Quantitative omics analyses of NCOA4 deficiency reveal an integral role of ferritinophagy in iron homeostasis of hippocampal neuronal HT22 cells

Abstract

Introduction: Neurons require iron to support their metabolism, growth, and differentiation, but are also susceptible to iron-induced oxidative stress and cytotoxicity. Ferritin, a cytosolic iron storage unit, mediates cellular adaptation to fluctuations in iron delivery. NCOA4 has been characterized as a selective autophagic cargo receptor facilitating the mobilization of intracellular iron from ferritin. This process named ferritinophagy results in the degradation of ferritin and the consequent release of iron into the cytosol.

Methods: Here we demonstrate that NCOA4 is important for the adaptation of the HT22 mouse hippocampal neuronal cell line to cellular iron restriction. Additionally, we determined the pathophysiological implications of impaired ferritinophagy via functional analysis of the omics profile of HT22 cells deficient in NCOA4.

Results: NCOA4 silencing impaired ferritin turnover and was cytotoxic when cells were restricted of iron. Quantitative proteomics identified IRP2 accumulation among the most prominent protein responses produced by NCOA4 depletion in HT22 cells, which is indicative of functional iron deficiency. Additionally, proteins of apoptotic signaling pathway were enriched by those responsive to NCOA4 deficiency. Transcriptome profiles of NCOA4 depletion revealed neuronal cell death, differentiation of neurons, and development of neurons as potential diseases and bio functions affected by impaired ferritinophagy, particularly, when iron was restricted.

Discussion: These findings identify an integral role of NCOA4-mediated ferritinophagy in the maintenance of iron homeostasis by HT22 cells, and its potential implications in controlling genetic pathways of neurodevelopment and neurodegenerative diseases.

Keywords: IRP2; RNA-seq; ferritin; ferritinophagy; iron deficiency; proteomics.

FIGURE 1

NCOA4 post-transcriptionally responds to iron availability in HT22 mouse hippocampal neuronal cells. HT22 cells were treated with ferric ammonium citrate (Fe, 200 μM; Fe⁺, 400 μM) and desferrioxamine (DFO, 100 μM) for cellular iron overload and restriction, respectively. (A) Responses of ferritin and NCOA4 to iron availability, measured by western blot analyses. x, non-specific bands. (B) Quantitation of ferritin and NCOA4 western blots. HT22 cells were treated with Fe or DFO for 24 h (n = 4 independent experiments). Protein abundance was normalized to GAPDH. (C,D) Effects of cellular iron status on transcript abundance of Ncoa4 (C) and Tfrc (D), normalized to Tbp levels in HT22 cells (n = 3 biological replicates). Data are presented as mean ± SD (n = 3 biological replicates). *P < 0.05, compared with control levels.

FIGURE 2

HT22 neuronal cells require NCOA4 for ferritin turnover and survival during iron restriction. NCOA4 was knocked-down via liposome-mediated siRNA delivery, and DFO (100 μM) was added to induce iron deficiency. (A) siRNA-induced Ncoa4 knockdown confirmed by qPCR, normalized to Tbp levels. (B) Loss in ferritin protein by NCOA4 depletion. Ferritin quantitation was normalized to GAPDH. (C) NCOA4 deficiency impairs the repression of ferritin by iron restriction. *P < 0.05 by DFO; ^#P < 0.05 by NCOA4 depletion. (D,E) Morphological changes (D) and reduced viability (E) of NCOA4-depleted HT22 neuronal cells by iron restriction. Data presented as mean ± SD of n = 3–6 independent experiments. *P < 0.05 compared with control levels. x, non-specific band. siRNA, small interfering RNA; qPCR, quantitative real-time PCR.

FIGURE 3

Quantitative proteomics identify roles of NCOA4 in apoptosis, translation, and neuronal functioning and development. Protein expression profiles of control and NCOA4-depleted HT22 cells were compared using the tandem mass tag quantitative proteomics approach. (A) Expression of proteins with FC values above 2.0 or below -2.0, and P < 0.05. Heatmap was generated with Z-scores of Log₂-transformed normalized expression values. (B) KEGG Pathways enriched by DE proteins of NCOA4-depleted cells with |FC| > 1.5 and P < 0.05, which were identified using the DAVID functional annotation analysis. EASE Score is a modified Fisher’s exact P-value from the DAVID tool. Numbers next to each bar indicate fold enrichment for DE proteins. (C) Normalized enrichment scores (NES) of gene sets significantly enriched by proteins DE by NCOA4 depletion (P < 0.05) from GSEA. Numbers in each bar indicate P-value of enrichment by the DE proteins. (D) Effects of NCOA4 depletion on real-time OCRs after treatments of oligomycin, FCCP, and antimycin A/rotenone. Individual well OCR values were normalized to cell density. Data from n = 3 independent cultures and are presented as mean ± SD. (E) Basal respiration, ATP-coupled respiration, and spare respiratory capacity (Spare Cap) calculated from OCR measurements (D). *P < 0.05 compared with control levels. FC, fold-change; KEGG, Kyoto Encyclopedia of Genes and Genomes; DE, differentially expressed; DAVID, Database for Annotation, Visualization, and Integrated Discovery; GSEA, Gene Set Enrichment Analysis; OCR, oxygen consumption rate.

FIGURE 4

NCOA4 depletion impairs the adaptive responses of iron genes to cellular iron restriction in HT22 cells. HT22 cells were treated with Ncoa4 siRNA to knockdown NCOA4 expression. (A) Accumulation of IRP2 protein by NCOA4 depletion. (B) Schematic model of IRP2-mediated ferritin repression by NCOA4 depletion. (C) Loss in IRP2 responses to iron restriction in NCOA4-depleted cells. DFO was added at 100 μM. IRP2 quantitation was normalized to GAPDH. (D) NCOA4 depletion represses the responses of Tfrc mRNA to iron restriction (DFO, 100 μM). Transcript abundance is relative to control siRNA-treated cells without DFO, and was normalized to that of Tbp. (E) IRP2 and ferritin remain responsive to supplemental iron in NCOA4-depleted HT22 cells. Iron was added as ferric ammonium citrate at 200 μM for 24 h. Data in panels (C,D) are presented as mean ± SD, and are from n = 3–5 independent experiments. *P < 0.05 by DFO; ^#P < 0.05 by NCOA4 depletion. DFO, desferrioxamine; LIP, labile iron pool.

FIGURE 5

Transcriptome analysis reveal a protective role of NCOA4 against apoptosis induced by neuronal iron deficiency. Transcripts responsive to NCOA4 depletion by siRNA (siNCOA4), iron deficiency (DFO), and a combination of the two (siNCOA4 and DFO) were identified by RNA-seq (n = 3 biological replicates). DFO was treated at 100 μM for 24 h. (A) Numbers of differential expressions shared by or exclusive to each treatment are shown in the black Venn diagram. Transcripts with | FC| > 2.0 and FDR-adjusted P < 0.05 vs control siRNA (siControl) were considered DE. Venn diagrams in red and blue indicate the numbers of upregulated and downregulated genes, respectively, by each treatment. (B) Expression pattern of DE genes exclusively identified in cells treated with siNCOA4 plus DFO. For each gene, expression values of all treatment groups were standardized by Z-score transformation to yield a mean of zero and SD of one. (C) Comparisons among the IPA activation Z-scores of neuron-related Diseases and Bio Functions by genes responsive to siNCOA4, DFO, and siNCOA4 plus DFO. IPA core analyses was performed for DE by each treatment (vs siControl), and further analyzed by comparison analysis. Neuron-related Diseases and Bio Functions with significant activation Z-scores by siNCOA4 plus DFO vs siControl are shown (P < 0.05). (D) Relative transcript abundance of genes for cellular iron homeostasis determined by RNA-seq FPKM. Gene expression data was ranked by mean FPKM values of siControl cells, and presented as Log₁₀(FPKM + 1). *P < 0.05 by DFO. (E) Less Tfrc and Slc11a2 transcript responses to iron restriction in NCOA4-depleted cells. Data are shown as mean ± SD of FPKM from n = 3 biological replicates. *P < 0.05 by NCOA4 depletion; ns, not significantly different; siNCOA4, Ncoa4 siRNA; siControl, control siRNA; DFO, desferrioxamine; DE, differentially expressed; FC, fold-change; FDR, false discovery rate; vs, versus.