Shedding light on the role of extracellular vesicles in renal fibrosis

Abstract

Renal fibrosis is the final common pathway of chronic kidney disease (CKD); thus, preventing the fibrotic process is essential to retarding CKD progression. Extracellular vesicles (EVs) are cell-derived membranous particles comprising exosomes and microvesicles. They carry unique biomolecular cargo from donor to recipient cells and are recognized as a novel mechanism for intercellular communication. There is evidence that EV-mediated cellular crosstalk across the glomerular and tubulointerstitial regions potently participates in the pathogenesis of renal fibrosis. Moreover, due to their outstanding stability and delivery capacity in the body, EVs also show tremendous promise in the diagnosis and treatment of kidney diseases, ushering in a new era of precision theranostics. In this review, we discuss the pathological functions of EVs and highlight the applications of EVs as biomarkers, standalone therapeutics, and novel drug delivery systems for renal fibrosis.

1. Introduction

Chronic kidney disease (CKD) is a crucial health problem due to the high incidence and high mortality. It is estimated that globally, around 800 million people have CKD, and by 2040, CKD will become the fifth leading cause of death worldwide [[1], [2], [3]]. CKD is progressive and can induce irreversible damage to the kidneys. The pathological hallmark of CKD progression to end-stage kidney disease is renal fibrosis, which is characterized by excessive extracellular matrix (ECM) accumulation, immune cell infiltration, tubular atrophy, and capillary rarefaction [[4], [5]]. Given that renal fibrosis is the common endpoint of CKD of varied etiologies, it is important to understand the underlying mechanisms of fibrosis, advancing the early diagnosis and treatment of CKD.

Intercellular communication in the injured kidney is a crucial process causing fibrotic pathogenesis [6]. Cells can usually communicate via direct cell-cell contact as well as the secretion and uptake of soluble factors, such as cytokines, hormones and growth factors. Recent evidence suggests a new paradigm of communication, involving intercellular signal exchange by secreted vesicles. Extracellular vesicles (EVs) are membrane-bound particles secreted by virtually all types of cells [[7], [8], [9]]. EVs are initially assumed as a trash bag to rid the cell of waste. But within the past decade, they have emerged as pivotal signalosomes in cell-to-cell communication and play essential roles in numerous (patho-)physiological conditions via the transfer of bioactive cargo between cells [[10], [11]]. Furthermore, EVs also show tremendous promise as diagnostic biomarkers, natural therapeutic agents and drug delivery carriers [[12], [13], [14]]. In this review, we discuss the role of EVs in the development and progression of renal fibrosis and highlight current theranostic applications of EVs in kidney disease.

2. EV classes and biogenesis

EVs comprise a heterogeneous family of vesicles of diverse origins. As determined by their biogenesis pathways, the two main subtypes of EVs are exosomes and microvesicles [[7], [8]]. In this section, a brief overview of their main features and vesicle generation in the kidney are provided below.

2.1. Exosomes

Exosomes are small vesicles (50–150 nm in diameter) derived from the endosomal system (Fig. 1). Their biogenesis begins with the invagination of the plasma membrane, leading to the de novo formation of early-sorting endosomes, which can mature into late-sorting endosomes and ultimately generate multivesicular bodies (MVBs) that contain intraluminal vesicles (ILVs, future exosomes) formed by inward budding of the endosomal membrane. Then MVBs either end up in lysosomes for degradation or fuse with the plasma membrane to release ILVs as exosomes [[15], [16]]. To date, two mechanisms have been proposed to regulate ILV generation, including the endosomal sorting complex required for transport (ESCRT)-dependent or ESCRT-independent machinery. The canonical ESCRT machinery consists of four complexes and acts in a stepwise manner wherein ESCRT-0, -I, and -II orchestrate cargo sorting into the lumen of endosomes and ESCRT-III together with associated proteins, such as ALG­2 interacting protein X (ALIX), participate in the formation of ILVs in MVBs [17]. In contrast, the ESCRT-independent pathway relies on the activation of neutral sphingomyelinase-2, which hydrolyzes sphingomyelin to generate ceramide, promoting ILV formation [18]. As for exosome secretion, RAS‑related protein (RAB) RAB27a and RAB27b, actin and SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) proteins work together to mediate the docking and fusion of MVBs with plasma membrane followed by releasing exosomes into the extracellular milieu [[19], [20]].

Fig. 1.

Generation and characterization of EVs. Microvesicles are formed by the outward pinching of the plasma membrane; exosomes are derived from the intraluminal vesicles (ILV) within the multivesicular bodies (MVB), which then fuse with the plasma membrane to release ILV as exosomes. Exosomes are characterized by specific surface markers, such as tetraspanins, adhesion molecules, receptors, and transporters, and carry various cargoes, including proteins, lipids and nucleic acids.

Exosomes carry diverse cargoes, including nucleic acids, proteins, and lipids of their parental cells. Due to their endosomal origin, exosomes inherently package proteins involved in the endosomal biogenesis of ILVs such as the tetraspanins (CD63, CD81, and CD9) and ESCRT proteins (TSG101 and ALIX) [21]. However, sorting machineries that drive specific cargo targeting to exosomes are still ill-defined. Tetraspanins have been shown to contribute to the packaging and loading of various proteins to exosomes [[22], [23], [24]]. For example, β-catenin could be sorted into ILVs by the tetraspanins CD82 and CD9 and subsequently exported from the cell, thereby reducing the intracellular activity of Wnt/β-catenin signaling [24]. Moreover, some microRNAs (miRNAs) also seem to be scheduled for secreting within exosomes, probably depending on their specific sorting sequences (EXOmotifs) (e.g., CGGGAG) that bind to RNA-binding proteins (e.g., Alyref and Fus), which perform the sorting of these miRNAs into vesicles [[25], [26], [27]]. Other RNA-binding proteins implicated in nucleic acid sorting are being revealed, including the ESCRT-II [28], heterogeneous ribonucleoprotein A2B1 [29], and Y-box-binding protein 1 [30].

2.2. Microvesicles

Microvesicles (also known as ectosomes) are commonly referred to as large EVs (100 nm to 1 μm in diameter) and are shed from living cells through the outward budding and fission of the cell membrane [[15], [16]]. Unlike exosomes, the process of microvesicle biogenesis is still unclear but involves the rearrangement in phospholipid asymmetry, which is driven by Ca²⁺-dependent activation of floppase, scramblase, and calpain [31]. Additional molecular regulators, such as the small GTPase ADP­ribosylation factor 6 that contributes to actin reorganization, appear to be required for the shedding of microvesicles [32]. During microvesicle biogenesis, the membrane proteins surrounding the region are pinched off and the nearby cytoplasmic components can be packaged as microvesicle cargo. There is evidence that microvesicle membrane contains high levels of cholesterol, sphingomyelin, and receptors reflecting their cellular origin. For example, epidermal growth factor receptors on microvesicles have a high similarity to the originating tumor cells in terms of expression level and function [33]. Regarding cytosolic proteins and nucleic acids fated for microvesicle secretion, the mechanisms underlying their targeting into microvesicles remain to be unraveled.

2.3. EV biogenesis in the kidney

The kidney is a complex organ containing more than 20 specialized cell types [34], resulting in the complexity and diversity of EVs in the kidney. Evidence from the urine indicates that tubular epithelial cells (TECs) and podocytes are the two primary sources of kidney-derived EVs [35]. However, so far much less is known about the process of EV biogenesis and secretion in these cells, especially in vivo. Taking TECs as an example, their EV production may vary according to different pathological situations. Feng et al. found that albumin overload could promote tubular exosome release via interferon regulatory factor 1/Rab27a pathway in several murine models of proteinuria nephropathy [36]. While in the context of diabetic kidney disease, tubular secretion of exosomes was decreased due to the downregulation of Rab27b caused by FOXO1 inactivation [37]. Interestingly, studies by Matsui et al. reported that polarized renal epithelial cells could release exosomes from their apical and basolateral membrane but employ different machineries: apical release of exosomes is mediated by the ALIX/Syntenin1/Syndecan1 pathway, whereas basolateral exosome release is associated with the sphingomyelinase-dependent ceramide production [38]. Correspondingly, two independent regulators, Rab27 and Rab39, have been suggested in the transport of MVBs toward apical and basolateral membrane, respectively [39]. Given that certain EVs are able to provoke renal inflammation and fibrosis, further understanding how EVs are produced in the kidney will favor the development of treatments to inhibit their biogenesis and release. Currently some pharmacological agents have been suggested as vesicle inhibitors, such as the neutral sphingomyelinase-2 inhibitor (GW4869) [18], calcium-chelating agent (BAPTA-AM), and calpain inhibitor (ALLM) [40].

3. EVs as vicious signalosomes in renal fibrosis

The pathological role of EVs has been well-documented in a variety of kidney diseases. In response to insults, such as hypoxia, toxins, and oxidative stress, dysfunctional kidney cells will hijack EVs to discharge pathogenic materials, which are intentionally (or not) spread into neighboring or distant cells and thereby amplify renal injury. In the following part, we focus on the EV-mediated intercellular communication, particularly tubular cell-to-interstitial cell crosstalk that drives kidney fibrosis (Fig. 2).

Fig. 2.

The roles of EV-mediated cellular crosstalk in kidney damage. Damaging stimuli trigger proximal tubular epithelial cells (PTEC) to release EVs into the interstitial space, which localize to fibroblasts and macrophages, leading to their activation and promoting tubulointerstitial inflammation and fibrosis. In the glomeruli, mesangial cell and endothelial cell-derived EVs can deliver disease-causing factors to podocyte and consequently induce podocyte apoptosis and transdifferentiation. Moreover, EVs from podocyte have been shown to induce PTEC apoptosis, whereas EVs from PTEC can also travel along the tubule segment and are taken up by adjacent or distant PTECs via kidney injury molecule-1 (Kim-1).

3.1. Tubule-interstitium communication

The tubulointerstitium, comprising multiple cell types, including TECs, fibroblasts, endothelial cells, and immune cells, makes up a significant portion of the kidney. Crosstalk among these cells has an important impact on renal pathological processes [6]. TECs have been recognized as the core of such cellular communication network [41]. After injury, tubular production of EVs is elevated, along with the specific loading of disease-related nucleic acids and proteins. Once released into interstitial space, they can be taken up by fibroblasts and immune cells, enabling them to acquire a proinflammatory and profibrotic phenotype to promote tubulointerstitial inflammation and fibrosis [[10], [42]].

There is strong evidence of TEC/macrophage crosstalk via EVs. In various murine models, tubular exosomes have been demonstrated to be internalized by macrophages and to induce tubulointerstitial inflammation via regulating macrophage polarization [[43], [44], [45], [46]]. For example, TECs under albumin overload released exosomes packaged with CCL2 mRNA, contributing to macrophage infiltration and activation [43]. Hypoxic TEC-derived exosomal miRNA-23a could reprogram macrophages into a pro-inflammatory state via targeting ubiquitin editor A20 [44]. Moreover, multi-omics analysis of EVs secreted from the basolateral side of inflammatory TECs revealed a specialized molecular composition that links to immune cell trafficking and cell-to-cell signaling [47], offering further evidence that tubular EVs released into the interstitial milieu are important mediators in tubular-immune cell communication. Conversely, macrophages also produce EVs that affect TECs fate. Exosomes derived from M1 phenotype were reported to induce pyroptosis in TECs, whereas M2-derived exosomes enriched with miRNA-93–5p showed an anti-pyroptosis activity and supplement of such exosomes could alleviate septic acute kidney injury (AKI) via targeting thioredoxin-interacting protein [48].

Injured epithelia also utilize EV's delivery service to mobilize interstitial fibroblasts and robustly promote renal fibrogenesis. This process can be supported by the genetical or pharmacological inhibition of tubular exosome secretion via Rab27a knockdown or GW4869 treatment, which protects against renal fibrotic lesions in various models of CKD [49]. Several profibrotic growth factors, such as transforming growth factor (TGF)-β [50] and Sonic hedgehog [49], could be shuttled in tubular exosomes and subsequently migrated to fibroblasts to induce their activation and proliferation. Chen et al. reported that β-catenin is a master controller for the tubular secretion of exosomal osteopontin, especially its N-terminal fragment, which binds to CD44 on fibroblast, aggravating the development and progression of renal fibrosis [51]. Apart from proteins, tubular exosomes also bear nucleic acids, including miRNA-21 [52], miRNA-196–5p [53], and miRNA-150–5p [54] that have been suggested in fibroblast activation and collagen synthesis.

Peritubular capillary rarefaction due to endothelial cell damage is another hallmark of advanced renal fibrosis. When endothelial injury occurs, epithelial cells produce proangiogenic factors, such as vascular endothelial growth factor (VEGF), to maintain the normal structure and function of the capillary network [55]. Zhong et al. recently demonstrated that hypoxic TECs can secrete exosomal VEGFA from the basolateral membrane, favoring peritubular endothelial proliferation after ischemia/reperfusion injury (IRI). Interestingly, exogenous supplementation of exosomal VEGF-A was able to recuse vascular rarefaction, improve renal perfusion, and ultimately halt the AKI to CKD transition [56]. Although VEGF exhibits a vascular protective effect, its level needs to be exactly controlled as VEGF can lead to inflammation and fibrosis at the same time.

3.2. Intra-nephron communication

Within the kidney, EVs released into the urinary lumen can travel along the nephron, mediating intra-nephron communication across the glomerular and tubular regions. Mesangial cells have been shown to transfer miRNA-4455 to podocytes via exosomes, resulting in the disrupted autophagy and apoptosis of podocytes [57]. High glucose-treated glomerular endothelial cells could secrete exosomal TGF-β1 mRNA that induce podocytes undergoing epithelial-mesenchymal transition and dysfunction [58]. Exosomes derived from macrophages were found to activate mesangial cells via TGF-β1 pathway [59]. These data suggest that EV-mediated intra-glomerular crosstalk is very likely the important mechanism of glomerulopathy. Moreover, EVs also have a hand in conveying messages between tubular cells, from the upper to downstream segments of tubules. Proximal tubular cell-derived EVs could be internalized by cultured distal tubule and collecting duct cells [60], but direct in vivo evidence for intra-tubular communication is still lacking. Recently, Chen et al. found that kidney injury molecule-1 (Kim-1), a tubular damage marker present on injured proximal tubules, contributes to the cellular uptake of EVs via recognizing phosphatidylserine [61]. As a result, proinflammatory signals shuttled in the EVs were efficiently passed between adjacent or distant TECs, leading to the amplification of tubulointerstitial inflammation in AKI.

EV-mediated crosstalk between the glomeruli and the tubules is supported by reports showing that EVs from podocytes can deliver pathological molecules to tubules and consequently cause tubular injury. Podocytes exerted an increased secretion of EVs after exposure to damaging stimuli, such as mechanical stretch, high glucose, and puromycin [[62], [63]], and EVs from the later condition could promote apoptosis and p38 phosphorylation of TECs via the transfer of miRNA-149 and miRNA-424 in vitro [63]. Even EVs from healthy podocytes were supposed to induce pro-fibrotic responses in cultured TECs [64]. However, this issue should be tested and validated in vivo. A study published in preprint form employed a sequence-based metabolic approach to realize specific labelling of podocyte transcripts, which were subsequently detected in proximal tubular cells, providing strong evidence of the occurrence of physiological transfer of extracellular RNAs from podocytes to TECs [65], but whether EVs are involved in this transmission needs to be confirmed. Furthermore, a recent study reported a potential role of tubule-derived EVs in glomerular lesion. High glucose triggered TECs to secrete EVs containing miR-92a-1-5p, which induced endoplasmic reticulum stress and myofibroblast transdifferentiation in glomerular mesangial cells, resulting in the progression of diabetic nephropathy [66].

3.3. Inter-organ communication

Aside from intrarenal cellular crosstalk, EVs released in the blood may reach any organ and serve as endocrine signals underlying inter-organ communication. Circulating EVs from blood cells or other organs could influence the kidney. For example, in early diabetic nephropathy, activated platelets secreted microvesicles carrying CXC motif chemokine ligand 7, which signal through CXC motif chemokine receptor 1/2 on glomerular endothelium to promote the function and structure damage of endothelial cells [67]. Another study uncovered a natural tropism of cardiac exosome to the kidney using an inducible cardiac-specific exosome-tracking mouse model [68]. However, the potential significance of this heart-kidney communication is still unknown. Besides, whether EVs originating from renal cells have a role in inter-organ crosstalk remains an intriguing question.

Overall, EVs play a substantial role in the flow of biomolecular information inside or outside the kidney. But confined to live imaging methods and super-resolution techniques, it is therefore very challenging to monitor the processes of EV biogenesis, movement, and cellular uptake in the kidney. Furthermore, it would be exciting to illustrate the rules that govern EVs selecting their recipient cells, as we can manipulate this knowledge to intercept noxious cellular communication.

4. EVs as biomarkers for renal fibrosis

The development of renal fibrosis is silent, bringing great difficulties in diagnosis. The commonly used biomarkers in clinical practice are not optimal for detecting early-stage injury [69]. Urine, a window into kidney function and disease, contains a pool of EVs that mainly originate from the kidney, bladder, and prostate. Those EVs derived from kidney cells are able to precisely reflect the pathophysiological status of the kidney and, therefore, hold great promise as diagnostic or prognostic biomarkers for renal disorders [[70], [71], [72]]. This section outlines the potential proteomic and transcriptomic biomarkers within urinary EVs for kidney injury (Table 1).

Table 1.

Examples of urinary EVs as potential biomarkers in kidney diseases.

Disease	Marker	Patient cohorts	Level	Assay used	Outcome/utility	Ref.
Acute kidney injury (AKI)	ATF3	79 sepsis patients with AKI, 60 sepsis patients without AKI, and 8 healthy controls	Increased	Western blot	Elevated levels of uEV ATF3 predicted development of AKI	[73]
Kidney transplantation	AQP2	7 patients with kidney transplantation	Decreased	Western blot	Decreased levels of uEV AQP2 was correlated with urinary concentration defect in the early phase after renal allograft	[74]
	CD1c, CD105, CD133, SSEEA-4	58 patients with kidney transplant	Increased	Flow cytometry	uEV reflected repairing features of the graft and predicted 1-year renal outcome	[75]
Chronic kidney disease (CKD)	miRNA-29c	32 patients with CKD and 7 controls	Decreased	qRT-PCR	uEV miR-29c was correlated with renal function and degree of renal fibrosis	[76]
	CD2AP mRNA	32 patients with CKD and 7 controls	Decreased	qRT-PCR	uEV CD2AP mRNA detected both renal function and fibrosis of kidney disease.	[77]
Diabetic nephropathy (DN)	AQP2, AQP5	35 diabetic patients	Increased	ELISA and Western blot	uEV AQP5 and AQP2 were correlated with the histological class of DN	[78]
	WT1	48 patients with type 1 diabetes mellitus and 25 healthy controls	Increased	Western blot	uEV WT1 was associated with an increase of proteinuria and a decline in renal function	[79]
	ELF3	25 patients with DN and 25 patients with minimal change nephrotic syndrome	Increased	Western blot	uEV Elf3 was a marker for podocyte injury and predicted the decline in renal function	[80]
	miRNA-145	24 diabetic patients and 10 non-diabetic subjects	Increased	qRT-PCR	uEV miRNA-145 was increased in diabetic patients with incipient DN	[81]
	CCL21 mRNA	15 diabetic patients, 28 patients with DN, and 15 healthy controls	Increased	qRT-PCR	uEV CCL21 mRNA was an early marker of renal dysfunction and histological damage	[82]
Lupus nephritis (LN)	HMGB1, HLA-DR, CX3CR1	60 patients with systemic lupus erythematosus	Increased	flow cytometry	HMGB1+, HLA-DR+, and CX3CR1+ uEVs discriminated patients with and without LN, while only HMGB1+ uEV discriminated patients with active and inactive LN	[83]
	miR‐21, miR‐150	45 patients with lupus nephritis and 20 AND 20 healthy controls	Increased	qRT-PCR	uEV detected early renal fibrosis and predicted disease progression in LN	[84]
	miR‐29c		Decreased
IgA nephropathy	CCL2 mRNA	55 patients with IgA nephropathy and 24 healthy controls	Increased	qRT-PCR	uEV CCL2 mRNA reflected active renal histologic injury and renal function deterioration	[85]
Focal segmental glomerulosclerosis (FSGS)	WT1	25 patients with FSGS or SSNS (steroid-sensitive nephrotic syndrome) and 5 healthy controls	Increased	Western blot	uEV WT-1 was an early marker for podocyte injury, monitored progression, and predicted response to therapy	[86]

4.1. Protein biomarkers

Urinary protein biomarker for the early detection of tubular injury have been extensively investigated. Studies in rodent models and patients with AKI identified decreased abundance of ion and water transporters in urinary EVs due to tubule damage. For example, significant reductions in urinary exosomal aquaporin (AQP)-1 and -2 were observed in cisplatin- and IRI-induced AKI [[87], [88], [89]]. Urinary exosomal AQP-2 decline is also correlated with low urine osmolality in kidney transplant recipients [90]. Likewise, the decreased urine excretion of exosomal sodium-chloride cotransporter (NCC) and furosemide-sensitive sodium-potassium-chloride cotransporter (NKCC2) has been used to phenotypically differentiate hereditary salt-losing tubulopathies [74]. However, in the context of urine concentration defects, such as diabetic kidney disease, there were elevated secretion of urinary NKCC2 and AQPs, among which AQP-2 and -5 could predict the progression of diabetic nephropathy [91]. In addition, transcription factors are activated in the injured tubules following AKI and leave clues in the urinary EVs as an early sign of tubular injury. For example, activating transcription factor 3 (ATF3) was continuously present in urinary exosomes and detected earlier than the growth of serum creatinine in rat models of AKI and a small number of human samples, suggesting that urinary ATF3 might be a useful biomarker for the early diagnosis of AKI [78]. Subsequent work with a larger sample size confirmed that urinary exosomal ATF3 is an early sepsis-AKI biomarker with a higher specificity in comparison to urine neutrophil gelatinase-associated lipocalin [92].

Urinary EVs also offer insight into glomerular disease including podocyte dysfunction. There is evidence that Wilms tumour protein (WT1, a molecular marker for podocyte) shifts to the podocyte cytoplasm to be ready for secretion via urinary exosomes, which specifically exist in the urine of patients with glomerular disease such as focal segmental glomerulosclerosis [73]. Similarly, in a cohort of patients with diabetic kidney disease, increased urinary exosomal WT1 was associated with the increase of proteinuria and a decline in renal function [86]. Moreover, another epithelium-specific transcription factor, Elf3, was also measured only in urinary exsosomes in cases of diabetic kidney disease, reflecting an irreversible injury of podocytes [79].

A more recent study has profiled urinary EV surface antigen from a longitudinal cohort of transplanted patients and found that mesenchymal/progenitor cell markers, including CD1c, CD105, CD133, and SSEEA-4, in urinary EVs gradually increased in transplant recipients displaying renal recovery at follow-up. These markers were able to predict renal outcomes after kidney transplant [80]. Furthermore, immune cells also embed some impression in the urinary EV pool. Monocyte-derived microvesicles expressing HMGB1, CX3CR1, and HLA-DR were frequent in urine of patients with lupus nephritis, and HMGB1⁺ microvesicles showed a good diacritical power of disease activity [75].

4.2. Nucleic acid biomarkers

EVs innately carry abundant RNA species that could be exploited as RNA biomarkers for kidney disease. Several miRNAs within urinary EVs have been introduced as candidate biomarkers for renal fibrosis, CKD, and diabetic kidney disease. The miRNA-29 family is one of the most frequent miRNAs in the urine of kidney diseases. Lv et al. firstly reported that the levels of miRNA-29a and miRNA-29c in urinary EVs were reduced in CKD patients, which could distinguish CKD from healthy individuals [83]. Moreover, urinary miRNA-29c correlated with both renal function and the degree of tubulointerstitial fibrosis [83]. Another study that included 45 patients with lupus nephritis identified a urine exosomal miRNA signature of decreased miRNA-29c and increased miRNA‐21 and miRNA‐150, providing a multi-marker panel to detect early renal fibrosis in lupus nephritis [76]. Studies of urinary EVs from diabetic patients with microalbuminuria showed enrichment of miRNA-130a, miRNA-145 [84], miRNA-320c [81], and miR-15b-5p [93], suggesting a potential role of urinary miRNAs to report early-stage diabetic kidney disease.

The mRNA biomarker in urinary EVs also deserves high attention. It is evidenced that urinary EVs are enriched with chemokine mRNAs linked to inflammatory conditions. For example, the level of urinary exosomal CCL2 mRNA was upregulated in patients with IgA nephropathy and correlated with tubulointerstitial inflammation and renal function deterioration [94]. Increased CCL21 mRNA was observed in urinary EVs from rat model and patients of diabetic nephropathy, which might be an early biomarker of renal dysfunction and histological damage [85]. Additionally, podocyte specific CD2AP mRNA in urinary exosomes could be applied in detecting renal function and kidney fibrosis of CKD patients [82].

Although urinary EVs represent a new frontier in liquid kidney biopsy, some significant roadblocks must be passed before they can enter clinical practice. How to quantify urinary EVs and normalize their content is a critical problem to be solved. Urine creatinine has been proposed as a reliable normalization variable for urinary EV concentration in spot urines [77]. Moreover, except for proteins and nucleic acids, other cargoes such as metabolites in the urinary EVs are worthy of exploitation and utilization as renal fibrotic biomarkers. In addition to urine, EVs are also notably abundant in the bloodstream. These circulating EVs, despite their diverse origins, still offer valuable information about kidney damage and disease progression. A distinctive profile of circulating exosomal miRNAs correlated with the degree of albuminuria was identified among patients suffering from diabetic nephropathy [95]. Additionally, in patients with kidney transplants, high levels of miRNA-21 within plasma EVs emerged as a more accurate indicator of interstitial fibrosis and tubular atrophy compared to serum creatinine and proteinuria [96].

5. EVs as nanotherapeutics for renal fibrosis

Currently, there is no effective treatment for renal fibrosis, and the application of EV-based therapy holds promise to change this dilemma. As a matter of fact, EVs are evolving to an optimal vehicle in our bodies and are qualified for drug delivery because they inherently benefit from non-toxicity, immunotolerance, and biological barrier permeability [[12], [14]]. Potentially, certain EVs possess natural renoprotective effects, which are candidates as a standalone therapy in the management of kidney disease. In this section, we will explore therapeutic actions and mechanisms of native or bioengineered EV-based nanotherapeutics for renal fibrosis.

5.1. Native EVs against renal fibrosis

Native EVs from various sources, especially those derived from mesenchymal stem cells (MSC), have robust therapeutic potential. MSCs can be successfully isolated from several tissues, such as bone marrow, umbilical cord, and adipose tissue. Converging evidence from animal models of AKI and CKD shows that MSC-EVs exert pro-angiogenic, anti-apoptotic, immunosuppressive, and anti-fibrotic properties that, in turn, promote the regeneration of injured kidney (Fig. 3) [[97], [98], [99]].

Fig. 3.

Mechanisms of action of MSC-EVs. MSC-EVs exert cytoprotective, pro-angiogenic, immunosuppressive, and metabolic reprogramming effects depending on their cargoes, alleviating renal fibrosis and promoting kidney repair.

Progressive loss of the nephron and capillary network is critical to the pathology of renal fibrosis. Promoting cytoprotection and preventing cell death are thought to sustain the integrity of the structure and function of kidney. A pioneering study in a murine model of glycerol-induced AKI found that bone marrow MSC-EVs could be taken up by tubular cells through CD44 and CD29-dependent internalization and subsequently promote TEC proliferation while inhibiting apoptosis to favor kidney repair [100]. The aforementioned effect was abolished when EVs were pretreated with RNase, suggesting that RNA cargoes of MSC-EVs are responsible for tubule regeneration [100]. Similarly, Cao et al. reported that EVs derived from umbilical cord MSCs have a tropism to the injured kidney suffering IRI, leading to tubular repair through ameliorating cell cycle arrest and apoptosis via miR-125b-5p/p53 pathway [101]. Aside from RNAs, other protective payloads, including growth factors, cytokines, and hormones, have also been identified in MSC-EVs. For example, the specificity protein (Sp1) present in the EVs from induced pluripotent stem cell-derived MSCs was shown to be delivered into renal cells and thereby transcriptionally activated sphingosine kinase 1 and suppressed IRI-induced necroptosis [102]. Moreover, MSC-EVs from different sources were enriched with proangiogenic transcripts and factors (e.g., VEGF), which contributed to the increase of renal capillary density and reduce of renal fibrosis after ischemic AKI [[103], [104]].

Inflammation involving the activation of resident renal cells and infiltrated immune cells is another major driver in the pathogenesis of renal fibrosis. MSC-EVs can subdue such inappropriate inflammatory responses by multiple mechanisms. Data from a wide variety of studies indicate that the immunosuppressive effects of MSC-EVs are attributed to their delivery of anti-inflammatory molecules that minimize the renal production of pro-inflammatory factors and inhibits the infiltration of immune cells such as macrophages and T cells [[105], [106], [107], [108]]. On the other hand, MSC-EVs potently modulate immune cells into an anti-inflammatory state. In a murine model of lupus nephritis, bone marrow MSC-EVs conveyed miRNA‑16 and miRNA‑21 to macrophages and induced the M2 anti-inflammatory polarization. These macrophages could ease kidney injury by eliminating apoptotic debris and promoting Treg cell recruitment [109]. As another example, adipose MSC-EVs harvested from lean pigs could secrete EVs enriched with TGF-β signaling components, provoking Treg cell differentiation and ameliorating renal inflammation and fibrosis in pigs with metabolic syndrome and renal artery stenosis; however, those from pigs with metabolic syndrome failed to do so [110], indicating that the regenerative capability of MSC-EVs may decrease with disease status of the donor. Interestingly, there is evidence that the biologic action of MSC-EVs can be reinforced via cell preconditioning. EVs from erythropoietin-preconditioned bone marrow MSCs displayed an enhanced anti-fibrosis activity, partly because of the alteration of EV miRNAs [111].

Metabolic disorder also has a role in the fibrotic process of the kidney. Healthy kidney generally produces energy through fatty acid oxidation (FAO), but damaged kidney undergoes a metabolic shift toward aerobic glycolysis, which promotes renal fibrosis [112]. Recent evidence supports a potential role of MSC-EVs in metabolic regulation. In a mouse model of unilateral ureteral obstruction (UUO), bone marrow MSC-EVs carrying miRNA-21a-5p were shown to alleviate renal inflammation and fibrosis through inhibiting glycolysis of TECs [113]. In another study, placenta MSC-EVs restored mitochondrial homeostasis and thereby enhanced carnitine palmitoyl-transferase 1A-mediated FAO of TECs, resulting in a significant alleviation of renal fibrosis after IRI [114].

Despite the extensive preclinical research, the translation of MSC-EVs into practice has been relatively slow. To date, only one randomized, placebo-controlled, phase 2/3 clinical trial has been conducted and demonstrated the safety and efficacy of MSC-EVs in grade III-IV CKD patients, as evidenced by the improvement in glomerular filtration rate and the reduction of albuminuria, plasma creatinine and urea levels [115]. Moreover, EVs with renal efficacy can also be isolated from amniotic epithelial cells [116], endothelial progenitor cells [[117], [118]], tubular cells [[119], [120]], and immune cells [[121], [122]]. Study using EVs produced by amniotic epithelial cells reported renoprotection in ischemic AKI through limiting TEC apoptosis, preventing capillary rarefaction and immunoregulation [116]. A similar effect was seen with endothelial progenitor cell-derived EVs that induced kidney regeneration by miRNA-dependent reprogramming of resident renal cells [117]. However, whether EV-based cell-free therapy confers benefits above and beyond their parental cells remains a matter of debate.

5.2. Bioengineered EVs against renal fibrosis

EVs are highly tunable and can be harnessed for the delivery of diverse agents, including small molecule drugs, therapeutic RNAs and proteins. Studies demonstrated that macrophage-derived EVs may be an ideal nanocarrier for anti-inflammatory agents, such as dexamethasone [123] and IL-10 [124], as they could spontaneously migrate toward inflamed tissues via surface integrins. Dexamethasone administration within EVs not only enhanced its efficacy in repressing kidney inflammation and fibrosis but also reduced the side effects [123]. Whereas EVs engineered with IL-10 induced mitophagy of TECs and maintained mitochondrial homeostasis, promoting renal recovery from ischemic AKI [124]. Another anti-inflammatory protein, a super-repressor inhibitor of NF-κB (srIκB), was delivered to the tubulointerstitial macrophages and neutrophils through EVs and, as a result, ameliorated IRI-induced AKI and renal inflammation by inhibiting NF-κB pathway [125]. Moreover, genetically modulated T cells were employed to produce EVs displayed with adhesive molecules, secretory cytokines, and coinhibitory contact molecules, which had the capacity to restore graft function and alleviate renal pathological changes such as fibrosis and immune cell infiltration in a murine model of allogeneic transplantation [126]. These data support the enormous therapeutic potential of EV-based drug delivery in kidney disease.

To achieve a better renoprotective effect, different kinds of functional ligands such as targeting peptides have been introduced to the EV surface to modulate their bioactivity, targeting, and cell internalization [12]. For example, RVG peptide-modified EVs preferentially localized within the injured kidney and were utilized to deliver miRNA-29 to improve renal fibrosis in UUO mice [127]. Tang et al. reported a peptide LTHVVWL that showed a strong affinity to Kim-1. EVs assembled with this targeting peptide and siRNAs targeting P65 and Snai1 could prominently reach injured tubules, attenuating renal inflammation and fibrosis in ischemic and obstructive AKI [128]. Another study described a theranostic strategy based on P-selectin binding peptide-engineered MSC-EVs that exhibited a tendency to the injured kidney. Intriguingly, such versatile MSC-EVs not only provided a tool for the diagnosis of AKI by quantifying renal P-selectin but also mitigated IRI-induced kidney injury by the promotion of renal repair and inhibition of fibrosis [129]. Besides enhancing the targeting capacity of EVs, a study took advantage of the endogenous ligand binding strategy to engineer EVs expressing albumin-binding peptide, resulting in an elevated circulation time and renal retention through megalin-mediated endocytosis [130].

Kidney local delivery via biomaterials offers an alternative strategy that helps increase EV concentration to achieve the desired therapeutic effect. Hydrogels have been widely used to encapsulate MSC-EVs for controlled drug release. Injection of MSC-EVs within matrix metalloproteinase 2 (MMP2)-sensitive peptide nanofiber hydrogels under renal capsule allowed for sustained release of MSC-EVs in the presence of MMP2 and facilitated kidney repair during AKI [131]. Similar benefits were obtained by packaging MSC-EVs within Arg-Gly-Asp (RGD) modified hydrogels, which increased the retention and stability of MSC-EVs in the injured kidney, thereby improving tubular injury and limiting renal fibrosis after IRI [132]. Likewise, a bioresponsive EV-eluting stent was developed for the local delivery of MSC-EVs. After abdominal aorta stenting in rats with renal IRI, a large amount of MSC-EVs were detached from the stent in response to reactive oxygen species and aggregated in the kidney, promoting angiogenesis to support renal regeneration [133].

6. Conclusion and perspectives

EVs have evolved in the body over billions of years, but our understanding of them is just beginning. Still, the EV field has made great leaps forward in the past twenty years. In the development and progression of renal fibrosis, cellular communication mediated by EVs is efficient and robust because they facilitate intercellular information exchange in a high throughput fashion. Their applications in liquid biopsy and targeted therapy possibly bring a new revolutionary in the diagnosis and treatment of kidney diseases. However, numerous challenges remain to be tackled before turning this new theranostic platform into clinical use.

A largely unexplored question in the application of urinary EVs as biomarkers is the tissue and cellular origin of these vesicles. EVs usually carry specific proteins and RNAs from parent cells, and in turn, resolving such tissue/cell-specific information could give an indication of their origin. A recent study performed a genetic source tracking of exosomes in the urine from patients with cancer, showing that the primary exosomal origins are the bladder, endothelial cell, basal cell, monocyte, and dendritic cell [134]. Going forward, it is necessary to clarify the kidney-derived EVs in the total urinary EV pool to favor the discovery of precise kidney disease biomarkers. Furthermore, new equipment that has the potential for high throughput and rapid detection of vesicle biomarkers is certainly required to fully realize the clinical utilization of urinary EVs.

Despite the promising therapeutic application of EVs, many aspects, such as cell culturing, EV purification, formulation, and quality control, remain to be defined. In particular, large-scale manufacturing is a major bottleneck for high quality and sufficient quantity of clinical-grade EVs. Although GMP-grade exosomes have been produced and tested in animal models of pancreatic cancer and myocardial ischemia [[135], [136]], there is still a long way to full-scale production. Moreover, given that EV engineering offers benefit to their bioactivity, future studies will be committed to augment the therapeutic efficacy of EVs through enriching their renoprotective cargoes, through enhancing their targeting to specific kidney cells, through evading their nonspecific clearance by the mononuclear phagocytic system, and through regulating their release within stimuli-responsive biomaterials.

In addition to fundamental issues, the field is still in demand of optimal methods for the isolation, quantification, and characterization of EVs, as well as techniques to evaluate EVs at a single-vesicle scale. Preliminary work has started to apply microfluidic and micromachined technologies in single-vesicle array [137], but these methods still need further improvement and optimization. Overall, ongoing efforts are needed to extend our knowledge on the essential role of EVs in renal fibrosis, enabling us to make better use of them in the diagnosis and treatment of renal disorders.

Declaration of competing interest

The authors declare that they have no conflicts of interest in this work.

Biographies

Taotao Tang received his Ph.D. degree from Southeast University in 2021, and since then, he has worked as an associate research fellow at Southeast University School of Medicine. His current research focus is extracellular vesicle-based therapy for kidney diseases. He has published over 40 peer-reviewed SCI journal papers.

Bi-Cheng Liu is a chair professor and the director of the Institute of Nephrology at Southeast University. His research focuses on unraveling the pathogenesis of renal fibrosis and its translation to clinical science. As PI, he leads over 30 research projects funded by the National Natural Science Foundation of China and the Ministry of Science and Technology. Prof. Liu has published more than 260 papers in peer-reviewed SCI journals and has been acknowledged as a highly cited scientist by Elsevier. Moreover, he has been honored with multiple national and ministry-level government awards.