Lysosomal-Associated Protein Transmembrane 5, Tubular Senescence, and Progression of CKD

Abstract

Key Points

• Lysosomal-associated protein transmembrane 5 (LAPTM5) is increased in tubular epithelial cells in CKD.

• Conditional knockout of Laptm5 in tubules attenuates kidney fibrosis in mice with CKD.

• LAPTM5 contributes to tubular senescence by inhibiting WWP2-mediated ubiquitination of notch1 intracellular domain.

Background

Tubular senescence is a major determinant of CKD, and identification of potential therapeutic targets involved in senescent tubular epithelial cells has clinical importance. Lysosomal-associated protein transmembrane 5 (LAPTM5) is a key molecule related to T- and B-cell receptor expression and inflammation. However, the expression pattern of LAPTM5 in the kidney and the contribution of LAPTM5 to the development of CKD are unknown.

Methods

Laptm5^−/− mice and tubule specific–Laptm5 knockout mice were used to examine the role of LAPTM5 in tubular senescence by establishing different experimental mouse CKD models.

Results

LAPTM5 expression was significantly induced in the kidney, especially in proximal tubules and distal convoluted tubules, from mice with aristolochic acid nephropathy, bilateral ischemia/reperfusion injury–induced CKD, or unilateral ureter obstruction. Tubule-specific deletion of Laptm5 inhibited senescence of tubular epithelial cells and alleviated tubulointerstitial fibrosis in aged mice. Moreover, Laptm5 deficiency ameliorated kidney injury and tubular senescence in mice with CKD. Mechanistically, LAPTM5 inhibited ubiquitination of notch1 intracellular domain by mediating WWP2 lysosomal degradation and then leading to cellular senescence in tubular epithelial cells. We also observed a higher expression of LAPTM5 in tubules from patients with CKD, and the level of LAPTM5 was correlated with kidney fibrosis and tubular senescence in people with CKD.

Conclusions

LAPTM5 contributed to tubular senescence by regulating the WWP2/notch1 intracellular domain signaling pathway and exacerbated kidney injury during the progression of CKD.

Visual Abstract

Introduction

CKD is a global health burden affecting approximately 9.1% of the global population.¹ Kidney fibrosis is considered as the ultimate common pathway for the progression of all types of CKDs.² Among kidney parenchymal cells, tubular epithelial cells are not only recognized as victims but also as an important driver in CKD³ and are the most likely to transition to the senescent condition and promote kidney fibrosis in CKD.⁴ Despite exiting the cell cycle, senescent tubular epithelial cells remain metabolically active and secrete cytokines, chemokines, and growth factors, thereby developing senescence-associated secretory phenotype (SASP) that is involved in the maladaptive repair during the progression of CKD.⁵ Therefore, clarifying the mechanisms and identifying the key molecules involved in tubular senescence may provide clues to new therapeutic strategies for patients with CKD.

Lysosome, an organelle responsible for intracellular protein digestion, has gained notoriety as the recycling center of the cell.⁶ Lysosomal abnormalities are not only associated with the development of CKD but also induce tubular senescence and exacerbate aging‐related kidney fibrosis.⁷ Therefore, maintaining lysosomal homeostasis can repair damaged proximal tubular epithelial cells and delay the progression of kidney injuries.⁸ Among the lysosomal-related proteins, lysosomal-associated protein transmembrane 5 (LAPTM5) is a member of LAPTM family, which consists of LAPTM4A, LAPTM4B, and LAPTM5. LAPTM5 is initially identified as a regulator of protein transport and lysosomal degradation.⁹ Previous studies have revealed that LAPTM5 negatively regulated T-cell receptor and B-cell receptor levels¹⁰ but positively mediated proinflammatory signaling pathways in macrophages.¹¹ Recently, new aspects of LAPTM5 functions in parenchymal cells are emerging. LAPTM5 can ameliorate nonalcoholic steatohepatitis⁹ and improve cardiac hypertrophy.¹² However, the expression pattern of LAPTM5 in the kidney and the contribution of LAPTM5 to the development of CKD still are unknown. Thus, the aim of this study was mainly to explore the role and mechanism of LAPTM5 in CKD.

Methods

Details are provided in the Supplemental Materials and Methods.

Human Kidney Biopsy Samples

Kidney biopsies were collected with the detailed information presented in Supplemental Tables 1 and 2. The investigations were conducted in accordance with the principles of the Declaration of Helsinki and were approved by the Research Ethics Committee of Shandong University (Document No. ECSBMSSDU2021-1-116).

Animal Studies

All experimental protocols for animal studies were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee of School of Basic Medical Sciences, Shandong University (Document No. ECSBMSSDU2021-2-208). For aristolochic acid nephropathy, male C57BL/6 mice, aged 8 weeks (24–28 g), were used. Aristolochic acid nephropathy was induced by a one-time intraperitoneal injection of aristolochic acid (5 mg/kg body weight, A5512, Sigma-Aldrich) in PBS.

Generation of Global Laptm5 Knockout Mice

Global Laptm5 knockout (Laptm5^−/−) mice were provided by Shanghai Model Organisms.

Generation of Tubule-Specific Laptm5 Knockout Mice

Floxed Laptm5 mice (Shanghai Model Organism) were crossed with mice expressing Cre recombinase (Cre) under the cadherin 16 promoter to generate tubule-specific Laptm5 knockout mice (Cdh16-Cre⁺/Laptm5^fl/fl; Cre⁺/Laptm5^fl/fl).

Intrarenal Lentivirus Delivery

Briefly, the mice were anesthetized, and a 31G needle was inserted at the lower pole of the kidney parallel to the long axis and was carefully pushed toward the upper pole after a temporary occlusion of kidney pedicle. As the needle was slowly removed, purified lentivirus cocktail was injected.

Cell Culture and Treatments

Rat kidney tubule epithelial cells (NRK52E) and normal rat kidney fibroblast cells (NRK49F) were purchased from American Type Culture Collection and cultured in DMEM supplemented with 10% FBS.

Senescence-Associated β-Galactosidase Staining

Briefly, frozen kidney tissues (5-μm thick) were fixed in fixative for 15 minutes and then washed in PBS. Then, the fixed sections were incubated with reaction buffer overnight at 37°C.

Immunoprecipitation Assay

Specific primary antibodies were incubated with Protein A&G magnetic beads for 1 hour at room temperature with constant rotation. The magnetic beads and sample lysates were incubated overnight at 4°C. Then, the immunoprecipitated proteins were separated using the magnetic separation rack and detected by Western blot.

Proximity Ligation Assay

Briefly, NRK52E cells were blocked with DuoLink blocking buffer and then incubated with target specific primary antibodies and specific proximity ligation assay probes. Then, oligonucleotides labeled by fluorescence were added to hybridized with the concatemeric products. The signal was detected by an LSM880 laser scanning confocal microscope.

Statistical Analysis

Data are expressed as mean±SEM. Statistical analyses were performed with GraphPad Prism (version 8.0, GraphPad Software, San Diego, CA). Details are present in the Supplemental Materials and Methods.

Results

LAPTM5 Was Induced in Tubules from Mice with Different Experimental Mouse CKD Models

By mRNA analysis, we found that Laptm5 was expressed in isolated perfused organs, including the heart, liver, spleen, lung, kidney, and bone marrow in the adult murine (Supplemental Figure 1A). Meanwhile, we further confirmed Laptm5 expression in kidney parenchymal cells, including NRK52E, rat glomerular endothelial cells, rat mesangial cells, and mouse podocytes (Figure 1, A and B). Compared with macrophages, the basal expression level of Laptm5 was relatively low in tubular epithelial cells (Figure 1, A and C). The presence of Laptm5 in tubular epithelial cells was further confirmed by Sanger sequencing analysis (Figure 1C and Supplemental Figure 1B).

Figure 1.

LAPTM5 was increased in tubules from multiple types of experimental CKD models. (A and B) The expression of Laptm5 in various kinds of renal parenchymal cells, including rat proximal tubule epithelial cells (NRK52E), rat glomerular endothelial cells, rat mesangial cells, and mouse podocytes. RAW264.7 and rat renal macrophage cells as a positive control. **P < 0.01, ***P < 0.001 versus positive control (n=6 biologically independent experiments). (C) Representative agarose gel showed the expression of Laptm5 in NRK52E cells, rat renal macrophage cells, TCMK1 (mouse renal tubular epithelial) cells, and RAW264.7 cells. (D) Volcano plots showed the expression of LAPTM family members in the kidneys involving aristolochic acid nephropathy (n=4), bilateral IRI (n=3), and UUO (n=4). (E) The association between the expression of Laptm5 and the expression of Mmp7, Tgf-β1, and Collagen I in the kidneys from aristolochic acid nephropathy (n=4), bilateral IRI (n=3), and UUO (n=4) mice. (F) Relative mRNA level of Laptm5 in the kidney cortex from mice with aristolochic acid nephropathy. *P < 0.05 versus control mice (n=5 mice per group). (G) Protein levels of LAPTM5 in the kidney cortex from mice with aristolochic acid nephropathy. ***P < 0.001 versus control mice (n=6 mice per group). (H) LAPTM5 expression in tubules from aristolochic acid nephropathy mice. LTL was used as a proximal tubular marker; calbindin was used as a marker for distal convoluted tubule; AQP3 was used as a marker for collecting duct. Scale bar: white=50 μm. ***P < 0.001 versus control mice (n=6 mice per group). Data are expressed as mean±SEM (A, B, and F–H). One-way ANOVA followed by Tukey's post-test (A). Two-tailed Student's unpaired t test analysis (B and F–H). Spearman's correlation coefficient r with two-tailed P value (E). AAN, aristolochic acid nephropathy; AQP3, aquaporin-3; DAPI, 4′,6-diamidino-2-phenylindole; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GEC, glomerular endothelial cell; IRI, ischemia/reperfusion injury; LAPTM5, lysosomal-associated protein transmembrane 5; LTL, Lotus tetragonolobus lectin; MPC, mouse podocyte; RMC, rat mesangial cell; UUO, unilateral ureter obstruction.

To further clarify the expression pattern of LAPTM5, we performed global gene expression profiling in the kidney from mice with aristolochic acid nephropathy,¹³ ischemia/reperfusion injury (IRI),^14,15 and unilateral ureter obstruction (UUO),¹⁶ three different animal models related to CKD. Among LAPTM family members, Laptm5 was induced in the kidney from mice with different CKD (Figure 1D). Moreover, Laptm5 had correlation with Mmp7, Tgf-β1, and Collagen I (Figure 1E). The upregulation of LAPTM5 was further confirmed in the kidney from aristolochic acid nephropathy (Figure 1, F and G), IRI (Supplemental Figure 1, C and D), and UUO (Supplemental Figure 1, E and F) mice by real-time RT-PCR and Western blot analyses. Meanwhile, we observed that LAPTM5 was mainly induced in proximal tubules and distal convoluted tubules (Figure 1H). Compared with controls, LAPTM5 was also increased in macrophages in the interstitium from mice with aristolochic acid nephropathy, which is possibly due to the increased infiltration of macrophages in the kidney (Supplemental Figure 1G). In vitro, aristolochic acid (Supplemental Figure 1H) or TGF-β1 (Supplemental Figure 1I) treatment induced LAPTM5 expression in NKR52E cells.

LAPTM5 Accelerated Kidney Injury and Tubular Senescence

To explore the role of LAPTM5 in the kidney, Laptm5^−/− mice were generated (Supplemental Figure 2). Laptm5 knockout did not cause any obvious changes of physiologic index in mice at age 4 months (Supplemental Table 3). However, Laptm5 deficiency attenuated tubular injury and tubulointerstitial fibrosis in mice with aristolochic acid nephropathy (Supplemental Figure 3A), which was further confirmed by the decrease of Vimentin, α-smooth muscle actin (α-SMA), Collagen I, and Collagen IV (Supplemental Figure 3, B and C). In particular, immunofluorescence staining showed that LAPTM5 was increased in senescent tubular cells from mice with aristolochic acid nephropathy (Supplemental Figure 4A) and aged mice (Supplemental Figure 4B).

Next, tubule-specific Laptm5 knockout (Cre⁺/Laptm5^fl/fl) mice were generated by a Cre-LoxP recombination system to better elucidate the role of LAPTM5 in renal tubular epithelial cells (Figure 2A), which was confirmed by tail genotyping (Supplemental Figure 5A) and a reduction of LAPTM5 in the kidney (Supplemental Figure 5B). Although Cre⁺/Laptm5^fl/fl mice were indistinguishable from control littermates as analyzed by the physiologic index (Supplemental Table 4), serum creatinine (Figure 2B), BUN (Figure 2C), kidney fibrosis (Figure 2D) and senescent indexes (Figure 2E) at 4 months after birth, tubule-specific deletion of Laptm5 moderately reduced the serum creatinine (Figure 2B) and kidney injury in aged mice (Figure 2D).There was no significant difference in the level of BUN in aged mice (Figure 2C). Laptm5 deficiency reduced the level of senescence-associated β-galactosidase (SA-β-gal) (Figure 2D), a biomarker of cellular senescence,¹⁷ as well as senescence-regulating proteins, including p53, p21, and p16 (Figure 2E), and SASP markers, including Il-6, Tnf-α, and Ctgf (Figure 2F), in the kidney of aged mice.

Figure 2.

Deficiency of Laptm5 ameliorated tubular senescence and kidney fibrosis in aged mice. (A) Generation of conditional knockout mice in which Laptm5 is specifically ablated in tubular cells by using the Cre–LoxP recombination system. (B) Serum creatinine in different groups of mice. *P < 0.05 versus Cre⁻/Laptm5^fl/fl mice at 4 months after birth; #P < 0.05 versus Cre⁻/Laptm5^fl/fl mice at 24 months after birth (n=6 mice per group). (C) BUN in different groups of mice (n=6 mice per group). (D) H&E staining, Masson's trichrome staining, and sirius red staining were performed to assess kidney injury and fibrosis. Photomicrographs and quantifications of Collagen I and α-SMA staining were performed to assess kidney fibrosis. SA-β-gal staining was performed to assess senescence in tubular epithelial cells. Scale bar: black=50 μm. ***P < 0.001 versus Cre⁻/Laptm5^fl/fl mice at 4 months after birth; ###P < 0.001 versus Cre⁻/Laptm5^fl/fl mice at 24 months after birth (n=6 mice per group). (E) Protein levels of p53, p21, and p16 in the kidney cortex from different groups. ***P < 0.001 versus Cre⁻/Laptm5^fl/fl mice at 4 months after birth; ###P < 0.001 versus Cre⁻/Laptm5^fl/fl mice at 24 months after birth (n=6 mice per group). (F) Relative mRNA level of Il-6, Tnf-α, and Ctgf in the kidney cortex from 24-month-old mice. ***P < 0.001 versus Cre⁻/Laptm5^fl/fl mice at 4 months after birth; ###P < 0.001 versus Cre⁻/Laptm5^fl/fl mice at 24 months after birth (n=6 mice per group). Data are expressed as mean±SEM (B–F). Two-way ANOVA followed by Tukey's post-test (B–F). H&E, hematoxylin and eosin; SA-β-gal, senescence-associated β-galactosidase; SCr, serum creatinine; α-SMA, α-smooth muscle actin.

Under pathological conditions (Figure 3A), tubule-specific deletion of Laptm5 markedly reduced the levels of serum creatinine (Figure 3B) and BUN (Figure 3C); alleviated tubular injury and tubulointerstitial fibrosis (Figure 3D); and reduced Collagen I, Collagen IV, Vimentin, and α-SMA (Figure 3D and Supplemental Figure 6) in mice with aristolochic acid nephropathy. We further confirmed the detrimental role of tubular LAPTM5 in kidney fibrosis in UUO mice (Supplemental Figure 7).

Figure 3.

Laptm5 deficiency mitigated tubular senescence and tubulointerstitial fibrosis under aristolochic acid nephropathy. (A) Schematic diagram shows the experimental procedure. (B) Serum creatinine in different groups of mice. ***P < 0.001 versus control (Cre⁻/Laptm5^fl/fl) mice; ##P < 0.01 versus Cre⁻/Laptm5^fl/fl aristolochic acid nephropathy mice (n=6 mice per group). (C) BUN in different groups of mice. ***P < 0.001 versus control (Cre⁻/Laptm5^fl/fl) mice; #P < 0.05 versus Cre⁻/Laptm5^fl/fl aristolochic acid nephropathy mice (n=6 mice per group). (D) H&E staining, Masson's trichrome staining, and sirius red staining were performed to assess kidney injury and fibrosis. Photomicrographs and quantifications of Collagen I staining were performed to assess kidney fibrosis. Scale bar: black=50 μm. ***P < 0.001 versus control (Cre⁻/Laptm5^fl/fl) mice; ###P < 0.001 versus Cre⁻/Laptm5^fl/fl aristolochic acid nephropathy mice (n=6 mice per group). (E) SA-β-gal, Klotho, and γ-H2A.X staining were performed to assess senescence in tubular epithelial cells. Scale bar: black=50 μm. ***P < 0.001 versus control (Cre⁻/Laptm5^fl/fl) mice; ###P < 0.001 versus Cre⁻/Laptm5^fl/fl aristolochic acid nephropathy mice (n=6 mice per group). (F) Representative Western blot gel documents and summarized data showed the relative protein levels of p53, p21, p16, and TGF-β1 in the kidney cortex from different groups of mice. ***P < 0.001 versus control (Cre⁻/Laptm5^fl/fl) mice; ###P < 0.001 versus Cre⁻/Laptm5^fl/fl aristolochic acid nephropathy mice (n=6 mice per group). (G) Relative mRNA level of Il-1β, Il-6, and Ctgf in the kidney cortex from aristolochic acid nephropathy mice. *P < 0.05, ***P < 0.001 versus control (Cre⁻/Laptm5^fl/fl) mice; ##P < 0.01, ###P < 0.001 versus Cre⁻/Laptm5^fl/fl aristolochic acid nephropathy mice (n=6 mice per group). Data are expressed as mean±SEM (B–G). Two-way ANOVA followed by Tukey's post-test (B–G). AA, aristolochic acid; i.p., intraperitoneal.

Importantly, Laptm5 deficiency could inhibit tubular senescence as evidenced by the reduction of SA-β-gal; the upregulation of Klotho, an aging suppressor¹⁸; and the decrease of aging‐associated DNA double‐strand break marker γ-H2A.X¹⁹ (Figure 3E). Meanwhile, Laptm5 deficiency reversed senescence-regulating protein (Figure 3F and Supplemental Figure 8) and SASP marker (Figure 3, F and G) expression in the kidney from aristolochic acid nephropathy mice. In addition, tubule-specific deletion of Laptm5 attenuated kidney injury and tubular senescence in IRI mice (Supplemental Figure 9).

LATPM5 Was Associated with Tubular Senescence in Patients with CKD

To further elucidate the importance of LATPM5 in tubular senescence, kidney tissue specimens from patients with CKD, including diabetic kidney disease (DKD, n=8) and hypertensive nephropathy (n=4), were collected (Figure 4A and Supplemental Table 1). LAPTM5 was increased in the kidney, especially in senescent tubular cells, of patients with DKD and hypertensive nephropathy, which showed interstitial fibrosis (Figure 4B). Meanwhile, the senescent indexes (γ-H2A.X and SA-β-gal) were also higher in the kidney of patients with DKD and hypertensive nephropathy (Figure 4B). Notably, the level of LAPTM5 was positively correlated with Vimentin (Figure 4C), α-SMA (Figure 4D), γ-H2A.X (Figure 4E), and SA-β-gal (Figure 4F), indicating that LATPM5 was associated with tubular senescence and kidney fibrosis in patients with CKD. The upregulation of LAPTM5 was further confirmed in the kidney from elderly participants (Figure 4G and Supplemental Table 2). Importantly, the significant increase of LAPTM5 was correlated with eGFR in people with CKD (Figure 4H) and elderly participants (Figure 4I).

Figure 4.

LATPM5 was associated with tubular senescence in patients with CKD. (A) Schematic diagram shows the experimental procedure. (B) Photomicrographs and quantifications of LAPTM5, Masson's trichrome, sirius red, Vimentin, α-SMA, γ-H2A.X, and SA-β-gal staining in kidney sections of patients with CKD, including DKD and hypertensive nephropathy. Scale bar: black=50 μm, white=25 μm. ***P < 0.001 versus normal participants (n=5 for normal participants, n=8 for patients with DKD, n=4 for patients with hypertensive nephropathy). (C) Correlation between LAPTM5 expression and the degree of Vimentin staining in the patients with CKD (n=12). (D) Correlation between LAPTM5 expression and the degree of α-SMA staining in the patients with CKD (n=12). (E) Correlation between LAPTM5 expression and the degree of γ-H2A.X staining in the patients with CKD (n=12). (F) Correlation between LAPTM5 expression and the degree of SA-β-gal staining in the patients with CKD (n=12). (G) Photomicrographs and quantifications of LAPTM5 staining in kidney sections of elderly participants. Scale bar: black=50 μm. ***P < 0.001 versus young participants (n=5 for young participants, n=10 for elderly participants). (H) Correlation between LAPTM5 expression and eGFR in the patients with CKD (n=12). (I) Correlation between LAPTM5 expression and eGFR in all elderly participants (n=10). Data are expressed as mean±SEM (B and G). Two-tailed Student's unpaired t test analysis (B and G); Spearman's correlation coefficient r with two-tailed P value (C–F and H and I). DKD, diabetic kidney disease; HN, hypertensive nephropathy.

LAPTM5 Contributed to Kidney Fibrosis by Mediating Cell Crosstalk between Senescent Tubular Epithelial Cells and Interstitial Fibroblasts

To clarify the causal relationship between LAPTM5-mediated tubular senescence and tubulointerstitial fibrosis, we explored the cell crosstalk between senescent tubular epithelial cells and fibroblasts. In vitro, LAPTM5 overexpression (Supplemental Figure 10A) induced the expression of p21, p16, and SASP markers, but reduced Klotho and Lamin B1 expression in NRK52E cells (Figure 5, A and B). Notably, overexpression of LAPTM5 had no obvious effect on lysosomal pH (Supplemental Figure 10B) as well as ATP6V0D1 and ATP6V1G1 expression (Supplemental Figure 10C), which are two key V-ATPases in maintaining lysosomal acidification in tubular epithelial cells.⁷ However, in NRK52E cells with aristolochic acid treatment, Laptm5 knockdown (Supplemental Figure 10D) reduced SA-β-gal, p21, and p16 expression (Figure 5C); increased Klotho, p-Rb, and Lamin B1 expression (Figure 5C and Supplemental Figure 10E); and reduced the expression and secretion of SASP markers (Figure 5, C and D, and Supplemental Figure 10F). In addition, Laptm5 knockdown attenuated the loss of epithelial phenotype in NRK52E cells with aristolochic acid treatment (Supplemental Figure 10G).

Figure 5.

LAPTM5 contributed to kidney fibrosis by mediating cell crosstalk between senescent tubular epithelial cells and interstitial fibroblasts. (A) Representative Western blot gel documents and summarized data showed the relative protein levels of Klotho, Lamin B1, p21, and p16 in LAPTM5-overexpressed NRK52E cells. **P < 0.01, ***P < 0.001 versus OE-NC (n=6 biologically independent experiments). (B) Relative mRNA level of Tnf-α, Mcp-1, Mmp7, and Cxcl10 in LAPTM5-overexpressed NRK52E cells. *P < 0.05, ***P < 0.001 versus OE-NC (n=6 biologically independent experiments). (C) Representative Western blot gel documents and summarized data showed the relative protein levels of Klotho, SA-β-gal, p21, p16, and TGF-β1 in NRK52E with aristolochic acid treatment. ***P < 0.001 versus scramble group; ##P < 0.01, ###P < 0.001 versus scramble group of aristolochic acid treatment (n=6 biologically independent experiments). (D) The levels of TGF-β1, IL-1β, and TNF-α measured by ELISA. *P < 0.05, **P < 0.01, ***P < 0.001 versus scramble group; #P < 0.05, ###P < 0.001 versus scramble group of aristolochic acid treatment. (TGF-β1 and TNF-α: n=6 biologically independent experiments; IL-1β: n=4 biologically independent experiments). (E) Experimental scheme for cell treatment: after 36-hour treatment of NRK52E cells with aristolochic acid, the drug was washed out and the cells continued in culture for 24 hours. The CM was then collected and added to serum-starved NRK49F cells. (F) Protein levels of α-SMA and PCNA in fibroblasts treated with CM from control or senescent NRK52E cells. ***P < 0.001 versus CM from untreated NRK52E cells; ###P < 0.001 versus CM from aristolochic acid–treated NRK52E cells (n=6 biologically independent experiments). (G) Protein levels of α-SMA and PCNA in fibroblasts treated with CM from NRK52E cells with or without LAPTM5 overexpression. ***P < 0.001 versus CM from NRK52E cells with negative control (OE-NC) (n=6 biologically independent experiments). (H) Photomicrographs and quantifications showed PCNA expression in the kidney from different groups of mice (up panel). PCNA-positive cells per hpf are counted and shown. Scale bar: black=50 μm. Representative photomicrographs of kidney sections stained for α-SMA, PDGFRβ⁺, and DAPI (down panel) as well as quantitative analysis of α-SMA staining in the kidney. Scale bar: white=20 μm. ***P < 0.001 versus control (Cre⁻/Laptm5^fl/fl) mice; ###P < 0.001 versus Cre⁻/Laptm5^fl/fl aristolochic acid nephropathy mice (n=6 mice per group). Data are expressed as mean±SEM (A–D and F–H). Two-tailed Student's unpaired t test analysis (A, B, and G). Two-way ANOVA followed by Tukey's post-test (C, D, F, and H). CM, conditioned medium; hpf, high power field; OE-NC, negative control of LAPTM5 overexpression; PCNA, proliferating cell nuclear antigen; PDGFRβ, platelet-derived growth factor receptor β; shRNA, short hairpin RNA.

Considering that senescent tubular cells can lead to fibroblast activation by secreting the SASP component,⁴ we then cultured fibroblasts NRK49F using the conditioned medium (CM) from aristolochic acid–treated NRK52E cells (Figure 5E). CM promoted the activation of fibroblasts, which was reversed by gene silencing of Laptm5 (Figure 5F). Consistently, CM collected from LAPTM5-overexpression NRK52E cells could also promote the activation of fibroblast (Figure 5G). In vivo, Laptm5 deficiency reduced the protein levels of proliferating cell nuclear antigen and α-SMA, especially in platelet-derived growth factor receptor β-positive fibroblasts from fibrotic kidneys, suggesting that tubular LAPTM5 promoted fibroblast activation (Figure 5H).

LAPTM5 Inhibited Ubiquitination of Notch1 Intracellular Domain by Mediating WWP2 Lysosomal Degradation

The activation of Notch signaling plays an important role in tubular senescence and kidney fibrosis.^20,21 On the basis of the results of global gene expression profiling showing that Notch signaling was activated in the kidney from mice with aristolochic acid nephropathy as evidenced by upregulation of HeyL and Hes7 (Supplemental Figure 11A) and mass spectrometry analysis showing an increased tendency in the protein levels of Notch1 and Hes1 in NRK52E cells with overexpression of LAPTM5 (Supplemental Figure 11B), we therefore proposed that Notch might be a key target in LAPTM5-mediated kidney fibrosis.

We further found that LAPTM5 overexpression preferentially upregulated the expression of notch1 intracellular domain (NICD1), which is the active intracellular domain of Notch1, but had no effects on NICD2, NICD3, and NICD4 in NRK52E cells (Figure 6A). Laptm5 knockdown reduced NICD1 and Hes1 expression in NRK52E cells with aristolochic acid treatment (Figure 6B). Consistently, Laptm5 deficiency decreased NICD1 expression in the kidney from mice with aristolochic acid nephropathy (Supplemental Figure 11C). Moreover, we further identified that NICD1 was increased in senescent tubular cells (Supplemental Figure 11D) and positively correlated with LAPTM5 in the kidney of patients with CKD (Supplemental Figure 11E). Notably, there was no obvious change of Notch1 mRNA level in aristolochic acid–treated NRK52E (Supplemental Figure 11F) and in the kidney of Laptm5 conditional knockout mice with aristolochic acid nephropathy (Supplemental Figure 11G), suggesting that LAPTM5-mediated post-transcriptional modification may be involved in regulating the protein level of NICD1. Considering that NICD1 could be cleared by continuous ubiquitination (Ub) and proteasomal degradation,²² we then detected the NICD1 Ub and found that Laptm5 knockdown increased the Ub level of NICD1 in aristolochic acid–treated NRK52E cells (Figure 6C).

Figure 6.

LAPTM5 inhibited Ub of NICD1 by mediating WWP2 lysosomal degradation. (A) Representative Western blot gel documents and summarized data showed the relative protein levels of NICD1, NICD2, NICD3, and NICD4 in the LAPTM5-overexpressed NRK52E cells. *P < 0.05 versus OE-NC (n=3 biologically independent experiments). (B) Representative Western blot gel documents and summarized data showed the relative protein levels of NICD1 and Hes1 in NRK52E with aristolochic acid treatment. ***P < 0.001 versus scramble group; ###P < 0.001 versus scramble group of aristolochic acid treatment (n=6 biologically independent experiments). (C) IP demonstrated that Laptm5 knockdown increased the Ub of NICD1. (D) IP demonstrated that WWP2 bound to LAPTM5 and NICD1 in NRK52E cells under normal conditions. (E) PLA of the interaction between endogenous WWP2 and LAPTM5 or endogenous WWP2 and NICD1 in NRK52E cells under normal conditions. Scale bar: 5 μm. (F) IP demonstrated that WWP2 bound to LAPTM5 and NICD1 in the cortex of kidney from mice. (G) Protein levels of WWP2 in NRK52E with aristolochic acid treatment. ***P < 0.001 versus scramble group; ###P < 0.001 versus scramble group of aristolochic acid treatment (n=6 biologically independent experiments). (H) Protein levels of WWP2 in NRK52E with NH₄Cl treatment. ***P < 0.001 versus OE-NC; ###P < 0.001 versus OE-LAPTM5 group (n=6 biologically independent experiments). (I) Protein levels of WWP2 in NRK52E with chloroquine treatment. ***P < 0.001 versus OE-NC; ###P < 0.001 versus OE-LAPTM5 group (n=6 biologically independent experiments). (J). Colocalization of LAPTM5 and WWP2 in lysosomes in NRK52E cells. Scale bar: 10 μm. Data are expressed as mean±SEM (A, B, and G–I). Two-tailed Student's unpaired t test analysis (A). Two-way ANOVA followed by Tukey's post-test (B and G–I). CQ, chloroquine; IP, immunoprecipitation; NICD1, notch1 intracellular domain; OE-LAPTM5, LAPTM5-overexpressed; PLA, proximity ligation assay; Ub, ubiquitination.

To clarify the mechanisms by which LAPTM5 regulates NICD1 Ub, we performed immunoprecipitation (IP)-mass spectrometry with overexpression of FLAG-tagged LAPTM5 in NRK52E cells. Among the interacting proteins of LAPTM5, WWP2 was one of the enriched proteins that are involved in Ub (Supplemental Table 5). Co-IP showed that WWP2 could bind to LAPTM5 and NICD1 (Figure 6D), which was further confirmed by proximity ligation assay (Figure 6E), but there was no obvious direct interaction between LAPTM5 and NICD1 in NRK52E cells under normal conditions (Supplemental Figure 11H). Meanwhile, the interaction between WWP2 and LAPTM5 or NICD1 was further confirmed in the kidney of mice by IP (Figure 6F) and double immunostaining (Supplemental Figure 11I). Moreover, Laptm5 deficiency increased WWP2 expression in aristolochic acid–treated NRK52E cells (Figure 6G) and in the kidney from aristolochic acid nephropathy mice (Supplemental Figure 11J). Contrarily, LAPTM5 overexpression reduced WWP2 expression in NRK52E cells, which was abrogated by NH₄Cl (Figure 6H) and chloroquine (Figure 6I), two inhibitors of lysosomal activity. However, proteasome inhibitor MG-132 had no obvious effect on WWP2 expression in NRK-52E cells with LAPTM5 overexpression (Supplemental Figure 11K). Immunofluorescent staining further confirmed that LAPTM5 and WWP2 colocalized with LAMP1, a marker for lysosomes (Figure 6J). However, there was no change in Nedd4 expression in NRK52E cells with LAPTM5 overexpression (Supplemental Figure 11L). Meanwhile, overexpression of Nedd4 also had no effect on LAPTM5 expression (Supplemental Figure 11M). In addition, overexpression of WWP2 (Supplemental Figure 12A) reduced NICD1 expression (Figure 7A) but had no obvious effect on LAPTM5 expression in NRK52E cells with aristolochic acid treatment (Supplemental Figure 12B). Importantly, Wwp2 knockdown could decrease the Ub of NICD1 in Laptm5-knockdown NRK52E cells with aristolochic acid treatment (Figure 7B), suggesting LAPTM5 inhibits NICD1 Ub by mediating WWP2 degradation in the lysosome.

Figure 7.

WWP2/NICD1 axis was involved in LAPTM5-mediated cellular senescence. (A) Protein levels of NICD1 in the WWP2-overexpressed NRK52E cells with aristolochic acid treatment. ***P < 0.001 versus negative control of WWP2 overexpression (OE-NC); ###P < 0.001 versus negative control with aristolochic acid treatment (n=6 biologically independent experiments). (B) IP demonstrated that Wwp2 knockdown decreased the Ub of NICD1 in aristolochic acid–treated NRK52E cells with gene silencing of Laptm5. (C) Representative Western blot gel documents and summarized data showed the relative protein levels of p21 and p16 in NRK52E with the DAPT treatment. ***P < 0.001 versus OE-NC; ###P < 0.001 versus OE-LAPTM5 group (n=6 biologically independent experiments). (D) Relative mRNA level of Il-1β, Il-6, and Mmp9 in NRK52E with DAPT treatment. **P < 0.01, ***P < 0.001 versus OE-NC; #P < 0.05, ###P < 0.001 versus OE-LAPTM5 group (n=6 biologically independent experiments). (E) Representative Western blot gel documents and summarized data showed the relative protein levels of SA-β-gal and p21 in NRK52E with different treatments. ***P < 0.001 versus Laptm5-deficiency NRK52E cells with aristolochic acid treatment. (n=6 biologically independent experiments). (F) Representative Western blot gel documents and summarized data showed the relative protein levels of α-SMA and PCNA in fibroblasts treated with CM from NRK52E cells with different treatments. ***P < 0.001 versus CM/aristolochic acid/shRNA-Laptm5/OE-NC. (n=6 biologically independent experiments). (G) Schematic depicting LAPTM5 inhibited Ub of NICD1 by mediating WWP2 lysosomal degradation, then leading to senescence in tubular epithelial cells, finally contributing to the progression of CKD. Data are expressed as mean±SEM (A and C–F). Two-way ANOVA followed by Tukey's post-test (A and C–F). DAPT, N-[N-(3,5-difluorophenacetyl)-l-alanyl]-S-phenylglycine t-butyl ester; OE-NICD1, NICD1 overexpression; SASP, senescence-associated secretory phenotype.

WWP2/NICD1 Axis Was Involved in LAPTM5-Mediated Cellular Senescence

To further evaluate the relationship between NICD1 and LAPTM5-mediated tubular senescence, N-[N-(3,5-difluorophenacetyl)-l-alanyl]-S-phenylglycine t-butyl ester, an inhibitor of γ-secretase that could catalyze protein cleavage of Notch1 and induce the release of NICD1, was used in this study. We found that N-[N-(3,5-difluorophenacetyl)-l-alanyl]-S-phenylglycine t-butyl ester reduced p21 and p16 expression (Figure 7C) and attenuated SASP marker upregulation in NRK52E with LATPM5 overexpression by protein and mRNA analysis using specific antibodies and primers (Figure 7D and Supplemental Tables 6 and 7). Meanwhile, NICD1 overexpression (Supplemental Figure 13A) promoted cellular senescence (Figure 7E) and epithelial phenotype loss in aristolochic acid–treated NRK52E cells with Laptm5 knockdown (Supplemental Figure 13B). Wwp2 knockdown (Supplemental Figure 13C) also counteracted the effect of Laptm5 knockdown in NRK52E with aristolochic acid treatment (Supplemental Figure 13D). Finally, we found that NICD1 overexpression (Figure 7F) or Wwp2 knockdown (Supplemental Figure 13E) in Laptm5-knockdown NRK52E cells with aristolochic acid treatment promoted fibroblast activation.

Targeting LAPTM5 Ameliorated Kidney Injury in Mice with CKD

To examine genetic therapeutic efficiency by targeting Laptm5 in mice, recombinant lentivirus vectors GV493 harboring short hairpin RNA-Laptm5 were delivered into the kidney by means of intraparenchymal injections (Supplemental Figure 14A), which was confirmed by the reduction of LAPTM5 in the kidney (Supplemental Figure 14B). Laptm5 knockdown decreased serum creatinine (Supplemental Figure 15A) and BUN (Supplemental Figure 15B), alleviated kidney injury (Supplemental Figure 15C), inhibited tubular senescence (Supplemental Figure 15, D–F), and attenuated fibroblast activation in aristolochic acid nephropathy mice (Supplemental Figure 15G).

Discussion

Although recent studies reveal that LAPTM5 is a key molecule related to T- and B-cell receptor expression, inflammation, tumor, and cardiovascular disease,^{11,12,23–25} the role of LAPTM5 in the kidney is unclear. In this study, we found that LAPTM5 was induced in the kidney, especially in proximal tubules and distal convoluted tubules, from different mouse models of CKD. Furthermore, the upregulation of LAPTM5 was correlated with kidney fibrosis and tubular senescence in patients with CKD, suggesting that LAPTM5 is involved in kidney fibrosis and aging during the development of CKD.

The kidney is one of the most prominent organs affected by aging.²⁶ Moreover, the kidney’s pathophysiological changes in CKD share many similar characteristics with natural aging, suggesting that CKD is a clinical manifestation of premature aging or accelerated senescence.⁴ Senescence, as the key feature of aging at the cellular level, plays an important role in multiple physiological and pathological conditions.²⁷ Accumulating evidence has revealed that senescence of tubular epithelial cells is a key event in the progression of CKD.^28–30 Transplantation of a small number of senescent renal scattered tubular-like cells are sufficient to induce kidney injury.³¹ Meanwhile, clearance of senescent renal epithelial cells can alleviate kidney injury.²⁹ Knockout of p21 and p16 protects against tubular senescence and fibrotic changes in CKD.^32,33 Therefore, senotherapies by targeting tubular senescence may provide novel therapies to CKD. In this study, we demonstrated that tubular LAPTM5 contributed to fibroblast activation by mediating cell crosstalk, indicating the importance of LAPTM5 in driving tubular senescence and kidney fibrosis.

Mechanistically, LAPTM5 inhibited degradation of NICD1 by regulating Ub. The Notch signaling mainly comprises a family of transmembrane receptors and their ligands. Upon interaction with the ligand, Notch receptor can be cleaved by γ-secretase, releasing NICD into the nucleus, thereby inducing the transcription of target genes, such as Hes and Hey.³⁴ Among the Notch family, Notch1 is recognized as a detrimental factor in tubular epithelial cells during CKD.^35–37 Conditional expression of NICD1 in proximal tubules is associated with a prosenescent phenotype and maladaptive repair after AKI.²⁰ We found that LAPTM5 selectively upregulated the protein level of NICD1 and further demonstrated that LAPTM5 inhibited the degradation of NICD1 by reducing Ub of NICD1.

Among the interacting proteins of LAPTM5, WWP2 is one of the enriched proteins that are involved in Ub. WWP2, a member of the Nedd4 family of E3 ligases,³⁸ can facilitate the conjugation of ubiquitin to a number of proteins, including NICD1.³⁹ Moreover, a recent study has reported that LAPTM5 could regulate the lysosomal degradation of WWP2 in B cells.¹⁰ In this study, we confirmed that LAPTM5 could bind to WWP2 in renal tubular epithelial cells, followed by inducing WWP2 degradation in the lysosome. Meanwhile, we further proved that the WWP2/NICD1 axis was involved in LAPTM5-mediated cellular senescence. These results indicated that LAPTM5 inhibited Ub of NICD1 at least in part by promoting lysosomal degradation of WWP2, finally leading to tubular senescence. Notably, a very recent study reported that WWP2 can promote kidney fibrosis by targeting myofibroblasts,⁴⁰ inconsistent with the role of WWP2 in tubular epithelial cells in this study. This discrepancy may be because of the following reasons: First, recent studies have demonstrated that the role of WWP2 is cell specific. For example, WWP2 derived from the myocardium plays a protective role in myocardial fibrosis,⁴¹ but fibroblast WWP2 promotes cardiac fibrosis.⁴² Second, the difference in animal models may also cause this discrepancy. The data from single-cell RNA sequencing showed that Wwp2 was decreased in fibroblasts from IRI mice on day 14 and 42 after injury⁴³ but increased in fibroblasts from patients with DKD,⁴⁴ indicating that WWP2 expression was different in fibroblasts under different kidney disease. In addition, differences in the interval of observation may also lead to different conclusions. WWP2 was reduced in the S2 segment of the proximal tubule from IRI mice on day 14 after injury, but there was no obvious change in Wwp2 expression at day 42.⁴³ Collectively, controversy about the role of WWP2 in kidney fibrosis might be attributed to the difference in cell types, animal models, and interval of observation, which need to be clarified in the future.

It should be noted that there are some limitations in this study. First, further studies are required to delineate the mechanisms by which various stimuli increase LAPTM5 expression under pathologic conditions, although recent studies suggested that transcription factor EB might be one of the potential molecules in regulating LAPTM5 expression.^45,46 Second, because of the uncertain recombination taking place during development, doxycycline or tamoxifen-inducible Laptm5 knockout in tubules should be considered in the future study.^47,48

Collectively, our studies demonstrate that LAPTM5 contributes to tubular senescence at least in part by inhibiting WWP2-mediated Ub of NICD1 in CKD (Figure 7G). Genetic or pharmacological targeting of LATPM5 may provide a novel approach for the treatment of CKD.

Supplementary Material

jasn-35-1655-s001.pdf ^{(1.4MB, pdf)}

jasn-35-1655-s002.pdf ^{(4.2MB, pdf)}

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/JSN/E790.

Funding

F. Yi: National Natural Science Foundation of China (T2321004 and 82090024) and Natural Science Foundation of Shandong Province (ZR2019ZD40). M. Liu: National Natural Science Foundation of China (82090021) and Natural Science Foundation of Shandong Province (ZR2019BH030).

Author Contributions

Conceptualization: Min Liu, Jinpeng Sun, Fan Yi.

Data curation: Huiying Jin, Min Liu, Xiaohan Liu, Youzhao Wang, Yujie Yang, Fan Yi, Ping Zhan, Yang Zhang.

Formal analysis: Min Liu, Xiaohan Liu, Rong Sun, Fan Yi.

Funding acquisition: Min Liu, Fan Yi.

Investigation: Huiying Jin, Xiaohan Liu, Rong Sun, Xiaojie Wang, Youzhao Wang, Ziying Wang, Yujie Yang, Ping Zhan, Yang Zhang.

Methodology: Min Liu, Xiaojie Wang, Ziying Wang.

Project administration: Min Liu, Fan Yi.

Resources: Qianqian Xu, Junhui Zhen.

Supervision: Min Liu, Jinpeng Sun, Xiaojie Wang, Ziying Wang, Qianqian Xu, Fan Yi, Junhui Zhen.

Validation: Huiying Jin, Xiaohan Liu, Jinpeng Sun, Youzhao Wang, Yujie Yang, Ping Zhan, Yang Zhang.

Visualization: Xiaohan Liu.

Writing – original draft: Min Liu, Xiaohan Liu, Fan Yi.

Writing – review & editing: Min Liu, Fan Yi.

Data Sharing Statement

RNA-sequencing datasets have been deposited to Gene Expression Omnibus under accession code GSE254440 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE254440). The mass spectrometry proteomics data and immunoprecipitation-mass spectrometry data have been deposited to the ProteomeXchange Consortium (https://proteomecentral.proteomexchange.org) via the iProX partner repository with the dataset identifier PXD052845.

Supplemental Material

This article contains the following supplemental material online at http://links.lww.com/JSN/E789.

Materials and Methods.

Supplemental Table 1. Clinical characteristics in normal human control individuals or people with CKD.

Supplemental Table 2. Clinical characteristics in young participants or elderly participants.

Supplemental Table 3. Physical and biochemical parameters of mice with Laptm5 deficiency.

Supplemental Table 4. Physical and biochemical parameters of mice with Laptm5 deficiency in tubules.

Supplemental Table 5. Proteins interacting with LAPTM5 from mass spectrometry analysis.

Supplemental Table 6. Primer pairs of target genes used for PCR in this study.

Supplemental Table 7. Antibodies used in this study.

Supplemental Figure 1. Expression of LAPTM5 in the kidney and tubular epithelial cells.

Supplemental Figure 2. Establishment of Laptm5 knockout (Laptm5^−/−) mice.

Supplemental Figure 3. Laptm5 deficiency attenuated kidney fibrosis in aristolochic acid nephropathy.

Supplemental Figure 4. Expression of LAPTM5 in senescent cells from aged mice or mice with aristolochic acid nephropathy.

Supplemental Figure 5. Tubule-specific Laptm5 knockout mice were generated.

Supplemental Figure 6. Tubule-specific deletion of Laptm5 attenuated kidney fibrosis in mice with aristolochic acid nephropathy.

Supplemental Figure 7. Laptm5 deletion in tubules ameliorated tubulointerstitial fibrosis induced by unilateral ureter obstruction (UUO).

Supplemental Figure 8. Laptm5 deletion in tubules upregulated the expression of p-Rb in the kidney from mice with aristolochic acid nephropathy.

Supplemental Figure 9. Laptm5 deficiency alleviated tubulointerstitial fibrosis and premature senescence of tubular epithelial cells in bilateral ischemia/reperfusion injury (IRI)-induced CKD.

Supplemental Figure 10. Laptm5 knockdown attenuated cellular senescence and loss of epithelial phenotype in NRK52E cells with aristolochic acid treatment.

Supplemental Figure 11. LAPTM5 inhibited ubiquitination of NICD1 by mediating WWP2 lysosomal degradation.

Supplemental Figure 12. Expression of WWP2 and LAPTM5 in NRK52E cells with different treatments.

Supplemental Figure 13. WWP2/NICD1 axis was involved in LAPTM5-mediated cellular senescence.

Supplemental Figure 14. Gene silencing efficiency of Laptm5 in mice.

Supplemental Figure 15. Targeting LAPTM5 attenuated kidney injury in mice with CKD.