Comparison of substrate specificity of the ubiquitin ligases Nedd4 and Nedd4-2 using proteome arrays

Abstract

Target recognition by the ubiquitin system is mediated by E3 ubiquitin ligases. Nedd4 family members are E3 ligases comprised of a C2 domain, 2-4 WW domains that bind PY motifs (L/PPxY) and a ubiquitin ligase HECT domain. The nine Nedd4 family proteins in mammals include two close relatives: Nedd4 (Nedd4-1) and Nedd4L (Nedd4-2), but their global substrate recognition or differences in substrate specificity are unknown. We performed in vitro ubiquitylation and binding assays of human Nedd4-1 and Nedd4-2, and rat-Nedd4-1, using protein microarrays spotted with approximately 8200 human proteins. Top hits (substrates) for the ubiquitylation and binding assays mostly contain PY motifs. Although several substrates were recognized by both Nedd4-1 and Nedd4-2, others were specific to only one, with several Tyr kinases preferred by Nedd4-1 and some ion channels by Nedd4-2; this was subsequently validated in vivo. Accordingly, Nedd4-1 knockdown or knockout in cells led to sustained signalling via some of its substrate Tyr kinases (e.g. FGFR), suggesting Nedd4-1 suppresses their signalling. These results demonstrate the feasibility of identifying substrates and deciphering substrate specificity of mammalian E3 ligases.

Figure 1

Nedd4 proteins. (A) Schematic representation of rNedd4-1, hNedd4-1 and hNedd4-2 (not to scale), with % amino-acid identity between them depicted at the right hand side. The presence of WW3 domain in hNedd4-1 and hNedd4-2, but not rNedd4-1, should be noted. Additional unique regions are shown as blue and green thick lines. (B) Coomassie staining of purified Nedd4 proteins used for the protoarray screen.

Figure 2

Control protein arrays tested for ubiquitylation and binding of known substrates of Nedd4 proteins. (A) Decreasing concentration of the indicated proteins (1:5, 1:10, 1:50, i.e. 0.8–0.016 ng) were spotted multiple times on control PATH slides and the slides incubated with the indicated Nedd4 proteins (E3), E2 (UbcH5b), E1, FITC–Ub and Mg–ATP to test for substrate ubiquitylation (see Materials and methods section). An example of one sub-array incubated with hNedd4-1 is enlarged. (B) Parallel control protein arrays incubated with Alexa647–Nedd4 proteins to test for their binding to the spotted test proteins. Positive control proteins (all containing PY motifs and previously shown to be ubiquitylated by Nedd4 proteins): (1) a region within βENaC C terminus; (2) Rnf11; (3) the C terminus of LAPTM5; (4) the Xenopus xNedd4-HECT domain; (5) The C terminus of CNrasGEF. Negative controls (which lack PY motifs): (6) GST; (7) the C2 domain of rNedd4-1; (8) RNF11 with mutated PY motif; (9) the SH2 domain of Grb10 (which was previously shown to bind the C2 domain of rNedd4-1 but not to be ubiquitylated by it (Morrione et al, 1999)); (10) the PDZ domain of CNrasGEF.

Figure 3

Nedd4 proteins-mediated ubiquitylation of the human proteome. (A) Ubiquitylation of proteins spotted on the human protoarray by hNedd4-1, with an enlargement of one sub-array from this array/slide and from a parallel array/slide. Arrows point to several hits, which appear to be identical between the two parallel arrays/slides. Similar analysis of ubiquitylation of the protoarrays by rNedd4-1 and hNedd4-2 is depicted in Supplementary Figure S1. (B) As in (A), analysis of binding of hNedd4-1 to proteins spotted on the human protoarray. Similar analysis of binding by rNedd4-1 and hNedd4-2 is depicted in Supplementary Figure S2.

Figure 4

Sequence logos for the PY motif region. Sequence logos depicting the region around the L/PPxY motifs in hits containing one such motif from screens for ubiquitylation substrates of hNedd4-1, hNedd4-2 and rNedd4-1. Red arrows indicate the border between the Pro and Leu in the first position of the L/PPxY motif, and show that a Leu in this position is relatively more abundant in substrates for hNedd4-1 and hNedd4-2 than in substrates for rNedd4-1. The number of total sequences (n) used to generate the logo is indicated in each image.

Figure 5

Validation of hits common to the Nedd4 proteins. HEK293T cells were transfected (or not) with the indicated cDNA for hit protein (Flag tagged) alone or with co-transfected Nedd4 protein (either WT or a catalytically inactive CS mutant). The Nedd4 protein was N-terminally tagged with either T7 (rNedd4-1) or V5 (hNedd4-1 and hNedd4-2). After transfections, half of the lysate was boiled in SDS, SDS diluted out, the hit protein immunoprecipitated and immunoblotted with anti-ubiquitin (Ub) antibodies to test for substrates ubiquitylation (upper panels). The other half of the lysate (not boiled) was used for co-IP analysis to determine interactions between the hit protein and the indicated Nedd4s (middle panels). Lower panels are controls for protein loading. (A–C) Example of a hit protein, Yang-Yang1 (YY1), that was common to all three Nedd4s tested, as indicated. (D–F) Negative controls: examples of proteins that were not hits in the screens: (D) carbonic anhydrase 9 (CA9); (E) keratin 20 (KRT20); and (F) exosome component 8 (EXOCSC8). A full list of all validation assays performed in provided in Table III.

Figure 6

Validation of hits unique to each Nedd4 protein. HEK293T cells were co-transfected (or not) with the indicated hit cDNA and Nedd4 protein (WT or catalytically inactive CS mutant) and ubiquitylation and binding assays performed as described in the legend to Figure 5 and in the Materials and methods. (A) Ubiquitylation and binding by rNedd4-1 to annexin A9 (ANXA9, left panel) or aquaporin9 (AQP9, right panel). (B) Ubiquitylation and binding by hNedd4-1 to the large subunit of RNA Pol-II (RTF1, left panel) or the FGF receptor 1(FGFR1, right panel). (C) Ubiquitylation and binding by hNedd4-2 to the voltage-gated Ca²⁺ channel CACNB1 (left panel) or the activin receptor 1B (ACVR1B, right panel). A full list of all validation assays performed is provided in Table III. To ensure that binding between the Nedd4 proteins and their targets did not occur after cell lysis, independent transfections were also carried out on parallel plates and the lysates from these were mixed after lysis to test for co-IP. No co-IP was detected in any of them (see right-hand panels in A–C, middle blots).

Figure 7

Sustained FGF-mediated signalling by knockdown or knockout (KO) of Nedd4-1 in cells. (A) HEK293T cells were either not transfected (−) or transfected with non-specific (NS) shRNA or a mixture of two shRNAs that target Nedd4-1, as described earlier (Fouladkou et al, 2008). Where indicated, cells were also co-transfected with human Flag-tagged FGFR1 (Flag–hFGFR1). At 1 day after transfections, cells were serum-starved (or not (−)) for 36 h and then treated with human FGF1 (100 ng/ml) plus 10 μg/ml heparin for the indicated times. Cells were then lysed, Flag–hFGFR1 immunoprecipitated with anti-Flag antibodies and immunoblotted (WB) either with anti-pTyr antibodies to detect activated FGFR1, or with anti-Flag antibodies to determine total FGFR1 protein. Aliquots of the lysates were also blotted for hNedd4-1 to prove knockdown, for phospho-Akt (pAkt) and phospho-Erk1/2 (pErk) to determine activation of Akt and Erk, respectively, as well as for total Akt, Erk and β-actin, used as controls. (B) Wild-type (WT) and Nedd4-1^−/− knockout (KO) MEFs were incubated without (starved- time 0) or with mouse FGF2 plus heparin for 5, 30, 60 or 120 min. FGF-dependent signalling was tested by immunoblotting with phospho-specific antibodies against Erk1/2 (pErk) and Akt (pAkt) and compared with total expression of Akt and Erk1/2. Lower panels: quantification of kinetics of Erk and Akt activation (pErk and pAkt) normalized to total Erk and Akt expression. Two independently derived MEF cell lines for each genotype (E2 and E7 for WT and E3 and E5 for KO—see upper right panel) were used in three independent experiments. Error bars represent s.d. values of pooled data.

Figure 8

Summary of protein interactions and protein function for Nedd4 family substrates. (A) Network diagram showing 29 Nedd4 substrates identified in this study (coloured nodes) and their interactions with 85 other proteins (grey nodes)—only first neighbours are shown. (B) Functions of Nedd4 substrates. Each substrate was manually assigned to one of 11 categories on the basis of their GO categories. For both (A) and (B), nodes (proteins) are coloured according to their enzyme specificity. Source data is available for this figure at www.nature.com/msb.