UBE4B, a ubiquitin chain assembly factor, is required for MDM2-mediated p53 polyubiquitination and degradation

Abstract

Although MDM2 is known to be a critical negative regulator of p53, MDM2 only catalyzes p53 mono- or multiple monoubiquitination in vitro and in vivo, which is insufficient for the initiation of proteasomal degradation. MDM2 does not polyubiquitinate p53 in vitro, however, which indicates that the activity of other ubiquitin ligase(s) or cofactor(s) is required for MDM2-mediated p53 polyubiquitination and degradation. In our recent study, we demonstrated that UBE4B, an E3 and E4 ubiquitin ligase with a U-box domain, interacts physically with both p53 and MDM2. Our findings revealed that UBE4B negatively regulates the level of p53 and inhibits p53-dependent transactivation and apoptosis. We propose that inhibition of MDM2 binding to UBE4B may provide another approach to inhibit MDM2 E3 ligase activity for tumor suppressor p53. It could lead to novel anticancer therapies, with the possibility of reducing the public health burden from cancer.

The Tumor Suppressor p53 and Tumors

The p53 tumor suppressor is inactivated in more than 50% of all human tumors, and mutation of p53 is the most frequently observed genetic event in cancer cells.^1–3 The importance of p53 in safeguarding genomes has been clearly demonstrated by genetic and clinic studies. Mice deficient in p53 are highly susceptible to tumors: more than 75% of p53^−/− mice developed tumors in six months.⁴ Germline mutations in p53 result in Li-Fraumeni syndrome, a hereditary susceptibility to cancers of breast, brain and adrenal glands and to leukemias and sarcomas of bone and connective tissues.^2,5 The p53 gene regulates the cell cycle, the initiation of apoptotic cell death and DNA repair.^6–10 p53 mutation is associated with more aggressive disease and worse overall survival.^8–10 Inactivation of p53 correlates with poor prognosis and drug resistance in malignant tumors.^2,3 This suggests that loss of p53 plays an important role in the pathogenesis of cancer, and that the regulation of p53 expression and stability are essential for maintaining normal cell growth. Tight regulation of p53 expression is indeed essential for maintaining normal cell growth. It is highly likely that many upstream regulators and downstream effectors of p53 play a role in carcinogenesis in humans and other species.

Ubiquitin E3 Ligases and Disease

Ubiquitin (Ub) is an abundant and essential protein that mediates targeted protein degradation in eukaryotes.^11,12 Ubiquitin-mediated protein degradation is a three-step process involving three enzymes: E1 (Ub-activating enzyme), E2 (Ub-conjugating enzyme) and E3 (Ub protein ligase). A new class of ubiquitination enzyme, E4 (a Ub chain assembly factor), was recently shown to be necessary for the degradation of some proteins via the ubiquitin fusion degradation (UFD) pathway.¹³ Together, these enzymes catalyze the covalent attachment of one or more ubiquitin moieties to protein lysine residues. E3 enzymes play a central role in the recognition and ubiquitination of substrates¹¹ and fall into three subgroups: (1) HECT domain ligases such as E6-AP;¹⁴ (2) RING domain ligases,¹⁵ including Pirh2 and Mdm2 and (3) U-box domain ligases representing a non-canonical RING domain such as UFD2 and CHIP (carboxyl terminus of Hsp70-interacting protein).¹⁶ Defects in Ub-protein ligases or their substrates can give rise to human disease. Human conditions attributed to defective Ub-protein ligases include cancer, immune diseases, Angelman syndrome and Parkinson disease. Defective Ub substrates can lead to Alzheimer disease, cystic fibrosis, Liddle syndrome and diabetes mellitus.^17,18

Mdm2 and p53

The Mdm2 gene was originally identified through its amplification in a spontaneously transformed derivative of the mouse BALB/c cells.¹⁹ Mdm2 (known as MDM2 or Hdm2 in humans) binds to the p53 transactivation domain^20,21 and promotes p53 degradation via a ubiquitin-proteasome pathway.^22,23 Deletion of Mdm2 in the mouse results in a lethal phenotype in the embryo.^24,25 However, this early embryonic death of Mdm2-null mice is rescued by further deleting the p53 gene, thus indicating the importance of the negative regulatory function of Mdm2 on p53 during development.^24,25 Moreover, the Mdm2 gene itself is transcriptionally activated by p53,²⁶ forming an autoregulatory feedback loop that restrains p53 activity.²⁸ Mdm2 mediates mono- or multiple-monoubiquitination of p53,^28–30 but only polyubiquitin chains are efficiently recognized by the proteasome.³¹ Mdm2 does not polyubiquitinate p53, which suggests that the activity of other ubiquitin ligases is required for Mdm2-mediated p53 degradation.^28–30 The mechanisms of Mdm2-mediated p53 degradation are not fully understood.

Current Knowledge about UBE4B

The human UBE4B is a mammalian homolog of the protein UFD2 found in S. cerevisiae. Yeast UFD2 is involved in the UFD pathway and is encoded by a single-copy gene.³² It is not essential for viability, but it is linked to the survival of the cell under stress conditions.³³ Yeast UFD2 is required for a novel enzymatic activity in ubiquitin chain assembly and was the first known E4 ubiquitination factor.¹³ Single UFD2-homologs have been found in S. cerevisiae, C. elegans, D. discoideumand and A. thaliana. Two genes related to UFD2 have been found in Mus musculus (named Ube4a and Ube4b or Ufd2a) and Homo sapiens (named UBE4A and UBE4B).^34,35 NOSA, the UFD2 homolog from D. discoideum, is essential for cellular differentiation.³⁵ UBE4B and its homologs share a conserved domain of about 70 amino acids that have been named the U box. The E4 activity of UBE4B is now known to be a specialized type of E3 activity, and it has been reported that UBE4B functions as an E4 enzyme in the degradation of pathological forms of ataxin-3.³⁶ UBE4B also interacts with fasciculation and elongation protein ζ-1 (FEZ1) and mediates FEZ1 ubiquitination but does not affect its intracellular stability.³⁷ Ube4b has recently been reported to interact with p73: overexpression of Ube4b decreased p73 protein levels and inhibited p73-mediated transactivation. However, Ube4b did not promote p73 ubiquitination.³⁸ Ube4b is expressed predominantly in the neuronal tissues of adult mice. Unlike yeast UFD2, the deletion of Ube4b in the mouse results in very early embryonic lethality because of marked apoptosis.³⁹ Polyubiquitination activity for the E4 substrate is greatly reduced in Ube4b^−/− mouse embryonic fibroblasts (MEFs).³⁹ Ube4b^+/− mice develop a neurological disorder, manifesting axonal dystrophy in the nucleus gracilis as well degeneration of Purkinje cells accompanied by endoplasmic reticulum (ER) stress.³⁹ However, the underlying mechanisms for this lethality remain unclear.

CBP/p300, YY1 and UBE4B

The protein p300 is a transcriptional co-activator and histone acetyltransferase that functions in chromatin remodeling during transcription.^40–42 Several groups have demonstrated that p53 binds p300, and is acetylated by p300.^41,42 Recently, p300 was reported as a E4 ligase that mediates p53 polyubiquitination.²⁹ CBP, like p300, encoded an E3 activity within its N terminus.⁴³ Although CBP is required for p53 polyubiquitination in vivo, CBP shows an Mdm2-independent, p53-directed E4 ligase function.⁴³ Yin Yang (YY1) was described as a molecular clamp to mediate p53 polyubiquitination.⁴⁴ In contrast, UBE4B is a U-box-containing E3/E4 ubiquitin ligase.⁴⁵ We showed that UBE4B's E4 ligase activity for p53 is mainly dependent on MDM2.⁴⁵ CBP/p300 are E4 ligases and only act on prior monoubiquitinated substrates;⁴³ however, UBE4B requires functional MDM2 to degrade p53. To ascertain whether UBE4B can affect p53 ubiquitination mediated by MDM2 in vivo, H1299 cells were co-transfected with p53 and increased amounts of MDM2 expression plasmid then compared to those of similar cells co-transfected with a fixed amount of MDM2 expression plasmid with increased amounts of UBE4B-expressing plasmid along with HA-tagged ubiquitin. Interestingly, we found (1) the pattern of p53 ubiquitination was changed when cells co-expressed both MDM2 and UBE4B compared to the cells that expressed only MDM2 (Fig. 1, upper image) and (2) the level of p53 was markedly decreased in cells that co-expressed UBE4B and MDM2 compared with those that expressed MDM2 alone (Fig. 1, lower image). We further investigated whether Ube4b is required for Mdm2-mediated p53 degradation in vivo. Consistently, we observed that the p53 level was greatly decreased in the presence of transfected Ube4b, and to a lesser extent in the presence of Ube4b and Ube4bΔU or in the presence of Ube4b when Mdm2 was eliminated in Ube4b-null MEFs, indicating that Ube4bΔU may act as a dominant negative to interfere with Ube4b (Fig. 2). Therefore, these findings indicated that Ube4b is required for Mdm2-mediated p53 degradation in vivo.

Figure 1.

UBE4B changed the pattern of p53 ubiquitination mediated by MDM2 in vivo. H1299 cells were transfected with p53, HA-tagged ubiquitin or in combination with increased amounts of MDM2, or with fixed amount of MDM2 and increased amounts of UBE4B as indicated and immunoprecipitated with anti-p53 (DO-1) and analyzed by immunoblotting with anti-HA to detect ubiquitinated p53 or with anti-p53 (DO-1) to detect total p53.

Figure 2.

UBE4B is required for Mdm2-mediated p53 degradation in vivo. Ube4b-null MEFs were transfected with the indicated plasmids and analyzed by western blot using anti-Flag for Ube4b and Ube4bΔU, anti-Mdm2 (MD-219) or anti-p53 (CM5) antibodies, as indicated.

Notably, Ube4b-knockout mice are embryonic lethal, suggesting that Ube4b, CBP/p300 and YY1 play nonoverlapping roles in regulating p53 stability in vivo. An important question raised by our results is whether the deletion of p53 can rescue the early embryonic lethality observed in Ube4b-deficient mice. We believe that several possibilities exist. First, it is possible that the embryonic lethality of Ube4b^−/− mice may be overcome in the absence of p53, because the deletion of Ube4b in mice causes marked apoptosis that results in lethality.³⁹ Importantly, we show that the basal level of endogenous p53 is elevated in Ube4b-null MEFs. Second, it is also possible that the absence of p53 would only partially, or not at all, rescue Ube4b-null embryonic lethality because of redundancy in the genome. p63 and p73 are required for p53-dependent apoptosis in response to DNA damage,^46,47 and Ube4b binds to p73 and promotes its degradation.³⁸ Therefore, it is possible that p73 or p63, both in the p53 family, may compensate for the loss of p53. If apoptosis is, in fact, the cause of lethality in the Ube4b-null embryos, it will be important to examine whether these embryos are still dying as a result of apoptosis in the absence of p53. Therefore, to overcome the embryonic lethal phenotype and to analyze the role of Ube4b in embryonic development, Ube4b conditional knockout mice are being generated in our laboratory using the Cre-loxP system. Overall, we have identified UBE4B as an E4 that is essential and required for MDM2 to promote p53 polyubiquitination and degradation. It is conceivable that the discovery of UBE4B will be a great help to understand the function of MDM2 and will have significant implications in cancer therapy. Inhibition of the UBE4B-MDM2 interaction will stabilize p53 protein, thus these findings have significant implications for cancer therapy.