Neuralized family member NEURL1 is a ubiquitin ligase for the cGMP-specific phosphodiesterase 9A

Abstract

Neuralized functions as a positive regulator of the Notch pathway by promoting ubiquitination of Notch ligands via its E3 ligase activity, resulting in their efficient endocytosis and signaling. Using a yeast two-hybrid screen, we have identified a cGMP-hydrolysing phosphodiesterase, PDE9A, as a novel interactor and substrate of Neuralized E3 ubiquitin protein ligase 1 (NEURL1). We confirmed this interaction with co-immunoprecipitation experiments and show that both Neuralized Homology Repeat domains of NEURL1 can interact with PDE9A. We also demonstrate that NEURL1 can promote polyubiquitination of PDE9A that leads to its proteasome-mediated degradation mainly via lysine residue K27 of ubiquitin. Our results suggest that NEURL1 acts as a novel regulator of protein levels of PDE9A.

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1

NEURL1 and NEURL1B interact with PDE9A. (A) PDE9A-V5His was expressed alone or together with either NEURL1-Flag, Flag-NEURL1 or Flag-NEURL1B in HEK293-FT cells. (B) PDE9A-V5His was expressed alone or together with untagged NEURL1 in HEK293-FT cells. In A and B the proteins were immunoprecipitated and immunoblotted using the indicated antibodies (untagged NEURL1 was detected with a custom-raised polyclonal NEURL1 antibody). (C) PDE9A-V5His and Flag-NEURL1 or NEURL-Flag or Flag-NEURL1B were co-expressed in HEK293 cells and their localisation was visualised by indirect immunofluorescence using respective tag-specific antibodies (co-localisation of green and red signals results in yellow additive colour in the Merge column). Nuclei were visualised using Hoechst 33342 (blue). (D) Flag-NEURL1 and E2-tagged PDE5A, PDE9A and PDE11A were expressed alone or in combination in HEK293-FT cells. Cell lysates were immunoprecipitated and immunoblotted with the indicated antibodies. IB, immunoblot; IP, immunoprecipitation.

Figure 2

NEURL1-PDE9A interaction is mediated by the NHR1 and NHR2 domains of NEURL1 and the catalytic and regulatory domains of PDE9A. (A) Schematic diagram of the various NEURL1 and PDE9A constructs used. (B) NEURL1-Flag, NEURL1-Rm-Flag Flag (encoding the ubiquitin ligase defective protein), Flag-NHR2 and PDE9A-E2 were expressed in various combinations in HEK293-FT cells. (C) PDE9A-V5His protein was expressed alone or in combination with either EGFP-NHR1 + NHR2 or NHR1-EGFP in HEK293-FT cells. (D) PDE9A-cat-E2 or PDE9A-reg-E2 (encoding the catalytic and regulatory domains of PDE9A, respectively) were expressed alone or together with Flag-NEURL1 in HEK293-FT cells. In B-D the proteins were immunoprecipitated and immunoblotted using the indicated antibodies. NHR, Neuralized Homology Repeat; Rm, mutant RING domain; cat, catalytic domain; reg, regulatory domain; IB, immunoblot; IP, immunoprecipitation.

Figure 3

NEURL1 but not NEURL1B promotes ubiquitination of PDE9A. (A) HA-tagged Ubiquitin (HA-Ub), PDE9A-V5His, Flag-NEURL1B, Flag-NEURL1, NEURL1-Flag, NEURL1-Rm-Flag and untagged NEURL1 were expressed in various combinations in HEK293-FT cells in the presence of proteasomal inhibitor MG-132. The proteins were immunoprecipitated using the indicated antibodies. Exogenously and endogenously ubiquitinated PDE9A were detected with HA and UbVU-1 antibodies, respectively. The higher molecular weight smear detected with HA and UbVU-1 antibodies in lanes 4, 5 and 9 indicates polyubiquitination of PDE9A. (B) PDE9A-V5His was expressed alone or together with either NEURL1-Flag or Flag-NEURL1 in HEK293-FT cells. Following immunoprecipitation with the indicated antibodies, PDE9A was detected in immunoprecipitates of endogenous ubiquitin with V5 antibodies and endogenous ubiquitin was detected in PDE9A immunoprecipitate with UbVU-1 antibodies. For clearer depiction of polyubiquinated PDE9A in the inputs of V5 immunoblots, the tonal range (i.e. levels) was changed differently for non-ubiquinated and polyubiquinated PDE9A species. The border for such differential level change is marked with a dotted line. IB, immunoblot; IP, immunoprecipitation.

Figure 4

NEURL1 promotes polyubiquitination of PDE9A via lysines K27, K29 and K33 of ubiquitin. (A) Wild-type HA-tagged ubiquitin, K0 ubiquitin (i.e. ubiquitin mutant whose lysine residues had been substituted for arginine) and the various K only ubiquitin mutants were co-expressed with either PDE9A-V5His only or with PDE9A-V5His and NEURL1-Flag in HEK293-FT cells in the presence of proteasome inhibitor MG-132. The proteins were immunoprecipitated and immunoblotted using the indicated antibodies. (B) Wild-type HA-tagged ubiquitin, K0 ubiquitin and the various K only ubiquitin mutants were co-expressed with either PDE9A-V5His only or together with NEURL1-Flag in HEK293-FT cells in the absence of proteasome inhibitor MG-132. The lysates were immunoblotted using the indicated antibodies. (C) Quantification of NEURL1-mediated degradation of PDE9A in the presence of different ubiquitin mutants when overexpressed either alone or with NEURL1. Coomassie-normalised levels of PDE9A in cells co-expressing NEURL1 and the respective ubiquitin mutant were divided with the respective levels of PDE9A in cells where NEURL1 was not overexpressed. The calculated ratio in cells overexpressing the K0 mutant was arbitrarily set as 1. The average of 5 independent experiments is shown, with error bars representing SEM. Statistical significance was calculated relative to the ratio in cells expressing K0 ubiquitin. *p < 0.05, **p < 0.01, ***p < 0.001, two-tailed paired t-test. IB, immunoblot; IP, immunoprecipitation.

Figure 5

NEURL1 mediates proteasomal degradation of PDE9A in HEK293-FT cells. (A) PDE9A-V5His was expressed either alone or together with various amounts (−, 0 μg; + , 1 μg; ++, 2 μg; +++, 3 μg) of Flag-NEURL1 or NEURL1-Flag or NEURL1-Rm-Flag in HEK293-FT cells. 24 hours after transfection, the cells were lysed and the lysates were immunoblotted with the indicated antibodies. (B) PDE9A-V5His was co-expressed with empty vector (‘negative control’) or NEURL1-Flag or NEURL1-Rm-Flag in HEK293-FT cells. Following transfection, the cells were first incubated for 16 hours in the presence of proteasome inhibitor MG-132 and then treated with 100 µM of protein synthesis inhibitor cycloheximide (CHX) for the indicated time periods. The lysates were immunoblotted with the antibodies indicated and the protein levels from two independent experiments were quantified. (C,D) PDE9A-V5His was expressed alone or together with NEURL1-Flag in HEK293-FT cells. 24 hours after transfection, the cells were treated with lysosomal inhibitor chloroquine or ammonium chloride (C) or with proteasome inhibitor MG-132 (D) at the indicated concentrations for 24 hours. The cells were lysed and subjected to immunoblotting with the indicated antibodies. In all experiments shown in A through D, immunoblotting with anti-GAPDH antibody was performed to demonstrate equal loading. IB, immunoblot.

Figure 6

NEURL1 promotes degradation of PDE9A in primary neurons. PDE9A-V5His was expressed either alone or together with NEURL1-Flag in primary hippocampal (HC), cortical (CTX) or cerebellar (CBL) neurons. 2–3 days post transfection, the cells were lysed and subjected to immunoblotting with the indicated antibodies. NEURL1-Flag could not be unequivocally detected due to the emergence of high background signal in primary neurons when using anti-Flag antibodies (data not shown). Immunoblotting with anti-GAPDH antibody was performed to demonstrate equal loading. IB, immunoblot.