Loss of legumain induces premature senescence and mediates aging‐related renal fibrosis

Abstract Aging is an independent risk factor for acute kidney injury and subsequent chronic kidney diseases, while the underlying mechanism is still elusive. Here, we found that renal tubules highly express a conserved lysosomal endopeptidase, legumain, which is significantly downregulated with the growing of age. Tubule‐specific legumain‐knockout mice exhibit spontaneous renal interstitial fibrosis from the 3rd month. In the tubule‐specific legumain‐knockout mice and the cultured legumain‐knockdown HK‐2 cells, legumain deficiency induces the activation of tubular senescence and thus increases the secretion of profibrotic senescence‐associated cytokines, which in turn accelerates the activation of fibroblasts. Blockage of senescence mitigates the fibrotic lesion caused by legumain deficiency. Mechanistically, we found that silencing down of legumain leads to the elevated lysosome pH value, enlargement of lysosome size, and increase of lysosomal voltage dependent membrane channel proteins. Either legumain downregulation or aging alone induces the activation of nuclear transcription factors EB (TFEB) while it fails to further upregulate in the elderly legumain‐knockdown tubules, accompanied with impaired mitophagy and increased mitochondrial ROS (mtROS) accumulation. Therapeutically, supplementation of exosomal legumain ameliorated fibronectin and collagen I production in an in vitro coculture system of tubular cells and fibroblasts. Altogether, our data demonstrate that loss of legumain in combined with aging dysregulates lysosomal homeostasis, although either aging or legumain deficiency alone induces lysosome adaptation via stimulating lysosomal biogenesis. Consequently, impaired mitophagy leads to mtROS accumulation and therefore activates tubular senescence and boosts the interstitial fibrosis.

knockout mice exhibit spontaneous renal interstitial fibrosis from the 3rd month. In the tubule-specific legumain-knockout mice and the cultured legumain-knockdown HK-2 cells, legumain deficiency induces the activation of tubular senescence and thus increases the secretion of profibrotic senescence-associated cytokines, which in turn accelerates the activation of fibroblasts. Blockage of senescence mitigates the fibrotic lesion caused by legumain deficiency. Mechanistically, we found that silencing down of legumain leads to the elevated lysosome pH value, enlargement of lysosome size, and increase of lysosomal voltage dependent membrane channel proteins. Either legumain downregulation or aging alone induces the activation of nuclear transcription factors EB (TFEB) while it fails to further upregulate in the elderly legumain-knockdown tubules, accompanied with impaired mitophagy and increased mitochondrial ROS (mtROS) accumulation. Therapeutically, supplementation of exosomal legumain ameliorated fibronectin and collagen I production in an in vitro coculture system of tubular cells and fibroblasts. Altogether, our data demonstrate that loss of legumain in combined with aging dysregulates lysosomal homeostasis, although either aging or legumain deficiency alone induces lysosome adaptation via stimulating lysosomal biogenesis. Consequently, impaired mitophagy leads to mtROS accumulation and therefore activates tubular senescence and boosts the interstitial fibrosis.

K E Y W O R D S
aging-related renal fibrosis, autophagy, legumain (asparagine endopetidase), premature senescence

| INTRODUC TI ON
Aging is an independent risk factor for acute kidney injury and its consequent chronic kidney disease (CKD). During aging or CKD progress regardless of cause, renal interstitial fibrosis is not only major pathological manifestation but also pathogenic factor (O'Sullivan et al., 2017). Many analogies can be drawn between aging and CKD, which implies that understanding aging-related fibrosis is a constructive approach to decipher effective treatments for CKD (Ruiz-Ortega et al., 2020).
As the major cell population in the kidney, renal tubules play a central role in the fibrotic response (Bonventre, 2014). Cell cycle arrest at G2/M in proximal tubule cells after kidney injury results in the abnormal amplification of profibrogenic factors (Yang et al., 2010).
Cellular senescence is a permanent cell cycle arrest induced by diverse stressors including shortening of telomeres, oncogenes, DNA damage, and mechanical stress (Childs et al., 2015). Compelling evidences support that aspects of the senescence program are active during either renal aging or diverse disease conditions (Sturmlechner et al., 2017). Chronic senescence is thought to be deleterious and usually related to phenotypes of aging (Baker et al., 2011;Wolstein et al., 2010). Most recently, Mylonas et al verified the determinant role of senescent renal epithelia in fibrosis and regenerative capacity after renal injury. Targeting senescent cells, therefore, represent a potential treatment to protect aging and vulnerable kidneys (Mylonas et al., 2021). It is currently accepted that distinct phenotypic traits of senescent cells and the genomic profile that encodes the so-called "senescence-associated secreted phenotype (SASP)" act as determinants of senescent outcomes (Tchkonia et al., 2013).
Autophagy is a lysosome-dependent, self-clearance process responsible for degradation of cytoplasmic components (Klionsky, 2007). Renal tubules reabsorb most filtered solutes in a highly energy-consuming process and proliferation is rare in renal tubules under physiological conditions compared with other epithelial cells with an active transport function (Tojo, 2013;Yamamoto et al., 2017). Considering the combination of long cellular life and high protein load, it is logical to reason that well-being of tubular cells highly depends on adequate capacity of lysosomal adaptation. It has been reported that renal tubules of elderly mice have a higher basal autophagic flux compared with young mice, and elderly kidneys are more reliant on autophagy for removal of damaged proteins and organelles (Yamamoto et al., 2016). Dysfunctional lysosomes hinder autophagy, which results in accumulation of dysfunctional organelles, protein aggregates, and toxic reactive oxygen species (ROS) (Festa et al., 2018;Isaka et al., 2011). To date, the regulatory mechanism of the autophagy-lysosome response in elderly tubules remains uncharacterized.
As a highly conserved lysosomal endopeptidase, legumain belongs to a family of cysteine proteases and shows strict specificity for hydrolysis of asparaginyl bonds (Chen et al., 1997). The specific cleavage functions of legumain have been implicated in various contexts including antigen processing in dendritic cells and the pathogenesis of neurofibrillary diseases via its site-specific cleavage of toll-like receptors and tau (Maschalidi et al., 2012;Sepulveda et al., 2009;Zhang et al., 2014Zhang et al., , 2017. Expression of legumain in kidney proximal tubules is particularly abundant and mice lacking legumain develop automatous renal interstitial fibrosis (Miller et al., 2011). Our previous data showed that deletion of legumain aggravates renal fibrosis in the mouse model of unilateral ureteral obstruction (Wang et al., 2018). Although it is undefined whether this effect is related to the pro-aging feature triggered by legumain deficiency, these data indicate an intrinsic correlation between this unusual lysosomal protease and the pathogenesis of kidney fibrotic lesions.
Here, we demonstrate that loss of legumain occurs with aging in renal proximal tubules. Depletion of legumain accelerates premature senescence and promotes aging-related renal fibrosis. In the elderly kidney, loss of legumain disequilibrates the autophagy-lysosome balance, which impairs mitophagy and leads to accumulation of mitochondrial ROS (mtROS). In terms of application, supplementation of exosomal legumain demonstrates potential on alleviating premature senescence and aging-related renal fibrosis.
To obtain CAG-RFP-EGFP-LC3 labeled Lgmn WT and Lgmn KO mice, conventional legumain-knockout (Lgmn KO ) mice were established using homologous recombination methods and generated by Cyagen Biosciences Inc. Lgmn KO mice were mate with CAG-RFP-EGFP-LC3 transgenic mice (The Jackson Laboratory, #027139). Mice aged 3 and 18 months were used for further studies. All mice used in this study were on a pure C57/BL6J background and were housed with controlled temperature (22℃), humidity, and lighting. Mice were used under the University of Nankai's Institutional Animal Use and Care Committee's approval.

| Human kidney specimens
Kidney biopsies were obtained from the Second Hospital of Tianjin Medical University (Tianjin, China). Samples analyzed were normal para-carcinoma tissue of renal carcinoma patients aged from 35 to 86 years old (n = 16). Informed consents were obtained from all participants of this study. Briefly, fresh tissues were separated for two parts. One part of the samples was fixed in 4% paraformaldehyde for histological studies. The other part was directly store at −80℃ for RNA and protein extraction use. The information of patients is listed in Table S1. All studies involving human kidney sections were performed in accordance with the Chinese Ministry of Health national guidelines for biomedical research and approved by the Human Research Ethics Committee of Nankai University.

| Isolation of murine renal tubules and cell culture
Mouse primary tubular epithelial cells (PTECs) were isolated and cultivated following a modified previously established protocol (Luo et al., 2018). Briefly, kidneys were collected and the medullary regions were removed, then minced into 4-6 slices. Kidney slices were transferred in reaction tube containing 1 mg/ml collagenase type II in incubation solution (Sigma-Aldrich, St. Louis, MO) for 1 h at 37°C.
After neutralization with FBS, the tissues were filtered through 40and 70μm strainers sequentially and then centrifuged to collect the tubules.
Human proximal cell line HK-2 was cultured in DMEM/ F12 supplemented with 10% fetal bovine serum (FBS; Biological Industries, Israel) and 1% penicillin and streptomycin (P/S; Thermo Fisher Scientific, Waltham, MA). Human foreskin fibroblast cell line BJ was cultured in MEM supplemented with 10% FBS, 1% P/S, and 1% NEAA (Thermo Fisher Scientific). Mouse fibroblast cell line NIH-3T3 was cultured in high glucose DMEM supplemented with 10% FBS and 1% P/S. The cell lines were obtained from ATCC (Manassas, VA). Mouse primary tubular epithelial cells (PTECs) were isolated and cultivated as described previously. Briefly, kidneys collected were minced into small pieces, and then digested in 1 mg/ml collagenase (Sigma-Aldrich, St. Louis, MO) for 1 h at 37°C.

| Sirius red staining
Kidney tissues were fixed in 4% paraformaldehyde, dehydrated through a graded series of ethanol solutions, embedded in paraffin, and sectioned at 5 μm thicknesses. Sirius red staining was performed using a Picrosirius Red Stain Kit (Abcam), following the manufacturers' protocol. Kidney sections were deparaffinized, washed, and then sequentially stained with picric acid and Sirius red solutions.
Sections were washed, dehydrated, and then mounted with neutral resin. Quantification of the Sirius red-positive area was performed in random cortical images (×400, 10 fields/kidney) by counting the percentage of positively stained areas in each microscopic field.

| SA-β-Gal activity assay
SAβ-gal staining was performed in accordance with the manufacturer's instructions (Beyotime, Shanghai, China). Briefly, frozen kidney tissues (5 μm thick) or cells were fixed in 4% paraformaldehyde for 15 min and then washed in PBS for 3 times. Then, the fixed sections or cells were incubated with reaction buffer overnight at 37°C without CO 2 . The sections were counterstained with eosin for 5 min and then dehydrated. Quantification of the SAβ-gal-positive area was performed in random cortical images (×400, 10 fields) by counting the percentage of positively stained areas in each microscopic field.

| Immunohistochemistry and Immunofluorescence Staining
For IHC analysis, 5μm-thick kidney sections were used. After deparaffinization and rehydration, antigen retrieval was performed using citrate buffer (0.1 M citric acid and 0.1 M sodium citrate, pH 6.4). The sections were blocked with 5% goat serum and then incubated with primary antibodies overnight at 4°C. Sections were incubated with biotinylated anti-rabbit or mouse secondary antibodies (Vector Laboratories, Burlingame, CA) for 2 h at room temperature and then incubated with streptavidin-peroxidase complex for 1 h at room temperature. Antibody signals were detected using a DAB kit (Vector Laboratories). Sections were counterstained with hematoxylin and then dehydrated. Quantification of the antigenpositive area was performed in random cortical images (×400, 10 fields/kidney) by counting the percentage of positively stained areas in each microscopic field.
For IF staining, shCtrl/shLGMN HK-2 cells, and Lgmn WT and Lgmn KO PTECs cultured on glass-bottom dishes (Thermo Fisher) were washed with PBS and then fixed in 4% paraformaldehyde for 15 min. After washing, the fixed cells were blocked with 5% goat serum and then incubated with primary antibodies overnight at 4°C. Then, cells were incubated with Alexa Fluor TM 594 anti-rabbit secondary antibody (Invitrogen, Thermo Fisher Scientific) for 1 h at room temperature, washed, dried, and then mounted with mounting medium with DAPI (Vector Laboratories). The cells were imaged under a confocal microscope (model FV1000, Olympus).

| Hydroxyproline colorimetric assay
Hydroxyproline assay was performed according to the manufacturer's instructions (Biovision, Milpitas, CA). Kidney samples were prepared for an equal weight and then hydrolyzed in 6N HCl at 120°C for 3 h. Samples were reacted with the Chloramine T at room temperature for 5 min, then the products were incubated with DMAB reagent for 90 min, at 60°C. Absorbance at 560 nm was measured using Clariostar microplate reader (BMG Labtech, Offenburg, Germany).

| Real-time polymerase chain reaction
Kidney tissues and cell pellets were harvested and total RNA was extracted using TRIzol Reagent (Life Technologies, Grand Island, NY) in accordance with the manufacturer's instructions.  Table S2.

| Proliferation assay
The cell growth rate was measured by staining of crystal violet (Beyotime, Shanghai, China). Briefly, cells were seeded at 2 × 10 4 cells per well in a 6-well plate. After 2 days, the cells were fixed in methanol and stained with 0.5% crystal violet in 25% methanol for 10 min. The cell plates were dried and then washed in 1% SDS. Cell density was quantified by measuring absorbance of the wash solution at 570 nm using a microplate reader.

| Migration assay
A fibroblast migration assay was performed in a Corning BioCoat™ matrigel invasion chamber containing an 8μm-pore size PET membrane coated with a thin layer of extracellular matrix proteins. After coculture with tubular epithelial cells for 48 h, the same number of fibroblasts was seeded in the upper well and cultured for 15 h at 37°C with 5% CO 2 . Non-invading cells were removed from the upper surface of the chamber and the membrane was stained with crystal violet. Images of the membranes were obtained and cell numbers were measured.

| Contractility assay
A fibroblast contraction assay was performed as described previously. A cell suspension was prepared with serum-free medium mixed with 3 mg/ml type I collagen from rat tail tendon (BD Biosciences, San Jose, CA) at a ratio of 2:1. Then, the cell suspension at a density of 2 × 10 5 cells/ml was seeded onto a 24-well plate (300 μl/well).
Gels were allowed to polymerize at 37°C for 1 h before adding 1 ml of medium and detaching the edge of gels from the walls of the well.
After 48 h, images of the gels were obtained and the gel area was measured using ImageJ software.

| Measurement of mitochondrial ROS
Mitochondrial ROS production was detected by MitoSOX (Thermo Fisher Scientific). Briefly, Lgmn WT and Lgmn KO PTECs were seeded at 1500 cells per well in a 96-well plate. Cells were treated with 250 nM mitoquinone mesylate (Selleckchem, Houston, TX) for 24 h or exosomes for 48 h at 37°C with 5% CO 2 and then incubated with 5 μM MitoSOX for 15 min at 37°C with 5% CO 2 . The plates were washed with PBS three times and the fluorescence released by MitoSOx was measured at 510 nm (excitation) and 580 nm (emission) using a Clariostar microplate reader.

| Measurement of lysosomal pH
Lysosomal pH was measured using Lysosensor Yellow/Blue DND-100 (Invitrogen). Cells were incubated in medium containing 1 μM Lysosensor for 1 min at 37°C with 5% CO 2 . Fluorescence images were obtained at a wavelength range of 510-641 nm (yellow) and 404-456 nm (blue) under a confocal microscope (Olympus). The yellow and blue fluorescence intensities were measured using ImageJ.
The blue/yellow ratio was calculated by a division process in each section. The average blue/yellow ratio of a given sample was calculated in five sections. The intensity was measured in at least three independent experiments.

| Electron microscopy
Fresh mouse kidney tissues were cut into approximately 1 × 1 × 3 mm 3 cube-like pieces and fixed with 4% precooled glutaraldehyde for at least 2 h and 1% osmium tetroxide for 1-2 h at 4°C and then subjected to standard electron microscopic techniques as described previously. Ultrathin sections (60 nm) were observed under an electron microscopy (model HT7700, Hitachi).

| Enrichment of nuclear fractions
Nuclear and cytosolic fractions from shCtrl/shLGMN HK-2 cells were obtained using the NE-PER extraction reagents (Thermo Fisher Scientific). Briefly, cells were homogenized in cytosolic extraction reagent (CER I) using a Dounce homogenizer. Homogenates were then vortexed and let to stand on ice for 10 min. Following addition of CER II solution, samples were vortexed and centrifuged for 10 min at 16,000 g. The supernatant cytosolic fractions were then collected. The pellets, containing nuclei and cellular debris, were washed in cold PBS and suspended in nuclear extraction buffer (NER). Nuclear fractions were then sonicated three times for ~3 to 5 s at 30% power, and incubated on ice for 40 min. After centrifuged for 10 min at 16,000 g, the resulting supernatant nuclear fractions were collected and stored at −80°C for further analysis.

| Assessment of autophagic flux
To assess the autophagic flux in young and elderly Lgmn KO and wild-type mice. Kidneys from 3-and 18 months old CAG-RFP-EGFP-LC3 labeled Lgmn WT /Lgmn KO mice were collected. Kidney tissues were prepared for frozen sections (5 μm thick). Slides were washed with PBS, then mounted using the mounting medium with DAPI. The fluorescence images were collected using confocal microscope (FV1000, Olympus). GFP and RFP-MAP1LC3B dots per proximal tubule were counted in at least 10 fields (×400). The intensity of the positive staining area was measured using ImageJ software.

| Extraction of soluble and insoluble protein fractions
Extraction of soluble and insoluble proteins was performed as previously described (Festa et al., 2018). Briefly, kidney tissues were lysed in buffer (50 mM pH 7.5 Tris-HCl, 150 mM NaCl, 0.1% SDS, 1% Triton X-100, and 1% sodium deoxycholate) supplemented with protease and phosphatase inhibitors, and centrifuged at 13,300 g for 20 min at 4°C, to collect the soluble fraction (supernatant). The pellet was suspended in a buffer containing 4% SDS and 20 mM pH 7.5 HEPES, protease and phosphatase inhibitors, then centrifuged at 15,000 g for 10 min at room temperature, to collect the insoluble fraction (supernatant). Samples were then prepared for Western blotting analysis.

| Exosome isolation and cell treatment assay
Exosomes were purified from conditioned media of control and PTECs transduced with a lentiviral vector that overexpressed legumain (Lgmn OE) or blank control (MCS) separately. Cells were cultured in DMEM/F12 supplemented with 10% exosome-depleted FBS and 1% P/S. At 90% confluence, the medium was changed to DMEM/F12 and then collected after 48 h. Exosomes were purified by sequential centrifugation at 300 g for 10 min, 2,000 g for 10 min, and 10,000 g for 30 min to remove cell debris. The medium was then ultracentrifuged at 100,000 g for 2 h (model X90, Beckman Coulter).
Exosome pellets were resuspended in ice-cold PBS and used immediately or stored at −80°C. All steps were performed at 4°C.
For the cell treatment assay, elderly (18-month-old) PTECs were seeded in a 6-well plate. Cells were cultured in DMEM/F12 complete medium containing legumain overexpressing (Lgmn OE) or control (MCS) exosomes (200 μg) for 48 h at 37°C with 5% CO 2, and then, the supernatant was collected for further analysis.

| Statistical analysis
All results are presented as the mean ± SEM and analyzed using GraphPad Prism, Version 7.0 (GraphPad Software Inc., La Jolla, CA).
Comparisons of two groups were performed using the unpaired Student's t-test. For multiple group comparisons, ANOVA followed by Bonferroni's post hoc correction was used. The correlation between legumain expression in kidney and the level of renal interstitial fibrosis was analyzed by Pearson correlation analysis.
p < 0.05 was considered statistically significant.

| Aging-related loss of legumain correlates positively with renal fibrosis
To evaluate the change of legumain in renal tubules during aging, we first measured the expression of legumain in 3-, 10-, and 18-month-old mice. As shown in Figure 1a, renal tubules were the major sites of legumain expression, and tubular expression of legumain was decreased with aging. Analysis of semi-quantitative PCR and Western blot confirmed that expression of legumain was decreased in the elderly kidney (Figure 1b,c), which implied its role in the aging-related tubular lesion. We next compared the extent of renal interstitial fibrosis in the kidneys of wild-type (WT) and proxi- Center for Biotechnology Information (NCBI). It is revealed the distinct tissue-specific expressional pattern for cathepsin D, cathepsin B, and LGMN ( Figure S1B). Together, these data suggest the unique role of LGMN in the aging-related fibrotic lesions.
In human samples, we performed immunohistochemical staining and protein analysis of legumain in para-cancerous normal kidney sections obtained from patients of different ages with renal carcinoma. As shown in Figure S2A,B, expression of legumain was significantly downregulated in elderly human kidneys. Sirius red staining demonstrated more serious renal interstitial fibrosis in elderly kidneys ( Figure S2C). Moreover, negative correlation was exhibited between tubular expression of legumain and the extent of renal interstitial fibrosis in human kidney samples ( Figure S2D).

| Deficiency of legumain induces premature senescence and drives renal fibrosis
Accumulation of senescent cells has been observed in multiple elderly tissues and cellular senescence is thought to be a major driver of age-associated diseases. We then compared the senescent indexes, which included SAβ-gal activity and major senescencerelated signaling molecules p21 Cip1 and p16 Ink4a in kidneys of WT and Lgmn ΔTub mice. Our results showed that significantly increased activity of SAβ-gal was observed in 1-, 3-, 6-, 10-, and 18-month kidneys of Lgmn ΔTub mice compared with that of the control (Figure 2a).

| Legumain participates in the maintenance of lysosomal homeostasis
Legumain acts majorly in lysosomes and belongs to a unique enzyme family that cleaves after asparagine with high specificity. Therefore, we investigated the effect of legumain deficiency on the status of lysosomes. Lysosomal proton-transporting V-type ATPase (V-ATPase) is responsible for acidifying lysosomes, which is regulated through diverse subunits of integral membrane channel complex.
As shown in Figure 4a, Lysosomal protein LAMP1 and v-ATPase, ATP6V0D1, and ATP6V1G1 were increased in legumain-knockdown HK-2 cells compared with the control.

| Legumain deficiency in elderly tubular cells stagnates autophagic flux
Given the influence of legumain deficiency on the status of lysosomes, we then explored whether change of lysosomal status plays a role in the profibrotic effect of legumain deficiency. We measured the expression level of TFEB in the young and elderly kidneys of WT and Lgmn ΔTub mice separately. Western blotting and immunohistochemistry assay showed that TFEB expression was increased in Lgmn ΔTub mice at young age. In WT mice, TFEB was increased in elderly mice compared with the young mice.

Unexpectedly, activation of TFEB was significantly downregulated in
elderly Lgmn ΔTub mice compared with all other groups (Figure 5a,b).
A similar change in the expression of TFEB was found in renal tubular cells isolated from the mouse models. Indexes of lysosomal status, Lamp1, Atp6v0d1, Atp6v1g1, CtsB, and CtsD were increased in legumain-knockout and elderly wild-type PTECs, while there was no further increase in elderly legumain-knockout tubular cells (Figure 5c,d).
In addition to degradation on substances, lysosomes participate in multiple aspects of cell and tissue physiologies including autophagy, a major intracellular degradation system related closely to aging and age-related lesions. We therefore evaluated the effect of legumain-knockout on the initiation and processing of autophagy.
Our data showed that the ratio of LC3-II/LC3-I was increased in young legumain-knockout or elderly WT tubular cells, which indicated activation of autophagy.

| Exosomal legumain alleviates premature senescence and ameliorates fibrosis
Next, we made the preliminary attempt at the treatment of legumain supplementation against aging-related renal fibrosis. Exosomes derived from legumain-overexpressing primary mouse renal tubular cells were used as the resource of legumain supplementation.
First, we constructed legumain-or MCS-overexpressing tubular cells, respectively. Legumain could be detected in the exosomes collected from the legumain-overexpressing tubular cells (Figure 7a).
Application of legumain-overexpressing exosomes to young or el-

| DISCUSS ION
Progressive CKD and aging kidneys share many similarities in both manifestation and underlying molecular mediators, which suggests that CKD is a clinical model of premature kidney aging (Zhou et al., 2008). Tubular interstitial fibrosis, the common histological feature of end-stage renal diseases, appears to be prominent among the age-related kidney structural changes (Docherty et al., 2019). substrates (Sardiello et al., 2009). Moreover, TFEB modulates proteins involved in degradation of known autophagy substrates and overexpression of TFEB induces autophagy, which suggest its critical role in regulating autophagy (Settembre et al., 2011). TFEB is activated during the lysosomal response, which is essential to prevent oxalate nephropathy (Nakamura et al., 2020). Our data demonstrated that legumain deficiency or aging activated TFEB in tubular cells, whereas downregulated instead of further increased activation of TFEB was observed in elderly legumain-knockout cells, which was consistent with the tendency of lysosomal biogenesis and autophagy. A putative explanation was the failure of the lysosomal adaptive response due to overload of lysosome stress. Aging is the chronic and persistent stimuli of lysosomal stress in terms of a heavy protein load and highly autophagy-dependent tubular cells. Activation of TFEB during aging promotes lysosomal functions to maintain homeostasis. However, it also weakens the capacity of the lysosomal response when sudden and additional lysosome stress occurs.
Mitochondria are widely involved in cellular senescence as the major generator of ROS that play major roles in senescence by inducing genomic damage, accelerating telomere shortening, and acting as drivers of signaling networks important for maintenance of the senescent phenotype (Chan, 2020). Moreover, mitochondria are required for pro-oxidant and proinflammatory effects of cellular senescence, which suggest their role as drivers of age-related diseases (Sun et al., 2016). Maintenance of mitochondrial homeostasis depends on adequate mitophagy, a selective form of autophagy, which selectively removes redundant or damaged mitochondria (Ashrafi & Schwarz, 2013). Timely elimination of abnormal mitochondria in renal tubular cells represents an important quality control mechanism for cell homeostasis and survival during kidney injury and repair (Yamamoto et al., 2016). Although impaired mitophagy and consequent ROS accumulation have been associated with age-dependent mitochondrial degeneration, treatment of ROS elimination remains controversial because of the dual role for ROS in cell physiology (Bratic & Larsson, 2013). Specifically, high concentrations of ROS F I G U R E 6 Deficiency of legumain promotes mtROS accumulation and profibrotic effect of senescence in the elderly tubule. Kidney samples were collected from wild-type and Lgmn ΔTub mice of young (3-month-old) and elderly (18-month-old) (n = 3). (a) Representative images and quantitative analysis of mitochondria morphology detected by transmission electron microscopy. Scale bar, 1 μm. PTECs were isolated from wild-type and legumain-knockout mice aged 3 or 18 months. (b) Quantitative analysis of mitochondrial ROS detected by the MitoSOX probe. (c) Soluble and insoluble protein fractions of the kidney were extracted from young and elderly wild-type/Lgmn ΔTub mice. Western blot analysis of ubiquitin in the fractions. (d) Proteins were extracted from the kidney homogenates of young and elderly wild-type/Lgmn ΔTub mice (n = 6). Western blot and quantitative analysis of SQSTM1 and prohibitin in the kidneys. (e) Western blot and quantitative analysis of Parkin and Pink1 in the kidneys. (f) Mitochondria was isolated from freshly prepared renal tubules collected from wild-type and Lgmn ΔTub mice at 3 and 18 months of age. Western blotting of LC3-I/II, SQSTM1, Parkin, and Pink1. Tomm20 was used as internal control for the loading volume of mitochondria. (g) Representative images and quantitative analysis of SAβ-gal activity in PTECs after treatment with MitoQ. Scale bar, 50 μm. (h) Real-time PCR analysis of Ctgf, Pai-1, and Tgf-β1 in PTECs after treatment with MitoQ. Data are presented as the mean ± SEM. ***p < 0.001, **p < 0.01, *p < 0.05 are pathological, whereas moderate amounts of ROS are essential to maintain several biological processes including gene expression.
In this study, our data demonstrated that significantly increased mtROS had occurred in elderly legumain-knockout tubular cells with stagnant autophagic flux. A scavenger of mtROS effectively alleviated premature senescence and thus ameliorated fibrosis, which highlighted that mitochondrial ROS may be a decisive target in aging-related diseases.
Taken together, our data demonstrate that lysosomal proteinase legumain actively participates in the maintenance of lysosome homeostasis during aging of renal tubules. Loss of legumain in elderly tubular cells stagnates autophagic flux, impairs mitophagy, and activates senescence via accumulation of mtROS. Accelerated tubular senescence consequently promotes aging-related renal interstitial fibrosis. Therefore, this study reveals the potential role of legumain in aging-related renal fibrosis and sheds light on the treatment of CKD.

F I G U R E 7
Supplementation of legumain alleviates senescence and inhibits the profibrotic effect of elderly PTECs. Legumainoverexpressing EVs were prepared from Lgmn-overexpressing PTECs by ultracentrifugation. (a) Western blot analysis of legumain in legumain-overexpressing (Lgmn OE) and control (MCS) EVs. Expression of Alix and CD63 was used as internal control. PTECs were isolated from young (3-month-old) and elderly (18-month-old) wild-type mice and then treated with legumain-overexpressing or control EVs for 48 h. (b) Representative images and quantitative analysis of SAβ-gal activity in cultured cells. Scale bar, 50 μm. (c) Quantitative analysis of mitochondrial ROS detected by the MitoSOX probe. (d) Elderly PTECs were treated with Lgmn OE or MCS EVs for 48 h and then conditioned medium was collected after culture for another 24 h. MEFs were treated with the CMs described above for 48 h, and then, cell lysates were collected for analysis. Western blot and quantitative analysis of fibronectin, collagen I, and α-SMA expression. Data are presented as the 783 mean ± SEM. ***p < 0.001, **p < 0.01, *p < 0.05. (e) Schematic illustration of the hypothesized mechanism of agingrelated renal fibrosis

ACK N OWLED G M ENTS
We thank Cyagen, Suzhou Corp for constructing the transgene mice.
This work was supported by National Natural Science Foundation of China 81872254 (to Dr. Tan.) and 82000708 (to Dr. Wang).

CO N FLI C T O F I NTE R E S T
The authors declare no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.