Accumulation of ApoJ in renal tubules restricts autophagy and lysosomal function to accelerate renal lipid deposition and EMT
Since ApoJ is sensitive to glucose level17, 31, it was found to be correlated positively with fasting glucose (r=0.357, p=7.85E-05) and HbAlc levels (r=0.391, p=9.67E-09) (Supplementary Table 3), suggesting the role of ApoJ in the progression of T2DM (r=0.392, p=8.78E-09) (Fig. 1a). Notably, increment of serum ApoJ associated with deleterious renal function, including declined eGFR (r=-0.287, p=3.64E-05) and elevating levels of creatinine (r=0.251, p=3.19E-04), BUN (r=0.275, p=3.61E-04), and UACR (r=0.363, p=1.20E-07) (Supplementary Table 1). The result was further supported by increasing level of ApoJ in patients with DN, the single commonest cause of ESKD (24.3±5.8 vs. 30.6±6.5 mg/dl, p=7.10E-07, Fig. 1b). To dissect the role of ApoJ in the development of DKD, the expression of ApoJ in different cell types was first assessed in the kidney using the Human Protein Atlas database. The results showed that proximal tubular cells were the major cell type expressing ApoJ in the human kidney (Fig. S2a), and elevated ApoJ mRNA in renal tubules was found in patients with DN (Fig. S2b). The findings were further confirmed in mouse models of HFD and T2DM, which exhibited elevated serum ApoJ levels (Fig. 1c) and accumulated ApoJ in renal tubules but not in glomeruli (Fig. 1d and Fig. S3). Notably, both serum (r=0.768, p=1.15E-05) and renal ApoJ (r=0.632, p=9.30E-04) correlated positively with tubular injury in mice fed a HFD and mice with diabetes (Fig. 1e). The results indicated the pathogenic role of ApoJ in the progression of DKD.
The proteomic analysis was next performed in proximal tubule epithelial cell HK2 cells to identify ApoJ-related pathogenesis. Compared to parental cells, mTOR was enriched in ApoJ knockdown HK2 cells in the presence of OA (Fig. 1f). In addition, GSEA analysis showed that ApoJ silencing prevented OA-induced epithelial-mesenchymal transition (EMT) and restored fatty acid metabolism (Fig. 1g). In vitro functional evaluations further demonstrated that OA supplementation increased intracellular ApoJ levels and promoted mTOR activation (Fig. 2a and Fig. S4), resulting in impairment of autophagic flux (Fig. 2b), whereas ApoJ silencing reactivated autophagy through downregulation of mTOR (Figs. 2a, 2b, and Fig. S4). Moreover, ApoJ silencing upregulated the expression of the lysosomal vacuolar proton pump ATP6V0D1 (Fig. 2a) and restored lysosomal pH and activity (Figs. 2c-2e and Supplementary video S1). Thus, targeting ApoJ facilitated lysosomal degradation of LDs and ameliorated aberrant lipid in HK2 cells (Figs. 2f, and 2g). In addition, ApoJ silencing rebalances redox homeostasis in OA-treated HK2 cells, indicated by reduction of ROS and MDA accumulation and increasing in both non-enzymatic antioxidant GSH level and enzymatic antioxidant SOD activity (Fig. 2h). The results were confirmed by rescue experiment. The deleterious effects of the ApoJ-mTOR axis on ROS accumulation and EMT were also found in high glucose-treated HK-2 cells (Fig. S5) and in mice fed a high-fat diet (HFD) (Figs. S6). Notably, the circulating ApoJ in medium collected from hepatocytes over-expressed ApoJ was taken up by ApoJ-silenced HK2 cells to persist mTOR activation (Fig. 2i). Altogether, these findings suggested that accumulation of ApoJ disrupts lipid and redox homeostasis in a mTOR-dependent manner and promote EMT in nutrient-overloaded HK2 cells.
ApoJ silencing reduces mTOR/TFEB interaction and facilitates TFEB activation
The crosstalk between the ApoJ-mTOR axis and TFEB activation was next examined. In OA-supplemented HK2 cells, ApoJ silencing disrupts TFEB/mTOR interaction (Fig. 3a) and promotes nuclear translocation of TFEB (Fig. 3b and Fig. S7a) which induces activation of TFEB, indicated by upregulation of downstream gene expression, including CTSD, LAMP1, and UVRAG (Fig. 3c). In addition, antagonizing ApoJ chaperone by a MK53 peptide reduced mTOR-TFEB interaction (Figs. 3d and 3e). Compared with penetrating control peptide, the MK53 peptide promoted TFEB nuclear translocation (Fig. 3f), leading to reactivation of TFEB (Fig. 3g) and relieve aberrant lipid and ROS accumulation in nutrient-overloaded HK2 cells (Fig. 3h, and Fig. S7b). Additionally, the function of the MK53 peptide on the mTOR-TFEB-autophagy axis and LD accumulation was confirmed in mouse primary renal proximal tubule epithelial cells (PTECs, Fig. S8).
Renaloverexpression of ApoJ accelerates DKD progression
To confirm the pathological role of ApoJ in DKD, a mouse model with renal overexpression of ApoJ was generated by intrarenal injection of AAV-ApoJ (Figs. S9a- S9c). As expected, renal overexpression of ApoJ facilitated lipid deposition in the kidneys of db/db mice (Fig. 4a), as well as in in vitro HK-2 cell models (Fig. S9d). Increasing renal ApoJ levels promoted mTOR activation and increased cytosolic TFEB retention (Fig. 4b and Figs. S9e-S9h). Thus, ApoJ exaggerated kidney damages in db/db mice, indicating by the elevating serum BUN and creatinine levels (Fig. 4c), accumulating oxidative stress (Fig. 4d), and advancing renal fibrosis (Figs. 4e and 4f).
Accumulation of circulating ApoJ in the renal tubule sustains diabetes-induced nephropathy
Next, a tubular epithelial cells-specific (TEC-KO) ApoJ knockout mice was generated to elucidate the role of renal ApoJ in DKD progression. TEC-KO abolished the expression the precursor (psApoJ) in the renal tubular, but not the mature form of ApoJ (Fig. S10a). Therefore, obesity-induced renal pathogenesis persisted in TEC-KO mice (Figs. S10b-S10e). Since accumulated ApoJ in renal tubular epithelial cell might originate from the liver (Fig. 2i)20, the role of hepatocyte-secreted ApoJ in diabetic renal injury was therefore examined using hepatocyte-specific (HKO) ApoJ knockout mice (Fig. S11a). As expected, ApoJ HKO with completely abolishing the accumulation of secretory mature ApoJ in the kidney (Fig. S11b) reduced the levels of renal lipids (Figs. S11c), serum BUN, and creatinine (Fig. S11d). Additionally, ApoJ HKO restored mTOR-mediated suppression of autophagy and suppressed the expression of fibrotic markers in the kidneys of HKO-mice fed a HFD (Fig. S11b).
These findings were further examined in STZ-induced diabetic mice (Fig. S12a). Consistently, the levels of ApoJ in renal tubules and serum were comparable between flox control and TEC-KO mice, but barely detected in HKO mice (Figs. 5a-5c, and Fig. S12b), confirming liver is the primary organ for producing secretory ApoJ. The ApoJ putative receptor LRP2, primarily expressed in TEC (Fig. S13), was significantly increased in ApoJ TEC-KO mice, but not in those of HKO (Fig. 5d), suggesting a renal uptake of circulating ApoJ. Physiologically, ApoJ HKO exhibited a lower kidney-to-body ratio (Fig. S12c) and improved redox (Fig. 5e) and lipid homeostasis (Fig. 5f), supported by the downregulation of genes involved in glucose metabolism, cholesterol synthesis, lipogenesis and EMT in the kidneys of mice with HFD/STZ-induced diabetes (Fig. 5g). In addition, ApoJ HKO ameliorated diabetes-induced enlargement of glomerular size and mesangial expansion and prevented glomeruli and tubulointerstitial fibrosis (Fig. 5h, and Figs. S12m-S12o), therefore restored renal function in mice with HFD/STZ-induced diabetes (Fig 5i). Additionally, ApoJ HKO restored the levels of proteins involved in autophagy and lysosomal function (Fig. 5c), indicating reactivation of mTOR inhibited autophagy and lysosome activity in these mice.
MK53 peptide relieves DKD progression by rebalancing renal lipid and redox homeostasis in diabetic mice
Since ApoJ is potential therapeutic target for DKD, a MK53 peptide was next applied to relieve ApoJ-promoted DKD progression (Fig. 6a). Comparable to liraglutide, an FDA-approved antidiabetic peptide, the MK53 peptide improved renal lipid deposition (Fig. 6b) and renal function (Fig. 6c) of HFD-fed mice. Both liragutide and MK53 peptide reduced renal injuries (Figs. 6d-6g), thus attenuated obesity-induced renal fibrosis (Figs. 6d, 6h, and Fig. S14). Consistently, the MK53 peptide restored mTOR inhibited TFEB activity and autophagic flux (Fig. 6h and Fig. S14).
The function of the MK53 peptide on DKD was further examined in mouse model of STZ-induced diabetes (Fig. S15a). Notably, the MK53 peptide, but not liraglutide, restored STZ-induced weight loss (Fig. S15b), suggesting the protective effect of the MK53 peptide. Both treatments prevented renal accumulation of oxidative stress (Fig. 7a) and lipid deposition (Fig. 7b), thus relieving renal injuries (Figs. 7c-7g). In agreement with the results of HFD-treated mice, the MK53 peptide inhibited the mTOR pathway and reactivated lysosomal activity in STZ-induced diabetic mice (Fig. 7h and Figs. S15c-S15j). The results were also confirmed in db/db mice (Fig. 8a), in which the MK53 peptide significantly reduced lipid levels (Fig. 8b), restored renal function (Fig. 8c), and prevented renal injuries (Figs. 8d-8e and Figs. S16a-S16c) by rescuing autophagy and lysosome activity in the kidney (Figs. 8e and Figs. S16d-S16m). Altogether, these findings demonstrated that antagonizing ApoJ improves the progression of DKD by rebalancing renal lipid and redox homeostasis through reactivation of autophagy and lysosomal function.