Male Sprague–Dawley rats (150 $\pm$ 20 g body weight) were obtained from Shanghai Bikai Laboratory Animal Co., Ltd. (Shanghai, China) and maintained under specific-pathogen-free conditions. Following a 7-day acclimatization period, animals were randomly assigned to three experimental groups: normal control (control), diabetic model (DM), and emodin-treated diabetic groups (DM+emodin). To induce type 1 diabetes, 8-week-old rats received a single intraperitoneal dose of streptozotocin (STZ, 60 mg/kg; S0130, Sigma Aldrich, St. Louis, MO, USA) prepared in 0.1 M citrate buffer at pH 4.5, whereas control animals received an equal volume of the vehicle buffer. Successful induction of hyperglycemia was verified 72 h later by fasting blood glucose measurement ($>$16.7 mmol/L). Emodin-treated rats were orally administered 40 mg/kg emodin (S1312, Selleck, Houston, TX, USA) formulated in 0.5% sodium carboxymethyl cellulose (CMC-Na, S6703, Selleck) once daily by gavage; control and diabetic groups received the corresponding volume of vehicle. The dosage and route of emodin administration have been verified safe and effective [19, 20, 21]. Following 8 weeks of treatment, all animals were euthanized for tissue collection. Deep anesthesia was achieved prior to sacrifice through intraperitoneal administration of pentobarbital sodium (2% solution, 45 mg/kg). The animals were then euthanized under deep anesthesia by transcardial perfusion with PBS, which ensured death.

The human proximal tubular cell line HK-2, purchased from Wuhan Pricella Biotechnology Co., Ltd., was cultured as described [22]. The cell line was verified by STR profiling and tested negative for mycoplasma. Cells were randomly allocated into three groups: control, cobalt chloride (CoCl₂, 100 µmol/L), and CoCl₂+ emodin (100 µmol/L CoCl₂ and 20 µmol/L emodin) groups. Upon reaching 40–50% cellular confluence, cultures underwent 24 h of serum starvation to achieve cell cycle synchronization, followed by 6 h incubation with either emodin or vehicle (dimethyl sulfoxide, DMSO), and then stimulated with or without CoCl₂ (C8661, Sigma–Aldrich) for another 24 h.

Blood biochemical indicators, including blood glucose levels and lipid profiles, and urinary creatinine levels were determined using automatic analyzers (Rayto, Shenzhen, China). Urinary protein concentrations were quantified via commercial colorimetric assay (E-BC-K252-M, Elabscience, Wuhan, Hubei, China) and then multiplied by 24-h total urine volume to calculate the total protein content in 24-h urine. For albumin detection, a sandwich ELISA protocol was employed according to the manufacturer’s specifications. (E-EL-R0025c, Elabscience).

Following the procedures outlined in prior studies [23], specimens were processed for PAS staining and ultrastructural evaluation via electron microscopy. In PAS staining, five fields in each rat were randomly selected, and tubulointerstitial damage was quantified using Image-Pro Plus software 6.0 (Media Cybernetics, Rockville, MD, USA).

The lipid deposition in kidney tissue and HK-2 cells was determined using triglyceride (TG) quantitative assays and Oil Red O staining as previously described [9, 23].

Immunohistochemical staining of kidney tissue and immunofluorescence staining of HK-2 cells were performed by applying previously described methods [23]. Primary antibodies of transforming growth factor-beta 1 (TGF-$\beta$1, 1:400, bs-0086R, Bioss, Beijing, China), connective tissue growth factor (CTGF, bs-0743R, 1:400; Bioss), mitofusin 2 (Mfn2, 1:200, 9482S, Cell Signaling Technology, Danvers, MA, USA), and dynamin-related protein 1 (Drp1, 1:200, ab184247, Abcam, Cambridge, UK) were used for immunohistochemical staining. Primary antibody of HIF-1$\alpha$ (1:200, ab179483, Abcam) was used for immunofluorescence staining.

Total cellular RNA was extracted from HK-2 cells utilizing TRIzol reagent (9109, TaKaRa, Shiga, Japan) for subsequent RT-qPCR analysis. Then reverse transcripted to cDNA with a commercially available kit (R323-01, Vazyme, Nanjing, Jiangsu, China). Next, Quantitative amplification of target transcripts was subsequently performed on a Bio-Rad thermocycler with gene-specific primers and SYBR green PCR Master Mix (Q711-02, Vazyme). Relative mRNA expression was determined through the comparative 2^{-Δ⁢Δ⁢CT} approach. Primer sequences utilized for human gene detection are listed in Table 1.

We conducted Western blot experiments following protocols outlined in our earlier publication [22]. Incubation with primary antibodies targeted HIF-1$\alpha$ (1:1000, ab179483, Abcam), Drp1 (1:1000, ab184247, Abcam), Mfn2 (1:1000, 9482S, Cell Signaling Technology), TGF-$\beta$1 (1:1000, bs-0086R, Bioss), CTGF (1:1000, bs-0743R, Bioss), and $\beta$-actin (1:1000, AF5003, Beyotime).

Cell survival rates were evaluated through CCK-8 colorimetric assay (BS350B, Biosharp, Hefei, Anhui, China) following standard procedures. In brief, cells were seeded at roughly 2000 cells per well in 96-well plates and allowed to grow for 24 h prior to analysis. After synchronized, the cells were treated with emodin (0, 5, 10, 20, 40, 60, 80, and 100 µM) dissolved in DMSO for 6 h. Then, the supernatant containing emodin or DMSO was discarded, followed by adding 10 µL of CCK-8 reagent into each well for 2 h. Finally, optical density readings were then acquired using a multi-mode microplate reader (Thermo Fisher Scientific, Waltham, MA, USA).

The adenosine triphosphate (ATP) content in kidney tissues was evaluated using a commercial kit (E-BC-K157-M, Elabscience).

Total DNA isolation from renal tissues and HK-2 cells was accomplished utilizing a commercial extraction kit (DP304, Tiangen, Beijing, China). Then, the gene NADH dehydrogenase subunit-2 (ND2) of mitochondrial DNA (mtDNA) was amplified using the real-time quantitative PCR method previously described. The sequences of rat and human primers are shown in Table 2.

Mitochondrial transmembrane potential was evaluated through 5,5′,6,6′-Tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide (JC-1) staining methodology (E-CK-A301, Elabscience). Briefly, adherent cells from six-well plates were detached, pelleted, and suspended in loading buffer before exposure to JC-1 dye for 20 minutes at 37 °C. After dual washing steps, samples were re-suspended in chilled 1$\times$ assay buffer. Acquisition of fluorescent signals was performed on a BD Accuri C6 flow cytometer (BD, Franklin Lakes, NJ, USA). Red emission (575 nm, FL2) indicated JC-1 aggregate formation in viable cells with intact mitochondrial polarization, whereas green emission (525 nm, FL1) signified monomeric JC-1 accumulation in cells with compromised membrane potential.

Statistical analyses were conducted with GraphPad Prism 8.0 (GraphPad Software, San Diego, CA, USA) or SPSS 22.0 (IBM Corporation, Armonk, NY, USA). Data are presented as mean $\pm$ SD. Comparisons between two groups were performed using two-tailed unpaired t-test, while one-way ANOVA coupled with Tukey’s post-hoc test was employed for multi-group analyses. Statistical significance was set at p 0.05.

Emodin treatment significantly decreased the ratio of kidney weight to body weight, the 24-h urinary total protein levels, and the ratio of urinary albumin to creatinine in rats with diabetes (Fig. 1A–C), but did not affect blood urea nitrogen levels (Fig. 1D). Meanwhile, the body weights, and serum glucose levels in diabetic rats were not significantly changed by emodin treatment (Fig. 1E,F), suggesting that emodin did not improve kidney damage by affecting blood glucose level. PAS staining showed that emodin treatment rescued the pathological injury of renal tubules in diabetic rats (Fig. 1G,H), characterized by reduced glycogen deposition, attenuated luminal dilation, and less extensive epithelial cell detachment and necrosis. Consistently, the protein expression levels of TGF-$\beta$1 and CTGF in the kidneys of rats in the DM + emodin group significantly decreased compared with those in the DM group, suggesting that the degree of renal fibrosis in rats with diabetes was alleviated by emodin administration (Fig. 1I,J).

Fig. 1.

Emodin alleviated tubular damage in diabetic kidneys. (A) Ratio of kidney weight to body weight (n = 5–7), (B) 24-h total protein levels in urine (n = 5–7), (C) urinary albumin-to-creatinine ratio (n = 4–6), (D) blood urea nitrogen levels (n = 5–6), (E) body weights (n = 5–7), (F) blood glucose levels (n = 5–7) were measured. (G,H) PAS staining (original magnification 400$\times$, scale bar 200 µm). The tubular damage area was quantified using the software Image-Pro Plus (n = 5–6), IOD/area represents the integrated optical density per unit area, serving as an indicator of mean PAS-positive staining intensity. (I,J) For protein expression analysis, Western blot was employed to assess TGF-$\beta$1 and CTGF levels in renal tissue, with $\beta$-actin serving as an internal normalization standard. Densitometric analysis of protein bands was carried out using ImageJ (n = 4–6). Data are expressed as mean $\pm$ SD. Statistical significance relative to the Control group is indicated as *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001; comparisons against the DM group are denoted by ^#p 0.05, ^#⁢#p 0.01, ^{#⁢#⁢#⁢#}p 0.0001. DM, Diabetic Model; PAS, Periodic Acid-Schiff; IOD, Integrated Optical Density; TGF-$\beta$1, Transforming Growth Factor-beta 1; CTGF, Connective Tissue Growth Factor.

Hou and colleagues showed that emodin inhibited HIF-1$\alpha$ expression at the translational level [11]. As demonstrated by Western blot, emodin treatment markedly decreased HIF-1$\alpha$ protein expression in the kidneys of rats with diabetes (Fig. 2A,B). The serum lipid profile results showed that emodin administration significantly reduced low-density lipoprotein (LDL) and TG levels, but had no effect on the total cholesterol (TC) and high-density lipoprotein (HDL) levels in rats with diabetes (Fig. 2C). The results of TG quantitative assay and Oil Red O staining showed that emodin treatment decreased lipid accumulation in the kidneys of rats with diabetes (Fig. 2D,E). These results suggested that emodin reduced lipid deposition in diabetic kidneys, possibly by inhibiting HIF-1$\alpha$ protein expression.

Fig. 2.

Emodin decreased lipid deposition in the kidneys of rats with diabetes by inhibiting HIF-1$\alpha$ expression. (A,B) Western blot was conducted to evaluate HIF-1$\alpha$ abundance in renal tissue, with $\beta$-actin serving as an internal normalization standard. Densitometric analysis of protein bands was carried out using ImageJ (n = 4). (C) Serum lipid profiles were measured (n = 5–7). (D) TG quantification assay was performed to detect the TG content in the kidneys (n = 4–5). (E) Neutral lipid deposition in kidney sections was visualized by Oil Red O staining (400$\times$ magnification; scale bar = 200 µm). Data are expressed as mean $\pm$ SD. Statistical significance relative to the Control group is indicated as *p 0.05, **p 0.01, ****p 0.0001; comparisons against the DM group are denoted by ^#p 0.05, ^#⁢#p 0.01. DM, Diabetic Model; HIF-1$\alpha$, Hypoxia-Inducible Factor-1 alpha; HDL, High-density lipoprotein; LDL, low-density lipoprotein; TC, total cholesterol; TG, triglyceride.

It was revealed that the mitochondrial damage mediated by HIF-1$\alpha$ activation aggravated cellular injury [24]. It showed that most mitochondria in RTECs of rats with diabetes appeared fragmented, with fractured cristae and vesiculated matrix. Some mitochondria even appeared vacuolated, with swelling and rupture, losing their structural integrity (Fig. 3A). Emodin treatment improved the morphology and structure of mitochondria in RTECs of rats with diabetes (Fig. 3A). Additionally, we quantified the expression of mitochondrial dynamics regulators, specifically Mfn2 and Drp1. Immunohistochemical staining indicated that treatment with emodin normalized the suppressed expression of the fusion-associated protein Mfn2 in diabetic rat kidneys, though no comparable effect was observed on the fission-related protein Drp1 across experimental cohorts (Fig. 3B). The Western blot results were consistent with immunohistochemical staining findings (Fig. 3C,D). However, the ratio of Mfn2 to Drp1 expression levels markedly reduced in the kidneys of rats in the DM group compared with the control group, which was restored after emodin treatment (Fig. 3E). Besides, decreased ATP content and mtDNA copy number in rats with diabetes were rescued by emodin treatment (Fig. 3F,G), indicating that emodin restored energy production and mitochondrial biogenesis in rats with diabetes.

Fig. 3.

Emodin improved mitochondrial damage in the tubules of diabetic kidneys. (A) Representative transmission electron microscopy image of tubular epithelial cells in rats. The red arrows indicate damaged mitochondria and their cristae, whereas the yellow arrows indicate improved mitochondria and their cristae (4000$\times$ magnification, scale bar = 2 µm; 15,000$\times$ magnification scale bar = 1 µm). (B) Protein expression of Mfn2 and Drp1 in the kidneys was shown using immunohistochemistry staining (400$\times$ magnification, scale bar = 200 µm). (C,D) Western blot was conducted to evaluate the protein expression of Mfn2 and Drp1 in renal tissue, with $\beta$-actin serving as an internal normalization standard. Densitometric analysis of protein bands was carried out using ImageJ (n = 4–7). (E) Ratio of Mfn2 and Drp1 protein expression levels in kidneys was calculated to indicate the degree of mitochondrial fragmentation (n = 6). (F) ATP content of the kidneys was measured (n = 5–7). (G) mtDNA copy numbers of the kidneys were measured by real-time quantitative PCR of ND2 (n = 5). Data are expressed as mean $\pm$ SD. Statistical significance relative to the Control group is indicated as *p 0.05, **p 0.01, ****p 0.0001, NS (No Significance); comparisons against the DM group are denoted by ^#p 0.05, ^#⁢#p 0.01, NS (No Significance). DM, Diabetic Model; Mfn2, Mitofusin 2; Drp1, Dynamin-Related Protein 1; ATP, Adenosine Triphosphate; mtDNA, Mitochondrial DNA; ND2, NADH Dehydrogenase 2.

Based on CCK-8 assay results, emodin at a concentration of 20 µM was selected as the optimal intervention concentration for HK-2 cells (Fig. 4A). Western blot and immunofluorescence staining revealed that CoCl₂ stimulation markedly increased HIF-1$\alpha$ expression in the nucleus of HK-2 cells, which was significantly downregulated by emodin treatment (Fig. 4B–D). Meanwhile, emodin intervention substantially attenuated CoCl₂-triggered lipid deposition in HK-2 cells (Fig. 4E,F). Furthermore, the upregulated expression of profibrotic mediators TGF-$\beta$1 and CTGF in CoCl₂-exposed HK-2 cells was significantly suppressed following emodin application. (Fig. 4G,H). These results revealed that emodin effectively ameliorated lipid accumulation and cellular injury in CoCl₂-treated HK-2 cells, possibly by inhibiting HIF-1$\alpha$ expression.

Fig. 4.

Emodin attenuated lipid deposition and cellular damage in CoCl_2-treated HK-2 cells by inhibiting HIF-1$\alpha$ expression. (A) Cell viability of emodin-treated HK-2 cells with gradually increasing concentrations detected using the CCK-8 kit. (B,C) Western blot was conducted to evaluate HIF-1$\alpha$ abundance in HK-2 cells. (D) Immunofluorescence staining was used to reveal the cellular localization and protein expression level of HIF-1$\alpha$ (scale bar = 30 µm, original magnification 800$\times$). (E) TG quantification assay was performed to detect the TG content in HK-2 cells. (F) Oil Red O staining was used to observe neutral lipid accumulation in HK-2 cells (scale bar = 30 µm, original magnification 200$\times$). (G,H) Western blot analysis was conducted to evaluate the protein expression of TGF-$\beta$1 and CTGF in HK-2 cells. $\beta$-actin was used as a loading control. Densitometric analysis of protein bands was carried out using ImageJ. All experiments were performed with n $\ge$ 3 biological replicates, $\ge$2 technical replicates per biological replicate, and independently repeated three times. Results are presented as mean $\pm$ SD. Statistical significance relative to the Control group is indicated as **p 0.01, ***p 0.001, ****p 0.0001, NS (No Significance); comparisons against the CoCl₂ group are denoted by ^#p 0.05, ^#⁢#p 0.01, ^#⁢#⁢#p 0.001, ^{#⁢#⁢#⁢#}p 0.0001. CoCl₂, Cobalt Chloride; CCK-8, Cell Counting Kit-8; HIF-1$\alpha$, Hypoxia-Inducible Factor-1 alpha; TGF-$\beta$1, Transforming Growth Factor-beta 1; CTGF, Connective Tissue Growth Factor; TG, Triglyceride.

The mRNA expression levels of Drp1, Mfn2, and OPA1 significantly decreased in CoCl₂-treated HK-2 cells, which were rescued by emodin treatment (Fig. 5A–C). The protein expression of the mitochondrial fusion protein Mfn2 was reduced in CoCl₂-treated HK-2 cells but was markedly restored by emodin treatment. While no significant difference was observed in the expression levels of mitochondrial fission protein Drp1 among the three groups (Fig. 5D,E). However, the ratio of Mfn2 to Drp1 expression levels still significantly diminished in CoCl₂-treated HK-2 cells compared with those in the control group, which was restored after emodin treatment (Fig. 5F). These results suggested that emodin improved mitochondrial dynamics disorder by regulating the ratio of Mfn2 to Drp1 protein expression levels. Moreover, this treatment reversed the reduction of mtDNA copy number (Fig. 5G) and the loss of mitochondrial membrane potential (Fig. 5H,I) in CoCl₂-treated HK-2 cells.

Fig. 5.

Emodin ameliorated mitochondrial dysfunction in CoCl_2-treated HK-2 cells. (A–C) Real-time quantification PCR was performed to detect the mRNA expression of Drp1, Mfn2, and OPA1 in HK-2 cells. (D,E) Western blot was conducted to evaluate of Mfn2 and Drp1 protein expression in HK-2 cells, with $\beta$-actin serving as an internal normalization standard. Densitometric analysis of protein bands was carried out using ImageJ (F) The Mfn2-to-Drp1 expression ratio was computed to assess the extent of mitochondrial fragmentation in HK-2 cells. (G) For mtDNA quantification, real-time PCR targeting ND2 was performed in HK-2 cells. (H,I) Mitochondrial membrane potential of HK-2 cells was evaluated by flow cytometric analysis, with subsequent data processing conducted using FlowJo. All experiments were performed with n $\ge$ 3 biological replicates, $\ge$2 technical replicates per biological replicate, and independently repeated three times. Data are expressed as mean $\pm$ SD. Statistical significance relative to the Control group is indicated as *p 0.05, ****p 0.0001, NS (No Significance); comparisons against the CoCl₂ group are denoted by ^#p 0.05, ^#⁢#p 0.01, ^{#⁢#⁢#⁢#}p 0.0001, NS (No Significance). CoCl₂, Cobalt Chloride; Drp1, Dynamin-Related Protein 1; Mfn2, Mitofusin 2; OPA1, Optic Atrophy 1; mtDNA, Mitochondrial DNA; ND2, NADH Dehydrogenase 2; JC-1, 5,5′,6,6′-Tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide.