Exosomal microRNA‐16‐5p from human urine‐derived stem cells ameliorates diabetic nephropathy through protection of podocyte

Abstract Diabetic nephropathy (DN) remains one of the severe complications associated with diabetes mellitus. It is worthwhile to uncover the underlying mechanisms of clinical benefits of human urine‐derived stem cells (hUSCs) in the treatment of DN. At present, the clinical benefits associated with hUSCs in the treatment of DN remains unclear. Hence, our study aims to investigate protective effect of hUSC exosome along with microRNA‐16‐5p (miR‐16‐5p) on podocytes in DN via vascular endothelial growth factor A (VEGFA). Initially, miR‐16‐5p was predicated to target VEGFA based on data retrieved from several bioinformatics databases. Notably, dual‐luciferase report gene assay provided further verification confirming the prediction. Moreover, our results demonstrated that high glucose (HG) stimulation could inhibit miR‐16‐5p and promote VEGFA in human podocytes (HPDCs). miR‐16‐5p in hUSCs was transferred through the exosome pathway to HG‐treated HPDCs. The viability and apoptosis rate of podocytes after HG treatment together with expression of the related factors were subsequently determined. The results indicated that miR‐16‐5p secreted by hUSCs could improve podocyte injury induced by HG. In addition, VEGA silencing could also ameliorate HG‐induced podocyte injury. Finally, hUSC exosomes containing overexpressed miR‐16‐5p were injected into diabetic rats via tail vein, followed by qualification of miR‐16‐5p and observation on the changes of podocytes, which revealed that overexpressed miR‐16‐5p in hUSCs conferred protective effects on HPDCs in diabetic rats. Taken together, the present study revealed that overexpressed miR‐16‐5p in hUSC exosomes could protect HPDCs induced by HG and suppress VEGFA expression and podocytic apoptosis, providing fresh insights for novel treatment of DN.


| INTRODUC TI ON
Diabetic nephropathy (DN), one of the most common complications of diabetes mellitus, often results in end-stage renal disease, such as type I diabetes, type II diabetes and other secondary subtypes of diabetes. 1 Ageing, hyperglycaemia, obesity, hypertension, inheritance, sedentary lifestyle and smoking are considered to be the hallmark risk factors associated with the occurrence of DN. 2 Damage to the glomerular podocytes occurs with continually high blood glucose levels, with studies highlighting hyperglycaemia as a key event occurring in the early stages of DN. 3 The occurrence and development of DN are associated with damage of podocytes. 4 Moreover, studies have suggested that high glucose (HG) leads to a decrease in the number of podocytes, inducing the apoptosis of cultured podocytes.
HG also gives rise to proteinuria and speeds up the regression of foot processes. 5 Furthermore, protrusion of the foot process and detachment of podocytes from the glomerular basement membrane contribute to the occurrence of proteinuria, 6 highlighting the importance of inhibiting podocyte apoptosis for the treatment of renal diseases, including DN. 7 Stem cells, which can be self-renewed and differentiated into one or more pedigrees to produce specialized cell types, have been speculated as a potential therapeutic approach for DN. 8 Interestingly, consistent with the characteristics of marrow mesenchymal stem cells (MSCs), human urine-derived stem cells (hUSCs) isolated from human urine have been reported to secrete proangiogenic growth factors capable of maintaining the differentiation potential to endothelial cells, emphasizing their significance in the treatment of diabetic diseases. 9,10 Notably, hUSCs possess superior proliferative abilities when compared to adipose-derived stem cells, with studies indicating the ability of hUSCs to differentiate into osteoblasts, adipocytes and chondrocytes in the presence of optimal conditions, all of which highlight the potential of hUSCs as a promising cell source for regenerative medicine and novel cell therapies. 11 Exosomes, nanoscale vesicles secreted by a variety of cells, are involved in intracellular communication and material transport via directing action of signal molecules on the surface of the cell membrane and the mediation of cell content during membrane fusion. 12 Commonly, exosomes are nano-sized vesicles (40-100 nm) that can be transported between cells, carrying proteins, microR-NAs (miRNAs) and messenger RNAs (mRNAs). 13 A growing body of evidence has suggested that MSC-derived exosomes (MSC-Exo) are capable of carrying and secreting miRNA "marker signatures," which is a key factor in the treatment of chronic inflammation and the enhancement of the therapeutic effect of MSCs. 14 MicroRNAs are a group of small non-coding RNA molecules associated with the initiation and development of diabetes mellitus. 15 Intriguingly, a previous study demonstrated that circulating levels of miR- 16-5p are highly expressed in women with gestational diabetes mellitus (GDM) when compared to controls, suggesting the potential use of miR-16-5p as a non-invasive biomarker for GDM. 16 Furthermore, miR-16-5p directly targets vascular endothelial growth factor A (VEGFA), which has been widely reported as a critical mediator of DN, and participates in endogenous angiogenesis of endothelial cells. 17,18 Notably, the intravenous administration of hUSCs-Exo in streptozocin (STZ)-induced rat models has been found to reduce urine volume and urinary microalbumin excretion as well as inhibit podocyte apoptosis and promote angiogenesis. 19 However, as to whether miR-16-5p is present in hUSCs-Exo and manipulates VEGFA to protect podocytes from DN remains largely unknown. Hence, the current study set out to investigate the role of hUSCs-Exo combined with miR-16-5p in DN working in tandem with VEGFA.

| Ethical statements
The animal experiments in our study were strictly carried out in accordance with the instructions of animal protection and animal usage issued by American National Institutes of Health.

| Dual-luciferase reporter gene assay
The bioinformatics prediction website (microRNA.org) was employed to determine the target gene of miR-16-5p, which indicated that miR-16-5p could bind to the 3′ untranslated region (3′UTR) of VEGFA. A dual-luciferase reported gene assay was employed to further assess whether VEGFA was indeed a direct target gene of miR-16-5p. Next, miR-16-5p was co-transfected with reporter plasmids pMIR-VEGFA-wild-type (WT) or pMIR-VEGFA-mutant (MUT), respectively, into HEK-293T cells (CRL-1415, Shanghai Xinyu Biological Technology Co, Ltd). After a 48-hours period of transfection, the cells were collected, lysed and centrifuged with the supernatant collected. A luciferase assay kit (RG005, Beyotime Biotechnology Co, Ltd) was then used to dissolve the Renilla luciferase detection buffer and firefly luciferase detection reagent. The buffer (100 μL/sample) was added with substrate (1:100) to prepare the Renilla luciferase detection working solution. A fluorometer was used to determine the luciferase activity. In each group, 20-100 μL sample were added with 100 μL firefly luciferase detection reagent and 100 μL luciferase detection working solution, respectively, and fully mixed by means of pipetting followed by determination of the relative light unit (RLU). The reporter gene cell lysate was regarded as the blank control and renilla luciferase as an internal reference. Luciferase activity = RLU (firefly luciferase)/RLU (Renilla luciferase).

| In vitro HG-treated human podocytes (HPDCs) cultures
After the cryopreserved HPDCs had been thawed and centrifuged, the supernatant was removed and the cells at the bottom were resuspended with HPDC medium. The cells were then cultured in a 100-cm 2 plate with 8-10 mL HPDC medium in the incubator (33°C, 5% CO 2 ). When HPDCs had reached 70%-80% confluence, they were treated with 0.05% trypsin solution at 37°C for 3 minutes, centrifuged at 1120 g for 5 minutes and re-suspended in the HPDC medium. The cell suspension was then adjusted to a certain concentration and sub-planted into 2-3 culture plates (100 cm 2 ), which were then cultured in a 33°C incubator or a 37°C incubator (both containing 5% CO 2 ), respectively, to proliferate and differentiate.
After 8 days of culturing, mature HPDCs were incubated in eluent for 12 hours and then classified into three groups: control (normal HPDCs), MA (HPDCs treated with 25 mmol/L isotonic mannitol) and HG (HPDCs treated with 25 mmol/L HG for 24 hours).

| Cell counting kit-8 (CCK-8) assay
Human podocytes in the logarithmic growth were collected and inoculated into a 96-well plate and incubated in an incubator (37°C, 5% CO 2 ). After culturing for 24 hours, the corresponding reagents were added into each group (each group contained 6 duplicated wells).
After 24 hours of incubation, the CCK-8 reagent (10 μL/well) was added into the 96-well plate under aseptic conditions, after which the cells were cultured in the incubator for another 1.5 hours. The optical density (OD) values of each group were determined using an automatic microplate reader (Molecular Devices Corporation) at a wavelength of 490 nm. The measurement was repeated at least three times after which a mean OD value was obtained.

| hUSCs culture and screening
Fresh midstream urine (200 mL) from healthy male adults was collected and added with 5 mL Penicillin-Streptomycin (PS). The urine samples were then promptly sub-packed into 4 × 50 mL sterile centrifuge tubes and centrifuged. After that, 20 mL phosphate buffer saline (PBS) containing 1% PS was added into each tube with 6 mL LMM101 medium added for re-suspension purposes. The cell suspension was then transferred to 0.1% gelatin-coated 6-well plates and incubated at 37°C with 5% CO 2 . After 5-7 days, the cell adherence was observed, and the LMM101 medium was replaced to remove non-adherent cells. Finally, when the attached cells had reached 80% confluence, they were trypsinized and sub-cultured.

| Flow cytometry
The hUSCs at passage 3 were treated with 0.25% trypsin for 3 minutes, re-suspended with LMM101 medium and centrifuged ac-

| Osteoblasts and adipocytes differentiated from hUSCs
Directional differentiation of the hUSCs into osteoblasts was performed as follows: a total of 5 × 10 5 hUSCs at passage 3 were inoculated in 24-well plates. After the cell confluence reached more than 80%, 2 mL osteogenic inducing fluid (Gibco Company) or 2 mL adipogenic inducing fluid (Gibco Company) was added into each well.
Cells in the control group were added with 2 mL LMM101 medium, with the medium changed at the regular intervals every 3 days. The

| Exo quick-TC method
The hUSCs were collected 48 hours after lentiviruses infection, centrifuged at 28000 g for 15 minutes and added with Exo Quick-TC exosome sediment (every 1 mL Exo Quick-TC was added to 5 mL medium) at 4°C overnight. The Exo Quick-TC/body fluid mixture was centrifuged at 17920 g for 30 minutes. Exosome sediment located at the bottom of the tube was observed to be light brown or white in colour.

| Transmission electron microscopy (TEM)
The prepared exosomes were immediately fixed in 4% glutaraldehyde for at 4°C for 2 hours, washed three times with 0.1 mol/L PBS and then fixed with 1% osmium tetroxide for 2 hours. Semi-thin slices at a thickness of 0.5 μm were prepared and positioned under a light microscope. Then, ultra-thin slices with a thickness of 60 nm were prepared, stained with uranium acetate and lead citrate, and observed under a TEM. TEM was also used to observe the ultrastructure of podocytes in kidney tissues after the model was established. The podocyte, the average of foot process width (FPW) and the podocyte fusion rate were then calculated. (a) Podocytes: 3 glomeruli were observed in each case with 10 visual fields randomly selected to determine the number of podocytes; (b) FPW: the distance between the bilateral membrane of the foot process on the level of the pore membrane was measured in order to obtain the average value; (c) Podocyte fusion rate: the total length of the basement membrane was measured and regarded as X, then, after which the total length of the foot process fusion on the basement membrane was measured and regarded as Y. Finally, the fusion rate was calculated based on Y/X.

| Co-culture of HPDCs and hUSCs
Human podocytes and hUSCs were spread into a co-culture cham- The co-culture chamber was placed in a 6-well plate. During the coculturing process, the apical chamber was treated with Dulbecco's Modified Eagle Medium (DMEM) containing 10% serum, and the basolateral chamber was treated with 10% exosome-free foetal bovine serum (FBS) medium. The cells were co-cultured for 4-5 days, with the medium changed every 1-2 days.

| Confocal microscopy observation
The carboxyfluorescein diacetate succinimidyl ester (CFSE) was diluted at a ratio of 1:1000 and then mixed with the exosomes suspension secreted from 20 μg of lentivirus-infected hUSCs. After incubated at 37°C for 15 minutes, the exosomes labelled by CFSE were co-cultured with HPDCs. The uptake of exosomes by HPDCs was observed under a confocal fluorescence microscopy at 12 hours, 24 hours and 48 hours during co-culturing, respectively.

| Reverse transcription-quantitative polymerase chain reaction (RT-qPCR)
A TRIzol kit (Invitrogen) was applied to extract the total RNA from cells in each group. Next, 1 μg total RNA was reversely transcribed TA B L E 1 Primer sequences for RT-qPCR  was considered as the internal reference for VEGFA and nephrin.
Then, 2 −ΔΔCt was considered to be a reflection of the ratio of the expression of the target gene in the experimental group and the control group. All the primers were provided by Shanghai GenePharma Co Ltd (Table 1).

| Western blot analysis
Total protein was extracted from cells in each group. Total protein

| Terminal deoxynucleotidyl transferase (TdT)mediated 2′-Deoxyuridine 5′-Triphosphate (dUTP) nick-end labelling (TUNEL)
The  In the event, the random blood glucose was over 16.7 mol/L, and the 24 hours urine volume had doubled, the diabetes rat model was considered to be successfully established. 23

| Animal grouping and treatment
In addition to the normal group, the successfully established diabetic rats were randomly classified into three groups (at least 10 rats in each group): diabetes, diabetes + hUSCs-Exo control and diabetes + hUSCs-Exo miR-16-5p. Rats in the diabetes + hUSCs-Exo control and diabetes + hUSCs-Exo miR-16-5p groups were injected with 200 μL PBS containing 100 μg (protein content) exosomes (extracted from hUSCs infected with overexpressed miR-16-5p) and corresponding NC lentiviruses via tail vein, while rats in the diabetes group were injected with 200 μL pure PBS via tail vein. The above injections were given one time a week for consecutive 12 weeks. After that, metabolic cage was used to collect the urine at the 24th hours for determining urine protein (Upro) and urine creatinine (Ucr), and the rats in these four groups were euthanized. The blood glucose (Glu), kidney weight (KW) and body weight (BW) were measured, and an automatic biochemical analyzer (Hitachi) was used to detect renal functions, including that of blood urea nitrogen (BUN), serum creatinine (Scr) and creatinine clearance rate (CCR).

| Haematoxylin-eosin (HE) staining
After the collection of kidney tissues, routine fixation, dehydration, embedding, sectioning, staining and mounting were performed, after which the morphological changes of the podocytes were observed under an optical microscope (200×). A total of three visual fields were randomly selected for photographing. Afterwards, the surface area of renal tissue cells of the rats in each group was determined through the application of the automatic morphometery (HPIAS21000). The measurements were repeated five times in order to obtain average values.

| Periodic acid-Schiff (PAS) staining
After dewaxing and hydration, the paraffin slices were oxidized in periodic acid solution and stained with Schiff reagent (Beijing Leagene Biotech Co Ltd) for 30 minutes. The slices were then stained with haematoxylin for 5 minutes, dehydrated with 95% ethanol and absolute ethanol, cleared in xylene and mounted with neutral gum.
The hyperplasia area of the purple-red mesangial matrix of each rat was calculated using Image-Pro Plus software (Media Cybernetics Company), followed by observation and photography under an optical microscope. PAS staining results revealed the nucleus was blue, while the mesangial matrix, amyloid substance and fibrae sanguis were purple-red.

| TUNEL staining
The dewaxed slices were collected and incubated with 1% protease K diluent (50 μL) for 30 minutes at 37°C, followed by the addition of 0.3% H 2 O 2 methanol solution to eliminate endogenous peroxidase (POD) activity. Next, the slices were added with 50 μL converter-POD and developed by 2% diaminobenzidine (DAB).
After the cells were observed to have a brown-yellow nucleus under a microscope, the distilled water was added to terminate the reaction. Then, the slices were counterstained with haematoxylin, dehydrated by gradient alcohol, cleared, mounted and observed under an optical microscope (×400). Ten visual fields were randomly selected, after which the number of positive cells and podocytes were counted. The apoptotic positive cells had a brown-yellow nucleus, while normal cells presented blue nucleus. After the average values had been obtained, the apoptotic index (AI) of podocytes was expressed as a ratio of the number of brown-yellow cells to that of blue cells.

| Statistical analysis
All data were analysed using SPSS 21.0 software (IBM) with equality of variances assessed. Normal distribution was tested for all the data by using D'Agostino and Pearson omnibus normality test.
Measurement data subjected to normal distribution were expressed as mean ± standard deviation. Comparisons of mean values between F I G U R E 1 miR-16-5p targets VEGFA and inhibits its expression. (A,B) the miRNAs that could potentially be involved in the regulation of human or rat VEGFA predicted by miRNA-gene relationship prediction websites, including DIANA, miRDB, TargetScan, miRWalk and miRSearch websites. The comparison of the predicted results was displayed by a Venn diagram. C, the targeting relationship between miR-16-5p and VEGFA predicted by dual-luciferase reporter gene assay. D, Western blot analysis was performed to determine VEGFA expression in HPDCs stimulated by high glucose in the presence of miR-16-5p compared with that of miR-16-5p inhibitor. *P < .05 vs the VEGFA-MUT group. Measurement data were expressed as mean ± standard deviation. Data between two groups were analysed using t test, and the experiment was repeated three times. HPDCs, human podocytes; miR-16-5p, microRNA-16-5p; MUT, mutant; VEGFA, vascular endothelial growth factor A two groups were analysed using an independent sample t test, while comparisons among multiple groups were analysed using one-way analysis of variance (ANOVA). The post hoc test was applied for pairwise comparison of multiple groups with homogeneity of variance.
Rank-sum tests were used to compare the data failed to conform to normal distribution. A P value < .05 was considered to be significantly different.
The comparisons of the predicted results were illustrated using a Venn diagram, which identified two miRNAs, has-miR-16-5p and has-miR-195-5p in the intersection ( Figure 1A). Next, DIANA, miRDB, TargetScan, miRWalk and miRSearch were further used to predict the possible miRNAs that could regulate rat VEGFA. As depicted in Figure 1B In order to further verify the aforementioned results, a dual-luciferase reporter gene assay was performed. Compared with the NC group, the activity of luciferase in the VEGFA-WT group was significantly decreased in the presence of miR-16-5p mimic (P < .05), while no significant effect on the activity of luciferase in the VEGFA-MUT group was observed (P > .05) ( Figure 1C). HPDCs stimulated by HG were treated with overexpressed or inhibited miR-16-5p plasmids. Western blot analysis revealed that VEGFA expression in the miR-16-5p mimic group was significantly decreased compared with the mimic NC group (all P < .05). On the contrary, VEGFA expression in the miR-16-5p inhibitor group was significantly higher than that in the inhibitor NC group (all P < .05) ( Figure 1D). Taken together, the above results demonstrate that miR-16-5p can specifically bind to VEGFA-3′-UTR and down-regulate the expression of VEGFA.

| HG-stimulated HDPCs inhibits the expression of miR-16-5p and nephrin but promotes VEGFA expression
In

| hUSCs have specific molecular markers of MSCs and the ability of multidirectional differentiation
In order to assess whether hUSCs possessed the ability of multidirectional differentiation into osteoblasts or adipocytes, the cells were treated with osteogenic induction and adipogenic induction.
The cellular morphology was analysed first. After inoculation for 5 days, hUSCs were noted to have adhered to the wall in cobblestone-like shape ( Figure 3A-a). Subsequently, the cells exhibited a clone-like growth and reached 80%-90% confluence after 16 days ( Figure 3A-b). Next, 2-3 weeks after subculture, hUSCs retained slender morphology, grew and proliferated rapidly ( Figure 3A-c).
Then, the expression of hUSCs surface antigen was analysed by flow cytometry. As illustrated in Figure 3B,  Figure 3C-a). Besides, 14 days after adipogenic induction, hUSCs were observed to have a lipid sediment inside, with the lipid droplets gradually becoming larger and exhibiting a string-like structure, indicating that the hUSCs were capable of differentiating into adipocytes ( Figure 3C-b). Thus, based on the above results, we conclude that hUSCs have the specific molecular markers of MSCs and could differentiate into osteoblasts and adipocytes.

| hUSCs inhibit VEGFA expression in HGinduced podocytes through exosomes-secreted miR-16-5p
The size of the vesicles in the hHUSCs-Exo was varied when observed under a TEM, ranging from 30 nm to 120 nm ( Figure 4A).
The morphology was generally the same as the round or oval membranous vesicles. After staining, the complete envelope containing low-density electron-dense deposits was found in the vesicle.
Western blot analysis was employed to further verify the presence of hHUSCs-Exo and to detect the evolutionarily conserved proteins (CD9, CD63, CD81, TSG101 HSP90B1) as exosome surface markers.
In order to determine whether exogenous miR-16-5p could be transferred into HPDCs from hUSCs through exosomes, the expression of miR-16-5p in hUSCs and exosomes derived from hUSCs was determined by RT-qPCR. As displayed in Figure 4C, compared with miR-NC-treated cells, the expression of miR-16-5p was significantly higher in hUSCs infected with lentiviruses of miR-16-5p and in exosomes derived from the hUSCs (all P < .05). These results suggested that hUSCs could effectively release exosomes enriched with miR-16-5p. In order to further elucidate that exosomes released by hUSCs could transfer miR-16-5p to HPDCs, exosomes traced by CFSE were co-cultured with HPDCs. The uptake of exosomes by HPDCs was observed under a confocal fluorescence microscopy at 12, 24 and 48 hours after co-culture ( Figure 4D). Over the passage of co-culture, more and more HPDCs presented green fluorescence, suggesting that the number of CFSE-exosomes uptaken by HPDCs had increased in a progressive manner. After 48 hours of co-culture, the uptake of CFSE-exosomes by HPDCs was obvious, indicating that exosomes could be transferred from donor cells, hUSCs, to recipient cells, HPDCs.
In order to assess whether the in vitro transfer of miR-16-5p effectively inhibited endogenous VEGFA in HPDCs, the expression of miR-16-5p and VEGFA in HPDCs co-cultured with hUSCs was determined by RT-qPCR. Initially, whether exosomes were involved in the inhibition process was assessed, and the exosome-depleted medium was used to inhibit exosome secretion. As shown in Figure 4E, there was no significant change regarding the expression of miR- 16-5p and VEGFA in the Exo-depl group compared with the HPDC group

| Secretion of miR-16-5p from hUSC exosomes can ameliorate podocytes injury induced by HG
In order to investigate the effect of miR-16-5p secreted by hUSC exosomes on HG-induced podocyte injury, CCK-8 assay and TUNEL were applied to detect the podocyte viability and apoptosis rate after HG treatment. No significant difference was detected between the Exo-depl and hUSCs miR-16-5p-KD + VEGF-KD groups as well as the HPDC group regarding the podocyte viability and apoptosis rate (P > .05) ( Figure 5A-B). However, the podocyte viability was promoted, while the apoptosis rate was decreased in the hUSCs miR-16-5p and miR-16-5p mimic groups remarkably (P < .05). In contrast, Taken together, the aforementioned results indicated that the secretion of miR-16-5p from hUSC exosomes could ameliorate podocyte injury induced by HG.

| Silencing of VEGA can attenuate HG-induced podocyte injury
In order to detect the effect of VEGFA silencing on HG-induced podocyte injury, CCK-8 assay and TUNEL staining were conducted to detect the podocyte viability and apoptosis rate after VEGFA was depleted in hUSCs ( Figure 6A-B). The results showed that relative to the siRNA-NC group, the podocyte viability was enhanced and the apoptosis rate was reduced in the siRNA-VEGFA group (all P < .05).
Next, Western blot analysis was performed to determine the pro-

| Overexpressed miR-16-5p in hUSCs-Exo protects the podocytes in diabetic rats
Next, in order to assess whether the protective effect of hUSCs-Exo on nephrocytes exists in diabetic rats, DN rat models were established.
After STZ injection for 72 hours, the random blood glucose of rats in the model group was examined at the same time-point for three consecutive days, with the blood glucose detected to be higher than 16.7 mol/L (Table 2). Furthermore, in order to evaluate the protective effect of hUSCs-Exo on nephrocytes in diabetic rats, hUSC exosomes containing overexpressed miR-16-5p were injected into the tail veins of the DN rats. The rats were then assigned into normal, diabetes, diabetes + hUSCs-Exo control and diabetes + hUSCs-Exo miR-16-5p groups. The Glu, KW, BW, BUN, Scr, CCR, Upro and Ucr of rats were measured, with the results collected demonstrating that compared with the normal group Glu, KW, BUN, Scr, CCR, Ucr and Upro were increased, while BW was obviously decreased in the other three groups (all P < .05). Interestingly, Glu, KW, BUN, Scr, CCR, Ucr and Upro exhibited marked reductions in the diabetes + hUSCs-Exo miR-16-5p group compared with the diabetes + hUSCs-Exo control group (P < .05), but no significant changes were observed in BW (P > .05) ( Table 2).
Subsequently, RT-qPCR methods were employed to determine the expression of miR-16-5p in kidney tissue. The obtained results revealed that the expression of miR-16-5p was reduced in the diabetes, diabetes + hUSCs-Exo control and diabetes + hUSCs-Exo miR-16-5p groups compared with the normal group (P < .05).
Furthermore, the morphology changes of podocytes in each group were observed by a TEM ( Figure 7B). In the normal group, the podocyte structure was observed to be complete, with a neat and clear foot process arrangement, in addition to no observed thickening of the basement membrane. In the diabetes, diabetes + hUSCs-Exo and diabetes + hUSCs-Exo miR-16-5p groups, endothelial cell injury, basement membrane thickening and podocyte exfoliation and widened foot process space in addition to foot process fusion were detected. Remarkably, compared with the diabetes + hUSCs-Exo control group, the injury of podocytes in the diabetes + hUSCs-Exo miR-16-5p group was alleviated (all P < .05), and the injury of podocytes in the diabetes group was aggravated (all P < .05).
In order to further verify the protective effect associated with overexpressed miR-16-5p in hUSCs on rat nephrocytes, podocytes in each group were stained using HE, PAS and TUNEL. As illustrated in Figure 7C-D, the glomeruli of the rats in the normal group were observed to be clear in structure, regular in shape, with the cells in the glomeruli regularly arranged with clear a basement membrane.
However, the volume of glomerulus was increased, while the mesangial area was widened in addition to a greater matrix among the rats in the diabetes group. Moreover, the mesangial matrix hyperplasia of the rats in the diabetes + hUSCs-Exo control and diabetes + hUSCs-Exo miR-16-5p groups was more significantly ameliorated when compared to the diabetes group (P < .05). Compared with the diabetes + hUSCs-Exo control group, the hyperplasia of mesangial matrix in the diabetes + hUSCs-Exo miR-16-5p group was also alleviated (all P < .05). TUNEL assay results demonstrated that the apoptosis rate of nephrocytes in the diabetes + hUSCs-Exo control group was remarkably higher than that in diabetes + hUSCs-Exo miR-16-5p group (P < .05) ( Figure 7E).

| D ISCUSS I ON
Diabetic nephropathy commonly progressively manifests itself via haemodynamic and structural changes, which are often reflected by a decreased glomerular filtration rate and large amount of proteinuria. 22 Studies have implicated hUSCs-Exo in the process of wound repair in STZ-induced diabetic mice. 24 The key findings of the pre-  Mesenchymal stem cells regenerative therapy is a novel way of renal regeneration, where MSCs can repair the nephron through differentiation. 34 Studies have indicated that MSC-derived exosomes play a protective role in renal injury induced by ischaemia-reperfusion. 35 F I G U R E 5 Secretion of miR-16-5p from hUSC exosomes ameliorates podocyte injury induced by HG. A, the viability of podocytes was detected by CCK-8 assay. B, TUNEL was performed to measure the apoptosis rate of podocytes (scale bar = 25 μm). C, Western blot analysis was applied to determine the expression of nephrin, α-SMA, Bax and Caspase-3. D, the expression of VEGFA and relate factors MCP-1, TGF-β1 and TNF-α examined by Western blot analysis. *P < .05 vs the HPDC group. Measurement data were expressed as mean ± standard deviation. Multivariate analysis of variance was used for comparison among multiple groups, and the experiment was repeated 3 times. CCK-8, Cell Counting Kit-8; Exo-depl, hUSCs co-cultured with HPDCs in a exosome-depleted medium; HG, high glucose; HPDCs, human podocytes; hUSCs miR-16-5p-KD + VEGF-KD, hUSCs transfected with miR-16-5p inhibitor and siRNA-VEGFA plasmid and co-cultured with HPDCs; hUSCs miR-16-5p-KD, hUSCs transfected with miR-16-5p inhibitor and co-cultured with HPDCs; hUSCs, human urine-derived stem cells; MCP-1, monocyte chemoattractant protein-1; miR-16-5p, microRNA-16-5p; TGF-β1, transforming growth factor-β1; TNF-α, tumour necrosis factor-α; TUNEL, Terminal deoxynucleotidyl transferase (TdT)-mediated 2′-Deoxyuridine 5′-Triphosphate (dUTP) nick-end labelling; VEGFA, vascular endothelial growth factor A F I G U R E 6 Depletion of VEGFA contributes to relieved podocyte injury induced by HG. The expression of VEGFA was depleted. A, the viability of podocytes was detected by CCK-8 assay. B, TUNEL was performed to measure the apoptosis rate of podocytes (scale bar = 25 μm). C, Western blot analysis was applied to determine the expression of nephrin, α-SMA, Bax and Caspase-3. D, the expression of VEGFA and relate factors MCP-1, TGF-β1 and TNF-α examined by Western blot analysis. Measurement data were expressed as mean ± standard deviation. Multivariate analysis of variance was used for comparison among multiple groups, and the experiment was repeated three times. *P < .05 vs the siRNA-NC group. CCK-8, Cell Counting Kit-8; HG, high glucose; MCP-1, monocyte chemoattractant protein-1; TGF-β1, transforming growth factor-β1; TNF-α, tumour necrosis factor-α; TUNEL, Terminal deoxynucleotidyl transferase (TdT)mediated 2′-Deoxyuridine 5′-Triphosphate (dUTP) nick-end labelling; VEGFA, vascular endothelial growth factor A Interestingly, the paracrine mechanism, in which exosome is involved, has been shown to provide a therapeutic effect via stem cell transplantation. 36 Bearing in mind the advantages of both hUSCs and exosomes, hUSCs-Exo have been shown to be associated with a smaller rate of immune rejection, better differentiation, and more stable and sufficient supply, highlighting its potential role as a regenerative medicine Note: Measurement data were expressed as mean ± standard deviation, data among multiple groups were analysed using one-way analysis of variance, and the experiment was repeated three times. Abbreviations: BUN, blood urea nitrogen; BW, body weight; CCR, creatinine clearance rate; Glu, blood glucose; hUSCs, human urine-derived stem cells; KW, kidney weight; miR-16-5p, microRNA-16-5p; Scr, serum creatinine; Ucr, urine creatinine urine protein; Upro, urine protein. *P < .05 vs the normal group; # P < .05 vs the diabetes + hUSCs-Exo control group.
F I G U R E 7 hUSC-miR-16-5p overexpression has a protective effect on podocytes in diabetic rats. A, miR-16-5p expression was measured by RT-qPCR. B, TEM was used to observe the ultrastructure of podocytes, to count podocytes and to calculate the average of FPW and podocyte fusion rate (scale bar = 1 μm). C, the pathological changes of kidney tissue detected by HE staining (scale bar = 25 μm). D, PAS staining was used to observe the pathological changes of kidney tissue (scale bar = 25 μm). E, TUNEL was used to detect the apoptosis rate of kidney tissue (scale bar = 25 μm). F, the expression of VEGFA, MCP-1, TGF-β1 and TNF-α was determined by Western blot analysis. *P < .05 vs the normal group; #P < .05 vs the diabetes group. Measurement data were expressed as mean ± standard deviation. Data among multiple groups were analysed using one-way analysis of variance, and the experiment was repeated three times. FPW, foot process width; HE, haematoxylin-eosin; hUSCs, human urine-derived stem cells; MCP-1, monocyte chemoattractant protein-1; miR-16-5p, microRNA-16-5p; PAS, periodic acid-Schiff; RT-qPCR, reverse transcription-quantitative polymerase chain reaction; TEM, transmission electron microscopy; TGF-β1, transforming growth factor-β1; TNF-α, tumour necrosis factor-α; TUNEL, Terminal deoxynucleotidyl transferase (TdT)-mediated 2′-Deoxyuridine 5′-Triphosphate (dUTP) nick-end labelling; VEGFA, vascular endothelial growth factor A   19 Likewise, a previous study concluded that hUSCs-Exo can alleviate renal damage in type I diabetic rats, while demonstrating the therapeutic role played by exosomes through the suppression of podocyte apoptosis as well as promotion of angiogenesis and cell survival. 19

| CON CLUS IONS
All data together allow us to propose a model of molecular mechanisms underlying DN treatment by hUSCs (Figure 8). hUSCs-Exo overexpressing miR-16-5p could act to inhibit VEGFA while promoting proliferation and inhibiting the apoptosis of podocytes, all of which alleviate the damage inflicted by DN by protecting against podocyte injury. In conclusion, hUSCs-Exo overexpressing miR-16-5p presents a fresh perspective, for which future therapeutic approaches for DN may be premised upon. However, as the research is still in the preclinical stage, further investigation into the finer mechanisms is needed in future studies.

ACK N OWLED G EM ENTS
This study was supported by the Project of Henan Provincial Education Department (No. 16A320031).

CO N FLI C T O F I NTE R E S T
The authors have declared that no competing interests exist.

AUTH O R CO NTR I B UTI O N S
Yu-Rui Duan and Bao-Ping Chen designed the study. Fang Chen, Su-Xia Yang and Chao-Yang Zhu collated the data, carried out data analyses and produced the initial draft of the manuscript. Ya-Li Ma, Yang Li and Jun Shi contributed to drafting the manuscript. All authors have read and approved the final submitted manuscript.