A promiscuous liaison between IL-15 receptor and Axl receptor tyrosine kinase in cell death control

Abstract

Discrimination between cytokine receptor and receptor tyrosine kinase (RTK) signaling pathways is a central paradigm in signal transduction research. Here, we report a 'promiscuous liaison' between both receptors that enables interleukin (IL)-15 to transactivate the signaling pathway of a tyrosine kinase. IL-15 protects murine L929 fibroblasts from tumor necrosis factor alpha (TNFalpha)-induced cell death, but fails to rescue them upon targeted depletion of the RTK, Axl; however, Axl-overexpressing fibroblasts are TNFalpha-resistant. IL-15Ralpha and Axl colocalize on the cell membrane and co-immunoprecipitate even in the absence of IL-15, whereby the extracellular part of Axl proved to be essential for Axl/IL-15Ralpha interaction. Most strikingly, IL-15 treatment mimics stimulation by the Axl ligand, Gas6, resulting in a rapid tyrosine phosphorylation of both Axl and IL-15Ralpha, and activation of the phosphatidylinositol 3-kinase/Akt pathway. This is also seen in mouse embryonic fibroblasts from wild-type but not Axl-/- or IL-15Ralpha-/- mice. Thus, IL-15-induced protection from TNFalpha-mediated cell death involves a hitherto unknown IL-15 receptor complex, consisting of IL-15Ralpha and Axl RTK, and requires their reciprocal activation initiated by ligand-induced IL-15Ralpha.

Figure 1

Axl is required for resistance to TNFα cytotoxicity. L929 and L929R cells were transfected with Axl siRNA. (A) Axl mRNA expression was analyzed by RT–PCR (upper panel). β-Actin is used as an internal control (lower panel). (B) WB analysis of Axl protein expression (upper panel). IL-15Rα expression served as a control (lower panel). (C) Cells were treated for 18 h with TNFα (5 ng/ml), IL-15 (100 ng/ml), or both, and cell viability was assessed by PI and annexin V staining and FACS analysis. (D) L929, L929R, and Axl-overexpressing L929 cells were treated or not with TNFα (100 ng/ml) alone or together with 20 mM zVAD-fmk and ceramide generation was determined. All ceramide determinations were performed in duplicate.

Figure 2

IL-15 induces phosphorylation of IL-15Rα and Axl, which constitutively associate. (A) L929 cells were stimulated with IL-15 (100 ng/ml). Cell lysates were precipitated with specific Abs against Axl, IL-15Rα, or IL-2Rγ and analyzed by WB. (B) Axl/IL-15Rα coassociation was detected in a two-site ELISA. Plates were coated with anti-Axl Abs. Axl (white bars) or Axl/IL-15Rα complexes (gray bars) were detected in cell lysates using biotinylated anti-Axl or anti-IL-15Rα Abs, respectively. COS-7 cells served as a negative control. (C) L929 cells were treated for 15 min with IL-15, TNFα, or both. Axl was precipitated from lysates and its phosphorylation was assessed with anti-pTyr Abs (upper panel). Axl detection served as a loading control (lower panel). (D) Axl and IL-15Rα colocalization analysis by confocal microscopy. Cells were stained with anti-IL-15Rα Abs (red), anti-Axl Abs (green), and TOTO-3 for nuclei (blue). Yellow color indicates IL-15Rα and Axl colocalization (right panels).

Figure 3

Mapping analysis of IL-15Rα/Axl interaction. COS-7 cells were transfected with IL-15Rα, Axl, Axl-Bek, and Axl-DN expression constructs in different combinations. IL-15Rα (A) or Axl (B) was precipitated from the cell lysates using specific Abs and analyzed by WB using anti-IL-15Rα or Axl Abs (upper two panels). Transfection efficiency was proved by protein expression analysis in lysates (lower two panels). (C) Cells were stimulated with IL-15 or Gas6 for 15 min. IL-15Rα or Axl was precipitated from the cell lysates and its phosphorylation was analyzed by WB. IL-15Rα or Axl was detected as a control for transfection and loading.

Figure 4

IL-15 mimics Gas6-mediated signaling. (A) L929 cells were activated with IL-15 or Gas6. Axl (first panel) and PI3K (third panel) were precipitated from the cell lysates and analyzed by WB using anti-pTyr Abs. Phosphorylation of Akt (fifth panel) and ERK (seventh panel) was analyzed using phospho-specific Abs. For loading control, Axl, PI3K, Akt, or ERK was detected after stripping (even panels). (B) Bcl-x_L and Bcl-2 were detected in cells stimulated for 24 h.

Figure 5

Axl-mediated signaling is impaired in L929 cells upon Axl depletion. L929 cells were transfected with Axl siRNA. After 48 h, cells were harvested, serum-starved for 4 h, and activated with IL-15 or Gas6. Axl (first panel) and PI3K (third panel) were precipitated from the cell lysates and analyzed by WB using anti-pTyr Abs. Phosphorylation of Akt (fifth panel) and ERK (seventh panel) was analyzed by phospho-specific Abs. For loading control, Axl, PI3K, Akt, or ERK was detected on the same blots after stripping (even panels).

Figure 6

Gas6 does not affect IL-15-mediated Axl transactivation. (A) Cells were incubated for 10 min with 100 ng/ml Axl-Fc or control IL-3R-Fc proteins and then IL-15 or Gas6 (100 ng/ml) was added for another 5 or 15 min. Axl phosphorylation was detected by WB. (B) L929 cells were cultured in the presence of warfarin (100 μg/ml) for 24 h and stimulated with IL-15 or Gas6. Axl phosphorylation was detected by WB. (C) Recombinant Gas6 or IL-15 (10 nM) was incubated with 10 nM of Axl-Fc , IL-15Rα-Fc, or control IL-3R-Fc proteins. Complexes were captured on protein A-agarose and analyzed by WB using anti-Gas6 (upper panel) or anti-IL-15 (lower panel) Abs.

Figure 7

IL-15 and Gas6 signal in WT but not in Axl−/−, whereas Gas6 but not IL-15 induces signaling in IL-15Rα−/− MEFs. (A) WB analysis of Axl, IL-15Rα, and IL-2Rγ protein expression in WT, Axl−/−, and IL-15Rα−/− MEFs. (B) IL-15Rα expression was analyzed by FACS on cells labeled with IL-15Rα-IgG_2b. (C) MEFs were stimulated with IL-15 or Gas6. Axl was precipitated from the lysates and phosphorylation was analyzed using anti-pTyr Abs (first panel). Same blots were analyzed for the association of Axl with IL-15Rα (third panel). Equal loading was confirmed by re-probing of blots with anti-Axl Abs (second panel). PI3K (fourth and fifth panels) and IL-15Rα (last two panels) were precipitated and their phosphorylation was detected using anti-pTyr Abs. Phosphorylation of Akt and ERK was detected using phospho-specific Abs. Blots were re-probed with PI3K, IL-15Rα, Akt, or ERK Abs for loading control. (D) Bcl-x_L and Bcl-2 were detected in MEFs stimulated with IL-15 or Gas6 for 24 h. (E) Influence of IL-15 or Gas6 on serum deprivation-induced apoptosis in WT, Axl−/−, and IL-15Rα−/− MEFs. Apoptosis was induced by 48 h incubation in serum-free medium. IL-15 or Gas6 was added at a concentration of 100 ng/ml. Cells cultured in serum-free or serum-supplemented media were used as controls. Percentage of apoptotic cells was determined by PI and annexin V staining and FACS analysis. Data represent the mean±s.d. from duplicate samples.

Figure 8

Axl and IL-15Rα coassociate in mouse DCs. (A) DCs were stimulated with cytokines for 15 min. Axl was precipitated from the cell lysates and precipitates were analyzed by WB using anti-IL-4Rα (upper panel), anti-IL-2Rγ (second panel), and anti-IL-15Rα (third panel) Abs. IP using specific Abs served as a positive control. For loading control, blots were developed with anti-Axl Abs (lower panel). (B) Plates were coated with anti-Axl Abs, and biotinylated anti-Axl (specific) or anti-IL-15Rα (two-site) Abs were used as detection Abs. (C) Plates were coated with anti-IL-15Rα Abs, and biotinylated anti-IL-15Rα (specific) or anti-Axl (two-site) Abs served as detection Abs.

Figure 9

Hypothetical model of cross-talk between IL-15 and Axl signaling pathways suggested by the current study.