Six1 regulates leukemia stem cell maintenance in acute myeloid leukemia

Abstract

Molecular genetic changes in acute myeloid leukemia (AML) play crucial roles in leukemogenesis, including recurrent chromosome translocations, epigenetic/spliceosome mutations and transcription factor aberrations. Six1, a transcription factor of the Sine oculis homeobox (Six) family, has been shown to transform normal hematopoietic progenitors into leukemia in cooperation with Eya. However, the specific role and the underlying mechanism of Six1 in leukemia maintenance remain unexplored. Here, we showed increased expression of SIX1 in AML patients and murine leukemia stem cells (c-Kit⁺ cells, LSCs). Importantly, we also observed that a higher level of Six1 in human patients predicts a worse prognosis. Notably, knockdown of Six1 significantly prolonged the survival of MLL-AF9-induced AML mice with reduced peripheral infiltration and tumor burden. AML cells from Six1-knockdown (KD) mice displayed a significantly decreased number and function of LSC, as assessed by the immunophenotype, colony-forming ability and limiting dilution assay. Further analysis revealed the augmented apoptosis of LSC and decreased expression of glycolytic genes in Six1 KD mice. Overall, our data showed that Six1 is essential for the progression of MLL-AF9-induced AML via maintaining the pool of LSC.

Keywords: MLL-rearranged leukemia; Six1; acute myeloid leukemia; glycolysis; leukemia stem cell.

Figure 1

SIX1 expression patterns and its correlation with the acute myeloid leukemia (AML) patient prognosis. A, The expression level of SIX1 in normal and malignant human blood cells (the data were obtained from bloodspot specimens); one‐way ANOVA; ***P < 0.001. B and C, The expression pattern of Six1 in murine normal and leukemic stem and progenitor‐rich cells (defined as c‐Kit⁺ cells in panel B and LSK (Lin⁻c‐Kit⁺ Sca‐1⁺) and GMP (granulocyte‐macrophage progenitors) for normal hematopoietic stem/progenitor cell (HSPC)s, and L‐GMP for leukemic stem cells in panel C, respectively). MA9, MLL‐AF9; MN3, MLL‐NRIP3; n = 3, mean ± SD; one‐way ANOVA; ***P < 0.001. D and E, Overall survival between AML patients with higher or lower expression levels of SIX1 (n = 79, GSE12417‐GPL570 for panel D and n = 163, GSE12417‐GPL97 for panel E, analyzed data from PrognoScan). F and G, Overall survival between AML patients with (F) or without (G) MLL‐r with higher or lower expression levels of SIX1 (n = 30, for panel F and n = 115, for panel G analyzed data from TCGA). For panels D‐G, Mantel‐Cox test; minimal P‐value: *P < 0.05; **P < 0.01

Figure 2

Establishment of the Six1‐knockdown acute myeloid leukemia (AML) murine model. A, Schematic outline of the structure of the vector used for knockdown (KD) Six1. B, Diagram establishing the mixed‐lineage leukemia (MLL) fusion‐induced and Six1‐KD MLL‐AF9 AML mouse models. C, Quantitative RT‐PCR analysis of the KD efficiency of Six1 with 2 different shRNA (sh‐Six‐a and sh‐Six‐b) compared with scramble in GFP⁺BFP⁺ bone marrow (BM) cells; n = 3, mean ± SD; one‐way ANOVA; ***P < 0.001. D, Western blot analysis of the GFP⁺BFP⁺ cells from the secondary recipient mouse BM. E, Densitometry of the amount of Six1 in panel D (normalized to β‐tubulin, quantified by ImageJ). n = 3, mean ± SD; one‐way ANOVA; ***P < 0.001

Figure 3

Prolonged survival and reduced leukemic cell burden in Six1‐KD mice. A‐C, Peripheral blood (PB) count analysis of moribund secondary recipient mice transplanted with primary Six1‐KD acute myeloid leukemia (AML) cells or controls; A, white blood cells (WBC); B, red blood cells (RBC); C, platelets (PLT); n = 6, mean ± SD; one‐way ANOVA; *P < 0.05; **P < 0.01. D, Kaplan‐Meier survival curve of the secondary recipient mice (2 × 10⁴ GFP⁺BFP⁺ primary leukemic cells per group). The median survival was 20 d of control, 23 d of sh‐Six‐a groups and 26.5 d of sh‐Six‐b groups, respectively; n = 6, Mantel‐Cox test; **P < 0.01; ***P < 0.001. E, Representative image of spleens from secondary recipient mice. Scale bar = 1 cm. F, Spleen (SP) weight of secondary recipient mice. G, Spleen cellularity from secondary recipients. H, Bone marrow (BM) cellularity from secondary recipients. For panels F‐H, n = 6, mean ± SD; one‐way ANOVA, *P < 0.05. I, Wright‐Giemsa staining of the cytospin of BM cells from secondary recipients. Arrowheads show segmented nuclei in AML cells; Scale bar = 20 μm. n = 6, 2 independent experiments for Figure 3

Figure 4

Decreased frequency and absolute number of leukemic stem cells (LSC) in Six1‐KD mice. A and B, c‐Kit⁺ frequency (A) and absolute number (B) among GFP⁺BFP⁺ bone marrow (BM) leukemic cells. C and D, L‐GMP (Lin⁻IL‐7R⁻Sca‐1⁻c‐Kit⁺CD16/32⁺CD34⁺) frequency (C) and absolute number (D) among GFP⁺BFP⁺ BM leukemic cells. E and F, L‐GMP frequency (E) and absolute number (F) among GFP⁺BFP⁺ SP leukemic cells. For A‐D, the leukemic cells were collected from the femur, tibia and ilium, and live cells were counted with trypan blue exclusion; For A‐F, n = 5, 2 independent experiments; mean ± SD; one‐way ANOVA, **P < 0.01; ***P < 0.001. G, Colony‐forming assay of leukemic BM cells. In total, 500 live cells were plated per well in a 24‐well plate. n = 4, mean ± SD; one‐way ANOVA, *P < 0.05; **P < 0.01; ***P < 0.001. H and I, Limiting dilution assay analyzing the frequency of functional leukemia stem cells (LSC) in Six1‐KD and scramble groups. Logarithmic plot showing the percentage of negative recipients transplanted with different cell doses of GFP⁺ BM cells isolated from secondary recipients. Frequencies of LSC were evaluated on the website http://bioinf.wehi.edu.au/software/elda/

Figure 5

Increased apoptosis and altered glycolytic gene expression of leukemic stem cells (LSC) in Six1‐KD mice. A, Representative flow cytometric gating of the apoptosis of c‐Kit^high leukemic cells. B and C, Quantification of apoptotic c‐Kit^high bone marrow (BM) (B) or spleen (SP) (C) leukemic cells, and Annexin V⁺7‐AAD⁻ represents early apoptosis, while Annexin V⁺7‐AAD⁺ represents late apoptosis; n = 5, mean ± SD; one‐way ANOVA, *P < 0.05; **P < 0.01; ***P < 0.001. D, Western blot analysis of the Caspase 3 and cleaved Caspase 3 level in the BM cells from Six1‐KD and control cells; n = 3. E, Quantitative RT‐PCR of classic genes involved in leukemia initiation and/or maintenance. F, Quantitative RT‐PCR analysis of glycolytic genes and Six1. G, Cellular level of pyruvate in Six1‐KD and control cells. For panels E‐G, n = 3, one‐way ANOVA, *P < 0.05; **P < 0.01; ***P < 0.001