Mesenchymal stem cell-derived extracellular vesicles induce regulatory T-cells to ameliorate chronic kidney injury

Abstract

Metabolic syndrome (MetS) profoundly changes the contents of Mesenchymal stem cells (MSCs) and MSC-derived extracellular vesicles (EVs). The anti-inflammatory transforming growth factor (TGF)-β is selectively enriched in EVs from Lean but not from MetS pigs, but the functional impact of this endowment remains unknown. We hypothesized that Lean-EVs more effectively induce regulatory T-cells (Tregs) in injured kidneys. Five groups of pigs (n=7 each) were studied after 16 weeks of diet-induced MetS and unilateral renal artery stenosis (MetS+RAS). Two groups of MetS+RAS were treated 4 weeks earlier with an intra-renal injection of either Lean-EVs or MetS-EVs. MetS+RAS had lower renal volume, renal blood flow, and glomerular filtration rate than MetS pigs. Compared with Lean-EVs, MetS-EVs were less effective in improving renal function and decreasing tubular injury and fibrosis in MetS+RAS. Lean-EVs upregulated TGF-β expression in stenotic kidney and increased Tregs numbers more prominently. Furthermore, markedly up-regulated anti-inflammatory M2 macrophages, reduced pro-inflammatory M1 macrophages and CD8+ T-cells were detected in stenotic kidneys treated with Lean-EVs compared to MetS-EVs, and renal vein level of IL-1β were reduced. In vitro, co-culture of Lean-EVs with activated T-cells led to greater TGF-β-dependent Tregs induction than did MetS-EVs. Therefore, the beneficial effects of MSC-derived EVs on injured kidneys might be partly mediated by their content of TGF-β signaling components which permitting increased Treg preponderance. Modulating EV cargo and transforming their functionality might be useful for renal repair.

Graphical Abstract

Introduction

Mesenchymal stem cells (MSCs) are self-renewing, multipotent cells that reside in various tissues and possess the ability to confer tissue repair. Thanks to their regeneration potential, exogenous administration of MSCs promotes renal functional recovery both in animal models¹ and human subjects² with kidney diseases. For example, MSC delivery improved kidney function, structure, and microvascular remodeling, and reduced oxidative stress, apoptosis, and fibrosis in post-stenotic pig kidneys³. A series of complex mechanisms have been in renoprotection by MSCs, in particular via immunomodulation by way of paracrine factors and microvesicles release ⁴. However, the underlying molecular machinery remains to be elucidated.

Extracellular vesicles (EVs) release is an important mechanism by which MSCs confer renal protection. EVs arise from microvesicular bodies (exosome) or membrane blebbing (microparticles), are released into extracellular space, and internalized into recipient cells by docking or fusion with plasma membrane. Because they possess characteristics of their parental cells, MSC-derived EVs are renoprotective both in acute and chronic kidney disease^{5, 6}. A unique sorting process results in selectively enriched elements in EVs⁷, and we have recently found specific pro-angiogenic genes and proteins, and microRNA cargo that regulate angiogenesis enriched in healthy MSC-derived EVs^{8, 9}.

The metabolic syndrome (MetS) is characterized by a combination of metabolic disorders, including obesity, dyslipidemia, elevated blood pressure, and insulin resistance. MetS not only contributes to increased cardiovascular morbidity and mortality, but also affects MSC phenotype¹⁰, with decreased proliferation rate, higher susceptibility to apoptosis, and decreased multilineage differentiation potential, which might limit their therapeutic value¹¹. We have shown increased propensity for adipogenesis in adipose tissue-derived MSCs from a porcine model of MetS compared to Lean-MSCs, possibly driven by tumor necrotic factor (TNF)-α¹⁰. Furthermore, we found that the cargo of MSC-derived EVs from MetS pigs has enriched pro-inflammatory and decreased anti-inflammatory mediators, which might affect their ability to suppress inflammation in target cells^12-14.

Chronic inflammation is prevalent in patients with chronic kidney disease and contributes to cardiovascular disease progression. Local production of transcription factors like nuclear factor kappa-B (NF-κB)¹⁵ upregulates vascular cellular adhesion molecules, and thereby mediates production of proinflammatory cytokines like monocyte chemoattractant protein (MCP)-1, which recruits immune and inflammatory cells, mediating irreversible kidney damage.

Regulatory T-cells (Tregs), a subset of T-cells that often show a Foxp3+CD4+CD25+ phenotype, possess immunomodulatory ability and suppress multi-organ inflammation¹⁶. Several studies have suggested renoprotective effect of Tregs in both acute and chronic renal injury^{17, 18}, which might be mediated by preventing accumulation of inflammatory cells, increased secretion of inhibitory cytokines, and shifting macrophages towards an M2 anti-inflammatory phenotype¹⁹. Transforming growth factor (TGF)-β is necessary for Treg phenotype gain and function maintenance. TGF-β induces forkhead box-P3 (Foxp3) gene expression in precursor CD4⁺CD25⁻ naive T-cells, which mediates their transition toward a Treg phenotype with potent immunosuppressive potential²⁰, and the capacity to suppress effector T-cell function²¹.

MSC-derived EVs from Lean animals are enriched with both proteins and mRNAs along the TGF-β signalling pathway, but we have shown that these are relatively depleted in MetS-EVs^{12, 13, 22}. Given the central role of TGF-β in Treg expression and function, we hypothesized that Lean-EVs would increase the prevalence of Tregs in stenotic kidney and ameliorate inflammation, but that this function would be relatively attenuated in EVs derived from MetS-MSCs.

Methods

Female farm pigs were randomly fed standard (Lean, n=7) or high-cholesterol/carbohydrate (Mets, n=28) diet²³. Animal studies were approved by the Institutional Animal Care and Use Committee. Six weeks later, unilateral renal artery stenosis (RAS) induction (21 MetS pigs) or sham procedure were performed. Six weeks later, allogeneic Lean or MetS EVs (1×10¹⁰) (pre-labeled with PKH26, Sigma) were infused in the renal artery of MetS+RAS pigs (n=7 each), and vehicle in the remaining MetS+RAS, MetS, and Lean pigs (n=7 each) for controls.

Four weeks later, single-kidney hemodynamics and function were determined using multi-detector computed tomography (MDCT), and systemic blood samples collected for lipid panel, immune cells, fasting glucose, and insulin levels. Stenotic kidney venous blood was collected to measure inflammatory factors levels, such as interleukin (IL)-1α, IL-1β, IL-6, and tumor necrosis factor (TNF)-α. The pigs were then euthanized and kidneys were harvested.

MSC and EV studies

Our previous work has detailed the MSC and EV isolation and characterization techniques^{3, 13}. EVs were isolated from supernatants of MSCs (10^{^}6) using ultra-centrifugation⁸, and characterized based on the expression of EV (CD40, ß1, CD9, and CD81) and MSC (MHC-class-I and CD44) surface markers using flow-cytometry. Micro-RNAs targeting TGF-β signaling were analyzed using RNA-sequencing⁸. EVs were evaluated in stenotic kidney sections by immunofluorescence microscopy, and their distribution by flow cytometry in the heart, lungs, liver, spleen, and both kidneys at 2 days (in 2 additional pigs) and 4 weeks after intrarenal EV injection. EV retention was calculated as % of injected amount.

T-cell assays

Fresh blood was collected from healthy Lean pigs to isolate peripheral blood monocyte (PBMCs)²⁴, and T-cell activation induced in a phytohemagglutinin (PHA) assay. Then, 1×10⁶ activated T-cells were co-cultured with MetS-EVs, Lean-EVs, or Lean-EVs and TGF-β neutralizing antibody for 3 days.

Stenotic kidneys were dissociated to analyze distribution of Tregs by flow cytometry (Figures S1-S2), characterized by CD4+, CD25+, CD45RA−, CD127−/Low, FoxP3+. Immune cell distribution in the renal tissue and co-culture were also phenotyped by flow cytometry (Figure S3) for CD3, CD4, CD8, CD14 and CD16.

Tissue studies

Kidney cortical sections were assessed for tubular injury, fibrosis, inflammatory cell (M1/M2 macrophages) distribution and Tregs. Renal gene expression of kidney injury molecule-1 (KIM-1) and collagen-1 was detected by quantitative polymerase-chain-reaction, and forkhead box protein-3 (FoxP3) protein by Western blotting. Levels of TGF-β were assessed by immunostaining, Western blotting, and renal vein levels.

Statistical analysis

Statistical analysis was performed using JMP 14.0-Pro, and p<0.05 considered statistically significance.

See full details in Supplemental Methods. On reasonable request data, analytic methods, and study materials will be made available to other researchers for the purposes of reproducing the results.

Results

Compared to Lean, all other experimental groups had markedly increased body weight, mean arterial pressure (MAP), lipid levels, fasting insulin, renal volume, RBF and GFR (Table 1, Figure 1A). MetS+RAS had lower renal volume, RBF and GFR than MetS pigs. All RAS pigs demonstrated moderately hemodynamically significant stenosis (p<0.05 in all).

Table 1.

Systemic characteristics and single-kidney function in study groups at 16 weeks.

Characteristic	Lean	MetS	MetS+RAS	MetS+RAS+ Lean-EVs	MetS+RAS+ MetS-EVs
Body weight (Kg)	70.0±9.8	96.3±4.5*	91.5±5.6*	93.7±7.2*	93.2±6.0*
MAP (mmHg)	90.7±4.4	124.6±9.6*	131.7±17.5*	120.9±11.5*	126.7±6.9*
Degree of RAS (%)	0	0	66.7±18.3*†	66.8±8.2*†	65.0±5.5*†
Total cholesterol (mg/dl)	88.3 (76.7-90.8)	333.5 (314.3-383.0)*	338.4 (324.4-416.1)*	350.7 (311.3-417.5)*	323.1 (311.4-400.5)*
HDL cholesterol (mg/dl)	47.0 (44.5-50.3)	126.5 (110.3-160.5)*	132.5 (98.7-136.8)*	128.0 (116.8-134.0)*	125.5 (114.5-132.5)*
LDL cholesterol (mg/dl)	33.9±6.7	354.5±140.3*	347.8±89.3*	367.0±38.8*	349.4±41.4*
Triglycerides (mg/dl)	7.8 (5.8-9.0)	15.8 (13.8-23.3)*	14.6 (12.9-15.3)*	16.5 (13.3-18.1)*	16.7 (11.9-21.9)*
Fasting glucose (mg/dl)	125.0±16.9	120.7±18.7	111.8±11.6	109.6±12.9	121.3±17.0
Fasting insulin (μU/ml)	0.4 (0.4-0.5)	0.7 (0.7-0.8)*	0.8 (0.8-0.9)*	0.8 (0.7-0.8)*	0.8 (0.7-0.8)*
HOMA-IR score	0.6 (0.5-0.7)	1.9 (1.6-1.9)*	1.8 (1.7-1.9)*	1.7 (1.7-1.8)*	1.8 (1.7-1.9)*

MAP: mean arterial pressure; RAS: renal artery stenosis; HDL: high-density lipoprotein; LDL: low-density lipoprotein; HOMA-IR: homeostasis model assessment of insulin resistance.

^*p<0.05 vs. Lean

^†p<0.05 vs. MetS

^‡p<0.05 vs. MetS+RAS

^#p<0.05 vs. MetS+RAS+Lean-EVs.

Figure 1. Inflammatory cytokines levels in the renal vein.

A. Compared to Lean, kidney volume, renal blood flow (RBF) and glomerular filtration rate (GFR) were increased in MetS, while superimposed RAS decreased kidney volume, RBF and GFR. Kidney volume, RBF, and GFR were restored in Lean-Evs-treated pigs, but MetS-EVs failed to preserve them. B. Renal vein levels of interleukin (IL)-1α, IL-1β, tumor necrotic factor (TNF)-α and IL-6 were measured by Luminex. No difference was found in IL-1α among all groups. MetS+RAS increased IL-1β, IL-6, and TNF-α compared to Lean, but only IL-1β and IL-6 compared to MetS. Both Lean- and MetS-EVs markedly decreased IL-1β and TNF-α levels. *p<0.05 vs. Lean; †p<0.05 vs. MetS; ‡p<0.05 vs. MetS+RAS (n=7 each). C. MetS+RAS+Lean-EVs had significantly higher renal protein expression of TGF-β than the other groups (n=3 each). D. Treated and untreated MetS+RAS pigs had similarly higher level of TGF-β in the stenotic renal vein compared to Lean and MetS (n=7 each).

Mets-EVs are enriched with miRNAs targeting the TGF-β pathways

Forty miRNAs were differently expressed between Lean- and MetS-EVs, of which 27 were up-regulated in MetS-EVs (Figure S4). Nineteen miRNAs enriched in MetS-EVs targeted most of the genes involved in TGF-β pathway (Table S1), indicating that miRNAs enriched in MetS-EVs can negatively regulate TGF-β signaling.

EVs decreased inflammatory cytokines release

Levels of IL-1β, IL-6, and TNF-α were higher in the stenotic MetS+RAS compared to control renal vein, whereas IL-1α remained unchanged. Both Lean- and MetS-EVs decreased cytokines levels, but IL-1β was further lower in Lean-EV- compared to MetS-EV-treated groups (Figure 1B).

MetS-EVs did not attenuate renal injury

EVs can be internalized into others cells and exert their biological effects by delivering their gene and protein cargo. Both Lean- and MetS-EV were detected in stenotic kidneys at euthanasia (Figure 2A). Most injected EVs were retained in liver, lung and spleen, with ~9.1% retained within the stenotic kidney by 2 days and 2.3% by 4 weeks after injection, and 2.3% and 1%, respectively, in the contralateral kidney (Figure 2B). Tubular injury score was greater in MetS+RAS vs. Lean and MetS, indicated by greater number of tubules with brush border loss, dilation, atrophy, and vacuolization. Lean-EVs markedly reduced tubular injury, although it remained higher than Lean, whereas MetS-EVs failed to blunt it (Figure 2C). Renal fibrosis showed a similar pattern (Figure 2C). In addition, KIM-1 and Collagen-1 expression was upregulated in MetS+RAS compared to Lean and MetS, but downregulated by both Lean- and MetS-EV, yet more effectively by Lean-EVs (Figure 2D).

Figure 2. EV retained in stenotic kidney and Lean-EVs ameliorate tissue injury induced by renal artery stenosis (RAS) and metabolic syndrome (MetS).

A. Red immunofluorescent stained of both Lean- and MetS-EVs (PKH26, arrows, 40X) were detected in the stenotic kidney 4 weeks after administration, and none in untreated kidneys. B. EV distribution in solid organs at 2 days and 4 weeks. Most EVs were retained in the liver, lung, and spleen. Stenotic had more EVs than contralateral kidneys at 2 days and 4 weeks. EVs retention gradually decreased over time in all organs (n=2 each). C. Representative images (×40) and quantitative analysis of periodic acid–Schiff (PAS) and trichrome staining for cortical tubular injury. MetS+RAS show greater tubular brush border loss, dilation, atrophy, and vacuolization. Lean-EVs ameliorated tubular injury, although it remained higher than Lean, whereas MetS-EVs failed to blunt it. D. MetS+RAS up-regulated gene expression of both kidney injury molecule (KIM)-1 and Collagen-1 compared to Lean and MetS, which was markedly decreased by both Lean-EV and MetS-EV, but retained higher in MetS+RAS+MetS-EVs (n=7 each). *p<0.05 vs. Lean; †p<0.05 vs. MetS; ‡p<0.05 vs. MetS+RAS; #, p<0.05 vs. MetS+RAS+Lean-EVs.

Lean-EVs increased TGF-β immunoreactivity and enriched Tregs in the stenotic kidney

TGF-β immunoreactivity was significantly lower in MetS, MetS+RAS, and MetS+RAS+MetS-EVs than in Lean kidneys, whereas MetS+RAS+Lean-EVs up-regulated it substantially, and markedly higher than MetS-EVs (Figure 3A). Renal TGF-β expression by Western blotting was slightly but significantly elevated in MetS+RAS+Lean-EVs compared to Lean, and unchanged in the other groups (Figure 1C), whereas in the venous effluent TGF-β levels were increased in all MetS+RAS stenotic-kidneys, treated or untreated (Figure 1D).

Figure 3. Lean-EVs increase the expression of TGF-ß and Tregs phenotypes.

A. Representative images (×40) of and transforming growth factor (TGF)-β immunoreactivity. Both MetS and MetS+RAS had lower TGF-β immunoreactivity than Lean pigs; Lean-EVs restored it, while MetS-EVs further decreased it. B. Fluorescent staining for Tregs characterized as CD4+/CD25+/Foxp3+. Quantitative analysis of Tregs expression using cell-count and positive area indicated that MetS had more Tregs than Lean and MetS+RAS, and Lean-EVs induced more Tregs than the remaining groups. MetS-EVs also induced a slightly greater number of Tregs vs. Lean and MetS+RAS, while markedly increasing the abundance of CD4+CD25−Foxp3+ T-cells. *p<0.05 vs. Lean; †p<0.05 vs. MetS; ‡p<0.05 vs. MetS+RAS; #, p<0.05 vs. MetS+RAS+Lean-EVs (n=7 each).

The number of Tregs was higher in MetS kidneys compared to normal. Their number increased in both MetS+RAS+Lean-EVs and MetS+RAS+MetS-EVs, but markedly more in MetS+RAS+Lean-EVs. Treg positive area and cell count showed similar patterns (Figure 3B, S5). Conversely, there were more Treg precursor (CD4+CD25−Foxp3+) T-cells in MetS+RAS+MetS-EVs, but not in any other group, detected by both their positive area and T-cell count (Figure 3B).

Lean-EVs decreased inflammatory cells in the stenotic kidney

The number of M1 (iNOS+/CD68+) macrophages increased in MetS kidneys and further in MetS+RAS, whereas M2 (Arg-1+/CD68+) macrophages were decreased compared to Lean. Lean-EVs polarized macrophages balance toward M2, whereas the MetS-EVs did not, resulting in a higher M1/M2 ratio in MetS-EVs-treated pigs (Figure 4). Furthermore, the number of CD8+ T-cells was progressively increased in MetS and MetS+RAS vs. Lean, and significantly decreased by Lean-EVs, but not MetS-EVs (Figure 4).

Figure 4. Lean-EVs induce macrophage polarization towards M2 phenotype and decrease infiltration of inflammatory cells.

Representative images (×40) of immunofluorescence staining for M1[CD68+(red)/inducible nitric oxide synthase(green)], M2 [CD68+(red)/arginase-1(green)], and CD8+ (red). Quantitative analysis of M1 and M2 macrophages, M1/M2 ratio and CD8+ T-cells. The number of M1 macrophages and M1+ area progressively increased in MetS and MetS+RAS, while the M2 macrophage demonstrated opposite trend. Lean-EVs decreased M1 and increased M2 macrophages, whereas MetS-EVs failed to decrease M1 and induced fewer M2 than Lean-EVs. M1/M2 ratio was significantly lower after Lean-EVs than after MetS-EVs treatment. Furthermore, the number of CD8+ T-cells was considerately increased in MetS and MetS+RAS, and only Lean-EVs significantly reduced it. *p<0.05 vs. Lean; †p<0.05 vs. MetS; ‡p<0.05 vs. MetS+RAS; #, p<0.05 vs. MetS+RAS+Lean-EVs. (n=7 each).

Lean-EVs induced more Tregs in-vitro

Flow cytometry (unavailable in the MetS+RAS+Lean-EVs group) demonstrated a greater number of Tregs (CD4+CD25+FoxP3+) in MetS and MetS+RAS+MetS-EVs vs. Lean and MetS+RAS kidneys (Figure 5A, 5D). Western blotting demonstrated upregulated renal protein expression of FoxP3+ in MetS+RAS+Lean-EVs compared to other groups, yet MetS-EVs also induced Foxp3+ expression in MetS+RAS (Figure 5B).

Figure 5. Lean-EVs increase TGF-β -dependent Tregs induction.

A. Tregs and CD4+CD25−Foxp3+ cell detected by flow cytometry. Tregs were characterized as CD4+/CD25+/CD45RA-/CD127^−/Low/FoxP3+ and CD4+CD25−Foxp3+ cells were CD4+/CD25−/CD45RA−/CD127^−/Low/FoxP3+. MetS kidneys had a greater number of Tregs than Lean and MetS+RAS. MetS-EVs induced Tregs and a greater number of CD4+CD25−Foxp3+ cells in MetS+RAS. B. Forkhead box protein-3 (Foxp3) expression detected by western blot. Lean-EVs and MetS-EVs upregulated Foxp3 expression compared to MetS and MetS+RAS, while MetS+RAS+Lean-EVs had higher Foxp3+ expression than MetS+RAS+MetS-EVs. C. In vitro co-culture study. Lean-EVs induced Tregs, whereas MetS-EVs failed to do so. TGF-β neutralization abolished the effect of Lean-EVs (n=4each). D. Representative image of Tregs and CD4+CD25−Foxp3+ T-cells in imaging flow cytometry. *p<0.05 vs. Control; †p<0.05 vs. PHA; #, p<0.05 vs. Lean-EVs.

In co-culture with PBMCs, flow cytometry showed that Lean-EVs significantly increased the percentage of Tregs compared to control and MetS-EVs. Treg number increased four-fold from baseline (0.23% to 0.83%) in Lean-EVs, but remained 0.14% in the MetS-EVs group. When also treated with a TGF-β neutralized antibody, induction of Tregs by Lean-EVs was abolished (0.25%) (Figure 5C). We also analyzed sub-populations of immune cells in co-cultures. In fresh isolated PBMCs, CD3+ cells accounted for most mononuclear cells (~40%), CD8+ for 10%, CD14+ for 6%, CD4+ for 0.5%, and CD16+ for 0.3%. Co-culture induced small fluctuations in some of the sub-populations (Figure S6), whereas Tregs increased more than five-fold after incubation with Lean-EV, but not with MetS-EVs, or with Lean-EV co-incubated with anti-TGF-ß antibody (Figure S6).

Discussion

The present study shows that Lean-EVs significantly induce Tregs both in-vivo and in-vitro. This was associated with improved stenotic kidney function and blunted tissue injury and inflammation in the MetS+RAS model, and shifted the balance of macrophages from M1 to M2 phenotype. Therefore, the momentous immunomodulatory properties of Lean-EVs might be mediated by their cargo of TGF-β that induces Tregs. Contrarily, these beneficial effects were considerably attenuated after delivery of MetS-EVs carrying reduced TGF-β signaling components, and a large content of miRNAs that target this pathway and may counteract the effects of TGF-β on Tregs.

MSCs are potent anti-inflammatory and immunomodulatory cells, and a promising tool for kidney repair. Our previous work has shown that MSCs delivery in the RAS porcine model ameliorated inflammation and restored stenotic kidney function^{1, 3}, and that healthy EVs repaired the stenotic kidney and decreased inflammatory markers⁶. Herein, we extended our previous findings to show that the ability of MSC-derived EVs to protect the kidney in MetS+RAS might be mediated by upregulating Tregs, a capacity that is blunted in EVs obtained from Mets-MSCs.

EVs are membrane particles derived from their parent cells, and thereby possess many of their characteristics. As a regenerative tool, EVs can overcome concerns about extensive expansion or mal-differentiation of MSCs. A sorting process during EVs formation selectively enriches elements in EVs compared to their parent cells⁷. We previously showed that EVs are enriched in proteins involved in alternative splicing, apoptosis, chromosome organization, pro-angiogenic pathways, and Golgi apparatus genes⁸. Pertinently, EVs obtained from Lean-MSCs express high levels of TGF-β-related genes like TGFB1, TGFB3, and FURIN⁸, as well as TGF-β signaling pathway proteins⁹. MetS profoundly alters the contents of MSCs, with up-regulated factors in MetS-MSCs mainly involved in inflammation²⁵. Consequently, metabolic disorders shift the cargo of MSC-derived EVs towards mRNAs involved in translational regulation and modulation of inflammation, but devoid of mRNAs related to anti-inflammatory TGF-β signaling²². The current study shows that MetS-EVs are also enriched with miRNA that target many components of TGF-β pathways.

Interestingly, we found that TGF-β immunoreactivity decreased in MetS and MetS+RAS, and was restored after Lean-EV treatment, but not after MetS+RAS+MetS-EVs. This is likely due to TGF-β delivery by Lean-EVs, and of miRNAs suppressing TGF-β signaling by MetS-EVs. Contrarily, protein expression of TGF-β in homogenates of MetS+RAS+MetS-EVs kidneys was unchanged, and its venous effluent levels were elevated, possibly secondary to differences in cellular expression, latent, secreted, and interstitial TGF-β content detected by the different techniques. For example, Immunocytochemical detection of TGF-β1 is sensitive for activated TGF-β1, given its cell-surface association, whereas Western blot analysis detects primarily latent TGF-β1²⁶, and its secreted form is in a complex with latency-associated protein²⁷. Indeed, we have previously identified discordant expression patterns of TGF-β in the renal vein and tissue in patients with RAS²⁸. Furthermore, the interstitial space contains latent TGF-β, whereas EVs are often up-taken by target cells and release their content within them, thereby interacting directly with cells.

We found a greater number of Treg in Lean-EVs-treated stenotic kidneys compared to all other groups, which might be attributed to the TGF-β signaling components enriched in Lean-EVs. This postulate is also supported by our findings that co-culture of PBMCs with Lean-EVs induced more Tregs than MetS-EVs, which was blunted by TGF-β inhibition. TGF-β induces FoxP3 gene expression in CD4+/CD25− naive T-cells, mediating their transition toward a regulatory T-cell phenotype (CD4+/CD25+/FoxP3+) with potent immunosuppressive potential²⁰. As CD4+/CD25− T-cells constitute a reservoir of future Tregs²⁹, blockade of TGF-β pathway by miRNAs in MetS-EVs might hamper this conversion, consistent with the elevated number of CD4+/CD25− T-cells detected inthe MetS+RAS+MetS-EVs kidneys. Furthermore, Tregs can induce naïve CD4+ T-cells to become FoxP3+-induced-Tregs (iTregs) through TGF-β from Tregs³⁰, constituting self-induction.

MetS+RAS induced renal inflammation, manifested as release of inflammatory cytokines in the renal vein, and greater numbers of pro-inflammatory macrophages (M1) and CD8+ T-cells, accompanied by loss of kidney function, underscoring the importance of inflammation for kidney injury, comparable to observations in the post-stenotic human kidney³¹. Importantly, MSCs delivery decreases the inflammatory mediators and cells infiltration³², as do their EVs⁶. The current study shows that MetS-EVs do not ameliorate renal inflammation, likely due to paucity of TGF-β-mediated Tregs induction. Depletion of Tregs increases numbers of activated CD4+ T-cells, CD8+ T-cells, and natural killer cells³³. Indeed, a higher number of Tregs was accompanied by fewer CD8+ T-cells in the stenotic kidney after Lean-EV, but not after MetS-EVs delivery. Additionally, increased Tregs prevalence was accompanied with marked polarization of M2 macrophage by Lean-EVs, in accordance with previous reports describing opposing effects of CD4+/CD25− T-cells and CD4+/CD25+ Tregs on macrophage polarization³⁴. Notably, TGF-β blockade reversed the effects of CD4+CD25+ Tregs on M2-macrophage induction³⁴. We also observed that MetS-EV, but not Lean-EV, induced greater number of CD3+ and CD8+ cells compared to control, whereas only Lean-EV decreased CD4+, CD8+, and CD14+ cells compared to PHA. These results were congruent with our in-vivo observations that Lean-EVs, but not MetS-EVs, ameliorate renal inflammation. Additional effects of TGF-β neutralizing on CD3+, CD4+, CD14+ and CD16+ cells may be linked to its role in lymphocyte differentiation and survival^{35, 36}.

Lean-EVs, but not MetS-EVs, significantly increased RBF and GFR, and reduced tubular injury and fibrosis, likely secondary to amelioration of renal inflammation. Treg depletion enhances renal inflammation, acute tubular necrosis, and dysfunction in acute kidney injury¹⁷, and transfer of Tregs into mice with Adriamycin nephropathy reduces glomerular and interstitial injury, as well as macrophage numbers¹⁸. Given the elevated numbers of Tregs in Lean-EVs-treated kidneys, endogenous Tregs might have promoted repair after the ischemic injury. Polarization of macrophages to an M2 phenotype also protects kidneys from ischemia^{32, 37} and promotes tubular regeneration³⁸. Furthermore, relief of renal inflammation might also contribute to restoration of the renal microcirculation, RBF, and GFR.

There are several limitations of this study. Owing to limited availability of fresh samples, we could not analyze Treg numbers in the peripheral circulation, or in kidneys of MetS+RAS+Lean-EVs pigs. Nonetheless, the patterns observed in immunofluorescence and flow cytometry for the other groups were very comparable, supporting our conclusions. Given the complex bioactive components in EVs, we cannot rule out the contribution of other EV cargoes. We have previously shown that EVs are not prone to rejection even after xenogeneic delivery³⁹. EV retention decreases by 4 weeks after injection, yet they suffice to improve renal integrity. Persistent presence of M2 macrophage after renal ischemic injury⁴⁰ or kidney transplant is associated with more fibrosis⁴¹, and their role in chronic injury needs to be clarified. TGF-β has been linked to development of fibrosis, which was attenuated after Lean-EVs, likely by other molecular pathways. The discordant expression of TGF-β and renal fibrosis indicate TGF-β independent pathway, such as mitogen-activated protein kinase signaling⁴², might play a more important role in renal fibrosis of MetS+RAS. In patients with RAS renal vein levels of TGF-β do not differ between stenotic and contralateral kidneys, and do not correlate with kidney blood flow, TGF-β staining, or function²⁸. Clearly, regulation of fibrosis in this disease is complex, and may vary with the disease stage and potentially comorbidities. However, the current study focused on the immunomodulatory, rather than pro-fibrotic, effect of TGF-β. The mechanism by which Lean-EVs improve RBF and GFR may involve local effects on Tregs induction, but other regional effects likely contributed as well. Finally, we studied all female pigs; as Tregs can be induced by estrogen⁴³, we cannot exclude contribution of sex hormones to our results.

In summary, we found elevated numbers of Tregs in kidneys of Lean-EVs-treated MetS+RAS pigs, as well as after their co-culture with activated T-cells, suggesting that the renal protective effect of EVs might be partly mediated by immunomodulatory function of Tregs. This was accompanied by polarizing pro-inflammatory macrophages towards anti-inflammatory phenotype and decreased infiltrating inflammatory cells and cytokines. MetS-MSC showed a relatively modest impact on Tregs, accompanied by blunted capacity for renal protection and inflammation amelioration. Given that MetS-EVs devoid of TGF-β components and TGF-β blockade blunted Tregs induction, TGF-β-induced Tregs might play a pivotal in renal protection achieved by healthy MSC-derived EVs.

Perspective

The mechanisms by which MSC offer renoprotection remain incompletely understood. Our study provides novel insights into the immunomodulatory effect of their paracrine EVs vectors. We show that their marked beneficial effects might be partly mediated by enrichment with TGF-β pathway components and induction of Tregs, which in turn ameliorate kidney injury and inflammation. Furthermore, we demonstrate that this endowment is deficient in MSC-EVs obtained from obese pigs, implying that novel strategies are needed to restore the endogenous repair capacity in obese individuals, and to ensure effectiveness of autologous regenerative approaches.

Summary.

Our study may provide novel insights into different immunological modulatory effect of MetS-EVs and Lean-EVs, which is linked to enriched TGF-β pathway components in Lean-EVs leading to Tregs induction.

Novelty and Significance:

1). What Is New:

Lean-EVs enriched with TGF-β pathway components induced Tregs in stenotic kidneys, whereas MetS-EVs did not. Blocking TGF-β signaling undermines induction of Tregs.

2). What Is Relevant:

TGF-β-induced Tregs may be the novel mechanism exerted by MSC-EVs counteracting inflammation in stenotic kidneys.