Critical role of tripartite fusion and LBD truncation in certain RARA- and all RARG-related atypical APL

Atypical acute promyelocytic leukemia (aAPL) presents a complex landscape of retinoic acid receptor (RAR) fusion genes beyond the well-known PML::RARA fusion. Among these, 31 individually rare RARA and RARG fusion genes have been documented, often reported in the canonical X::RAR bipartite fusion form. Intriguingly, some artificially mimicked bipartite X::RAR fusions respond well to all-trans retinoic acid (ATRA) in vitro, contrasting with the ATRA resistance observed in patients. To unravel the underlying mechanisms, we conducted a comprehensive molecular investigation into the fusion transcripts in 27 RARA fusion gene-positive aAPL (RARA-aAPL) and 21 RARG-aAPL cases. Our analysis revealed an unexpected novel form of X::RAR::X- or X::RAR::Y-type tripartite fusions in certain RARA-aAPL and all RARG-aAPL cases, with shared features and notable differences between these 2 disease subgroups. In RARA-aAPL cases, the occurrence of RARA 3' splices was associated with their 5' fusion partner genes, mapping across the coding region of helix 11_12 (H11_12) within the ligand-binding domain (LBD), resulting in LBD-H12 or H11_12 truncation. In RARG-aAPL cases, RARG 3' splices were consistently localized to the terminus of exon 9, leading to LBD-H11_12 truncation. Significant differences were also observed between RARA and RARG 5' splice patterns. Our analysis also revealed extensive involvement of transposable elements in constructing RARA and RARG 3' fusions, suggesting transposition mechanisms for fusion gene ontogeny. Both protein structural analysis and experimental results highlighted the pivotal role of LBD-H11_12/H12 truncation in driving ATRA unresponsiveness and leukemogenesis in tripartite fusion-positive aAPL, through a protein allosteric dysfunction mechanism.