Alvespimycin is identified as a novel therapeutic agent for diabetic kidney disease by chemical screening targeting extracellular vesicles

Abstract

Extracellular vesicles (EVs) are important mediators of intercellular communication and play key roles in the regulation of pathophysiological processes. In diabetic kidney disease (DKD), it has been reported that macrophages recruited in the mesangial region may play pathogenic roles through inducing local inflammation in glomeruli. We focused on EV-mediated crosstalk between mesangial cells (MC) and macrophages as a novel therapeutic target for DKD. EVs released from MC induced inflammation in macrophages and the effect was enhanced under high-glucose conditions. For discovering novel therapeutic agents which can inhibit such EV-mediated mechanisms, drug repositioning is considered as an effective tool. We established a unique screening strategy and screened agents to aim at maximizing their specificity and potency to inhibit EV mechanisms, along with minimizing their toxicity. We succeeded in identifying alvespimycin, an HSP90 inhibitor. Treatment of diabetic rats with alvespimycin significantly suppressed mesangial expansion, inflammatory gene activation including macrophage markers, and proteinuria. The inhibitory effect on EV uptake was specific to alvespimycin compared with other known HSP90 inhibitors. MC-derived EVs are crucial for inflammation by intercellular crosstalk between MC and macrophages in DKD, and alvespimycin effectively ameliorated the progression of DKD by suppressing EV-mediated actions, suggesting that EV-targeted agents can be a novel therapeutic strategy.

Keywords: Diabetic kidney disease; Drug screening; Extracellular vesicle; Intraglomerular crosstalk; Macrophages; Mesangial cells.

Fig. 1

Effects of mesangial cell-derived EVs upon macrophages. (a) Evaluation of extracted EVs from mesangial cells by nanoparticle tracking analysis. (b) Evaluation of extracted EVs from mesangial cells by nanoparticle counts. (particle size 30–120 nm, n = 5). (c) NF-κB activity in macrophages stimulated with mesangial cell-derived EVs isolated by polymer method; ExoQuick and phosphatidylserine affinity method; MagCapture (n = 8). Graph data are mean ± s.e.m. N.S., not significant; EQ, ExoQuick; M-Cap, MagCapture. *P < 0.05, ***P < 0.001 versus control.

Fig. 2

EVs derived from mesangial cells are endocytosed by macrophages in vitro and in vivo. (a) Representative images of RAW264.7 macrophage uptaking DiO-labeled EVs. DiO-EV were extracted from DiO-labeled mesangial cells. Macrophages are stained with red fluorescent dye (CytoTrace Red CMTPX) and culture medium is labeled with blue water-soluble, cell-impermeant polar tracer (Cascade blue hydrazide). White-framed squares are magnified images of inside the cells. (b) Positivity of DiO-EV in RAW264.7 macrophages evaluated by FCM was reduced by cytochalasin D, an endocytosis inhibitor. N.C., untreated negative control; P.C., DiO-EV treated positive control; CytoD, DiO-EV and cytochalasin D-treated. (c) Percentages of DiO-positive macrophages / all macrophages in the peripheral blood of non-STZ and STZ-mice were evaluated by FCM at an hour after DiO-EV injection (n = 4–6). Graph data are mean ± s.e.m. Mϕ, macrophage; STZ, streptozotocin. *P < 0.05.

Fig. 3

Effects of EVs isolated from high-glucose and low-glucose conditioned-mesangial cells upon macrophages. (a) Particle counts of EVs derived from high-glucose (triangles, HG-EV) and low-glucose (squares, LG-EV) conditioned-mesangial cells (n = 6). (b) NF-κB activation evaluated by SEAP-reporter in macrophages stimulated with HG-EV or LG-EV (n = 4–6). (c) Expressions of TNF-α and IL-1β mRNA by real-time PCR in macrophages stimulated with HG-EV and LG-EV (n = 5–6). Graph data are mean ± s.e.m. N.S., not significant; low glucose: 5.6 mM, high glucose: 25 mM. *P < 0.05, **P < 0.01.

Fig. 4

EV-targeted strategy for exploring novel drug candidates for diabetic kidney disease. Schema of the part of our drug screening assay. EVs extracted from the supernatant of cultured mesangial cells were added to macrophages, and reacted with each compound. Flowchart of high-throughput drug screening strategy targeting EV is shown. From the 3267 compounds in the validated library, candidate drugs are selected through the indicated 5 steps. LPS, lipopolysaccharide.

Fig. 5

Results of each step of the drug screening (Step 1–5). (a) Step 1. Compounds whose inhibition rate of NF-κB pathway induction by MC-EV ≧ 40% were selected. (Grey bars; 453 compounds). (b) Step 3A. Compounds which showed dose-dependent NF-κB inhibition rate were selected. Representative compounds are indicated. Whole data are listed in the Supplementary Table. (c) Step 3B. Compounds which showed cell viability higher than 60% after treatment were selected. (Grey bars; 182 compounds). (d) Step 4. Compounds which specifically inhibit EV-induced inflammation compared to nonspecific inflammation such as LPS. (Inhibition ratio [EV / LPS] > 1.5) Grey bars; 44 compounds. LPS, lipopolysaccharide. (e) Step 5. Final step was to evaluate EV inhibitory effect and integrated with the results of NF-κB inhibition rate. Open circles: HSP90-inhibitory effect-exerting compounds, arrowhead: Alvespimycin.

Fig. 6

Protective effects of Alvespimycin on renal pathogenesis in STZ-induced diabetic rats. (a) Representative images of PAS-stained kidney sections from nondiabetic control with vehicle (non-STZ + Veh), nondiabetic control treated with alvespimycin (non-STZ + Alv), STZ rats treated with vehicle (STZ + Veh) and STZ rats treated with alvespimycin (STZ + Alv) for 6 weeks. Scale bars = 20 µm. (b) Evaluation of the average area of the mesangial region of the randomly selected 10 glomeruli. (c) Evaluation of urinary protein level between the four groups. (n = 4) (d) mRNA expression of inflammatory genes (IL-1β, TNF-α) and macrophage markers (CD68, CD11b) in the kidney glomeruli. (n = 4–7) STZ, streptozotocin; Alv, alvespimycin; ACR, albumin creatinine ratio. *P < 0.05, **P < 0.01.

Fig. 7

Flowcytometric evaluation of the effect of alvespimycin on EV uptake inhibition. (a) Alvespimycin inhibits DiO-stained EV uptake in macrophages. N.C., untreated negative control; P.C., DiO-EV treated positive control; CytoD: DiO-EV and cytochalasin D-treated; Alv, DiO-EV and alvespimycin-treated. (b) Alvespimycin exerts significantly higher inhibitory effect of EV-uptake than other HSP90 inhibitors. T-210623, T-210751 and T-210582 are HSP90 inhibitors shortlisted in the final candidates of the screening. (n = 4–6) Alv, alvespimycin; CytoD, cytochalasinD; Pim, pimitespib. *P < 0.05.