Suppressor of fused and Spop regulate the stability, processing and function of Gli2 and Gli3 full-length activators but not their repressors

Abstract

Gli2 and Gli3 are primary transcriptional regulators that mediate hedgehog (Hh) signaling. Mechanisms that stabilize and destabilize Gli2 and Gli3 are essential for the proteins to promptly respond to Hh signaling or to be inactivated following the activation. In this study, we show that loss of suppressor of fused (Sufu; an inhibitory effector for Gli proteins) results in destabilization of Gli2 and Gli3 full-length activators but not of their C-terminally processed repressors, whereas overexpression of Sufu stabilizes them. By contrast, RNAi knockdown of Spop (a substrate-binding adaptor for the cullin3-based ubiquitin E3 ligase) in Sufu mutant mouse embryonic fibroblasts (MEFs) can restore the levels of Gli2 and Gli3 full-length proteins, but not those of their repressors, whereas introducing Sufu into the MEFs stabilizes Gli2 and Gli3 full-length proteins and rescues Gli3 processing. Consistent with these findings, forced Spop expression promotes Gli2 and Gli3 degradation and Gli3 processing. The functions of Sufu and Spop oppose each other through their competitive binding to the N- and C-terminal regions of Gli3 or the C-terminal region of Gli2. More importantly, the Gli3 repressor expressed by a Gli3 mutant allele (Gli3(Delta699)) can mostly rescue the ventralized neural tube phenotypes of Sufu mutant embryos, indicating that the Gli3 repressor can function independently of Sufu. Our study provides a new insight into the regulation of Gli2 and Gli3 stability and processing by Sufu and Spop, and reveals the unexpected Sufu-independent Gli3 repressor function.

Fig. 1.

Sufu is required for the stabilization of Gli2 and Gli3 proteins in vivo. (A) Gross morphology of wild-type and Sufu^−/− embryos at E9.5. (B) Immunoblot showing that levels of Gli2 and Gli3 proteins are significantly reduced in Sufu^−/− embryos. The relative intensity of Gli2^FL, Gli3^FL and Gli3^Rep was quantified by NIH image software and is shown above each bar in the righthand graph. (C) Overexpression of Sufu stabilizes Gli2 and Gli3 proteins in HEK293 cells.

Fig. 2.

Spop promotes Gli2 and Gli3 degradation and Gli3 processing only in the presence of Sufu. All panels are immunoblots. (A) Inhibition of the proteasome activity with MG132 restores Gli2 and Gli3 protein levels in Sufu mutant MEFs. (B) Spop shRNA knockdown rescues levels of Gli2^FL and Gli3^FL, but not those of Gli3^Rep in Sufu^−/− MEFs (compare lane 1 with lane 4). (C) Introducing Sufu into Sufu^−/− MEFs rescues both Gli2 and Gli3 stability and Gli3 processing. Sufu^−/− MEFs were infected with control or Sufu retrovirus. After cells were lysed, Gli2, Gli3, Sufu and αtubulin expression was determined by western blot with the indicated antibodies. Relative levels of Gli3^FL and Gli3^Rep from three exposures are plotted on the right. P≤0.0073, Student's t-test. (D) Overexpression of Spop promotes Gli2 and Gli3 degradation in a proteasome-dependent manner in both HEK293 and C3H10T1/2 cells. Compare lane 4 with lanes 2 and 3, and lane 8 with lanes 6 and 7 for Gli2 and Gli3 levels. Expression constructs are shown at the top, antibodies for western blot and cell lines to the left. (E) Overexpression of Spop facilitates the processing of Gli3 but not the Gli3P1-6 mutant in transfected HEK293 cells (compare the Gli3^Rep levels in lane 3 with those in lane 5). (F) Gli2 and Gli3 are ubiquitylated by Spop. HEK293 cells were transfected with constructs as indicated. After cells were treated with MG132 and lysed, the lysates were subjected to immunoprecipitation with anti-Gli2 or anti-Gli3 antibodies, then immunoblotting with an anti-Myc antibody.

Fig. 3.

Overexpression of Spop inhibits Gli2 and Gli3 transcriptional activity. Gli2^−/−; Gli3^xt/xt MEFs were transfected with 8×Gli-binding site-luciferase, TK-renillar luciferase, and a control vector, Gli2, or Gli3 expression constructs, together with or without the HA-Spop expression construct as indicated. Gli2 and Gli3 transcription activity is significantly reduced by coexpression of Spop.

Fig. 4.

Spop interacts with the N- and C-terminal regions of Gli3 and the N-terminal region of Gli2. (A) Gli-binding oligonucleotide beads pulled down Spop when Spop was coexpressed with Gli2 or Gli3 in HEK293 cells treated with MG132. (B-F) Mapping of Spop-binding sites to the Gli2 C-terminus (B-D) and the N- and C-termini of Gli3 (E,F). Shown at the top are proteins expressed in HEK293 cells. Co-immunoprecipitation and GST pull-down were performed (upper panels in B, C and F). Shown in the middle and lower panels are immunoblots of protein lysates with indicated antibodies. Mapping results are summarized in D and F. (G) Immunoblot showing that Sufu and Spop antagonistically regulate Gli2 and Gli3 protein stability. Gli2 and Gli3 protein levels were reduced by coexpression with Spop, but increased or restored by coexpression of Sufu or both. (H) Sufu and Flag-Spop-MATH competitively bind to Gli2 and Gli3. HEK293 cells were transfected with HA-Sufu, various amount (μg) of Flag-Spop-MATH, and Gli2 or Gli3 expression constructs. Cell lysates were subjected to immunoprecipitation with Gli2 or Gli3 antibodies, then immunoblotting with an HA antibody (upper panel). Shown in other panels are immunoblots of cell lysates with αFlag, αHA, αGli2 or αGli3 antibodies.

Fig. 5.

Spop promotes the degradation of Gli2 and Gli3 full-length proteins but not their repressors or C-terminal fragments. (A,B) Immunoblots showing that overexpression of Spop degraded Gli2^FL and Gli3^FL but did not affect the stability of Gli2 and Gli3 repressors or their C-terminal fragments (compare lanes with Spop to those without it). The endogenous Gli2 and Gli3 proteins were clearly detected. As only a small fraction of cells were transfected, the decrease in endogenous Gli2 and Gli3 levels in cells transfected with Spop was not obvious.

Fig. 6.

Immunoblot showing that both Gli3^FL and Gli3^Rep, but not the Gli3^Δ699 repressor, were degraded in Sufu mutant mouse embryos. Compare lane 1 with lane 3, and lane 2 with lane 4 in the upper panel. Immunoblot in middle panel confirms the loss of Sufu expression in the mutant. The lower panel shows tubulin expression for loading controls.

Fig. 7.

Expression of the Gli3^Δ699 repressor rescues Sufu mutant neural tube phenotypes. (A-T) Neural tube sections prepared from E9.5 embryos with indicated genotypes (left) were immunostained with antibodies against the transcription factors shown at the top. In the Gli3^Δ699 mutant, only the Hb9 expression domain is much smaller than that in wild type; the patterns of the rest of the markers examined are similar (compare A-E with F-J). The difference between the left and right lateral Nkx2.2+ domains in B and G does not represent actual differences, rather it is the result of the position of the embryos while sectioning. Gli3^Δ699 rescues Sufu mutant neural tube phenotypes (compare K-O with P-T). Arrowheads in Q indicate Nkx2.2+ cells. (U-X) In situ hybridization of neural tube sections of E9.5 embryos (genotypes at the top) with the Olig2 probe.

Fig. 8.

Gli3^Δ699 expression largely restores the pattern of Ptc expression in the Sufu mutant neural tube. (A-D) E9.5 neural tube sections with the indicated genotypes were hybridized with the Ptc probe.