ABSTRACT: Purpose: the goal of this study is to investigate the consequences of ubiquitin specific protease 15 (USP15) knockdown on gene expression in KBM7 and K562 leukemia cells. Methods: KBM7 or K562 cells were transfected with USP15 ( siGENOME Human USP15 (9958) siRNA-SMART pool M-006066-01) or control (Ctrl) siRNAs (siGENOME Non-Targeting siRNA Control Pool#2 D-001206-14-05, Dharmacon). Cells were transfected using Lipofectamine RNAiMAX Reagent from Life Technologies following the manufacturer’s instructions and assayed at 72 hrs after transfection in Western blotting, RNA-seq and qRT-PCR. Western blotting confirmed knokdown of USP15 protein. mRNA profiles were generated by deep sequencing using Illumina Hiseq2500. Alignments of the sequence reads that passed quality filters were performed using TopHat2.1, Genome build 38 and Ensembl gtf version 77 and genecounts have been generated using Itreecount. https://github.com/NKI-GCF/itreecount. qRT–PCR validation was performed using SYBR Green assays.The amount of target, normalized to an endogenous reference (HPRT), was calculated by 2-DDCT. Results: We assigned about 30 million reads per sample uniquely to a gene of the human reference genome (hg38). We identified 26,227 genes in KBM7 and 28,329 genes in K562 cells using TopHat2.1 in combination with Itreecount (https://github.com/NKI-GCF/itreecount) workflow. Comparison of normalized gene expression data resulted in a total of 657 and 330 differential expressed genes between USP15 KD vs Ctrl in KBM7 and K562 cells (in two separate experiments), respectively, with a treshold of logFC_C1 ≥1. and adj.PVal_C1 value <0.05. Downregulation of USP15 was confirmed by Western blotting and by qRT-PCR. Ingenuity pathway analysis of USP15-dependent genes in the KBM7 and K562 datasets (cut-off logFC > 1; adjusted P value <0.05) uncovered activation of inflammation-related pathways, which involve JAK/STAT and PI3K signal transduction. In K562, we also measured significant down-modulation of TGF-beta signaling. Expression of a set of 17 genes was assessed by RT-qPCR in three independent experiments per cell line and altered expression was validated. Overlap between Top 50 significantly enriched Ingenuity Upstream Regulators in each cell line RNAseq datasets was found. The Upstream Regulators were calculated based on differential gene expression between USP15 and control siRNAs. The nine overlapping terms include: Ifnar, IRF3, STAT1, Interferon alpha, IFNG, CD40LG, IFNB1, TNF and poly rI:rC-RNA. Conclusions: Our results represent the first detailed analyis of the consequences of USP15 depletion on gene expression in malignant hematopoietic populations such as KBM7 and K562 chronic myeloid leukemia cells by genome wide expression profiling in USP15 KD vs Ctrl cells. We confirmed downregulation of USP15 and validated altered expression of a set of 17 genes by qRT-PCR. Ingenuity pathway analysis identified 657 and 330 differentially regulated genes in KBM7 and K562, respectively. The changes in gene expression, though not extensive, indicate that RNAi of USP15 led to activation of inflammation-related pathways, which involve JAK/STAT and PI3K signal transduction in both cell lines. In K562, we also measured significant down-modulation of TGF-beta signaling, which is required for USP15 pro-oncogenic role in human glioma cells (Eichhorn et al., 2012). Regulation of inflammatory signals and TGF-beta are relevant both in normal HSC and in malignant development (Blank and Karlsson, 2015).
