Platelet Microparticles Mediate Glomerular Endothelial Injury in Early Diabetic Nephropathy

Abstract

Background

Glomerular endothelium dysfunction, which plays a crucial role in the pathogenesis of early diabetic nephropathy, might be caused by circulating metabolic abnormalities. Platelet microparticles, extracellular vesicles released from activated platelets, have recently emerged as a novel regulator of vascular dysfunction.

Methods

We studied the effects of platelet microparticles on glomerular endothelial injury in early diabetic nephropathy in rats with streptozotocin-induced diabetes and primary rat glomerular endothelial cells. Isolated platelet microparticles were measured by flow cytometry.

Results

Plasma platelet microparticles were significantly increased in diabetic rats, an effect inhibited in aspirin-treated animals. In cultured glomerular endothelial cells, platelet microparticles induced production of reactive oxygen species, decreased nitric oxide levels, inhibited activities of endothelial nitric oxide synthase and SOD, increased permeability of the glomerular endothelium barrier, and reduced thickness of the endothelial surface layer. Conversely, inhibition of platelet microparticles in vivo by aspirin improved glomerular endothelial injury. Further analysis showed that platelet microparticles activated the mammalian target of rapamycin complex 1 (mTORC1) pathway in glomerular endothelial cells; inhibition of the mTORC1 pathway by rapamycin or raptor siRNA significantly protected against microparticle-induced glomerular endothelial injury in vivo and in vitro. Moreover, platelet microparticle–derived chemokine ligand 7 (CXCL7) contributed to glomerular endothelial injury, and antagonizing CXCL7 using CXCL7-neutralizing antibody or blocking CXCL7 receptors with a competitive inhibitor of CXCR1 and CXCR2 dramatically attenuated such injury.

Conclusions

These findings demonstrate a pathogenic role of platelet microparticles in glomerular endothelium dysfunction, and suggest a potential therapeutic target, CXCL7, for treatment of early diabetic nephropathy.

Visual Abstract

Diabetic nephropathy (DN) is one of the most common microvascular complications of diabetes.¹ Despite increasing research progress in the treatment of DN, the incidence and prevalence of ESRD are rapidly rising worldwide.^2–4

The glomerular endothelium is the first layer of the glomerular filtration barrier (GFB), consisting of glomerular endothelial cells (GEnCs), the endothelial surface layer (ESL) and fenestrations, and tight junctions among cells.⁵ GEnCs are highly specialized cells characterized by the presence of 60–80 nm transcellular fenestrations and are covered with ESL.^6,7 Glycocalyx is the main component of the ESL, which covers both fenestral and interfenestral domains of the GEnC luminal surface.^5–7 The fenestrations and covered glycocalyx with characteristics of the molecular and charge barrier are essential for selective permeability.^7,8 The glomerular endothelium is encountered by plasma and forms the first barrier to macromolecular substances. In addition, the glomerular endothelium makes an important contribution to maintaining the homeostasis within the glomerulus in response to hemodynamic changes by secreting vasoactive substances. It is increasingly recognized that glomerular endothelial dysfunction is an early and deleterious hallmark of DN, resulting in impairment of the GFB and the appearance of albuminuria.^7,9,10 The thickness of the glycocalyx and mean percentage of fenestrations in glomerular endothelium were reduced in DN, which was correlated tightly with the permeability of the GFB and albuminuria.^11,12 Improvement of glomerular glycocalyx by inhibiting the heparanase level reduced glomerular permselectivity and the development of albuminuria.¹³ Recent studies have shown that endothelial nitric oxide synthase (eNOS) knockout mice develop aggravated DN with extensive mesangial expansion and severe podocyte injury, suggesting a causal link between endothelial dysfunction and the onset of early DN.^14,15 However, the initiating factors and related mechanisms of glomerular endothelial injury in early DN are still unclear.

Platelet microparticles (PMPs) are extracellular vesicles (EVs) derived from activated platelets that range in size from 0.1 to 1 μm in diameter.^16–18 PMPs account for 70%–90% of all of the circulating microparticles (MPs).¹⁶ PMPs enriched in proteins, lipids, and genetic information derived from activated platelets have the ability to modify the phenotype and function of the target cells.^19–21 PMPs increase the expression of intercellular adhesion molecule-1 in endothelial cells by delivering AA and inflammatory cytokines, leading to monocyte adherence to the vessel wall.²² In the condition of diabetes, the levels of circulating MPs are highly elevated, strongly associated with the urinary albumin excretion rate.^23–26 However, whether PMPs contribute to glomerular endothelium dysfunction in early DN is still unknown.

Therefore, this study aimed to investigate the role of PMPs in glomerular endothelial injury at the early stage of DN and to explore its underlying mechanisms.

Methods

Animals

Male Sprague–Dawley rats weighing 200–220 g were obtained from Shanghai Bikai Company (China). The rats were maintained under a constant 12-hour photoperiod at temperatures ranging from 21°C to 23°C and allowed free access to food and water, acclimatized for at least a week before further operation. All protocols, including diabetes induction and euthanasia operation, were approved by the Institutional Animal Care and Use Committee of Southeast University.

Diabetes was induced by streptozotocin (STZ) injection as described previously.²⁷ Briefly, rats were injected with 60 mg/kg body wt of STZ (Sigma) once, intraperitoneally. After 72 hours of STZ injection, blood glucose levels were measured, and rats with a blood glucose level higher than 16.7 mmol/L were diagnosed as diabetic rats. For the normal controls, rats were injected with 0.1 mol/L sodium citrate buffer once, intraperitoneally. At week 3 after diabetes induction, the rats were divided into four groups (n=10): normal controls treated with vehicle (Ctrl), diabetic rats treated with vehicle (DM), diabetic rats treated with aspirin (15 mg/kg) (DM+Asp), and diabetic rats treated with rapamycin (1 mg/kg) (DM+Rapa). Aspirin was given by intragastric administration daily in the whole study period. Body weight and 24-hour urine volume were determined weekly until death, at week 4, 8, and 12. The concentrations of BUN and serum creatinine were determined using automatic analyzers (Hitachi). Urinary albumin levels were measured using an ELISA.

GEnC Isolation, Culture, and Identification

Glomeruli were isolated from rat kidneys under sterile conditions by differential sieving, as previously described.²⁸ Briefly, male Sprague–Dawley rats weighing 150–200 g were anesthetized by an intraperitoneal injection of chloral hydrate (400 mg/kg). Kidneys were isolated and placed into cold sterile Hank solution. After tearing the renal capsule from the lavaged kidney, the renal cortex was cut into 1- to 2-mm³-thick fragments. The kidney tissues were ground with a 100-mesh stainless steel sieve, followed by filtering through 150- and 200-mesh steel sieves, respectively. Glomeruli in the 200-mesh steel sieve were collected and centrifuged at 460×g for 5 minutes. After two washes with serum-free DMEM, the glomeruli were treated with type IV collagenase (1 g/L) for 30 minutes at 37°C and then centrifuged at 460×g for 5 minutes. The supernatant was eliminated, and the pellet was washed twice with serum-free DMEM to collect the glomeruli. The glomeruli were seeded in a culture flask precoated with 10 μg/ml collagen I and the prepared endothelial cell medium. The endothelial cell medium was replaced every 3 days, and the GEnCs were grown in a confluent monolayer during the second week. Identification of GEnCs was made using immunofluorescence staining (Supplemental Figure 1).

For the mTORC1 pathway inhibition, GEnCs were treated with 10 ng/ml rapamycin (Sigma) or transfected with raptor siRNA using lipofectamine 2000 (Invitrogen) per the recommendations of the manufacturer.

The Isolation of Circulating MPs

Circulating MPs were separated according to the method reported by Dasgupta et al.²⁹ Briefly, circulating MPs were isolated by differential centrifugation. Blood was sampled by right heart puncture using a 10 ml syringe and then added to polypropylene tubes with 0.129 mol/L sodium citrate. Platelet-rich plasma (PRP) was obtained by centrifugation at 1000×g for 15 minutes at 18°C. The PRP was then centrifuged at 5000×g for 15 minutes at 18°C to obtain platelet-free plasma. The platelet-free plasma was further centrifuged using an Eppendorf 5424R centrifuge machine at 20,000×g for 30 minutes to precipitate MPs at 18°C. The pellets containing MPs were suspended in modified Tyrode buffer (MTB) and stored at −80°C.

Measurement of Circulating PMPs

To label PMPs, resuspension solution containing MPs was incubated with allophycocyanin-labeled Annexin V (Vazyme Biotech) and phycoerythrin (PE)-conjugated CD61 antibody (BD Biosciences) for 30 minutes. As a negative control, a subpopulation of MPs was resuspended in Annexin V binding buffer lacking calcium and containing a PE-conjugated IgG control antibody. The MPs were then centrifuged at 20,000×g for 30 minutes to remove excess fluorescent dye and washed twice by centrifugation for 30 minutes at 20,000×g. The pellet was resuspended in 300 μl PBS, followed by the addition of 1×10⁶ beads with a diameter of 3 μm. A tube containing beads with a diameter of 0.8 μm was used to gate the MPs. Finally, circulating PMPs were calculated by flow cytometry.

PMP Isolation from Activated Platelets In Vitro

Blood was collected from rats into centrifuge tubes with acid citrate dextrose buffer. After resting for 10 minutes at room temperature, the blood was centrifuged at 282×g for 10 minutes at 22°C. The PRP was then centrifuged for 5 minutes at 400×g, and the supernatant was harvested and further centrifuged for 5 minutes at 1600×g to collect the platelets. After one wash of the pellet with citrate glucose saline buffer, the platelets were resuspended in MTB. Platelets were counted and adjusted to a density of 3×10⁸ cells/ml, followed by stimulation with 30 mmol/L of high glucose, 20 μg/ml of collagen (type I collagen; Sigma), or high glucose plus collagen for 30 minutes at 37°C. Ethylenediaminetetraacetic acid (20 mmol/L) was added to stop the platelet activation, and the residual platelets were removed by centrifuging twice for 5 minutes each at 2000×g. The supernatant was harvested and centrifuged for 90 minutes at 20,000×g and 18°C to collect PMPs. Finally, the pellet containing the PMPs was resuspended in MTB. PMPs were stored at −80°C. To remove background dust and crystals, all reagents were double-filtered with 0.22 μm filters.

Protein Microarray Assay

Cytokines and chemokines in PMPs derived from platelets that were untreated or treated with collagen plus high glucose were measured using the Proteome Profiler (R&D Systems) according to the manufacturer’s recommendations.

Histology and Immunofluorescence Staining

For histologic analysis, tissues were dissected and fixed in 4% paraformaldehyde and then embedded in paraffin. Sections were stained with periodic acid–Schiff after deparaffinization. To evaluate the glomerular ESL glycocalyx, sections were stained with rhodamine-labeled wheat germ agglutinin lectin (10 μg/ml) for 2 hours at room temperature after deparaffinization and analyzed by Olympus fluorescence microscopy. For immunofluorescence, GEnCs and kidney sections were fixed with 4% paraformaldehyde for 30 minutes and treated with 0.05% Triton X-100 for 10 minutes. After blocking with 5% BSA, samples were stained with antibodies against CD61 (Novusbio), CD63 (Abcam), glypican-1, syndecan-1, CD31 (Santa Cruz Biotechnology), and p-S6K1 (Abcam). Secondary antibodies conjugated to Alexa Fluor 488, 555, or 594 were used (Life Technologies). Nuclei were labeled with 4′,6-diamidino-2-phenylindole. Coverslips were mounted on slides with glycerin and analyzed by Olympus fluorescence microscopy. Fluorescence intensity analysis was performed with ImageJ.

PMP Internalization Assay

Purified PMPs were labeled with PE-CD61 and incubated with GEnCs for 2, 4, or 12 hours. For uptake inhibition experiments, GEnCs were coincubated with different inhibitors. Cells were fixed, permeabilized, and stained for CD31 with Alexa Fluor 488 (Invitrogen) and with the nuclear dye 4′,6-diamidino-2-phenylindole. Imaging was performed on a fluorescence microscopy (Olympus, Japan) and fluorescence intensity analysis was performed with ImageJ.

The Ultramicrostructure Observation of PMPs

Rat kidneys were dissected and fixed in a mixture of 4% paraformaldehyde and 2.5% glutaraldehyde fixative buffered with PBS. Postfixation with 0.4% OsO4, the samples were dehydrated with a graded series of acetone before embedding in Epon 812. After stained with uranyl acetate and lead citrate, the sections were observed with a JEM1230. The morphology of the isolated PMPs was determined using electron microscopy. PMPs pellets were fixed in 4% paraformaldehyde for 30 minutes. The pellets were then loaded on grids coated with formvar carbon and stained with 0.75% uranyl acetate. Sections were observed with a Philips Tecnai-10 transmission electron microscopy operating at 80 kilovolts (Phillips Electronic Instruments). We further checked PMPs deposition in GEnCs by immunoelectron microscopy assay. Briefly, 70 nm kidney sample sections were transferred to formvar carbon–coated copper grids. Then, 1% H₂O₂ etching was used to remove osmium acid and then blocked with 5% BSA at room temperature for 60 minutes. The samples were then incubated with anti-CD61 antibody followed by secondary antibody conjugated with 10 nm gold. Finally, sections were observed with a Philips Tecnai-10 transmission electron microscopy.

Real-Time PCR

Total RNAs of GEnCs were extracted using TRIzol, referring to the manufacturer’s instructions. The mRNA expression was detected by real-time PCR using the SYBR Green PCR Master Mix (Takara, Japan). The primer sequences are listed in Supplemental Table 1.

Western Blot Analysis

Identical amounts of total protein extracts from renal cortex or cells were loaded and separated by 10% SDS-PAGE, and the proteins were then transferred onto a nitrocellulose membrane. After treatment with blocking solution, the membranes were incubated with anti-rat antibodies against CD61, glypican-1, syndecan-1, ZO-1, occludin, CD31, mTOR, p-mTOR, S6K1, p-S6K1, eukaryotic initiation factor 4E-binding protein 1 (4EBP1), and p-4EBP1 overnight at 4°C, followed by horseradish peroxidase–conjugated secondary antibodies for 2 hours at 4°C. Finally, the results were determined using the ECL Advance system (Amersham Biosciences).

Determination of Reactive Oxygen Species, Nitric Oxide, SOD, and eNOS

Cellular reactive oxygen species (ROS) were monitored as previously described.³⁰ Briefly, GEnCs were incubated with 10 μmol/L fluorescence probe DCFH-DA at 37°C for 30 minutes. After washing with PBS, the cells were harvested and resuspended in DMEM for flow cytometry. The nitric oxide (NO) levels in the renal cortex and cellular supernatant were measured using a commercially available NO assay kit. Nitric oxide synthase activity in GEnCs was measured by detecting the conversion of l-[14C] arginine to l-[14C] citrulline using a commercial NOS Detect Assay kit. The antioxidative enzyme activity of SOD was measured using a commercial kit. The procedure was performed according to the manufacturer’s instructions.

GEnCs Permeability Assay

The permeability assay of the GEnCs was performed as previously described.³¹ GEnCs were cultured in transwell plates. When cells in transwell membranes were fully confluent and formed an intact monolayer, the cells were treated for 24 hours. The cell culture medium was then removed, and 100 μl DMEM alone in the presence of FITC-BSA (0.25 mg/ml) was added to the upper compartment. The medium of the lower compartment was collected after 60 minutes. FITC-BSA was quantified in a fluorescence spectrofluorophotometer. The BSA flux was valued as a ratio between fluorescence intensities in the lower compartment and the upper compartment.

Statistical Analyses

Results are expressed as the mean±SD. Statistical analysis was performed using SPSS v16.0. Statistical comparisons among multiple groups were analyzed for significance by one-way ANOVA followed by Tukey or Games–Howell post hoc tests, or the Mann–Whitney U test for a nonparametric test. P<0.05 was considered a significant difference.

Results

Inhibition of PMPs Decreased Albuminuria and Ameliorated Glomerular Hypertrophy and Mesangial Matrix Accumulation in Early DN

We first determined whether PMPs were increased in a STZ-induced DM rat model. Circulating total MPs were isolated and detected by electron microscopy and flow cytometry assays. As shown in Figure 1, A and B, vesicles from plasma ranged in size from 0.1 to 0.8 μm with clear, intact, double-layer membrane structure. Using electron microscopy, we also found that DM rats produced more MPs than untreated controls. Treatment with aspirin significantly decreased MP formation at week 12 (Figure 1B). To determine the plasma levels of PMPs, the specific marker CD61 of platelets with Annexin V were used to detect PMPs. Flow cytometry showed that the majority of these MPs were CD61 and Annexin V–positive PMPs, which gradually increased from week 4 to week 12 in diabetic rats compared with the controls (Figure 1, C and D). Furthermore, by the end of week 12, the mean PMP count in the plasma of diabetic rats reached 20×10⁶/ml (Figure 1D). Aspirin is one of the most commonly used antiplatelet agents and has potential effects on the prevention of platelet activation. The results showed that plasma levels of PMPs in diabetic rats treated with aspirin were significantly decreased starting from week 8 (Figure 1D). We further observed the PMP infiltration in the glomeruli of rats. The results showed quantities of CD61-positive particles accumulated in the glomeruli of diabetic rats compared with the controls, suggesting that larger clusters of PMPs were deposited in the glomerulus (Figure 1, E and F). In order to distinguish platelets and PMPs, kidney sections were incubated with freshly isolated platelets and stained with anti-CD61 antibody. The results showed the complete platelet structure in glomerulus and the PMPs were much smaller than the intact platelets resided around the nuclei (Supplemental Figure 2A). Double immunofluorescence staining showed that CD63, a biomarker for EVs, overlapped with CD61 (Figure 1G). The colocalization confirmed that the CD61-positive structures represent PMPs. To further confirm that PMPs were internalized by GEnCs, we ultrastructurally observed the deposition of PMPs in GEnCs by immunoelectron microscopy assay. Results showed that there were quantities of CD61-positive dark black particles deposited in the cells, suggesting that PMPs were uptaken by GEnCs (Supplemental Figure 2B). Moreover, Western blotting further demonstrated that CD61 protein expression was markedly upregulated in the kidneys of diabetic rats at week 12 (Figure 1, H and I), suggesting that PMPs were incorporated into the glomerulus during the progression of DN. Interestingly, the deposition of PMPs in diabetic kidney was reduced after aspirin treatment (Figure 1, E and F).

Figure 1.

Inhibition of PMPs decreased albuminuria and ameliorated glomerular hypertrophy and mesangial matrix accumulation in early DN. Diabetes was induced by one intraperitoneal injection of STZ at 60 mg/kg body wt in 0.1 mol/L of sodium citrate buffer. Normal control rats were treated with vehicle (Ctrl) and diabetic rats were treated with vehicle (DM) or aspirin (15 mg/kg) (DM+Asp), and then euthanized at weeks 4, 8, and 12. (A and B) Plasma MPs were isolated according to a circulating MP isolation procedure and observed by electron microscopy (scale bar, 200 and 500 nm). (C) The plasma PMP level was analyzed by flow cytometry. MPs from DM rats or annexin V buffer controls, isotype controls, and MPs after lysis were analyzed by flow cytometry. PMPs were defined as CD61+/Annexin V+ events in Q2 window and analyzed for quantification. (D) The results are expressed as events per milliliter of plasma (mean±SD, n=8). *P<0.05; **P<0.01 compared with Ctrl; ^#P<0.05 compared with DM group. (E and F) The presence of CD61⁺ PMPs (green dots) within the glomeruli in the renal cortexes of rats at week 12 was determined by immunofluorescent staining. Nuclei were stained with DAPI (blue) (scale bar, 20 μm). (G) Double immunofluorescence staining of CD61 (green) and CD63 (red) in glomeruli of DM rats at week 12. Nuclei were stained with DAPI (blue) (scale bar, 20 μm). (H and I) Protein levels of CD61 in renal cortexes of rats at week 12 were detected by Western blotting. The relative density of protein bands were quantified and normalized to β-actin (mean±SD, n=5). *P<0.05; **P<0.01 compared with Ctrl; ^#P<0.05 compared with DM group. Blood glucose levels (J), BUN (K), serum creatinine (L), the urinary albumin/creatinine ratio (M), and the ratio of kidney weight to body weight (N) in rats were measured (mean±SD, n=8). *P<0.05; **P<0.01 compared with Ctrl; ^#P<0.05 and ^##P<0.01 compared with DM group. (O) Histopathological changes were evaluated by periodic acid–Schiff staining. Original magnification, ×400 (scale bar, 50 μm). (P and Q) The glomerular area and mesangial expansion score were determined from histology sections. *P<0.05; **P<0.01 compared with Ctrl; ^#P<0.05 and ^##P<0.01 compared with DM group. (R) Correlation analysis between the plasma PMP level and urinary ACR was determined by Spearman rank-order correlation analysis.

We then examined the role of the increased PMPs in the kidney injury of diabetic rats. Aspirin treatment did not affect the blood glucose levels of diabetic rats (Figure 1J). The results demonstrated no differences in serum creatinine in diabetic compared with control rats, although an increase in BUN was detected in diabetic rats from week 8 to week 12 (Figure 1, K and L). Interestingly, inhibition of PMPs by aspirin was able to override the elevated urinary albumin-to-creatinine ratio and kidney body weight ratio in diabetic rats (Figure 1, M and N). Periodic acid–Schiff staining indicated that glomerular hypertrophy and mesangial matrix expansion in early DN were markedly ameliorated by aspirin treatment (Figure 1, O–Q). Furthermore, after aspirin treatment, decreased plasma PMP levels in diabetic rats were positively associated with a significant reduction of albumin-to-creatinine ratio (Figure 1R).

PMPs Induced Glomerular Endothelial Injury in Early DN

As the glomerular endothelium directly contacts circulating PMPs, we checked the effects of PMPs on glomerular endothelial injury. Immunofluorescence staining showed that with an extended time of PMP coincubation with GEnCs, endothelial cells showed increased PMP uptake (Figure 2A). To test whether PMPs were internalized by endothelial cells, we labeled PMPs with a R-PE–tagged CD61 and coincubated with GEnCs. High magnification pictures and the Z-section analysis indicated that most of the labeled PMPs were uptaken by GEnCs, and a few of them were combined on the membrane (Figure 2B, Supplemental Figure 3). To test whether uptake of PMPs into GEnCs was through endocytic pathways, we coincubated GEnCs with PMPs alone or together with a series of known endocytosis inhibitors. Dynasore is an inhibitor of dynamin-2 required both for clathrin-mediated and caveolin-dependent endocytosis. Chlorpromazine is an inhibitor of clathrin-mediated endocytosis. We tested dynasore, chlorpromazine, the actin filament inhibitors latrunculin A and cytochalasin D, and the microtubule inhibitor nocodazole. We found that all these inhibitors had significant effects on PMPs uptake (Figure 2, C and D), strongly suggesting that PMPs internalization by GEnCs is indeed via endocytosis.

Figure 2.

PMPs induced glomerular endothelial injury in vitro. (A and B) Primary GEnCs were incubated with diabetic rat-derived PMPs labeled with PE-conjugated anti-CD61 (red) antibody for 0, 2, 4, and 12 hours at 37°C. The GEnCs were labeled with Alexa 488-conjugated anti-CD31 (green) antibody and nuclei were stained with DAPI (blue), and uptake of PMPs was observed by confocal microscopy (scale bars, 20 and 10 μm). (C and D) GEnCs were coincubated with PMPs labeled with PE-CD61 and inhibitors for 4 hours, and PMP uptake was quantified compared with control. The results are expressed as mean number of PMPs per cell (scale bars, 5 μm). **P<0.01 compared with Ctrl. (E and F) Primary GEnCs were incubated without or with 0, 10³/ml, 10⁴/ml, 10⁵/ml, and 10⁶/ml of MPs derived from diabetic rats for 24 hours. The NO level in the culture supernatant (E) and intracellular ROS in GEnCs (F) were measured (mean±SD, n=4 experiments). *P<0.05 compared with control; **P<0.05 compared with DM-PMPs (10³); ***P<0.05 compared with DM-PMPs (10⁴); ****P<0.05 compared with DM-PMPs (10⁵). (G) PMPs were isolated from platelets stimulated without or with collagen (Col; 25 μg/ml), high glucose (HG; 30 mmol/L), or collagen plus high glucose (Col+HG) and quantified by flow cytometry. The results are expressed as the mean fold changes versus unstimulated platelets, which were used as a reference (mean±SD, n=3 experiments). *P<0.05 compared with Ctrl; ^#P<0.05 compared with Col-PMPs group. (H–K) Primary GEnCs were incubated without or with 5×10⁶/ml of different types of PMPs for 24 hours. NO level (H) in culture supernatant, activity of eNOS (I) and SOD (J), and intracellular ROS (K) were measured in GEnCs (mean±SD, n=3 experiments). *P<0.05 compared with Ctrl; ^#P<0.05 compared with Col-PMPs group. (L) Expression of glypican-1 (green) and syndecan-1 (red) in GEnCs was determined by immunofluorescence staining. Nuclei were stained with DAPI (blue) (scale bars, 20 μm). (M) The permeability of GEnCs was assessed by measuring the ratio of FITC-labeled albumin in the bottom after 1 hour (mean±SD, n=4 experiments). *P<0.05 compared with Ctrl; ^#P<0.05 compared with Col-PMPs group. Chlo, chlorpromazine (10 μg/ml); Ctrl, control; Cyto, cytochalasin D (600 nmol/L); Dyn, dynasore (80 μmol/L)l Lat, latrunculin A (5 μmol/L); Noco, nocodazole (20 μmol/L).

The PMPs isolated from the plasma of diabetic rats decreased NO production and increased ROS levels in GEnCs in a dose-dependent manner (Figure 2, E and F), indicating that the PMPs from diabetic rats induced GEnCs dysfunction. To further confirm the effects of PMPs on endothelial injury, we isolated PMPs in vitro from cultured rat platelets activated with different agonists. The results demonstrated that the generation of PMPs from activated platelets stimulated by high glucose plus collagen (Col+HG-PMPs) was remarkably increased compared with high glucose or collagen alone (Figure 2G), suggesting that a high glucose background is a critical factor inducing and amplifying PMP generation in diabetes. A significant reduction of the NO levels was observed, with inhibition of eNOS and SOD activities, in contrast to increased production of ROS in GEnCs during stimulation with Col+HG-PMPs (Figure 2, H–K). Consistently, immunofluorescence analysis confirmed that Col+HG-PMPs reduced the expression of glycocalyx-associated core proteins glypican-1 and syndecan-1 (Figure 2L). In addition, GEnC permeability was enhanced by Col+HG-PMPs, as shown by a significant increase in trans-endothelial FITC-BSA flux (Figure 2M).

The in vivo results revealed a gradually reduction of the ESL thickness, as assessed by rhodamine-wheat germ agglutinin staining in diabetic rats compared with controls, which were reversed by aspirin (Figure 3, A and B). Immunofluorescence and Western blot assays also revealed that expression levels of the glycocalyx-associated core proteins glypican-1 and syndecan-1 as well as the tight junction-associated protein ZO-1 and occludin in glomerular endothelium were downregulated in diabetic rats, whereas aspirin provided a protective role for the GEnC glycocalyx and tight junction in early DN (Figure 3, C–H). Moreover, inhibition of PMPs by aspirin caused a reduction of ET-1 expression and improvement of GEnC fenestration in diabetic rats (Figure 3, G–K). To observe the effect of PMPs on GEnCs diastolic function, we detected the NO level and SOD activity in the kidney cortex. As shown in Figure 3, L and M, the NO level and SOD activity were decreased in diabetic rats; however, this response was mitigated by aspirin treatment.

Figure 3.

Inhibition of PMPs protected glomerular endothelium against injury in early DN. Diabetes was induced by one intraperitoneal injection of STZ at 60 mg/kg body wt in 0.1 mol/L of sodium citrate buffer. Normal control rats were treated with vehicle (Ctrl) and diabetic rats were treated with vehicle (DM) or aspirin (15 mg/kg) (DM+Asp), and then euthanized at weeks 4, 8, and 12. (A) Glomerular ESL was determined by staining with rhodamine-labeled WGA. A coat of glomerular endothelial glycocalyx lined the luminal surface of the endothelial cells. Original magnification, ×400 (scale bars, 20 μm); enlarged image magnification, ×800 (scale bars, 5 μm). (B) Quantitation of WGA staining in the glomerulus. *P<0.05; **P<0.01 compared with Ctrl; ^#P<0.05; ^##P<0.01 compared with DM. (C and D) The expression of glypican-1 (green) and syndecan-1 (red) in renal tissues of rats was determined by immunofluorescence staining. Original magnification, ×400 (scale bars, 20 μm). (E and F) Quantitation of immunofluorescence staining for glypican-1 and syndecan-1. *P<0.05; **P<0.01 compared with Ctrl; ^#P<0.05; ^##P<0.01 compared with DM. (G and H) Protein levels of glypican-1, syndecan-1, ZO-1, occludin, and ET-1 in renal cortexes of rats were detected by Western blotting. The relative density of the protein bands was quantified and normalized to β-actin (mean±SD, n=5). **P<0.01 compared with Ctrl; ^#P<0.05; ^##P<0.01 compared with DM. (I) The ultramicrostructure of the GFB, including the endothelial fenestration, glomerular basement membrane, and podocyte foot processes, was analyzed by transmission electron microscopy (scale bars, 0.5 μm). (J and K) The glomerular endothelial fenestration and glomerular basement membrane thickness were determined from ultramicrostructure images. *P<0.05 compared with Ctrl; ^#P<0.05 compared with DM group. (L) NO level and (M) SOD activity in the renal cortex of rats were measured (mean±SD; n=5). *P<0.05 compared with Ctrl; ^#P<0.05 compared with DM group.

PMPs Activated the mTORC1 Pathway in GEnCs

We next investigated potential mechanisms of PMP-induced glomerular endothelial injury. The mTORC1 is a rapamycin-sensitive protein kinase complex that regulates processes of cell growth, metabolism, proliferation, and autophagy by phosphorylating p70 S6 kinase 1 (S6K1) and 4EBP1. Hyperactivation of mTORC1 in kidney cells is associated with early changes in DN. Although we did not observe a significant difference in mTOR, S6K1, and 4EBP1 at the mRNA level between the control and PMP-stimulated group (Figure 4, A–C), immunofluorescence staining and Western blot analysis showed that PMPs significantly increased protein phosphorylation of mTOR, S6K1, and 4EBP1, indicating a hyperactivation of the mTORC1 pathway (Figure 4, D–F). In the in vivo study, double immunofluorescence staining of p-S6K1 and CD31 showed that S6K1 phosphorylation was increased in the cytoplasm of CD31-positive cells in diabetic rats, suggesting an upregulation of mTORC1 activity in the glomerular endothelium (Figure 4G). Aspirin treatment prevented S6K1 phosphorylation in GEnCs of diabetic rats (Figure 4G). Consistent with these findings, Western blot analysis also showed higher levels of p-mTOR, p-S6K1, and p-4EBP1 in the glomeruli of diabetic rats, which were reversed by aspirin, confirming the contribution of PMPs to mTORC1 activation (Figure 4, H and I).

Figure 4.

PMPs activated the mTORC1 signaling in GEnCs. (A–F) Primary GEnCs were incubated without or with 5×10⁶/ml of different types of PMPs for 24 hours. (A–C) mRNA level of mTOR, S6K1, and 4EBP1 in GEnCs was detected by real-time PCR. (D) Expression of p-S6K1 in GEnCs was detected by immunofluorescence staining. Original magnification, ×400 (scale bars, 20 μm). (E and F) Protein levels of mTOR, p-mTOR, S6K1, p-S6K1, 4EBP1, and p-4EBP1 in GEnCs were detected by Western blotting. The relative intensity of protein bands was quantified, and the ratio of phosphoprotein to total protein was expressed (mean±SD, n=3 experiments). **P<0.01 compared with Ctrl. (G–I) Diabetes was induced by one intraperitoneal injection of STZ at 60 mg/kg body wt in 0.1 mol/L of sodium citrate buffer. Normal control rats were treated with vehicle (Ctrl) and diabetic rats were treated with vehicle (DM) or aspirin (15 mg/kg) (DM+Asp) for 12 weeks. (G) Coexpression of p-S6K1 and CD31 in renal tissues of rats was determined by immunofluorescence staining. Original magnification, ×400 (scale bars, 20 μm). (H and I) Protein levels of mTOR, p-mTOR, S6K1, p-S6K1, 4EBP1, and p-4EBP1 in renal cortexes of rats were detected by Western blotting. The relative intensity of protein bands was quantified, and the ratio of phosphoprotein to total protein was expressed (mean±SD, n=5). **P<0.01 compared with Ctrl; ^#P<0.05; ^##P<0.01 compared with DM group.

PMPs Mediated Glomerular Endothelial Injury through Activation of the mTORC1 Pathway

To confirm whether mTORC1 signaling is involved in PMP-mediated glomerular endothelial injury, rapamycin or raptor siRNA was used to block this pathway. Western blotting showed that the increased protein expression of p-mTOR, p-S6K1, and p-4EBP1 in GEnCs was significantly inhibited by rapamycin or raptor-specific siRNA treatment (Figure 5, A and B). Under these conditions, inhibition of mTORC1 reversed the PMP-induced decrease in NO levels and eNOS activity, and decreased ROS production (Figure 5, C–E). Immunofluorescence and Western blot analysis indicated that rapamycin or raptor siRNA prevented PMP-induced upregulation of the ET-1 expression and downregulation of glypican-1, syndecan-1, ZO-1, and occludin (Figure 5, F–H). Furthermore, inhibition of mTORC1 reversed the increased permeability of the GEnC barrier induced by PMP treatment (Figure 5I). To evaluate the role of mTORC1 signaling in glomerular endothelial injury of early DN, we treated diabetic rats with rapamycin. PMP deposition in the kidneys was not altered in rapamycin-treated compared with vehicle-treated diabetic rats (Figure 6, A and B). As shown in Figure 6, C–H, rapamycin prevented glomerular endothelial injury by improving the glomerular endothelial fenestration, ESL glycocalyx, and tight junctions, as well as by decreasing ET-1 expression. Moreover, rapamycin enhanced NO production and SOD activity in the diabetic rats (Figure 6, I and J).

Figure 5.

PMPs mediated glomerular endothelial injury through activation of the mTORC1 pathway in vitro. Primary GEnCs were incubated without (Ctrl) or with 5×10⁶/ml of PMPs, PMPs plus 10 ng/ml of rapamycin (PMPs+Rapa), or pretreated with raptor siRNA (PMPs+siRaptor) or negative control siRNA (PMPs+siCtrl) for 24 hours. (A and B) Protein levels of mTOR, p-mTOR, S6K1, p-S6K1, 4EBP1, and p-4EBP1 in GEnCs were detected by Western blotting. The relative intensity of the protein bands was quantified, and the ratio of phosphoprotein to total protein was expressed (mean±SD, n=3 experiments). **P<0.01 compared with Ctrl; ^##P<0.01 compared with PMPs group. The NO level (C), eNOS activity (D), and intracellular ROS (E) in GEnCs were measured (mean±SD, n=3 experiments). *P<0.05 compared with Ctrl; ^#P<0.05 compared with PMPs group. (F) The expression of glypican-1 (green) and syndecan-1 (red) in GEnCs was determined by immunofluorescent staining. Nuclei were stained with DAPI (blue). Original magnification, ×400 (scale bars, 20 μm). (G and H) Protein levels of glypican-1, syndecan-1, ZO-1, occludin, and ET-1 in renal cortexes of rats were detected by Western blotting. The relative intensity of the protein bands was quantified and normalized to β-actin (mean±SD, n=3 experiments). *P<0.05; **P<0.01 compared with Ctrl; ^#P<0.05; ^##P<0.01 compared with PMPs group. (I) The permeability of GEnCs was assessed by measuring the ratio of FITC-labeled albumin in the bottom after 1 hour (mean±SD, n=4 experiments). **P<0.01 compared with Ctrl; ^##P<0.01 compared with PMPs group.

Figure 6.

PMPs mediated glomerular endothelial injury through activation of the mTORC1 pathway in vivo. Diabetes was induced by one intraperitoneal injection of STZ at 60 mg/kg body wt in 0.1 mol/L of sodium citrate buffer. Normal control rats were treated with vehicle (Ctrl) and diabetic rats were treated with vehicle (DM) or rapamycin (1 mg/kg) (DM+Rapa), and then euthanized at week 12. (A and B) The presence of CD61⁺ PMPs (green dots) within the glomerulus in the renal cortexes of rats was detected by immunofluorescence staining followed by fluorescence semiquantitative analysis. Nuclei were stained with DAPI (blue) (scale bar, 20 μm) (mean±SD, n=5). **P<0.01 compared with Ctrl. (C and D) The ultramicrostructure of endothelial fenestration was observed by electron microscopy and the endothelial fenestrations were quantified. (scale bars, 0.5 μm). **P<0.01 compared with Ctrl; ^#P<0.05 compared with DM. (E) Glomerular ESL was determined by staining with rhodamine-labeled WGA. A coat of glomerular endothelial glycocalyx lined the luminal surface of endothelial cells. Original magnification, ×400 (scale bars, 20 μm); enlarged image magnification, ×800 (scale bars, 5 μm). (F) Expression of glypican-1 (green) and syndecan-1 (red) in renal tissues of rats was determined by immunofluorescence staining. Original magnification, ×400 (scale bars, 20 μm). Nuclei were stained with DAPI (blue). (G and H) Protein levels of glypican-1, syndecan-1, ZO-1, occludin, and ET-1 in renal cortexes of rats were detected by Western blotting. The relative intensity of the protein bands was quantified and normalized to β-actin (mean±SD, n=5). **P<0.01 compared with Ctrl; ^#P<0.05; ^##P<0.01 compared with DM. (I) NO levels and SOD activity (J) in renal cortexes of rats were measured (mean±SD, n=5). **P<0.01 compared with Ctrl; ^#P<0.05; ^##P<0.01 compared with DM.

PMP-Derived CXCL7 Promoted GEnC Damage

PMPs contain various proinflammatory and anti-inflammatory cytokines and chemokines deriving from platelets. PMPs can act as messengers and deliver these signals through soluble mediators, regulating lymphocyte rolling and activating endothelial cells, leukocytes, and other platelets. Given the efficient uptake of PMPs by GEnCs (Figure 2B), we considered whether the binding of PMPs to GEnCs could deliver cytokines that target the mTORC1 signaling pathway. To characterize PMP-derived cytokines, PMP lysates were analyzed with cytokine protein microarrays. We surveyed a range of cytokines elicited from activated PMPs and observed a large amount of the chemokine CXCL7 (Supplemental Figure 4, A and B). For the following experiments, we focused on CXCL7 because this is the most abundant cytokine in PMPs. Consistently, CXCL7 expression was significantly higher in activated PMPs than normal PMPs by ELISA analysis (Figure 7A). Because CXCL7 mediates its effect through the G protein–coupled receptors CXCR1 and CXCR2, we analyzed the expression of CXCR1 and CXCR2 in PMPs treated GEnCs by real-time PCR. The results revealed higher expression of CXCR1 and CXCR2 in GEnCs that received activated PMPs treatment compared with those treated with normal PMPs (Figure 7B). Next, we sought to determine whether the abundant CXCL7 could recapitulate the effect on the GEnCs mTORC1 pathway of PMPs. Western blot analysis showed that CXCL7 increased the phosphorylation of mTORC1, S6K1, and 4EBP1 in GEnCs in a dose-dependent manner (Figure 7, C and D). However, SB225002, a competitive inhibitor of CXCR1 and CXCR2, reduced the PMP-induced mTORC1 activation in GEnCs (Figure 7, E and F), indicating that CXCL7 is responsible for mTORC1 activation induced by PMPs in GEnCs.

Figure 7.

PMP-derived CXCL7 contributed to GEnC injury. (A) Concentrations of CXCL7 in different types of PMPs lysate were measured by ELISA (mean±SD, n=4 experiments). *P<0.05 compared with Ctrl. (B) Primary GEnCs were incubated with Ctrl or Col+HG-PMPs for 24 hours. The mRNA levels of the CXCR1 and CXCR2 in GEnCs were determined by real-time PCR (mean±SD, n=3 experiments). *P<0.05 compared with Ctrl. (C and D) GEnCs were treated with CXCL7 (1, 10, 20, 40, or 100 ng/ml) for 24 hours. Protein levels of mTOR, p-mTOR, S6K1, p-S6K1, 4EBP1, and p-4EBP1 in GEnCs were detected by Western blotting. The relative intensity of the protein bands was quantified, and the ratio of phosphoprotein to total protein was expressed (mean±SD, n=3 experiments). *P<0.05; **P<0.01 compared with Ctrl. (E–O) GEnCs were pretreated without or with SB225002 (1 mmol/L) for 2 hours and then incubated with 5×10⁶/ml PMPs for 24 hours. (E and F) Protein levels of mTOR, p-mTOR, S6K1, p-S6K1, 4EBP1, and p-4EBP1 in GEnCs were detected by Western blotting. The relative intensity of the protein bands was quantified, and the ratio of phosphoprotein to total protein was expressed (mean±SD, n=3 experiments). **P<0.01 compared with Ctrl; ^#P<0.05; ^##P<0.01 compared with PMPs. NO level (G), eNOS (H), and SOD (I) activity, and intracellular ROS (J) in GEnCs were measured (mean±SD, n=4 experiments). **P<0.01 compared with Ctrl; ^#P<0.05; ^##P<0.01 compared with PMPs. (K) The GEnC glycocalyx was evaluated by WGA staining (scale bars, 20 μm). (L) Expression of glypican-1 (green) and syndecan-1 (red) in GEnCs was determined by immunofluorescence staining. Nuclei were stained with DAPI (blue). Original magnification, ×400 (scale bars, 20 μm). (M and N) Protein levels of glypican-1, syndecan-1, ZO-1, occludin, and ET-1 in renal cortexes of rats were detected by Western blotting. The relative intensity of the protein bands was quantified and normalized to β-actin (mean±SD, n=3 experiments). *P<0.05; **P<0.01 compared with Ctrl; ^#P<0.05; ^##P<0.01 compared with PMPs. (O) Permeability of GEnCs was assessed by measuring the ratio of FITC-labeled albumin in the bottom after 1 hour (mean±SD, n=4 experiments). **P<0.01 compared with Ctrl; ^#P<0.05 compared with PMPs.

To ascertain that CXCL7-mediated PMPs induced GEnCs damage, CXCR1/2 in GEnCs was antagonized by SB225002, and diastolic function, glycocalyx, cell junctions, and permeability were analyzed in GEnCs. We found that pretreatment with the SB225002 reversed the reduction of the NO level, eNOS and SOD activity, and decreased ROS production in GEnCs induced by PMPs (Figure 7, G–J). Consistently, the destruction of the GEnC glycocalyx and cell junctions by PMPs was reversed by SB225002 (Figure 7, K–N). Furthermore, blockade of CXCL7 by SB225002 preserved the permeability of the GEnCs barrier disrupted by PMPs (Figure 7O). Moreover, to test whether the PMPs-induced glomerular endothelial injury was dependent on CXCL7, CXCL7 blockade using CXCL7-neutralizing antibody was performed. Interestingly, anti-CXCL7 treatment led to a marked reduction of mTORC1 activity, increase of glycocalyx, and recovery of glomerular endothelium function under the stimulation of PMPs (Figure 8). These results indicated that CXCL7/CXCR1/2 pathway played a crucial role in PMP-induced glomerular endothelial injury.

Figure 8.

CXCL7-neutralizing antibody blocked PMP-derived CXCL7 induced GEnC injury. GEnCs were treated with 5×10⁶/ml PMPs for 24 hours and anti-rat CXCL7 antibodies (1 μg/ml) for 24 hours. (A and B) Protein levels of mTOR, p-mTOR, S6K1, p-S6K1, 4EBP1, and p-4EBP1 in GEnCs were detected by Western blotting. The relative intensity of the protein bands was quantified, and the ratio of phosphoprotein to total protein was expressed (mean±SD, n=3 experiments). **P<0.01 compared with Ctrl; ^#P<0.05 compared with PMPs. NO level (C), eNOS (D), and SOD (E) activity in GEnCs were measured (mean±SD, n=4 experiments). *P<0.05; **P<0.01 compared with Ctrl; ^#P<0.05 compared with PMPs. (F) The GEnC glycocalyx was evaluated by immunofluorescence staining of glypican-1 (green) and syndecan-1 (red) and WGA staining. Nuclei were stained with DAPI (blue). Original magnification, ×400 (scale bars, 10 μm).

Discussion

The release of PMPs after platelet activation is considered to be a novel effector that may play a key role in the development of DN. Here we report that PMPs participate in the intercellular communication between platelets and GEnCs via the delivery of platelet cytokines and regulation of protein expression, leading to glomerular endothelial injury and progression of early DN.

Our data showed that plasma PMPs level was significantly elevated in STZ-induced diabetic rats from week 4, and approximately 80% of circulating MPs were derived from platelets. This result was further confirmed in an in vitro study showing that high glucose effectively enhanced the ability of platelets to generate PMPs in the presence of agonists, suggesting that high glucose is a crucial contributor to the rise in PMP levels in diabetes. By immunofluorescence and Western blot analyses, we observed a large amount of accumulated PMPs in the glomeruli of diabetic rats, but rarely in the glomeruli of normal rats. This difference likely reflected alterations in the quantity and function of PMPs, and the abnormal glomerular endothelial activation in diabetes, favoring the aggregation of PMPs. Importantly, the significant increase in the PMP level in plasma and the glomerulus was inhibited by aspirin treatment, indicating that aspirin is a potential PMP inhibitor in diabetes. Furthermore, inhibition of PMPs by aspirin significantly ameliorated DN by reducing albuminuria, glomerular hypertrophy and mesangial matrix expansion, and it had no effect on blood glucose level. These results suggested a strong association between the increased PMP level and the progression of early DN.

Considering that the glomerular endothelium is in direct contact with blood and is vulnerable to PMPs, we then studied the effect of PMPs on the glomerular endothelium. First, we demonstrated that most of plasma-derived PMPs from diabetic rats were uptaken by GEnCs when coincubated in vitro, and then PMPs were internalized by GEnCs via endocytosis in a variety of forms. The role of PMPs in diseases has recently begun to emerge and varies with different disease conditions and the different types of cells.^29,32,33 PMP stimulation in vitro induced tumor cell apoptosis,³⁴ inhibited regulatory T cell production of IL-17,³⁵ and induced fibroblast-like synoviocytes to release IL-8.³⁶ Our results indicated that plasma PMPs led directly to the decrease in NO synthase activity and increase in ROS production in GEnCs. Therefore, we further investigated the effect of PMPs on GEnCs by using purified PMPs derived from activated platelets stimulated by agonists in vitro. Consistently, PMPs released from activated platelets effectively inhibited NO synthase activity and SOD production, but increased ROS production, in GEnCs. In addition, PMPs reduced expression of the glycocalyx-associated core proteins glypican-1 and syndecan-1, suggesting a disruption of GEnC structural integrity. The permeability assay confirmed that PMPs reduced permeability of the GEnC barrier, as shown by increased FITC-BSA passage through GEnCs. These alterations of GEnCs induced by PMPs are consistent with the manifestations of glomerular endothelium injury in DN. To confirm these findings from the in vitro study, we treated diabetic rats with aspirin, which effectively reduced the level of PMPs from week 8. Consistently, our in vivo study revealed that inhibition of PMPs protected against GEnC injury in early DN, favoring the restoration of GEnC fenestration, improvement of the glycocalyx and tight junctions, and increased NO synthesis and SOD activity. A previous clinical study reported that the circulating PMP level was associated with micro- and macroangiopathy in young patients with type 1 diabetes.²³ Here, we first demonstrated that PMPs contributed to the progression of early DN by directly inducing glomerular endothelial damage. It suggests that PMPs could be a therapeutic target for the treatment of early DN. Tarnow et al.³⁷ found that nephropathy in type 1 diabetes was associated with increased circulating activated platelets and platelet hyperreactivity. Some clinical and experimental studies reported that aspirin provided a renoprotective role in DN, which was mainly through reducing urinary protein, decreasing the expression of inflammatory factors and collagen in kidneys.^38,39 However, clinical application of aspirin in DN treatment is still relatively insufficient and its mechanism for DN improvement is not very clear. Our findings evoked a novel therapeutic concept for DN, implying that the pharmacological inhibition of PMPs released from activated platelets might prevent glomerular endothelial injury. Thus, besides traditional pharmacological efficacies, aspirin could be exploited as a drug for the prevention of glomerular endothelial injury in the progression of early DN.

Next, we explored potential mechanisms of glomerular endothelial injury induced by PMPs. It has been reported that PMPs regulate the signaling pathway in target cells both at the gene and protein levels depending on its contents. Ceroi et al.⁴⁰ demonstrated that PMPs stimulated liver X-receptor activation in plasmacytoid dendritic cells. Laffont et al.³⁴ showed that PMPs regulated the gene expression of F-box/WD repeat-containing protein 7 and ephrin A1 in human umbilical vein endothelial cells. Increasing evidence has indicated that mTORC1 plays a crucial role in the regulation of kidney cell homeostasis and has been implicated in the development of glomerular lesions in DN.⁴¹ The activation of mTORC1 induces oxidative stress, apoptosis, and phenotypic changes in podocytes.^42,43 In addition, downregulation of the mTORC1 pathway alleviates hindlimb ischemic injury in diabetic mice by improving the vascular structure as well as attenuating the inflammation, oxidative stress, and apoptosis in endothelial cells.⁴⁴ Our previous studies indicated that mTORC1 activation induced by inflammation contributed to atherosclerosis through disruption of the LDL receptor pathway in vascular smooth muscle cells.⁴⁵ In this study, we demonstrated that PMPs activated the mTORC1 pathway by increasing the phosphorylation levels of mTOR and their downstream proteins S6K1 and 4EBP1 in GEnCs in vitro. In contrast, inhibition of PMPs by aspirin decreased activation of the mTORC1 pathway in GEnCs of diabetic rats. Disruption of mTORC1 signaling by the knockdown of raptor or rapamycin protected against the PMP-induced decrease in NO synthases, SOD and eNOS activity, increase of ROS production, loss of glycocalyx and tight junctions, and hyperpermeability in GEnCs. Consistent with these in vitro results, rapamycin improved the above alterations of GEnCs in early DN, although the PMP level in kidney was not changed. Thus, the mTORC1 pathway modulated PMP-induced GEnC injury.

It is known that PMPs are the basic storage units for different active proteins in platelets, including resting platelet membrane proteins and a variety of active substances. Previous studies have verified the capacity of PMPs to elicit or disseminate a range of cytokines and chemokines, transcription factors, and microRNAs to recipient cells.^34,46 For example, Michael et al.⁴⁷ showed that PMPs infiltrated solid tumors and transferred platelet RNA to tumor cells in vivo and in vitro. Using a protein microarray technique, we found that activated platelets released the vast majority of inflammatory cytokines in PMPs. CXCL7 was richly expressed and significantly increased in PMPs released by activated platelets. CXCL7 is one kind of cytokine in the CXCL family that signal through the G protein–coupled receptors CXCR1 and CXCR2, which leads to activation of intracellular signaling pathways.⁴⁸ In this study, we found that the expression of CXCR1 and 2 in GEnCs significantly increased after treatment with PMPs, suggesting that PMP-derived CXCL7 interacted with GEnCs by binding to CXCR1 and CXCR2. Interestingly, CXCL7 activated the mTORC1 pathway in GEnCs in a dose-dependent manner, which was inhibited by CXCL7 blockade using CXCL7-neutralizing antibody or an inhibitor of CXCL7 receptors, SB225002. These findings suggest that CXCL7 is a major PMP-derived regulator of glomerular endothelial injury.

In summary, increased PMPs contributed to glomerular endothelial injury in early DN through the CXCL7 released from PMPs. The potential underlying mechanism was correlated with activation of the mTORC1 pathway in GEnCs induced by CXCL7. Persistent PMP release finally resulted in functional and structural damage to the glomerular endothelium, an increase in permeability, urinary albumin leakage, and the progression of DN (Figure 9). Our findings herein may provide a potential drug therapeutic target for the prevention of early DN onset and progression.

Figure 9.

Schematic diagram showing that PMPs contributed to glomerular endothelial injury in early diabetic nephropathy. (A) In early DN, circulating PMPs released from activated platelets under hyperglycemia conditions infiltrated the glomerulus and bound to the glomerular endothelium, thereby promoting glomerular endothelial injury and albumin leakage. (B) PMP-derived CXCL7 induced a reduction of the NO levels and eNOS and SOD activities, production of ROS in GEnCs, and destruction of GEnC glycocalyx and cell junctions, and hyperpermeability in GEnCs by activating the mTORC1 pathway.

Disclosures

None.