J Cell Sci
January 1, 2018;
Phosphorylation of XIAP at threonine 180 controls its activity in Wnt signaling.
X-linked inhibitor of apoptosis (XIAP
) plays an important role in preventing apoptotic cell death. XIAP
has been shown to participate in signaling pathways, including Wnt signaling. XIAP
regulates Wnt signaling by promoting the monoubiquitylation of the co-repressor Groucho/TLE family proteins, decreasing its affinity for the TCF/Lef family of transcription factors and allowing assembly of transcriptionally active β-catenin-TCF/Lef complexes. We now demonstrate that XIAP
is phosphorylated by GSK3
at threonine 180, and that an alanine mutant (XIAPT180A) exhibits decreased Wnt activity compared to wild-type XIAP
in cultured human cells and in Xenopus embryos. Although XIAPT180A ubiquitylates TLE3 at wild-type levels in vitro, it exhibits a reduced capacity to ubiquitylate and bind TLE3 in human cells. XIAPT180A binds Smac
(also known as DIABLO) and inhibits Fas
-induced apoptosis to a similar degree to wild-type XIAP
. Our studies uncover a new mechanism by which XIAP
is specifically directed towards a Wnt signaling function versus its anti-apoptotic function. These findings have implications for development of anti-XIAP
therapeutics for human cancers.
J Cell Sci
Wnt signaling pathway
[+] show captions
XIAP is phosphorylated by GSK3 at T180. (A) Endogenous XIAP co-immunoprecipitates with endogenous GSK3. XIAP was immunoprecipitated (IP) from whole-cell lysates (WCL) of HEK293STF cells incubated in the absence or presence of recombinant Wnt3a followed by immunoblotting. The asterisk indicates the heavy chain of IgG. (B) Purified XIAP is phosphorylated by GSK3 in vitro in a [γ32P]ATP kinase assay. Reactions were analyzed by SDS/PAGE followed by autoradiography. (C) CK1 enhances XIAP phosphorylation by GSK3. XIAP bound to beads was incubated with CK1 in a kinase reaction containing non-radioactive (cold) ATP (lanes 4 and 6). CK1 was washed away and XIAP-bound beads incubated with GSK3 in a kinase reaction containing [γ32P]ATP. Reactions were analyzed by SDS/PAGE followed by autoradiography. Kinase only lanes show CK1 and GSK3 autophosphorylation (lanes 1 and 2). Results in A–C were replicated at least three times. (D) The predicted phosphorylation site at T180 of XIAP is an in vitro GSK3 phosphorylated site. Predicted phosphorylation sites on XIAP were curated by the PhosphoSitePlus online resource tool (Hornbeck et al., 2015). LC-MS analysis of XIAP identifies T180 (asterisk) as a prominent in vitro GSK3 phosphorylation site.
The XIAPT180A mutant exhibits decreased Wnt activity compared to wild-type XIAP in cultured human cells and in Xenopus embryos. (A) GSK3 associates with XIAP and the kinase mutant, XIAPT180A, to a similar extent. HEK293STF cells were transfected with vector, HA–GSK3, MYC–XIAP and MYC-XIAPT180A as indicated, whole cell lysates (WCL) were collected, and immunoprecipitation (IP) performed with anti-MYC antibody. The asterisk indicates endogenous XIAP. (B) The XIAPT180A mutant does not exhibit a localization that is distinct from XIAP. Left, HEK293STF cells were transfected with MYC–XIAP or MYC–XIAPT180A, fixed, and immunostained for MYC and DNA (DAPI). Scale bars: 12.5 µm. Right, cytoplasmic and nuclear fractionations were isolated from transfected cells and immunoblotted for MYC, Histone H3 (a nuclear marker) and GAPDH (a cytoplasmic marker). (C) In contrast to wild-type XIAP, XIAPT180A does not potentiate Wnt3a signaling. HEK293STF cells were transfected with vector, MYC–XIAP or MYC–XIAPT180A as indicated, for 24 h, and cells incubated in the absence or presence of recombinant Wnt3a for 24 h. The asterisk indicates endogenous XIAP. Graph shows mean±s.d. of TOPflash activation normalized to the cell titer. ***P<0.0001 (one-tailed Student's t-test). All experiments were repeated at least three times. (D) The XIAPT180A mutant does not potentiate Wnt8 (Xwnt8)-induced axis formation in Xenopus embryos, in contrast to wild-type XIAP. Embryos (four-cell stage) were injected ventrally with control, XIAPT180A or XIAP mRNA (2 ng each) with or without Wnt8 mRNA (0.1 ng) and allowed to develop. Representative Wnt8-, XIAP- and XIAPT180A-injected embryos are on the left. The percentage of embryos with secondary axis formation is shown on the right (n≥31 per group). ***P<0.0001 (Fisher's exact test, with Bonferroni correction). All results were replicated at least three times.
The XIAPT180A mutant shows decreased binding and ubiquitylation of TLE3. (A) XIAPT180A ubiquitylates TLE3 to a similar extent to wild-type XIAP in vitro. In vitro-translated HA–TLE3 was incubated in an in vitro ubiquitylation reaction containing recombinant E1/E2, ubiquitin, and XIAP or XIAPT180A. Ubiquitylated TLE3 was visualized by immunoblotting with anti-HA antibody. (B) XIAPT180A exhibits reduced capacity to ubiquitylate TLE3 in cultured human cells compared to wild-type XIAP. HEK293STF cells were transfected as indicated, lysed under denaturing conditions, and His-Ub-modified proteins isolated by nickel affinity chromatography. XIAP and TLE3 were detected by immunoblotting with anti-MYC and anti-HA antibodies, respectively. The asterisk indicates endogenous XIAP. WCL, whole-cell lysates. (C) XIAPT180A exhibits decreased affinity for HA–TLE3 compared to wild-type XIAP. HEK293STF cells were transfected for 24 h as indicated followed by incubation in the absence or presence of recombinant Wnt3a for 24 h. Lysates were collected and immunoprecipitated (IP) with anti-MYC antibody. Co-immunoprecipitated HA–TLE3 was detected by anti-HA antibody. All results were replicated at least three times.
In contrast to its activity in the Wnt pathway, XIAPT180A has a similar activity to that of wild-type XIAP in the apoptotic pathway. (A) Smac co-immunoprecipitates (IP) with XIAP and XIAPT180A to a similar extent. HEK293STF cells were transfected as indicated with vector or HA–Smac plus MYC–XIAP or MYC–XIAPT180A. Lysates were collected and immunoprecipitation performed with anti-MYC antibody followed by immunoblotting with anti-HA antibody. WCL, whole cell lysates. Results were replicated at least three times. (B) Overexpression of XIAP and XIAPT180A decreases the percentage of Fas ligand (FasL)-induced apoptosis to a similar extent. HeLa cells were transfected for 24 h as indicated and incubated in the absence or presence of recombinant Fas ligand (100 ng/ml) for 24 h. Cells were then stained with Annexin V and propidium iodide. Graph shows the mean±s.d. of percentage of apoptotic cells. Data was analyzed and processed using FlowJo. **P<0.001, *P<0.01; n.s., not significant (one-tailed Student's t-test). All experiments were repeated at least three times. (C) A model for the regulation of XIAP activity in the Wnt pathway by phosphorylation at position T180. See text for more information.