LEF1 intragenic deletions: is Wnt’er coming for T-ALL? Free

In this issue of Blood, Delafoy et al¹ demonstrate that intragenic exon 2 and 3 deletions are the most frequent LEF1 alterations in T-cell acute lymphoblastic leukemia (T-ALL), originating a dominant-negative LEF1 form that is associated with increased sensitivity to calcineurin inhibition and glucocorticoids (see figure).

LEF1, a member of the LEF/TCF family of transcription factors, participates in Wnt signaling by interacting with β-catenin in the nucleus. LEF1 and Wnt signaling are critical for normal T-cell development, but the portrait of their involvement in T-ALL is complex, possibly context-dependent, hence remains incomplete. Evidence that T-ALL stem cells rely on Wnt signaling,² that β-catenin activation predisposes thymocytes to transformation,³ and that high LEF1 levels in adult patients with T-ALL are associated with higher white blood cell counts⁴ point toward Wnt signaling being oncogenic in T-ALL. Pointing in the opposite direction is the knowledge that LEF1 inactivation is recurrent in T-ALL.^5,6 Now, Delafoy et al used almost 500 pediatric and adult T-ALL cases, enrolled in the FRALLE 2000 and GRAALL-2003-2005 trials, to dive deeper into the incidence and the biological and clinical impact of LEF1 lesions. In agreement with Pölönen et al,⁶ who found LEF1 lesions in 15% of pediatric T-ALL, they identified LEF1 alterations in 13% of the cases, of which 14.3% were mutations (mostly truncating), 28.6% large deletions, and 57.1% focal deletions. These results confirm that LEF1 lesions are recurrent and inactivating.

LEF1 alterations are associated with a mature cortical αβ immunophenotype, and SIL::TAL1 and TLX1 genetic subgroups, but are rare in early T-cell precursor ALL. Patients with LEF1 alterations showed lower levels of PRC2 complex mutations (particularly in SUZ12) and higher frequency of alterations in CDKN2A/B and PI3K signaling pathway members (namely in PIK3R1 and PTEN).

Delafoy et al then assessed the clinical impact of the LEF1 alterations. They did not find statistical differences in age, sex ratio, or median leukocyte count at diagnosis in patients with and without LEF1 lesions. Interestingly, patients with the LEF1 inactivating lesions displayed a higher prevalence of central nervous system involvement, a surprising observation given the evidence that LEF1 silencing impairs T-ALL–cell migration.⁷ However, LEF1 inactivation could still provide a competitive advantage to T-ALL cells that have spread to the brain. Although a more granular view did not reveal associations between specific LEF1 lesions and clinical parameters, this may be due to the small size of each LEF1 group. Patients with LEF1 lesions had better prednisone response and increased chemosensitivity than LEF1 wild-type patients. However, LEF1 alterations did not impact long-term survival outcomes with similar survival to wild-type patients.

A major finding is that the most frequent LEF1 alterations in T-ALL cells are focal deletions that lead to a LEF1 isoform lacking exons 2 and 3, thus juxtaposing exon 1 with exon 4 (hereafter, LEF1 del2/3). These somatic intragenic lesions result in a shorter LEF1 protein that acts in a dominant-negative manner, that is, it blocks the activity of wild-type LEF1 and impairs Wnt signaling. How this deletion results in a dominant-negative role remains unclear. LEF1 del2/3 is unable to bind β-catenin in coimmunoprecipitation assays in T-ALL cells. However, how does this inability to bind β-catenin prevent wild-type LEF1 from fulfilling its normal function? And how does exon 2-3 deletion prevent LEF1 binding to β-catenin, since the β-catenin binding domain is encoded by exon 1? Apart from these questions, transcriptomic analyses confirmed downregulation of Wnt signaling, repression of pathways normally activated by T-cell receptor signaling (eg, MAPK, PI3K, and calcium signaling), and enrichment of cell cycle–related gene signatures in T-ALL cells with LEF1 del2/3.

The authors then reasoned that residual Wnt signaling is essential for LEF1 del2/3 T-ALL cells, thus providing a potential therapeutic vulnerability. They showed that cyclosporin A, a calcineurin inhibitor, is more effective against LEF1 del2/3 than wild-type T-ALL cells. This is consistent with the concomitant downregulation of Wnt and calcium signaling in LEF1 del2/3 T-ALL and the known cross talk between Wnt (including noncanonical signaling) and calcineurin signaling. As noted above, LEF1 alterations were associated with increased prednisone responsiveness in patients. The link between LEF1 del2/3 and response to glucocorticoids in T-ALL cells was explored in patient-derived xenograft (PDX) models. The combination of cyclosporin A and dexamethasone was synergistic in a LEF1 del2/3 but not in a LEF1 wild-type PDX sample in vitro. Combined administration of cyclosporin A and dexamethasone prolonged the survival of mice (as compared to each treatment alone and untreated controls) transplanted with a PDX sample carrying a LEF1 del2/3. This survival benefit was not observed in mice transplanted with PDX samples that were LEF1 wild-type or harbored a large LEF1 deletion. These results, although based on a very limited number of samples, suggest that combination treatment may be particularly indicated for T-ALL patients with LEF1 focal deletions.

Several important questions remain. For example, what is the selective advantage driven by Wnt signaling downregulation via LEF1 inactivation? Which mechanisms explain the cooperation between calcineurin signaling pharmacological inhibition and dexamethasone specifically in LEF1 del2/3 T-ALL and not in large LEF1 deletions? Could it relate, at least in part, to signaling pathways other than Wnt that LEF1 can crosstalk with (for instance TCR-dependent)? Are there other therapeutic combinations that may be used to selectively target T-ALL cells with other types of LEF1 lesions?

These and other questions must be substantiated by evidence arising from broader studies with larger numbers of PDXs and clinical samples. Nonetheless, the studies by Delafoy et al definitively clarify that LEF1 inactivation is a common event in T-ALL. And they further expose a therapeutic vulnerability, which arises from LEF1 focal deletions affecting exons 2 and 3 and the consequent generation of a dominant-negative LEF1 isoform. This merits further clinical investigation. Perhaps Wnt’er is coming for T-ALL.

Conflict-of-interest disclosure: The authors declare no competing financial interests.