Introduction

Acute kidney injury (AKI) is characterized by a rapid loss of renal functions1,2, and it affects 20–50% of the hospitalized patients3. The pathogenesis of AKI is quite complex, and previous publications indicate that multiple bioprocesses, such as, apoptosis, necroptosis, redox injury, cellular senescence, and hypoxia, are involved in the pathogenesis of AKI4,5,6,7. AKI may be the result of various events, including infection, hypovolemic shock, cardiac surgery, and the administration of nephrotoxins8. Cisplatin is a commonly used as an anti-cancer drug, but it is also well-known for its severe nephrotoxicity5. Previous literature indicated that ~30% of patients receiving high dose of cisplatin treatment had variable extent of renal injuries9. Morphologically, cisplatin-induced AKI is characterized by severe tubular damage, while glomerular damage is relatively mild. Despite decades of investigations, the precise mechanism(s) of cisplatin-induced AKI remains elusive.

Apoptosis, a regulated cell death, has been extensively investigated in the pathogenesis of cisplatin-induced AKI, and it is characterized by DNA fragmentation and apoptotic bodies10,11. Apoptosis is initiated in multiple ways, and intrinsic apoptosis has drawn the most attention, which occurs by the disruption of the outer mitochondrial membrane12. Subsequently, mitochondrial proteins, such as, Cytochrome C and SMAC, leaks out into the cytoplasm, leading to the activation of multiple caspases and ultimately cell death13. The permeability of the outer mitochondrial membrane is modulated by the pro-apoptotic factors (BAX and BAK) and the anti-apoptotic factors (BCL2 and BCL-XL)12. It is noteworthy that the bio-process of apoptosis is usually accompanied with excessive oxidative stress, and the attenuation of intracellular ROS generation conceivably can mitigate the progression of intrinsic apoptosis in AKI14. Interestingly, it has been well established that oxidative stress and intrinsic apoptosis were interdependent and mutually promotive in the progression of cisplatin-induced AKI15.

Ubiquitination-mediated proteasomal degradation, a process for the maintenance of internal homeostasis, is associated with the progression of intrinsic apoptosis and ROS generation. Interestingly, the role of proteasomal degradation in apoptosis is dependent upon the given settings. Previous literature data suggested that proteasome inhibition accentuated cisplatin-induced intrinsic apoptosis in multiple cancer cell lines16,17, while a protective effect of proteasome inhibition was observed in cisplatin-induced AKI18. E3 ligase-mediated ubiquitination of targeted proteins is a necessary prelude to proteasomal degradation, and various types of E3 ligases have been investigated in the progression of cisplatin-induced AKI. Carboxy-terminus of Hsc70-interacting protein (CHIP), a ubiquitously expressed U-box E3 ligase, is involved in the maintenance of protein turnover in various tissues19. It has been well-documented that CHIP participated in the modulation of damage in various organs20,21 and cisplatin-induced cellular injury in cancer cell lines22,23. In kidneys, previous studies indicated that CHIP regulated water handling in distal convoluted tubules24 and oxidative stress25 in proximal tubular cells. However, the role of CHIP in cisplatin-induced AKI has been inadequately investigated.

NUR77, a member of the nuclear receptor superfamily, is widely expressed in various types of tissues and cells. The induction of NUR77 can be observed in multiple pathological conditions, including inflammation, stress, and hormone treatment, and it can participate in various bioprocesses as a transcription factor or co-regulator26. In kidneys, it has been demonstrated that NUR77 inhibited TGF-β signaling to mitigate obstruction- and age-related renal fibrosis27,28. Besides, it was also reported that NUR77 deficiency accentuated macrophage-induced renal injuries in hypertensive nephropathy29. However, NUR77 conceived a detrimental phenotype in AKI states via its interaction with BCL230.

In this study, we presented supportive evidence that CHIP was downregulated in cisplatin-induced AKI, and CHIP indirectly interacted with NUR77 with its CC domain to catalyze the proteasomal degradation of NUR77, leading to the attenuation of cisplatin-related renal injuries, which shed light on the pathogenesis of cisplatin-related nephropathy.

Results

CHIP, abundantly expressed in proximal tubular cells, was down-regulated in cisplatin-induced AKI in mice

Four types of E3 ligases (including HECT, RING-finger, U-box, and PHD-finger) have been identified31, and the role of U-box E3 ligases is not adequately investigated in AKI states. RPKM value analysis indicated that CHIP had the highest expression in kidneys among the eight types of U-box E3 ligase identified in human tissues (including UFD2a, UFD2b, CHIP, UIP5, CYC4, PRP19, WDSUB1, and ACT1)32 (Fig. 1A). Similar profile was observed by qRT-PCR analyses (Fig. 1B, C). Co-immunofluorescence staining of CHIP and MEGALIN (a biomarker of renal proximal tubules) showed that CHIP was abundantly expressed in renal proximal tubules (Fig. 1D). In addition, we observed that CHIP was downregulated in cisplatin-treated BUMPT cells (Boston University mouse renal proximal tubular cells) and kidneys, as indicated by immuno-blotting, immunohistochemistry, and immunofluorescence staining studies (Fig. 1E–H). Besides, the changes of CHIP were also investigated at different time points in cisplatin-induced AKI. Our western blot studies and qRT-PCR analysis showed that CHIP was downregulated in cisplatin-treated BUMPT cells (from 12 h to 24 h) or kidneys (from day 2 to day 3) (Supplemental Fig. 1). Taken together, these data indicates that CHIP is highly expressed in renal proximal tubules, and it is downregulated in cisplatin-induced AKI.

Fig. 1: CHIP was downregulated in cisplatin-induced AKI.
figure 1

A RPKM value analysis showed that CHIP had the highest value than other U-box E3 ligases in human or mouse kidneys. B, C qRT-PCR data confirmed that CHIP had the highest mRNA levels in BUMPT cells and mouse kidneys, as compared with other U-box E3 ligases. D Immunofluorescence staining showed that CHIP (green) co-stained with MEGALIN (red, a marker of proximal tubules), suggesting that CHIP is highly expressed in proximal renal tubular cells of mouse kidneys. E Immunofluorescence staining showed that the expression of CHIP was decreased in cisplatin-treated BUMPT cells. F Immunohistochemistry staining indicated that CHIP was downregulated in cisplatin-treated mouse kidneys. G, H Western blot studies revealed that CHIP was downregulated in cisplatin-treated BUMPT cells and mouse kidneys. (n = 3 *P = 0.0023 vs. the WT BUMPT cells; n = 3 *P = 0.0269 vs. the WT mouse kidneys) Scale bars: 50 μm, Error bars show SD, WT: wide type, CP: cisplatin.

CHIP overexpression attenuated, while its gene disruption accentuated, cisplatin-induced cellular injury in BUMPT cells

Lentivirus was used to modulate the expression profile of CHIP in BUMPT cells (Figs. 2B & 3B). There was shrinkage of BUMPT cells following cisplatin treatment, which was ameliorated by CHIP overexpression but worsened by CHIP gene disruption (Figs. 2A & 3A). TUNEL staining showed that cisplatin-induced DNA damage was attenuated by CHIP overexpression, and exacerbated by CHIP knockdown (Figs. 2A, D, 3A, and D). The positive control for TUNEL staining is included in Supplemental Fig. 2. ROS generation, as indicated by DCF and DHE staining, was readily increased following cisplatin treatment, and these cellular changes were modulated by CHIP expression (Figs. 2A, E, 3A, and E). The findings of MTT assay revealed that cell death was induced by cisplatin treatment, which was also modulated by the expression profile of CHIP (Figs. 2F & 3F). It is noteworthy to indicate that cisplatin treatment led to an increased expression of cleaved caspase 3 and BAX, and a decreased expression of BCL2. These changes were mitigated by CHIP overexpression but were exacerbated by CHIP gene disruption (Figs. 2C & 3C). Altogether, these results suggest that CHIP alleviates cellular injury and apoptosis in cisplatin-treated BUMPT cells.

Fig. 2: CHIP overexpression alleviated cisplatin-induced BUMPT cell injury.
figure 2

A Light microscopy showed that cisplatin-induced cellular shrinkage, which was ameliorated by the CHIP overexpression. TUNEL staining revealed that CHIP overexpression attenuated cisplatin-induced DNA damage in BUMPT cells. DHE and DCF staining demonstrated that cisplatin-induced increase in ROS generation was mitigated by CHIP overexpression. B Western blot studies related to CHIP and FLAG confirmed lentivirus-mediated overexpression of CHIP in BUMPT cells. C Western blotting studies showed that CHIP overexpression attenuated cisplatin-induced upregulation of cleaved-caspase 3 and BAX and downregulation of BCL2 in BUMPT cells. D Quantification of TUNEL staining. (n = 3 *P = 0.0002 vs. the Lv-NC group.) E Quantification of DCF staining by flow cytometry. (n = 3 *P = 0.0117 vs. the Lv-NC group.) F MTT assay revealed that CHIP overexpression increased the cell survival in cisplatin-treated BUMPT cells. (n = 5 *P < 0.0001 vs. the Lv-NC group.) WT: wide type, CP: cisplatin, Lv-NC: empty vector lentivirus for CHIP overexpression, Lv-CHIP: overexpression of CHIP by lentivirus. Scale bars: 100 μm, Error bars show SD.

Fig. 3: CHIP knockdown aggravated cisplatin-induced BUMPT cell injuries.
figure 3

A Light microscopy showed that CHIP knockdown accentuated cisplatin-induced cellular morphological shrinkage. TUNEL staining indicated that CHIP knockdown enhanced cisplatin-induced DNA damage in BUMPT cells. DCF and DHE staining demonstrated that CHIP knockdown accentuated cisplatin-induced increase in ROS generation. B Western blot studies demonstrated lentivirus-mediated downregulation of CHIP in BUMPT cells. C Western blot studies showed that CHIP knockdown accentuated cisplatin-induced upregulation of cleaved caspase 3 and BAX and downregulation of BCL2 in BUMPT cells. D Quantification of TUNEL staining. (n = 3 *P = 0.0002 vs. the shNC group.) E Quantification of DCF staining by flow cytometry. (n = 3 *P = 0.0069 vs. the shNC group.) F MTT assay revealed that CHIP knockdown exacerbated cisplatin-induced BUMPT cell death. (n = 5 *P < 0.0001 vs. the shNC group.) WT: wide type, CP: cisplatin, shNC: empty vector lentivirus for CHIP knockdown, shCHIP: knockdown of CHIP by lentivirus. Scale bars: 100 μm, Error bars show SD.

CHIP overexpression alleviated, while its gene disruption worsened, cisplatin-induced AKI in mice

Renal intra-parenchymal injection of adenovirus was used to modulate the expression of CHIP in kidneys (Figs. 4F & 5F). Immunofluorescence staining of GFP confirmed the successful delivery of adenovirus into renal parenchyma (Supplemental Fig. 3). H & E staining revealed that severe morphological alterations were induced by cisplatin treatment, which were mitigated by CHIP overexpression but were worsened by CHIP knockdown (Figs. 4A & 5S). DHE staining indicated that CHIP overexpression decreased cisplatin-induced ROS generation in renal tubules, while CHIP gene disruption enhanced their generation (Figs. 4A & 5A). Besides, cisplatin induced deterioration of renal functions, as indicated by increased levels of serum creatinine and BUN. While the accentuated specific tubular target injury was indicated by increased levels of KIM1 and NGAL, and these levels were modulated by the expression profile of CHIP (Figs. 4B–E & 5B–E). Finally, the immuno-blotting studies revealed a decreased expression of BCL2 and an increased expression of cleaved caspase 3 and BAX in cisplatin-treated kidneys, and these changes were attenuated by CHIP overexpression but accentuated by CHIP knockdown (Figs. 4G & 5G). Taken together, the findings of these experiments indicate that CHIP suppresses apoptosis to alleviate cisplatin-induced AKI in mice kidneys.

Fig. 4: CHIP overexpression mitigated cisplatin-induced AKI.
figure 4

A H & E staining revealed that CHIP overexpression mitigated cisplatin-induced morphological changes (including tubular dilatation, cast formation, and interstitial edema) in mouse kidneys. DHE staining showed that CHIP overexpression alleviated cisplatin-induced increased ROS generation in renal tubules. B CHIP overexpression alleviated cisplatin-induced increase in serum creatinine levels. (n = 5 *P < 0.0001 vs. the Ad-NC group; *P = 0.043 vs. the Ad-CHIP group; #P < 0.0001 vs. the Ad-NC + CP group.) C CHIP overexpression alleviated cisplatin-induced increase in serum BUN levels. (n = 5 *P = 0.028 vs. the Ad-NC group; *P < 0.0001 vs. the Ad-CHIP group; #P = 0.0031 vs. the Ad-NC + CP group.) D qRT-PCR data showed that cisplatin-induced increase in the mRNA levels of KIM1 was mitigated by CHIP overexpression. (n = 4 *P < 0.0001 vs. the Ad-NC group; *P = 0.0009 vs. the Ad-CHIP group; #P < 0.0001 vs. the Ad-NC + CP group.) E qRT-PCR data showed that cisplatin-induced increase in the mRNA levels of NGAL was mitigated by CHIP overexpression. (n = 4 *P < 0.0001 vs. the Ad-NC group; *P < 0.0001 vs. the Ad-CHIP group; #P < 0.0001 vs. the Ad-NC + CP group.) F Western blot studies confirmed that CHIP was overexpressed in Ad-CHIP adenovirus-treated kidneys. G Western blot studies revealed that CHIP overexpression attenuated cisplatin-induced upregulation of cleaved caspase 3 and BAX, and downregulation of BCL2 in kidneys. CP: cisplatin, Ad-NC: empty vector for CHIP overexpression, Ad-CHIP: overexpression of CHIP by adenovirus. Scale bars: 100 μm, Error bars show SD.

Fig. 5: CHIP knockdown accentuated cisplatin-induced AKI.
figure 5

A H & E staining revealed that CHIP knockdown exacerbated cisplatin-induced renal morphological changes (including tubular dilatation, cast formation and interstitial edema) in mouse kidneys. DHE staining showed that CHIP knockdown accentuated cisplatin-induced increased ROS generation in renal tubules. B Cisplatin-induced increase in serum creatinine levels were further accentuated by CHIP gene disruption. (n = 5 *P = 0.0003 vs. the Ad-shNC group; *P < 0.0001 vs. the Ad-shCHIP group; #P = 0.0001 vs. the Ad-shNC+CP group.) C Cisplatin-induced increase in serum BUN levels were further accentuated by CHIP gene disruption. (n = 5 *P < 0.0001 vs. the Ad-shNC group; *P < 0.0001 vs. the Ad-shCHIP group; #P < 0.0001 vs. the Ad-shNC+CP group.) D CHIP knockdown accentuated cisplatin-induced increase of KIM1 mRNA levels, as validated by qRT-PCR studies. (n = 4 *P < 0.0001 vs. the Ad-shNC group; *P < 0.0001 vs. the Ad-shCHIP group; #P < 0.0001 vs. the Ad-shNC+CP group.) E CHIP knockdown accentuated cisplatin-induced increase of NGAL mRNA levels, as validated by qRT-PCR studies. (n = 4 *P = 0.0098 vs. the Ad-shNC group; *P < 0.0001 vs. the Ad-shCHIP group; #P = 0.0037 vs. the Ad-shNC+CP group.) F Western blot studies showed that CHIP was downregulated by Ad-shCHIP adenovirus in mouse kidneys. G CHIP knockdown accentuated cisplatin-induced upregulation of cleaved caspase 3 and BAX and downregulation of BCL2 in kidneys. CP: cisplatin, Ad-shNC: empty vector for CHIP knockdown, Ad-shCHIP: knockdown of CHIP by adenovirus. Scale bars: 100 μm, Error bars show SD.

Pharmacological activation of CHIP attenuated cisplatin-induced AKI in mice

YL-109, an activator of CHIP, was used to induce CHIP expression in BUMPT cells (Fig. 6B). The morphological changes and the DNA damage in cisplatin-treated cells were attenuated by YL-109 treatment (Fig. 6A & E). The immuno-blotting studies revealed that cisplatin-induced aberrant changes in the expression of apoptosis markers (cleaved caspase 3, BAX and BCL2) were partially mitigated by YL-109 treatment (Fig. 6C). MTT assay showed that cisplatin-induced cell death was also mitigated by the administration of YL-109 in BUMPT cells (Fig. 6D). YL-109 (15 mg kg1, every 2 days) was also used in in vivo studies where the treatment of Yl-109 led to an upregulation of CHIP in mice kidneys (Fig. 6F). The morphological changes (assessed by light microscopy), renal damage (assessed by functional parameters, i.e., serum creatinine and BUN levels) and specific tubular injuries (assessed by KIM-1 and NGAL levels) were partially attenuated by the administration of YL-109 to mice treated with cisplatin (Fig. 6A & H–K). In addition, cisplatin-induced changes in the apoptosis-related markers in kidneys were also alleviated by YL-109 treatment, as indicated by immuno-blotting studies (Fig. 6G). Taken together, these data indicate pharmacological activation of CHIP by YL-109 may be effective in ameliorating cisplatin-induced AKI.

Fig. 6: YL-109 alleviated cisplatin-induced AKI.
figure 6

A Light microscopic analysis showed that YL-109 alleviated cisplatin-induced cellular shrinkage. TUNEL staining revealed that YL-109 mitigated cisplatin-induced DNA damage in BUMPT cells. H & E staining demonstrated that YL-109 attenuated cisplatin-induced morphological changes in mouse kidneys. B Western blot studies demonstrated the upregulation of CHIP by YL-109 in BUMPT cells. C Cisplatin-induced upregulation of cleaved caspase 3 and BAX and downregulation of BCL2 was attenuated by YL-109 treatment in BUMPT cells, as indicated by Western blot procedures. D MTT analysis showed that YL-109 treatment alleviated cisplatin-induced cell death in BUMPT cells. (n = 6 *P = 0.0002 vs. WT group.) E Quantification of TUNEL staining. F Western blotting studies showed that CHIP was upregulated in YL-109-treated kidneys. G YL-109 attenuated cisplatin-induced upregulation of cleaved caspase 3 and BAX and downregulation of BCL2 in kidneys, as indicated by Western blot procedures. H YL-109 alleviated cisplatin-induced increase in serum creatinine levels; (n = 5 *P < 0.0001 vs. WT group; *P = 0.0123 vs. YL-109 group; #P = 0.0017 vs. WT + CP group.) I YL-109 alleviated cisplatin-induced increase in serum BUN levels; (n = 5 *P < 0.0001 vs. WT group; *P < 0.0001 vs. YL-109 group; #P < 0.0001 vs. WT + CP group.) J qRT-RT studies showed that the increased KIM1 mRNA levels induced by cisplatin were partially reduced by YL-109; (n = 4 *P < 0.0001 vs. WT group; *P < 0.0001 vs. YL-109 group; #P = 0.0017 vs. WT + CP group.) K qRT-RT studies showed that the increased NGAL mRNA levels induced by cisplatin were partially reduced by YL-109; (n = 4 *P < 0.0001 vs. WT group; *P < 0.0001 vs. YL-109 group; #P < 0.0001 vs. WT + CP group.) WT: wide type, CON: control, CP: cisplatin. Scale bars: 100 μm, Error bars show SD.

CHIP interacted with NUR77 to trigger its ubiquitination and proteasomal degradation

String analysis indicated that CHIP has the potential to bind with NUR77, and this was validated by CO-IP analysis (Fig. 7A, B, and Supplemental Fig. 4). Since CHIP is an E3 ligase, it is conceivable that CHIP may promote the proteasomal degradation of NUR77. With this notion in perspective, immunoblot studies revealed that NUR77-HA expression was negatively correlated with the transfected dose of CHIP-FLAG plasmid in 293t cells (Fig. 7C). A similar pattern was observed in BUMPT cells with CHIP overexpression or following its gene disruption (Supplemental Fig. 5). Furthermore, CHIP overexpression accelerated the time-dependent degradation of NUR77 in cycloheximide (CHX, an inhibitor for protein synthesis)-treated cells (Fig. 7D). Besides, co-transfection of NUR77-HA plasmid and CHIP-FLAG plasmid led to an increased binding between ubiquitin and NUR77 when MG132 was used to inhibit the proteasomal degradation (Fig. 7E). The types of ubiquitination were then investigated. Our CO-IP studies showed that increased extent of ubiquitination in NUR77 was observed when ubiquitin plasmids or K48 ubiquitin plasmids, not K63 ubiquitin plasmids, were applied, suggesting that K48 ubiquitination was involved in the present state (Supplemental Fig. 6). Previous publications indicated that the E3 ligase activity of “human CHIP” relied on its H260 site endowed with enzymatic activity and K30 site that modulated its interaction with HSP7033,34. It is noteworthy that mouse CHIP plasmid used in our study had an additional “glycine” at the 17th amino acid residue, as compared with human CHIP (Supplemental Fig. 7). In this regard, point mutation of H261 and K31 sites of “mouse CHIP plasmid” was carried out. Expectedly, the CO-IP analysis demonstrated that the mutated mouse CHIP (H261Q or K31A) failed to ubiquitinate NUR77 (Fig. 7F). Notably, NUR77 is a nuclear transcription factor, we then investigated the role of NUR77 on the transcription of CHIP. Our qRT-PCR data suggested that NUR77 overexpression failed to drive the expression of CHIP in BUMPT cells, with or without cisplatin treatment (Supplemental Fig. 8). All in all, the data indicate that CHIP facilitates ubiquitination-dependent proteasomal degradation of NUR77.

Fig. 7: CHIP interacted with and ubiquitinated NUR77.
figure 7

A String analysis showed that CHIP may interact with NUR77. B CO-IP analysis revealed that CHIP bound with NUR77 in BUMPT cells. C Western blot studies revealed that CHIP overexpression (FLAG tag) led to a decreased expression of NUR77 (HA tag). D Immunoblot studies showed that CHIP overexpression enhanced NUR77 degradation in a dose-dependent manner when protein synthesis was inhibited by cycloheximide (CHX). E CO-IP analysis revealed that forced expression of CHIP promoted the binding of NUR77 with ubiquitin in MG132-treated 293t cells. F CO-IP studies showed that H261Q or K31A mutation in CHIP failed to ubiquitinate NUR77. G CHIP was truncated at TPR domain, ΔTPR domain and U-box domain. CO-IP analysis revealed that only ΔTPR domain of CHIP co-precipitated with NUR77. H NUR77 was truncated at TAD domain, ΔTAD domain and LBD domain. The CO-IP data showed that ΔTAD domain and LBD domain of NUR77 co-precipitated with CHIP. ΔTPR: a truncated protein containing CC domain and U-box domain; TAD: transactivation domain; DBD: DNA binding domain; LBD: ligand-binding domain; ΔTAD: a truncated protein containing DBD domain and LBD domain; Lv-NC: empty vector lentivirus for CHIP overexpression; Lv-CHIP: overexpression of CHIP by lentivirus.

In order to investigate the exact nature of interaction, CHIP and NUR77 were truncated as indicated in Fig. 7G, H, and the CO-IP analyses revealed that the NUR77-HA signal (immunoblot band) was only detected in ΔTPR immunoprecipitates, and CHIP-FLAG signal (immunoblot bands) were detected in ΔTAD and LBD immunoprecipitates (Fig. 7G, H), suggesting that CC domain of CHIP and LBD domain of NUR77 are involved in the interaction. Then, we proceeded to investigate the protein-protein docking prediction to explore the interactive site. In this regard, the prediction indicated that E179, R195, and Q197 are the sites of CHIP that has multiple potential hydrogen bonds with NUR77 (Supplemental Fig. 9). These sites were mutated for further CO-IP analysis. Surprisingly, the CO-IP data indicated that alternative or combined mutation in these sites failed to disrupt the binding between CHIP and NUR77 (Supplemental Figs. 10A–C & G). Two other potential binding sites of NUR77 (R494 and R557) were also selected and mutated. However, the mutated NUR77 (R494A, R557A or both) still interacted with CHIP (Supplemental Figs. 10D–F). Taken these data together, we can surmise that CC domain of CHIP binds with LBD domain of NUR77, but their precise binding site still remains to be determined.

CHIP inhibited NUR77-mediated intrinsic apoptosis

The changes in the expression of NUR77 and the status of intrinsic apoptosis-related markers were investigated. Western blot studies showed that NUR77 was upregulated in cisplatin-treated BUMPT cells and kidneys (Fig. 8A, B). Previous publications had indicated that NUR77 exerted an anti-apoptotic effect when localized in the nuclei, while it exhibited pro-apoptotic potential when localized in the cytoplasm35. The immunofluorescence studies indicated that NUR77 was localized in the cytoplasm in cisplatin-treated cells (Fig. 8B). Interestingly, some of the investigators have reported that cytoplasmic NUR77 worsens AKI via its interaction with BCL2 in the mitochondria, leading to the exacerbation of intrinsic apoptosis30. In this regard, the localization of NUR77 in the mitochondria may indicate the induction of intrinsic apoptosis. The data of this investigation indicated that cisplatin-induced translocation into the mitochondria (Mitotracker-red in BUMPT cells and TOMM20-red in renal tissue sections) of NUR77 was modulated by the expression profile of CHIP (Fig. 8C–F). In view of these observations, mitochondrial extracts and mitochondria-deficient extracts were prepared from BUMPT cells, and immunoblotting studies were performed. No expression of NUR77 was observed in the mitochondria of the untreated cells, and its expression increased following cisplatin treatment (Fig. 8G). It is noteworthy that the expression of NUR77 in the mitochondrial and mitochondria-deficient extracts was modulated by the expression profile of CHIP (Fig. 8G, H). The expression of CHIP in the mitochondria was barely observed (Supplemental Fig. 11), suggesting that CHIP degraded NUR77 before its entry into the mitochondria. BAX was upregulated and BCL2 was downregulated in the mitochondrial extracts and mitochondria-deficient extracts of cisplatin-treated BUMPT cells, and these changes were modulated by CHIP expression (Fig. 8G, H). Taken together, one can speculate that CHIP conceivably inhibits the translocation of NUR77 into the mitochondria, thus leading to the amelioration of cisplatin-induced intrinsic apoptosis.

Fig. 8: CHIP alleviated NUR77-induced intrinsic apoptosis.
figure 8

A Western blot studies revealed that NUR77 was upregulated in cisplatin-treated BUMPT cells and mouse kidneys. B Immunofluorescence staining showed that NUR77 was upregulated in cisplatin-treated BUMPT cells, and a notable increase in fluorescence intensity was observed in the cytoplasm following cisplatin treatment. C, D Immunofluorescence staining revealed that the co-localization (yellow) between NUR77 (green) and mitochondria (red, stained with Mitotracker) was reduced by CHIP overexpression but increased by CHIP gene disruption in cisplatin-treated BUMPT cells. E, F Immunofluorescence studies demonstrated that the colocalization (yellow) of NUR77 (green) and mitochondria (red, stained with TOMM20) was modulated by the expression profile of CHIP in cisplatin-treated mouse kidneys. G Western blot studies showed that no expression of NUR77 was noted in the mitochondria of the untreated cells, and cisplatin-treatment led to notable expression of NUR77 in the mitochondria, which was modulated by the expression profile of CHIP. Besides, cisplatin-induced downregulation of BCL2 and upregulation of BAX in the mitochondria were attenuated by CHIP overexpression but accentuated by CHIP knockdown. H Cisplatin treatment led to the downregulation of BCL2 and the upregulation of NUR77 and BAX in mitochondria-deficient extracts, which was mitigated by CHIP overexpression but aggravated by CHIP knockdown. Lv-NC: empty vector lentivirus for CHIP overexpression; Lv-CHIP: overexpression of CHIP by lentivirus; shNC: empty vector lentivirus for CHIP knockdown; shCHIP: knockdown of CHIP by lentivirus.

Discussion

Cisplatin, a commonly used anti-cancer drug, is also a classic nephrotoxin for the induction of AKI in laboratory investigations. DNA damage was widely accepted as the major cause of cisplatin-induced AKI since cisplatin crosslinks with DNA strands in the nuclei5. However, this notion has been challenged, since only ~1% of cisplatin accumulates in cellular nuclei36. Various mechanisms, including mitochondrial damage, autophagy, ferroptosis, and apoptosis, have been proposed in the pathogenesis of cisplatin-induced AKI, and among which, intrinsic apoptosis has drawn the most attention4,5,37,38. Recent literature data indicate that overexpression of CHIP, a U-box E3 ligase, attenuates various pathological perturbations in several organ systems, including heart, brain and liver20,21,39. Eight types of U-box E3 ligase have been identified in human tissues32, and CHIP has the highest expression in renal proximal tubules, and latter are the main target of cisplatin (Fig. 1). Previous publications indicated that CHIP participated in renal water handling by ubiquitinating aquaporin-224, and it also exhibited anti-oxidative effect by promoting proteasomal degradation of NOX4 in Angiotesion II-treated proximal tubular cells25. Besides, CHIP enhanced cisplatin resistance of prostate cancer cells by degrading Mammalian sterile 20–like kinase 1 (Mst1)23. Similarly, our data revealed that CHIP overexpression or activation attenuated, while its gene disruption accentuated, cisplatin-induced intrinsic apoptosis and oxidative stress in renal tubular cells (Figs. 26). Interestingly, during the preparation of our manuscript, a related publication by Yanting et al. appeared, which described the beneficial effect of CHIP in cisplatin-treated HK2 cells but with somewhat dissimilar changes in CHIP expression40. Our data suggest that CHIP is down-regulated, while Yanting et al. noted that it is up-regulated, in cisplatin-induced AKI40. These differing results may be related to the differences in the administration of cisplatin in mice (25 mg/kg for two days vs 20 mg/kg for three days) and usage of cell lines (BUMPT cells vs HK2 cells). Nevertheless, these data suggest that CHIP decelerates intrinsic apoptosis to ameliorate cisplatin-induced AKI.

The next question that needed to be addressed was how CHIP attenuates oxidative stress and intrinsic apoptosis in cisplatin-induced AKI. Our initial work revealed that CHIP interacts with NUR77 in renal proximal tubular cells (Fig. 7). NUR77 is a unique orphan nuclear receptor, which can orchestrate cellular responses to various stimuli41,42. It is noteworthy that the phenotype of NUR77 in renal injuries is context-dependent. In chronic states, NRU77 presented as a beneficial participator, and it could alleviate obstructive, senile, or hypertensive renal fibrosis27,28,29. In comparison, one publication showed that genetic ablation or pharmaceutical inhibition of NUR77 attenuated renal injury in ischemia/reperfusion-induced AKI, and its detrimental effect was related to its interaction with BCL2, which led to the exposure of BH3 of BCL230. It has been well-documented that CHIP, a chaperone-associated U-box E3 ligase, interacted with HSP70 to recruit and degrade its target proteins to maintain intracellular homeostasis33,34,43. Along these lines, we speculated CHIP might degrade NUR77, a protein detrimental to the homeostasis of kidney, to exert its beneficial effect. Expectedly, the results of our experiments indicated that CHIP caused proteasomal degradation of NUR77 (Fig. 7). Of note, previous publications have indicated that the E3 ligase activity of CHIP (human) rely on the H260 site of its U-box domain, while HSP70 interacted with CHIP (human) at the K30 site to recruit its substrates33,34. Our data revealed that the mutated CHIP at these sites failed to ubiquitinate NUR77, suggesting that the predominant modulatory effect of CHIP on NUR77 was dependent on HSP70 and the enzymatic activity of its U-box domain (Fig. 7). Collectively, our data indicate that CHIP interacts with and degrades NUR77 to alleviate cisplatin-induced AKI.

The protein:protein interactive pattern between CHIP and NUR77 was also investigated in the current study. CHIP, a highly conservative protein, contains a tetratricopeptide-repeat (TPR) domain at the N-terminal end, a CC domain in the center, and a U-box domain at its C-terminal end44. Previous publications revealed that CHIP preferentially bound to the substrates via its U-box or TPR domain45,46. However, our present data revealed that CHIP interacted with the LBD domain of NUR77 via its CC domain (Fig. 7), a rarely observed pattern as previously reported for its interaction with OTUD347. In order to explore the exact binding site of the two proteins, bioinformatic predication was employed, and three highly potential binding sites (E179, R195, and Q197) in CHIP were selected for mutation. Surprisingly, point mutations in these sites failed to disrupt their interaction (Supplemental Fig. 10). One possible explanation could be that CHIP forms homodimers prior to the interaction with its substrates, as previously reported by Brenda48. In this regard, the exogenous mutated CHIP might dimerize with endogenous CHIP, and this unique type of “heterodimerization” would likely preserve the capacity to bind with NUR77. Therefore, two potential binding sites in NUR77 were selected and mutated, and CO-IP analysis showed that the mutated NUR77 (R494A and R557A) could still bind with CHIP, suggesting that none of the predicted sites were predominately involved in their interaction (Supplemental Fig. 10). Taken together, we reckon with the idea that CC domain of CHIP interacts with LBD domain of NUR77, while their precise binding site needs to be yet explored in future investigations.

Another noteworthy aspect of NUR77 is that its phenotypic characteristics depend upon the subcellular localization.35. When present in the nuclei, NUR77 operates as an inducible transcription factor to exhibit anti-apoptotic and pro-survival effects35,49. While, NUR77 exerts pro-apoptotic effects when it is translocated into the mitochondria. NUR77 disables BCL2 in the mitochondria to disrupt the outer mitochondrial membrane, leading to the generation of ROS and apoptosis50. Our data indicated that NUR77 was strongly up-regulated in cisplatin-induced AKI (Fig. 8), similar to the observations made in renal ischemia/reperfusion injury30. Interestingly, NUR77 was only observed in the nuclei in untreated cells, and its expression in the mitochondria was conceivably induced in cisplatin-treated cells (Fig. 8). Besides, the expression of NUR77 and intrinsic apoptosis-related markers in the mitochondrial and the mitochondria-deficient extracts were stringently modulated by the expression profile of CHIP (Fig. 8). Taken these data together, one can conclude that CHIP inhibits NUR77-mediated intrinsic apoptosis to shield against cisplatin-induced AKI.

In summary, this study provides new insights into the critical role of CHIP in cisplatin-induced AKI and its relevance with NUR77-mediated renal tubular injury. These findings highlight novel aspects of mechanism for the progression of cisplatin-induced AKI.

Materials & methods

Cell culture studies

BUMPT cells, a mouse proximal tubular cell line, was initially obtained from Drs. William Lieberthal and John Shwartz at Boston University. Cells were cultured in high glucose DMEM supplemented with 10% FBS and antibiotics, and they were maintained at 37 °C in a humidified atmosphere of 5% CO2. Various lentiviruses (purchased from Vigene Biosciences), including CHIP overexpression lentivirus (Lv-CHIP), NUR77 overexpression lentivirus (Lv-NUR77), CHIP knockdown lentivirus (shCHIP) and empty vector lentivirus (Lv-NC and shNC), were employed to transfect the cells. Puromycin (3 μg ml1) was used to generate stably transfected cell lines following 24 h treatment. The cells were seeded into 96-well plates or 6-well plates containing culture medium supplemented with 2% FBS. The cells were treated with 20 μM cisplatin (Sigma-Aldrich, catalog P4394) for 16–24 h. They were then harvested for various studies. Polyethylenimine was used for transfection of 293t cells. Cyclohexamide (50 μg ml1, MCE, catalog HY-12320) was used to treat the cells for 2–8 h, while the treatment with 15 μM of MG132 (MCE, catalog HY-13259) was carried out for 2–12 h, as indicated previously 51,52. For CHIP activation, 2 μM YL-109 (MCE, catalog HY-18619) was used to pre-treat the BUMPT cells for 8 h as indicated before53, and they were then co-treated with cisplatin.

Animal experiments

All experiments were performed by following the guidelines of the Animal Ethical and Welfare Committee of the Second Xiangya hospital at Central South University (Approval number: 2022620). Male C57BL/6 J mice (8–10 weeks old) were used, and they were grouped into various groups (6 mice in each group). Cisplatin was intraperitoneally administered to mice at the dose of 25 mg kg1 to induced AKI. Two days later, kidneys and blood samples were then harvested. In some studies, YL-109 (MCE, catalog HY-18619) was subcutaneously injected in the scruff of the neck (15 mg kg1, every 2 days) three times before cisplatin injection, as indicated in previous publications53,54.

For the modulation of CHIP expression in the kidneys, adenovirus (Vigene Biosciences) was used in in vivo studies. Mice were anesthetized with pentobarbital sodium, and their dorsum of the flank was shaved. The left kidney was exposed with a dorsal incision, and the renal pedicle was clamped. Following clamping, a total of 100 μl adenovirus solution (at the concentration of 1.5–2 × 1012 particles ml1) was injected in four different regions of the kidney (25 μl at each site) with a 31-gauge needle. The clamp was then removed from the renal pedicle 5 min after the injection. The adenoviral vectors used in the present studies included: CHIP overexpressing adenovirus (Ad-CHIP), CHIP knockdown adenovirus (Ad-shCHIP), and empty adenovirus vector (Ad-NC and Ad-shNC). Two days after the administration of adenoviral vectors, the mice were injected with cisplatin. All animal experiments in this article were approved by the Animal Research Institute of the Second Xiangya Hospital at Central South University (Approval number: 2022620). We have complied with all relevant ethical regulations for animal use.

RPKM value analysis

For the initial evaluation of expression of the U-box E3, the RPKM values were obtained from https://www.ncbi.nlm.nih.gov/mesh/. Gene need to be selected, and the gene name can be entered to search values of the gene in different species.

Western blotting procedures

The protocol for immunoblotting has been described in ref. 4. Briefly, cells or renal cortices were homogenized in RIPA buffer, and the protein concentration was measured by BCA assay (Biyotime, catalog P0012S). Equal amounts of proteins were then subjected to SDS-PAGE, and the fractionated proteins were electroblotted onto the PVDF membranes. The membranes were immersed in 5% milk solution, and then incubated with diluted primary antibody overnight at 4 °C. The membranes were washed and incubated with secondary diluted antibodies at 22 °C for 1 hr. Finally, the membranes were subjected to ECL chemiluminescence, and the signal was detected by exposing the membranes to X-ray film. The antibodies used in these studies included: anti–CHIP (Abcam, catalog ab134064; 1:2,000), anti–β-actin (proteintech, catalog 66009-1-Ig; 1:5,000) and anti-GAPDH (proteintech, catalog 60004-1-Ig; 1:5,000), anti-tubulin (proteintech, catalog 66031-1-Ig; 1:5,000), anti-BAX (proteintech, catalog 60267-1-Ig; 1:2,000), anti-cleaved caspase3 (CST, catalog 9664; 1:1,000), anti-BCL2 (proteintech, catalog 26593-1-AP; 1:1,000), anti-FLAG (Sigma, catalog F1804; 1:5,000), anti-HA (proteintech, catalog 51064-2-AP; 1:5,000, and Santa Cruz, catalog sc-7392; 1:5,000), anti-NUR77 (proteintech, catalog 25851-1-AP; 1:1,000), anti-ubiquitin (CST, catalog 3936; 1:1,000), anti-MYC (proteintech, catalog 60003-2-Ig; 1:2000) and anti-COXIV (proteintech, catalog 11242-1-AP; 1:500).

qRT-PCR analysis

The mRNA levels were detected by qRT-PCR, as previously described55. Briefly, the cells or renal cortex were homogenized with RNA Trizol (Invitrogen, catalog 15-596-026), and the mRNA was isolated. After removal of the contaminating DNA, ~1 μg mRNA was used for reverse transcription with PrimeScriptTM RT reagent kit (Takara). The cDNA samples thus generated were diluted with an equal volume of water and subjected to an ABI PRISM 7900 Sequence Detector System (Applied Biosystems) with ChamQTM Universal SYBR® qPCR Master Mix (Vazyme). For the quantitative analysis in Fig. 1, the relative expression of the genes was obtained using ΔCt values, and their relative expression were compared with the mean value of CHIP to reveal their differences in expression. The quantitative analysis were calculated with the ΔΔCt method. The primers used in our studies were given in Supplemental table 1.

Immunohistochemistry methods

4 μm thick paraffin embedded tissue sections were used for carrying out immuno-histochemical procedures4. Briefly, the sections were deparaffinized, rehydrated, and rinsed with PBS. Following which, the tissue sections were subjected to antigen retrieval. After blocking of endogenous peroxidase, the sections were immersed in goat serum for 1 h at 22 °C. The tissue sections were then immersed in diluted primary CHIP antibody (Abcam, catalog ab134064; 1:200) solution and kept them overnight at 4 °C. After a brief rinse with PBS, they were incubated with diluted secondary antibody solution for 1 h at 22 °C. Subsequently, the sections were treated with DAB solution, followed by staining with hematoxylin for 30 s. The sections were re-rinsed with PBS, dehydrated and coverslip mounted and evaluated by light microscopy.

Immunofluorescence staining methods

For in vitro studies, the cells were seeded on the coverslips prior to cisplatin treatment. They were washed, fixed with 4% paraformaldehyde, and permeabilized. For in vivo studies, ~4 μm paraffin embedded thick sections were deparaffinized, rehydrated, permeabilized, and subjected to antigen-retrieval. The sections were immersed in 5% BSA solution for 1 h at 22 °C. They were then incubated with diluted primary antibody solution overnight at 4 °C. The sections were rinsed with PBS and incubated with diluted secondary antibody solution for 1-2 h at 22 °C. The sections were stained with DAPI to visualize the nuclei by fluorescence microscopy. The antibodies used in these studies were as follows: anti–CHIP (Abcam, catalog ab134064; 1:1,000), anti–TOMM20 (Abcam, catalog ab565783; 1:500), anti–MEGALIN (Abcam, catalog ab 184676; 1:500), anti–NUR77 (proteintech, catalog 25851-1-AP; 1:200), and anti-GFP (proteintech, catalog 66002-1-Ig; 1:200).

Morphological analysis

Cellular morphology was evaluated by light microscopy immediately after cisplatin treatment and following H & E staining. Briefly, 4 μm paraffin embedded thick sections were deparaffinized and rehydrated. They were then treated with Hematoxylin for 3 min and eosin for 30 s. The sections were dehydrated and coverslip mounted for microscopic evaluation.

Renal functional analysis

Renal functions were assessed by measuring serum creatinine and blood urea nitrogen (BUN). Briefly, the blood samples were kept at 22 °C for ~ 20 min; following which they were centrifuged to collect the clear serum. Serum creatinine kit (Jiancheng, China, catalog C011-2-1) and BUN kit (Jiancheng, China, catalog C013-2-1) were used to measure the blood concentrations of creatinine and BUN, per the manufacturer’s instructions.

Preparation of mitochondrial and mitochondria-deficient cellular extracts

For the isolation of mitochondria from BUMPT cells, a special kit (Proteintech, catalog PK10016) was used, as per vendor’s instructions. Briefly, cells were washed with cold PBS solution after various treatments and then collected in a centrifuge tube. Mitochondria Isolation Solution A ( ~ 1 ml for 20 million cells) was added, and the cells were homogenized at 4 °C. The same amount of Mitochondrial Isolation Solution B was added in another tube. The cellular homogenate was then carefully added into 2nd tube. The resulting mixture was centrifuged at 600 g for 10 min, and the supernatant was collected. The supernatant was then centrifuged at 10,000 g for 10 min. Finally, the mitochondria at the bottom of the tube were collected and designated as mitochondrial extracts while the supernatant was designated as mitochondria-deficient extracts. The mitochondrial extracts were lysed in the Mitochondrial Lysis Buffer, and the lysate was processed for immuno-blotting procedures.

Cell death evaluation

MTT assay was used to detect cell viability. Briefly, 3,000–5,000 cells were seeded into each well of 96-well plates. After cisplatin treatment, the culture medium was removed, and 100 μl MTT mixture (cell culture medium with 1 mg ml1 Thiazolyl Blue Tetrazolium Bromide) was added into the wells. The cells were then incubated in this mixture for 4 h. Afterwards, the mixture was removed and 150 μl DMSO was added into the wells to dissolve the formazan. The plates were then gently shaken for 5 min on an orbital shaker, and then readings were recorded using a microplate reader at an OD of 495 nm. The background readings were subtracted. The mean values of the relative data of untreated groups were also calculated. Finally, the values of cisplatin-treated groups were divided by the mean values of their related-untreated groups, and thus the relative cell survival rate was calculated. For studies related to apoptosis, TUNEL assay Kit (UElandy, catalog T6014L) was used to evaluate the extent of cellular DNA damage, per the manufacturer’s instructions. The tissue sections were photographed with a fluorescent microscope equipped with UV illumination, and the quantitation of cellular damage was assessed.

Evaluation of intracellular ROS

Dihydroethidium (DHE) and 2′,7′-dichlorodihydrofluorescein diacetate (DCF) were used to assess the intracellular ROS in cells and kidney tissues. For DHE staining, BUMPT cells or cryosections of kidney tissues were incubated with 10 μM DHE at 37 °C for 30 min and then evaluated by fluorescence microscopy. For DCF staining, BUMPT cells were incubated with 10 μM DCF for 60 min at 37 °C, and the sections were then subjected to microscopic evaluation, and flow cytometry was used to evaluate the fluorescence intensity.

Evaluation of protein-protein interaction

The interaction between CHIP and NUR77 was initially predicted by String analysis. The website of String analysis is as follows: STRING: functional protein association networks (string-db.org). Co-immunoprecipitation (CO-IP) assay was used to confirm their interaction, as previously described in ref. 55. Briefly, the cells were homogenized with CO-IP lysis buffer (150 mM NaCl, 50 mM Tris-HCl, pH 8.0, 5% glycerol, 1.0% NP40, and 1 mM MgCl2), supplemented with protease inhibitors. The protein concentration was measured, and 500–1000 μg protein in a total volume of 250–350 μl was used for these studies. The primary antibodies were added into the cell lysate and gently mixed using an orbital shaker at 4 °C. 25 μl protein A/G beads (Santa-cruz) were added into the lysate mixtures, and they were gently shaken for another 4 h. The beads were washed, centrifuged, and boiled with the SDS-PAGE loading buffer. The final mixtures were then subjected to SDS-PAGE analysis. The antibodies used in these studies included: anti-FLAG (Sigma, catalog F1804; 1:5,000), anti-MYC (proteintech, catalog 60003-2-Ig; 1:2,000) and anti-HA (proteintech, catalog 51064-2-AP and Santa, catalog sc-7392; 1:1,000).

Protein-protein docking prediction

The protein sequence of CHIP (PDB, PDB sequence number: 2C2L) and NUR77 (PDB, PDB sequence number: 2QW4) were obtained from the Protein Data Bank. Subsequently, Cluspro was used to generate their interaction model56,57,58,59,60, and the potential protein-protein docking sites were predicted by using LigPlus software. Finally, the biomolecular structures of the interaction were acquired from PyMOL.

Plasmid construction

CHIP and NUR77 CDS nucleotide sequence were obtained from Miaolingbio (Wuhan, China). The CDS sequence was used as the template to duplicate or truncate CHIP and NUR77 with the primers as indicated in Supplemental table 1. The PCR products generated were subcloned into PCDH-3 × FLAG or PCDH-3 × HA plasmid. For point mutation analysis, the primers were designed as indicated in Supplemental table 1. The full-length CHIP plasmid was used as the template, and PCR amplification was carried out. Subsequently, the PCR products were mixed and used as the template for the secondary fusion PCR. The products of fusion PCR were finally subcloned into PCDH-3 × FLAG plasmid or PCDH-3 × HA plasmid.

Statistics and reproducibility

Statistical significances were tested by independent sample t-test and one-way analysis of variance (ANOVA) with the Dunn’s multiple comparison test. P  <  0.05 considered statistically significant. All data were analyzed using GraphPad Prism 9 software. The values were presented as mean ± SD. All experiments were performed at least three times.