The receptor tyrosine kinase (RTK) AXL, activated by its ligand growth arrest specific protein 6 (Gas6), regulates cell survival, proliferation, and other cellular processes. While mutations in AXL are rare, AXL is widely overexpressed in solid cancers and is predictive of poor outcomes. Upregulated AXL has been shown to promote cancer cell survival and has been linked to chemo-and RTK drug resistance (Tanaka, Int J Mol Sci 2021). It has been suggested that acute myeloid leukemia (AML) cells express AXL, and induce GAS6 in bone marrow stromal cells, which in turn mediates cell proliferation and survival of AML cells (Ben-Batalla, Blood 2013). However, the mechanisms of AXL upregulation are still largely unknown. The role of AXL in hematologic malignancies is also not well established.

Loss-of-function mutations in CBL (Casitas B-lineage Lymphoma), an E3 ubiquitin ligase, are found in a wide range of myeloid malignancies most frequently occurring in chronic and juvenile myelomonocytic leukemia (CMML and JMML). CBL mutations are associated with poor prognosis; however, the therapeutic targets are poorly established. Here, we used a human GM-CSF-dependent hematopoietic progenitor cell line (TF1) to study CBL signaling and regulation. Dual deletion of CBL and CBLB (CBL-DKO) or overexpression of the E3-dead C381A mutant CBL renders TF1 cells cytokine-independent. To explore therapeutic targets for CBL-mutant myeloid malignancies, we performed CRISPR-Cas9 screens targeting functional domains of human kinome, in CBL-DKO versus control TF1 cells. This screen identified AXL as specifically required for the growth of CBL-DKO cells.

TF1 cells express negligible level of AXL, while AXL level is marked upregulated in CBL loss or mutant cells. CBL family E3 ligases are known to ubiquitinate and promote the degradation of various active tyrosine kinases. Both CBL and CBLB have been reported to ubiquitinate AXL in vitro (Paolino, Nature 2014; Valverde, BBRC 2005). However, our RNAseq studies revealed that AXL mRNA was also upregulated in CBL-DKO and mutant CBL cells compared with control or wildtype CBL expressing cells. ChIPseq analysis revealed increased active epigenetic marks (H3K27Ace, H3K4me3) at the AXL locus in CBL-DKO and mutant CBL cells, implying a potential positive feedback loop composed of protein stability, signaling, and transcription regulation.

We found that serum induced the phosphorylation of AXL as well as AKT and ERK at 15 min, while STAT5 phosphorylation occurred at 3hrs, in CBL-DKO and mutant CBL cells but not in control cells. Importantly, AXL depletion by CRISPR/Cas9-mediated knockout diminished serum-induced AXL/AKT/ERK /STAT5 phosphorylation in CBL-DKO cells, and consequently, abrogated CBL-DKO cell growth in serum. Interestingly, AXL depletion minimally affected GM-CSF-induced AKT/ERK/STAT5 signaling. Identification of AXL upregulation upon CBL loss implicates it as a potential therapeutic target in CBL mutant myeloid malignancies. Indeed, AXL inhibitors potently diminished AKT/ERK/ STAT5 phosphorylation in CBL-DKO and CBL mutant cells. Importantly, AXL inhibitors blunted CMML development in tamoxifen-induced Cbl-/-;Cbl-b-/- mice in vivo.

Half-life studies showed that the glycosylated slow-migrating AXL form is more stable than the under-glycosylated fast-migrating AXL form. It has been shown that AXL is an HSP90 client protein that stabilizes glycosylated AXL (Krishnamoorthy, JBC 2013). Indeed, we found that HSP90 inhibitors significantly reduced AXL phosphorylation and protein levels, resulting in a concurrent decrease in downstream signaling and cell growth. Moreover, combined inhibition of AXL and HSP90 showed a more pronounced signaling suppression than single agents and displayed synergic effects in restraining CBL-DKO cell growth.

In summary, we discovered that AXL is upregulated and maintained at high levels in CBL loss and CBL-mutant cells. AXL-mediated signaling is critical to cytokine-independent growth, serving as a promising therapeutic target for the treatment of CBL-mutant myeloid malignancies. This study also sheds light on the mechanisms of AXL upregulation for improved precision medicine for myeloid malignancies.

Disclosures

No relevant conflicts of interest to declare.

This content is only available as a PDF.
Sign in via your Institution