Three-dimensional culture of MSCs produces exosomes with improved yield and enhanced therapeutic efficacy for cisplatin-induced acute kidney injury

Exosomes derived from mesenchymal stem cells (MSC-exos) have been demonstrated with great potential in the treatment of multiple human diseases including acute kidney injury (AKI) by virtue of their intrinsic cargoes. However, there are major challenges of low yield and the lack of an established biomanufacturing platform to efficiently produce MSC-exos, thereby limiting their therapeutic application. Here, we aimed to establish a novel strategy to produce MSC-exos with a hollow fiber bioreactor-based three-dimensional (3D) culture system and evaluate the therapeutic efficacy of 3D-exosomes (3D-exos) on AKI. Mesenchymal stem cells (MSCs) were isolated from fresh human umbilical cord and cultured in two-dimensional (2D) flasks. 2 × 108 MSCs were inoculated into the hollow fiber bioreactor for 3D culture. The culture supernatants were collected every 1 or 2 days for isolating exosomes. Exosomes from 2D (2D-exos) and 3D cultures were characterized by transmission electron microscopy, nanoparticle tracking analysis, and western blotting analysis of exosome markers. The yield of exosomes from 2 × 108 MSCs seeded in 2D and 3D culture system was compared, based on protein quantification. The therapeutic efficacy of 2D-exos and 3D-exos was investigated in a murine model of cisplatin-induced AKI in vivo and in vitro. 3D culture did not significantly change the surface markers of MSCs, as well as the morphology, size, and exosomal markers of 3D-exos when compared to those of 2D-exos. Compared with conventional 2D culture, the 3D culture system increased total exosome production up to 19.4-fold. 3D-exos were more concentrated in the harvested supernatants (15.5-fold) than 2D-exos, which led to a higher exosome collection efficiency of 3D culture system. In vivo, both 2D-exos and 3D-exos significantly alleviated cisplatin-induced murine AKI evidenced by improved renal function, attenuated pathological changes of renal tubules, reduced inflammatory factors, and repressed T cell and macrophage infiltration. Impressively, 3D-exos were more effective than 2D-exos. Moreover, 3D-exos were taken up by tubular epithelial cells (TECs) with improved efficiency, thereby exhibiting superior anti-inflammatory effect and improved viability of TECs in vitro. In summary, our findings demonstrate that the hollow fiber 3D culture system provides an efficient strategy for the continuous production of MSC-exos which has enhanced therapeutic potential for cisplatin-induced AKI.

Exosomes, a major class of extracellular vesicles (EVs) that are 30 to 150 nm in diameter, are released by almost all kinds of cells and play important roles in both physiological and pathological conditions [10]. They provide a short-to long-distance form of intercellular communication by shuttling bioactive molecules, including DNA fragments, mRNAs, non-coding RNAs, proteins, and lipids [10,11]. Very recently, remarkable advances in our understanding of exosomes have spurred a renewed interest in their utility as delivery vehicles for various therapeutic agents ranging from chemotherapeutics to gene therapy [12][13][14].
An increasing number of studies have demonstrated MSC-derived exosomes (MSC-exos) have innate therapeutic potential by virtue of their intrinsic cargoes. MSCexos have shown excellent efficacy in tissue repair and regeneration of many organs, including liver, lung, cartilage, myocardium, brain, spinal cord, and kidney [15][16][17][18][19][20][21][22]. Taking acute kidney injury (AKI) as an example, MSC-exos exert a series of renoprotective and regenerative effects through various mechanisms, including anti-inflammatory, immunomodulatory, anti-necroptosis, anti-apoptosis, and promoting cell proliferation [15,23]. To date, administration of MSC-exos effectively improves clinical outcomes in patients suffering from graft-versus-host disease and chronic kidney diseases [24,25]. In particular, it is shown that the therapeutic effect of MSC-exos is similar to that of MSCs [1], without the drawbacks of MSC-based therapy, such as unwanted differentiation of engrafted MSCs and embolism. Additionally, MSC-exos possess the characteristic of lower immunogenicity because they lack the antigens on the surface membrane [1,15]. Therefore, they are likely to represent a novel cell-free therapeutic agent for many human diseases.
However, preclinical and clinical research of exosomebased therapy requires large quantities of MSC-exos.
Generally, it requires a dose of 20-200 μg per mouse (10 9 -10 11 particles) to achieve biological outcomes [26][27][28]. For clinical testing, one patient needs approximately 100 μg/kg exosomes per treatment [8,25]. Meanwhile, unlike immortal cells lines, the expansion of MSCs is limited in culture and only cells within passage 6 can be used according to most studies [29,30]. Therefore, the low yield and the lack of an established biomanufacturing platform to produce sufficient MSC-exos remain major challenges, which currently limit their therapeutic application [8,15,31].
In the current study, we reported a novel strategy to produce MSC-exos continuously and efficiently in a hollow fiber bioreactor-based three-dimensional (3D) culture system. Furthermore, we compared the efficacy of the MSC-exos produced from this 3D culture system with those obtained from the conventional 2D flask culture in a murine model of cisplatin-induced AKI.

Materials and methods
Culture and identification of human umbilical cord mesenchymal stem cells (hucMSCs) Primary hucMSCs were donated by Shenzhen Wingor Biotechnology Co, Ltd. In brief, sterile and fresh human umbilical cords were rinsed twice with phosphatebuffered saline (PBS) supplemented with antibioticantimycotic (Gibco) to remove cord blood. Then, the umbilical cords were cut into 1-mm 3 -sized pieces and floated in configured media for traditional 2D flask culture. Culture media contained mixed (1:1 ratio) DMEM/ F12 (Gibco) and mesenchymal stem cell medium (Scien-Cell), supplemented with 5% fetal bovine serum (FBS; ScienCell) and 1% antibiotic-antimycotic (Gibco). Half of the medium was replaced every 3 days, and welldeveloped colonies appeared after 10 days. Early-passage hucMSCs (passages 2-6) were used for downstream experiments.
To determine the multidirectional differentiation potential of hucMSCs, the cells of passage 3 were seeded into six-well plates at a density of 1 × 10 5 cells/well. The culture media were replaced with osteogenic or adipogenic differentiation media (Cyagen Biosciences) in the next day and every 3 days thereafter. Three weeks later, Alizarin Red S and Oil Red O staining were performed to evaluate the extent of osteogenesis and adipogenesis, respectively. 5 × 10 5 hucMSCs were resuspended in a 15 mL polypropylene tube with 0.5 mL chondrogenic differentiation medium (StemCell technologies). It was capped tightly and centrifuged at 300×g for 5 min at room temperature, then the cap was gently loosened, and the tube was incubated at 37°C with 5% CO 2 . Chondrogenic differentiation medium was replaced every 3 days. Three weeks later, histological sections of the chondrogenic pellet were generated by fixing the pellets in 10% formalin for 30 min, following standard paraffin embedding methods and staining 6 μm sections with Alcian Blue/Nuclear Fast Red.

2D culture of hucMSCs
HucMSCs were cultured in T225 cm 2 flasks (Fig. 1a). The glucose concentration of 2D culture media was monitored every 12 h. The culture media were replaced, or cells were passaged when the glucose concentration reduced to 50%. 2 × 10 8 hucMSCs (passages 4-6) in 2D flasks were washed with PBS and cultured for an additional 48 h in serum-free media. The serum-free culture supernatants were collected for isolating 2D exosomes (2D-exos).

3D culture of hucMSCs
The hollow fiber bioreactor FiberCell System (C2011; FiberCell Systems Inc.) was adopted for 3D culture of hucMSCs. This 3D culture system consisted of a pulsatile perfusion pump, an oxygenator, a cartridge containing thousands of hollow fibers, a bottle of culture medium, and the connecting tube. The membrane material of hollow fiber was hydrophilic polysulfone with an internal diameter of approximately 200 μm and a molecular weight cutoff (MWCO) of 20 kDa at 50% retention. The surface area of the hollow fiber bioreactor was 3000 cm 2 within a volume of 20 mL.
Prior to cell culture, the external and internal spaces of 3D hollow fiber bioreactor were conditioned with PBS or culture media for circulating 24 h, respectively. Then, 2 × 10 8 hucMSCs within 15 mL suspension were seeded into the external space of 3D culture system. The culture media flowed through the internal space. The flow rate of the pulsatile perfusion pump was 22-28 times per minute.
After 24 h, almost all living cells had attached to the hollow fibers, and the culture media in the external space were replaced with serum-free media. The media in the internal space were still the same with 2D. The The general schematic of the hollow fiber bioreactor-based 3D culture system. The system was composed of a pulsatile perfusion pump, an oxygenator, a cartridge containing thousands of hollow fibers, a bottle of culture medium, and the connecting tube. c A general schematic of cross-sectional view through the bioreactor. d The photograph of our 3D culture system glucose concentration in the culture medium bottle was monitored every 12 h and used as a gauge for media replacement. The culture media were replaced immediately when the glucose concentration dropped by 50%. According to the glucose consumption of the 3D culture system, the serum-free supernatants in external space were collected every 24-48 h.

Trypan blue staining
Cell death in the 2D and 3D systems was determined by trypan blue staining. The cells of 2D system and the exfoliated cells from 3D system were suspended and stained with trypan blue solution (Servicebio). Then, the number of living and dead cells was counted by hemocytometry.

Isolation and purification of exosomes
The 2D and 3D supernatants were centrifuged at 2000×g for 20 min and 13,500×g for 30 min at 4°C to eliminate the cells and debris, followed by filtration with a 0.22-μm filter (Millipore) to remove microvesicles. Then, the supernatants were centrifuged at 200,000×g for 120 min (Type 70 Ti rotor, Beckman Coulter Optima L-80 XP) at 4°C to deposit 2D-exos or 3D exosomes (3D-exos). The exosome pellets were resuspended in PBS and filtered using a 0.22-μm filter once again. The resuspended exosomes were concentrated using 100 KDa MWCO (Millipore) at 4000×g for 30 min. Finally, purified 2D-exos and 3D-exos were harvested in 200-400 μL of PBS and stored at − 80°C.

Transmission electron microscopy (TEM)
The exosome pellets were fixed with 2.5% glutaraldehyde and post-fixed with 1% osmium for 1 h at room temperature. Samples were then dehydrated with a graded series of ethanol solutions (30%, 50%, 70%, 90%, and 100%) for 15 min each. All samples were infiltrated with and embedded in Epos 812 resin. After polymerization, the coverslips were removed from the bottom of the sample blockers then processed the samples by 70-nm-thick ultrathin sectioning. Every section was double-stained with uranyl acetate and lead citrate [32]. After air drying, specimens were viewed and photographed with TEM (FEI Tecnai G2 Spirit).

Nanoparticle tracking analysis (NTA)
The particle size of exosomes was measured by NTA with ZetaView PMX 110 (Particle Metrix). Samples were appropriately diluted using PBS, and NTA measurements were recorded and analyzed at 11 positions. The ZetaView system was calibrated using 110 nm polystyrene particles.

Comparisons of 2D and 3D exosomal production
The exosomes were quantified by protein quantification using the BCA protein assay kit, according to the manufacturer's instructions (ThermoFisher). To compare the yield of exosomes derived from 2 × 10 8 hucMSCs seeded in 2D and 3D culture systems, both the total output and the concentration of exosomes per mL of supernatant were assessed.

Animal models and therapeutic experiments
Adult male C57BL/6 N mice (21-24 g body weight) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd., China. Experiments described in this manuscript used protocol approved by the Animal Experimentation Ethics Committees of Southeast University (No. 20190410016). Mice were fed in a specific pathogen-free animal facility with a 12-/12-h light/dark cycle and free access to food and water. For the cisplatin (Cis)-induced injury, a single dose of 18 mg/kg cisplatin was injected intraperitoneally. At 24 h and 48 h after cisplatin injection, mice were injected intravenously with PBS, 2D-exos (100 μg), or 3D-exos (100 μg) (n = 6-7, respectively) ( Fig. 5a). Mice were euthanized at 96 h post cisplatin injection.

Immunohistochemistry and immunofluorescence staining
For immunohistochemistry analysis, the paraformaldehydefixed and paraffin-embedded kidney sections were incubated with primary antibodies to CD3 (ab16669, Abcam) or CD68 (ab955, Abcam), overnight at 4°C. Then, the reaction was monitored with an ultrasensitive streptavidin peroxidase detection system (Maixin), which contained secondary goat anti-mouse or anti-rabbit antibody. The sections were then counterstained with hematoxylin. Immunofluorescence staining of paraformaldehyde-fixed kidney sections were performed with primary antibody against KIM-1 (MA5-28211, Invitrogen), followed by incubation with a secondary antibody. Cell nuclei were stained with DAPI. Immunostained sections were visualized under a confocal microscope (FV1000, Olympus).

Quantitative real-time polymerase chain reaction (PCR) assay
The total RNA from cells or kidney cortex was extracted using the RNAiso plus reagent (Takara), and cDNA was then synthesized using PrimeScript RT reagent kit (Takara) according to the instructions. Quantitative realtime PCR was performed using a 7300 real-time PCR System (Applied Biosystems) to determine the levels of TNF-α, MCP-1, IL-6, and IL-1β. The relative expression levels of mRNA were normalized by β-actin. Primers were synthesized by Generay (Generay Biotech Co., Ltd.). All the primer sequences were listed in Supplementary Table 1.

Cell treatment
TECs were plated in 6-well plates at a density of 2 × 10 5 cells/well and incubated until they reached approximately 70% confluence for experiment. The cells were cultured in FBS-free medium for 12 h. Further, cells in cisplatin group were treated with 2.5 μg/mL cisplatin for 6 h, and the culture media were replaced with complete culture medium for 24 h. For the 2D-exos or 3D-exos treatment, cells were treated with 2.5 μg/mL cisplatin for 6 h and then incubated with 2D-exos or 3D-exos (15 μg) in complete culture medium for 24 h. The cells in control group were treated with FBS-free medium for 6 h and incubated in complete culture medium for 24 h.

Cell counting kit-8 assay
Cell viability was determined by the cell counting kit-8 (CCK-8) assay kit (APExBIO Technology LLC). Briefly, TECs were cultured to reach 70-80% confluence in 96-well plates with different interventions. Ten-microliter CCK-8 reagent was added to the medium and incubated for 2 h. The absorbance was detected at 450-nm wavelength.

Western blotting
The protein lysates from exosomes or kidney tissues were prepared following standard protocols, and the protein content was determined using the BCA protein assay kit. Then, protein samples were separated by Bis-Tris Gel (Invitrogen) and transferred onto PVDF membranes (Millipore). Membranes were blocked in 5% BSA in TBST for 1 h at room temperature and were incubated with primary antibodies overnight at 4°C. Then, membranes were washed three times and incubated with secondary antibodies for 2 h at room temperature, and the signals were detected using an ECL advanced system (GE Healthcare). Intensity values expressed as the relative protein expression were normalized to GAPDH (AB2000, Abways). Primary antibodies used were anti-Alix [35] (sc-53540, Santa), anti-CD63 (ab193349, Abcam), anti-CD9 (ab92726, Abcam), and antiphosphorylated NF-κB p65 (3033, Cell Signaling Technology). Secondary HRP-conjugated antibodies used were anti-mouse IgG and anti-rabbit IgG (Abcam).

Statistical analysis
Data were expressed as the mean ± SD. Statistical analysis was performed using two-tailed Student's t test or one-way ANOVA. p < 0.05 was considered statistically significant.
A schematic diagram of conventional 2D cell culture, for the extraction of 2D-exos, is shown in Fig. 1a. Trypan blue staining showed that the proportion of living hucMSCs (passage 6) was 98%.

3D culture and identification of hucMSCs
The general schematic of the hollow fiber bioreactorbased 3D culture system is shown in Fig. 1b. 2 × 10 8 2Dcultured hucMSCs were seeded into the external space through port A. The culture medium was continuously perfused into the system through the internal space via the lateral port B. Figure 1c shows a general schematic of cross-sectional view through the bioreactor. Figure 1d shows the photograph of our hollow fiber bioreactorbased 3D culture system. This 3D culture system was kept running for a total of 55 days. In the first 10 days, the glucose consumption rate gradually increased, indicating that hucMSCs adapted to the 3D culture system and grew steadily. From the 10th to the 48th day, glucose consumption rate was stable. Thereafter, glucose consumption started to decline (Fig. 2a). Trypan blue staining showed that the living cell ratio of exfoliated hucMSCs from the 3D system was 95% on the 55th day. During the first 10 days, the serum-free supernatants of the external space were collected every 48 h for isolating 3D-exos. After the 10th day, supernatants were collected every 24 h.
To explore whether the phenotype of husMSCs from 3D culture system had changed, on the 10th, 23rd, 34th, and 55th days, the exfoliated cells in the supernatants from the external space were collected and cultured in a 2D culture flask (Fig. 2c). Due to the small number of exfoliated cells, it took 2 weeks of 2D culture to expand optimal cells for MSC and HSC surface markers' detection. Interestingly, the exfoliated hucMSCs from 3D system were still highly positive (> 97%) for MSC surface markers including CD29, CD44, CD73, and CD90, and negative (< 1%) for CD34 and CD45, suggesting that 3D culture system did not change the phenotype of husMSCs. Figure 2b shows the flow cytometry analysis of surface markers of husMSCs on the 23rd and 55th days.
High-yield exosome production from 3D culture 2 × 10 8 hucMSCs were cultured in conventional 2D flasks, and 420 mL supernatants were collected. The same number of hucMSCs was inoculated into 3D Yields of 2D-exos and 3D-exos were detected by protein quantification. The yield of exosomes extracted from 2D-cultured supernatants was 0.42 mg, while that from 3D-cultured supernatants was 19.4-fold higher, with a maximum yield of 8.15 mg (Fig. 4a). 3D-exos were more concentrated in the harvested supernatants (15.5fold) than 2D-exos, which led to much higher efficiency of isolation through labor-intensive differential ultracentrifugation. Figure 4b shows the number of particles and protein contents of 3D-exos harvested from 3D culture system at different time points.

Enhanced renoprotective efficacy of 3D-exos in cisplatintreated mice
To compare the therapeutic efficacy of 2D-exos and 3Dexos, a murine model of cisplatin-induced AKI was established, and the treatment protocol was described in detail in the "Materials and methods" section (Fig. 5a). Treatment with 3D-exos caused a more obvious reduction in the level of sCr compared with 2D-exos (Fig. 5b). Mice treated with cisplatin showed extensive epithelial cell swelling, vacuolar degeneration, necrosis, detachment, and casts' formation occurring predominantly in the proximal tubules, all of which were markedly attenuated by treatment with either 2D-exos or 3D-exos.
However, 3D-exos-treated group improved more significantly (Fig. 5c, d). The mouse body weight of 3D-exostreated group was significantly higher than that of 2Dexos-treated group (Fig. 5e). Besides, kidney sections were stained with anti-KIM-1 antibody for detecting injured tubules. The results demonstrated that KIM-1 was expressed lower in 3D-exos-treated group (Fig. 5f, g). Collectively, these data clarified that both 2D-exos and 3D-exos can alleviate cisplatin-induced murine AKI, but 3D-exos were more effective.

Enhanced anti-inflammatory efficacy of 3D-exos in cisplatin-treated mice
To fully assess the treatment efficacy of 2D-exos and 3D-exos, we also measured the mRNA expressions for TNF-α, MCP-1, IL-6, and IL-1β, which showed decreased expression after 3D-exos treatment (Fig. 6a), accompanied by decreased expression of NF-κB p-p65 (Fig. 6b, c). In addition, the kidney interstitial infiltration of inflammatory cells, such as macrophages (CD68 positive) and T cells (CD3 positive), was significantly diminished by 3D-exos treatment (Fig. 6d-f). These inflammatory mediators were also blunted with 2D-exos, however, to a lesser extent than that with 3D-exos. Meanwhile, levels of IL-6 and TNF-α in mouse serum of 3D-exos-treated group were lower than 2D-exos-treated group (Fig. 6g, h). These data suggested the anti- Representative NTA analysis of the particle size distribution of 2D-exos and 3D-exos. The difference between the median diameter of 2D-exos (n = 4) and 3D-exos (n = 18) was not statistically significant (p > 0.05). c The particle size distribution of 3D-exos was detected at different time points, and the median diameters of 3D-exos were 118.1-141.3 nm. d Western blotting analysis of exosomal surface markers (CD9, CD63, and Alix) of 2D-exos and 3D-exos inflammatory efficacy of 3D-exos was significantly higher than that of 2D-exos.

Enhanced therapeutic efficacy of 3D-exos in cisplatintreated TECs
Considering that TECs are extremely sensitive to toxins, protecting TECs from such insults largely determines the prognosis of AKI [37], we speculated that MSC-exos would protect against cisplatin-induced AKI, at least in part, through targeting TECs. Therefore, TECs were used to explore the therapeutic effect of 2D-exos and 3D-exos in vitro. Higher levels of DiI-labeled 3D-exos were found to be internalized by cisplatin-treated TECs (Fig. 7a-c). The uptake efficiency of DiI-labeled 2D-exos and 3D-exos by cisplatin-treated TECs were approximately 60% and 80%, respectively. Furthermore, compared with 2D-exos, 3D-exos markedly inhibited cisplatin-induced mRNA expression of TNF-α, MCP-1, IL-6, and IL-1β (Fig. 7d). Moreover, the cell viability of cisplatin-injured TECs was significantly improved by treatment with 3D-exos (Fig. 7e). Above all, these findings suggested that 3D-exos were more easily taken up by TECs and had a more potent ability to ameliorate cisplatin-induced inflammation and to improve the viability of TECs.

Discussion
Adult stem cells are extensively used for the treatment of multiple human diseases, with the secretion of soluble factors and EVs that can enhance the repair of damaged tissues through paracrine regulation of local cells [1,15,38]. As a major class of EVs, MSC-exos show innate therapeutic potential by virtue of their intrinsic cargoes [15,19]. MSC-exos represent a sweet spot for cell-based therapies. They can be used as nature's own delivery tool, act as biotherapeutics, and mimic the action of cellular therapies without the side effects of embolism and proliferation [8]. However, the lack of a platform to obtain sufficient MSC-exos remains a challenge of therapeutic development. In this study, we discovered a novel solution by using a hollow fiber bioreactor-based 3D culture system, which could culture large quantities of husMSCs over a period of 55 days, thereby resulting in a 19.4-fold improvement of MSC-exos yield than that of 2D culture. Furthermore, we explored that 3D-exos showed superior renoprotective efficacy by alleviating inflammation in vivo and in vitro.
3D culture platform presents more advantages over conventional 2D culture. First, this 3D culture system can culture a multitude of cells (estimated to be in the order of 10 9 cells) within a standard cartridge [39]. Second, high-yield of exosomes can be achieved. Compared with 2D culture, the exosome production of 3D culture is 19.4 times higher. This is mainly due to the fact that the external space of the 3D system fills with serum-free medium, and the supernatants for extracting 3D-exos can be collected continuously. Third, 3D-exos are more concentrated in the harvested supernatants (15.5-fold) than 2D-exos, which leads to a higher exosome collection efficiency. In order to obtain equivalent output from the 3D system, approximately 230 T225 cm 2 flasks are required to extract 2D-exos. 3D culture reduces the use of plastic consumables such as cell culture flasks and centrifuge tubes. Furthermore, compared with the 2D culture system that requires several operating hours to harvest an equivalent quantity of exosomes, it takes less than 30 min daily in the 3D culture system. In these terms, 3D culture is both human resource-efficient and environmentally friendly.
In recent years, other researchers have used scaffolds, spheroid culture, or microcarrier-based 3D to culture MSCs [40][41][42]. Previously, Zhang et al. [40] demonstrated that compared with 2D culture, human bone marrow-derived MSCs (hBM-MSCs) seeded in the 3D collagen scaffolds generated more exosomes with improved repair function in rats after traumatic brain injury. They transferred 3 × 10 6 hBM-MSCs per scaffold into 200 μL of culture medium for exosome isolation, Fig. 4 The exosomes yield of 2D-culture and 3D-culture. a The total output of exosomes from 2 × 10 8 MSCs seeded in 2D and 3D culture system. b The number of particles and protein contents of 3D-exos harvested from 3D culture system at different time points which was harvested only once. Haraszti et al. proved that scalable microcarrier-based 3D cultures could double the density of MSCs and yield more exosomes than 2D cultures [41]. Although these microcarrierbased 3D cultures had a large total surface area (1150 cm 2 ) which can culture millions of MSCs, the culture media could only be collected once. In contrast to the other 3D culture systems, hundreds of millions of hucMSCs could be cultured in this hollow fiber bioreactor 3D culture system over a longer time period, which could produce 3D-exos in a continuous and efficient manner.
Despite advances in medical care, AKI remains an independent predictor of in-hospital mortality [43,44]. Currently, there are no reliable therapies existing to prevent or treat established AKI [44]. MSC-exos may . e Effects of 2D-exos and 3D-exos on the body weight (n = 6-7). f Representative confocal images of kidney injury molecular-1(Kim-1) in tubules. Scale bar, 25 μm. g The quantification of Kim-1 + tubules per HPF (n = 6). Data are presented as mean ± SD, *p < 0.05, **p < 0.01, ***p < 0.001 vs. Cis group, # p < 0.05, ### p < 0.001 represent a novel cell-free therapeutic agent for AKI. A wealth of researches corroborated that 2D-exos derived from MSCs could ameliorate AKI [45,46]. Zhou et al. [45] indicated hucMSC-derived exosomes could repair cisplatin-induced AKI through ameliorating oxidative stress and cell apoptosis and promoting cell proliferation Fig. 6 Enhanced anti-inflammatory efficacy of 3D-exos in cisplatin-treated mice. a Real-time PCR analysis of inflammatory cytokine mRNA levels in kidney tissues (n = 6-7). b Western blotting analysis of p-p65 in kidney tissues. c Quantification of p-p65 in kidney tissues (n = 4). d Representative immunostaining images of CD68 + macrophages or CD3 + T cells in the tubulointerstitium. Scale bar, 50 μm. e Quantification of CD68 + macrophages in the tubulointerstitium (n = 6). f Quantification of CD3 + T cells in the tubulointerstitium (n = 6). g Levels of IL-6 in the mouse serum was detected by ELISA (n = 6-7). h Levels of TNF-α in the mouse serum was detected by ELISA (n = 6-7). Data are presented as mean ± SD, *p < 0.05, **p < 0.01, ***p < 0.001 vs. Cis group, # p < 0.05, ## p < 0.01 in rats. Collino et al. [46] showed EVs derived from bone marrow MSCs protected the host from glycerol-induced AKI by carrying microRNAs. In the present study, we compared the therapeutic efficacy between 2D-exos and 3D-exos in cisplatin-induced AKI and found that 3Dexos exhibited superior renoprotective effects than 2Dexos, probably because 3D-exos were more easily taken up by TECs. Previous studies have shown that EVs could be internalized by clathrin-mediated or clathrinindependent endocytosis, such as macropinocytosis and phagocytosis as well as through endocytosis via caveolae and lipid rafts [47]. The process of endocytosis is complex and depends on the EVs and the recipient cells. Considering that the surface proteins of EVs allow them to be targeted and captured by recipient cells [48], 3D culture may alter exosomal membrane proteins, thus making them more easily be taken up by TECs. Interestingly, one study demonstrated that clathrin could mediate the endocytosis of TECs [49]. Changes in the expression of endocytosis-related proteins in damaged TECs are worthy of attention in the future.
On the other hand, the superior renoprotective ability of 3D-exos was also dependent on the components of their cargos. Yang et al. [50] demonstrated that 195 kinds of miRNAs and proteins, such as neprilysin, insulin-degrading enzyme, and heat shock protein 70, in exosomes derived from hucMSCs cultured in 3D scaffold are dramatically different from those in 2D culture. These 3D-exos exerted enhanced therapeutic effects on ameliorating the memory and cognitive deficits in Alzheimer's disease mice through their special cargo. The results showed that the content of exosomes was affected by the culture conditions. In future studies, it would be of interest to employ comparative proteomic and RNA-seq to dissect the differences in proteins and nucleic acids between 2D-exos and 3D-exos.
Additional file 1: Supplementary Table 1. Primers used in this study. Figure S1. Identification of hucMSCs. a HucMSCs exhibited a spindle fibroblast-like morphology. Scale bar: 50 μm. b HucMSCs could differentiate into osteoblasts, adipocytes or chondroblasts, evidenced by Alizarin Red S staining (b-a), Oil Red O staining (b-b), and Alcian Blue/Nuclear Fast Red staining (b-c), respectively. Scale bar: 50 μm. c Flow cytometry analysis of MSC (CD29, CD44, CD73 and CD90) and HSC (CD34 and CD45) surface markers. Blue solid peaks represent the isotype controls and the red solid peaks represent the marker indicated.