Introduction: TBLR1-RARα is the tenth fusion gene of acute promyelocytic leukemia (APL) first identified in a rare case of APL with t(3;17)(q26;q21) chromosomal translocation in our previous study. The characteristics of its basic structure and functions had been clarified in our previous study. In this study, we successfully established a novel TBLR1-RARα leukemia mouse model (TR mouse) which fully recapitulated the most relevant features of human APLs. The therapeutic effects of retinoic acid (ATRA), arsenic trioxide (As2O3), cytarabine (Ara-C) and histone deacetylase inhibitors (HDACi) on TR mice were examined. The differentially expressed genes (DEGs) between TR mice and normal mice were compared to explore the possible mechanisms and better therapeutic targets for this kind of APL.

Methods: pMSCV-TBLR1-RARα-Flag-IRES-GFP (MSCV-TR) and pMSCV-IRES-GFP (vehicle) retroviral plasmids were constructed and transfected 293T packaging cells to produce retroviruses. Lin- cells from C57BL/6 mice bone marrow were purified and infected with MSCV-TR and vehicle retroviral supernatant. For in vitro assay, the GFP+ lin- cells sorted and incubated with or without different concentrations of ATRA were analyzed for the differentiation and proliferation capacity by cell morphology, myeloid markers (CD11b and GR-1) and colony formation assay. For the in vivo experiment, GFP+ lin- cells transfected with indicated retroviral vectors were injected intravenously to lethally irradiated C57BL/6 mice to establish an APL mouse model. Cell surface markers were analyzed by flow cytometry. In treatment assays, GFP+ spleen cells from TR leukemia mice were injected intravenously into recipient mice. The mice were randomly separated into groups and received different treatment with ATRA, As2O3, As2O3 in combination with ATRA, Ara-C, Ara-C in combination with ATRA, chidamide and NL101, respectively. The percentage of GFP+ cells in peripheral blood and body weight were measured dynamically. The survival time of every group was recorded and compared. RNA-seq assay was used to identify DEGs between TR mice and normal mice.

Results: In vitro assays indicated that TBLR1-RARα could either block the differentiation of HSCs at a relatively early stage or enhanced the clonogenic potential of cells. The TBLR1-RARα leukemia mouse model was successfully established. During the ten-month observational period, 3 out of 15 mice transplanted with TBLR1-RARα expressing cells developed an APL-like disease. Development of leukemia was not observed in any of the mice in control group. All the leukemia mice had a body weight loss as well as splenomegaly and hepatomegaly. The phenotype analysis revealed that the progenitor markers Sca-1, CD34 and C-kit were positive, the myeloid lineage markers Gr-1 and CD11b were also positive, erythroid lineage marker Ter119 was weekly positive, but the lymphatic lineage marker B220, CD3,CD4 and CD8 were all negative. TR mice treated with 1.5-2.5 mg/kg ATRA alone or together with 2.0 mg/kg As2O3 didn't survive longer than that of control group, although in vitro differentiation experiment showed that the leukemia cells were sensitive to ATRA. Leukemic mice receiving Ara-C treatment had a much longer survive time. Surprisingly, HDAC inhibitors (12.5 and 25 mg/kg chidamide and 30 mg/kg NL-101) could significantly prolong the survival time of TR mice. Thousands of DEGs had been identified between TR mice and wild type mice, which were widely involved in multiple pathways and participated in various biological functions.

Conclusion: The TBLR1-RARα leukemia mouse model was successfully established for the first time, and its main characteristics were clarified. Although the leukemia cells were sensitive to ATRA in vitro, TR mice didn't benefit from ATRA or As2O3 treatment in vivo. Besides Ara-C, HDAC inhibitors, such as chidamide and NL-101 exhibited potency therapeutic values for TR mice, which provided a new strategy for this kind of refractory APL. What' more, lots of genes that might be related with the process of leukemogenesis and new therapeutic targets for TR leukemia were identified. This model would serve as a versatile tool to study the mechanisms of leukemogenesis and help to design better strategies for APLs in further studies.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.