A case of acute myeloid leukemia with promyelocytic features characterized by expression of a novel RARG-CPSF6 fusion

Christopher A. Miller, Christopher Tricarico, Zachary L. Skidmore, Geoffrey L. Uy, Yi-Shan Lee, Anjum Hassan, Michelle D. O’Laughlin, Heather Schmidt, Ling Tian, Eric J. Duncavage, Malachi Griffith, Obi L. Griffith, John S. Welch and Lukas D. Wartman

Data supplements

Article Figures & Data


  • Figure 1.

    Histopathologic and genomic characterization of a case of AML with promyelocytic features. (A) Wright-Giemsa staining of the peripheral blood smear highlighting promyelocytes (original magnification ×100). (B) Wright-Giemsa staining of the aspirate from the diagnostic bone marrow biopsy highlighting numerous blasts and promyelocytes (original magnification ×100). The images were captured by an Olympus BX53 microscope with an Olympus UPlanSApo 100×/1.4 oil objective and an Olympus DP26 digital camera with Olympus cellSens standard software (version 1.16; Tokyo, Japan). (C) Representative schematic of the main protein domains of RARG isoform 1 (top), CSPF6 isoform 2 (middle), and the predicted RARG-CPSF6 fusion (bottom) with the fusion breakpoints highlighted in red (for RARG at amino acid 392 and for CPSF6 at amino acid 231). Isoform choice was based on read support from the RNA-Seq data. (D) Schematic of the highly rearranged region on chromosome 12 where the reciprocal inversion occurred. In panel D, regions A and K bookend regions B-J that were involved in distinct deletions or rearrangement events. The segments are not to scale. (E) RNA-Seq reads based on a pseudo-alignment to the RARG-CPSF6 predicted fusion transcript. The coverage of reads uniquely aligned to the predicted fusion (spanning the fusion breakpoint) is displayed in red on the top track in relation to the exons for the predicted fusion (bottom track). We observed only three reads supporting wild-type RARG expression, which are likely indicative of a small number of contaminating benign cells.

  • Figure 2.

    Protein analysis and functional characterization of the RARG-CPSF6 gene fusion. (A) N-terminal and C-terminal anti-RARG antibodies were used to probe immunoblots prepared from NB4 cells (an APL cell line), a t(15;17) APL sample, and a sample from our patient. Both experiments were repeated independently with similar results. (B) Supervised hierarchical clustering of the top 500 dysregulated genes in t(15;17) APL (compared to all of the other non-t(15;17) AML cases included in the TCGA AML analysis) clusters the case (bright red) with other APLs (dark red). (C) A schematic of the experimental platform (top). The Gal4-RARG*395 truncation did not activate the UAS-GFP reporter when treated with either ATRA or BMS961 (a RARG agonist). Gal4-RARG and Gal4-RARA both activated the UAS-GFP reporter in response to ATRA, and this was not inhibited by coexpression of Gal4-RARG*395, indicating that the truncated RARG did not act as a dominant-negative against RARA or RARG in this assay (* and § P values calculated by ANOVA with Bonferroni correction for multiple comparisons). This experiment was repeated independently with similar results.