One hundred patients diagnosed with CKD from the Second Affiliated Hospital of Qiqihar Medical University and 30 healthy people were collected for inclusion in this study. Patient inclusion exclusion criteria: Patients attending our clinic were identified as having chronic kidney disease after routine urinalysis and renal function tests. Renal transplant recipients and patients with acute kidney injury were excluded. The baseline data of the patients was recorded at the time of diagnosis (Supplementary Table 1), and blood samples were collected simultaneously. Based on the guidelines developed by Kidney Disease: Improving Global Outcomes (KDIGO) [3], patients with CKD can be categorized into five stages based on estimated glomerular filtration rate (eGFR): CKD1 (eGFR, $>$90 mL/min), CKD2 (eGFR, 60–89 mL/min), CKD3 (eGFR, 30–59 mL/min), CKD4 (eGFR, 15–29 mL/min), and CKD5 (eGFR, 15 mL/min); Where eGFR 60 mL/min/1.73 m² indicates abnormal renal function and possible chronic kidney disease. Based on the eGFR values in the patients’ baseline data, CKD patients can be categorized into two groups, CKD 1-2 and CKD 3-5, for subsequent studies. The study was approved by the Second Affiliated Hospital of Qiqihar Medical University Ethics Committee (ethics number: [2022]0815-8-2) and subjects participating in the study gave informed consent. This study strictly adheres to the ethical principles and guidelines of the Declaration of Helsinki.

The study procured male C57BL/6 mice, aged 6–8 weeks, with an average weight of 25–30 grams, and THBS4-knockdown mice (shRNA-THBS4) were generated by intravenous injection of AAV9 particles expressing shRNA targeting THBS4 (Guangzhou Ruige Bio-Tech Co., Ltd., Guangzhou, China), while control mice (scramble) received AAV9-scramble-shRNA (Guangzhou Ruige Bio-Tech Co., Ltd., Guangzhou, China). A total of 12 mice were used in this study. All animal experiments comply with the international ethical standards for experimental animals and the regulations designated by the Second Affiliated Hospital of Qiqihar Medical University Ethics Committee (ethics number: [2022]0815-8-2). Human-derived renal tubular proximal epithelial cell line HK2 was purchased from Hunan Fenghui Biotechnology Co., Ltd. (CL0144, Changsha, Hunan, China).

Six 6–8 week-old healthy mice were prepared for each group. The mice were anesthetized by intraperitoneal injection of 2% sodium pentobarbital (4 mL/kg). In the experimental group, an incision was made on the dorsal side of the mice to reveal the left ureter. The ureter was then tied off using two 4.0-gauge silk sutures (0.35 mm diameter). The sham-operated group was operated only without ureteral ligation. 14 days later, mice were anesthetized as described above and then cervical dislocation was performed on the mice and kidney tissues were collected. Some kidneys were fixed with 4% formalin, followed by dehydration and paraffin embedding, and the remaining tissues were stored in a cryogenic liquid nitrogen environment for preservation [18].

HK2 cells were cultured in Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F12; C11330500BT, Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS; 10270106, Gibco, Grand Island, NY, USA) and 1% penicillin/streptomycin (15140122, Gibco, Grand Island, NY, USA) at 37 °C under 5% CO₂. For experimental interventions, cells were divided into six groups: Control (untreated), si-NC (transfected with scramble siRNA), si-THBS4 (transfected with THBS4-specific siRNA; sequence: 5^′-CGCCCUGAAUGAUCUCUAUTT-3^′; L31520, Beyotime, Shanghai, China), si-NC + TGF-$\beta$1 (scramble siRNA + 10 ng/mL TGF-$\beta$1; 100-21, PeproTech, Cranbury, NJ, USA), si-THBS4 + TGF-$\beta$1 (THBS4 siRNA + TGF-$\beta$1), and Rescue (THBS4 siRNA + TGF-$\beta$1 + 100 ng/mL IGF-1; HY-P7018, MedChemExpress, Shanghai, China). Transfection was performed using Lipofectamine 3000 (L3000015, Invitrogen, Carlsbad, CA, USA) with 50 nM siRNA at 60% cell confluence. After 6 hours, the medium was replaced, followed by TGF-$\beta$1 treatment for 48 hours or insulin-like growth factor 1 (IGF-1) treatment for 24 hours as indicated. The short tandem repeat (STR) validation and Mvcoplasma detection results for all cell lines were confirmed. The mycoplasma test results were negative.

Sections were routinely deparaffinized and heat repaired in antigen repair solution for 35 min, and then immunohistochemical (IHC) assays were performed according to the instructions of IHC kit (36311ES50, Yeasen Biotechnology, Shanghai, China). The sections were treated with 3% H₂O₂ at a temperature of 4 °C for a duration of 10 minutes, followed by rinsing with phosphate belanced solution (PBS) for three cycles, each lasting 2 minutes; incubated with Avidin at room temperature for 15 min and rinsed with PBS for three cycles, each lasting 2 minutes/3 times; exposed to Biotin for a duration of 15 minutes at ambient temperature, followed by a 5-minute rinse with PBS; subsequently, the blocking solution was introduced and the sections were subjected to a 20-minute blocking process at ambient temperature. The blocking solution was removed, and the sections were exposed to the primary antibody (THBS4, 1:100, CSB-PA562708, Wuhan Huamei BIOTECH, Wuhan, China). for an extended period at a low temperature of 4 °C. Wash with PBS for a duration of 5 minutes, followed by incubation with the secondary antibody (Horseradish Peroxidase (HRP) Goat Anti-Rabbit, 1:2000, ab205718, abcam, Cambridge, UK) at ambient temperature for a period of 30 minutes. Rinse with PBS for 5 min, add 3-3^′-diaminobenzidine (DAB) chromogenic solution for 5–10 min, rinse with water, re-stain, dehydrate, clear and seal. For the quantitative analysis of immunohistochemical staining, the intensity of staining was evaluated using the H-score method.

Paraffin sections were deparaffinized and subjected to PAS staining as described below. The sections were subjected to a 15-minute immersion in a 1% solution of periodic acid, followed by rinsing with water for 2–3 cycles. The sections underwent staining by being immersed in Schiff’s solution for a duration of 10–30 minutes, followed by rinsing with water 2–3 times. Stained with Haematoxylin for 1–2 min, washed 2–3 times in water. Immersion in 1% hydrochloric acid-ethanol for 5 sec, wash in water 2–3 times. Ammonia counterblue for 5–10 sec, rinsed in water, dehydrated, transparent, and sealed. Masson (ALP87019, Thermo Fisher Scientific, Waltham, MA, USA) and Picro Sirius Red (PSR) (36324ES60, Yeason Biotechnology, Shanghai, China) staining was performed according to the operating instructions of the kit. Staining results were quantitatively analyzed using Image-Pro Plus software (5.0.2, Media Cybernetics, Rockville Pike, MD, USA).

Patient serum was obtained by collecting patient whole blood and centrifuging the supernatant. The amount of THBS4 was detected based on the instructions of the ELISA kit (XG-E990828, Shanghai Sig Biotechnology Co., Ltd., Shanghai, China). The absorbance value of the test wells at 450 nm was read by an enzyme meter. The formula for calculating the concentration (C) of THBS4 in the samples based on the OD value (OD_sample) is as follows:

$$C={\mathrm{e}}^{\left(\frac{\left(O{D}_{\text{sample}}-O{D}_{\text{blank}}\right)\times \left({\log }_{10}\left(C_{\text{max}}\right)-{\log }_{10}\left(C_{\text{min}}\right)\right)}{\left(O{D}_{\text{max}}-O{D}_{\text{min}}\right)}+{\log }_{10}\left(C_{\text{min}}\right)\right)}$$

Where:

C is the concentration of THBS4 in the sample,

OD_{𝑠𝑎𝑚𝑝𝑙𝑒} is the optical density of the sample,

OD_{𝑏𝑙𝑎𝑛𝑘} is the optical density of the blank (background),

C_𝑚𝑎𝑥 and C_𝑚𝑖𝑛 are the maximum and minimum concentrations of the standard curve, respectively,

OD_𝑚𝑎𝑥 and OD_𝑚𝑖𝑛 are the corresponding optical densities of the standard curve.

Total RNA was extracted from tissues and cells using Trizol reagent (15596018, Invitrogen, Carlsbad, CA, USA). First strand DNA was synthesized by cDNA reverse transcription kit using RNA as a template. The reaction solution was prepared by SYBR Select Master Mix kit (4472919, ABI, Foster City, CA, USA) using DNA as a template. The reaction solution was used to complete quantitative fluorescence assay in a quantitative PCR instrument (Bio-Rad, Hercules, CA, USA) to accomplish quantitative fluorescence detection. The assay results were calculated using the 2^{-Δ⁢Δ⁢Ct} method. The amplification primer sequences of genes are shown in Table 1 (Human) and Table 2 (mouse).

The Radio Immunoprecipitation Assay (RIPA) lysis buffer was utilized to extract the entire cellular proteins, while the protein concentration was determined using the Bromocresol Green with Albumin (BCA) kit (A55864, Thermo Fisher Scientific, Waltham, MA, USA). The separation of proteins was achieved through the utilization of SDS-PAGE gel electrophoresis. The gel that underwent electrophoresis was transferred onto a polyvinylidene fluoride (PVDF) mold and sealed with 5% skimmed milk for 1 hour at ambient temperature. The gel was subsequently subjected to overnight incubation at a temperature of 4 °C with the primary antibody GAPDH (5174S; 1:1000), collagen type I alpha 1 (COL1A1, 72026; 1:1000), Fibronectin (26836; 1:1000), $\alpha$-SMA (19245; 1:1000), PI3K (4257; 1:1000), AKT (4691; 1:1000), p-PI3K (4292; 1:1000) and p-AKT (4060; 1:1000) (Cell Signaling Technology, Danvers, MA, USA) and THBS4 (ZY65162R, Shanghai Zeye Biotechnology Co., Ltd., Shanghai, China). After the membrane was washed, it underwent incubation with a secondary antibody (7074; 1:1000, Cell Signaling Technology, Danvers, MA, USA) for 1 hour at room temperature. Developed and photographed. The grayscale values of the protein bands were calculated employing Image 5.0 software (5.0.2, Media Cybernetics, Rockville Pike, MD, USA).

The statistical analysis of the experimental data was conducted utilizing IBM SPSS statistics 26 software (IBM Corp., Armont, NY, USA). Continuous variables are presented as mean $\pm$ standard deviation (SD) or median with interquartile range (IQR) depending on their distribution. The drawing tool is GraphPad Prism8 software (GraphPad, La Jolla, CA, USA). Data that were tested to be normally distributed and significant differences between two groups of data were analyzed by Student’s t-test. Significant differences between more than two groups of data were analyzed by one-way analysis of variance (ANOVA). Those that were tested not to follow normal distribution were analyzed for significant differences using Mann-Whitney U’s nonparametric test. p 0.05 represents significant differences (*p 0.05, **p 0.01, ***p 0.001, ****p 0.0001). The linear relationship between the two variables was accomplished by Pearson correlation analysis. Pearson correlation coefficient r 0 indicates negative correlation, r $>$ 0 represents positive correlation, and no correlation is indicated when r = 0. For post hoc analysis following one-way ANOVA, the Tukey’s Honestly Significant Difference (HSD) test was used to determine which specific groups differed significantly from each other when the ANOVA showed a significant result. The Tukey’s HSD test controls the family-wise error rate, ensuring that the overall Type I error rate is maintained at the desired level (commonly $\alpha$ = 0.05) across all multiple comparisons.

Serum was collected from 100 patients diagnosed with CKD as well as from 30 healthy people. ELISA assay showed significantly higher serum levels of THBS4 in CKD patients compared to healthy control population (Fig. 1A; p 0.05). The estimated glomerular filtration rate (eGFR) is an important indicator for determining CKD patients. The staging of CKD is mainly based on the patient’s eGFR, the lower the eGFR, the higher the CKD stage, CKD stage 5 is the uremic stage, the patient’s eGFR is extremely low. Correlation analysis was done between serum THBS4 levels and eGFR in CKD patients, and the findings revealed that serum THBS4 concentration was negatively associated with eGFR (Fig. 1B; r = –0.77, p 0.05). This suggests that THBS4 expression is elevated as eGFR decreases, i.e., as CKD stage increases. Next, we analyzed the relationship between THBS4 concentration and patient’s stage, and the results showed that THBS4 expression in the serum of CKD patients was significantly higher than that of controls, and in serum of CKD 3-5 patients was significantly higher than that of CKD 1-2 (Fig. 1C; p 0.05 for both comparisons). The above results indicated that THBS4 was closely related to renal fibrosis.

Fig. 1.

Expression analysis of THBS4 in clinical CKD patients. (A) Expression analysis of THBS4 in serum of CKD patients and healthy population. (B) Correlation analysis of THBS4 content with eGFR. (C) Relationship between THBS4 content and staging of CKD patients. Control, n = 30; CKD, n = 100. ****p 0.0001. THBS4, thrombospondin-4; CKD, Chronic kidney disease; eGFR, estimated glomerular filtration rate.

A mouse UUO model was constructed to reveal the expression of renal fibrosis-related indexes. qPCR showed that COL1A1, encoding Fibronectin (FN1), as well as encoding $\alpha$-SMA (ACTA2) were significantly elevated after UUO surgery compared with the sham-operated (Sham) group (Fig. 2A; p 0.05 for all comparisons). WB results revealed that protein expression of COL1A1, $\alpha$-SMA as well as Fibronectin was significantly increased in the UUO group compared to the Sham group (Fig. 2B,C; p 0.05 for all comparisons). These results suggested that our mouse renal fibrosis model was successfully constructed. PAS and Masson staining also showed renal mesenchymal injury with fibrotic lesions. IHC findings indicated a notable elevation in THBS4 expression within the kidneys of the UUO group (Fig. 2D,E, p 0.05). The qPCR results showed that the mRNA expression of THBS4 was significantly increased in the kidney tissues of UUO mice (Fig. 2F; p 0.05). These findings indicate THBS4 is significantly upregulated during the progression of renal fibrosis.

Fig. 2.

Expression analysis of THBS4 in the mouse UUO model. (A) qPCR detection of mRNA expression changes of renal fibrosis indexes after UUO. (B,C) WB detection of protein expression changes of renal fibrosis indexes after UUO. (D) PAS, Masson staining, and IHC for detection of renal fibrotic lesions; Magnification factor: 200$\times$. Scale bar = 50 µm. (E) Quantitative analysis of the figure (D). (F) THBS4 mRNA expression in mouse kidney tissue. n = 6. There is a significant difference compared to the Sham group, with ***p 0.001 and ****p 0.0001. UUO, unilateral ureteral obstruction; qPCR, real-time quantitative PCR; WB, Western Blotting; PAS, Periodic Acid-Schiff Stain; IHC, immunohistochemical.

To ensure the efficacy of THBS4 knockdown, both protein and mRNA levels were assessed. qPCR analysis showed a significant decrease in THBS4 mRNA expression in the si-THBS4 group (Fig. 3A; p 0.05). These validation experiments confirm that si-THBS4 effectively knocks down THBS4 expression, supporting the reliability of the observed anti-fibrotic effects. Similarly, Western Blot analysis demonstrated that THBS4 protein expression was significantly reduced in the si-THBS4 group compared to the si-NC group (Fig. 3B,C; p 0.05). TGF-$\beta$1, a potent fiber-forming factor, was used to induce HK2 cell fibrosis model. Cells were transfected with si-NC and si-THBS4 vectors and given TGF-$\beta$1 intervention as an experimental group (si-NC+TGF-$\beta$1, si-THBS4+TGF-$\beta$1), and the control group (si-NC) was transfected with si-NC vector only without TGF-$\beta$1 intervention. The qPCR results revealed a notable upregulation in the expression of COL1A1, FN1, ACTA2 and THBS4 after TGF-$\beta$1 stimulation contrasted with the si-NC group. COL1A1, FN1, ACTA2 and THBS4 was suppressed in the si-THBS4+TGF-$\beta$1 group contrasted with the si-NC+TGF-$\beta$1 group (Fig. 3D–G; p 0.05 for all comparisons). Consistent with the qPCR results, WB showed that interfering with THBS4 expression inhibited COL1A1, Fibronectin, $\alpha$-SMA protein expression (Fig. 3H; p 0.05 for both comparisons). This suggests that TGF-$\beta$1 stimulation induced HK2 cell fibrosis, and interfering with THBS4 expression could inhibit the process of HK2 cell fibrosis.

Fig. 3.

Effect of inhibiting THBS4 expression on HK2 cell fibrosis. (A) qPCR validation of THBS4 mRNA knockdown. (B,C) WB validation of THBS4 protein knockdown. (D–G) qPCR detection of the effect of inhibiting THBS4 expression on TGF-$\beta$1-induced renal fibrosis. (H) WB detection of the effect of inhibiting THBS4 expression on TGF-$\beta$1-induced renal fibrosis. n = 3. **p 0.01, ***p 0.001, ****p 0.0001. The ns indicates no significant difference. HK2, Human Kidney-2; TGF-$\beta$1, transforming growth factor $\beta$1.

Control mice (scramble) and THBS4 knockdown mice (shRNA-THBS4) were purchased. To confirm the efficacy of THBS4 knockdown in vivo, both mRNA and protein levels were assessed in kidney tissues from control mice and THBS4 knockdown mice. qPCR analysis revealed a significant reduction in THBS4 mRNA expression in knockdown mice compared to control mice (Fig. 4A; p 0.05). Consistent with these findings, Western Blot analysis demonstrated that THBS4 protein expression was significantly lower in knockdown mice compared to control mice (Fig. 4B,C; p 0.05). All mice were taken 14 days after UUO modeling, with the contralateral kidney as a control (Sham). PSR and Masson staining was performed on kidney tissue sections, which revealed increased renal interstitial fibrosis after UUO modeling. The semi-quantitative assessment of PSR and Masson findings revealed a significant increase in the extent of renal fibrosis within the UUO group contrasted with the Sham group. Knockdown of THBS4 followed by UUO modeling showed that the PSR and MTC-positive area were significantly lower in knockdown mice compared with control mice (Fig. 4D–G; p 0.05 for both comparisons). This indicated that renal fibrosis was alleviated in mice after knocking down THBS4. The qPCR assay revealed a significant upregulation in the expression of COL1A1, FN1, and ACTA2 in the kidneys of the UUO model compared to the Sham group. Whereas knockdown of THBS4 followed by the construction of the UUO model resulted in down-regulation of the expression levels of these fibrosis indicators in kidney tissues, which was statistically different compared with control mice (Fig. 4H; p 0.05 for both comparisons).

Fig. 4.

Mechanism study of THBS4 regulating the process of renal fibrosis. (A) qPCR validation of THBS4 mRNA knockdown in kidney tissues. (B,C) WB validation of THBS4 protein knockdown in kidney tissues. (D–G) PSR and Masson staining to detect the fibrosis of mouse kidney after THBS4 knockdown. Scale bar = 50 µm. (H) qPCR to detect the effect of THBS4 knockdown on the expression of renal fibrosis indicators in mice. n = 6. ***p 0.001, ****p 0.0001. PSR, Picro Sirius Red.

The above experiments have confirmed that interfering with THBS4 expression can inhibit the renal fibrosis process. We next explored its mechanism of action. In vitro, WB analysis showed that the protein expression ratios of p-PI3K/PI3K and p-AKT/AKT were significantly increased in HK2 cells following TGF-$\beta$1 intervention compared with the si-NC group. After THBS4 interference treatment, the levels of p-AKT/AKT and p-PI3K/PI3K exhibited a significant decrease (Fig. 5A,B; p 0.05 for all comparisons), indicating that THBS4 knockdown inhibits TGF-$\beta$1-induced activation of the PI3K/AKT pathway. To further validate this finding, additional experiments were conducted where HK2 cells were treated with TGF-$\beta$1 and transfected with si-NC or si-THBS4, followed by IGF-1 (100 ng/mL) or solvent. Western Blot analysis revealed that the protein expression ratios of p-PI3K/PI3K and p-AKT/AKT were significantly reduced in the si-THBS4 + activator control group compared to the si-NC + activator control group (p 0.05). Furthermore, IGF-1 treatment (si-THBS4 + activator) partially rescued the suppression of PI3K/AKT pathway activation caused by THBS4 knockdown, as evidenced by increased p-PI3K/PI3K and p-AKT/AKT ratios compared to the si-THBS4 + activator control group (p 0.05) (Fig. 5C,D). Meanwhile, we assessed the expression of COL1A1, FN1, ACTA2 and THBS4 via qPCR (Supplementary Fig. 1), which showed that their expression was significantly decreased in the si-THBS4 + activator control group and partially restored by IGF-1 treatment in the si-THBS4 + activator group. These results indicate that THBS4 knockdown inhibits the activation of the PI3K/AKT pathway, thereby reducing renal fibrosis. In vivo, the UUO model group exhibited significantly elevated p-AKT/AKT and p-PI3K/PI3K ratios compared to the Sham group, confirming PI3K/AKT pathway activation after surgery. Knockdown of THBS4 significantly decreased these ratios in UUO mice compared to control mice (Fig. 5E,F; p 0.05), suggesting that THBS4 regulates renal fibrosis by suppressing PI3K/AKT pathway activation through inhibition of PI3K and AKT phosphorylation.

Fig. 5.

Reduced THBS4 expression inhibits PI3K/AKT pathway activation. (A) WB detection of the effect of inhibition of THBS4 expression on PI3K/AKT pathway protein expression in cells, n = 3. (B) Quantitative analysis of Fig. 5A. (C) WB analysis of IGF-1 activator-treated cells with TGF-$\beta$1: THBS4 silencing suppressed PI3K/AKT pathway activation. (D) Quantitative analysis of Fig. 5C. (E) WB detection of the effect of THBS4 knockdown on PI3K/AKT pathway protein expression in mouse kidney, n = 6. (F) Quantitative analysis of Fig. 5E. *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001. The ns indicates no significant difference. PI3K/AKT, phosphatidylinositol 3-kinase/protein kinase B; IGF-1, insulin-like growth factor 1.