Highlights CRLF2 overexpression associates with IKZF1 deletions that lead to a dominant-negative effect and with IKZF1 plus. Paediatric patients with a high load expression of IK4 isoform presented higher CRLF2 transcript levels. CRLF2 overexpression and IKZF1 deletions conferred poorer prognosis both to paediatric patients treated with RELLA05 protocol as well as to adult patients.

T‐cell acute lymphoblastic leukemia (T‐ALL) is an aggressive hematopoietic malignancy with few molecular alterations showing a consensual prognostic value. CRLF2 overexpression was recently identified in high‐risk T‐ALL patients. For these cases, no genomic abnormality was found to be associated with CRLF2 overexpression. IKZF1 has been recently shown to be a direct transcriptional regulator of CRLF2 expression. Moreover, it is known that NOTCH1 antagonizes IKZF1 in T‐ALL. In light of these pieces of evidence, we reasoned that IKZF1 binding perturbation and CRLF2 upregulation could be associated in T‐ALL. We evaluated two independent series of pediatric T‐ALL cases (PHOP, n = 57 and TARGET, n = 264) for the presence of common T‐ALL molecular abnormalities, such as NOTCH1/FBXW7 mutations. We also assessed CRLF2 and IKZF1 gene expression. CRLF2 overexpression was observed in 14% (PHOP) and 16% (TARGET) of T‐ALL patients. No correlation was found between mRNA expression of CRLF2 and IKZF1 in both cohorts. Interestingly, we show that patients with mutations affecting NOTCH1‐PEST domain and/or FBXW7 had higher CRLF2 expression (P = .04). In summary, we demonstrate for the first time that only mutations resulting in ICN1 (intracellular domain of NOTCH1) stabilization are associated with CRLF2 overexpression.

CRLF2 overexpression has been described as a biomarker of poor prognosis in T‐cell acute lymphoblastic leukemia (T‐ALL). In the present study, we aimed to unravel the genomic profile underlying CRLF2 overexpression (CRLF2‐high) by analyzing RNA‐seq, WES, and SNP‐array data from 264 T‐ALL patients and five cell lines deposited on the TARGET initiative, Cancer Cell Line Encyclopedia and Gene Expression Omnibus. These data allowed us to delineate the genomic landscape of CRLF2‐high in T‐ALL, which was associated with PTEN, JAK3, PHF6, EZH2, and RUNX1 mutations. We also observed an enrichment of CRLF2‐high in early T‐precursor (ETP)‐ALL (23.08% vs. 4.02%, P = 7.579e−06) and a very similar gene upregulation profile between these two entities. The inhibition of BET (iBET) proteins is a strategy previously demonstrated to reverse the gene upregulation pattern of ETP cells through restoration of polycomb repressive complex 2 (PRC2) activity. While CRLF2 expression was rescued by using this strategy in LOUCY (untreated vs. iBET P = 0.0095, DMSO vs. iBET P = 0.0286), a classical ETP‐ALL cell line, PRC2 loss was not sufficient to promote CRLF2 upregulation in JURKAT, a more mature T‐ALL cell line. Considering the role of IKZF1 in CRLF2 regulation and in recruitment of PCR2, we evaluated IKZF1 status according to CRLF2‐expression subgroups. We identified that IKZF1 transcripts with intron retention were upregulated in the CRLF2‐high subgroup. Here, we delineated the gene expression profile of CRLF2‐high T‐ALL samples and unraveled the crucial role of PRC2 in CRLF2 regulation in ETP‐ALL.

No abstract

No abstract

IKZF1 deletions are associated with an increased risk of relapse in B-cell precursor acute lymphoblastic leukemia (B-ALL), and their accurate detection has great clinical impact. Here, we included four international cohorts of pediatric and adult patients with B-ALL, and reviewed literature to illustrate the recombination map of IKZF1 deletions, with a focus at non-recurrent deletions. We provide a substantial basis for the improvement of diagnostic methods based on MLPA and multiplex PCR for the identification of IKZF1 deletions, and also demonstrate that rare IKZF1 deletions increase the incidence of relapse in these patients. Of note, non-recurrent deletions comprised a wide range of alterations, but the majority were Δ1 and Δ1–3. They were often associated with reciprocal IKZF1 fusions. So far, a total of 23 IKZF1 gene fusions were identified in B-ALL. We also verified the occurrence of the heptamer sequence (E-value: 9.9 x 10− 9) and an enrichment of GC nucleotides (71% versus 56%; P value = 4.9 x 10− 3) exclusively within breakpoint clusters, suggesting that RAG recombination and TdT activity may promote the majority of IKZF1 deletions, although rare types of alterations may be associated with other molecular mechanism of leukemogenesis, such as microhomology-mediated end joining.

No abstract