1. Cell Biology

Receptor-mediated mitophagy regulates EPO production and protects against renal anemia

  1. Guangfeng Geng
  2. Jinhua Liu
  3. Changlu Xu
  4. Yandong Pei
  5. Linbo Chen
  6. Chenglong Mu
  7. Ding Wang
  8. Jie Gao
  9. Yue Li
  10. Jing Liang
  11. Tian Zhao
  12. Chuanmei Zhang
  13. Jiaxi Zhou
  14. Quan Chen  Is a corresponding author
  15. Yushan Zhu  Is a corresponding author
  16. Lihong Shi  Is a corresponding author
  1. State Key Laboratory of Medicinal Chemical Biology, Frontiers Science Center for Cell Responses, Tianjin Key Laboratory of Protein Science, College of Life Sciences, Nankai University, China
  2. State Key Laboratory of Experimental Hematology, National Clinical Research Center for Blood Diseases, Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, China
https://doi.org/10.7554/eLife.64480
Share this article
Cite this article
  1. Guangfeng Geng
  2. Jinhua Liu
  3. Changlu Xu
  4. Yandong Pei
  5. Linbo Chen
  6. Chenglong Mu
  7. Ding Wang
  8. Jie Gao
  9. Yue Li
  10. Jing Liang
  11. Tian Zhao
  12. Chuanmei Zhang
  13. Jiaxi Zhou
  14. Quan Chen
  15. Yushan Zhu
  16. Lihong Shi
(2021)
Receptor-mediated mitophagy regulates EPO production and protects against renal anemia
eLife 10:e64480.
https://doi.org/10.7554/eLife.64480
  2. Download BibTeX
  3. Download .RIS

Abstract

Erythropoietin (EPO) drives erythropoiesis and is secreted mainly by the kidney upon hypoxic or anemic stress. The paucity of EPO production in renal EPO-producing cells (REPs) causes renal anemia, one of the most common complications of chronic nephropathies. Although mitochondrial dysfunction is commonly observed in several renal and hematopoietic disorders, the mechanism by which mitochondrial quality control impacts renal anemia remains elusive. In this study, we showed that FUNDC1, a mitophagy receptor, plays a critical role in EPO-driven erythropoiesis induced by stresses. Mechanistically, EPO production is impaired in REPs in Fundc1-/- mice upon stresses, and the impairment is caused by the accumulation of damaged mitochondria, which consequently leads to the elevation of the reactive oxygen species (ROS) level and triggers inflammatory responses by up-regulating proinflammatory cytokines. These inflammatory factors promote the myofibroblastic transformation of REPs, resulting in the reduction of EPO production. We therefore provide a link between aberrant mitophagy and deficient EPO generation in renal anemia. Our results also suggest that the mitochondrial quality control safeguards REPs under stresses, which may serve as a potential therapeutic strategy for the treatment of renal anemia.

Introduction

Erythropoiesis is a highly dynamic and tightly regulated process in which red blood cells are generated from immature progenitors in the bone marrow (An et al., 2015; Nandakumar et al., 2016). This process is mainly driven by the cytokine glycoprotein erythropoietin (EPO) (Kuhrt and Wojchowski, 2015). Upon the engagement of EPO with its receptor, erythropoietin receptor (EPOR), the Janus kinase (JAK) signal transducer and activator of transcription (STAT) pathway is activated to prevent committed erythroid progenitors from apoptosis and to facilitate their proliferation and differentiation into red blood cells (Nandakumar et al., 2016). During the final maturation step of erythropoiesis, reticulocytes need to remove all of the remaining organelles, including mitochondria, endocytic vesicles, ribosomes, Golgi cisternae, and endoplasmic reticulum (Koury et al., 2005; Gronowicz et al., 1984). Among them, mitochondria are the most abundant organelles for removal (Zhang et al., 2015). Prior studies have demonstrated that defects in mitochondrial removal are linked to the dysfunction of red blood cells and can lead to anemia (Kundu et al., 2008; Mortensen et al., 2010; Honda et al., 2014; Nishida et al., 2009; Schweers et al., 2007; Sandoval et al., 2008). Therefore, a deeper understanding of the molecular basis of mitochondrial quality control is of great significance for dissecting normal erythropoiesis and could elucidate the pathogenesis of dyserythropoietic diseases.

As one of the key regulators of erythropoiesis, EPO is primarily secreted by the kidneys during adulthood by the renal EPO-producing cells (REPs) located in the renal cortex and outer medulla (Pan et al., 2011; Yamazaki et al., 2013). REPs possess functional plasticity for EPO production in response to hypoxia and other microenvironmental changes (Souma et al., 2015). For instance, hypoxic stresses stabilize hypoxia-inducible factors (HIFs), especially HIF2α, which binds to the enhancer region of HIF binding element of the EPO gene and upregulates its expression (Kuhrt and Wojchowski, 2015). Upon inflammation or with chronic diseases, however, proinflammatory cytokines, such as TNFa and IL1b, promote the myofibroblastic transition of REPs (Asada et al., 2011), concomitantly repress EPO production in REPs (Frede et al., 1997), and ultimately lead to anemia (Weiss and Goodnough, 2005; Guo et al., 1999; Inoue et al., 2010). Although mitochondrial quality control exerts protective roles in a variety of renal disorders (Liu et al., 2014), whether and how it might be implicated in REPs during chronic or inflammatory diseases remain unclear.

Mitochondria are essential organelles that maintain cellular energy and redox homeostasis and are the major source and target of intracellular oxidative stress (Wei et al., 2015). They also play central roles in regulating innate immunity and apoptosis (Wei et al., 2015). Maintaining normal mitochondrial activities and homeostasis is of great importance for cellular and organismal metabolisms and functions (Simon et al., 2000; Yakes and Van Houten, 1997). Mitophagy, a selective type of autophagy that allows the elimination of unwanted or damaged mitochondria, is the major mechanism sustaining mitochondrial homeostasis and quality control. Mitophagy can be mediated by the PTEN-induced putative kinase 1-Parkin (PINK1-PAKN) pathway and/or by mitophagic receptors residing in the outer mitochondrial membrane (Wei et al., 2015; Wu and Chen, 2015), such as BCL2 Interacting Protein 3 Like (BNIP3L, also named as NIX) (Novak et al., 2010), BCL2 interacting protein 3 (BNIP3) (Hanna et al., 2012), FUN14 domain containing 1 (FUNDC1) (Liu et al., 2012a; Chen et al., 2014), Prohibitin2 (PHB2) (Wei et al., 2017), and FKBP Prolyl Isomerase 8 (FKBP8) (Saita et al., 2013) in mammalian cells. It has been demonstrated that receptor-mediated mitophagy is important for erythroid maturation (Kundu et al., 2008; Mortensen et al., 2010; Honda et al., 2014; Nishida et al., 2009; Schweers et al., 2007; Sandoval et al., 2008). For example, Bnip3l-/- mice exhibit anemia and reticulocytosis with defective mitophagy due to the persistence of the mitochondrial membrane potential and the resultant depolarization in the mitochondria (Schweers et al., 2007; Sandoval et al., 2008). However, in the reticulocytes of the Bnip3l-/- mice, mitochondrial removal still occurs, raising the possibility that additional factors are also required for mitochondrial degradation during red blood cell maturation.

We have previously shown that FUNDC1 functions as a specific mitophagy receptor in response to mitochondrial stresses, such as hypoxia (Liu et al., 2012a; Chen et al., 2017). FUNDC1 regulates mitochondrial homeostasis, including mitochondrial fission and fusion, content alteration, quality surveillance and physical interactions with other organelles (Li et al., 2019). Deletion of Fundc1 results in the accumulation of dysfunctional mitochondria and the activation of inflammasomes and is associated with a wide range of pathological conditions, such as hepatocarcinogenesis (Li et al., 2018) and heart failure (Wu et al., 2017). Although the role of FUNDC1 has been studied in an array of cellular processes, whether it functions during EPO-driven erythropoiesis is still unexplored. In this study, using Fundc1 germline deletion mice (Fundc1-/- mice), we found that despite being dispensable for erythroid maturation, FUNDC1 regulates EPO production during stress erythropoiesis. Mechanistically, enhanced inflammatory responses in REPs lead to insufficient EPO production as a consequence of compromised mitophagy. Together, our studies reveal a link between deficient mitophagy and reduction in EPO production and demonstrate that mitochondrial quality control safeguards REPs under stresses.

Results

Fundc1 is implicated in stress but not steady-state erythropoiesis

Previous studies have shown that BNIP3L, an important mitophagy receptor, is involved in mitochondrial elimination in reticulocytes and proper functions in red blood cells (Kundu et al., 2008; Mortensen et al., 2010; Honda et al., 2014; Nishida et al., 2009; Sandoval et al., 2008). These findings led us to dissect the function of FUNDC1, another key mitophagy receptor we previously characterized (Liu et al., 2012a), in erythroid development and differentiation. We found that in Fundc1-/- mice, the hemogram parameters of the peripheral blood, including erythroid parameters such as red blood cell (RBC) count, hemoglobin amount, hematocrit (HCT, measuring the volume percentage of red blood cells in whole blood), and reticulocytes, were comparable to those in the corresponding wild-type (WT) littermates at 8–10 weeks old (Figure 1—figure supplement 1A and Supplementary file 1). Despite evident mitochondrial retention in erythroid cells of Bnip3l-/- mice (Sandoval et al., 2008), negligible mitochondrial accumulation was detected in the Fundc1-/- cells collected from the peripheral blood, spleen and bone marrow (Figure 1—figure supplement 1B–D). In addition, Fundc1 deletion did not affect erythroid lineage commitment and differentiation (Figure 1—figure supplement 1E), generation of hematopoietic stem and progenitors (Figure 1—figure supplement 1F), or differentiation of other non-erythroid, hematopoietic lineages (Figure 1—figure supplement 1G,H) in peripheral blood, spleen, and bone marrow, respectively. Together, these observations suggest that FUNDC1 is dispensable for steady-state erythropoiesis and possibly for hematopoiesis.

The role of FUNDC1 in mitophagy induced by hypoxia and other stresses led us to speculate that FUNDC1 might play a role in stress erythropoiesis. We treated the Fundc1-/- mice (Zhang et al., 2016) with phenylhydrazine (PHZ, 100 mg/kg), a common agent to induce acute hemolytic anemia via the induction of lysis of mature red blood cells, and then tracked hemogram in peripheral blood for a week (Figure 1A). We found that when compared with WT counterparts, Fundc1-/- mice exhibited low survival with PHZ treatment (p = 0.04, n = 30) (Figure 1B). Time-course hemogram examination showed reduced RBC numbers, lower HGB levels, and decreased HCT values in PHZ-Fundc1-/- mice than those of controls (Figure 1C, Figure 1—figure supplement 2A and Supplementary file 1).

Figure 1 with 2 supplements see all
Download asset Open asset
Defects of splenic erythroid progenitor production in Fundc1-/- mice during stress erythropoiesis.

(A) Schematic diagram of PHZ-induced stress erythropoiesis. Mice around 8–10 weeks old were treated with a single dose of PHZ (100 mg/kg body weight) to induce acute anemia, and multiple examinations were taken in the following week. (B) Survival curves of PHZ- treated WT control littermates and Fundc1-/- mice (n = 30 for each group). (C) Hemogram parameters of red blood cell (RBC) counts and hematocrit (HCT) in the peripheral blood of WT controls and Fundc1-/- mice from day 0 to 7 after PHZ treatment. (D) Total erythroid cell number in spleen of WT and Fundc1-/- mice at 0, 24, 48, and 72 hr after PHZ treatment, respectively. (E, F) Flow cytometry analysis of erythroid differentiation with using CD71 and Ter119 from the spleen of WT and Fundc1-/- mice after 48 hr of PHZ treatment. The representative flow cytometry diagram (E) and the corresponding quantitative analysis (F) are shown. (G–I) The bar graphs showing the normalized intensity of fluorescence by using flow cytometry of mitochondrial mass assayed by MitoTracker Green (G), membrane potential by TMRM (H) and ROS level by MitoSOX (I) within splenic R2 erythroid cells of WT or Fundc1-/- mice after 48 hr of PHZ treatment, respectively. PBS acts as the control group for PHZ treatment. For each experiment (C–I), n = 3–5 mice for each group. Individual mice are represented by symbols. Data shown are representative of at least three independent experiments. Similar results were found in each experiment. All data are mean ± SEM; *p<0.05, **p<0.01. Statistical significance was analyzed by using the two-tailed unpaired Student’s t-test.

Figure 1—source data 1

Erythroid progenitor production during stress erythropoiesis.

https://cdn.elifesciences.org/articles/64480/elife-64480-fig1-data1-v2.xlsx

Given the established function of Fundc1 in mitophagy, we next asked whether reduced RBCs in PHZ-Fundc1-/- mice might be caused by the defects in mitophagy, which may block the maturation of reticulocytes to red blood cells. Given that the reticulocytosis occurred in WT and Fundc1-/- mice upon PHZ treatment (Figure 1—figure supplement 2A), we then FACS-sorted Ter119+CD71+ reticulocytes in the peripheral blood of WT and Fundc1-/- mice on day 5 after PHZ treatment (Figure 1—figure supplement 2B). The isolated reticulocytes were morphologically confirmed by Giemsa staining (Figure 1—figure supplement 2C, left) and brilliant cresyl blue staining (Figure 1—figure supplement 2C, right). Next, the reticulocytes were cultured to maturation in vitro for 3 days using a previously described method (Koury et al., 2005). The brilliant cresyl blue staining showed progressive maturation of reticulocytes in WT mice after 3 days of culture (Figure 1—figure supplement 2C). Unexpectedly, there was no detectable difference in the maturation of reticulocytes between WT and Fundc1-/- mice (Figure 1—figure supplement 2C). Consistent with these findings, the mitochondrial mass, as measured with MitoTracker Green, exhibited little difference between WT and Fundc1-/- mice during reticulocyte maturation (Figure 1—figure supplement 2D). Thus, Fundc1-mediated mitophagy seems dispensable for reticulocyte maturation.

To further decipher the cause of reduced RBCs, we assessed erythropoiesis in the spleen of PHZ-Fundc1-/- mice because stress erythropoiesis primarily occurs in the mouse spleen. We found that the total number of erythroid cells in the spleen of PHZ-Fundc1-/- mice was reduced compared to WT mice after 48 hr of treatment (p<0.05, Figure 1D), while there were no apparent differences in spleen size, weight, and color (Figure 1—figure supplement 2E). Subsequent analysis revealed a transient reduction of the erythroid compartment such as the R2 population (CD71hiTer119+) (mean 8% vs 2%, p<0.01) (Figure 1E,F and Figure 1—figure supplement 2F). Nonetheless, the decreased R2 population was unlikely to be due to elevated apoptosis (Figure 1—figure supplement 2G), accumulation of mitochondrial mass (Figure 1G), aberrant mitochondrial membrane potential (Figure 1H), or enhanced mitochondrial reactive oxygen species (ROS) level (Figure 1I). Additionally, we examined splenic progenitors including CMP, MEP, and GMP in PHZ-Fundc1-/- mice and found no obvious alterations in either the percentage or the total cell number (Figure 1—figure supplement 2H,I).

Together, our results suggest that FUNDC1-mediated mitophagy is dispensable for steady-state erythropoiesis, whereas it is involved in PHZ-induced stress erythropoiesis.

Hypoactivation of JAK-STAT signaling underlies the reduction of erythroid population

To uncover the molecular mechanism governing the reduction of R2 population, we next conducted high-throughput transcriptome profiling using FACS-purified R2 cells from the spleen of WT and Fundc1-/- mice after 48 hr of PHZ treatment. We identified 1065 significant differentially expressed genes (DEGs) in total (Figure 2A). Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis with DEGs revealed differential regulation of key signaling pathways and biological processes (Figure 2B). We found that the JAK-STAT signaling pathway was among the top five significantly enriched pathways (Figure 2B), in agreement with gene set enrichment analysis (GSEA) (Figure 2C) and gene set quantitative analysis (Figure 2D). The key components and downstream targets of JAK-STAT signaling pathway, such as Crebbp, Ep300, Irf9, and Spred2, were also downregulated in Fundc1-/- R2 cells (Figure 2E).

Figure 2 with 1 supplement see all
Download asset Open asset
Hypoactivation of JAK-STAT signaling underlies the comprised erythroid progenitor production in Fundc1-/- mice during stress erythropoiesis.

(A) Volcano Plot illustrates the differentially expressed genes (DEGs) of FACS-purified erythroid progenitors (R2 populations) in WT (n = 3) and Fundc1-/- (n = 3) spleen at 48 hr after PHZ treatment. Top 10 genes in each group were labeled. (B) Top 5 KEGG functional enrichment pathways analysis of downregulated DEGs in Fundc1-/- R2 cells. (C) GSEA showing the enrichment of the JAK-STAT signaling pathway in splenic R2 cells from WT and Fundc1-/- mice. Normalized enrichment score (NES) and false discovery rate (FDR) are shown. (D) The box plot depicts the quantitative analysis of the genes in JAK-STAT signaling pathway in splenic R2 cells of WT and Fundc1-/- mice. The boxes represent the median and quartile of the sum of log2 transformed (FPKM +1). p-value was determined by using paired Wilcoxon signed-rank Test. (E) The expression of key components in the JAK-STAT pathway in splenic R2 cells from WT and Fundc1-/- mice. (F) Western blotting of STAT3, STAT5 and phosphorylated STAT3 and STAT5 in FACS-sorted R2 populations in spleen of WT and Fundc1-/- mice after 48 hr of PBS or PHZ treatment, respectively. ACTIN was used as a loading control. n = 3. (G) The quantification of the signal intensity of phosphorylated STAT3 and STAT5 out of the total STAT3 and STAT5 in (F). (H) The expression of Epor mRNA in the spleen of WT and Fundc1-/- mice after 48 hr of PBS or PHZ treatment. (I) The serum EPO concentration in Fundc1-/- and WT mice after 48 hr of PBS or PHZ treatment. (J) Schematic illustrating the administration of EPO to PHZ-Fundc1-/- and PHZ-WT mice. (K) RBC number (upper) and HGB level (bottom) after EPO supplement in PHZ-Fundc1-/- and PHZ-WT mice. (L) Erythroid differentiation in spleen of PHZ-Fundc1-/- and PHZ-WT after EPO administration. The left panel shows representative FACS plots and right bar graph shows the statistical quantification of the percentage of the R2 compartment. For each experiment (E–L), n = 3–5 mice for each group. Individual mice are represented by symbols. Data shown are representative of at least three independent experiments. Similar results were found in each experiment. All data are mean ± SEM; *: p<0.05; ns: no statistical significance. Statistical significance was analyzed by using the two-tailed unpaired Student’s t-test unless stated otherwise.

Figure 2—source data 1

The JAK-STAT signaling in WT and Fundc1-/- mice during stress erythropoiesis.

https://cdn.elifesciences.org/articles/64480/elife-64480-fig2-data1-v2.xlsx

To further confirm the hypoactivation of JAK-STAT signaling, we analyzed the total and phosphorylated levels of STAT5 and STAT3 proteins in FACS-purified R2 cells, which are central players in the JAK-STAT pathway during erythropoiesis (Richmond et al., 2005). In WT R2 cells, PHZ treatment had limited effects on the total levels of STAT5 and STAT3 (Figure 2F). However, the levels of phosphorylated STAT5 and STAT3 were markedly elevated (Figure 2F,G), indicating that the JAK-STAT signaling pathway in R2 cells was activated upon PHZ-induced hemolytic stress. In striking contrast, the levels of phosphorylation of STAT5 and STAT3 were significantly lower in PHZ-Fundc1-/- R2 cells than those in the WT counterparts (Figure 2F,G). Thus, the JAK-STAT signaling pathway is hypoactivated in PHZ-treated Fundc1-/- R2 cells.

During erythropoiesis, the JAK-STAT signaling pathway acts downstream of the EPO receptor (EPOR) and is regulated by the serum EPO level (Kuhrt and Wojchowski, 2015). EPOR, a transmembrane glycoprotein, is expressed in erythroid progenitors (colony-forming unit-erythroid: CFU-E), proerythroblasts, and early basophilic erythroblasts (Kuhrt and Wojchowski, 2015). These erythroid progenitors and precursors are the major components of the R2 population (Socolovsky et al., 2001), which may account for the sensitivity of the R2 compartment to EPO paucity. We then investigated the Epor expression in R2 cells in WT and Fundc1-/- mice with or without PHZ treatment and found no apparent changes when Fundc1 was deleted (Figure 2H). However, enzyme-linked immunosorbent assay (ELISA) experiments showed that the serum EPO level was significantly reduced in PHZ-Fundc1-/- mice compared with the corresponding controls (6,100 vs 4,400 and 4,500 vs 3,500 pg/mL after 24 and 48 hr of PHZ treatment, respectively, p<0.05) (Figure 2—figure supplement 1A and Figure 2I). In contrast, a large number of inflammatory cytokines in the serum, such as IL6 and IL1, exhibited a modest increase in PHZ-Fundc1-/- mice (Figure 2—figure supplement 1B). More importantly, exogenous administration of EPO (1,000 IU/kg) to PHZ-Fundc1-/- mice (Figure 2J) showed the comparable peripheral RBC number and HGB level with their WT counterparts (Figure 2K). Erythroid differentiation in the spleen of PHZ-Fundc1-/- mice was restored to the levels of PHZ-WT controls (particularly of R2) as well (Figure 2L). These results suggested that insufficient EPO supply underlies the reduced RBC number in PHZ-Fundc1-/- mice. In contrast, no difference in EPO production was observed between Fundc1-/- mice and the control mice in the pure anemic model as determined via phlebotomy of 500 μL blood for three consecutive days (Figure 2—figure supplement 1C–G), suggesting that the reduced EPO production in PHZ-Fundc1-/- mice most likely arises from the oxidative stress of PHZ.

Taken together, these results suggest the following scenario: Fundc1 deletion reduces EPO secretion and consequently leads to hypoactivation of JAK-STAT signaling in erythroid R2 cells in the spleen during stress erythropoiesis, which ultimately results in the reduction of the red blood cell count in Fundc1-/- mice with PHZ treatment.

Fundc1 deletion causes renal injury

The kidney is the predominant organ for EPO production in adults. EPO is produced by REPs that are responsible for approximately 90% of total EPO synthesis (Nolan and Wenger, 2018). We therefore examined the impact of PHZ on the kidney in Fundc1-/- mice after 48 hr of treatment. We found that Fundc1 was highly and ubiquitously expressed in the kidney (Figure 3—figure supplement 1A,B). In addition, the mRNA level of Epo in kidney of Fundc1-/- mice was indeed significantly lower than that of its WT counterparts at 48 hr after PHZ treatment (130 vs 61, p<0.05) (Figure 3A). We also detected the elevated expression of two kidney injury markers, Lipocalin-2 (Lcn2) (Moschen et al., 2017) and kidney injury molecule-1 (Kim1) (Humphreys et al., 2013; Figure 3B), although hematoxylin and eosin (H&E) staining illustrated no major structural alterations in the kidneys of PHZ-Fundc1-/- mice (Figure 3—figure supplement 1C).

Figure 3 with 1 supplement see all
Download asset Open asset
MtDNA cytosolic translocation caused by insufficient mitophagy induces elevated inflammation and severer renal injury in Fundc1-/- mice upon PHZ treatment.

(A) qRT-PCR analysis showing the kinetics of Epo mRNA expression in kidneys of WT and Fundc1-/- mice after PBS or PHZ treatment. (B) Relative mRNA levels of kidney injury marker Lcn2 and Kim1 in kidneys of WT and Fundc1-/- mice with or without PHZ treatment. (C) Western blotting showing the HIF2α expression in whole kidney lysate in WT and Fundc1-/- mice with or without PHZ treatment. ACTIN was used as a loading control. (D) Immunohistochemistry of mitochondrial protein PHB2 in kidneys of WT and Fundc1-/- mice with or without PHZ treatment, respectively. The indicated area (blue dash box) is zoomed in and intense signals of PHB2 are marked with red arrows. The right bar graph presents the quantitative assay measured by ImageJ. Scale bars, 200 mm; original magnification, 10×. (E, F) Western blotting (E) and corresponding quantification (F) showing the protein levels of mitochondrial membrane proteins (TIMM23, COXIV, and TOMM20) and autophagic proteins (P62, LC3II/I, and their ratio) in kidneys of WT and Fundc1-/- mice with or without PHZ treatment. TUBULIN was used as a loading control. (G) Relative mRNA levels of Atp5d, Nrf1, and Tfam in kidneys of WT and Fundc1-/- mice with or without PHZ treatment. (H, I) qRT-PCR shows the total and cytosolic mtDNA in kidneys of WT and Fundc1-/- mice with or without PHZ treatment. (J) The mRNA levels of Il1b, Il6, and Tnfa in kidneys treated with PBS or PHZ. For each experiment, n = 3–5 mice for each group. Individual mice are represented by symbols. Data shown are representative of at least three independent experiments. Similar results were found in each experiment. All data are mean ± SEM; *p<0.05, **p<0.01, ns: no statistical significance. Statistical significance was analyzed by using the two-tailed unpaired Student’s t-test.

Figure 3—source data 1

MtDNA cytosolic translocation and severer renal injury in Fundc1-/- mice upon PHZ treatment.

https://cdn.elifesciences.org/articles/64480/elife-64480-fig3-data1-v2.xlsx

It has been shown previously that EPO production is almost exclusively controlled at the transcription level by HIFs, predominantly by HIF2α (Kapitsinou et al., 2010; Rankin et al., 2007; Gruber et al., 2007). These findings led us to examine the potential link between EPO and HIFs in PHZ-treated Fundc1-/- mice. To this end, we first measured the expression of HIF2α and its downstream targets, Vegf and Slc2a1, in the kidneys of PHZ-treated Fundc1-/- mice. Although HIF2α protein levels were sharply enhanced in kidneys of both WT and Fundc1-/- mice after PHZ treatment, the levels between WT and Fundc1-/- kidneys were similar (Figure 3C). Similar results were observed for the HIF2α targets Vegf and Slc2a1 (Figure 3—figure supplement 1D). Thus, the reduction of EPO secretion is unlikely to be caused by the aberrant transcriptional regulation of Epo. We also excluded the possibility of apoptosis in the kidneys of PHZ-Fundc1-/- mice (Figure 3—figure supplement 1E).

How is EPO supply attenuated in PHZ-Fundc1-/- mice? The key role of FUNDC1 in mitophagy led us to further test the efficiency of mitophagy. In immunohistochemical analysis of mitochondrial membrane protein Prohibitin 2 (PHB2), we found that PHZ treatment generally reduced the accumulation of total mitochondrial mass (Figure 3D), suggesting that PHZ stress could promote the mitophagic process. In contrast, Fundc1 deletion resulted in more retainment of mitochondria in the kidneys than in the WT controls after PHZ treatment (Figure 3D), indicating that mitophagy was attenuated in Fundc1-/- kidneys under PHZ-induced stresses. To further support this notion, we found that the mitochondrial membrane proteins TIMM23, COXIV, and TOMM20 were more abundant in PHZ-Fundc1-/- renal cells than in the control cells (Figure 3E,F).

The excessive mitochondrial accumulation may be due to either enhanced mitochondrial biogenesis or inefficient mitochondrial elimination. To distinguish between these two possibilities, we next examined mitochondrial biogenesis by determining the expression of the mitochondrial constitutive components of ATP synthase, including F1 subunit delta (Atp5d), nuclear respiratory factor 1 (Nrf1), and transcription factor A mitochondria (Tfam), and found no apparent differences between the kidneys of the PHZ-Fundc1-/- mice and the PHZ-treated control mice (Figure 3G). Thus, Fundc1 deletion has little impact on mitochondrial biogenesis. Instead, the expression of autophagic proteins of LC3-II and P62 in PHZ-Fundc1-/- renal cells was reduced (Figure 3E,F), suggesting that the accumulation of mitochondrial content is caused by defective autophagy.

The mitochondria, when damaged, are retained in the cytoplasm and inclined to release their DNA (mtDNA). Thus, to further confirm that deficient mitophagy was the cause of the accumulation of damaged mitochondria in PHZ-Fundc1-/- mouse kidneys, we examined the level of total and leaked cytosolic mtDNA from the lysed kidneys in PHZ-Fundc1-/- mice. With this aim, we removed the nuclei (to exclude the genomic DNA contamination) and mitochondria (to exclude intrinsic mitochondrial DNA contamination) (see Materials and methods) and discovered that the total level of mtDNA was elevated, while the cytosolic mtDNA was indeed induced in PHZ-Fundc1-/- renal cells (Figure 3H,I). Because cytosolic mtDNA acts as a critical trigger for cellular inflammatory responses (Allison, 2019), we then determined the level of inflammatory markers in renal cells and observed significant upregulation of proinflammatory cytokines, including Il6, Tnfa, and Il1b in PHZ-treated kidneys of Fundc1-/- mice (Figure 3J). Thus, Fundc1 deletion causes renal injury upon PHZ treatment by activating inflammation as a result of defective mitophagy.

Deficient mitophagy leads to reduced EPO production and increased inflammation

The inflammatory milieu in the kidney is one of the major contributors to renal fibrosis (Lv et al., 2018), which in turn facilitates REP conversion into myofibroblasts and the concomitant loss of EPO production capability (Asada et al., 2011; Souma et al., 2013). Therefore, it is reasonable to propose that the low EPO yield in REPs might result from their fibrotic transformation in PHZ-Fundc1-/- mice, which is induced by activation of the inflammatory milieu in the cytosol. To test this hypothesis, we co-stained α-smooth muscle actin (α-SMA), a broadly used myofibroblast marker (Duffield, 2014), and platelet-derived growth factor receptor beta (PDGFRβ), a REP marker (Asada et al., 2011; Gerl et al., 2016), in the renal tissue of PHZ-treated mice. We found that REPs exhibited stronger fibrotic signatures in PHZ-Fundc1-/- mice than in the PHZ-WT controls (Figure 4A).

Figure 4 with 1 supplement see all
Download asset Open asset
Impaired mitophagy triggers inflammation and reduced Epo expression in REPs.

(A) Representative immunofluorescence images (left) of PDGFRβ and α-SMA co-staining in renal tissue sections from WT and Fundc1-/- mice after 48 hr of PBS or PHZ treatment, respectively. The right bar graph shows the percentage of α-SMA positive fibrotic cells out of PDGFRβ labeled REPs. For quantification, approximately 10 fields for each mouse (200×) were randomly selected to evaluate the frequency of α-SMA+ cells out of PDGFRβ+ cells. Scale bars, 30 μm. (B) The schematic diagram showing the procedures of isolating REPs for flow cytometry assay. (C) The representative FACS plots showing the sorting strategy for PDGFRβ+ REPs. (D) Epo mRNA expression in PDGFRβ- and PDGFRβ+ cells by qRT-PCR. (E) Western blotting showing protein levels of mitochondrial membrane proteins (TIMM23, COXIV, and TOMM20) and autophagic proteins LC3 in FACS-sorted PDGFRβ+ REPs from WT and Fundc1-/- mice under the treatment of hypoxia (1% O2). (F) The quantification of the signal intensity of indicated mitochondrial membrane proteins (TIMM23, COXIV and TOMM20) and ratio of LC3II/I from (E). (G) Mitochondrial mass examined by using MitoTracker Green in FACS-sorted PDGFRβ+ REPs in WT and Fundc1-/- mice with or without PHZ treatment. (H) Flow cytometry analysis of mitophagy in REPs of mt-Keima/WT and mt-Keima/Fundc1-/- mice after 48 hr of PBS or PHZ treatment. (I) mtDNA copy number in FACS-purified REPs in WT and Fundc1-/- mice with or without PHZ treatment. (J, K) Mitochondrial membrane potential detected by using TMRM (J) or ROS level (K) measured by using MitoSOX in FACS-purified REPs of WT or Fundc1-/- mice after 48 hr of PBS or PHZ treatment. (L, M) Relative mRNA levels of proinflammatory cytokines Il1b, Il6 (L) and Epo (M) in sorted PDGFRβ+ REPs in WT or Fundc1-/- mice after 48 hr of PBS or PHZ treatment, respectively. For each experiment, n = 3–5 mice for each group. Individual mice are represented by symbols. Data shown are representative of at least three independent experiments. Similar results were found in each experiment. All data are mean ± SEM; *p<0.05, **p<0.01. Statistical significance was analyzed by using the two-tailed unpaired Student’s t-test.

Figure 4—source data 1

Inflammation and Epo expression in REPs.

https://cdn.elifesciences.org/articles/64480/elife-64480-fig4-data1-v2.xlsx

We next sought to dissect the molecular mechanism underlying the fibrotic transformation of REPs. We sorted PDGFRβ+ REPs from WT and Fundc1-/- mice (Asada et al., 2011; Lemos et al., 2018; Figure 4B,C) and then examined Epo mRNA levels in PDGFRβ+ and PDGFRβ- cells by using quantitative real-time PCR (qRT-PCR) analysis. As expected, PDGFRβ+ cells exhibited more Epo production (Figure 4D). To elucidate the potential role of FUNDC1-mediated mitophagy in REPs, we cultured purified PDGFRβ+ cells from WT and Fundc1-/- mice in vitro for 24 hr upon hypoxic stress stimulation (1% O2). We found that in WT REPs, the levels of mitochondrial membrane proteins (TIMM23, COXIV, and TOMM20) were sharply reduced upon hypoxic stress, concomitant with the enhanced conversion from LC3 I to II, a hallmark of autophagy (Kabeya et al., 2000), indicating that hypoxia promotes the degradation of mitochondria by enhancing mitophagy (Figure 4E,F). In contrast, these mitochondrial membrane proteins persisted at high levels in REPs of Fundc1-/- mice under hypoxic conditions when compared with the WT counterparts (Figure 4E,F), suggesting that mitophagy is impaired in REPs of Fundc1-/- mice under hypoxic stress.

To unravel the mitophagic activity in PDGFRβ+ REPs of PHZ-treated Fundc1-/- mice, we cultured purified PDGFRβ+ cells from WT and Fundc1-/- mice with or without PHZ treatment, respectively. Similar to hypoxia, PHZ also induced the accumulation of mitochondrial mass as revealed by MitoTracker staining in Fundc1-deleted REPs (Figure 4G). Furthermore, to quantify the efficiency of mitophagy more directly, we crossed Fundc1-/- mice with the mitophagy reporter (mito-Keima) (Sun et al., 2015), a mitochondrial-targeted transgenic mice, and generated the mito-Keima/Fundc1-/- mice (Figure 4H). Because Keima, a coral-derived fluorescent protein, exhibits both pH-dependent excitation and resistance to lysosome proteinases, the mito-Keima mice allow for in vivo assessment of mitophagy based on the state of mitochondria. In mito-Keima mice, the free mitochondria (pH ~8.0) without undergoing mitophagy are green, while damaged or superfluous mitochondria are encapsulated in the lysosomes (pH ~4.5) through mitophagy, and with the altered pH value (from pH 8.0 to 4.5) the mitochondria become red. In a single cell, both green and red colors can be seen if substantial mitochondrial biosynthesis and mitophagy occur. Thus, the ratio of red and green fluorescent intensity reflects the efficiency of mitophagy in vivo.

We next used flow cytometry to examine PDGFRβ+ REPs isolated from the mito-Keima/Fundc1-/- and mito-Keima/WT mice treated with either PBS or PHZ, respectively. Although the fraction of mitochondria in the lysosome was elevated in PHZ-REPs from both mito-Keima/Fundc1-/- (34.1 vs 48.4%) and mito-Keima/WT (32.7 vs 35.8%) mice, the proportion of encapsulated mitochondria was lower in PHZ-mito-Keima/Fundc1-/- than in PHZ-mito-Keima/WT mice (48.4 vs 35.8%) (Figure 4H). Thus, Fundc1 deletion in REPs impairs mitophagy and results in the retention of a large number of damaged mitochondria under PHZ-induced stresses.

To examine the effects (direct or indirect) of PHZ on EPO production in Fundc1-/- REPs, we FACS-sorted PDGFRβ+ REPs from the kidneys of mito-Keima/WT and mito-Keima/Fundc1-/- mice and treated them with various concentrations of PHZ in vitro for 12 hr (Figure 4—figure supplement 1A). Although a low dosage of PHZ (10 μM) had little effect on mitophagy (Figure 4—figure supplement 1B,C) and EPO production (Figure 4—figure supplement 1D), a high dosage (100 μM) of PHZ markedly impaired mitophagy and reduced EPO expression (Figure 4—figure supplement 1D) in Fundc1-/- REPs when compared with their WT counterparts. Thus, the high dosage of PHZ might have, at least in part, a direct impact on the Fundc1-/- REPs.

Consistent with and likely the result of the retention of the damaged mitochondria, we observed the accumulation of mtDNA copy number, reduction of mitochondrial membrane potential, and elevation of mitochondrial ROS level in PHZ-Fundc1-/- REPs (Figure 4I–K). At the molecular level, the level of proinflammatory cytokines, such as Il1b and Il6, was elevated (Figure 4L). Under such an inflammatory milieu, Epo production in PHZ-Fundc1-/- REPs was reduced (Figure 4M). These results suggest the following scenario: the reduction of EPO generation in PHZ-Fundc1-/- REPs is caused by deficient mitophagy, which in turn triggers the inflammatory responses and induces the fibrotic transformation of REPs. Thus, FUNDC1-mediated mitophagy plays a protective role in REPs under stresses.

Fundc1 deletion results in renal fibrosis

We next examined the potential link between mitochondrial quality in REPs and renal fibrosis and reduced RBCs. To achieve this goal, we applied a commonly used renal disease model, unilateral ureteral obstruction (UUO) mice that manifest inflammatory fibrosis (Guo et al., 1999; Inoue et al., 2010; Figure 5A). As shown in Figure 5B, Masson’s trichrome staining of the renal sections from UUO-Fundc1-/- mice revealed more positive blue staining of collagenous fibers, indicating that renal fibrosis was exacerbated with Fundc1 ablation. Consistently, fibrotic markers, such as Acta2 (encoding αSma) and collagen type 1 alpha one chain (Col1a1), were also significantly upregulated in kidneys of UUO-Fundc1-/- mice, while little difference was seen in the Fundc1-/- control (Figure 5C). In addition, the mitochondrial content was greatly reduced in kidneys of WT mice with UUO induction, as revealed by immunoblotting analysis of the mitochondrial proteins of TOMM20, COXIV, TIMM23, and FUNDC1 itself (Figure 5D). In contrast, mitochondrial elimination was partially blocked in the kidneys of UUO-Fundc1-/- mice (Figure 5D).

Figure 5
Download asset Open asset
More severe renal fibrosis and increased inflammation in Fundc1-/- mice with UUO model.

(A) Schematic diagram showing the generation of UUO pathological model from WT and Fundc1-/- mice. (B) Masson’s trichrome staining showing collagenous fiber (blue staining) of the renal section from UUO-Fundc1-/- and WT mice. (C) The expression of fibrosis related genes, α-Sma and Col1a1, in kidneys of sham or UUO-induced WT and Fundc1-/- mice. (D) Western blotting of the protein levels of mitochondrial membrane proteins (TOMM20, TIMM23, COXIV, and FUNDC1) and autophagy associated proteins (LC3 and P62) in kidney of sham, UUO-induced WT and Fundc1-/- mice. (E) The mRNA expression levels of proinflammatory cytokines Il6 and Il1b in kidney of sham, UUO-induced WT and Fundc1-/- mice. (F–H) Flow cytometry analysis of mitochondrial ROS level by MitoSOX (F), membrane potential by TMRM (G) and mitochondrial mass by MitoTracker Green (H) in FACS sorted, PDGFRβ+ REPs of kidneys in Sham, UUO-induced WT and Fundc1-/- mice, respectively. (I) Relative mRNA levels of Epo in sham, UUO-induced WT and Fundc1-/- mice. (J) The mRNA expression of DNA methyltransferase Dnmt1, Dnmt3a, and Dnmt3b in sham, UUO-induced WT and Fundc1-/- mice. For each experiment, n = 3–5 mice for each group. Individual mice are represented by symbols. Data shown are representative of at least three independent experiments. Similar results were found in each experiment. All data are mean ± SEM; *p<0.05; ns: no statistical significance. Statistical significance was analyzed by using the two-tailed unpaired Student’s t-test.

Figure 5—source data 1

Fibrosis and inflammation in UUO model.

https://cdn.elifesciences.org/articles/64480/elife-64480-fig5-data1-v2.xlsx

Furthermore, the level of autophagic genes, such as P62 and LC3-II, showed a trend of reduction in UUO-Fundc1-/- renal cells, suggesting the accumulation of dysfunctional mitochondria due to deficient autophagy (Figure 5D). Proinflammatory cytokines of Il6 and Il1b were upregulated in the UUO-Fundc1-/- renal cells, likely as a consequence of defective mitochondrial clearance (Figure 5E).

We next focused on mitophagy in the REPs of UUO-Fundc1-/- mice. We found that Fundc1 deletion induced an elevation of the mitochondrial ROS level (Figure 5F) and an alteration of the mitochondrial membrane potential (Figure 5G) in REPs of UUO-Fundc1-/- mice, with minor changes in mitochondrial mass (Figure 5H). As a result of the accumulation of dysfunctional mitochondria in REPs, Epo transcription was significantly impaired in Fundc1-/- mice (Figure 5I). Thus, the reduction of EPO production in Fundc1-/- REPs may result from active inflammation caused by abnormal mitophagy.

We next asked how aberrant mitochondrial dysfunction resulted in the suppression of Epo transcription. Given that the dysregulation of mitochondrial metabolism may alter the transcription program epigenetically (Matilainen et al., 2017), we then examined the expression of epigenetic regulators, such as DNA methyltransferases Dnmt1, Dnmt3a, and Dnmt3b, in REPs. We found that these DNA methyltransferases were upregulated in UUO-Fundc1-/- REPs when compared with their WT counterparts (Figure 5J). Thus, impaired Epo transcription might be associated with the changes in epigenetic modifications.

Fundc1 deletion aggravates renal anemia

To confirm the notion that Fundc1 deficiency contributes to renal anemia, we utilized cisplatin in Fundc1-/- mice to induce renal anemia (Figure 6A). Cisplatin is a widely used chemotherapeutic drug with notorious toxicity in the kidneys, leading to renal anemia (Kuzur and Greco, 1980; Wood and Hrushesky, 1995). We found that cisplatin-treated Fundc1-/- mice showed a fewer number of RBCs, lower HGB levels in the peripheral blood (Figure 6B), and a lower serum EPO level (Figure 6C) than in cisplatin-treated WT mice. Western blotting assay of mitochondrial membrane proteins (e.g. TOMM20 and COXIV) and autophagic proteins (e.g. P62, LC3II/I) revealed that Fundc1 ablation impaired the clearance of damaged mitochondria (Figure 6D) and caused elevated inflammatory signals (Figure 6E) in the kidneys of cisplatin-treated Fundc1-/- mice. More importantly, deficient mitophagy was pronounced in REPs, as revealed by the direct assessment of the encapsulated mitochondria in PDGFRβ+ cells in cisplatin-treated mito-Keima/Fundc1-/- mice (Figure 6F,G). We also detected increased mitochondrial mass, dysregulated mitochondrial membrane potential (indicated by TMRM), and enhanced mitochondrial ROS levels in the PDGFRβ+ REPs in cisplatin-Fundc1-/- mice (Figure 6H–J). As a consequence of the abnormal mitophagy, enhanced inflammation (e.g. Il6 and Tnfa) and reduced EPO generation were found in PDGFRβ+ REPs in cisplatin-Fundc1-/- mice (Figure 6K,L). Thus, Fundc1 deficiency impairs the clearance of damaged mitochondria and activates inflammation, which results in decreased EPO production in PDGFRβ+ REPs and, ultimately, aggravates the renal anemia (Figure 6M).

Figure 6
Download asset Open asset
Impaired mitophagy and reduced EPO production in Fundc1-/- mice during cisplatin-induced renal anemia.

(A) Schematic diagram showing the generation of renal anemia disease model by cisplatin. (B) Hemogram parameters of red blood cell (RBC) counts and hemoglobin (HGB) in the peripheral blood of WT controls and Fundc1-/- mice after PBS or cisplatin treatment. (C) The serum EPO concentration in Fundc1-/- and WT mice after PBS or cisplatin treatment. (D) Western blotting (left) and the corresponding quantification (right) showing the protein levels of mitochondrial membrane proteins (TOMM20 and COXIV) and autophagic proteins (P62 and ratio of LC3II/I) in kidneys of WT and Fundc1-/- mice with PBS or cisplatin treatment. (E) Relative mRNA levels of Il6, Il1b, and Tgfb1 in kidneys of WT and Fundc1-/- mice with or without cisplatin treatment. (F, G) Flow cytometry analysis of mitophagy in PDGFRβ+ REPs of mt-Keima/WT and mt-Keima/Fundc1-/- mice after PBS or cisplatin treatment. The representative FACS plots (F) and the quantification (G) are shown. (H) Mitochondrial mass detected by using MitoTracker Green via FACS in PDGFRβ+ REPs of WT or Fundc1-/- mice after PBS or cisplatin treatment. (I) Mitochondrial membrane potential detected by using TMRM by FACS in PDGFRβ+ REPs of WT or Fundc1-/- mice after PBS or cisplatin treatment. (J) Mitochondrial ROS level detected by using MitoSOX via FACS in PDGFRβ+ REPs of WT or Fundc1-/- mice after PBS or cisplatin treatment. (K) Relative mRNA levels of proinflammatory cytokines Il6 (left) and Tnfa (right) in sorted PDGFRβ+ REPs in WT or Fundc1-/- mice after PBS or cisplatin treatment. (L) Relative mRNA levels of Epo in sorted PDGFRβ+ REPs in WT or Fundc1-/- mice after PBS or cisplatin treatment, respectively. (M) A hypothetic model depicting the role of FUNDC1 in REPs. FUNDC1-mediated mitophagy is required for mitochondrial steady-state homeostasis. However, under stresses, the damaged mitochondria are accumulated in Fundc1-/- REPs due to impaired mitophagy. Consequently, elevated ROS levels from these damaged mitochondria of REPs incur inflammatory responses by enhancing the expression of proinflammatory cytokines TNFa, IL6 and IL1b, which in turn promote myofibroblastic transformation of REPs, resulting in the loss of EPO generation and subsequently anemia. For each experiment, n = 3–5 mice for each group. Individual mice are represented by symbols. Data shown are representative of at least three independent experiments. Similar results were found in each experiment. All data are mean ± SEM; *p<0.05, **p<0.01, ***p<0.001. Statistical significance was analyzed by using the two-tailed unpaired Student’s t-test.

Figure 6—source data 1

Mitophagy and EPO production during cisplatin-induced renal anemia.

https://cdn.elifesciences.org/articles/64480/elife-64480-fig6-data1-v2.xlsx

Discussion

FUNDC1 in stress versus steady-state erythropoiesis

Our current study showed that FUNDC1 is required for EPO production upon PHZ-induced stresses (Figure 6M). Specifically, we found that FUNDC1-mediated mitophagy is essential for EPO production in REPs during stress-induced kidney injuries and that damaged mitochondria accumulate in Fundc1-/- REPs as a result of impaired mitophagy. Consequently, an elevation of the ROS level and the release of mtDNA from damaged mitochondria in REPs incur inflammatory responses by enhancing the expression of proinflammatory cytokines including TNFa, IL6, and IL1b. These changes in the inflammatory milieu promote myofibroblastic transformation of REPs and the concomitant loss of EPO production capacity and, ultimately, lead to renal anemia. To the best of our knowledge, we provide the first link between mitophagy and the inflammatory state in REPs, demonstrating that mitophagy and mitochondrial quality control play protective roles in REP function and renal anemia.

Additionally, it is somewhat surprising that FUNDC1 is dispensable for the programmed mitochondrial elimination in the final step of erythroid maturation, whereas another mitophagy receptor BNIP3L is required for the removal of mitochondria in reticulocytes during terminal erythroid differentiation (Schweers et al., 2007; Sandoval et al., 2008). However, during cardiac lineage differentiation, both FUNDC1 and BNIP3L are required for the remodeling of the mitochondrial network (Lampert et al., 2019). Thus, it appears that the function of FUNDC1 and BNIP3L is highly context- and cell-type dependent.

Mitochondrial quality and EPO production in REPs

Renal EPO production is central for erythropoiesis. It has been well-established that HIF2α plays essential roles in hypoxia-induced EPO gene expression (Gruber et al., 2007; Semenza et al., 1991). In this study, we found that the expression of HIF2α is largely unaffected by Fundc1 deletion, indicating that reduced EPO production in Fundc1-/- mice is likely to be unrelated to the oxygen-sensing mechanism. Instead, the low yield of EPO may be due to the reduced number of REPs and/or compromised function during REP transformation to myofibroblasts. Nonetheless, we found that the fraction of PDGFRβ+ REPs is unchanged in Fundc1-/- mice when compared with the WT controls under stresses (data not shown). However, in the purified REPs from Fundc1-/- mice, we showed that the increase in the mitochondrial ROS levels, as a result of defective mitophagy, provokes inflammatory responses, including the upregulation of TNFa, IL6 and IL1b in REPs, and subsequently lead to fibrotic transformation. Although prior studies demonstrated that activation of inflammatory signals, such as NFκB (Souma et al., 2013; Liu et al., 2012b), TNFa (Frede et al., 1997), IL1b (Frede et al., 1997), IL6 (Stenvinkel et al., 2005), repress EPO production in REPs and promote their transition to myofibroblasts (Guo et al., 1999; Inoue et al., 2010; Souma et al., 2013; Ernandez and Mayadas, 2009), in this study we provide the first link between deficient mitophagy and EPO production in REPs with the induction of inflammation as the intermediate step. We argue that inefficient mitophagy caused by Fundc1 deletion results in the loss of EPO production in REPs, resembling the clinical manifestations of renal anemia. Therefore, by selectively removing the damaged mitochondria, mitophagy represents an essential element of mitochondrial quality control in response to renal injury or anemia caused by UUO, PHZ, or cisplatin. Although PDGFRβ is the most frequently used surface marker to identify or isolate REPs (Gerl et al., 2016; Greenwald et al., 2019; Broeker et al., 2020), PDGFRβ+ cells are heterogenous and dynamic (Pan et al., 2011). Therefore, the role of mitophagy in REPs awaits future investigation.

In addition, defective mitophagy might also result in metabolic reprogramming, which is widely observed in renal epithelial cells or myofibroblast transformation (Xu et al., 2019; Zhao et al., 2017; Livingston et al., 2019), and the alteration in mitochondrial metabolism may regulate EPO production epigenetically. In line with this speculation, DNMTs, including DNMT1, DNMT3a, and DNMT3b, are upregulated, suggesting that altered epigenetic modifications might be responsible for the reduced EPO production in Fundc1-/- REPs, consistent with previous observations in mouse myofibroblast-transformed REPs (Souma et al., 2013; Chang et al., 2016) and the cell lines generated from EPO-producing cells without the capabilities of EPO production (Yin and Blanchard, 2000; Steinmann et al., 2011; Sato et al., 2019). Thus, our current study further highlights the promise of applying demethylating agents to the restoration of EPO expression in patients with renal anemia.

Mitochondrial dysfunction, renal fibrosis, and renal diseases

Renal anemia is one of the most common complications among various chronic kidney diseases (Quaggin and Kapus, 2011). Renal anemia adversely affects the quality of patients’ life and is strongly associated with poor clinical outcomes (Collins et al., 1998). The treatment of renal anemia has long been dependent on erythropoiesis-stimulating agents and iron supplement in chronic kidney diseases, raising the concerns of increased patients’ death and cardiovascular risk (Solomon et al., 2010; Bohlius et al., 2009). In the past decades, tremendous efforts have been focused on understanding the cellular basis and molecular regulatory mechanism of REPs, in order to offer more efficient and safer treatment options.

Myofibroblastic transformation of REPs, which causes REPs to lose their EPO-producing ability (Koury, 2014), is one of the major sources of myofibroblasts contributing to renal fibrosis (Souma et al., 2013). Despite the established correlation between inflammation and REP fibrotic conversion, it is largely unclear how the upstream pathological inflammatory pathways are triggered in REPs and how mitochondrial quality control is linked to REPs in chronic kidney disorders or injuries. In this study, we showed that the kidney of Fundc1-/- mice displays a more severe fibrotic phenotype in UUO model. Concomitant with the inability of REPs to secret EPO, these cells undergo myofibroblastic conversion. Thus, receptor-mediated mitophagy protects against renal anemia and possible renal fibrosis. In addition, we found that Fundc1 is abundantly and widely expressed in the kidney. It is therefore highly likely that FUNDC1-mediated mitophagy might also play crucial roles in other types of renal cells besides REPs, and that these cells might also contribute to the death of Fundc1-/- mice under PHZ stresses. For example, during this study, it was reported that FUNDC1-dependent mitophagy in tubule cells plays protective roles in ischemic acute kidney injury (Wang et al., 2020). It appears that targeting FUNDC1-mediated mitophagy should have pleiotropic therapeutic impacts by ‘killing two birds with one stone’. In renal fibrosis, enhanced mitophagy might not only ameliorate anemia by attenuating inflammation in REPs, but also alleviate fibrosis by preserving and protecting the mitochondrial quality from other fibrogenic contributing cells. These exciting possibilities should be examined in future investigations.

Materials and methods

Mouse experiments

Request a detailed protocol

Mice were maintained in the animal core facility of College of Life Sciences, Nankai University, Tianjin, China. All experiments involving animals were reviewed and approved by the Animal Care and Use Committee of Nankai University and were performed in accordance with the university guidelines. All mice in this study were of the C57BL/6 background. Fundc1 whole-body knockout mice (Fundc1-/-) were generated as previously described (Zhang et al., 2016). The WT and Fundc1-/- mice around 8–10 weeks old were injected intraperitoneally with phenylhydrazine (PHZ, cat# P26252, Sigma, 100 mg/kg body weight) to induce acute hemolysis anemia. For the complementary experiments, after 12 hr of PHZ treatment, EPO (cat# 46130, Epiao, 1000 IU/kg body weight) was injected subcutaneously into PHZ-Fundc1-/- and PHZ-WT mice three times with an interval of 12 hr (Nai et al., 2016; Patel et al., 2004; Kwak et al., 2020). These mice were sacrificed at 48 hr after PHZ treatment for further analyses. We induced pure anemia via phlebotomy. In detail, 500 μL of blood was taken from the tail vein of mice for three consecutive days, and they were sacrificed for further analyses (Casanovas et al., 2013; Kautz et al., 2014). To generate the unilateral ureteral obstruction (UUO) mice model, the left ureter was obstructed by ligation with a suture or by the application of a ligating clip in adult mice. Kidneys were harvested on day 7 after unilateral ureteral obstruction for further analysis, as previously described (Hesketh et al., 2014). We also injected cisplatin (cat# S1166, Selleckchem, 10 mg/kg body weight) intraperitoneally into WT and Fundc1-/- mice around 8–10 weeks old to induce renal anemia. These mice were sacrificed at day 14 after cisplatin injection for further analyses (Kuzur and Greco, 1980; Wood and Hrushesky, 1995; Wang et al., 2018). Mitochondrial-targeted transgenic mice (mt-Keima) were crossed with Fundc1-/- mice to breed the mt-Keima/Fundc1-/- mice.

Flow cytometry and cell sorting

Request a detailed protocol

Primary antibodies used for flow cytometry were as follows: CD71 (clone RI7217, BioLegend), Ter-119 (clone TER-119, BioLegend), CD3 (clone 17A2, BioLegend), B220 (clone RA3-6B2, BioLegend), LY6G (clone 1A8, BioLegend), LY6C (clone HK1.4, BioLegend), F4/80 (clone BM8, BioLegend), Sca1 (clone D7, BioLegend), c-Kit (clone 2B8, BioLegend), CD34 (clone RAM34, eBioscience), CD16/32 (clone 93, BioLegend), PE/Cy7-Streptavidin (cat# 405206, BioLegend), Biotin-CD140b (clone APB5, BioLegend), and Brilliant Violet 605 Streptavidin (cat# 405229, BioLegend). The erythroid cells were stained with CD71 and Ter119; hematopoietic stem and progenitor cells were immunophenotypically labeled as LSK (Lin-Sca1+c-Kit+) and LK (Lin-c-Kit+); common myeloid progenitor (CMP) cells were labeled as Lin-Sca1+c-Kit+CD34+CD16/32−; granulocyte-macrophage progenitor (GMP) cells were labeled as Lin−c-Kit+CD34+CD16/32+; and megakaryocyte-erythroid progenitor (MEP) cells were labeled as Lin−c-Kit+CD34−CD16/32−. Neutrophils, T cells, B cells, and monocytes were marked as Gr1+, CD3+, B220+, and F4/80+, respectively.

Fluorescence-activated Cell Sorting (FACS) was conducted as previously described (Liu et al., 2018). Briefly, the cells were collected, washed, and resuspended in ice-cold phosphate-buffered saline (PBS) supplemented with 2% FBS. For each assay, approximately 106 cells were stained with antibodies at 4°C for 30 min in the dark. Cells were then washed and resuspended in 300 μL ice-cold PBS–FBS and subjected to flow cytometry analysis on a FACS LSRII flow cytometer (BD Biosciences).

Mitochondrial assays

Request a detailed protocol

The mitochondrial membrane potential was analyzed using tetramethylrhodamine methyl ester perchlorate (TMRM, cat# T668; Invitrogen) staining, following the manufacturer’s instructions. Briefly, cells were stained with TMRM at a final concentration of 200 nM for 30 min at 37°C and then stained with Ter119 and CD71 antibodies before flow cytometry. MitoTracker Green (MTG; cat# M7514, Thermo Fisher Scientific) was used to analyze mitochondrial content following the manufacturer’s instructions. Cells were stained with MTG at a final concentration of 200 nM for 30 min at 37°C and then stained with Ter119 and CD71 antibodies before flow cytometry analysis. The mitochondrial ROS level was analyzed with the MitoSOX red mitochondrial superoxide indicator (cat# M36008, Thermo Fisher Scientific) according to the manufacturer’s instructions. Cells were stained with MitoSOX red mitochondrial superoxide indicator at a final concentration of 5 μM for 30 min at 37 °C and then stained with Ter119 and CD71 antibodies before flow cytometry analysis. Data were analyzed with FlowJo (version 7.6.1).

Ex vivo maturation of reticulocytes

Request a detailed protocol

FACS-sorted Ter119+CD71+ reticulocytes from the peripheral blood were cultured for 3 days in IMDM medium (I3390, Sigma) containing 2 mM l-glutamine (cat# 25030, Gibco) and 100 U penicillin-streptomycin (cat# DE17-602E, Lonza), 30% FBS (cat# 10270, Gibco), 1% BSA (cat# A1595, Sigma), and 0.001% monothioglycerol. In the final stage, 10 μM FCCP (HY-100410, MCE) was added to the cultured reticulocytes.

Western blotting

Request a detailed protocol

Western blotting was performed as previously described (Liu et al., 2018). Briefly, cells were lysed in Laemmli sample buffer (cat# 161–0737, Bio-Rad) before SDS–PAGE. Primary antibodies used for western blotting were as follows: anti-FUNDC1 (home-made) (Liu et al., 2012a), anti-LC3 (cat# L8918, Sigma), anti-TIMM23 (cat# 611222, Biosciences), anti-TOMM20 (cat# 612278, Biosciences), anti-STAT3 (cat# 9139, Cell Signaling Technology), anti-phosphorylated STAT3 (cat# 9134, Cell Signaling Technology), anti-STAT5 (cat# 94205, Cell Signaling Technology), anti-phosphorylated STAT5 (cat# 4322, Cell Signaling Technology), anti-HIF2α (cat# NB100-122, Novus Biologicals), anti-βACTIN (cat# 8H10D10, Cell Signaling Technology), and anti-αTUBULIN (cat# ab52866, Abcam). Horseradish peroxidase-conjugated secondary antibodies were used, and signals were detected using an ECL kit (cat# WP20005, Invitrogen).

ELISA of cytokines

Request a detailed protocol

The sera collected from the mice with distinct genotypes were stored at −80°C. Inflammatory cytokines were measured using Mouse Inflammation Panel (13-plex) with V-bottom Plate (cat# 740446, BioLegend), according to the manufacturer’s instructions.

The concentrations of EPO were measured using ELISA kit (cat# ab119593, Abcam). Briefly, the standards and samples were added to the wells and incubated for 90 min at 37°C. After samples were washed, a biotinylated goat polyclonal antibody specific for EPO was supplemented. Finally, avidin-biotin-peroxidase complex was incubated for 30 min for the enzymatic reaction. The signal was assessed by using a microplate reader (Synergy H4, BIO-TEK).

Histological analysis of the renal tissue sections

Request a detailed protocol

Kidneys were fixed in 4% paraformaldehyde overnight at room temperature and dehydrated in 20% sucrose overnight on the next day. The paraffin-embedded tissues were then cut into 4 μm thick sections. For staining, the slides were dewaxed and incubated with mouse anti-α-Smooth Muscle-Cy3 antibody (cat# 1A4, Sigma, 1:200) and anti-PDGFRβ antibody (cat# 28E1, Cell Signaling Technology, 1:200) at 4°C overnight, followed by staining with an Alexa Fluor-conjugated secondary antibody. Fluorescent images were obtained using a confocal microscope (UltraVIEW Vox, PerkinElmer).

A TUNEL assay was used to detect apoptotic cells (cat# 12156792910, Roche Applied Science), according to the manufacturer’s instructions. In brief, tissue sections were deparaffinized and permeabilized with 0.1 M sodium citrate and incubated with a TUNEL reaction mixture for 1.5 hr at 37°C in a humidified, dark chamber. For quantification, approximately 20 fields (200×) were randomly selected to evaluate the TUNEL-positive cells.

Measurement of cytosol-releasing mitochondrial DNA (mtDNA) in kidney

Request a detailed protocol

Fifty to 100 mg renal tissues were collected and washed with PBS. The tissues were then cut into small pieces before the addition of 10 volumes of pre-chilled mitochondrial separation reagent A and homogenized for 10–15 times. The produced homogenates were centrifuged at 600 g at 4°C for 5 min to remove the cell nuclei, while the supernatant was centrifuged at 12,000 g at 4°C for 10 min to further remove the intact mitochondria. The remaining cytosolic free mtDNA was purified and quantified by using qRT-PCR with mtDNA-specific primers. The primers are listed in Supplementary file 2.

RNA extraction and qRT-PCR

Request a detailed protocol

Total RNA was extracted with TRIzol (cat# 15596026, Invitrogen) and reversely transcribed using TransScript II One-Step gDNA Removal and cDNA Synthesis SuperMix (cat# AH311-03, TransGen Biotech). The qRT-PCR was completed using the SYBR Green Master Mix kit (cat# A25742, Thermo Fisher Scientific), and the assay was conducted with the QuantStudio5 Real-Time PCR System (Life Technologies). The primers are listed in Supplementary file 2.

Isolation of PDGFRβ+ REPs

Request a detailed protocol

Kidneys were decapsulated, diced on ice, and then incubated in RPMI 1640 medium containing 0.5 mg/mL liberase (cat# 423610, Roche) and 100 U/mL DNase I (cat# D8071, Solarbio) at 37°C for 45 min, followed by supplementation of 0.1% bovine serum albumin to stop the enzymatic reaction. After filtration and red blood cell lysis, the remaining single cells were incubated with PDGFRβ+ antibody for 20 min at 4°C before flow cytometry analysis.

RNA-sequencing and data analysis

Request a detailed protocol

Total RNA from FACS-purified R2 cells of the spleen of PHZ-treated WT (n = 3) and Fundc1-/- mice (n = 3) were extracted with TRIzol (cat# 15596026, Invitrogen). Library construction and data processing were performed by Novogene (Beijing, China), as described previously (Liu et al., 2018). In brief, quality control was conducted by using FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/), and clean reads were aligned by HISAT2 (Pertea et al., 2016) to the mouse reference genome GRCm38 (https://www.gencodegenes.org/) of gencode.vM17 (https://www.gencodegenes.org/). Next, the expression of genes was quantified by using StringTie (Pertea et al., 2016) with default parameters, and differentially expressed genes (DEGs) were identified by using DESeq2 (Love et al., 2014) within R package. Genes with fold change > 1.5 and p- value < 0.05 were considered as significant DEGs.

Functional annotation of DEGs

Request a detailed protocol

Functional annotation of significant DEGs was performed using ‘Investigate Gene Sets’ in GSEA (https://www.gsea-msigdb.org/gsea/msigdb/annotate.jsp). We selected KEGG (Kyoto Encyclopedia of Genes and Genomes) as our target database, while the p-values were adjusted using the Benjamini and Hochberg method with significant enrichment as FDR < 0.05.

Gene set enrichment analysis

Request a detailed protocol

The JAK-STAT gene set derived from MSigDBwas used for GSEA (http://software.broadinstitute.org/gsea/index.jsp). All of the expressed genes (average FPKM > 0.01) from different groups were used as the input dataset. The numbers of permutations were set to 1000. p-values were corrected for multiple testing methods, and the threshold for significantly enrichment was FDR < 0.05.

Statistical analysis

Request a detailed protocol

Statistical significance was calculated using the two-tailed unpaired Student’s t-test unless stated otherwise. Data in bar graphs were represented as mean ± SEM, and statistical significance was expressed as follows: *p<0.05; **p<0.01; ***p<0.001; ns: no significant.

Data availability

RNA-Sequencing data is deposited at GEO Accession number GSE 158361. Information on replicates is presented in Materials and Methods as well as in figure legend. Replicate numbers are mentioned in figure legends. All data generated or analysed during this study are included in the manuscript and supporting files.

The following data sets were generated
    1. Geng G
    (2021) NCBI Gene Expression Omnibus
    ID GSE158361. Receptor-mediated mitophagy regulates EPO production and protects against renal anemia.
    https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE158361

References

Decision letter

  1. Jian Xu
    Reviewing Editor; University of Texas Southwestern Medical Center, United States
  2. Matt Kaeberlein
    Senior Editor; University of Washington, United States

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

Mitochondrial dysfunction is often observed in renal and red blood cell disorders, however the role of mitochondrial quality control in renal anemia remains unclear. This study describes that FUNDC1, a mitophagy receptor, is required for EPO production upon stresses due to impaired mitochondrial quality control in renal EPO-producing cells. These findings support a new mechanism that FUNDC1-mediated mitophagy regulates EPO production under stresses to protect anemia of kidney disease.

Decision letter after peer review:

Thank you for submitting your article "Receptor-mediated mitophagy regulates stress erythropoiesis and confers protection against renal anemia" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Matt Kaeberlein as the Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

As the editors have judged that your manuscript is of interest, but as described below that additional experiments are required before it is published, we would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). First, because many researchers have temporarily lost access to the labs, we will give authors as much time as they need to submit revised manuscripts. We are also offering, if you choose, to post the manuscript to bioRxiv (if it is not already there) along with this decision letter and a formal designation that the manuscript is "in revision at eLife". Please let us know if you would like to pursue this option. (If your work is more suitable for medRxiv, you will need to post the preprint yourself, as the mechanisms for us to do so are still in development.)

Summary:

All the reviewers find the work of interest by analyzing the erythroid defects caused by deletion of mitophagy receptor FUNDC1 in mice. The study reveals that FUNDC1 loss results in the accumulation of impaired mitochondria in renal EPO producing cells, inflammation, myofibroblastic transformation, and reduced EPO production and anemia in experimental models of acute anemia or kidney injury. The results suggest that FUNDC1 is required for mitochondrial removal during erythropoiesis, and provide new insights into the anemia of kidney disease. The reviewers have raised several major concerns related to the strengths of the main conclusions. The complete comments are included.

Reviewer 1 also suggested that the title of the manuscript and other relevant texts need to be revised to better reflect the main findings or additional evidence is needed to support the conclusions related to stress erythropoiesis.

Essential revisions:

1. It is important to provide more convincing evidence to support the conclusions related to severe anemia or stress erythropoiesis observed in FUNDC1-deleted mice, or the major conclusions should be significantly revised.

2. Evidence is needed to show whether defective mitophagy results in reduced Epo production, thus supporting the conclusion that the defective mitophagy observed in FUNDC1-deleted kidney is responsible for the decreased Epo production and associated phenotypes. A more direct experiment is required to assess the effect of PHZ on Epo production in Fundc1-/- kidney cells. To further support the hypothesis, supplemental Epo administration should be performed to examine the effects on the abnormalities observed in the Fundc1-/- mice.

3. It is important to provide evidence for mitochondrial retention in a model for kidney disease that is accompanied by anemia. Such studies will strengthen the conclusion that FUNDC1 deficiency contributes to anemia associated with kidney disease by regulating mitophagy.

4. Given the modest effects on blood cell production, the conclusions that FUNDC1-mediated mitophagy is not required for reticulocyte maturation need to be qualified or significantly revised.

5. PHZ leads to significant reduction of Epo transcript in FUNDC1-deleted kidney cells but only a mild reduction of serum Epo level. This discrepancy should be explained.

6. Given the modest effects observed in FUNDC1-deleted mice and the limitations of the PHZ-induced acute anemia, it is important to include studies using a model for chronic anemia, and examine the effects of modest Epo deficiency effects over time. For example, phlebotomy may help to distinguish the effects of oxidative stress versus pure anemia. Repeated phlebotomy may help to assess the role of Fundc1 in chronic anemia could have a more profound effect than what was observed after a single PHZ treatment.

7. Given the limitations of analyzing p62 and LC3 expression to represent altered autophagy, additional experiments are required to support conclusions related to autophagy.

Reviewer #1:

The authors have contributed substantially to the understanding of FUNDC1 as a mitophagic receptor in multiple cell types and disease conditions including hypoxia-mediated mitophagy in mammalian cells, inhibition of inflammasome activation, ischemia-reperfusion injury to the heart. Here the authors turned their attention to the kidney and showed that lack of FUNDC1 results in the accumulation of impaired mitochondria in EPO producing cells (REPs), leading to an inflammatory response, a myofibroblastic transformation of REPs, resulting in reduced EPO production and anemia in two experimental models including PHZ and UUO. Work presented here indicates that FUNDC1 does not play a role in mitochondrial removal during erythropoiesis. Overall, the paper is well written and presents strong evidence showing the importance of FUNDC1-mediated mitophage in renal fibrosis-mediated anemia by reducing the Epo production. However, the paper can be improvde by addressing following major concern:

The title and manuscript's emphasis of FUNDC1-mediaetd mitopahge regulting "stress eythropoiesis" are misleading. One of major complications of renal fibrosis is anemia due to the reduced production of Epo. Thus, simply because a particular protein deficiency results in renal fibrosis and reduced EPO production does not mean that such a protein should be considered as a regulator of "stress erythropoiesis" as implied by the title and multiple places in the manuscript. The findings showed here support that FUNDC1 protects against the accumulation of impaired mitochondria, inflammation, renal fibrosis and reduced EPO production and thus the development of major complication of anemia but not stress erythropoesis. Please consider to revise the title and text to avoid misleading and confusion.

Reviewer #2:

The manuscript by Shi et al., analyzes the effect of deletion of mitophagy receptor FUNDC1 in mice on erythropoiesis and kidney production of erythropoietin (Epo). The authors have performed a considerable amount of work and generated a large set of data. Since the authors did not find any effect of FUNDC1 deletion on hematopoiesis or erythropoiesis, they subjected mice to erythroid stress by inducing hemolytic anemia using the established phenylhydrazine (PHZ) treatment. Mice survival in response to phenylhydrazine was reduced and attributed to severe anemia. They noticed a mild reduction in Epo production in response to phenylhydrazine. They then focused on kidney production of Epo. They confirm that FUNDC1-deleted kidney cells exhibit defective mitophagy, and PHZ treatment enhanced defective mitophagy in FUNDC1-deleted kidney cells. They attributed this defect to reduced Epo production. There are two major problems in this manuscript. One is that in contrary to the authors assertion, it is unclear whether FUNDC1-deleted mice exhibit anemia (hemoglobin is not reported in PHZ-treated mice and other RBC indices are very mild) but possibly only a reduced compensated red blood cell production in response to phenylhydrazine treatment. The other is that although the authors confirm that FUNDC1-deleted kidney cells exhibit defective mitophagy that is more pronounced in PHZ-treated animals, they do not show that reduced Epo production results from defective mitophagy. My specific comments are below:

– There is no evidence of mitochondrial retention in the spleen, BM and PB. The authors use CD71 as a marker but it is unclear on which cells.

– Although the survival in response to phenylhydrazine is reduced which was achieved by using 100mg/kg (instead of the usual dose of 50mg/kg) and including 30 mice (p=0.04), there was no severe anemia. In fact Hb levels are normal and the slight reduction in red blood cell indices suggest that there was a mild reduced red blood cell production that did not reach the definition of anemia. So, in the absence of anemia the claim that Fundc1-mediated mitophagy is not required for reticulocyte maturation is not accurate.

– It remains unclear what animals die from.

– The authors FACS-sort reticulocytes 5 days post phenylhydrazine. This is an interesting experiment. What conditions for the FACS sorting do they use?

– From Figure S2 it seems that the authors FACS-sorted immature erythroblasts but not reticulocytes, the data on mitochondria in these cells also seems to be from immature erythroid cells. The authors should clarify how they distinguish reticulocytes from immature erythroblasts.

– The authors state: "Together, our results suggest that FUNDC1- mediated mitophagy is

dispensable for programmed mitochondrial elimination in reticulocytes during

erythroid maturation." It is difficult to assess the validity of this statement given that it is unclear which cells were used in these experiment.

– The authors examine differential gene expression in R2 erythroblast subpopulation between cells from mice subjected to PHZ treatment or not. They use RNA-Seq analysis of R2 and identify among others JAK-STAT pathway to be differentially expressed in PHZ treated Fundc1-/- spleens. They conclude that JAK-STAT is hypoactivated in Fundc1-/- R2 cells. Given that the reduced RBC production seem to be compensated (see R3, R4) does not seem to be highly affected this result seems to indicate compensated JAK-STAT hypoactivation.

– How do the authors explain that Epo reduction in Fundc1-/- cells has an effect solely on R2 spleen cells?

– The significant reduction of Epo transcript in kidney cells in response to PHZ is interesting but how does this correlate with a mild reduction of Epo in PB?

– To determine the role of defective mitophagy on Epo production, the authors selected PDGFRB+ and – cells and subjected to hypoxia. A more direct experiment is to assess the effect of PHZ on Epo production in Fundc1-/- kidney cells instead they have measured mitochondrial mass which does not address their hypothesis.

– How are experiments in 4M and 3A different? What is missing is the effect of PHZ on Fundc1-/- PDGFRB+ cells?

Reviewer #3:

The authors show that loss of the mitophagy receptor Fundc1 in mice causes defects in stress erythropoiesis by impairing the production of Epo by the kidney, with defective mitophagy of Epo-producing cells. This is interesting work that has the potential to provide insights into the anemia of kidney disease. However, the overall effects seem to be fairly modest and several of the studies must be investigated further to support the claims to a reasonable level. Notably, modest impairment of Epo production could lead to significant anemia over time, but this is not shown. Perhaps the most important problem is that study does not examine a model for anemia of kidney disease.

1. Following PHZ administration to the Fundc1-/- mice, the reductions in blood HCT and RBC, splenic erythroblast number and serum Epo concentration relative to WT mice are very mild. So while Fundc1 may have a role in maintaining Epo production during stress, the effect seems to be relatively minor, at least as revealed in the current assay. As the authors point out, Fundc1 may have a role in chronic stress erythropoiesis. In this regard, they may want to examine a more chronic model for anemia to examine the effects of modest Epo deficiency effects over time (see point #3).

2. To support further their hypothesis that delayed recovery from PHZ is not due to erythroid-intrinsic problems, the investigators should show that supplemental Epo administration eliminates the abnormalities observed in the Fundc1-/- mice.

3. The effects of Fundc1 loss on serum Epo levels after PHZ are relatively modest (Figure 2I), while the effects on kidney Epo mRNA are more extreme (Figure 3A). How can this discrepancy be explained? Perhaps measuring the kinetics of these parameters at a few timepoints after PHZ administration would help to reconcile these differences and be informative.

4. It might be useful to examine a different model for stress anemia. PHZ is a strong oxidant. Therefore, its effects on REPs could be due to oxidative stress rather than the erythropoietic demand of anemia. Investigating the effects of phlebotomy might be useful to distinguish the effects of oxidative stress from pure anemia. Additionally, repeated phlebotomy might allow the investigators to assess the role of Fundc1 in chronic anemia where the effects of Epo deficiency over time could have a more profound effect than what was observed after a single PHZ treatment.

5. The authors use one week of UUO treatment as a model for acute kidney disease. They show in this model that loss of Fundc1 causes enhanced accumulation of inflammatory markers, increased mitochondrial ROS and modest reductions in Epo compared to WT mice. However, they do not show that the model causes anemia in WT or Fundc1-/- mice. In order to substantiate a major claim of this manuscript, it is essential that the investigators examine the effects of Fund1c loss in a model for anemia of kidney disease, preferably a chronic one.

6. One central hypothesis is that Fundc1 deficiency promotes retention of abnormal mitochondria, inflammation and transformation to myofibroblast (abstract and line 470). However, the data do not show this. Rather, they report induction of some fibroblastic markers but constant numbers of REPs. Would it be possible to show this data?-- specifically co-staining of kidney cells for Epo mRNA or protein and PDGFb to quantify the number of REPs and their Epo induction.

7. In general, all of the histology slides in this manuscript are not well explained. For example, the histology in Figure S4C should be interpreted. What are the signs of renal injury? Is there evidence of injury throughout the kidney or specifically in Epo producing cells? The figure legends of all micrographs should include interpretation of the histopathology with the abnormal regions indicated on the micrograph.

8. Figure 3 Western blots – it is difficult to appreciate some of the differences in protein expression between WT vs mutant kidney. All western blots should include quantification of multiple experiments.

9. In several experiments, the authors interpret altered expression of p62 and LC3 as representing altered autophagy. This is not correct. Autophagy rates cannot be inferred by steady state measurements of these proteins. Enhanced autophagy could also deplete these proteins! Use of the Keima reporter seems more correct. Along the same lines, if Fundc1 is the mitochondrial autophagy adaptor in kidney, then what is the role of LC3 or p62 and how would their altered expression relate to mitophagy versus other autophagy pathways?

https://doi.org/10.7554/eLife.64480.sa1

Author response

Summary:

All the reviewers find the work of interest by analyzing the erythroid defects caused by deletion of mitophagy receptor FUNDC1 in mice. The study reveals that FUNDC1 loss results in the accumulation of impaired mitochondria in renal EPO producing cells, inflammation, myofibroblastic transformation, and reduced EPO production and anemia in experimental models of acute anemia or kidney injury. The results suggest that FUNDC1 is required for mitochondrial removal during erythropoiesis, and provide new insights into the anemia of kidney disease. The reviewers have raised several major concerns related to the strengths of the main conclusions. The complete comments are included.

We appreciate that the reviewers suggest our work provides “new insights into the anemia of kidney diseases”. Below, we have performed new experiments to address each of the reviewers’ comments and articulated the point-by-point responses.

Reviewer 1 also suggested that the title of the manuscript and other relevant texts need to be revised to better reflect the main findings or additional evidence is needed to support the conclusions related to stress erythropoiesis.

We agree with the reviewers and have changed the title from “Receptor-mediated mitophagy regulates stress erythropoiesis and confers protection against renal anemia” to “Receptor-mediated mitophagy regulates EPO production and protects against renal anemia”. We also generated the renal anemia (Figure 6) and chronic anemic mice model (Figure 2—figure supplement 1C-G) to further support our conclusions. In addition, the text was revised accordingly to reflect our findings.

Essential revisions:

1. It is important to provide more convincing evidence to support the conclusions related to severe anemia or stress erythropoiesis observed in FUNDC1-deleted mice, or the major conclusions should be significantly revised.

Thanks for the insightful comment. To strengthen anemia observed in Fundc1-/- mice upon stresses, we incorporated an additional renal anemia mouse model generated via the treatment with cisplatin (Figure 6). We indeed detected severer anemia in cisplatin-Fundc1-/- mice than -WT controls (Figure 6A-C) (page 17, lines 436-442 (highlighted in blue)). We also revised the text to describe our results more precisely. For example, instead of stating “severe anemia”, we revised the text as “reduced RBC numbers and HCT values” in Fundc1-/- mice upon PHZ-induced acute hemolytic anemia. (page 8, lines 156-159).

2. Evidence is needed to show whether defective mitophagy results in reduced Epo production, thus supporting the conclusion that the defective mitophagy observed in FUNDC1-deleted kidney is responsible for the decreased Epo production and associated phenotypes. A more direct experiment is required to assess the effect of PHZ on Epo production in Fundc1-/- kidney cells. To further support the hypothesis, supplemental Epo administration should be performed to examine the effects on the abnormalities observed in the Fundc1-/- mice.

We do appreciate these thoughtful suggestions. To directly assess the effects of PHZ on Fundc1-/- renal cells, we FACS-sorted PDGFRb+ renal EPO producing cells (REPs) from kidneys of mito-Keima/WT and mito-Keima/Fundc1-/- mice and treated them with PHZ, as described previously (1). We observed impaired mitophagy and reduced EPO production in Fundc1-/- cells treated with a high dose of PHZ (100 mM), but not lower dose (10 mM) (Figure 4—figure supplement 1), suggesting that high dosage of PHZ might have, at least in part, a direct impact on the Fundc1-/- REPs. The new data was incorporated in the revised manuscript on page 15, lines 376-385.

We also administrated EPO to PHZ-Fundc1-/- mice (Figure 2J) and found that EPO supplement restored the RBC number, HGB and splenic erythroid differentiation to the similar level of WT controls (Figure 2J-L). These new data have been included in the revised manuscript on page 10-11, lines 238-244. We thus conclude that the defective mitophagy observed in Fundc1-deleted kidney is responsible for the decreased Epo production and associated phenotypes.

3. It is important to provide evidence for mitochondrial retention in a model for kidney disease that is accompanied by anemia. Such studies will strengthen the conclusion that FUNDC1 deficiency contributes to anemia associated with kidney disease by regulating mitophagy.

Thank you for the valuable comment. We utilized cisplatin in Fundc1-/- mice to induce renal anemia (Figure 6). We found that cisplatin-treated Fundc1-/- mice showed less RBCs and lower HGB in their peripheral blood (Figure 6B), which was accompanied by reduced serum EPO (Figure 6C). Also, we observed accumulated mitochondrial mass (Figure 6H), deficient mitophagy (Figure 6F-G), enhanced inflammation (Figure 6K) and decreased EPO production (Figure 6L) in PDGFRb+ REPs of cisplatin-treated mito-keima/Fundc1-/- mice. This new data was incorporated in the revised manuscript on page 17-18, lines 436-457.

4. Given the modest effects on blood cell production, the conclusions that FUNDC1-mediated mitophagy is not required for reticulocyte maturation need to be qualified or significantly revised.

The results from hemogram, including reduced RBCs and HGB as well as increased reticulocytes indicated that both PHZ-treated Fundc1-/- and WT mice were anemic (Figure 1—figure supplement 2A and Supplementary File 1), although little difference was found in RBCs, HGB and reticulocytes between them. We then isolated reticulocytes (CD71+Ter119+) in peripheral blood of PHZ-treated Fundc1-/- and WT mice (Figure 1—figure supplement 2B) with morphologically confirmed by Giemsa and Brilliant Cresyl blue staining (Figure 1—figure supplement 2C), as previously described (2, 3). We cultured them to maturation in vitro for 3 days, using a previously described method (4). The brilliant cresyl blue staining showed progressive maturation of reticulocytes in WT mice after 3 days of culture (Figure 1—figure supplement 2C), while there were no detectable differences in the maturation of reticulocytes between WT and Fundc1-/- mice (Figure 1—figure supplement 2C). Consistent with these findings, mitochondrial mass showed little difference between WT and Fundc1-/- mice during reticulocyte maturation (Figure 1—figure supplement 2D). Thus, our results suggest that Fundc1-mediated mitophagy might be dispensable for reticulocyte maturation. The text for these data can be found on page 8, lines 163-169.

5. PHZ leads to significant reduction of Epo transcript in FUNDC1-deleted kidney cells but only a mild reduction of serum Epo level. This discrepancy should be explained.

We traced the kinetics of Epo mRNA and protein in kidney and serum, respectively, at multiple time points after PHZ treatment (Figure 2I, Figure 2—figure supplement 1A and Figure 3A). The data suggested that the discrepancy between EPO protein and mRNA likely arises from their asynchronization. Please see the text on page 10, lines 234-236.

6. Given the modest effects observed in FUNDC1-deleted mice and the limitations of the PHZ-induced acute anemia, it is important to include studies using a model for chronic anemia, and examine the effects of modest Epo deficiency effects over time. For example, phlebotomy may help to distinguish the effects of oxidative stress versus pure anemia. Repeated phlebotomy may help to assess the role of Fundc1 in chronic anemia could have a more profound effect than what was observed after a single PHZ treatment.

These comments are insightful. As suggested, we generated the chronic anemic model using phlebotomy. As shown in Figure 2—figure supplement 1C, we observed no significant differences in peripheral RBC counts, peripheral HGB level (Figure 2—figure supplement 1D), EPO production (Figure 2—figure supplement 1E) and erythroid differentiation in the spleen and BM (Figure 2—figure supplement 1F-G) between WT and Fundc1-/- mice after phlebotomy. These results imply that Fundc1 is inclined to respond to the oxidative stress of PHZ rather than pure anemia. Please see the text on page 11, lines 244-249.

7. Given the limitations of analyzing p62 and LC3 expression to represent altered autophagy, additional experiments are required to support conclusions related to autophagy.

We agree with the reviewers that the expression of P62 and LC3 are the biochemical hallmarks of general autophagy. To overcome this limitation, we have developed mt-Keima/Fundc1-/- mice to examine mitophagy in vivo under various stress conditions (Figure 4H and 6F). We also measured the accumulation of mitochondrial membrane proteins (Figure 3E, 5D and 6D) by Western blotting. These evidences collectively reflect mitophagy in our system.

Reviewer #1:

The authors have contributed substantially to the understanding of FUNDC1 as a mitophagic receptor in multiple cell types and disease conditions including hypoxia-mediated mitophagy in mammalian cells, inhibition of inflammasome activation, ischemia-reperfusion injury to the heart. Here the authors turned their attention to the kidney and showed that lack of FUNDC1 results in the accumulation of impaired mitochondria in EPO producing cells (REPs), leading to an inflammatory response, a myofibroblastic transformation of REPs, resulting in reduced EPO production and anemia in two experimental models including PHZ and UUO. Work presented here indicates that FUNDC1 does not play a role in mitochondrial removal during erythropoiesis. Overall, the paper is well written and presents strong evidence showing the importance of FUNDC1-mediated mitophage in renal fibrosis-mediated anemia by reducing the Epo production. However, the paper can be improvde by addressing following major concern:

We deeply appreciate the reviewer’s encouragement of our work and insightful comments.

The title and manuscript's emphasis of FUNDC1-mediaetd mitopahge regulting "stress eythropoiesis" are misleading. One of major complications of renal fibrosis is anemia due to the reduced production of Epo. Thus, simply because a particular protein deficiency results in renal fibrosis and reduced EPO production does not mean that such a protein should be considered as a regulator of "stress erythropoiesis" as implied by the title and multiple places in the manuscript. The findings showed here support that FUNDC1 protects against the accumulation of impaired mitochondria, inflammation, renal fibrosis and reduced EPO production and thus the development of major complication of anemia but not stress erythropoesis. Please consider to revise the title and text to avoid misleading and confusion.

We agree with the reviewer. We have accordingly revised the title as “Receptor-mediated mitophagy regulates EPO production and protects against renal anemia”. Please also refer to earlier comment for additional explanation.

Reviewer #2:

The manuscript by Shi et al., analyzes the effect of deletion of mitophagy receptor FUNDC1 in mice on erythropoiesis and kidney production of erythropoietin (Epo). The authors have performed a considerable amount of work and generated a large set of data. Since the authors did not find any effect of FUNDC1 deletion on hematopoiesis or erythropoiesis, they subjected mice to erythroid stress by inducing hemolytic anemia using the established phenylhydrazine (PHZ) treatment. Mice survival in response to phenylhydrazine was reduced and attributed to severe anemia. They noticed a mild reduction in Epo production in response to phenylhydrazine. They then focused on kidney production of Epo. They confirm that FUNDC1-deleted kidney cells exhibit defective mitophagy, and PHZ treatment enhanced defective mitophagy in FUNDC1-deleted kidney cells. They attributed this defect to reduced Epo production.

We appreciate the positive comments on our manuscript.

There are two major problems in this manuscript. One is that in contrary to the authors assertion, it is unclear whether FUNDC1-deleted mice exhibit anemia (hemoglobin is not reported in PHZ-treated mice and other RBC indices are very mild) but possibly only a reduced compensated red blood cell production in response to phenylhydrazine treatment.

Thanks for the thoughtful comments. We have provided the hemogram of PHZ-treated Fundc1-/- mice, indicating that these mice were anemic (Supplementary File 1 and Figure 1—figure supplement 2A). To further strengthen the anemia observed in the Fundc1-/- mice upon stresses, we treated Fundc1-/- mice with cisplatin, which induced the renal anemia. We indeed detected severer anemia in Fundc1-/- mice than WT controls (please refer to Essential revision 2 above). However, to describe our results more precisely, we also revised the text as “reduced RBC number and HCT values” in Fundc1-/- mice upon PHZ-induced acute hemolytic anemia.

The other is that although the authors confirm that FUNDC1-deleted kidney cells exhibit defective mitophagy that is more pronounced in PHZ-treated animals, they do not show that reduced Epo production results from defective mitophagy. My specific comments are below:

We thank the reviewer for pointing it out. The EPO expression was reduced in the PDGFRb+ REPs of PHZ- (Figure 4M), UUO- (Figure 5I) and cisplatin-treated (Figure 6L) Fundc1-/- mice, respectively. Meanwhile, all these PDGFRb+ REPs exhibited defective mitophagy upon PHZ- (Figure 4H,I), UUO- (Figure 5D) and cisplatin-treated (Figure 6F,G). In addition, the direct effect of PHZ on the EPO production was also examined (Figure 4—figure supplement 1). Please refer to Essential revision 2 above.

– There is no evidence of mitochondrial retention in the spleen, BM and PB. The authors use CD71 as a marker but it is unclear on which cells.

We replaced CD71 with Ter119 to better distinguish erythroid cells (Figure 1—figure supplement 1B-D). We found that there were no evident changes between Fundc1-/- mice and WT control despite substantial mitochondrial retention exhibited in Bnip3l-/- mice (Figure 1—figure supplement 1B-D).

– Although the survival in response to phenylhydrazine is reduced which was achieved by using 100mg/kg (instead of the usual dose of 50mg/kg) and including 30 mice (p=0.04), there was no severe anemia. In fact Hb levels are normal and the slight reduction in red blood cell indices suggest that there was a mild reduced red blood cell production that did not reach the definition of anemia. So, in the absence of anemia the claim that Fundc1-mediated mitophagy is not required for reticulocyte maturation is not accurate.

Please refer to Essential revision 4.

– It remains unclear what animals die from.

FUNDC1 is expressed in a broad range of tissues and organs, playing an important role in cardiac tissues, liver and kidney (8-10). Currently, the cause leading to the death of the animal is yet to be determined. We speculate that the death of PHZ-Fundc1-/- mice might result from the dysfunction of multiple tissues and/or organs beyond anemia.

– The authors FACS-sort reticulocytes 5 days post phenylhydrazine. This is an interesting experiment. What conditions for the FACS sorting do they use?

Please refer to Essential revision 4.

– From Figure S2 it seems that the authors FACS-sorted immature erythroblasts but not reticulocytes, the data on mitochondria in these cells also seems to be from immature erythroid cells. The authors should clarify how they distinguish reticulocytes from immature erythroblasts.

Please refer to Essential revision 4.

– The authors state: "Together, our results suggest that FUNDC1- mediated mitophagy is

dispensable for programmed mitochondrial elimination in reticulocytes during

erythroid maturation." It is difficult to assess the validity of this statement given that it is unclear which cells were used in these experiment.

Please refer to Essential revision 4.

– The authors examine differential gene expression in R2 erythroblast subpopulation between cells from mice subjected to PHZ treatment or not. They use RNA-Seq analysis of R2 and identify among others JAK-STAT pathway to be differentially expressed in PHZ treated Fundc1-/- spleens. They conclude that JAK-STAT is hypoactivated in Fundc1-/- R2 cells. Given that the reduced RBC production seem to be compensated (see R3, R4) does not seem to be highly affected this result seems to indicate compensated JAK-STAT hypoactivation.

The R3 and R4 populations appears to be normal in the spleen of PHZ-Fundc1-/- and -WT erythroid cells, suggesting that other signaling pathways may compensate the hypoactivation of the JAK-STAT signaling pathway. However, since we have not performed the RNA-Seq analysis of R3 and R4 cells from Fundc1-/- and WT mice, we do not know yet what signal(s) may compensate the hypoactivated JAK-STAT signaling pathway. This topic should be addressed in future investigations.

– How do the authors explain that Epo reduction in Fundc1-/- cells has an effect solely on R2 spleen cells?

It has been reported that EPOR, a transmembrane glycoprotein, is expressed in erythroid progenitors (colony-forming unit-erythroid: CFU-E), proerythroblasts, and early basophilic erythroblasts (5, 6). The proerythroblasts and early basophilic erythroblasts are the major components of R2 population, as indicated by Giemsa staining of sorted R2 cells (Author response image 1) as well as prior studies (7). Given that the major components of R2 are EPO dependent, which may account for the sensitivity of R2 cells to EPO paucity. Please see the text on page 10, lines 224-229.

Author response image 1
Download asset Open asset
Giemsa staining of FACS-sorted R2 populations.

The red and green arrow indicate proerythroblasts and basophilic erythroblasts, respectively. The photographs were taken at an original magnification of 400×.

– The significant reduction of Epo transcript in kidney cells in response to PHZ is interesting but how does this correlate with a mild reduction of Epo in PB?

Please refer to Essential revision 5.

– To determine the role of defective mitophagy on Epo production, the authors selected PDGFRB+ and – cells and subjected to hypoxia. A more direct experiment is to assess the effect of PHZ on Epo production in Fundc1-/- kidney cells instead they have measured mitochondrial mass which does not address their hypothesis.

Please refer to Essential revision 2.

– How are experiments in 4M and 3A different? What is missing is the effect of PHZ on Fundc1-/- PDGFRB+ cells?

In Figure 3A the Epo level was tested with the lysates of whole kidney, while in Figure 4M the Epo level was measured in FACS-sorted PDGFRb+ REPs.

Please refer to Essential revision 2 for the direct effect of PHZ on Fundc1-/- PDGFRb+ cells.

Reviewer #3:

The authors show that loss of the mitophagy receptor Fundc1 in mice causes defects in stress erythropoiesis by impairing the production of Epo by the kidney, with defective mitophagy of Epo-producing cells. This is interesting work that has the potential to provide insights into the anemia of kidney disease. However, the overall effects seem to be fairly modest and several of the studies must be investigated further to support the claims to a reasonable level. Notably, modest impairment of Epo production could lead to significant anemia over time, but this is not shown. Perhaps the most important problem is that study does not examine a model for anemia of kidney disease.

We thank the reviewer for the insightful comments and suggestions.

1. Following PHZ administration to the Fundc1-/- mice, the reductions in blood HCT and RBC, splenic erythroblast number and serum Epo concentration relative to WT mice are very mild. So while Fundc1 may have a role in maintaining Epo production during stress, the effect seems to be relatively minor, at least as revealed in the current assay. As the authors point out, Fundc1 may have a role in chronic stress erythropoiesis. In this regard, they may want to examine a more chronic model for anemia to examine the effects of modest Epo deficiency effects over time (see point #3).

Please refer to Essential revision 6.

2. To support further their hypothesis that delayed recovery from PHZ is not due to erythroid-intrinsic problems, the investigators should show that supplemental Epo administration eliminates the abnormalities observed in the Fundc1-/- mice.

Please refer to Essential revision 2.

3. The effects of Fundc1 loss on serum Epo levels after PHZ are relatively modest (Figure 2I), while the effects on kidney Epo mRNA are more extreme (Figure 3A). How can this discrepancy be explained? Perhaps measuring the kinetics of these parameters at a few timepoints after PHZ administration would help to reconcile these differences and be informative.

Please refer to Essential revisions.

4. It might be useful to examine a different model for stress anemia. PHZ is a strong oxidant. Therefore, its effects on REPs could be due to oxidative stress rather than the erythropoietic demand of anemia. Investigating the effects of phlebotomy might be useful to distinguish the effects of oxidative stress from pure anemia. Additionally, repeated phlebotomy might allow the investigators to assess the role of Fundc1 in chronic anemia where the effects of Epo deficiency over time could have a more profound effect than what was observed after a single PHZ treatment.

Please refer to Essential revision 6.

5. The authors use one week of UUO treatment as a model for acute kidney disease. They show in this model that loss of Fundc1 causes enhanced accumulation of inflammatory markers, increased mitochondrial ROS and modest reductions in Epo compared to WT mice. However, they do not show that the model causes anemia in WT or Fundc1-/- mice. In order to substantiate a major claim of this manuscript, it is essential that the investigators examine the effects of Fund1c loss in a model for anemia of kidney disease, preferably a chronic one.

We have generated a cisplatin-induced renal anemic model to measure the role of Fundc1. Please refer to Essential revision 1 and Essential revision 3.

6. One central hypothesis is that Fundc1 deficiency promotes retention of abnormal mitochondria, inflammation and transformation to myofibroblast (abstract and line 470). However, the data do not show this. Rather, they report induction of some fibroblastic markers but constant numbers of REPs. Would it be possible to show this data?-- specifically co-staining of kidney cells for Epo mRNA or protein and PDGFb to quantify the number of REPs and their Epo induction.

It is an excellent idea. We found the percentage of REPs show no difference between WT and Fundc1-/- mice after PHZ treatment in kidney as Author response image 2, despite the capacity of EPO production was reduced (Figure 4M, 5I and 6L). As suggested, we had tried the co-staining of Epo mRNA and PDGFRb protein in FACS-sorted PDGFRb+ REP cells in PHZ-Fundc1-/- and -WT mice. However, the EPO RNA-FISH experiments did not work out. We speculate that this might result from the lengthy procedure for renal sample preparation, which led to the degradation of EPO mRNA.

Author response image 2
Download asset Open asset
The percentage of REPs after PHZ treatment in kidney.

7. In general, all of the histology slides in this manuscript are not well explained. For example, the histology in Figure S4C should be interpreted. What are the signs of renal injury? Is there evidence of injury throughout the kidney or specifically in Epo producing cells? The figure legends of all micrographs should include interpretation of the histopathology with the abnormal regions indicated on the micrograph.

The interpretation of histopathology was incorporated in the figure legends (Figure 3—figure supplement 1C) of the revised manuscript. (Page 44, lines 1208-1215).

8. Figure 3 Western blots – it is difficult to appreciate some of the differences in protein expression between WT vs mutant kidney. All western blots should include quantification of multiple experiments.

We have included the quantification of multiple experiments in all Western blots.

9. In several experiments, the authors interpret altered expression of p62 and LC3 as representing altered autophagy. This is not correct. Autophagy rates cannot be inferred by steady state measurements of these proteins. Enhanced autophagy could also deplete these proteins! Use of the Keima reporter seems more correct. Along the same lines, if Fundc1 is the mitochondrial autophagy adaptor in kidney, then what is the role of LC3 or p62 and how would their altered expression relate to mitophagy versus other autophagy pathways?

Please refer to Essential revision 7.

With respect to “if Fundc1 is the mitochondrial autophagy adaptor in kidney, then what is the role of LC3 or P62”, the expression of P62 and LC3 is the biochemical hallmark of general autophagy and selective mitophagy also needs general autophagy machinery. For example, our prior studies demonstrated that P62 (15) and LC3 (16) are involved in FUNDC1-mediated mitophagy.

With respect to “how would their altered expression relate to mitophagy versus other autophagy pathways?”, although we cannot completely rule out the possibility that P62 and LC3 might also be involved in other autophagic pathways, the collective results from mt-Keima/Fundc1-/- mice (Figure 4H and 6F) and the accumulation of mitochondrial membrane proteins (Figure 3E, 5D and 6D) suggest that the altered expression of P62 and LC3 could partially reflect deficient mitophagy with Fundc1 deletion.

1. Merle NS, Grunenwald A, Figueres ML, Chauvet S, Daugan M, Knockaert S, et al. Characterization of Renal Injury and Inflammation in an Experimental Model of Intravascular Hemolysis. Front Immunol. 2018;9:179.2. Sandoval H, Thiagarajan P, Dasgupta SK, Schumacher A, Prchal JT, Chen M, et al. Essential role for Nix in autophagic maturation of erythroid cells. Nature. 2008;454(7201):232-5.3. Gothwal M, Wehrle J, Aumann K, Zimmermann V, Grunder A, and Pahl HL. A novel role for nuclear factor-erythroid 2 in erythroid maturation by modulation of mitochondrial autophagy. Haematologica. 2016;101(9):1054-64.4. Koury MJ, Koury ST, Kopsombut P, and Bondurant MC. In vitro maturation of nascent reticulocytes to erythrocytes. Blood. 2005;105(5):2168-74.5. Wickrema A, Bondurant MC, and Krantz SB. Abundance and stability of erythropoietin receptor mRNA in mouse erythroid progenitor cells. Blood. 1991;78(9):2269-75.6. Broudy VC, Lin N, Brice M, Nakamoto B, and Papayannopoulou T. Erythropoietin receptor characteristics on primary human erythroid cells. Blood. 1991;77(12):2583-90.7. Socolovsky M, Nam H, Fleming MD, Haase VH, Brugnara C, and Lodish HF. Ineffective erythropoiesis in Stat5a(-/-)5b(-/-) mice due to decreased survival of early erythroblasts. Blood. 2001;98(12):3261-73.8. Wu S, Lu Q, Wang Q, Ding Y, Ma Z, Mao X, et al. Binding of FUN14 Domain Containing 1 With Inositol 1,4,5-Trisphosphate Receptor in Mitochondria-Associated Endoplasmic Reticulum Membranes Maintains Mitochondrial Dynamics and Function in Hearts in Vivo. Circulation. 2017;136(23):2248-66.9. Li W, Li Y, Siraj S, Jin H, Fan Y, Yang X, et al. FUN14 Domain-Containing 1-Mediated Mitophagy Suppresses Hepatocarcinogenesis by Inhibition of Inflammasome Activation in Mice. Hepatology. 2019;69(2):604-21.10. Wang J, Zhu P, Li R, Ren J, and Zhou H. Fundc1-dependent mitophagy is obligatory to ischemic preconditioning-conferred renoprotection in ischemic AKI via suppression of Drp1-mediated mitochondrial fission. Redox Biol. 2020;30:101415.11. Greenwald AC, Licht T, Kumar S, Oladipupo SS, Iyer S, Grunewald M, et al. VEGF expands erythropoiesis via hypoxia-independent induction of erythropoietin in noncanonical perivascular stromal cells. J Exp Med. 2019;216(1):215-30.12. Broeker KAE, Fuchs MAA, Schrankl J, Kurt B, Nolan KA, Wenger RH, et al. Different subpopulations of kidney interstitial cells produce erythropoietin and factors supporting tissue oxygenation in response to hypoxia in vivo. Kidney Int. 2020;98(4):918-31.13. Gerl K, Nolan KA, Karger C, Fuchs M, Wenger RH, Stolt CC, et al. Erythropoietin production by PDGFR-beta(+) cells. Pflugers Arch. 2016;468(8):1479-87.14. Pan X, Suzuki N, Hirano I, Yamazaki S, Minegishi N, and Yamamoto M. Isolation and characterization of renal erythropoietin-producing cells from genetically produced anemia mice. PLoS One. 2011;6(10):e25839.15. Li Y, Xue Y, Xu X, Wang G, Liu Y, Wu H, et al. A mitochondrial FUNDC1/HSC70 interaction organizes the proteostatic stress response at the risk of cell morbidity. EMBO J. 2019;38(3).16. Liu L, Feng D, Chen G, Chen M, Zheng Q, Song P, et al. Mitochondrial outer-membrane protein FUNDC1 mediates hypoxia-induced mitophagy in mammalian cells. Nat Cell Biol. 2012;14(2):177-85.17. Chang YT, Yang CC, Pan SY, Chou YH, Chang FC, Lai CF, et al. DNA methyltransferase inhibition restores erythropoietin production in fibrotic murine kidneys. J Clin Invest. 2016;126(2):721-31.18. Schweers RL, Zhang J, Randall MS, Loyd MR, Li W, Dorsey FC, et al. NIX is required for programmed mitochondrial clearance during reticulocyte maturation. Proceedings of the National Academy of Sciences of the United States of America. 2007;104(49):19500-5.

https://doi.org/10.7554/eLife.64480.sa2

Article and author information

Author details

  1. Guangfeng Geng

    State Key Laboratory of Medicinal Chemical Biology, Frontiers Science Center for Cell Responses, Tianjin Key Laboratory of Protein Science, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Conceptualization, Data curation, Writing - original draft
    Contributed equally with
    Jinhua Liu and Changlu Xu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-0518-406X

  2. Jinhua Liu

    State Key Laboratory of Experimental Hematology, National Clinical Research Center for Blood Diseases, Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China
    Contribution
    Data curation, Formal analysis, Validation
    Contributed equally with
    Guangfeng Geng and Changlu Xu
    Competing interests
    No competing interests declared

  3. Changlu Xu

    State Key Laboratory of Experimental Hematology, National Clinical Research Center for Blood Diseases, Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China
    Contribution
    Data curation, Software, Formal analysis, Validation
    Contributed equally with
    Guangfeng Geng and Jinhua Liu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-1581-7027

  4. Yandong Pei

    State Key Laboratory of Medicinal Chemical Biology, Frontiers Science Center for Cell Responses, Tianjin Key Laboratory of Protein Science, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Resources, Methodology
    Competing interests
    No competing interests declared

  5. Linbo Chen

    State Key Laboratory of Medicinal Chemical Biology, Frontiers Science Center for Cell Responses, Tianjin Key Laboratory of Protein Science, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Resources, Validation
    Competing interests
    No competing interests declared

  6. Chenglong Mu

    State Key Laboratory of Medicinal Chemical Biology, Frontiers Science Center for Cell Responses, Tianjin Key Laboratory of Protein Science, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Resources, Validation, Methodology
    Competing interests
    No competing interests declared

  7. Ding Wang

    State Key Laboratory of Experimental Hematology, National Clinical Research Center for Blood Diseases, Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China
    Contribution
    Software, Formal analysis
    Competing interests
    No competing interests declared

  8. Jie Gao

    State Key Laboratory of Experimental Hematology, National Clinical Research Center for Blood Diseases, Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China
    Contribution
    Resources, Supervision, Methodology
    Competing interests
    No competing interests declared

  9. Yue Li

    State Key Laboratory of Experimental Hematology, National Clinical Research Center for Blood Diseases, Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China
    Contribution
    Formal analysis
    Competing interests
    No competing interests declared

  10. Jing Liang

    State Key Laboratory of Experimental Hematology, National Clinical Research Center for Blood Diseases, Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China
    Contribution
    Resources, Supervision, Visualization
    Competing interests
    No competing interests declared

  11. Tian Zhao

    State Key Laboratory of Medicinal Chemical Biology, Frontiers Science Center for Cell Responses, Tianjin Key Laboratory of Protein Science, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Software, Validation, Methodology
    Competing interests
    No competing interests declared

  12. Chuanmei Zhang

    State Key Laboratory of Medicinal Chemical Biology, Frontiers Science Center for Cell Responses, Tianjin Key Laboratory of Protein Science, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Validation, Methodology
    Competing interests
    No competing interests declared

  13. Jiaxi Zhou

    State Key Laboratory of Experimental Hematology, National Clinical Research Center for Blood Diseases, Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China
    Contribution
    Supervision, Methodology
    Competing interests
    No competing interests declared

  14. Quan Chen

    State Key Laboratory of Medicinal Chemical Biology, Frontiers Science Center for Cell Responses, Tianjin Key Laboratory of Protein Science, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Resources, Funding acquisition, Project administration, Writing - review and editing
    For correspondence
    chenq@ioz.ac.cn
    Competing interests
    No competing interests declared

  15. Yushan Zhu

    State Key Laboratory of Medicinal Chemical Biology, Frontiers Science Center for Cell Responses, Tianjin Key Laboratory of Protein Science, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Conceptualization, Data curation, Funding acquisition, Writing - review and editing
    For correspondence
    zhuys@nankai.edu.cn
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-5648-0416

  16. Lihong Shi

    State Key Laboratory of Experimental Hematology, National Clinical Research Center for Blood Diseases, Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China
    Contribution
    Formal analysis, Funding acquisition, Writing - original draft, Writing - review and editing
    For correspondence
    shilihongxys@ihcams.ac.cn
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-8876-0802

Funding

National Key Research and Development Program of China (2019YFA0508601)

  • Quan Chen

National Natural Science Foundation of China (91849201)

  • Quan Chen

National Key Research and Development Program of China (2019YFA0508603)

  • Yushan Zhu

National Natural Science Foundation of China (32030026)

  • Yushan Zhu

National Natural Science Foundation of China (81870089)

  • Lihong Shi

Chinese Academy of Medical Sciences (2016-I2M-3-002)

  • Lihong Shi

National Key Research and Development Program of China (2017YFA0103102)

  • Quan Chen

National Key Research and Development Program of China (2016YFA0102300)

  • Quan Chen

National Natural Science Foundation of China (31790404)

  • Quan Chen

National Natural Science Foundation of China (81890990)

  • Lihong Shi

National Natural Science Foundation of China (81700105)

  • Lihong Shi

Chinese Academy of Medical Sciences (2019-I2M-1-006)

  • Lihong Shi

Chinese Academy of Medical Sciences (2016-I2M-1-018)

  • Lihong Shi

Chinese Academy of Medical Sciences (2017-I2M-1-015)

  • Lihong Shi

National Natural Science Foundation of China (91754114)

  • Yushan Zhu

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Ethics

Animal experimentation: Mice were maintained in the animal core facility of College of Life Sciences, Nankai University, Tianjin, China. All experiments involving animals were reviewed and approved by the Animal Care and Use Committee of Nankai University and were performed in accordance with the university guidelines (NO. 2021-SYDWLL-000410).

Senior Editor

  1. Matt Kaeberlein, University of Washington, United States

Reviewing Editor

  1. Jian Xu, University of Texas Southwestern Medical Center, United States

Version history

  1. Received: October 30, 2020
  2. Accepted: May 2, 2021
  3. Accepted Manuscript published: May 4, 2021 (version 1)
  4. Version of Record published: May 14, 2021 (version 2)

Copyright

© 2021, Geng et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,223
    Page views
  • 246
    Downloads
  • 6
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Guangfeng Geng
  2. Jinhua Liu
  3. Changlu Xu
  4. Yandong Pei
  5. Linbo Chen
  6. Chenglong Mu
  7. Ding Wang
  8. Jie Gao
  9. Yue Li
  10. Jing Liang
  11. Tian Zhao
  12. Chuanmei Zhang
  13. Jiaxi Zhou
  14. Quan Chen
  15. Yushan Zhu
  16. Lihong Shi
(2021)
Receptor-mediated mitophagy regulates EPO production and protects against renal anemia
eLife 10:e64480.
https://doi.org/10.7554/eLife.64480

Categories and tags

Research organism

Further reading

    1. Biochemistry and Chemical Biology
    2. Cell Biology

    Fluorescein-based sensors to purify human a-cells for functional and transcriptomic analyses

    Sevim Kahraman, Kimitaka Shibue ... Rohit N Kulkarni
    Tools and Resources

    Pancreatic a-cells secrete glucagon, an insulin counter-regulatory peptide hormone critical for the maintenance of glucose homeostasis. Investigation of the function of human a-cells remains a challenge due to the lack of cost-effective purification methods to isolate high-quality a-cells from islets. Here, we use the reaction-based probe diacetylated Zinpyr1 (DA-ZP1) to introduce a novel and simple method for enriching live a-cells from dissociated human islet cells with ~ 95% purity. The a-cells, confirmed by sorting and immunostaining for glucagon, were cultured up to 10 days to form a-pseudoislets. The a-pseudoislets could be maintained in culture without significant loss of viability, and responded to glucose challenge by secreting appropriate levels of glucagon. RNA-sequencing analyses (RNA-seq) revealed that expression levels of key a-cell identity genes were sustained in culture while some of the genes such as DLK1, GSN, SMIM24 were altered in a-pseudoislets in a time-dependent manner. In conclusion, we report a method to sort human primary a-cells with high purity that can be used for downstream analyses such as functional and transcriptional studies.

    1. Cell Biology
    2. Physics of Living Systems

    Force propagation between epithelial cells depends on active coupling and mechano-structural polarization

    Artur Ruppel, Dennis Wörthmüller ... Martial Balland
    Research Article Updated

    Cell-generated forces play a major role in coordinating the large-scale behavior of cell assemblies, in particular during development, wound healing, and cancer. Mechanical signals propagate faster than biochemical signals, but can have similar effects, especially in epithelial tissues with strong cell–cell adhesion. However, a quantitative description of the transmission chain from force generation in a sender cell, force propagation across cell–cell boundaries, and the concomitant response of receiver cells is missing. For a quantitative analysis of this important situation, here we propose a minimal model system of two epithelial cells on an H-pattern (‘cell doublet’). After optogenetically activating RhoA, a major regulator of cell contractility, in the sender cell, we measure the mechanical response of the receiver cell by traction force and monolayer stress microscopies. In general, we find that the receiver cells show an active response so that the cell doublet forms a coherent unit. However, force propagation and response of the receiver cell also strongly depend on the mechano-structural polarization in the cell assembly, which is controlled by cell–matrix adhesion to the adhesive micropattern. We find that the response of the receiver cell is stronger when the mechano-structural polarization axis is oriented perpendicular to the direction of force propagation, reminiscent of the Poisson effect in passive materials. We finally show that the same effects are at work in small tissues. Our work demonstrates that cellular organization and active mechanical response of a tissue are key to maintain signal strength and lead to the emergence of elasticity, which means that signals are not dissipated like in a viscous system, but can propagate over large distances.

    1. Biochemistry and Chemical Biology
    2. Cell Biology

    Inositol pyrophosphate dynamics reveals control of the yeast phosphate starvation program through 1,5-IP8 and the SPX domain of Pho81

    Valentin Chabert, Geun-Don Kim ... Andreas Mayer
    Research Article

    Eukaryotic cells control inorganic phosphate to balance its role as essential macronutrient with its negative bioenergetic impact on reactions liberating phosphate. Phosphate homeostasis depends on the conserved INPHORS signaling pathway that utilizes inositol pyrophosphates and SPX receptor domains. Since cells synthesize various inositol pyrophosphates and SPX domains bind them promiscuously, it is unclear whether a specific inositol pyrophosphate regulates SPX domains in vivo, or whether multiple inositol pyrophosphates act as a pool. In contrast to previous models, which postulated that phosphate starvation is signaled by increased production of the inositol pyrophosphate 1-IP7, we now show that the levels of all detectable inositol pyrophosphates of yeast, 1-IP7, 5-IP7, and 1,5-IP8, strongly decline upon phosphate starvation. Among these, specifically the decline of 1,5-IP8 triggers the transcriptional phosphate starvation response, the PHO pathway. 1,5-IP8 inactivates the cyclin-dependent kinase inhibitor Pho81 through its SPX domain. This stimulates the cyclin-dependent kinase Pho85-Pho80 to phosphorylate the transcription factor Pho4 and repress the PHO pathway. Combining our results with observations from other systems, we propose a unified model where 1,5-IP8 signals cytosolic phosphate abundance to SPX proteins in fungi, plants, and mammals. Its absence triggers starvation responses.