Journal of Molecular Biology
Volume 385, Issue 2, 16 January 2009, Pages 507-519
Journal home page for Journal of Molecular Biology

E2–c-Cbl Recognition Is Necessary but not Sufficient for Ubiquitination Activity

https://doi.org/10.1016/j.jmb.2008.10.044Get rights and content

Abstract

The E2 ubiquitin-conjugating enzymes UbcH7 and UbcH5B both show specific binding to the RING (really interesting new gene) domain of the E3 ubiquitin-protein ligase c-Cbl, but UbcH7 hardly supports ubiquitination of c-Cbl and substrate in a reconstituted system. Here, we found that neither structural changes nor subtle differences in the E2–E3 interaction surface are possible explanations for the functional specificity of UbcH5B and UbcH7 in their interaction with c-Cbl. The quick transfer of ubiquitin from the UbcH5B∼Ub thioester to c-Cbl or other ubiquitin acceptors suggests that UbcH5B might functionally be a relatively pliable E2 enzyme. In contrast, the UbcH7∼Ub thioester is too stable to transfer ubiquitin under our assay conditions, indicating that UbcH7 might be a more specific E2 enzyme. Our results imply that the interaction specificity between c-Cbl and E2 is required but not sufficient for transfer of ubiquitin to potential targets.

Introduction

Posttranslational modification by ubiquitin is a major mechanism for regulating protein function in eukaryotes. The general enzymatic cascade of this pathway comprises three enzymes known as E1 (ubiquitin-activating enzyme), E2 (ubiquitin-conjugating enzyme), and E3 (ubiquitin-protein ligase enzyme). The first step in this cascade is the ATP-dependent formation of a thioester bond between ubiquitin and E1, after which ubiquitin is transferred to the catalytic cysteine of an E2 enzyme forming an E2∼Ub thioester. Next, the E3 catalyzes the transfer of the ubiquitin from the E2 to a lysine residue of the substrate, with which it forms an isopeptide bond. While HECT (homologous to the E6AP C-terminus) domain E3 ligases mediate this step by formation of a HECT-ubiquitin thioester intermediate, RING (really interesting new gene) domain E3 ligases facilitate a direct transfer of ubiquitin from E2 to the substrate via noncovalent interactions with the E2∼Ub and the substate.1 In yeast, the substrate can be tagged by a single ubiquitin or ubiquitin chains, which can vary in length and linkage specificity. The functions of chains linked through K48 and K63 of ubiquitin have been established. The K48-linked chains are shown to target proteins for degradation via the 26S proteasome, while K63-linked chains recruit binding partners during inflammation or DNA repair.2

The E2 enzymes are the key enzymes in the ubiquitin and ubiquitin-like pathways. More than 30 structures of E2s from various organisms are available, and all these E2 proteins share a topologically conserved α/β-fold core domain of ∼ 150 residues. Four α-helices (α1–α4) compose one face of the protein, and a four-stranded antiparallel β-sheet (β1–β4) sits on the back of the enzyme between helices α1 and α2 (Fig. 4). Every E2 core domain can bind three other proteins: the E1 ubiquitin-activating enzyme, a cognate E3 ligase enzyme, and activated ubiquitin via a labile thioester linkage. E2 enzymes can gain additional functionality via association with additional domains or subunits.3, 4

The complex of c-Cbl and UbcH7 was the first RING E3–E2 structure to be solved,5 but further structural insight into E2–E3 interactions comes from the HECT E3 E6AP complexed to UbcH7,6 the NMR-based model of the RING E3 CNOT4 bound to UbcH5B,7 and the U-box domain E3 CHIP/Ubc13–Uev1a complex.1, 8 All these complex structures show equivalent elements in the E2–E3 interface and similarities in overall structure. The core interface is formed by a hydrophobic groove on the E3 and two main regions on E2: loop L1, located between strands β3 and β4, and loop L2, connecting strand β4 and helix α2. In sharp contrast to the wealth of knowledge on E2 structures, the best structural insight of the E2∼Ub thioester intermediate is limited to models based on NMR chemical shift perturbation data due to the inherent instability of the thioester complex.4, 9

The human genome encodes a few E1s,1, 10 around 30 E2s, several hundreds of E3s, and thousands of substrates for ubiquitin, which makes the ubiquitination pathway very versatile and highly complex. An E3 may have more than one substrate and some substrates can be recognized by multiple E3s.11 The large abundance of E3s relative to E2s implies that E2s can function with multiple E3 ligases, but the inverse situation also exists. For example, the RING E3 complex SCF shows in vitro polyubiquitination activity with both CDC34 and Ubc4,12, 13 and the activity of the E3 APC is observed with both UbcH10 and UbcH5A, C.14, 15

It is generally assumed that the binding ability of E2–E3 is the determinant of functional E2–E3 pairs, because decreasing the interaction between E2 and E3 also diminishes their poly- or monoubiquitination activity in vitro. Interestingly, Brzovic et al. found that the BRCA1/BARD1 E3 heterodimer can interact with UbcH5C and UbcH7 with similar affinity and a similar interface, but only UbcH5C was active in Ub-ligase activity assays.16 Other examples of E3s that can physically interact with some E2s but fail to support ubiquitin transfer to targets include the RING E3 heterodimer Ring1b/Bmr117 and the U-box E3 CHIP.8 These results suggest that binding between E2–E3 pairs does not suffice for function. To address this paradox, we have analyzed and compared the c-Cbl/UbcH5B and c-Cbl/UbcH7 complexes both at a structural and functional level.

The E3 ubiquitin ligases of the Cbl family are key regulators of signaling by many surface receptors. From the N- to the C-terminus, c-Cbl contains a tyrosine kinase binding (TKB) domain, a linker region, and a RING domain, followed by an extensive proline-rich region and a ubiquitin-associated (UBA) domain. The UBA domain of Cbl-b, a homolog of c-Cbl, has recently been shown to bind ubiquitin noncovalently and to promote Cbl-b UBA dimerization and is required for RTK ubiquitination in vivo.18 The RING domain can recruit the E2 and function as an E3 ligase.19 The c-Cbl RING E3 ligase can regulate receptor kinases via their ubiquitin conjugation and catalyze the ubiquitination of c-Cbl itself as well. This autoubiquitination of E3 enzymes is suggested to represent a regulatory mechanism to control the abundance of Ub protein ligases in cells, since E3 autoubiquitination mediates their proteasome-dependent degradation.20, 21 Here, we studied the binding and functional specificity of UbcH5B and UbcH7 to c-Cbl RING using NMR structural analysis, chimeric UbcH5B/UbcH7 proteins, and in vitro activity assays, mostly by studying autoubiquitination of the c-Cbl RING domain. We found that although both UbcH5B and UbcH7 can bind specifically to the c-Cbl RING domain, only UbcH5B can facilitate ubiquitination of c-Cbl and substrate, which suggests that the binding specificity between E2–E3 pairs does not necessarily suffice for their functional specificity.

Section snippets

Ub-charged UbcH7 cannot mediate the autoubiquitination of c-Cbl

Early studies indicated that c-Cbl can ubiquitinate epidermal growth factor receptor and Src in cooperation with UbcH7,22, 23 while more recent work showed that c-Cbl promoted epidermal growth factor receptor and Src ubiquitination in a UbcH5B-dependent fashion.24, 25 We tested if c-Cbl can form functional complexes with UbcH7 and/or UbcH5B using an in vitro autoubiquitination assay, which is a characteristic commonly used to identify functional E2–E3 pairs.26 To enable accurate quantitation of

Discussion

In our experiments, we found that UbcH5B can support the in vitro autoubiquitination of c-Cbl and the c-Cbl-facilitated substrate ubiquitination but UbcH7 cannot. NMR titrations of c-Cbl to UbcH5B and UbcH7 did not show structural distinctions that could explain the functional differences. More likely, the different reactivity of the thioesters UbcH5B∼Ub and UbcH7∼Ub might explain why c-Cbl/UbcH7 is almost “nonproductive” in vitro. The E2∼Ub thioester intermediate is a key intermediate in the

Construction of plasmids

The construct of wild-type human UbcH5B was amplified from a human cDNA library and cloned into the plasmid pLICHIS, a pET15B-derived expression vector, by enzyme-free cloning.34 The chimera 7H1, 7L1, and 7L2 (Table 2) plasmids were amplified from the plasmid pLICHIS-UbcH5B, and further details are provided as Supplementary material. The fragment encoding the c-Cbl linker and RING domain (358–437) was cloned from a human cDNA library into pLISHISGST vector.34 Correct constructions of these

Acknowledgements

We would like to thank Dr. E. AB for help with structure calculations, Dr. Tammo Diercks and Dr. Rainer Wechselberger for expert NMR assistance, and Dr. Gert E. Folkers, M. Hilbers, and J. van der Zwan for laboratory and technical support. This work was supported by grants from the Netherlands Organization for Scientific Research (NWO-CW TOP 700-52-303 and 700-53-103) and by European Commission funding through the SPINE2-COMPLEXES project LSHG-CT-2006-031220. G.S.W. and H.Th.M.T. were supported

References (43)

  • YokouchiM. et al.

    Src-catalyzed phosphorylation of c-Cbl leads to the interdependent ubiquitination of both proteins

    J. Biol. Chem.

    (2001)
  • LevkowitzG. et al.

    Ubiquitin ligase activity and tyrosine phosphorylation underlie suppression of growth factor signaling by c-Cbl/Sli-1

    Mol. Cell

    (1999)
  • GeethaT. et al.

    Lysine 63 polyubiquitination of the nerve growth factor receptor TrkA directs internalization and signaling

    Mol. Cell

    (2005)
  • WinklerG.S. et al.

    An altered-specificity ubiquitin-conjugating enzyme/ubiquitin-protein ligase pair

    J. Mol. Biol.

    (2004)
  • EletrZ.M. et al.

    Sequence determinants of E2-E6AP binding affinity and specificity

    J. Mol. Biol.

    (2007)
  • SattlerM. et al.

    Heteronuclear multidimensional NMR experiments for the structure determination of proteins in solution employing pulsed field gradients

    Prog. Nucl. Magn. Reson. Spectrosc.

    (1999)
  • HerrmannT. et al.

    Protein NMR structure determination with automated NOE assignment using the new software CANDID and the torsion angle dynamics algorithm DYANA

    J. Mol. Biol.

    (2002)
  • HoubenK. et al.

    Solution structure of the ubiquitin-conjugating enzyme UbcH5B

    J. Mol. Biol.

    (2004)
  • PickartC.M.

    Mechanisms underlying ubiquitination

    Annu. Rev. Biochem.

    (2001)
  • KerscherO. et al.

    Modification of proteins by ubiquitin and ubiquitin-like proteins

    Annu. Rev. Cell Dev. Biol.

    (2006)
  • AndersenP.L. et al.

    Distinct regulation of Ubc13 functions by the two ubiquitin-conjugating enzyme variants Mms2 and Uev1A

    J. Cell Biol.

    (2005)
  • Cited by (37)

    • In Vitro Ubiquitination Platform Identifies Methyl Ellipticiniums as Ubiquitin Ligase Inhibitors

      2021, SLAS Discovery
      Citation Excerpt :

      In our autoubiquitination assay, these compounds do not inhibit CBLB E3 activity (Suppl. Fig. S7). UbcH7 has been crystallized in complex with CBL but does not function with CBL proteins in E3 assays, possibly explaining their lack of activity in our hands.13,54–57 In addition, a pharmaceutical company has described a CBLB inhibitor in a conference abstract in which no chemical or screening assay information was provided.45

    • RINGs hold the key to ubiquitin transfer

      2012, Trends in Biochemical Sciences
      Citation Excerpt :

      In addition, the complexes formed between the RING domain of CBL and either Ube2D2 or Ube2L3 (two E2 enzymes) are comparable, although only Ube2D2 allows ubiquitin transfer. This suggests that no major structural change in the E2 protein is associated with activity [19,23], and the interaction of an E2 enzyme and an E3 ligase, although necessary, is insufficient to predict activity. To elicit ubiquitin transfer, the E3 protein must interact with an E2∼ubiquitin conjugate, yet most studies have focused on the interaction between RING or U-box domains and E2 enzymes in the absence of ubiquitin.

    View all citing articles on Scopus
    1

    Present address: R. N. de Jong, Genmab B.V., Yalelaan 60, 3584 CM Utrecht, The Netherlands.

    2

    Present address: G. S. Winkler, School of Pharmacy, University of Nottingham, University Park, NG7 2RD Nottingham, UK.

    View full text