pubmed.ncbi.nlm.nih.gov

Characterization of the mammalian family of DCN-type NEDD8 E3 ligases - PubMed

  • ️Fri Jan 01 2016

. 2016 Apr 1;129(7):1441-54.

doi: 10.1242/jcs.181784. Epub 2016 Feb 18.

Affiliations

Characterization of the mammalian family of DCN-type NEDD8 E3 ligases

Matthew J Keuss et al. J Cell Sci. 2016.

Abstract

Cullin-RING ligases (CRL) are ubiquitin E3 enzymes that bind substrates through variable substrate receptor proteins and are activated by attachment of the ubiquitin-like protein NEDD8 to the cullin subunit. DCNs are NEDD8 E3 ligases that promote neddylation. Mammalian cells express five DCN-like (DCNL) proteins but little is known about their specific functions or interaction partners. We found that DCNLs form stable stoichiometric complexes with CAND1 and cullins that can only be neddylated in the presence of a substrate adaptor. These CAND-cullin-DCNL complexes might represent 'reserve' CRLs that can be rapidly activated when needed. We further found that all DCNLs interact with most cullin subtypes, but that they are probably responsible for the neddylation of different subpopulations of any given cullin. This is consistent with the fact that the subcellular localization of DCNLs in tissue culture cells differs and that they show unique tissue-specific expression patterns in mice. Thus, the specificity between DCNL-type NEDD8 E3 enzymes and their cullin substrates is only apparent in well-defined physiological contexts and related to their subcellular distribution and restricted expression.

Keywords: CRL; Cullin-RING ligases; DCUN1D; NEDD8; SCCRO; Ubiquitin.

© 2016. Published by The Company of Biologists Ltd.

PubMed Disclaimer

Conflict of interest statement

Competing interests

The authors declare no competing or financial interests.

Figures

Fig. 1.
Fig. 1.

Characterization of DCNL interactors. (A) Silver-stained gels of FLAG immunoprecipitations from HEK293 cells stably expressing wild-type (WT) or cullin-binding-deficient DAD patch mutants of all mammalian DCNLs (DCNL1–5). Asterisks denote FLAG–DCNL proteins. WT, but not DAD mutants, specifically co-immunoprecipitate proteins that are not present in the empty FLAG or control, including a large protein with molecular mass >98 kDa (arrow). (B) Heatmap of DCNL interactors identified by mass spectrometry of immunoprecipitations in A. Interactors with >3 unique peptides detected are plotted. All DCNLs co-immunoprecipitate most cullins, as well as cullin-associated proteins RBX1, RBX2 (RNF7), the Cul4A adaptor protein DDB1 and the Cul2/5 adaptor elongin B (TCEB1). The strongest interactor of all DCNLs is CAND1, the CRL substrate adaptor exchange factor. (C) Immunoprecipitation of FLAG–DCNL1, 2, 3 and 4 (top panels) and FLAG-DCNL5 (bottom panels) from stably expressing HEK293 cells followed by western blot analysis. The slower migrating of the two cullin-reactive bands is the neddylated form (asterisk). Only WT DCNLs co-immunoprecipitate mostly non-neddylated cullins, CAND1 and RBX1. Expression levels of FLAG–DCNL4 were too low in this experiment to detect co-immunoprecipitating proteins.

Fig. 2.
Fig. 2.

DCNLs forms a stable complex with cullin and CAND1. (A–E) His6–DCNL1, 2, 3, 4 and 5 were bound to magnetic Ni-NTA beads, incubated for 20 min at 4°C with CAND1 or Cul3–CAND1 complexes and washed with 10 mM imidazole to reduce nonspecific binding. Bound proteins were eluted with 300 mM imidazole and separated by SDS-PAGE followed by Coomassie staining. CAND1, Cul3 and DCNLs form a stoichiometric complex (lane 7) and CAND1 does not bind to any His6–DCNL in the absence of Cul3 (lane 6). (F–J) In vitro neddylation reactions of Cul3 show inhibition by CAND1 and rescue with substrate adaptor KLHL3. Each reaction contained 1 µM Cul3 and 1 µM DCNL and as indicated 1 µM CAND1, 1 µM KLHL3 (1×), or 10 µM KLHL3 (10×). Neddylation was induced by addition to a final concentration of 34 µM NEDD8, 4 µM UBE2M, and 0.2 μM NAE in 50 mM Tris-HCl pH 8 with 0.15 mM ATP, 1.5 mM MgCl2 and 20% glycerol. Reactions were performed for 2 min at 30°C and quenched by the addition of SDS loading buffer. Samples were resolved by SDS-PAGE and processed for immunoblotting with the indicated antibodies. Cul3 is readily neddylated in the absence of CAND1 as seen from the ∼10 kDa band shift in lane 2 (asterisk). Addition of CAND1 inhibits the neddylation reactions, and inhibition is relieved upon the addition of Cul3 substrate adaptor KLHL3.

Fig. 3.
Fig. 3.

Tissue-specific expression and subcellular localization of DCNLs. (A) DCNLs have unique tissue expression profiles and upregulation in cancer cell lines. Western blot analysis of mouse tissue lysates and cancer cell lysates (30 μg each) and where indicated recombinant protein (50 ng). DCNL1 is widely expressed in all tissue samples as well as cancer cell lines. DCNL2 expression is more restricted with highest expression in the brain, liver and kidneys. At least one of the three isoforms of DCNL4 is expressed in most tissue, albeit at varying levels. DCNL3 expression is restricted to testis and brain with weak expression in cancer cell lines. DCNL5 has low expression in testis, skin, thymus and spleen but is drastically upregulated in the three cancer lines. (B) Subcellular localization of DCNLs is independent of cullin binding. U2OS cells expressing N-terminally FLAG tagged DCNL1–5 were analysed by indirect immunofluorescence with an anti-FLAG antibody, rhodamine-conjugated Phalloidin (actin) and DAPI staining (scale bars: 10 µm). DCNL1 and DCNL2 are localized throughout the cell in both the cytoplasm and nucleus. DCNL3 is located throughout the cell, but also at the plasma membrane. DCNL4 and DCNL5 are restricted to the nucleus. Mutation of any of the DCNLs DAD patch domains does not cause a change in subcellular localization, suggesting DCNLs localization is independent of cullin binding and instead determined by their unique N-terminal domain. (C) Diagram of each DCNL family member depicting their conserved C-terminal PONY domain and specific N-terminal domain. Depicted on DCNL3 is a myristoylation site and depicted on DCNL5 is a nuclear localization sequence (NLS). (D) Immunofluorescence analysis of U2OS cells stably expressing either GFP–DCNL5 WT or GFP–DCNL5 with mutations in the NLS (scale bar: 10 µm). Mutation of the NLS sequence results in a delocalization of some DCNL5 into the cytoplasm.

Fig. 4.
Fig. 4.

Effects of DCNL downregulation on cullin neddylation. (A) Knockdown of DCNLs by siRNA has only mild effects on the overall level of individual cullin neddylation. U2OS cells were treated with specific DCNL siRNAs or control siRNA for 72 h, and where indicated the neddylation inhibitor MLN4924 was added for 3 h at 3 μM. Following treatment, cells where harvested and processed for western blotting for the cullin family members to assess the fraction of cullin that was modified by NEDD8. Knockdown of DCNL1 reduces the amount of neddylated Cul1, 4A and 5, whereas knockdown of DCNL2 or DCNL4 had no effect on neddylation levels. Knockdown of DCNL3 increased Cul1 and Cul4A neddylation and DCNL5 knockdown reduced Cul4A and Cul5 neddylation levels. Adjacent graphs show the means±s.e.m. of quantified percentage of neddylated cullins. Statistical significance was determined by one-way ANOVA with Newman–Keuls multiple comparison test. *P≤0.05; n≥3. (B) Schematic of DCNL1 and DCNL5 sequences targeted by CRISPR/Cas9-mediated gene knockout with Protospacer adjacent motif (PAM) indicated (purple) and Cas9 cleavage site marked with an arrow. Depicted below each WT sequence are the mutations (in red) as determined by DNA sequencing. (C) Reduced cullin neddylation in U2OS knockout cells (individual clones) for DCNL1 and DCNL5 generated by CRISPR/Cas9. Whole cell lysates were prepared from WT, DCNL1 KO and DCNL5 KO and processed for SDS-PAGE and immunoblotting for cullin family members to assess the changes in the fraction of neddylated cullins upon loss of DCNL1 or DCNL5. DCNL1 knockout has the strongest effect on Cul4A neddylation and caused mild reduction of Cul3 and Cul5 neddylation. DCNL5 knockout caused reduced neddylation of all cullins tested with the greatest effect on Cul4A, Cul4B and Cul2. Adjacent graphs show the means±s.e.m. of quantified percentage of neddylated cullins. Statistical significance was determined by one-way ANOVA with Newman–Keuls multiple comparison test. *P≤0.05, **P≤0.01, ***P≤0.001; n≥3. Asterisks denote the slower migrating neddylated form of the two cullin-reactive bands.

Fig. 5.
Fig. 5.

DCNL1 and DCNL5 independently contribute to Cul4A neddylation. (A) Cul4A neddylation partly depends on DCNL1 expression. Western blot analysis of lysates from U2OS cell lines stably expressing siRNA-resistant transgenes for FLAG–empty, FLAG–DCNL1 (WT), or FLAG–DCNL1 (DAD mutant). Cells were treated with control or DCNL1 siRNA for 72 h. Transgene expression was induced by doxycycline (0.5 ng/ml) for 24 h. Where indicated, MLN4924 was added for 3 h at 3 μM. Cell lysates were resolved by SDS-PAGE and immunoblotted with anti-Cul4A antibody to assess the fraction of neddylated cullin. siRNA depletion of DCNL1 reduces Cul4A neddylation, which can be rescued by re-expression of DCNL1 (WT) but not cullin-binding-deficient DAD mutant. (B) Cul4A neddylation partly depends on DCNL5. Whole cell lysates were prepared as in A but cell lines stably expressed FLAG–empty, FLAG–DCNL5 (WT) or FLAG–DCNL5 (DAD mutant) and siRNA was targeted against DCNL5 or control. siRNA-mediated depletion of DCNL5 reduces the fraction of Cul4A that is neddylated. Upon re-expression of WT DCNL5, Cul4A neddylation is rescued to wild-type levels. The cullin-binding-deficient DAD patch mutant of DCNL5 fails to rescue Cul4A neddylation, suggesting DCNL5 directly interacts with Cul4A to promote neddylation. (C) Cul4A neddylation in DCNL1-KO cells directly depends on DCNL1 expression. Cul4A neddylation was assessed by western blot analysis of lysates from U2OS DCNL1 KO cell lines that stably express Flp-In-generated GFP, GFP–DCNL1 (WT) or GFP–DCNL1 (DAD mutant). Similar to the Cul4A neddylation reduction by siRNA depletion of DCNL1, knockout of DCNL1 can only be rescued by re-expression of WT DCNL1, but not DAD patch mutant DCNL1. (D) Cul4A neddylation in DCNL5-KO cells directly depends on DCNL5 expression. Same as in C except the U2OS knockout cell line was DCNL5 and rescue lines were Flp-In for GFP, GFP–DCNL5 (WT) or GFP–DCNL5 (DAD mutant). Cul4A neddylation is rescued by re-expression of WT but not DAD patch mutant DCNL5. (E) DCNL5 depletion has an additive effect on Cul4A neddylation in DCNL1 KO-cells. The fraction of neddylated Cul4A was determined by western blot analysis of cell lysates from WT or DCNL1 KO treated with control siRNA or siRNA against DCNL5. Additional depletion of DCNL5 in DCNL1-KO cells further reduces the amount of neddylated Cul4A. This additive effect of DCNL5 depletion suggests there are separate pools of Cul4A that are independently neddylated by DCNL1 or DCNL5. (F) DCNL1 depletion has an additive effect on Cul4A neddylation in DCNL5-KO cells. Same as in E, but the cell lines were WT or DCNL5 KO and treatment was with control siRNA or siRNA against DCNL1. The additional depletion of DCNL1 in DCNL5-KO cells mirrors the result in E and further supports that model that separate pools of Cul4A are independently regulated by DCNL1 or DCNL5. All graphs plot the means±s.e.m. of quantified percentage of neddylated Cul4A. Statistical significance was determined by one-way ANOVA with Newman–Keuls multiple comparison test. **P≤0.01, ***P≤0.001; n≥3. Asterisks denote the slower migrating neddylated form of the two cullin-reactive bands.

Fig. 6.
Fig. 6.

DCNL5 but not DCNL1 is involved in the DNA damage response. (A) Clonogenic survival analysis of U2OS cells (WT, DCNL1 KO, DCNL5 KO and DCNL5 KO re-expressing WT DCNL5) carried out after UV exposure. 500 cells/well were plated in six-well plates and 24 h later exposed to indicated amounts of UV radiation. Cells were allowed to grow for 14 days and colony formation was assessed by Crystal Violet staining. Cells able to form colonies are interpreted as having repaired the UV-induced damage. DCNL5-KO cells have an impaired ability to form colonies after UV exposure, suggesting defects in DNA repair mechanisms. DCNL1-KO cells are not sensitive to DNA damage. Each datapoint is the mean±s.e.m. of three experiments undertaken in at least three replicates. (B) DCNL5 is recruited to sites of DNA damage. Immunofluorescence of U2OS cells treated with BrdU (10 µM, 24 h; upper panels) or mono-psoralen (25 µM, 3 h; lower panels). Cells were fixed 5–10 min after laser micro-irradiation and stained for endogenous DCNL5 or Cul4A, phosphorylated γ-H2AX and DAPI. The histone variant γ-H2AX, a marker of DNA damage, serves as a positive control. DCNL5 and Cul4A are also recruited to the site of DNA damage. (C) DCNL5 is recruited to DNA damage sites independently of cullin binding. Live-cell analysis of U2OS cells stably expressing GFP, GFP–DCNL5 (WT) or GFP–DCNL5 (DAD mutant). Cells were BrdU-treated (10 µM, 24 h) and images captured at indicated times after micro-irradiation with a 355 nm UV laser. Both WT DCNL5 and mutant DCNL5 are recruited to DNA damage sites. (D) DCNL recruitment to sites of DNA damage. U2OS cells were transfected with N-terminally GFP-tagged DCNLs. 24 h after transfection, cells were treated as in C. All slides were treated identically and images were taken with the same microscope with identical settings. All DCNLs are readily recruited to sites of DNA damage with the exception of DCNL3, which displays only very weak recruitment. (E) U2OS cells transiently transfected with FLAG–DCNL3 or chimeric FLAG–DCNL3 where the N-terminal domain (aa 1–85) was replaced with the N-terminus domain of DCNL5 (aa 1–46). Cells were treated as in C and D. Replacement of the N-terminus readily targets DCNL3 to sites of DNA damage. Scale bars: 10 µm.

Similar articles

Cited by

References

    1. Bommelje C. C., Weeda V. B., Huang G., Shah K., Bains S., Buss E., Shaha M., Gönen M., Ghossein R., Ramanathan S. Y. et al. (2014). Oncogenic function of SCCRO5/DCUN1D5 requires its Neddylation E3 activity and nuclear localization. Clin. Cancer Res. 20, 372-381. 10.1158/1078-0432.CCR-13-1252 - DOI - PMC - PubMed
    1. Bornstein G., Ganoth D. and Hershko A. (2006). Regulation of neddylation and deneddylation of cullin1 in SCFSkp2 ubiquitin ligase by F-box protein and substrate. Proc. Natl. Acad. Sci. USA 103, 11515-11520. 10.1073/pnas.0603921103 - DOI - PMC - PubMed
    1. Broderick S. R., Golas B. J., Pham D., Towe C. W., Talbot S. G., Kaufman A., Bains S., Huryn L. A., Yonekawa Y., Carlson D. et al. (2010). SCCRO promotes glioma formation and malignant progression in mice. Neoplasia 12, 476-484. 10.1593/neo.10202 - DOI - PMC - PubMed
    1. Brown J. S., Lukashchuk N., Sczaniecka-Clift M., Britton S., le Sage C., Calsou P., Beli P., Galanty Y. and Jackson S. P. (2015). Neddylation promotes ubiquitylation and release of Ku from DNA-damage sites. Cell Rep. 11, 704-714. 10.1016/j.celrep.2015.03.058 - DOI - PMC - PubMed
    1. Chen X., Zhang Y., Douglas L. and Zhou P. (2001). UV-damaged DNA-binding proteins are targets of CUL-4A-mediated ubiquitination and degradation. J. Biol. Chem. 276, 48175-48182. - PubMed

Publication types

MeSH terms

Substances