The lncRNA CASC15 regulates SOX4 expression in RUNX1-rearranged acute leukemia

Abstract

Background: Long non-coding RNAs (lncRNAs) play a variety of cellular roles, including regulation of transcription and translation, leading to alterations in gene expression. Some lncRNAs modulate the expression of chromosomally adjacent genes. Here, we assess the roles of the lncRNA CASC15 in regulation of a chromosomally nearby gene, SOX4, and its function in RUNX1/AML translocated leukemia.

Results: CASC15 is a conserved lncRNA that was upregulated in pediatric B-acute lymphoblastic leukemia (B-ALL) with t (12; 21) as well as pediatric acute myeloid leukemia (AML) with t (8; 21), both of which are associated with relatively better prognosis. Enforced expression of CASC15 led to a myeloid bias in development, and overall, decreased engraftment and colony formation. At the cellular level, CASC15 regulated cellular survival, proliferation, and the expression of its chromosomally adjacent gene, SOX4. Differentially regulated genes following CASC15 knockdown were enriched for predicted transcriptional targets of the Yin and Yang-1 (YY1) transcription factor. Interestingly, we found that CASC15 enhances YY1-mediated regulation of the SOX4 promoter.

Conclusions: Our findings represent the first characterization of this CASC15 in RUNX1-translocated leukemia, and point towards a mechanistic basis for its action.

Keywords: B-all; CASC15; ETV6-RUNX1; Non-coding RNA; SOX4.

Fig. 1

CASC15 expression is highest in acute leukemia samples that carried a translocation involving the RUNX1/AML1. a RT-qPCR of the B-ALL patient bone marrow samples using primer set #1 normalized to Actin, showing differential expression between cytogenetic subtypes based on translocations (n = 125). Comparisons were made using 1-way ANOVA (p < 0.02) and a two-tailed T-test; statistically significant differences are denoted as follows: P < 0.005 (**)). b RT-qPCR analysis of CASC15 expression in different molecular and cytogenetic subtypes of AML patient samples, using primer set #1 and normalized to GUS (n = 48). Comparisons were made using a two-tailed T-test. c Top: Schematic showing the genomic conservation of the syntenic block among vertebrates. Highly conserved region of the human CASC15 was used for the analysis. Bottom: Chip-seq histone modification map from the ENCODE/Broad institute, taken from UCSC genome browser, shows H3K4me3 and H3K36me3 patterns at the SOX4 and CASC15 locus in two different cell types indicating active transcription of the lncRNA

Fig. 2

Over-expression of CASC15 leads to increased apoptosis in B-ALL cells and myeloid expansion in mice. a RT-qPCR of mouse pre-B 70Z/3 cell line transduced with either human CASC15(s) or murine Casc15. b Annexin V staining of 70Z/3 stably transduced with CASC15(S) and mouse CASC15, either at baseline (DMSO) or treated with prednisolone (concentrations as indicated). c RT-qPCR or primary murine bone marrow cells transduced with murine Casc15. d Methylcellulose colony formation assay using MYC (positive control) or murine Casc15. e-f FACS analysis of peripheral bleed from mice 4 weeks after bone marrow transfer showing successful engraftment and transduction (GFP+) in vector (e) and CASC15 (f) over expression mice. g RT-qPCR showing Casc15 overexpression in bone marrow at 16 weeks after transplantation. h FACS analysis of peripheral bleeds from the mice 4–20 weeks after bone marrow transplantation showing increased percentage of myeloid cells in the Casc15 overexpression mice. i FACS analysis of peripheral blood showing percentage of B220+ and CD11b + cells at 16 weeks post-transplantation. j FACS analysis of bone marrow showing percentage of GFP+, B220+ and GFP+, CD11b + cells at 16 weeks post-transplantation. In experiments depicted for g-j, n = 8 mice/group, and experiments were repeated three times. Statistically significant differences are shown as follows: p < 0.05 (*); p < 0.01 (**); p ≤ 0.0005 (***). All qPCR analyses for CASC15 were performed with primer set #1 and normalized to GAPDH or actin, except where otherwise noted. Experiments were repeated three times for validation

Fig. 3

CASC15 expression is strongly correlated with and regulates SOX4 expression. a-b Correlation between CASC15 and SOX4 expression in various B-ALL and AML cell lines. c-d RT-qPCR of CASC15 and SOX4 in REH cells (c) and RS4;11 (d) following knockdown of CASC15 using three independent siRNA sequences. Statistically significant differences from control are noted as follows: p < 0.05 (*); p < 0.01 (**); p < 0.0005 (***). e-f Western blot analysis of SOX4 protein levels in RS4;11 (e) and REH (f) cell lines upon CASC15 knockdown. (g) RT-qPCR of CASC15 and SOX4 in REH cells following CRISPR/Cas9 mediated targeting. Statistically significant differences from control are denoted as follows: p < 0.05 (*); p < 0.01 (**). h RT-qPCR of murine Casc15 and Sox4 on RNA extracted from FACS-purified hematopoietic progenitor fractions shows a rough correlation in their expression, particularly in the B-cell subsets. i RT-qPCR for murine Sox4 following transduction of bone marrow cells with murine Casc15 shows upregulation of expression. All qPCR analyses for CASC15 were performed with primer set #1 and normalized to GAPDH or actin, except where otherwise noted. Experiments were repeated three times for validation. Experiments were repeated three times for validation

Fig. 4

Casc15 knockout in murine cells results in Sox4 downregulation. a Schematic of Casc15 gene in the mouse showing exonic sequences 1 and 2, along with the placement of short guide RNAs to induce deletion of these exons. b Schematic of approach to generation of genomic-level knockouts of Casc15. c FACS analysis of mCherry expression ion bulk transductants from this approach. d PCR analysis of mutant (DEL) and wild-type (WT) Casc15 alleles in bulk transductants cells. e FACS analysis of mCherry expression in single cell clones generated by limiting dilution plating. f PCR analysis of mutant (DEL) and wild-type (WT) Casc15 alleles in single cell clones generated by limiting dilution plating. g RT-qPCR analysis of Casc15 and Sox4 expression. Statistically significant differences from control are noted as follows: p < 0.05 (*); p < 0.01 (**)

Fig. 5

CASC15 knockdown leads to enrichment for the transcriptional program of SOX4 and YY1. a Unsupervised hierarchical gene clustering of differentially expressed genes upon CASC15 siRNA mediated knockdown in RS4;11 cells (PPDE >99%, fold change >2). b-g RT-qPCR confirmation of differentially expressed genes from the microarray in RS4;11 (b-d) and REH (e-g) cell lines. h To narrow down transcription factors that might be responsible for the observed changes in gene expression, we overlapped transcription factors that were associated with the differentially expressed gene sets in CASC15 KD RS4;11 cells as well as CASC15 KO REH cells. Shown are the numbers of genes in the differentially expressed gene set for each transcription factor that showed an association. i-j Enrichment plots from gene set enrichment analysis (GSEA). The differentially regulated gene set from CASC15 KD RS4;11 cells showed a positive enrichment score when compared to genes up-regulated in ACC3 cells with SOX4 knockdown (i; Enrichment Score = 0.5, FDRq = 0.0 and P value = 0.0), and showed a negative enrichment score when compared to genes downregulated in ACC3 cells with SOX4 knockdown (j; Enrichment Score = −0.38,FDRq = .017 and P value = 0.01). k-l Enrichment plots from gene set enrichment analysis (GSEA) showing that the differentially regulated gene set showed a positive enrichment score when compared with upregulated genes upon YY1 knockdown (k; Enrichment Score = 0.5, FDRq = 0 and P value = 0.0) and a negative enrichment score with downregulated genes upon YY1 knockdown (l; Enrichment Score = -0.39, FDRq = 0 and P value = 0.0)

Fig. 6

CASC15 regulates the activity of YY1 on the SOX4 promoter. a Transcriptional activity of SOX4 promoter upon CASC15 (L) and/or YY1 overexpression, as measured by dual luciferase assay. Luciferase values are normalized to the empty vector. b-d RT-qPCR with primers directed against SOX 4 (b) CASC15 (c) and YY1 (d) and Western blot for YY1 (e) in 293 T cells transiently transfected with YY1 and CASC15. All qPCR analyses for CASC15 were performed with primer set #1 and normalized to GAPDH or actin, except where otherwise noted. Experiments were repeated three times for validation