Abstract

(Introduction) Acute myeloid leukemia (AML) originates from self-renewing leukemic stem cells (LSCs). Self-renewal capacity is one of the most important biological features of LSCs, and therefore, targeting self-renewal machineries of LSCs should be necessary for the eradication of LSCs. We have recently identified the autocrine loop consisted of LSCs-specific surface molecule TIM-3 and its ligand galectin-9 (Gal-9) in human myeloid malignancies. TIM-3/Gal-9 autocrine loop constitutively activates the nucleus accumulation of β-catenin in primary AML LSCs. However, the precise mechanism how this autocrine loop induces β-catenin accumulation has been still unknown. Here, we extended the analysis to clarify the molecular basis of the enhanced β-catenin activity in TIM-3+LSCs.

(Results) We tested whether TIM-3 signaling could interact with the canonical Wnt pathway, which plays a central role in controlling the self-renewal capacity in normal hematopoietic stem cells (HSCs) through regulating intracellularβ-catenin accumulation. LDL receptor-related protein 6 (LRP6) is a key component of canonical Wnt pathway and its phosphorylation is essential for the signal transduction of Wnt signaling. Surprisingly, TIM-3 signaling induced by Gal-9 (10 ng/ml) led to phosphorylation of LRP6and accumulation of β-catenin in primary TIM-3+ AML cells. Furthermore, the phosphorylation of LRP6 induced by Gal-9 was completely abrogated in the presence of an anti-TIM-3 antibody (F38-2E2), which blocks the Gal-9 ligation to TIM-3. These results suggested that the ligation of TIM-3 by Gal-9 could activate canonical Wnt pathway independent of Wnt ligands in TIM-3+AML cells. We then tried to clarify the molecular machinery for canonical Wnt pathway activation by TIM-3/Gal-9 interaction. It has been reported that Wnt signaling promptly induces formation of the protein complex described as LRP6-signalosome preceding LRP6 phosphorylation. Therefore, we tested whether TIM-3 signaling could affect the signalosome formation. In the presence of Dickkopf-1 (Dkk-1; 200 ng/mL), which could inhibit LRP6-signalosome formation, TIM-3 signaling totally failed to phosphorylate LRP6 and to activate β-catenin accumulation, indicating that TIM-3 signaling induced Wnt signaling activation through the LRP6-signalosome formation. Since it has been shown that Src family kinases are cytoplasmic mediator of TIM-3 signaling, we focused on p120-catenin because the molecule has been identified as a substrate of Src family kinases as well as an indispensable molecule in LRP6-signalosome formation. We found that the phosphorylation of p120-catenin, a very early step of canonical Wnt signaling, was promptly induced at Tyr228 by TIM-3/Gal-9 interaction as well as Wnt ligand stimulation (Wnt3a; 200 ng/mL), leading to the enhanced β-catenin accumulation in primary TIM-3+LSCs. These results collectively suggested that the TIM-3 signaling mimicked the canonical Wnt signaling from the very early phase of its signaling cascade in LSCs.

(Discussion) In this study, we showed TIM-3/Gal-9 signal "mimick" canonical Wnt signaling and lead to the aberrant accumulation of intracellular β-catenin in AML LSCs. Considering that TIM-3 is commonly expressed in LSCs of myeloid malignancies, but not in normal HSCs, TIM-3/Gal-9 signal can serve as a promising therapeutic target for the selective eradication of LSCs because inhibition of TIM-3 signaling can impair the self-renewal capacity of LSCs without affecting normal hematopoiesis.

Disclosures

Akashi:Bristol Meyers Squibb: Research Funding; Astellas Pharma: Research Funding; Celgene: Research Funding; Kyowa Hakko Kirin: Consultancy, Research Funding; Chugai Pharmaceutical Co., Ltd.: Research Funding; Asahi Kasei Pharma Corporation: Research Funding; Shionogi & Co., Ltd: Research Funding; Sunitomo Dainippon Pharma: Consultancy.

Author notes

*

Asterisk with author names denotes non-ASH members.