HIRA loss transforms FH-deficient cells

Abstract

Fumarate hydratase (FH) is a mitochondrial enzyme that catalyzes the reversible hydration of fumarate to malate in the tricarboxylic acid (TCA) cycle. Germline mutations of FH lead to hereditary leiomyomatosis and renal cell carcinoma (HLRCC), a cancer syndrome characterized by a highly aggressive form of renal cancer. Although HLRCC tumors metastasize rapidly, FH-deficient mice develop premalignant cysts in the kidneys, rather than carcinomas. How Fh1-deficient cells overcome these tumor-suppressive events during transformation is unknown. Here, we perform a genome-wide CRISPR-Cas9 screen to identify genes that, when ablated, enhance the proliferation of Fh1-deficient cells. We found that the depletion of the histone cell cycle regulator (HIRA) enhances proliferation and invasion of Fh1-deficient cells in vitro and in vivo. Mechanistically, Hira loss activates MYC and its target genes, increasing nucleotide metabolism specifically in Fh1-deficient cells, independent of its histone chaperone activity. These results are instrumental for understanding mechanisms of tumorigenesis in HLRCC and the development of targeted treatments for patients.

Fig. 1.. Fh1 loss in kidney epithelial mouse cells compromises proliferation enhancing migration and invasion.

(A) 2D growth measured using the Incucyte system of Fh1-proficient (Fh1^fl/fl), Fh1-deficient (Fh1^−/−CL1 and Fh1^−/−CL19), and Fh1-reconstituted (Fh1^−/−CL1 + pFH) cell lines (n = 4). Data were normalized to time 0. The last time point was used for the statistical comparison. (B) Cell cycle comparisons were performed identifying differences between percentage of propidium iodide staining in G₁ (physical cell growth interphase) and S (DNA synthesis) phases of cell cycle (n = 3 to 4). (C) DNA synthesis by means of BrdU incorporation into the DNA. Percentage of BrdU-positive cells relative to total nucleus number is plotted for each condition (n = 3). (D) Transwell migration of cells normalized by starting cell number (n = 4). (E) Analysis of increased spheroid area in 48 hours (n = 3). At least 20 spheroids were analyzed per experiment. Dots represent experiment 1, squares represent experiment 2, and triangles represent experiment 3. (F) Representative pictures of mice injected with Fh1-proficient (Fh1^fl/fl) and Fh1-deficient (Fh1^−/−CL1) cells (n = 5). No tumors were visible in any of the mice. Error bars represent SEM. Statistic tests performed: two-tailed Student’s T test (A, D, and E) and one-tailed Student’s t test (B and C). Numbers represent P value for all comparisons.

Fig. 2.. A genome-wide CRISPR-Cas9 screen identifies Hira loss as an oncogenic factor in Fh1-deficient cells.

(A) Schematic of the CRISPR-Cas9 screen carried out. Fh1-proficient (Fh1^fl/fl), Fh1-deficient (Fh1^−/−CL1), and Fh1-reconstituted (Fh1^−/−CL1 + pFH) cell lines expressing Cas9 were transduced with a pool mouse library containing ~90,000 guide RNAs (gRNAs) (51). Cells were grown for 18 days, and at least 90 million cells were harvested for consequent DNA extraction and high-throughput sequencing. (B) Venn diagram of the comparisons performed. Two enriched gRNAs dependent on Fh1 and plasmid expression were identified (highlighted in blue). (C) Volcano plots showing the significantly depleted and enriched gRNAs for each comparison performed (Fh1^−/−CL1 versus Fh1^fl/fl) and (Fh1^−/−CL1 versus Fh1^−/−CL1 + pFH). Blue and pink colors represent the comparison between Fh1 expression conditions. Shapes refer to whether a gene is significant in the comparison treatment condition versus plasmid. This comparison gets rid of significant depleted or enriched gRNAs depend on the corresponding gene basal expression in the cells. (D) 2D growth analysis of Fh1-deficient cells (Fh1^−/−CL1 and Fh1^−/−CL19) under Hira depletion (Fh1^−/−CL1 g1Hira and Fh1^−/−CL19g1Hira) (n = 4). Data normalized to time 0. Statistics performed comparing the values of the last time point. (E) Representation of the increase in spheroid area for 48 hours (n = 3). At least 20 spheroids were analyzed per experiment. Dots represent experiment 1, squares represent experiment 2, and triangles represent experiment 3. Error bars represent SEM. Statistic tests performed: two-tailed Student’s T test (D) and one-tailed Student’s t test (E). Numbers represent P value for all comparisons. LFC, log₂ fold change.

Fig. 3.. Hira- and Fh1-deficient cells promote tumor initiation, growth, and invasion in vivo.

(A) Scheme of xenograft injections in the flank of nude mice. Two million cells were injected in each flank (5 mice, 10 injections in total), and tumor initiation/growth was monitored for 11 weeks by In Vivo Imaging System (IVIS) bioluminescence imaging (BLI). (B and C) Xenograft tumor growth by means of average BLI and flux intensity normalized to day 0 (n = 10 tumors) of Hira- and Fh1-deficient cells (Fh1^−/−CL1 g1Hira and Fh1^−/−CL19 g1Hira). (D) Scheme of orthotopic experiments carried out. Cells were injected in the kidney capsule (n = 4 kidneys per condition). Tumor initiation, growth, and invasion were analyzed for 8 weeks by IVIS BLI. (E) Representation of average luminescence signal by means of BLI and flux intensity normalized to day 1 (day after surgery) of Fh1-deficient (Fh1^−/−CL1 Cas9) and Hira- and Fh1-deficient cells (Fh1^−/−CL1 g1Hira). (F) Representative hematoxylin and eosin images of the kidney injected with Hira- and Fh1-deficient cells. Tumors attached to the kidney capsule and invasive lesions within the kidney can be observed. Small square represents the whole sections of the kidney and adjacent tumors. Scale bars, 1 mm. (G) Analysis of FH and HIRA expression data from patients with HLRCC (25). Numbers on the bars represent P values. (H) Gene expression of HIRA in KIRP II comparing normal and primary tumor samples. Tumor samples with low FH expression are represented in blue. (I) Overall survival data associated to HIRA expression from KIRP using Gene Expression Profiling Interactive Analysis (GEPIA) (26). Low/High HIRA represents top and bottom 50%. Error bars represent SEM. Statistic test performed: two-tailed Mann-Whitney U test. Numbers represent P value for all comparisons. TPM, transcripts per million, HR, hazard ratio; FC, fold change.

Fig. 4.. Hira- and Fh1-deficient cells activate an EMT program and Myc and E2f-target signatures.

(A) Volcano plot representing the GSEA for Fh1^−/−g1Hira versus Fh1^−/− Cas9 cells. Signatures are colored depending on database represented. (B) Volcano plot representing the GSEA from the comparison between HLRCC and normal patient transcriptomic data. Signatures are colored depending on database represented. (C) Volcano plots of the genes present in the significantly up-regulated signatures in the GSEA for Fh1^−/−CL19 g1Hira versus Fh1^−/−CL19 Cas9 cells (MYC_Targets_V1 and E2F_Targets). Orange circles on the top up-regulated target-Kpna2. (D) Transcriptomic expression of FH, HIRA, and EMT genes (VIM and CDH1) and MYC targets (MYC, KPNA2, PPM1D, CCT3, MCM5, and SMC3) in HLRCC patient cohort. Numbers over the bars represent P value. (E) Lollipop graph representing the mean TF change in the transcriptomic data comparing Hira- and Fh1-deficient cells with Fh1-deficient cells alone. (F) Volcano plot with the untargeted metabolomics performed. Nucleotides up-regulated in Hira- and Fh1-deficient cells. CTP, cytidine 5′-triphosphate; GTP, guanosine 5′-triphosphate; dATP, 2′-deoxyadenosine 5′-triphosphate; dCTP, 2′-deoxycytidine 5′-triphosphate; UTP, uridine 5′-triphosphate. (G) Metabolite abundance of the nucleotides shown for Fh1-deficient (Fh1^{−/-CL1/CL19}) and Hira- and Fh1-deficient cell lines (Fh1^{−/−CL1/CL19} g1Hira) (n = 5). NES, normalized enrichment score. Error bars represent SEM. Cutoff for transcriptomic volcano plots: NES = ±0.5 and P_adj = 0.25 (=25%). Statistic test performed: two-tailed Student’s T test (G). For comparisons between Fh1-deficient cells and Hira- and Fh1-deficient cells, a paired comparison was performed. Numbers represent P value for all comparisons. a.u., arbitrary units.

Fig. 5.. The activation of Myc and E2f target signatures is independent of H3.3 deposition.

(A) Normalized ChIP-seq signal associated with MYC and E2F target signature expression for all conditions. Shadows represent the SEM. (B) Quantitative reverse transcription polymerase chain reaction showing expression levels for Myc, Kpna2, and Ppm1d in control (Fh1^fl/fl), Fh1-deficient (Fh1^−/−CL1), and Hira- and Fh1-deficient cells (Fh1^{−/−CL1/CL19} g1Hira) (n = 4). (C) Confocal representative images for Myc and nuclear (4′,6-diamidino-2-phenylindole) colocalization comparing the effect of two independent gRNAs for Hira (g1Hira and g4Hira) in Fh1-deficient cells (Fh1^−/−CL1) (n = 3). Dotted line represents Fh1-deficient cells (Fh1^−/−CL1 Cas9) as a control. β-Actin was used as a housekeeping gene. TSS, transcription starting site; TESs, transcription end sites. Error bars represent SEM. Statistic test performed: two-tailed Student’s T test. For comparisons between Fh1-deficient cells and Hira- and Fh1-deficient cells, a paired comparison was performed. Numbers represent P value for all comparisons.