal signal transduction. Endocr Rev. 1992, 13: 536-565. 10.1210/er.13.3.536.PubMedView ArticleGoogle Scholar
- Exton JH: Cell signalling through guanine-nucleotide-binding regulatory proteins (G proteins) and phospholipases. Eur J Biochem. 1997, 243: 10-20. 10.1111/j.1432-1033.1997.t01-1-00010.x.PubMedView ArticleGoogle Scholar
- Jiang M, Pandey S, Tran VT, Fong HK: Guanine nucleotide-binding regulatory proteins in retinal pigment epithelial cells. Proc Natl Acad Sci USA. 1991, 88: 3907-3911.PubMedPubMed CentralView ArticleGoogle Scholar
- Susa M: Heterotrimeric G proteins as fluoride targets in bone. Int J Mol Med. 1999, 3: 115-126.PubMedGoogle Scholar
- Fuertes G, Villarroya A, Knecht E: Role of proteasomes in the degradation of short-lived proteins in human fibroblasts under various growth conditions. Int J Biochem Cell Biol. 2003, 35: 651-664. 10.1016/S1357-2725(02)00382-5.PubMedView ArticleGoogle Scholar
- Kisselev AF, Goldberg AL: Proteasome inhibitors: from research tools to drug candidates. Chem Biol. 2001, 8: 739-758. 10.1016/S1074-5521(01)00056-4.PubMedView ArticleGoogle Scholar
- Chau V, Tobias JW, Bachmair A, Marriott D, Ecker DJ, Gonda DK, Varshavsky A: A multiubiquitin chain is confined to specific lysine in a targeted short-lived protein. Science. 1989, 243: 1576-1583.PubMedView ArticleGoogle Scholar
- Ciechanover A, Schwartz AL: The ubiquitin-proteasome pathway: the complexity and myriad functions of protein death. Proc Natl Acad Sci USA. 1998, 95: 2727-2730. 10.1073/pnas.95.6.2727.PubMedPubMed CentralView ArticleGoogle Scholar
- Zwickl P, Voges D, Baumeister W: The proteasome: a macromolecular assembly designed for controlled proteolysis. Philos Trans R Soc Lond B Biol Sci. 1999, 354: 1501-1511. 10.1098/rstb.1999.0494.PubMedPubMed CentralView ArticleGoogle Scholar
- Verma R, Deshaies RJ: A proteasome howdunit: the case of the missing signal. Cell. 2000, 101: 341-344. 10.1016/S0092-8674(00)80843-0.PubMedView ArticleGoogle Scholar
- Coffino P: Degradation of ornithine decarboxylase, a ubiquitin-independent proteasomal process. In Proteasomes: The World of Regulatory Proteolysis. 2000, Georgetown, TX: Landes Bioscience Press, 254-263.Google Scholar
- Li X, Stebbins B, Hoffman L, Pratt G, Rechsteiner M, Coffino P: The N terminus of antizyme promotes degradation of heterologous proteins. J Biol Chem. 1996, 271: 4441-4446. 10.1074/jbc.271.8.4441.PubMedView ArticleGoogle Scholar
- Glotzer M, Murray AW, Kirschner MW: Cyclin is degraded by the ubiquitin pathway. Nature. 1991, 349: 132-138. 10.1038/349132a0.PubMedView ArticleGoogle Scholar
- Rogers S, Wells R, Rechsteiner M: Amino acid sequences common to rapidly degraded proteins: the PEST hypothesis. Science. 1986, 234: 364-368.PubMedView ArticleGoogle Scholar
- Rechsteiner M, Rogers SW: PEST sequences and regulation by proteolysis. Trends Biochem Sci. 1996, 21: 267-271. 10.1016/0968-0004(96)10031-1.PubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.