Multi-omics analysis identifies RFX7 targets involved in tumor suppression and neuronal processes

Abstract

Recurrently mutated in lymphoid neoplasms, the transcription factor RFX7 is emerging as a tumor suppressor. Previous reports suggested that RFX7 may also have a role in neurological and metabolic disorders. We recently reported that RFX7 responds to p53 signaling and cellular stress. Furthermore, we found RFX7 target genes to be dysregulated in numerous cancer types also beyond the hematological system. However, our understanding of RFX7's target gene network and its role in health and disease remains limited. Here, we generated RFX7 knock-out cells and employed a multi-omics approach integrating transcriptome, cistrome, and proteome data to obtain a more comprehensive picture of RFX7 targets. We identify novel target genes linked to RFX7's tumor suppressor function and underscoring its potential role in neurological disorders. Importantly, our data reveal RFX7 as a mechanistic link that enables the activation of these genes in response to p53 signaling.

Fig. 1. Generation and validation of U2OS RFX7 knock-out cell lines.

a Western blot analysis of RFX7, PDCD4, PIK3IP1, p53, and actin (loading control) levels in parental and RFX7 knock-out (RFX7^−/−) U2OS cells treated with Nutlin-3a or dimethyl sulfoxide (DMSO) solvent control. Uncropped western blot images available in Supplementary Fig. S1a. b RT-qPCR data of the RFX7 target PIK3IP1 in parental and RFX7 knock-out U2OS cells treated with Nutlin-3a, Actinomycin D, 5-FU, Doxorubicin, or DMSO solvent control. MDM2 served as a positive control for p53 induction. Normalized to DMSO treatment and ACTR10 negative control. Mean and standard deviation is displayed; n = 3 technical replicates. With the exception of the 5-FU treatment in RFX7-KO2 (adj. p = 0.43), all comparisons of PIK3IP1 expression between parental cells and RFX7 knock-out lines under non-DMSO treatment conditions were significant (adj. p < 0.001). Expression was not significantly different under DMSO treatment (adj. p > 0.99). All tests have been performed using a one-way ANOVA. c Brightfield images of parental and RFX7 knock-out U2OS cells. d Cell number of parental and RFX7 knock-out U2OS cells following indicated days of cell culture. N = 4 biological replicates per time point. Statistical significance assessed by a two-sided unpaired t-test. Statistical significance displayed for clone2 compared with parental. Clone1 compared with parental displayed no significant difference.

Fig. 2. Transcriptome analysis of RFX7 knock-out U2OS cells.

a Volcano plot of differential gene expression data from RFX7 knock-out clone 1 (left panel) and 2 (right panel) compared to parental U2OS cells under Nutlin-3a treatment condition. Data has been obtained using DESeq2 and is available in Supplementary Table S1. Selected known RFX7 targets are highlighted. b Transcripts per kilobase million (TPM) expression values of PDCD4 and PIK3IP1 obtained from RNA-seq analysis from parental and RFX7 knock-out U2OS cells treated with Nutlin-3a or DMSO solvent control. Statistical significance obtained through a one-way ANOVA test, n = 3 biological replicates. c Log₂(fold-change) values of 57 established RFX7 targets [23] from RNA-seq analyses of parental and RFX7 knock-out U2OS cells treated with Nutlin-3a or DMSO solvent control. Fold-changes normalized to DMSO-treated parental U2OS cells. Mean and standard deviation is indicated.

Fig. 3. Integrative analysis data identifies novel RFX7 target genes.

a Heatmap of RNA-seq data for established and novel direct RFX7 target genes that bind RFX7 within 5 kb from their TSS according to ChIP-seq data and are significantly (FDR < 0.01) down-regulated (log₂FC ≤ –0.25) in both Nutlin-3a-treated RFX7 knock-out U2OS cell lines compared with parental U2OS cells. Significant (FDR < 0.01) p53-dependent regulation is indicated at the left. b–d Genome browser images displaying RFX7 ChIP-seq signals and predicted X-boxes at the (b) MAFTRR, (c) ARRDC3, and (d) FAM111A-DT gene loci.

Fig. 4. Proteome data validates RFX7 targets.

a Volcano plot of differential protein expression data from RFX7 knock-out clone 2 compared to parental U2OS cells under Nutlin-3a treatment condition. Data have been obtained using Spectronaut and are available in Supplementary Table S2. Selected known RFX7 targets are highlighted. b–d Fold-change expression values of protein abundance. Normalized to DMSO-treated parental U2OS cells. Statistical significance obtained through a one-way ANOVA test, n = 10 biological replicates. b The established RFX7 targets PDCD4 (left) and ABAT (right). c The established RFX7 target, cell cycle gene, and DREAM target CKS2. d The novel RFX7 targets JUN (left) and SYNPO (right). e Log₂(fold-change) values for 19 out of 57 established RFX7 targets [23] and 11 out of 33 novel RFX7 targets (Fig. 3a) from mass spectrometry analyses of RFX7 knock-out U2OS cells compared with parental U2OS cells under Nutlin-3a treatment condition. Mean and standard deviation is displayed.

Fig. 5. Generation of RPE-1 RFX7 knock-out cell lines and validation of novel RFX7 targets.

a Western blot analysis of RFX7, PDCD4, PIK3IP1, p53, and actin (loading control) levels in parental and RFX7 knock-out (RFX7^−/−) RPE-1 cells treated with Nutlin-3a or DMSO solvent control. Uncropped western blot images available in Supplementary Fig. S1b. b RT-qPCR data of the established RFX7 targets PIK3IP1, PDCD4, MXD4, and PNRC1 in parental and RFX7 knock-out RPE-1 cells treated with Nutlin-3a or DMSO solvent control. MDM2 served as a positive control for p53 induction. c RT-qPCR data of the novel RFX7 targets MAFTRR, MIR22HG, and SYNPO in parental and RFX7 knock-out RPE-1 cells treated with Nutlin-3a or DMSO solvent control. b, c Normalized to DMSO treatment and ACTR10 negative control. Mean and standard deviation is displayed. Statistical significance obtained through a one-way ANOVA test, n = 6 (two biological replicates with three technical replicates each).