NCOA2 promotes the return of hematopoietic stem cells to quiescence after irradiation stress by regulating FOXO3a‐dependent mitophagy

Abstract

Hematopoietic stem cells (HSCs) are responsible for replenishing blood cells under stress conditions through increasing proliferation and differentiation. After the hematopoietic function has reconstructed, HSCs must re‐enter a quiescent state to avoid their depletion, whereas the underlying mechanisms remain to be elucidated. Here, we show that the translocation of nuclear receptor coactivator 2 (NCOA2) into the nucleus is gradually increased in HSCs during the hematopoietic recovery phase after sub‐lethal dose irradiation (IR). Although deletion of NCOA2 only slightly affects the steady state hematopoiesis, its deficiency leads to HSC pool exhaustion and delayed hematopoietic recovery after exposure to IR. Further investigations reveal that loss of NCOA2 decreases the quiescence, survival, and long‐term reconstituting ability of HSCs following IR due to increased mitochondria‐derived oxidative stress. Mechanistically, NCOA2 promotes the clearance of activated or damaged mitochondria by coactivating FOXO3a‐dependent transcription of PINK1, which drives HSCs to return to quiescence after being activated by IR stress. Collectively, our findings demonstrate a critical role of NCOA2 in facilitating the restoration of HSC homeostasis after IR via the FOXO3a‐PINK1‐mediated mitophagy axis and thus provide an additional strategy to prevent hematopoietic failure induced by IR.

INTRODUCTION

Hematopoietic stem cells (HSCs) reside at the apex of the hematopoietic hierarchy and can generate all blood cell lineages throughout the lifetime. ¹ , ² Under normal conditions, HSCs are mainly retained in a quiescent status, which is required to promote their maintenance. ³ , ⁴ In response to stresses induced by irradiation (IR), chemotherapy, or blood loss, HSCs can be reversibly activated and then rapidly proliferate and differentiate to reconstruct hematopoiesis.5, 6, 7 Of note, cycling HSCs will return to dormancy to preserve their self‐renewal potential when the stress injury is repaired. ⁸ The insufficient ability to re‐enter quiescence may result in HSC exhaustion and myelosuppression. ⁹ At present, numerous studies have clarified the regulatory networks governing HSC cell cycle under homeostasis or during transition from a steady state to a stress state.10, 11, 12 However, the intrinsic mechanisms that terminate HSCs' activation and re‐entry quiescence after stress are still not fully understood.

It was reported that cellular metabolic status is closely associated with HSC quiescence and activation. ¹³ , ¹⁴ Dormant HSCs primarily locate in the hypoxic bone marrow (BM) niche, where they rely on anaerobic glycolysis for energy supply. ¹⁵ , ¹⁶ Following stress‐induced injury, mitochondrial metabolism is highly activated in HSCs, characterized by increased biogenesis, oxidative phosphorylation (OXPHOS), and reactive oxygen species (ROS) levels. ¹⁴ , ¹⁷ This metabolic activation provides the necessary energy support for HSC proliferation and differentiation, whereas continuous activation may cause damage to mitochondria themselves. ⁹ During the hematopoietic recovery phase after stress, how the activated or damaged mitochondria are removed and whether they are involved in the process of HSC return to dormancy are still unclear.

Nuclear receptor coactivator 2 (NCOA2; also known as SRC‐2/GRIP1/TIF2) is a member of the p160 SRC coactivator family. ¹⁸ Studies have shown that NCOA2 regulates target gene transcription through coactivation of nuclear receptors or certain transcription factors, thereby participating in diverse biological processes.19, 20, 21, 22 In particular, NCOA2 functions as a key regulator of energy metabolism, and its abnormal expression has been implicated in metabolism‐related diseases. ²³ , ²⁴ For instance, NCOA2 is known to enhance the expression of G6Pase by coactivating with RORα, eventually influencing glucose metabolism. ²⁵ NCOA2 also interacts with MEF2, GATA‐4 and Tbx5 to mediate cardiac function and metabolism. ²⁶ Despite its established roles in cellular metabolism, whether NCOA2 can modulate HSC function via specific metabolic processes remains unexplored.

In this study, through establishing a sub‐lethal IR stress injury model, we found that the NCOA2 nuclear translocation was evidently increased in HSCs during the late stage of hematopoietic recovery after IR. Importantly, knockout of NCOA2 significantly decreased the HSC pool and impaired their survival and function due to the inability to return to quiescence following IR stress. The mechanism study showed that NCOA2 acted as a coactivator of FOXO3a to regulate PINK1‐mediated mitophagy, therefore facilitating HSC maintenance after exposure to IR. In summary, our findings uncover a novel role of NCOA2 in HSC biology and provide a deeper insight into the radioprotection of the hematopoietic system.

MATERIALS AND METHODS

Animals

NCOA2 knockout mice (NCOA2^−/−) were commercially acquired from Shanghai Model Organisms (China), and littermate wild‐type (WT) mice were used as controls. C57BL/6 SJL mice (CD45.1) were kindly bred and maintained at our research institute. All the mice in the study were male and 8–10 weeks old. The animal experiment was approved by the Animal Ethics Committee of the Third Military Medical University (Chongqing, China).

IR or 5‐fluorouracil treatment

Mice were exposed to ⁶⁰Co γ‐ray at the Irradiation Center of our research institute. For 5‐fluorouracil (5‐FU) treatment, mice were intraperitoneally injected with a single dose of 5‐FU (150 mg/kg; #F6627, Sigma, St. Louis, MO, USA).

Flow cytometry

Single‐cell suspensions from mouse BM and peripheral blood (PB) were prepared according to the previous protocols. ²⁷ , ²⁸ For hematopoietic cell phenotype detection or sorting, samples were stained with the following antibodies: lineage cocktail (#88‐7772‐72, 22‐7770‐72), c‐Kit (#47‐1171‐82, 25‐1171‐82), Sca‐1 (#45‐5981‐82, 17‐5981‐82), Flk2 (#17‐1351‐82, #12‐1351‐82), CD34 (#50‐0341‐82, 11‐0341‐82, and 56‐0341‐82), CD16/32 (#48‐0161‐82), CD127 (#12‐1271‐82), CD150 (#115904), CD48 (#17‐0481‐82), B220 (#12‐0452‐82), CD3e (#47‐0031‐82), Gr‐1 (#11‐5931‐82), CD11b (#45‐0112‐82), CD45.1 (#48‐0453‐82), and CD45.2 (#25‐0454‐82). The above antibodies were purchased from eBioscience (San Diego, CA, USA) or Biolegend (San Diego, CA, USA). For cell cycle, apoptosis, and 5‐bromodeoxyuridine (BrdU) incorporation analysis in HSCs, a series of antibodies were used: anti‐Ki67 (#25‐5698‐82, eBioscience), anti‐Annexin V (#640937, Biolegend), and anti‐BrdU (#559619, #560809, BD Biosciences, San Diego, CA, USA), as described previously. ¹² , ²⁷ To test intracellular proteins expression, the primary and secondary antibodies were used as follows: anti‐NCOA2 (#sc‐365827, Santa Cruz, Dallas, TX, USA), anti‐PINK1 (#ab23707, Abcam, Cambridge, UK), anti‐FOXO3a (#12829, Cell Signaling Technology, Danvers, MA, USA), anti‐Phospho‐FOXO3a^Ser253 (#9466, Cell Signaling Technology), and Goat anti‐Rabbit immunoglobulin G (IgG), Alexa Fluor 488 (#A‐11034, Invitrogen, Carlsbad, CA, USA). Flow cytometric analysis was performed by ID7000 (Sony Biotechnology, Japan), FACSverse (BD Biosciences), or FACSymphony^TM A1 (BD Biosciences), and the data were analyzed using the FlowJo v.10.0 software (TreeStar, San Carlos, CA, USA). Cell sorting was conducted using a FACSAria III sorter (BD Biosciences). The number of BM and hematopoietic stem/progenitor cells (HSPCs) per femur and tibia was counted in mice.

Transplantation assays

For the noncompetitive transplantation, BM cells (1 × 10⁶) from WT or NCOA2^−/− mice (CD45.2) after 5.0 Gy IR were intravenously injected into lethally irradiated (10.0 Gy) CD45.1 recipients. After 16 weeks, BM cells (1 × 10⁶) isolated from the primary recipients were transplanted into another lethally irradiated CD45.1 recipients. For competitive transplantation, long‐term HSCs (LT‐HSCs) (5 × 10²) purified from the BM of WT and NCOA2^−/− mice (CD45.2) at steady state or after 5.0 Gy IR, along with the BM cells (5 × 10⁵) from competitor mice (CD45.1), were transplanted into lethally irradiated CD45.1 recipients. After 16 weeks, BM cells (1 × 10⁶) isolated from the primary recipients were transplanted into another lethally irradiated CD45.1 recipients. For reciprocal BM transplantation (BMT), BM cells (1 × 10⁶) obtained from WT mice (CD45.1) were transplanted into lethally irradiated WT or NCOA2^−/− mice, which were pre‐treated with 5.0 Gy IR. Then, the percentage of donor‐derived cells in recipients was analyzed at the indicated times post transplantation.

Homing assay

LSKs (1 × 10⁵) obtained from WT or NCOA2^−/− mice after 5.0 Gy IR were stained with 5‐ and 6‐carboxyfluorescein succinimidyl ester (CFSE, #65‐0850‐84, Invitrogen). Additional procedures were performed according to a previous study. ²⁷

HSC serial replating assay

LT‐HSCs (1 × 10²) were sorted from WT and NCOA2^−/− mice after 5.0 Gy IR, and then plated into MethoCult^TM GF M3434 medium (#03434, Stem Cell Technologies, Vancouver, Canada). At Day 7 after culture, colonies were counted, and then, 1 × 10⁴ culture‐derived cells were harvested for replating. A similar protocol was used for the third round of replating and colonies counting.

Quantitative polymerase chain reaction

The assay was conducted as previously described, ⁶ and primer sequences are provided in Table S1.

Bulk RNA sequencing

Total RNA was isolated from WT or NCOA2^−/− HSCs after 5.0 Gy IR exposure. Then, RNA extraction and quality control were carried out at Sinotech Genomics Co., Ltd (Shanghai, China). Genes with fold change > 1.5 and P < 0.05 were defined as differential expressed. Additional methods and gene set enrichment analysis (GSEA) in this study were based on a previous publication. ¹² All of the raw data have been deposited in the Sequence Read Archive (SRA) database (no. PRJNA1302003).

Single‐cell RNA sequencing

LSKs were sorted from normal WT mice and immediately subjected to single‐cell RNA sequencing (scRNA‐seq) using the 10× Genomics platform at OE Biotech. Co., Ltd. (Shanghai, China). Data were analyzed using the Seurat integration method (https://satijalab.org/seurat/articles/integration_introduction.html), as described. ²⁹ Cell clustering was performed by matching to a published HSPC scRNA‐seq dataset. ³⁰ All of the raw data have been deposited in the Gene Expression Omnibus (GEO) database (no. GSE306557).

Mitochondrial mass, mitochondrial membrane potential, and ROS analysis

BM cells isolated from WT or NCOA2^−/− mice after 5.0 Gy IR were stained with HSC surface markers. Afterwards, the samples were incubated with MitoTracker Green (MTG; #M46750, Invitrogen), tetramethylrhodamine methyl ester (TMRM; #I34361, Invitrogen), dichlorodihydrofluorescein diacetate (DCFH‐DA; #D6883, Sigma), or MitoSox (#M36009, Invitrogen) as previously described. ²⁹ , ³¹

Mitophagy flux detection

BM cells were stained with HSC surface markers. Then, mitophagy flux was detected using the Mitophagy Detection Kit (#MD01, Dojindo, Kumamoto, Japan) according to the manufacturer's instructions, followed by flow cytometric analysis.

Seahorse analysis

The oxygen consumption rate (OCR) of HSCs was measured using the Agilent Seahorse XFp Cell Mito Stress Test Kit (#103022, Seahorse Bioscience, Billerica, MA), as previously described. ²⁷ , ³¹

Immunofluorescence

Freshly sorted LT‐HSCs were plated onto poly‐l‐lysine‐coated slides. After fixation and permeabilization, the samples were stained with anti‐NCOA2 (#sc‐365827, Santa Cruz), anti‐TOMM20 (#ab186735, Abcam), anti‐Parkin (#39‐0900, Invitrogen), or anti‐LAMP1 (#MA1‐164, Invitrogen) antibodies. Subsequently, the cells were incubated with appropriate fluorescein‐conjugated secondary antibodies and DAPI (#D9542, Sigma). Images were acquired using a Zeiss LSM800 confocal microscope (Carl Zeiss, Jena, Germany). The nuclear translocation of NCOA2 was analyzed by Image J software (National Institutes of Health, Bethesda, MD, USA), and the relative colocalization of TOMM20 with LAMP1 or Parkin was analyzed using Fiji software (National Institutes of Health).

Co‐immunoprecipitation

Lin⁻ cells derived from WT mice after 5.0 Gy IR were lysed in IP buffer. Immunoprecipitation was performed using the Dynabeads™ Protein G Immunoprecipitation Kit (#10007D, Invitrogen). After that, the protein samples were detected by Western blot using anti‐NCOA2 (#NUCR2‐101AP, Invitrogen) or anti‐FOXO3a (#MA5‐14932, Invitrogen) antibodies.

Chromatin immunoprecipitation

LSKs (4–5 × 10⁵) were sorted from WT or NCOA2^−/− mice after 5.0 Gy IR. Subsequently, chromatin immunoprecipitation (ChIP) assay was conducted using a NovoNGS® CUT&Tag 4.0 High‐Sensitivity Kit (#N259‐YH01‐01A, Novoprotein Scientific, Inc., Shanghai, China). The relative binding of FOXO3a to the PINK1 promoter region was detected by quantitative polymerase chain reaction (qPCR). The primer sequences are provided in Table S2.

Lentiviral transduction

LT‐HSCs were sorted from WT or NCOA2^−/− mice after 5.0 Gy IR and then were infected with lentivirus carrying PINK1 or control according to our previous studies. ²⁷ , ³² Two days later, the mitophagy, proliferation, and apoptosis were analyzed in transduced cells (GFP⁺) by flow cytometry. Meantime, transduced cells were purified by flow cytometry and then used for a serial replating assay.

Metformin or N‐acetyl‐L‐cysteine treatment

The irradiated WT or NCOA2^−/− mice were given Metformin (250 mg/kg; #317240, Sigma) and vehicle by oral gavage every 2 days for a successive 28 days. For N‐acetyl‐L‐cysteine (NAC) administration, irradiated WT or NCOA2^−/− mice were treated with NAC (100 mg/kg; #A9165, Sigma) and vehicle via intraperitoneal injection once a day for 28 days.

Statistical analysis

Data analysis was performed by GraphPad Prism 8.0.2 (La Jolla, CA, USA). Statistical comparisons between two groups were carried out using an unpaired Student's t‐test, while multi‐group analyses were employed using one‐way analysis of variance (ANOVA) with Tukey's test. Mouse survival curves were compared by the log‐rank test. All experiments were independently repeated at least three times and displayed as mean ± standard deviation (SD). P < 0.05, P < 0.01, and P < 0.001 were considered statistically significant.

RESULTS

NCOA2 knockout slightly affects homeostatic hematopoiesis

It was reported that NCOA2 is essential for T cell activation, Treg cell differentiation, and macrophage polarization,33, 34, 35 whereas its role in HSCs is still unknown. By analyzing the public database, we observed that NCOA2 was also abundantly present in HSPCs in mice (Figure S1A). To strengthen this finding, we performed a scRNA‐seq on purified LSKs, which are enriched with HSPCs. Consistently, the results showed that HSPCs, especially LT‐HSCs and multipotent progenitor 1 (MPP1), expressed high levels of NCOA2 (Figure 1A,B). A previous study has discovered that another SRC family member, NCOA3, plays a critical role in sustaining HSC quiescence. ²⁷ Based on these findings, we then used NCOA2^−/− mice for subsequent experiments (Figure S1B). The knockout efficiency of NCOA2 in HSPCs was verified at both mRNA and protein levels (Figure S1C,D). Consistent with a previous study, adult NCOA2^−/− mice were viable and did not display apparent abnormalities in development at the gross levels. ³⁶ It was found that knockout of NCOA2 increased the cellularity of spleen rather than thymus in mice (Figure 1C,D). Meantime, NCOA2 deficiency modestly reduced the percentage of T cells and enhanced the percentage of myeloid cells in the PB, while the white blood cell (WBC), red blood cell (RBC), and platelet (PLT) counts were comparable between WT and NCOA2^−/− mice (Figures 1E,F and S1E–G). Further flow cytometric analysis displayed that loss of NCOA2 did not significantly alter the percentages and numbers of HSPCs in the BM (Figure 1G–L). Similarly, transplantation assays exhibited that NCOA2 knockout did not significantly affect HSC function (Figure S1H,I). The results indicate that NCOA2 deletion only mildly affects hematopoiesis under homeostatic conditions.

Figure 1.

Nuclear receptor coactivator 2 (NCOA2) knockout slightly affects homeostatic hematopoiesis. (A) Single‐cell RNA sequencing (scRNA‐seq) analysis of LSKs freshly sorted from the bone marrow (BM) of normal wild‐type (WT) mice. Hematopoietic stem/progenitor cell (HSPC) clustering was shown by the t‐distributed stochastic neighbor embedding (tSNE) plot. (B) Violin plots revealing NCOA2 expression in HSPC populations from (A). (C, D) The total cell numbers of (C) spleen and (D) thymus in WT and NCOA2^−/− mice at steady state (n = 6). (E) Representative flow cytometric plots revealing the percentages of T cells, B cells, and myeloid cells in the peripheral blood (PB) of WT and NCOA2^−/− mice at steady state. (F) The percentages of T cells, B cells, and myeloid cells in the PB of WT and NCOA2^−/− mice at steady state (n = 6). (G) The BM numbers of WT and NCOA2^−/− mice at steady state (n = 6). (H) Representative flow cytometric plots revealing the percentages of myeloid progenitors (MPs, Lin⁻ Sca1⁻ c‐Kit⁺), LSKs (Lin⁻ Sca1⁺ c‐Kit⁺), long‐term hematopoietic stem cells (LT‐HSCs, Lin⁻ Sca1⁺ c‐Kit⁺ CD34⁻ Flk2⁻), short‐term HSCs (ST‐HSCs, Lin⁻ Sca1⁺ c‐Kit⁺ CD34⁺ Flk2⁻), multipotent progenitors (MPPs, Lin⁻ Sca1⁺ c‐Kit⁺ CD34⁺ Flk2⁺), and signaling lymphocyte activation molecules‐labeled HSCs (SLAM‐HSCs, Lin⁻ Sca1⁺ c‐Kit⁺ CD150⁺ CD48^‐) in the BM of WT and NCOA2^−/− mice at steady state. (I, J) The numbers of (I) MPs, LSKs, (J) LT‐HSCs, ST‐HSCs, MPPs, and SLAM‐HSCs in the BM of WT and NCOA2^−/− mice at steady state (n = 6). (K) Representative flow cytometric plots revealing the percentages of common myeloid progenitors (CMPs, Lin⁻ Sca1⁻ c‐Kit⁺ CD16/32⁻ CD34⁺), megakaryocyte erythroid progenitors (MEPs, Lin⁻ Sca1⁻ c‐Kit⁺ CD16/32⁻ CD34⁻), granulocyte monocyte progenitors (GMPs, Lin⁻ Sca1⁻ c‐Kit⁺ CD16/32⁺ CD34⁺), and common lymphoid progenitors (CLPs, Lin⁻ CD127⁺ Sca1^med c‐Kit ^med) in the BM of WT and NCOA2^−/− mice at steady state. (L) The numbers of CMPs, MEPs, GMPs, and CLPs in the BM of WT and NCOA2^−/− mice at steady state (n = 6). NS, not significant; *P < 0.05, **P < 0.01, and ***P < 0.001.

NCOA2 deficiency leads to HSC pool exhaustion following IR

We then set out to examine whether NCOA2 is involved in the regulation of emergency hematopoiesis. Through establishing the total body IR mouse model, we observed that the expression of NCOA2 was unchanged in HSCs of mice subjected to 5.0 Gy IR (Figure S2A,B). Interestingly, the nuclear translocation of NCOA2 was elevated especially at the late stage of hematopoietic recovery after IR, and then gradually returned to normal levels (Figure 2A,B), which indicates that NCOA2 may influence HSC function during stress. In fact, NCOA2 ablation delayed the recovery of hematopoiesis following 5.0 Gy IR, yet it did not induce mortality in mice (Figures 2C–E and S2C). In view of the peak nuclear translocation of NCOA2 and the significant difference in blood cell counts between WT and NCOA2^−/− mice at Day 28 after exposure to IR, we selected this time point for subsequent analysis of HSPC phenotype. Notably, the number of HSPCs, including myeloid progenitors (MPs), LSKs, LT‐HSCs, short‐term HSCs (ST‐HSCs), and MPPs, was evidently reduced in irradiated NCOA2^−/− mice, despite there was a modest decrease in the BM cell number (Figure 2F–I). The loss of HSC number was further confirmed by using SLAM (CD150 and CD48) markers (Figure 2F–I). Concurrently, the committed hematopoietic progenitor cells, including common myeloid progenitors (CMPs), megakaryocyte erythroid progenitors (MEPs), granulocyte monocyte progenitors (GMPs), and common lymphoid progenitors (CLPs), were also diminished after IR when NCOA2 was knocked out (Figure 2J,K). However, we found that deletion of NCOA2 did not affect HSPC phenotype after 5‐FU stress (Figure S2D,E). These results suggest that NCOA2 is required for the maintenance of the HSC pool, especially upon IR stress.

Figure 2.

Nuclear receptor coactivator 2 (NCOA2) deficiency leads to hematopoietic stem cell (HSC) pool exhaustion following irradiation (IR). (A) Immunofluorescence analysis of NCOA2 nuclear translocation in HSCs at the indicated time points after 5.0 Gy IR. (B) Quantitative analysis of the ratio of nuclear/whole‐cell fluorescence intensity in (A) by Image J (n = 20 cells). “Day 0” in (A) and (B) represents the unirradiated control group. (C–E) The counts of (C) white blood cell (WBC), (D) red blood cell (RBC), and (E) platelet (PLT) in the peripheral blood (PB) of wild‐type (WT) and NCOA2^−/− mice at the indicated time points after 5.0 Gy IR (n = 10). (F) The bone marrow (BM) numbers of WT and NCOA2^−/− mice at Day 28 following 5.0 Gy IR (n = 6). (G) Representative flow cytometric plots revealing the percentages of myeloid progenitors (MPs), LSKs, long‐term HSCs (LT‐HSCs), short‐term HSCs (ST‐HSCs), multipotent progenitors (MPPs), and signaling lymphocyte activation molecules‐labeled HSCs (SLAM‐HSCs) in the BM of WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR. (H, I) The numbers of (H) MPs, LSKs, (I) LT‐HSCs, ST‐HSCs, MPPs, and SLAM‐HSCs in the BM of WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR (n = 6). (J) Representative flow cytometric plots revealing the percentages of common myeloid progenitors (CMPs), megakaryocyte erythroid progenitors (MEPs), granulocyte monocyte progenitors (GMPs), and common lymphoid progenitors (CLPs) in the BM of WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR. (K) The numbers of CMPs, MEPs, GMPs, and CLPs in the BM of WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR (n = 6). *P < 0.05, **P < 0.01, and ***P < 0.001.

Loss of NCOA2 decreases the return of HSCs to quiescence and inhibits their survival after IR

It has been well established that the cell cycle status is closely linked to the behavior of HSCs. ¹⁰ , ³⁷ As expected, Ki67 and Hoechst33342 staining showed that there were no obvious changes in the cell cycle of HSCs from WT and NCOA2^−/− mice under homeostasis and 5‐FU treatment conditions (Figure S3A,B). Importantly, compared to WT controls, the frequency of G0 phase was distinctly decreased, and the frequency of G1 and S/G2/M phases was increased in NCOA2^−/− HSCs at the recovery stage of IR‐induced stress (Figure 3A). Similar results were obtained from the in vivo BrdU incorporation assay (Figures 3B and S3C,D). In line with these data, the cell cycle‐related genes (Ccnd1, Ccnd2, Ccnd3, Ccne1, and Ccne2) were upregulated, and cyclin‐dependent kinase inhibitors (Cdkn1a and Cdkn1c) were downregulated in HSCs from irradiated NCOA2^−/− mice (Figure 3C). On the other hand, NCOA2^−/− HSCs displayed a significant increase in the apoptosis rate after IR exposure, whereas this was unchanged at steady state and after 5‐FU stress (Figures 3D and S3E,F). Moreover, both WT and NCOA2^−/− mice were subjected to 7.5 Gy IR, which can cause the death of mice by severely damaging the hematopoietic system. ³⁸ Consequently, NCOA2 deletion weakens the survival rate of mice after 7.5 Gy IR (Figure 3E). The above observations demonstrate that NCOA2 knockout results in the failure of HSCs to re‐enter quiescence and concomitantly decreases HSC survival after IR.

Figure 3.

Loss of nuclear receptor coactivator 2 (NCOA2) decreases the return of hematopoietic stem cells (HSCs) to quiescence and inhibits their survival after irradiation (IR). (A) Cell cycle analysis of LSKs and long‐term HSCs (LT‐HSCs) in the bone marrow (BM) of wild‐type (WT) and NCOA2^−/− mice at Day 28 following 5.0 Gy IR (n = 6). Representative flow cytometric plots are shown on the left. (B) The percentage of bromodeoxyuridine (BrdU)⁺ cells in LSKs and LT‐HSCs from the BM of WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR (n = 6). Representative flow cytometric plots are shown on the left. (C) Quantitative polymerase chain reaction (qPCR) analysis of the mRNA expression of the cell cycle‐associated genes in LT‐HSCs from the BM of WT and NCOA2^−/− mice at Day 28 following 5.0 Gy IR (n = 3). (D) The apoptosis rate of LSKs and LT‐HSCs in the BM of WT and NCOA2^−/− mice at Day 28 following 5.0 Gy IR (n = 6). Representative flow cytometric plots are shown on the left. (E) The survival rates of WT and NCOA2^−/− mice after subjected to 7.5 Gy IR (n = 10). *P < 0.05, **P < 0.01, and ***P < 0.001.

NCOA2 ablation aggravates the impairment in HSC long‐term reconstitution ability after exposure to IR

As reported, the disruption of quiescence maintenance usually has an adverse impact on the function of HSCs. ⁸ To this end, we first conducted a noncompetitive transplantation assay and found that mice reconstructed with NCOA2^−/− BM cells deceased earlier than control mice (Figure 4A–C). Subsequently, the competitive transplantation assay was performed to determine the effects of NCOA2 deficiency on the long‐term reconstitution ability of HSCs after IR (Figure 4D). Compared with WT controls, recipients transplanted with IR‐exposed NCOA2^−/− HSCs exhibited lower levels of chimerism in the PB, and the discrepancy was gradually enlarged from 4 to 16 weeks after primary and secondary transplantation (Figure 4E,F). Consistently, impaired donor reconstruction was also found in the BM subpopulations of recipients transplanted with NCOA2^−/− HSCs (Figure 4G,H). In addition, NCOA2‐null HSCs exhibited myeloid‐biased differentiation (Figure S4A,B), which is in accordance with the NCOA2^−/− mice themselves.

Figure 4.

Nuclear receptor coactivator 2 (NCOA2) ablation aggravates the impairment in hematopoietic stem cell (HSC) long‐term reconstitution ability after exposure to irradiation (IR). (A) The schematic of a noncompetitive transplantation assay. (B, C) The survival rates of recipient mice after (B) primary and (C) secondary transplantation (n = 10). (D) The schematic of the competitive transplantation assay. (E) Representative flow cytometric plots revealing the percentage of donor‐derived cells in the peripheral blood (PB) of recipient mice at 16 weeks after primary and secondary transplantation. (F) The percentage of donor‐derived cells in the PB of recipient mice at 4, 8, 12, and 16 weeks after primary and secondary transplantation (n = 6). (G, H) The percentages of donor‐derived bone marrow (BM) cells, Lin⁻, LSKs, and long‐term HSCs (LT‐HSCs) in the recipients at 16 weeks after (G) primary and (H) secondary transplantation (n = 6). (I) The strategy of serial replating assay. (J) Serial colony‐forming analysis of LT‐HSCs (1 × 10²) isolated from wild‐type (WT) or NCOA2^−/− mice at Day 28 after 5.0 Gy IR (n = 6). (K) The strategy of reciprocal BM transplantation (BMT). (L) The percentage of donor‐derived cells in the PB of recipient mice at 16 weeks after reciprocal BMT (n = 6). **P < 0.01, ***P < 0.001.

Given that NCOA2 was systematically deleted in mice, we then isolated HSCs to assess their function in vitro, which could eliminate the interference from the BM microenvironment. Compared with WT HSCs, NCOA2‐null HSCs generated a smaller total number of colonies after IR, although they modestly tended to form colony‐forming unit‐granulocyte/macrophage (CFU‐GM) colonies (Figures 4I,J and S4C). After the second and third rounds of replating, a more significant decline in the total number of colonies was observed in the absence of NCOA2 (Figure 4I,J). Consistent with these data, the reciprocal BMT confirmed that the role of NCOA2 in HSCs was cell intrinsic (Figures 4K,L and S4D). Additional experiments revealed that the compromised hematopoietic reconstitution capacity of irradiated NCOA2^−/− HSCs was not due to defective homing efficiency (Figure S4E,F). These results indicate that NCOA2 intrinsically preserves the long‐term reconstitution ability of HSCs after IR.

NCOA2 deletion results in the accumulation of abnormal mitochondria in HSCs post IR

We next investigated the potential molecular mechanism by which NCOA2 modulates HSC function during stress. RNA sequencing (RNA‐seq) analysis identified 462 downregulated genes and 518 upregulated genes in irradiated HSCs when NCOA2 was deleted (Figure 5A–C), suggesting that the transcriptional alterations were pronounced in HSCs after NCOA2 knockout. Moreover, GSEA displayed that the gene sets about HSC signature, quiescence, stemness, and long‐term hematopoiesis were evidently decreased, whereas gene sets linked to HSC proliferation and cell cycle were markedly increased in irradiated NCOA2^−/− HSCs (Figure 5D), further consolidating the above findings. Importantly, we noticed that NCOA2 deletion increased mitochondrial metabolic activity and oxidative stress response in HSCs after IR (Figure 5D). Subsequently, the elevated mitochondrial mass, ROS levels, and OCR were verified in irradiated NCOA2^−/− HSCs (Figures 5E,F and S5A,B). The high activation of mitochondria can produce excessive by‐product ROS, which in turn damages mitochondrial integrity and reduces mitochondrial membrane potential, ultimately leading to HSC dysfunction. ³⁹ , ⁴⁰ Previous researches have shown that mitochondria play a vital role in the HSC fate determination. ⁴ , ²⁷ , ³¹ , ⁴¹ Expectedly, NCOA2^−/− HSCs exhibited a decreased TMRM/MTG ratio at Day 28 post IR (Figure 5G). The accumulation of injured mitochondria became more pronounced over time, which further impaired HSC quiescence in irradiated NCOA2^−/− mice (Figure S5C–F). Furthermore, our data showed that the expression of NCOA3 was comparable between WT and NCOA2^−/− HSCs, indicating that NCOA3 may not compensate for the loss of NCOA2 (Figure S5G,H). Hence, NCOA2 may be involved in the clearance of abnormal mitochondria in HSCs undergoing IR.

Figure 5.

Nuclear receptor coactivator 2 (NCOA2) deletion results in the accumulation of abnormal mitochondria in hematopoietic stem cells (HSCs) post irradiation (IR). (A) The strategy of RNA sequencing (RNA‐seq) analysis. (B) Heatmap and (C) volcano plots of the differentially expressed genes (DEGs) in long‐term HSCs (LT‐HSCs) from wild‐type (WT) and NCOA2^−/− mice at Day 28 after 5.0 Gy IR (n = 3). (D) Gene set enrichment analysis (GSEA) of hematopoiesis, HSC signature, quiescence, proliferation, mitochondria metabolism, and oxidative stress‐associated gene sets in LT‐HSCs from WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR. (E) Flow cytometric analysis of mitochondrial mass in the LT‐HSCs from WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR by MitoTracker Green (MTG) staining (n = 6). Representative flow cytometric plots are shown on the left. MFI, mean fluorescence intensity. (F) Flow cytometric analysis of reactive oxygen species (ROS) levels in LT‐HSCs from WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR by dichlorodihydrofluorescein diacetate (DCFH‐DA) staining (n = 6). Representative flow cytometric plots are shown on the left. (G) The ratio of MFI of tetramethylrhodamine methyl ester (TMRM) to MTG in LT‐HSCs from WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR (n = 6). **P < 0.01, ***P < 0.001.

NCOA2 regulates PINK1 expression in irradiated HSCs by coactivation of FOXO3a

It has been reported that mitochondrial autophagy, also known as mitophagy, contributes mainly to remove redundant or damaged mitochondria. ⁴² , ⁴³ Through in‐depth analysis of RNA‐seq data, we found that the target genes of FOXO3a were negatively enriched in NCOA2^−/− HSCs following IR (Figure 6A,B). As we know, FOXO3a is a critical transcription factor that controls mitophagy‐related gene expression in various types of cells. ⁴⁴ However, the mRNA and protein expression of FOXO3a were comparable between WT and NCOA2^−/− HSCs with IR stress (Figures 6C and S6A). Besides, the levels of p‐FOXO3a were unchanged in irradiated HSCs after NCOA2 knockout (Figure S6B). Considering that NCOA2 is a coactivator of several transcription factors, we speculated that NCOA2 may participate in regulating FOXO3a activity. As anticipated, the co‐immunoprecipitation (Co‐IP) assay substantiated that NCOA2 could interact with FOXO3a in HSPCs (Figures 6D and S6C). By analyzing the expression of recognized mitophagy‐related genes, we observed that only PINK1, one of FOXO3a's target genes, was significantly downregulated in NCOA2‐deleted HSCs experiencing IR exposure, which was further validated at the protein level (Figure 6E,F). Importantly, the reduced binding of FOXO3a to the promoter region of PINK1 was obtained in NCOA2^−/− HSCs post IR by ChIP analysis (Figure 6G). Furthermore, we found that NCOA2 deficiency decreased the recruitment of Parkin, a key downstream factor of PINK1, to mitochondria in HSCs after IR, indicating that NCOA2 regulates mitophagy via the PINK1/Parkin pathway (Figure S6D). In line with these results, impaired mitophagy was observed in irradiated NCOA2‐null HSCs by flow cytometry and immunofluorescence analysis (Figure 6H,I). To determine whether PINK1 downregulation mediates the effect of NCOA2 deficiency on HSCs, we transduced cells with lentivirus carrying PINK1. Overexpression of PINK1 significantly rescued the mitophagy and then reversed the defects in NCOA2^−/− HSCs after IR (Figures 6J and S6E–G). Taken together, our data imply that NCOA2 may serve as a coactivator of FOXO3a, thereby modulating PINK1/Parkin‐mediated mitophagy and maintaining HSC function following IR.

Figure 6.

Nuclear receptor coactivator 2 (NCOA2) regulates PINK1 expression in irradiated hematopoietic stem cells (HSCs) by coactivation of FOXO3a. (A) Gene set enrichment analysis (GSEA) of FOXO3a target genes in long‐term HSCs (LT‐HSCs) from wild‐type (WT) and NCOA2^−/− mice at Day 28 after 5.0 Gy IR. (B) Heatmap analysis of FOXO3a target genes in LT‐HSCs from WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR. (C) Flow cytometric analysis of the protein expression of FOXO3a in LT‐HSCs from the bone marrow (BM) of WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR (n = 6). Representative flow cytometric plots are shown on the left. (D) Co‐immunoprecipitation (Co‐IP) analysis of the interaction of NCOA2 and FOXO3a protein in Lin⁻ cells from the BM of WT mice at Day 28 after 5.0 Gy IR. Whole‐cell lysate (WCL) served as a loading control. IB, immunoblotting. (E) Quantitative polymerase chain reaction (qPCR) analysis of the mRNA expression of mitophagy‐related gene in LT‐HSCs from WT and NCOA2^−/− mice at Day 28 post 5.0 Gy IR (n = 3). (F) Flow cytometric analysis of the protein expression of PINK1 in LT‐HSCs from WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR (n = 6). Representative flow cytometric plots are shown on the left. (G) Chromatin immunoprecipitation (ChIP)‐qPCR analysis of the binding of FOXO3a to PINK1 promoter region in the LSKs from WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR (n = 3). Immunoglobulin G (IgG) served as a negative control (n = 3). (H) Flow cytometric analysis of mitophagy in the LT‐HSCs from WT and NCOA2^−/− mice at the indicated time points after 5.0 Gy IR by Mtphagy Dye staining (n = 6). Representative flow cytometric plots at Day 28 post 5.0 Gy IR are shown on the left. (I) The colocalization of TOMM20 and LAMP1 in the LT‐HSCs from WT and NCOA2^−/− mice at Day 28 after 5.0 Gy IR (n = 20 cells). Representative immunofluorescence images are shown on the left. (J) LT‐HSCs were sorted from WT or NCOA2^−/− mice at Day 28 after 5.0 Gy IR and then were transduced with lentivirus carrying PINK1 or control. Then, the mitophagy levels in LT‐HSCs were detected by Mtphagy Dye staining (n = 5). Oe, overexpression; Ctrl, control. **P < 0.01, ***P < 0.001.

Inhibition of OXPHOS or elimination of ROS rescues the function of NCOA2‐deficient HSCs suffering from IR

To determine whether the mitochondrial changes are accountable for the defects of NCOA2^−/− HSCs, we then employed metformin (Figure 7A), which is a known inhibitor of OXPHOS and can also activate mitophagy. ⁴⁵ , ⁴⁶ Treatment of metformin did not affect the expression of NCOA2 in irradiated HSCs (Figure S7A). As a result, metformin administration significantly reduced ROS production and mitochondrial injury in NCOA2^−/− HSCs suffering from IR (Figure 7B,C). In addition, metformin treatment alleviated the decreases in HSC quiescence and survival, therefore protecting them from IR‐induced exhaustion in NCOA2^−/− mice (Figure 7D–F). More importantly, the addition of metformin significantly rescued the function of irradiated HSCs with NCOA2 deletion (Figures 7G and S7B). Finally, we employed a classical ROS scavenger NAC to directly evaluate whether the abnormalities of NCOA2^−/− HSCs are due to the elevated ROS levels (Figure 7H). Likewise, NCOA2 expression levels remained unchanged in irradiated HSCs treated with NAC (Figure S7C). Furthermore, NAC injection partially reversed mitochondrial changes as well as the phenotypic and functional defects in NCOA2‐null HSCs during IR‐induced emergency hematopoiesis (Figures 7I–L and S7D,E). Collectively, our data illustrate that NCOA2 promotes HSC maintenance after IR mainly via the PINK1‐mediated mitophagy by coactivating FOXO3a (Figure 7M).

Figure 7.

Inhibition of oxidative phosphorylation (OXPHOS) or elimination of reactive oxygen species (ROS) rescues the function of nuclear receptor coactivator 2 (NCOA2)‐deficient hematopoietic stem cells (HSCs) suffering from irradiation (IR). (A–G) Wild‐type (WT) and NCOA2^−/− mice were subjected to 5.0 Gy IR and subsequently administered metformin or vehicle by oral gavage every 2 days for 28 days. (A) The strategy of metformin treatment. (B–G) Flow cytometric analysis of (B) ROS levels, (C) tetramethylrhodamine methyl ester (TMRM)/MTG ratio, (D) apoptosis, (E) cell cycle in long‐term HSCs (LT‐HSCs) from irradiated WT and NCOA2^−/− mice with or without metformin treatment (n = 6). (F) The number of LT‐HSCs in the bone marrow (BM) of irradiated WT and NCOA2^−/− mice with or without metformin treatment (n = 6). (G) At Day 28 after IR, LT‐HSCs (5 × 10²) from WT or NCOA2^−/− mice with or without metformin treatment, together with BM cells (5 × 10⁵) from CD45.1 mice, were transplanted into lethally irradiated CD45.1 recipients. The percentage of donor‐derived cells in the peripheral blood (PB) of recipient mice was detected at the indicated time points after transplantation (n = 6). (H–L) WT and NCOA2^−/− mice were subjected to 5.0 Gy IR and subsequently administered N‐acetyl‐L‐cysteine (NAC) or vehicle by intraperitoneal injection once a day for 28 days. (H) The strategy of NAC treatment. (I, J) Flow cytometric analysis of (I) apoptosis and (J) cell cycle in LT‐HSCs from WT and NCOA2^−/− mice after NAC treatment (n = 6). (K) The number of LT‐HSCs in the BM of irradiated WT and NCOA2^−/− mice after NAC treatment (n = 6). (L) At Day 28 after IR, LT‐HSCs (5 × 10²) from WT or NCOA2^−/− mice with or without NAC treatment, together with BM cells (5 × 10⁵) from CD45.1 mice, were transplanted into lethally irradiated CD45.1 recipients. The percentage of donor‐derived cells in the PB of recipient mice was detected at the indicated time points after transplantation (n = 6). (M) Schematic diagram describing the role of NCOA2 in regulating the return of HSCs to quiescence after IR via the FOXO3a‐PINK1‐mediated mitophagy axis. NS, not significant; *P < 0.05, **P < 0.01, and ***P < 0.001; ^#P < 0.05, ^##P < 0.01, and ^###P < 0.001; and ^†P < 0.05, ^††P < 0.01, and ^†††P < 0.001.

DISCUSSION

HSCs are characterized by the capacity to self‐renew and differentiate into all blood cell lineages, which is necessary for maintaining hematopoietic homeostasis and facilitating hematopoietic regeneration after stress. ⁶ , ⁴¹ These processes are precisely controlled by diverse intrinsic and extrinsic factors. ⁸ , ¹⁵ Once the hematopoietic function recovers, activated HSCs will return to quiescence to prevent HSC pool depletion and subsequent BM failure. ⁹ Previous studies have identified the role of mitochondria in the regulation of stress hematopoiesis, ¹¹ , ⁴⁰ whereas the underlying molecular mechanism is not fully delineated. Here, we show for the first time that NCOA2 promotes mitophagy via the FOXO3a‐PINK1 signaling pathway, therefore driving HSCs to return to a resting state after IR stress.

As a transcriptional coactivator, NCOA2 is widely expressed in tissues or cells and participates in various biological processes, such as embryonic development, reproduction, and immune response. ¹⁸ , ²⁴ In recent years, NCOA2 has attracted great attention due to its overexpression in many types of tumor cells. ⁴⁷ , ⁴⁸ It was reported that the fusion of NCOA2 with histone acetyltransferase MOZ may promote the malignant transformation of HSPCs and ultimately induce acute myeloid leukemia. ⁴⁹ Meanwhile, ETV6‐NCOA2 gene rearrangement has been shown to bring about a rare T/myeloid mixed‐phenotype leukemia. ⁵⁰ Nevertheless, the role of NCOA2 in HSC biology is largely unknown. In the present study, we found that NCOA2 deficiency slightly but statistically significantly increased total cell number in the spleen, accompanied with elevated percentage of myeloid cells in the PB, which is consistent with previous reports that NCOA2 may directly regulate the function of mature hematopoietic and immune cells. ³³ , ³⁵ However, NCOA2 was dispensable for HSC behavior at steady state. Intriguingly, the nuclear translocation of NCOA2 was gradually increased during the late hematopoietic recovery stage in HSCs after IR, although its expression was unchanged. Consequently, NCOA2 ablation resulted in HSC pool exhaustion and delayed hematopoietic recovery when exposed to IR. The differences in HSCs are greater than those in PB, which might be due to the high compensatory capacity of the progenitor cells within a short period of time, as described previously. ⁹ , ⁵¹ These findings reveal that NCOA2 is a potential hematopoietic regulator, primarily functioning during the late recovery period following IR stress injury.

Previous studies focused more on how to maintain HSC quiescence during homeostasis or how HSCs transform from a dormant state to an activated state during emergency hematopoiesis. ³ , ⁵ Up to now, many positive and negative factors in regulating cell cycle progression, such as KAT7, Ptpn21, CDK19, and Gabbr1, have been identified.32, 52, 53, 54 However, the exact molecular pathways driving HSC re‐entry into quiescence following stress‐induced activation have not been deeply studied. It was reported that an ERK‐dependent feedback pathway promotes HSC quiescence after 5‐FU treatment. ⁹ In this work, we demonstrated that NCOA2 was a key factor in enabling the return of HSCs to quiescence when the hematopoiesis recovered from IR injury. The deletion of NCOA2 obviously decreased the quiescence and survival of irradiated HSCs, thus impairing their colony‐forming ability in vitro and long‐term repopulation ability in vivo. As confirmed by numerous studies, the maintenance of quiescence is essential for sustaining the self‐renewal ability of HSCs. ¹⁰ , ¹² , ⁵² On the other hand, although NCOA2 deletion markedly promoted HSC proliferation, it did not accelerate the recovery of WBC, RBC, and PLT counts after IR. This is most likely attributed to the increased apoptosis rate observed in irradiated mice in the absence of NCOA2. Actually, loss of quiescence can also accelerate HSC apoptosis induced by various harmful factors.

Mitochondria are the center of energy metabolism, which play a crucial role in the fate determination of HSCs. ¹³ , ¹⁷ , ²⁷ Under physiological conditions, most HSCs exhibit low metabolic activity to facilitate HSC maintenance. ¹⁴ During stress conditions, the mitochondrial biogenesis and metabolism are robustly activated to prompt HSCs to enter the cell cycle. ¹³ When HSCs return to quiescence after stress, the highly activated mitochondria must be suppressed or eliminated, whereas the underlying mechanism is unelucidated. NCOA2 has been recognized as an important metabolic regulator, which is involved in diverse metabolic processes via tissue‐specific pathways. ²³ , ³³ , ⁴⁷ In prostate cancer, NCOA2 supports tumor cell survival and metastasis primarily through modulating lipogenesis and glutamine metabolic reprogramming. ⁴⁷ In contrast, NCOA2 mediates hepatic beta‐oxidation via regulating PPARα activity. ¹⁹ In this study, we revealed that NCOA2 deletion led to increased mitochondrial mass and activity in irradiated HSCs, accompanied with elevated ROS levels. In turn, excessive ROS caused mitochondrial injury, as determined by decreased TMRM/MTG ratio in NCOA2^−/− HSCs. The accumulation of these mitochondria may disturb cellular metabolic balance, ultimately causing functional impairment in HSCs. Altogether, our data indicate that NCOA2 promotes the clearance of activated or damaged mitochondria in HSCs, thereby contributing to HSC maintenance after stress.

Accumulating studies have established the critical role of autophagy in HSC maintenance. As reported, deletion of autophagy‐related genes, such as Atg7, Atg5, and Atg12, leads to elevated ROS and impaired HSC function. ⁵⁵ , ⁵⁶ Additionally, FOXO3a is a multifunctional transcription factor that regulates diverse biological processes including autophagy, aging, and metabolism. ⁴⁴ As a master regulator of mitophagy that selectively eliminates redundant or damaged mitochondria, FOXO3a preserves mitochondrial integrity and cellular energy homeostasis. ⁵⁷ Among FOXO3a target genes, PINK1 is an indispensable functional regulator, which mediates mitophagy via the ubiquitin ligase Parkin‐dependent pathway. ⁵⁸ NCOA2 was initially identified as a steroid receptor coactivator, but in fact, it can also coactivate many transcription factors. ¹⁹ , ²¹ , ³⁴ In this research, we found that FOXO3a target genes, including PINK1, were significantly downregulated in irradiated NCOA2^−/− HSCs, while FOXO3a expression was unchanged. Subsequent assays showed that NCOA2 could interact with FOXO3a, and the binding of FOXO3a to the promoter region of PINK1 was substantially decreased. These data suggest that NCOA2 may be recruited by FOXO3a to regulate the transcription of PINK1 in HSCs. It was reported that the injury of mitochondria can activate mitophagy, which is a self‐protective strategy to maintain cellular homeostasis. ⁵⁹ Combined with our results, we speculate that NCOA2 nuclear translocation may be triggered by mitochondrial damage that is gradually accumulated after mitochondrial activation. Meanwhile, it is possible that high levels of ROS may directly or indirectly induce the nuclear localization of NCOA2. Finally, either inhibition of OXPHOS or elimination of ROS partially rescued the defects of NCOA2‐null HSCs after IR, further verifying that mitochondrial changes affect HSC function. Combined with the above results, we propose that NCOA2 facilitates mitophagy in HSCs after IR through regulating the FOXO3a‐PINK1 signaling axis. However, at Day 14 after IR, the PB recovery may not be a direct effect of NCOA2 nuclear translocation, or it may be caused by other functions of NCOA2 that are independent of nuclear translocation. After all, at that time, most of NCOA2 was present in the cytoplasm. Furthermore, given that SRC family members are able to promote histone acetylation and chromatin remodeling, we speculate that NCOA2 may also be involved in this process, while the exact mechanism still requires further investigation.

In conclusion, we illuminate that NCOA2 promotes the return of HSCs to quiescence following IR exposure via the modulation of FOXO3a‐PINK1‐mediated mitophagy. Our findings not only identify a previously unrecognized function of NCOA2 but also provide a new insight into the mechanism of hematopoietic regeneration under stress conditions. More importantly, the study offers a new therapeutic target to promote hematopoietic recovery in patients suffering from IR in the clinic.

AUTHOR CONTRIBUTIONS

Fang Chen: Writing—original draft; formal analysis; methodology; data curation; software. Naicheng Chen: Methodology; software; data curation; formal analysis. Linjie Wang: Software; methodology; formal analysis. Hao Zeng: Methodology; software. Yukai Lu: Methodology; validation. Yanying Liu: Methodology. Jing Chen: Software. Yang Xu: Methodology; validation. Mo Chen: Methodology. Song Wang: Methodology; validation. Mingqiang Shen: Methodology. Junping Wang: Validation; formal analysis; supervision; funding acquisition; project administration; writing—review and editing. Mengjia Hu: Validation; formal analysis; supervision; funding acquisition; project administration; writing—review and editing.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

The animal experiment was approved by the Animal Ethics Committee of the Third Military Medical University (AMUWEC20230133).

FUNDING

This work was supported by grants from the National Natural Science Foundation of China (Nos. 82430103 and 82574026), Natural Science Foundation of Chongqing City (Nos. CSTB2024NSCQ‐JQX0002 and CSTB2023NSCQ‐MSX0284), and Science Foundation of State Key Laboratory of Trauma and Chemical Poisoning (No. 2024K004).

Supporting information

Supporting Information.

DATA AVAILABILITY STATEMENT

scRNA‐seq data have been deposited in the Gene Expression Omnibus (GEO) database (No. GSE306557). Bulk RNA‐seq data have been deposited in the Sequence Read Archive (SRA) database (No. PRJNA1302003). All other data are available within the article and its Supplementary Files or from the corresponding authors upon reasonable request.