Exocytosis-Mediated Urinary Full-Length Megalin Excretion Is Linked With the Pathogenesis of Diabetic Nephropathy

Diabetic nephropathy (DN) is the leading cause of end-stage renal disease (1) and increases morbidity and mortality in patients with type 2 diabetes mellitus (T2DM) (2). Phenotypic changes in proximal tubule epithelial cells (PTECs) are the initial signs of DN (3), and the consequent tubulointerstitial damage plays a crucial role in the progression to end-stage renal disease (4). Albuminuria has been used as a conventional biomarker for DN, but its clinical relevance as a surrogate marker has been controversial (5). Thus, efficient biomarkers associated with the development and progression of DN have not been established.

Megalin, a large (∼600 kDa) glycoprotein member of the LDL receptor family (6), plays a pivotal role in the endocytosis of diverse glomerular-filtered substances into PTECs (7,8). We have recently established that megalin mediates autolysosomal dysfunction in PTECs by taking up pathogenic ligands, leading to the development of tubuloglomerular alterations in a high-fat diet (HFD)–induced diabetes model (9). Advanced glycation end products (AGEs) are also filtered by glomeruli and reabsorbed by PTECs, causing glycotoxicity (10), and megalin is involved in the uptake of filtrated AGEs by PTECs (11). Previously, we established a sandwich ELISA to measure the ectodomain (A-megalin) and full-length (C-megalin) forms of urinary megalin using specific monoclonal antibodies against the amino-terminal and carboxy-terminal of megalin, respectively. Compared with normal control subjects (NCs), urinary C-megalin excretion levels in T2DM patients were found to be elevated and positively correlated with the development and progression of DN (12). In addition, in our recent longitudinal analysis of normoalbuminuric patients with T2DM, measurement of urinary C-megalin was found to be useful for predicting the early progression of DN (T.Ii., M.Ho., H.Ka, A.S., unpublished data). However, the molecular mechanism of urinary C-megalin excretion is not understood clearly, although it is known to be excreted in association with extracellular vesicles (EVs) (12).

EVs are nano-sized membrane vesicles (30–2,000 nm in diameter) that include mainly exosomes, microvesicles, and apoptotic bodies, which are either excreted by exocytosis or direct budding and blebbing of plasma membranes into extracellular spaces (13,14). Exosomes (30–130 nm) are excreted from almost all types of cell upon fusion of multivesicular bodies (MVBs) with cell membranes, and these are being studied extensively for the discovery of potential biomarkers in different diseases (15).

Hence, here we studied urinary EVs (UEVs) in connection with their C-megalin content and PTEC injury during the progression of DN.

Fresh second urine samples were collected from adult outpatients with T2DM (n = 56) at Niigata University Medical and Dental Hospital from 2013 to 2016 and from NC volunteers (n = 19) who participated in public medical examinations at Co-op Chemicals Co., Ltd. (Niigata, Japan) from 2014 to 2016, with written informed consent, based on the approval of the ethics committee of Niigata University, performed in accordance with the principles embodied in the Declaration of Helsinki. UEVs were collected from protease inhibitor–treated urine samples (16). As described previously (12,16), the urine samples were centrifuged at 17,000g for 15 min at 4°C, followed by high-speed centrifugation of the supernatant at 200,000g for 1 h at 4°C using a P40ST rotor and Himac CP-80β ultracentrifuge (Hitachi-Koki Co., Ltd.). UEVs in the second pellet were lysed or suspended in PBS, depending on the purpose, and standardized by creatinine levels that were measured using an enzymatic method. The protein concentration of EVs was determined by a bicinchoninic acid protein assay (Pierce). Physical examination and blood test data of T2DM patients were collected during the same visit as their urine samples and those of NCs were obtained <1 year before urine collection. The albumin-to-creatinine ratio (ACR; in milligrams per gram) was determined using the collected urine samples.

For observation of UEVs, the urinary second pellet was suspended in PBS, applied to a carbon-coated 200-mesh copper grid (Nisshin EM Co., Ltd.), fixed with 4% paraformaldehyde phosphate buffer solution (Gibco) for 10 min, washed with distilled water, contrasted with 2% uranyl acetate (Wako) for 10 min, and visualized under a JEOL 1400 Transmission Electron Microscope (JEOL USA, Inc.) at different magnifications at 100 KV. For EV-bound beads, the pellet was fixed in a mixture of 4% paraformaldehyde, 0.1% glutaraldehyde in cacodylate-HCl (Sigma-Aldrich) buffer at pH 7.2 for 45 min and postfixed with 1% osmium tetroxide (Sigma-Aldrich) for 1 h. Subsequently, it was dehydrated in an ethanol series, treated with 400 µL propylene oxide (Sigma-Aldrich), and exposed to 50% epoxy resin (epon; Polysciences, Inc.) for 30 min and to 75% resin for 1 h at room temperature. Further, it was exposed to 100% resin and dried in an oven. The pellet was sliced into ultrathin sections (80 nm thick), treated with tannic acid, and stained with lead acetate.

Nanoparticle tracking analysis (NTA) was performed using a NanoSight LM10 instrument (NanoSight Ltd.) equipped with NTA 2.3 software 3.0 at a particle concentration within the optimal working range of the system (17). EVs were collected from creatinine-standardized urine and from culture media as described above. Size distributions are presented as the mean and SEM of three video recordings per sample.

The diluted EV fraction was fixed on a copper grid as described above and then blocked using 0.05 mol/L glycine/PBS and 1% BSA/PBS, respectively, for 10 min each. The grid was rewashed and incubated overnight with a primary antibody diluted in 0.02% Triton X-100 in PBS. The grid was washed with PBS and incubated with 5 nm gold-conjugated anti-mouse IgG antibody G7527 (1:25; Sigma-Aldrich) for 60 min at room temperature. The grid was washed again, fixed with 2.5% glutaraldehyde for 5 min, washed once more, and embedded with 0.5% uranyl acetate (18). The primary antibodies and their dilutions were as follows: mouse monoclonal anti-megalin antibody C-25 (22.2 µg/mL, final concentration) (12), mouse monoclonal anti-CD81 antibody sc-7637 (1:50; Santa Cruz Biotechnology), and mouse monoclonal anti-podocalyxin antibody 22A4 (15 µg/mL, final concentration) (19).

Immunoblotting was carried out as described previously (9). The primary and secondary antibodies and their dilutions are mentioned in the Supplementary Data.

Exosome-enriched UEVs were purified from the total population of UEVs using a human CD63 or CD81 isolation/detection kit (Invitrogen) according to the manufacturer instructions. After purification, the beads were either processed for transmission electron microscopy (TEM) as described above or lysed by adding 2× sample buffer for immunoblot analysis. For comparative analysis of C-megalin content in exosomes, we prepared the total UEVs from NCs (n = 11) and T2DM patients (n = 33), and suspended in variable volumes of PBS, depending on their urinary creatinine levels. Exosomes were purified from the same volume of those UEV samples using an excess amount of anti-CD81 antibody-coated paramagnetic beads, and the beads were lysed by adding the same volume of 2× sample buffer, the same volume of samples was loaded in each well of an SDS-PAGE gel, and immunoblotting was performed against C-megalin and tumor susceptibility gene 101 (TSG101).

C-megalin–containing UEVs were purified using recombinant rat receptor–associated protein (RAP), a high-affinity Ca²⁺-dependent ligand of megalin (20), which was prepared using a prokaryotic expression system as a fusion protein with glutathione S-transferase (GST), as described previously (20). The specificity of RAP binding to megalin in UEVs was verified by ligand blotting analysis, as described previously (21), using GST-RAP as a ligand of megalin, which was further detected using an anti-GST antibody. For purification of C-megalin–containing UEVs, glutathione Sepharose 4B beads (Amersham Bioscience) were incubated with recombinant GST-RAP or GST in the presence of 1 mmol dithiothreitol for 1 h at room temperature. The beads coated with the recombinant proteins were incubated with UEVs in the presence of binding buffer containing 10 mmol HEPES and 150 mmol NaCl with or without Ca²⁺ overnight at 4°C. The next day, the beads were collected and washed with the respective binding buffer and used for further experiments.

Total RNA was isolated from GST-RAP–captured UEVs and cultured cells using a Total Exosome RNA and Protein Isolation Kit (ThermoFisher Scientific) and GenElute Mammalian Total RNA Miniprep Kit (Sigma-Aldrich), respectively. MicroRNA (miRNA)-192 and miRNA-16 were both quantified using a TaqMan MicroRNA Assay (Applied Biosystems), and their relative quantities were calculated with the 2^-ΔΔCt method (22) using the normalization control RNA RNU6. See the Supplementary Data for further details.

AGE-BSA was prepared with a modified method that has been described previously (23). Briefly, fatty acid–free (FAF)-BSA (Calbiochem) (200 µg/mL in PBS) was treated with 33 mmol/L methylglyoxal (Sigma-Aldrich), incubated at 37°C for 7 days, dialyzed three times with PBS, and filtered through a 0.2-µm filter.

We cultured immortalized rat proximal tubule cells (IRPTCs) (a gift from Professor Julie R. Ingelfinger, Massachusetts General Hospital, Boston, MA) as described previously (9). The cells were allowed to grow for 24 h; were serum starved for 12 h; and were incubated with PBS (vehicle), FAF-BSA (1 mg/mL, final concentration), or AGE-BSA (1 mg/mL) for 48 h. Then, culture media were collected, and EVs excreted by the cells were isolated using a differential ultracentrifugation method optimized for cultured cells (24). The exosome excretion inhibitor GW4869 (4 μmol/L; Sigma-Aldrich) and lysosome inhibitors bafilomycin A1 (50 nmol/L; Sigma-Aldrich) and chloroquine phosphate (50 μmol/L; Wako) were also used for experiments with the cells.

IRPTCs were cultured in the medium described above. Subsequently, the cells were starved for 12 h and treated for 24 h with the substances described above. The cells were stained with LysoTracker Deep Red (Life Technologies) and NucBlue Live Cell Stain (Life Technologies), as described previously (9). LysoTracker Deep Red–positive vacuoles and cell surface area were evaluated using Image-Pro Plus version 7.0 in 20 cells selected randomly for each condition.

C57BL/6J mice (n = 12) were fed an HFD (60% of total calories from fat; Research Diets) by paired feeding with those (n = 12) fed a normal-fat diet (NFD) (10% of total calories from fat; Research Diets) from 11 weeks of age for 4 weeks. Urine was collected for 72 h in metabolic balance cages, after which it was used for collecting UEVs by the ultracentrifugation method described above. Briefly, collected urine samples were subjected to ultracentrifugation at 17,000g twice and then at 100,000g for 60 min. The resulting pellets were suspended in PBS and used for NTA as described above. Also, urine samples were collected for 24 h, and UEVs were prepared from urine collected for 48 h from mice fed an HFD or NFD for 6 weeks were used for immunoblotting using anti-megalin C-25 antibody as described above after urinary volume standardization.

The data were analyzed by one-way ANOVA followed by Bonferroni and Tukey multiple-comparisons tests using GraphPad Prism statistical software version 5 (GraphPad Software, Inc.) and SPSS statistical software version 22 (SPSS, Inc.). The level of significance was P < 0.05. For animal studies, comparisons between two groups were made by Student t test (two-tailed). For single-regression analysis, Spearman rank correlation tests were performed using SPSS statistical software version 22 (SPSS, Inc.).

We collected urine from NCs (n = 19) and patients with T2DM (n = 56) at different albuminuric stages who were classified into the following three groups depending on their ACR: normoalbuminuric T2DM (Nor-DM) with ACR <30 mg/g (n = 20); microalbuminuric T2DM (Mi-DM) with ACR ≥30 and <300 mg/g (n = 17); and macroalbuminuric T2DM (Ma-DM) with ACR ≥300 mg/g (n = 19). UEVs were collected using the classic differential centrifugation method (12,16). The NC criteria and clinical parameters of the NCs and T2DM patients are shown in Supplementary Tables 1 and 2, respectively.

In the differential centrifugation method, UEVs are expected to be enriched in the urinary second pellet. To characterize the UEVs, we visualized the second pellet by TEM (Supplementary Fig. 1A) and investigated their frequency distribution by NTA using NanoSight (Supplementary Fig. 1B). NanoSight analysis confirmed that the second pellet was exosome enriched, as the mode size of UEVs (113 nm) was within the size range of exosomes (15). Immunoelectron microscopy against the exosomal marker CD81 (25) (Supplementary Fig. 1C) and immunoblot analysis against exosomal markers, such as CD81, apoptosis-linked gene 2–interacting protein X (Alix) (26), and TSG101 (26) further confirmed the presence of exosomes in the second pellet and their scarcity in the first pellet (Supplementary Fig. 1D). Further, NTA revealed a positively correlated increase in the total number of UEVs in creatinine-standardized urine samples of T2DM patients with the development and progression of DN (Fig. 1A). Because of the lack of a comprehensive size range of different kinds of UEVs, we considered the smaller UEVs (<130 nm in diameter) preferentially as exosomes and the relatively larger ones (≥130 nm in diameter) as microvesicles. According to this classification, further NTA data showed that increased excretion of both exosome-like small UEVs as well as microvesicle-like large UEVs from patients with T2DM was positively correlated with the progression of the albuminuric stages (Fig. 1B and C). On single-regression analysis, the number of total, small, and large UEVs was also found to be positively correlated with the ACR (Fig. 1D–F) and negatively correlated with the estimated glomerular filtration rate (eGFR) in all the subjects (Fig. 1G–I). However, the number of total UEVs was not correlated with the age, BMI, systolic blood pressure, or HbA_1c levels of the subjects (Supplementary Fig. 2A–D), indicating their marginal impact on the excretion of UEVs in the subjects.

UEVs are known to contain the endocytic receptors megalin and cubilin, in addition to other transporter or channel proteins (12,16,25). C-megalin immunoblotting against urinary first and second pellets confirmed UEV-mediated C-megalin excretion (Fig. 2A). Since creatinine-normalized total UEV protein concentration was correlated significantly with creatinine-normalized total UEV number (Spearman rank correlation coefficient = 0.678, P < 0.001), we carried out standard immunoblotting using the same amount of UEV proteins to evaluate megalin or other tubular segment marker (16,25) protein content per total UEV protein in different sample types (Fig. 2B). We found that megalin protein content per total UEV protein appeared to increase along with the progression of the albuminuric stages in DN, but the other tubular segment markers did not (Fig. 2B). The protein content of CD9, an EV marker (25), (as well as CD81 and TSG101, more specific exosome markers) per total UEV protein was not also changed along with the development and progression of DN (Fig. 2B and Supplementary Fig. 3A–C). Thus, using CD9 as an EV internal control, the densitometry ratio between megalin protein content and CD9 protein content in UEVs (i.e., relative megalin protein content per UEV) was evaluated, showing that the ratio increased along with the progression of the disease (Fig. 2C). On the other hand, the densitometry ratio between other tubular segment marker proteins and CD9 protein content did not change along with the progression of DN (Supplementary Fig. 3D–H). The densitometry ratio between megalin protein content and CD9 protein content was also correlated with the ACR (Fig. 2D) and eGFR (Fig. 2E) but not with the age, BMI, systolic blood pressure, or HbA_1c levels of the subjects (Supplementary Fig. 2E–H). These data indicate that a gradual increase in UEV-mediated C-megalin excretion is characteristically associated with the progression of DN.

When we calculated total C-megalin excretion via UEVs by multiplying the relative C-megalin content per UEV with the corresponding numbers of UEVs, we found that it increased gradually along with the development and progression of DN from the normoalbuminuric stage (Fig. 2F), which is consistent with our previous data of direct urinary C-megalin ELISA measurement in T2DM patients (12), and was positively correlated with the ACR (Fig. 2G) and negatively correlated with the eGFR (Fig. 2H) in single-regression analysis.

As megalin is an endocytic receptor, it is expected to be excreted more via exosomes rather than other types of EVs. We purified exosomes from the total population of UEVs using paramagnetic beads coated with antibodies against CD63, a classic exosomal marker (18), and CD81, a more specific exosomal marker (26), and visualized the beads by TEM to evaluate purification efficiency followed by immunoblotting. TEM showed that the captured UEVs were small and within the size range of exosomes (<130 nm) (Fig. 3A). Immunoblot analysis of the lysates of corresponding beads further confirmed the presence of C-megalin within the captured exosomes, indicating exosome-mediated C-megalin excretion (Fig. 3B and C). We prepared UEVs from the same volume of creatinine-normalized urine samples, applied them to an excess of beads coated with antibodies against CD81, and analyzed purified urinary exosomes by immunoblotting. Megalin content in the purified exosomes was confirmed to increase gradually along with the progression of DN (Fig. 3D and E).

Immunoelectron microscopy against C-megalin and the podocyte marker podocalyxin (Fig. 3F), followed by size distribution analysis of C-megalin–positive and podocalyxin-positive UEVs from NCs and patients with T2DM, indicated that excretion of urinary C-megalin and podocalyxin was mediated preferentially via exosomes and microvesicles, respectively (Fig. 3G and H).

Next, we purified C-megalin–containing UEVs (mostly exosomes) from the total population of UEVs using GST-RAP–coated glutathione beads in the presence of Ca²⁺, because RAP is a high-affinity Ca²⁺-dependent ligand of the extracellular domain of megalin (27). The specificity of GST-RAP binding to megalin in the population of UEVs was confirmed by ligand-blotting analysis (Supplementary Fig. 4A). By TEM of the glutathione beads (Supplementary Fig. 4B) and subsequent immunoblot analysis of the lysates of the corresponding beads (Supplementary Fig. 4C), we showed that the beads effectively and specifically purified C-megalin–containing UEVs.

Exosomes contain large amounts of functional small RNAs including miRNAs for cell-cell communication (28). Therefore, we assessed levels of miRNA-192, which is one of the best characterized miRNAs expressed preferentially and highly in the kidney, in GST-RAP–purified C-megalin–containing UEVs from NCs and patients with T2DM. In comparison with NCs, RNU6-standardized miRNA-192 levels were lower in the early phase of DN and tended to increase with the progression of DN (Fig. 3I). However, miRNA-16, an endogenous control miRNA, did not change with disease progression (Fig. 3J), indicating that there are DN stage-related differences in miRNA content in C-megalin–containing UEVs with putative pathophysiological relevance.

To establish the molecular mechanism underlying the increased excretion of UEVs in patients with T2DM, we treated IRPTCs with FAF-BSA as a protein overload model and AGE-BSA as a glycated protein load model. Both FAF-BSA and AGE-BSA are the endocytic ligands of megalin (7,8). Thereafter, we collected IRPTC-derived EVs from the conditioned media. As with UEVs, TEM and NanoSight were used to characterize the IRPTC-derived EVs. TEM revealed that the majority of EVs had a typical size of 50–130 nm, which identified them as exosomes isolated from different cell lines (Fig. 4A and B) (15,29). By NTA, those EVs were confirmed to be exosome enriched as evidenced by their mode size of 117 nm (Fig. 4B). Total EV protein concentration data indicated that EV excretion was increased by treatment of the cells with FAF-BSA and were further elevated with AGE-BSA, compared with vehicle (PBS) treatment (Fig. 4C). These findings were further confirmed by NTA (Fig. 4D). It also confirmed that the number of exosome-like small EVs (<130 nm) and microvesicle-like large EVs (≥130 nm) was increased by FAF-BSA treatment and further elevated by AGE-BSA treatment (Fig. 4E and F).

C-megalin immunoblotting followed by densitometry analysis revealed a similar increase in C-megalin excretion per the content of exosomal marker Alix when the cells were treated with FAF-BSA and AGE-BSA (Fig. 5A and B), whereas the cellular expression of megalin was not changed by either treatment (Fig. 5C and D). To prove exosome-mediated C-megalin excretion by such treatments, an established exosome excretion inhibitor, GW4869 (30), which acts on neutral sphingomyelinase to inhibit exosome production, was added to IRPTCs, leading to decreased EV protein concentration in the culture media, even when the cells were treated with AGE-BSA (Fig. 5E). As a result of exosome inhibition, EV-mediated excretion of C-megalin, standardized by Alix, was also inhibited, indicating that increased C-megalin excretion by AGE-BSA treatment is mediated via exosomes (Fig. 5F and G), whereas the cellular expression of total megalin was not changed (Fig. 5H and I). Exosome inhibitor treatment along with FAF-BSA also showed similar results (data not shown). Increased mRNA expression of platelet-derived growth factor subunit B and monocyte chemoattractant protein 1, which are profibrotic factors, as well as kidney injury molecule 1, a PTEC injury marker protein, in IRPTCs treated with AGE-BSA was detected by exosome inhibition (Fig. 5J–L), indicating that increased exosome release by PTECs might be a protective mechanism of the cells.

It is generally accepted that, in PTECs, filtrated albumin undergoes megalin/cubilin-mediated endocytosis and is transferred to lysosomes for further degradation (8). We performed LysoTracker Deep Red staining for lysosomal morphology, which showed an accumulation of giant lysosomes (>1 µm) around the nucleus of IRPTCs treated with AGE-BSA, but not FAF-BSA, for 24 h (Fig. 6A). On quantification of the giant lysosomes standardized by cellular area, we found a significant increase in their number in AGE-BSA–treated IRPTCs, indicating lysosomal dysfunction (Fig. 6B). In correlation with the increased accumulation of giant lysosomes, the augmented expression of lysosomal-associated membrane protein 1 was also found in AGE-BSA–treated cells, indicating an increased number and size of lysosomes, a hallmark of lysosomal dysfunction (Fig. 6C and D) (31,32). Furthermore, AGE-BSA treatment for 48 h caused increased MVB formation, as indicated by the augmented cellular expression of the endosomal sorting complex required for transport–associated proteins TSG101 (Fig. 6E and F) and Alix (Fig. 6G and H) (33). Indeed, megalin-containing vesicles were found to be recruited into MVBs in those cells treated particularly with AGE-BSA (and also with FAF-BSA) for 48 h (Supplementary Fig. 5). Thus, we speculated that AGE-BSA treatment leads to apparent lysosomal dysfunction in IRPTCs within 24 h, which thereafter results in increased MVB formation and excretion of exosomes containing C-megalin.

To verify the lysosomal dysfunction–mediated increase of exosomal C-megalin excretion, we treated IRPTCs with two potent lysosomal inhibitors with different modes of actions, namely bafilomycin A1, an inhibitor of vacuolar-type H⁺-ATPase (34), and chloroquine phosphate, a lysosomotropic agent that increases lysosomal pH (35), and collected excreted EVs from the culture media. We found an increase in the cellular expression of TSG101 (Fig. 7A and B) and Alix (Fig. 7C and D) and increased C-megalin excretion per exosome (Fig. 7E and F) but not overall cellular C-megalin expression (Fig. 7G and H), indicating that lysosomal dysfunction caused an increase in MVB formation and exosomal C-megalin excretion.

Megalin mediates autolysosomal dysfunction in PTECs, which leads to tubuloglomerular alterations in mice fed an HFD (9). In this obesity-related diabetes model, we found an increased number of total, small, and large UEVs compared with mice fed an NFD (Fig. 8A–C). Also, urinary C-megalin excretion increased via UEVs in HFD-fed mice compared with NFD-fed mice (Fig. 8D and E).

Previously, we established a urinary C-megalin ELISA that was potentially useful for the early diagnosis and evaluation of the severity of DN in patients with T2DM (12). Here, we established that elevated urinary C-megalin excretion in patients with T2DM is associated with an increased number of urinary exosomes as well as C-megalin content per exosome. Besides, patients with T2DM may also excrete an increased number of urinary microvesicles, probably partly from podocytes (17,19).

Low–molecular-weight protein markers for PTEC injury, such as α₁-microglobulin, β₂-microglobulin, and liver-type fatty acid–binding protein, are filtered by glomeruli and reabsorbed by PTECs via megalin (8). Not only physiologically important substrates, including insulin, albumin, and vitamin D–binding protein, but also some substances, such as AGEs, myeloma light chain, and fatty acid–rich proteins, have been identified as toxic ligands of megalin, leading to PTEC injury (9,36).

Serum AGE levels reflect the severity of DN (37) and are reportedly involved in the pathogenesis of DN (38). High-concentration protein overload in cultured PTECs was found to lead to lysosomal dysfunction (39,40); however, the protein concentrations used (8 and 20 mg/mL, respectively) in these experiments were much higher than in ours (1 mg/mL). Although protein overload with FAF-BSA did not result in significant lysosomal dysfunction compared with vehicle-treated cells in the current study, overload with AGE-modified BSA caused apparent lysosomal dysfunction, again confirming AGE-mediated cytotoxicity in PTECs via megalin (41,42). Hence, the increased excretion of EVs by FAF-BSA–treated IRPTCs in this study may be associated with increased endocytosis.

Recently, we reported that megalin-mediated autolysosomal dysfunction in PTECs is primarily important for the development of kidney disease in an HFD-induced diabetes model (9). In the current study, we found that exocytosis-mediated urinary C-megalin excretion increased along with the development and progression of DN and that lysosomal dysfunction in AGE-BSA–treated IRPTCs leads to an increase in MVB formation and exocytosis-mediated C-megalin excretion. Indeed, we found that urinary C-megalin excretion increased via UEVs in HFD-fed mice compared with NFD-fed mice. As exosomes are formed within the endosomal network and late endosomes (MVBs) are targeted to fuse either with lysosomes for further degradation or with the plasma membrane for excreting exosomes, lysosomal dysfunction should result in increased MVB production, megalin recruitment in MVBs, and finally the release of exosomes containing megalin. Thus, urinary C-megalin excretion via exocytosis is linked with the megalin-mediated pathogenesis underlying the development and progression of DN.

Exosomal miRNAs and exosome-mediated miRNA transfer are increasingly recognized as critical players in gene regulation and various diseases including DN (28,43). After treatment with transforming growth factor-β, miRNA-192 expression is increased in cultured mesangial cells (44) but is decreased in PTECs (22). miRNA-192 was also found to be increased in human kidney in both hypertensive nephrosclerosis (45) and IgA nephropathy (46). Compared with NCs, PTEC-derived C-megalin–containing exosomes in normoalbuminuric T2DM patients contained less miRNA-192, which might be a protective mechanism against miRNA-192 loss in PTECs. In contrast, miRNA-192 might be produced more in PTECs with the progression of DN, resulting in increased miRNA-192 excretion via exosomes in the advanced stages of DN. Exosome inhibition in cultured IRPTCs also increased pathological reactions, indicating that exosome excretion may be a beneficial phenomenon for PTECs. It remains to be studied whether C-megalin–containing exosomes have a role in the transfer of miRNA-192 to cells in the downstream tubules or urinary tract.

In conclusion, protein overload or AGE-mediated PTEC injury should increase exocytosis-mediated urinary C-megalin along with the progression of DN. Thus, megalin-mediated quantitative or qualitative protein metabolic load to residual functioning nephrons should be the mechanism underlying the increase in urinary C-megalin excretion in T2DM patients, which could be a novel biomarker associated directly with the pathogenesis of DN.

Acknowledgments. The authors thank Dr. Julie R. Ingelfinger (Massachusetts General Hospital) for the gift of the experimental material described in 2research design and methods and Dr. Masaaki Komatsu, Dr. Reika Kaneko, Akiko Seino, Nanako Sugita, Koichi Komochi, Masaaki Nameta, and Kazumasa Sato at Niigata University for their technical assistance. The authors also thank ThinkSCIENCE for help in the preparation of the manuscript.

Funding. This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan to M.Ho. (grant 26860630) and A.S. (grant 26461216) and by Grants-in-Aid for the Practical Research Project for Renal Diseases from the Japan Agency for Medical Research and Development.

Duality of Interest. S.K. and A.S. received research grants from DENKA SEIKEN Co., Ltd. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. S.D. designed the study, researched data, and wrote the manuscript. S.K. and P.S. researched data, contributed to the discussion, and reviewed the manuscript. M.Ho., T.Is., R.K., H.Ka., T.Ii., S.G., and Y.S. collected clinical data and urine samples, contributed to the discussion, and reviewed the manuscript. Y.Y. and T.O. performed NTA of EVs and reviewed the manuscript. K.T. and Y.Hig. contributed to the animal experiments. M.Ha., H.Ku., and Y.Hir. contributed to the experiments for podocalyxin. I.N. contributed to the discussion and reviewed the manuscript. A.S. designed the study and wrote the manuscript. S.D. and A.S. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented in abstract form at Kidney Week 2015—American Society of Nephrology 48th Annual Meeting, San Diego, CA, 3–8 November 2015.