Degradation of E2A Proteins through a Ubiquitin-conjugating Enzyme, UbcE2A*

(Received for publication, October 22, 1996)

Choon-Joo Kho , Gordon S. Huggins Dagger , Wilson O. Endege , Chung-Ming Hsieh , Mu-En Lee §par and Edgar Haber Dagger §**

From the Cardiovascular Biology Laboratory, Harvard School of Public Health, Boston, the § Department of Medicine, Harvard Medical School, Boston, the  Cardiovascular Division, Brigham and Women's Hospital, Boston, Massachusetts 02115, and the Dagger  Cardiac Unit, Massachusetts General Hospital, Boston, Massachusetts 02114

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
Acknowledgments
REFERENCES


ABSTRACT

The helix-loop-helix E2A proteins (E12 and E47) govern cellular growth and differentiation. To identify binding partners that regulate the function of these ubiquitous transcription factors, we screened for proteins that interacted with the C terminus of E12 by the yeast interaction trap. UbcE2A, a rat enzyme that is highly homologous to and functionally complements the yeast ubiquitin-conjugating enzyme UBC9, was identified and cloned. UbcE2A appears to be an E2A-selective ubiquitin-conjugating enzyme because it interacts specifically with a 54-amino acid region in E47-(477-530) distinct from the helix-loop-helix domain. In contrast, most of the UbcE2A protein is required for interaction with an E2A protein. The E2A proteins appear to be degraded by the ubiquitin-proteasome pathway because the E12 half-life of 60 min is extended by the proteasome inhibitor MG132, and E12 is multi-ubiquitinated in vivo. Finally, antisense UbcE2A reduces E12 degradation. By participating in the degradation of the E2A proteins, UbcE2A may regulate cell growth and differentiation.


INTRODUCTION

By alternative splicing, the E2A gene encodes two proteins, E12 and E47, through two adjacent exons encoding a basic helix-loop-helix (HLH)1 motif (1). These proteins belong to a family of eukaryotic transcription regulators distinguished by the highly conserved HLH motif, which mediates dimerization, and by the adjacent basic region, which mediates site-specific DNA binding (2, 3). The ubiquitous E2A proteins form heterodimers with tissue-specific HLH proteins that then bind to DNA and up-regulate the transcription of target genes. Tissue-specific HLH proteins include the MyoD family involved in skeletal muscle differentiation (4), the achaete-scute family involved in neuronal differentiation (5), and SCL/TAL, which is involved in hematopoiesis (6). The E2A proteins also form homodimers that are linked by intermolecular disulfide bonds in B cells but not muscle cells (7). These homodimers are thought to be the predominant DNA-binding species in B cells (8). In mice carrying a null mutation in E2A, immunoglobulin gene segments do not rearrange and the animals lack mature B lymphocytes (9, 10).

In addition to its role in cellular differentiation, the E2A gene is the breakpoint of two translocations associated with childhood lymphoid leukemia. A truncated E2A gene fuses to the PBX1 homeobox gene (11) and to the HLF basic leucine zipper gene (12). Because the E2A portion is required for transformation in both instances, E2A proteins appear to play a role in growth control. Peverali et al. (13) have shown that overexpression of E12 or E47 inhibits cell proliferation and mediates arrest of growth a few hours before the G1-S transition of the cell cycle. The level of E2A proteins at different stages of the cell cycle could also determine whether cells proliferate or differentiate.

Certain transcription factors and cell cycle regulators are degraded rapidly in vivo (14). For example, c-Fos and c-Jun, which have half-lives of about 30 and 90 min, respectively, cause uncontrolled cell proliferation if their expression goes unchecked (15, 16). Also, the cyclins and cyclin-dependent kinase inhibitors must undergo programmed destruction if the cell cycle is to continue (17, 18). The ubiquitin-proteasome pathway fosters the rapid turnover of many cell regulators. These include the transcription factors MATalpha 2 (19) and GCN4 (20) from yeast, c-Fos (21) and c-Jun (22), and the cell cycle regulators cyclin B (23) and cyclin-dependent kinase inhibitor p27 (24). The ubiquitin-proteasome pathway also mediates processing of the p105 precursor of NF-kappa B and degradation of its inhibitor protein Ikappa Balpha (25, 26).

The ubiquitin-proteasome pathway involves covalent conjugation of a target protein to ubiquitin molecules, degradation of that protein, and release of reusable ubiquitin, as reviewed by Ciechanover (27). Ubiquitin is activated initially by ATP in a reaction catalyzed by the enzyme E1. The activated ubiquitin is then transferred to a ubiquitin-conjugating enzyme, E2, that catalyzes formation of an isopeptide bond between the C-terminal glycine of ubiquitin and the epsilon -amino group of a lysine residue on the target protein. For many proteins, conjugation to ubiquitin also requires a specific ubiquitin-protein ligase, E3. A mono-ubiquitinated target protein then undergoes further ubiquitination to produce multi-ubiquitinated chains (28). These ubiquitin conjugates are recognized by a multisubunit regulatory complex on the proteasome that also unfolds and translocates them to the barrel-shaped 20 S core, where they are degraded (29, 30).

Despite the importance of the E2A proteins in cell growth and differentiation, little is known about the mechanisms regulating their stability. While isolating proteins that bound to E12 by the yeast interaction trap, we cloned UbcE2A, the rat homologue of the yeast ubiquitin-conjugating enzyme UBC9. We found that E12 turns over rapidly and is multi-ubiquitinated and that its half-life is extended by a proteasome inhibitor. Moreover, antisense UbcE2A reduces E12 degradation. These observations suggest that E12 is regulated by the ubiquitin-proteosome pathway. By regulating the level of E12, UbcE2A may regulate the cell cycle.


EXPERIMENTAL PROCEDURES

Plasmids

Escherichia coli and nucleic acids were manipulated as described by Ausubel et al. (31). We were given the following cDNAs: E12 and E47 (11), deletions and point mutants of E47 generated by PCR (13), and mouse c-Myc (32). We cloned the following cDNAs by reverse transcriptase PCR and confirmed their sequences: rat Id3 (33): rat Max (34), human OCT-1 (35), and rat c-Jun (36). Mathias Treier (European Molecular Biology Laboratory, Heidelberg, Germany) provided the ubiquitin construct pCMVHA-Ub (22). The pCR3 vector (Invitrogen) containing the cytomegalovirus enhancer and promoter and a bovine growth hormone polyadenylation signal was used for expression in eukaryotic cells. Full-length E12, UbcE2A, and c-Jun cDNAs were amplified by PCR and ligated into pCR3 by TA cloning. cDNA authenticity was confirmed by dideoxy sequencing and translation of the appropriate protein in vitro. Various E12, E47, and UbcE2A deletion mutants were generated by standard PCR techniques and sequenced. Hemagglutinin (HA)-tagged UbcE2A contained the sequence MASYPYDVPDYASPEF added to the N terminus of full-length UbcE2A. The pGEX4T vector (Pharmacia Biotech Inc.) was used to express glutathione S-transferase (GST) fusion proteins in E. coli.

Cell Culture and Antibodies

All cells were maintained in Dulbecco's modified Eagle's medium (DMEM) plus 10% fetal calf serum (Hyclone), 100 units/ml penicillin, and 100 mg/ml streptomycin in a humidified atmosphere at 37 °C with 5% CO2. Mouse monoclonal anti-HA antibody (12CA5) was purchased from Berkeley Antibody Co. (Richmond, CA), anti-human E12/E47 monoclonal antibody from Pharmingen (San Diego, CA), anti-human E12 rabbit polyclonal antibody, and anti-mouse c-Jun antibody from Santa Cruz Biotechnology (Santa Cruz, CA), goat anti-mouse IgG-HRP from Amersham Corp., and rhodamine-conjugated anti-mouse IgG from Kirkegaard & Perry Laboratories (Gaithersburg, MD). Normal rabbit and mouse sera were purchased from ICN Biochemicals (Costa Mesa, CA).

Yeast Interaction Trap Experiments

We screened for E12-interacting proteins by the yeast two-hybrid interaction trap according to Gyuris et al. (37). EGY48 (MATalpha trp1 ura3 his3 LEU2::pLexop6-LEU2) was used as host yeast strain in all interaction experiments. All bait plasmids were constructed by inserting the corresponding cDNA (in-frame) downstream of the lexA gene contained in pEG202 (38). The oligo(dT)-primed rat aortic cDNA library used for screening had been constructed with the yeast galactose-inducible expression plasmid pJG4-5 (37). This library comprises 4.5 × 106 members, 88% of which contain a cDNA insert whose average size ranges between 0.6 and 2.3 kilobase pairs. We began the interaction screen with an EGY48-p1840-pLexA-E12-(477-654) (amino acids 477-654 of human E12) strain. pLexA-E12-(477-654) did not spontaneously activate transcription of either reporter gene (lacZ or LEU2) used in this system. We confirmed expression of the appropriate bait protein by Western blotting with both anti-LexA antibody (gift of Barak Cohen, Massachusetts General Hospital, Boston) and anti-E12/E47 antibody. We introduced the rat aortic cDNA library into the EGY48-p1840-pLexA-E12-(477-654) strain according to the procedure of Gietz et al. (39), with modification. A total of 4 × 106 transformants was obtained. Plasmids were screened and recovered as described by Gyuris et al. (37). Library plasmids were classified by their restriction patterns after digestion with EcoRI and XhoI and either HinfI or HaeIII. Plasmid DNAs from each class were retested in the interaction-trap assay with pEG202 and pLexA-E12-(477-654). Galactose-inducible expression of an HA-tagged fusion protein in the transformants was also confirmed with the anti-HA antibody 12CA5.

To assess the specificity of interaction and map the interaction domains, we transformed yeast of the EGY48/pSH18-34 strain with the library/interactant plasmids and the bait constructs indicated in Fig. 3 and applied them to glucose ura-his-trp- plates. The bait constructs used in the specificity test were LexA-Id3 (which contains all of the rat Id3 coding sequence), LexA-c-Myc (which contains the C-terminal 137 amino acids of mouse c-Myc), LexA-Max (which contains all of the rat Max coding sequences), and LexA-OCT-1 (amino acids 294-429 of human OCT-1 containing the POU domain). Eight to twelve colonies from each bait/interactant combination were picked and applied in duplicate to ura-his-trp- plates containing 5-bromo-4-chloro-3-indolyl-beta -D-galactoside, and either 2% glucose or 2% galactose, and 1% raffinose. We checked the color of the yeast 48 h later.


Fig. 3. Specificity of UbcE2A interactions in yeast by beta -galactosidase assay. Cells of the S. cerevisiae strain EGY48/pSH18-34 were sequentially transformed with the indicated LexA fusion plasmid (Baits) and the activation domain (AD-UbcE2A) library isolate. At least three independent colonies from each AD-UbcE2A/LexA fusion protein pair were used to inoculate a galactose-containing liquid culture. beta -Galactosidase activity expressed from the lacZ reporter gene (normalized units) was measured; error bars indicate standard deviations. Expression of the appropriate LexA fusion proteins was confirmed by Western blotting (data not shown).
[View Larger Version of this Image (14K GIF file)]


Crude extracts were assayed for yeast beta -galactosidase activity as described by Kaiser et al. (40). Cells bearing the appropriate bait and interaction plasmids were grown to saturation (overnight at 30 °C) in minimal ura-his-trp- medium with 2% glucose. The next day, cells were diluted 1:50 into medium containing 2% galactose and 1% raffinose and allowed to grow overnight. Lysates were then prepared and permeabilized as described (41). Cell concentrations were determined by measuring absorbance at 600 nm. beta -Galactosidase units were calculated by the equation 1000(A420)/(time(min)·vol(ml)·A600).

Transfection and Immunofluorescence

NIH3T3 fibroblasts were transfected by the calcium phosphate method. Colonies were picked with cloning cylinders after 18-21 days and expanded. Integration of transfected DNA in the transformants was confirmed by Southern blot analysis. COS7 cells were transfected transiently by electroporation. For immunofluorescence studies, transfected COS7 cells were grown to 75% confluence on chamber slides (Nunc). Cells were washed once with PBS and fixed for 20 min in 2% sucrose with 4% paraformaldehyde at room temperature. Fixed and permeabilized cells were hydrated in PBS for 5 min and incubated with 10% nonimmune rabbit serum in PBS with 0.1% Triton X-100 at room temperature for 20 min to suppress nonspecific binding of IgG. The slides were stained with anti-HA antibody 12CA5 (1:400 dilution) in a moist chamber for 1 h at room temperature. After three washes in PBS with 0.1% Triton X-100, the slides were incubated with 250 µl of rhodamine-conjugated goat anti-mouse IgG diluted 1:200 for 45 min at room temperature. The slides were washed extensively again and counterstained with Hoechst 33258 for 5 min, mounted, and viewed in a Nikon fluorescence microscope. 12CA5 staining and Hoechst staining were visualized and photographed in the same fields by changing filter sets.

In Vitro Binding Assays

GST fusion proteins were expressed and purified essentially as described by Smith and Johnson (42). Fresh, overnight cultures of E. coli (HB101) transformed with pGEX4T or pGEX4T E12-(477-654) were diluted 1:10 in LB medium containing ampicillin (100 µg/ml) and incubated (with shaking) for 3-5 h at 37 °C until the A600 reached 0.8. Isopropyl beta -D-thiogalactopyranoside was then added to a final concentration of 0.4 mM, and incubation was allowed to continue for another 3 h. Bacterial cultures were pelleted and resuspended in PBS with 1 mM phenylmethylsulfonyl fluoride and 1% (v/v) aprotinin. The bacteria were then lysed on ice by mild sonication, mixed with Triton X-100 to a final concentration of 1%, and centrifuged at 14,000 × g for 5 min at 4 °C. Aliquots (1 ml) of bacterial supernatant were rocked for 30 min at 4 °C with 25 µl of glutathione-Sepharose 4B (Pharmacia). The Sepharose beads were then washed three times with PBS. 35S-Labeled proteins were generated by using the TNT T7-coupled reticulocyte lysate system (Promega) with the Id3 or UbcE2A expression construct in pCite4 (Novagen). 35S-Labeled protein (3 µl) was incubated with the beads (25 µl) in 50 mM NaCl and bovine serum albumin (1 mg/ml) at 4 °C for 1 h (43). The beads were then washed four times with 0.1% Nonidet P-40 in PBS. Protein on the beads was fractionated by SDS-PAGE, stained with Coomassie Blue, and exposed to Kodak x-ray film.

Pulse-Chase Experiments and Immunoprecipitation

COS7 cells in 100-mm dishes (at about 80% confluence) were starved in Met-free DMEM (supplemented with 5% dialyzed fetal bovine serum) for 60 min at 37 °C. Cells were then pulse-labeled at 37 °C with 100 µCi/ml [35S]Met for 60 min at 37 °C. Cells were chased in warm DMEM supplemented with 100 µg/ml Met. For the experiment with the proteasome inhibitor MG132 (a gift from Alfred Goldberg, Harvard Medical School, Boston, MA), the inhibitor (at a concentration of 50 µM) was added 1 h before pulse-chase and was present throughout the pulse-chase periods. After the chase, dishes were washed three times with PBS and then lysed with 3 ml of ice-cold RIPA (PBS, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, and CompleteTM protease inhibitor mixture (Boehringer Mannheim)) for 20 min at 4 °C. Lysates were cleared of nuclei and debris by centrifugation at 14,000 × g at 4 °C for 15 min. Spun samples were then cleared with normal mouse serum and protein G-agarose (Pierce) for 1 h at 4 °C. 35S incorporation in the total protein pool was determined by trichloroacetic acid precipitation. Lysate volumes were adjusted so that each extract contained equivalent amounts of radioactivity (trichloroacetic acid-precipitable counts/min). For E12 immunoprecipitation, lysates were incubated overnight at 4 °C with 1-2 µg of purified anti-E12/E47 antibody and immobilized protein G. E12 bound to the beads was washed four times with RIPA and subjected to SDS-PAGE followed by fluorography. E12 reactivity in the bands was measured on a PhosphorImager (Molecular Dynamics).

In Vivo Ubiquitination Assay

COS7 cells were electroporated with 6 µg of the E12 or c-Jun expression construct and 20 µg of the HA-tagged ubiquitin expression vector. After 48 h, cells were lysed on ice in RIPA buffer with 10 mM N-ethylmaleimide. The cells were harvested, and cysteine was added to a final concentration of 0.1% to inactivate the N-ethylmaleimide. Cell extracts were immunoprecipitated as described for the pulse-chase experiments; proteins were separated by 10% SDS-PAGE and blotted onto Immobilon-P membranes (Millipore). Blots were immunostained successively with anti-HA antibody 12CA5 and anti-E12 antibody. Reactive products were visualized with an ECL kit (Amersham Corp.).


RESULTS

A Ubiquitin-conjugating Enzyme Cloned by the Yeast Interaction Trap

We used the yeast interaction trap cloning system (37) to identify proteins that interact with the C terminus of E12. A bait expression vector was constructed by fusing the LexA-binding domain to the C terminus of E12 (amino acids 477-654), which includes the basic and HLH domains. We screened a rat aorta cDNA expression library with LexA-E12-(477-654) and identified 42 positive clones from 3.5 × 106 transformants. Of the 42 positive clones, 29 encoded Id3 (33) and five encoded Id1 (44), indicating that specific protein-protein interactions were detectable in yeast with our E12 construct. Of the eight remaining clones, five encoded a ubiquitin-conjugating enzyme containing the highly conserved active site. We named this gene ubcE2A.

A comparison of the predicted amino acid sequence of UbcE2A with known ubiquitin-conjugating enzyme sequences revealed that ubcE2A is most homologous to Saccharomyces cerevisiae UBC9 (75% similarity) (45), Schizosaccharomyces pombe hus5 (82% similarity) (46), and the recently published human homologue of UBC9 (100% similarity) (47). S and M phase cyclins are degraded by UBC9, an essential nuclear ubiquitin-conjugating enzyme in budding yeast (45), and Schizosaccharomyces pombe do not grow when hus5 is mutated (46). Also, in yeast harboring the temperature-sensitive ubc9-1 mutation (45), ubcE2A rescued the ubc9-1 mutant from growth inhibition (data not shown). Because of the sequence homology and the functional complementation of the ubc9-1 mutant, we conclude that ubcE2A is the rat homologue of S. cerevisiae UBC9.

Cellular Localization of UbcE2A Protein

To localize the UbcE2A protein, we transfected monkey COS7 cells with a plasmid expressing HA-tagged UbcE2A and analyzed them by indirect immunofluorescence. A monoclonal anti-HA antibody (12CA5) and a rhodamine-tagged secondary antibody were used to detect HA-UbcE2A in transfected cells. The UbcE2A protein was expressed primarily in the nuclei (Fig. 1, left), as confirmed by counterstaining with Hoechst 33258. No staining was visible in vector-transfected cells (not shown). The immunoblot of nuclear extract from HA-UbcE2A-transfected cells (Fig. 1, right) showed a protein of 20 kDa, consistent with the expected molecular mass of UbcE2A. Because the UbcE2A protein localizes to the nucleus, it may act on E2A nuclear factors.


Fig. 1. Nuclear localization and functional activity of UbcE2A. COS7 cells transfected with hemagglutinin (HA)-tagged UbcE2A (HA-UbcE2A) were fixed and stained with anti-HA antibody 12CA5 and rhodamine-conjugated secondary antibody. Staining with Hoechst 33258 shows the position of the nuclei. The immunoblot on the right shows nuclear extracts prepared from HA-UbcE2A-transfected or mock-transfected cells stained with 12CA5. Control was total extract prepared from yeast expressing HA-UbcE2A. The arrow indicates HA-UbcE2A protein.
[View Larger Version of this Image (52K GIF file)]


UbcE2A Binds to the E2A Proteins in Vitro

To confirm the interaction observed in yeast, we performed an in vitro binding assay. Radiolabeled, in vitro translated Id3 or UbcE2A was bound to GST-E12-(477-654) that had been immobilized on glutathione-Sepharose beads (Fig. 2). As anticipated, UbcE2A associated with GST-E12-(477-654) but not with GST. The interaction of Id3 with GST-E12-(477-654) served as a positive control. [35S]Methionine-labeled, in vitro translated UbcE2A was also immunoprecipitated with an antibody to E12 in the presence of in vitro translated E12 protein (data not shown). We conclude that a specific interaction takes place between E12 and UbcE2A.


Fig. 2. Specificity of UbcE2A interactions in vitro by GST assay. Bacterially produced GST-E12-(477-654) or GST alone was bound to glutathione-Sepharose beads and incubated with [35S]methionine-labeled Id3 or UbcE2A produced by in vitro translation. Specifically bound protein was resolved by SDS-PAGE. There was at least 10 times more GST than GST-E12-(477-654) on a gel stained with Coomassie Blue (data not shown).
[View Larger Version of this Image (45K GIF file)]


Interaction between UbcE2A and the E2A Proteins Is Specific

To study the specificity of the interaction between E12 and UbcE2A, we introduced full-length UbcE2A fused to the B42 transcription activation domain (AD-UbcE2A) into yeast cells containing various LexA fusion proteins. beta -Galactosidase expression from the lacZ reporter gene increased by 20-fold in lysates from yeast bearing AD-UbcE2A and LexA-E12-(477-654) or LexA-E47-(477-651) (Fig. 3; E12 and E47 versus Vector). This observation indicates that E12 and E47 interacted with UbcE2A equally well and that the primary amino acid sequence within the differentially spliced region was not crucial for binding. We examined the specificity of the interaction partners further by transforming yeast harboring expression plasmids encoding LexA fused to known HLH proteins. No interaction was detected with LexA fused to Id3 (33), the leucine zipper protein Max (34), or the homeodomain protein OCT-1 (35) (Fig. 3). Weak promoter activity was detected after introduction of LexA-c-Myc; however, LexA-c-Myc has been shown to cause higher background LacZ expression when studied with other proteins (48). In addition, we have observed no interaction between E12-(477-654) and UBCH5 (49), the human ubiquitin-conjugating enzyme involved in the ubiquitination of p53 (data not shown).

Map of UbcE2A-Interacting Regions

To map the E2A protein domain that binds to UbcE2A, we generated deletion mutants and assayed transcriptional activity by the yeast interaction trap. As anticipated, deletion of the basic or the HLH region had no effect on UbcE2A binding to E47 (Fig. 4, left). More extensive mapping localized the binding site to a 54-amino acid region, E47-(477-530), 5'-proximal to the basic HLH domain (Fig. 4, left). This region is conserved in both E12 and E47. By itself this region conferred specific binding to UbcE2A; moreover, a construct lacking the E47-(477-538) region bound to Id3 but had no affinity for UbcE2A (Fig. 4, right; E47Delta -(477-538)). In contrast with this small interaction domain on E12/E47, almost the entire UbcE2A protein, including the conserved catalytic site, was required for binding to E12; only about 29 amino acids at the C terminus were dispensable (Fig. 5, left and right).


Fig. 4. Binding of LexA-E47 fusion baits to full-length AD-UbcE2A interactant protein. Left, E47 regions used as baits in yeast interaction trap assay. The basic domain is black; the HLH domain is striped. The asterisk above the Ala mutant marks the five amino acid substitutions in the basic domain. In each case, at least six independent transformants were scored for the intensity of galactose-inducible blue that developed in the presence of 5-bromo-4-chloro-3-indolyl-beta -D-galactoside: +++, dark blue; +/-, faint blue flecks in some colonies; -, white colonies only. Right, beta -galactosidase activity in lysates prepared from transformants harboring the indicated protein pairs. beta -Galactosidase levels were measured in duplicate from three independent isolates; average value is shown.
[View Larger Version of this Image (18K GIF file)]



Fig. 5. Binding of UbcE2A fusion proteins to the E12-(477-654) fusion protein bait in yeast interaction trap assay. Left, UbcE2A regions used as interactants. Striped box indicates conserved catalytic domain of ubiquitin-conjugating enzymes. Right, beta -galactosidase levels in yeast expressing the indicated protein pairs. beta -Galactosidase levels were measured in duplicate from three independent isolates; average value is shown.
[View Larger Version of this Image (15K GIF file)]


The E12 Protein Is Unstable

The specific binding of E12 by a ubiquitin-conjugating enzyme suggested that the E2A protein half-life may be regulated by proteolysis. We first studied E12 turnover to test the possibility that the protein was metabolically labile. COS7 cells transfected with a human E12 expression plasmid were pulse-labeled with [35S]methionine for 60 min and then chased with unlabeled methionine for up to 120 min. E12 was immunoprecipitated from the lysates with an antibody to human E12 and analyzed by SDS-PAGE (Fig. 6A). Immunoprecipitation of the lysate revealed an Mr 72,000 band migrating at the same position as E12 protein translated in vitro (Fig. 6A, E12IVT). These experiments showed that E12 is labile in vivo and has a half-life of about 60 min (Fig. 6B). Similar results were obtained with NIH3T3 cells (data not shown). Thus, E12 appears to be the target of an intracellular degradation pathway.


Fig. 6. Pulse-chase analysis of E12 in transfected COS7 cells. A, cells expressing human E12 were labeled with [35S]methionine for 1 h and then chased with unlabeled methionine for the indicated times. Clarified cell lysates (3 × 105 cpm each) were subjected to immunoprecipitation with an anti-E12 antibody and analyzed by SDS-PAGE fluorography. [35S]Methionine-labeled, in vitro translated E12 (E12IVT) marks the position of the E12 protein. No signal was obtained from vector-transfected cells immunoprecipitated with anti-E12 antibody (Vector) or E12-transfected cells immunoprecipitated with preimmune serum (Preimmune). The fluorogram is from a typical experiment. B, profile of the E12 half-life obtained by PhosphorImaging analysis of the bands in A.
[View Larger Version of this Image (13K GIF file)]


E12 Is Degraded through the Ubiquitin-Proteasome Pathway

We used the method described by Palombella et al. (26) without modification to test whether E12 is degraded by a proteasome. Forty-eight hours after COS7 cells had been transfected with a human E12 expression plasmid, they were treated for 1 h with the proteasome inhibitor MG132 or the protease inhibitor leupeptin (a negative control). MG132 stabilized the E12 protein, whereas leupeptin had no effect (Fig. 7). We conclude that degradation of E12 involves a proteasome.


Fig. 7. The proteasome inhibitor MG132 blocks degradation of E12 in vivo. Monkey COS7 cells were electroporated with a human E12 expression plasmid. After 48 h, cells were treated with the proteasome inhibitor MG132, dimethyl sulfoxide (DMSO) (an MG132 diluent), or leupeptin for 1 h. They were then pulse-chased for 3 h with [35S]methionine. Cell extracts were immunoprecipitated with an anti-E12 antibody and analyzed by SDS-PAGE fluorography. Inhibitors were present throughout the pulse-chase period. The signal intensity of the E12 bands was measured by PhosphorImaging.
[View Larger Version of this Image (13K GIF file)]


Because protein degradation through a proteasome requires tagging of the protein by covalent attachment of multiple ubiquitin molecules (50), we next investigated whether E12 could be ubiquitinated in vivo by a method used to show ubiquitination of c-Jun (22). In these assays (Fig. 8), the E12 expression plasmid (pCR3E12) together with an HA-tagged ubiquitin expression plasmid (HA-Ub) were introduced into COS7 cells by transient transfection. c-Jun (pCR3jun), which is known to be multi-ubiquitinated (22), was used as a control. Equivalent amounts of lysate were immunoprecipitated with an antibody to E12 or c-Jun. The precipitated proteins were separated by SDS-PAGE, blotted onto Immobilon filters, and probed with a monoclonal antibody to HA (12CA5). The faint ladder of bands visible for c-Jun-transfected lysates above Mr 39,000 (relative molecular mass of c-Jun) indicated formation of multiple ubiquitin conjugates (Fig. 8). A more distinct ladder of bands was visible for E12-transfected lysates. In both, most of the reactivity appeared above Mr 200,000, which indicates significant multi-ubiquitination. Because E12 is ubiquitinated in vivo and proteasome inhibitors block its degradation, the ubiquitin-proteasome pathway appears to regulate the abundance of this transcription factor.


Fig. 8. Ubiquitination of nontagged c-Jun and E12. COS7 cells were electroporated with cytomegalovirus vectors (pCR3) directing synthesis of HA-tagged ubiquitin (HA-Ub) and nontagged c-Jun (pCR3jun) or E12 (pCR3E12). Cell extracts were immunoprecipitated with an anti-E12 antibody, eluted, separated by SDS-PAGE, and immunoblotted with a mouse monoclonal antibody to HA (12CA5). The arrow marks mouse immunoglobulin heavy chain that cross-reacted with the secondary antibody. Brackets to the right of the panels mark a smear of cross-reactive material corresponding to the size of ubiquitinated c-Jun and E12. The bottom right panel shows E12 expression in transfected cells.
[View Larger Version of this Image (31K GIF file)]


Overexpression of Antisense UbcE2A mRNA Stabilizes E12

To demonstrate the role of UbcE2A in E12 degradation more directly, we tested E12 expression by stably transfecting NIH3T3 cells with antisense ubcE2A cDNA (two antisense clones were studied, Asc3 and Asc6). Levels of 1.1-kilobase ubcE2A mRNA decreased in Asc3- and Asc6-transfected cells, to about 30 and 32%, respectively, the level in vector-transfected (control) cells, as measured by Northern blotting with a 32P-labeled ubcE2A antisense riboprobe (data not shown). Pulse-chase analysis was performed 48 h after these cells had been transiently transfected with an E12 expression plasmid. In both antisense clones, the E12 protein was stabilized by approximately 2-fold in comparison with the vector clone (Fig. 9). We conclude that UbcE2A plays an important role in regulating the level of E12 protein in cells.


Fig. 9. Inhibition of E12 degradation in cells transfected with antisense UBCE2A. Cells were stably transfected with pCR3 (Vector) or antisense ubcE2A expression plasmids (Asc3 and Asc6). The cells were then transiently transfected with a human E12 expression plasmid and analyzed by pulse-chase as described for Fig. 6. Results from a representative experiment are shown. The experiment was repeated twice with similar results.
[View Larger Version of this Image (55K GIF file)]



DISCUSSION

UbcE2A Is a Novel Binding Partner of the E2A Proteins

UbcE2A, the ubiquitin-conjugating enzyme we isolated by the yeast two-hybrid system, interacts specifically with the HLH E2A proteins E12 and E47. Because of sequence homology and functional complementation, rat UbcE2A appears to be a homologue of yeast UBC9. Although UbcE2A does not contain an HLH domain, it binds E12 and E47 specifically (Figs. 3 and 4). Most of the UbcE2A molecule is necessary for binding to the E2A proteins; however, a 54-amino acid region in E47 (amino acids 477-530) located 5' of the basic and HLH domains is sufficient to bind UbcE2A. Thus we have identified a novel E2A interaction domain, amino acids 477-530, that binds to UbcE2A and may regulate E12/E47 turnover.

The E2A Proteins Are Highly Unstable

The specific binding of E12 by a ubiquitin-conjugating enzyme suggests that proteolysis may regulate the half-life of the E2A proteins. Two observations suggest that the E2A proteins may have a short half-life. First, increases in the amount of E12 or E47 in the mid-G1 phase prevent entry into the S phase in serum-stimulated fibroblasts (13). Because the E2A proteins must be down-regulated if the cell cycle is to progress, they must have a short half-life. Second, a feature of rapidly degraded proteins is the presence of PEST sequences, polypeptide chains rich in proline, glutamate/aspartate, serine, and threonine (51). Using the PEST-FIND program (14), we identified three PEST sequences in E12 (amino acids 47-67, 169-189, and 521-537). Indeed, we found that E12 protein expression declined within 3 h of serum stimulation and became undetectable after 9 h (data not shown). By pulse-chase analysis, we found that E12 has a half-life of 60 min (Fig. 6B).

The E2A Proteins Are Degraded by the Ubiquitin-Proteasome Pathway through UbcE2A

Our observations indicate that the instability of the E2A proteins is mediated by the ubiquitin-proteasome pathway. The 20 S proteasome inhibitor MG132 completely blocked degradation of E12 in experiments in which the protease inhibitor leupeptin was used as a control (Fig. 7), and E12 was multiply ubiquitinated in an in vivo assay (Fig. 8). To our knowledge this is the first demonstration that the E2A proteins are degraded by the ubiquitin-proteasome pathway, which has also been shown to regulate other important transcription factors and cell cycle regulators (27).

The specificity of substrate recognition by the ubiquitin-proteasome pathway appears to be mediated by ubiquitin-conjugating enzymes, sometimes in conjunction with ubiquitin ligases. For example, a complex forms between the UBC6 and UBC7 enzymes in the ubiquitination pathway targeting degradation of the yeast transcription factor MATalpha 2 (52). Only two genes encoding ubiquitin ligases have been cloned so far, S. cerevisiae UBRI and human E6-AP. The UBRI protein interacts with the RAD6 ubiquitin-conjugating enzyme to form a complex that targets substrates bearing "destabilizing" N-terminal residues (N-end rule substrates) (53). The E6-AP protein interacts with the E6 oncoprotein of human papilloma virus and induces ubiquitination and subsequent degradation of p53 (54). Although we detected a direct interaction between E12 and its ubiquitin-conjugating enzyme, we cannot completely rule out the possibility that E12 ubiquitination requires an unknown ubiquitin ligase.

We found that ubcE2A mRNA expression is regulated differentially in fibroblasts (data not shown). Expression is maximal at the mid-G1 phase, before the onset of the S phase. Our finding is consistent with the observation that E12 is degraded during progression of the cell cycle (13). To maintain oscillating levels of regulatory proteins, the ubiquitin-conjugating machinery would have to be activated only at specific points in the cell cycle. For example, the yeast CDC34 ubiquitin-conjugating enzyme, which is required for the transition from the G1 to the S phase, is regulated by phosphorylation and ubiquitination (55). Although it is conceivable that UbcE2A is subject to similar modification (UbcE2A contains four putative phosphorylation sites), our observation that it is up-regulated during G1 suggests that the abundance of UbcE2A, and hence ubiquitination, may determine the rate of E12 turnover.

Specific ubiquitin-conjugating enzymes are necessary for the degradation of many cellular substrates. p53 degradation requires the human homologue of UBC4 but not that of UBC2 (54), and p27 degradation specifically involves the human homologues of UBC2 and UBC3 (24). These observations suggest that it may be possible to inhibit degradation of a substrate in vivo by inhibiting its specific ubiquitin-conjugating enzyme. Indeed, microinjection of an antisense UBC4 expression plasmid into human tumor cells containing high levels of the p53 protein inhibited E6-stimulated degradation of p53 (54). We demonstrate here by an antisense approach that down-regulation of UbcE2A expression inhibits degradation of E12 (Fig. 9). Because the tissue-specific gene transcription that moves cells from a proliferative to a differentiated state involves the ubiquitous E2A proteins, it may be possible to regulate cellular differentiation by targeting UbcE2A.


FOOTNOTES

*   This work was supported in part by a grant from Bristol-Myers Squibb. The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) U54632[GenBank].


par    Supported by National Institutes of Health Grant R01GM 53249.
**   To whom all correspondence should be addressed: Cardiovascular Biology Laboratory, Harvard School of Public Health, 677 Huntington Ave., Boston, MA 02115. Tel.: 617-432-1010; Fax: 617-432-4098; E-mail: haber{at}cvlab.harvard.edu.
1    The abbreviations used are: HLH, helix-loop-helix; UbcE2A, ubiquitin-conjugating enzyme that binds the E2A proteins; AD-UbcE2A, full-length UbcE2A fused to the B42 transcription activation domain; HA, hemagglutinin; HA-Ub, HA-tagged ubiquitin expression plasmid; GST, glutathione S-transferase; DMEM, Dulbecco's modified Eagle's medium; E1, ubiquitin-activating enzyme; E2, ubiquitin-conjugating enzyme; E3, ubiquitin-protein ligase; E12IVT, E12 protein translated in vitro; Asc3 and Asc6, antisense ubcE2A clones; PCR, polymerase chain reaction; PAGE, polyacrylamide gel electrophoresis.

Acknowledgments

We thank Roger Brent and Russell Finley (Massachusetts General Hospital, Boston) for plasmids, yeast strains, and antibody for the interaction trap; Alfred Goldberg and Olivier Coux (Harvard Medical School, Boston, MA) for proteasome inhibitors and helpful suggestions; Stefan Jentsch and Petra Hubbe (Zentrum für Molekulare Biologie, Heidelberg University, Germany), Fiorenzo Peverali (European Molecular Biology Laboratory, Heidleberg, Germany), and Mathias Treier (EMBL) for plasmids and yeast strains; David Baltimore (Massachusetts Institute of Technology, Cambridge) and Zhengsheng Ye (The Rockefeller University, New York) for plasmids and technical assistance; and Martin Rechsteiner (University of Utah School of Medicine, Salt Lake City) for the PEST-FIND program. We are also grateful to Bonna Ith for cell culture and Thomas McVarish for editorial assistance.


REFERENCES

  1. Sun, X.-H., and Baltimore, D. (1991) Cell 64, 459-470 [Medline] [Order article via Infotrieve]
  2. Murre, C., McCaw, P. S., and Baltimore, D. (1989) Cell 56, 777-783 [Medline] [Order article via Infotrieve]
  3. Murre, C., McCaw, P. S., Vaessin, H., Caudy, M., Jan, L. Y., Jan, Y. N., Cabrera, C. V., Buskin, J. N., Hauschka, S. D., Lassar, A. B., Weintraub, H., and Baltimore, D. (1989) Cell 58, 537-544 [Medline] [Order article via Infotrieve]
  4. Weintraub, H. (1993) Cell 75, 1241-1244 [Medline] [Order article via Infotrieve]
  5. Guillemot, F., Lo, L.-C., Johnson, J. E., Auerbach, A., Anderson, D. J., and Joyner, A. L. (1993) Cell 75, 463-476 [Medline] [Order article via Infotrieve]
  6. Hsu, H. L., Cheng, J. T., Chen, Q., and Baer, R. (1991) Mol. Cell. Biol. 11, 3037-3042 [Medline] [Order article via Infotrieve]
  7. Benezra, R. (1994) Cell 79, 1057-1067 [Medline] [Order article via Infotrieve]
  8. Murre, C., Voronova, A., and Baltimore, D. (1991) Mol. Cell. Biol. 11, 1156-1160 [Medline] [Order article via Infotrieve]
  9. Bain, G., Robanus Maandag, E. C., Izon, D. J., Amsen, D., Kruisbeek, A. M., Weintraub, B. C., Krop, I., Schlissel, M. S., Feeney, A. J., van Roon, M., van der Valk, M., te Riele, H. P. J., Berns, A., and Murre, C. (1994) Cell 79, 885-892 [Medline] [Order article via Infotrieve]
  10. Zhuang, Y., Soriano, P., and Weintraub, H. (1994) Cell 79, 875-884 [Medline] [Order article via Infotrieve]
  11. Kamps, M. P., Murre, C., Sun, X., and Baltimore, D. (1990) Cell 60, 547-555 [Medline] [Order article via Infotrieve]
  12. Yoshihara, T., Inaba, T., Shapiro, L. H., Kato, J.-Y., and Look, A. T. (1995) Mol. Cell. Biol. 15, 3247-3255 [Abstract]
  13. Peverali, F. A., Ramqvist, T., Saffrich, R., Pepperkok, R., Barone, M. V., and Philipson, L. (1994) EMBO J. 13, 4291-4301 [Abstract]
  14. Rogers, S., Wells, R., and Rechsteiner, M. (1986) Science 234, 364-368 [Medline] [Order article via Infotrieve]
  15. Lee, W. M., Lin, C., and Curran, T. (1988) Mol. Cell. Biol. 8, 5521-5527 [Medline] [Order article via Infotrieve]
  16. Schutte, J., Viallet, J., Nau, M., Segal, S., Fedorko, J., and Minna, J. (1989) Cell 59, 987-997 [Medline] [Order article via Infotrieve]
  17. Amon, A., Irniger, S., and Nasmyth, K. (1994) Cell 77, 1037-1050 [Medline] [Order article via Infotrieve]
  18. Schwob, E., Böhm, T., Mendenhall, M. D., and Nasmyth, K. (1994) Cell 79, 233-244 [Medline] [Order article via Infotrieve]
  19. Hochstrasser, M., Ellison, M. J., Chau, V., and Varshavsky, A. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 4606-4610 [Abstract]
  20. Kornitzer, D., Raboy, B., Kulka, R. G., and Fink, G. R. (1994) EMBO J. 13, 6021-6030 [Abstract]
  21. Tsurumi, C., Ishida, N., Tamura, T., Kakizuka, A., Nishida, E., Okumura, E., Kishimoto, T., Inagaki, M., Okazaki, K., Sagata, N., Ichihara, A., and Tanaka, K. (1995) Mol. Cell. Biol. 15, 5682-5687 [Abstract]
  22. Treier, M., Staszewski, L. M., and Bohmann, D. (1994) Cell 78, 787-798 [Medline] [Order article via Infotrieve]
  23. Glotzer, M., Murray, A. W., and Kirschner, M. W. (1991) Nature 349, 132-138 [CrossRef][Medline] [Order article via Infotrieve]
  24. Pagano, M., Tam, S. W., Theodoras, A. M., Beer-Romero, P., Del Sal, G., Chau, V., Yew, P. R., Draetta, G. F., and Rolfe, M. (1995) Science 269, 682-685 [Medline] [Order article via Infotrieve]
  25. Chen, Z., Hagler, J., Palombella, V. J., Melandri, F., Scherer, D., Ballard, D., and Maniatis, T. (1995) Genes Dev. 9, 1586-1597 [Abstract]
  26. Palombella, V. J., Rando, O. J., Goldberg, A. L., and Maniatis, T. (1994) Cell 78, 773-785 [Medline] [Order article via Infotrieve]
  27. Ciechanover, A. (1994) Cell 79, 13-21 [Medline] [Order article via Infotrieve]
  28. Chau, V., Tobias, J. W., Bachmair, A., Marriott, D., Ecker, D. J., Gonda, D. K., and Varshavsky, A. (1989) Science 243, 1576-1583 [Medline] [Order article via Infotrieve]
  29. Goldberg, A. L. (1995) Science 268, 522-523 [Medline] [Order article via Infotrieve]
  30. Jentsch, S., and Schlenker, S. (1995) Cell 82, 881-884 [Medline] [Order article via Infotrieve]
  31. Ausubel, F. M., Brent, R., Kingston, R., Moore, D., Seidman, J. S., and Struhl, K. (1993) Current Protocols in Molecular Biology, John Wiley & Sons, Inc., New York
  32. Stanton, L. W., Fahrlander, P. D., Tesser, P. M., and Marcu, K. B. (1984) Nature 310, 423-425 [Medline] [Order article via Infotrieve]
  33. Christy, B. A., Sanders, L. K., Lau, L. F., Copeland, N. G., Jenkins, N. A., and Nathans, D. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 1815-1819 [Abstract]
  34. Blackwood, E. M., and Eisenman, R. N. (1991) Science 251, 1211-1217 [Medline] [Order article via Infotrieve]
  35. Sturm, R. A., Das, G., and Herr, W. (1988) Genes Dev. 2, 1582-1599 [Abstract]
  36. Bohmann, D., Bos, T. J., Admon, A., Nishimura, T., Vogt, P. K., and Tjian, R. (1987) Science 238, 1386-1392 [Medline] [Order article via Infotrieve]
  37. Gyuris, J., Golemis, E., Chertkov, H., and Brent, R. (1993) Cell 75, 791-803 [Medline] [Order article via Infotrieve]
  38. Zervos, A. S., Gyuris, J., and Brent, R. (1993) Cell 72, 223-232 [Medline] [Order article via Infotrieve]
  39. Gietz, R. D., St. Jean, A., Woods, R. A., and Schiestel, R. H. (1992) Nucleic Acids Res. 20, 1425 [Medline] [Order article via Infotrieve]
  40. Kaiser, C., Michaelis, S., and Mitchell, A. (1994) Methods in Yeast Genetics, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York
  41. Guarente, L. (1983) Methods Enzymol. 101, 181-191 [Medline] [Order article via Infotrieve]
  42. Smith, D. B., and Johnson, K. S. (1988) Gene (Amst.) 67, 31-40 [CrossRef][Medline] [Order article via Infotrieve]
  43. Shrivastava, A., Saleque, S., Kalpana, G. V., Artandi, S., Goff, S. P., and Calame, K. (1993) Science 262, 1889-1892 [Medline] [Order article via Infotrieve]
  44. Benezra, R. D., Davis, R. L., Lockshon, D., Turner, D. L., and Weintraub, H. (1990) Cell 61, 49-59 [Medline] [Order article via Infotrieve]
  45. Seufert, W., Futcher, B., and Jentsch, S. (1995) Nature 373, 78-81 [CrossRef][Medline] [Order article via Infotrieve]
  46. Al-Khodairy, F., Enoch, T., Hagan, I. M., and Carr, A. M. (1995) J. Cell Sci. 108, 475-486 [Abstract/Free Full Text]
  47. Yasugi, T., and Howley, P. M. (1996) Nucleic Acid Res. 24, 2005-2010 [Abstract/Free Full Text]
  48. Cuomo, C. A., Kirch, S. A., Gyuris, J., Brent, R., and Oettinger, M. A. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 6156-6160 [Abstract]
  49. Scheffner, M., Huibregtse, J. M., and Howley, P. M. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 8797-8801 [Abstract]
  50. Wilkinson, K. D. (1995) Annu. Rev. Nutr. 15, 161-189 [CrossRef][Medline] [Order article via Infotrieve]
  51. Rechsteiner, M. (1990) Semin. Cell Biol. 1, 433-440 [Medline] [Order article via Infotrieve]
  52. Chen, P., Johnson, P., Sommer, T., Jentsch, S., and Hochstrasser, M. (1993) Cell 74, 357-369 [Medline] [Order article via Infotrieve]
  53. Varshavsky, A. (1992) Cell 69, 725-735 [Medline] [Order article via Infotrieve]
  54. Rolfe, M., Beer-Romero, P., Glass, S., Eckstein, J., Berdo, I., Theodoras, A., Pagano, M., and Draetta, G. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 3264-3268 [Abstract]
  55. Goebl, M. G., Goetsch, L., and Byers, B. (1994) Mol. Cell. Biol. 14, 3022-3029 [Abstract]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.