The pancreatic β-cells are responsible for insulin production. Insulin begins as pre-proinsulin and is subsequently modified and cleaved in the endoplasmic reticulum and Golgi apparatus, respectively, to yield insulin and C-peptide (Halban, 1994; Fu et al., 2013; Bunney et al., 2017). Insulin is then stored in intracellular granules until the β-cells receive a signal to release insulin. This signal is commonly triggered by glucose entering the cell via glucose transporter 2 (GLUT2); however, it can also be triggered by amino acids, fatty acids, and hormones (Boland et al., 2017). Once glucose enters the cells, the ATP to ADP ratio rises, triggering the closure of ATP-dependent potassium channels, membrane depolarization, and calcium intracellular influx, which ultimately triggers insulin exocytosis (Seino et al., 2011; Boland et al., 2017). This process can also be triggered by cyclic AMP and is amplified through ryanodine receptor activation (Islam, 2002; Cuíñas et al., 2016).

T2DM is characterized by hyperglycemia resulting from impaired insulin secretion, insulin resistance, or a combination of both (Roden and Shulman, 2019). Insulin regulates glucose metabolism and maintains normal blood sugar levels in the human body by promoting glucose uptake and utilization in peripheral tissues (Stout, 1979). Insulin levels must accurately meet the metabolic demand to avoid hypoglycemia or hyperglycemia. For this reason, the synthesis and release of insulin and the ensuing insulin response in tissues are tightly regulated (Roden and Shulman, 2019). Defects in this regulation can lead to a metabolic imbalance that leads to the pathogenesis of T2DM. Unlike the β-cells that are affected in type 1 diabetes mellitus through autoimmunity, impairment of insulin secretion in T2DM occurs in a multifactorial, non-autoimmune process (Halban et al., 2014; Kawasaki, 2014).

DA is a common underlying pathology in diabetic patients. It is subdivided into macroangiopathy, which affects larger arteries, and microangiopathy, which includes endothelial damage to the vessels between primary arterioles and venules and is largely associated with chronic T2DM complications. The pathophysiology of DA is not well understood; however, it is believed to involve several pathways that are affected by reactive oxygen species (Hammes, 2003). Of importance, is the involvement of angiogenesis in DA as its unequal balance (i.e., of pro-angiogenic and anti-angiogenic factors) is known to be pathogenic for several T2DM complications (Xu et al., 2012).While the majority of exosomal studies focused on organ-specific diabetic complications (e.g., DCM), a few focused on DA itself. These studies specifically focused on the role of exosomes in vascular smooth muscle cells (VSMCs). VSMCs surround the endothelial cells of blood vessels and are known to contribute to oxidative damage in T2DM (Hattori et al., 2000; Inoguchi et al., 2000).

Wu et al. was able to identify that in-vitro VSMCs in high-glucose conditions contained higher levels of carnitine palmitoyltransferase 1 (CPT1A), and that this enzyme contributed to inflammation in the blood vessels as measured by levels of several proinflammatory markers (e.g., TNF-α, IL-6, etc.). Furthermore, the group identified exosomal miR-324-5p to be an inhibitor of CPT1A and then confirmed the findings by comparing VSMCs harvested from patients with and without T2DM (Wu G. et al., 2020). Using the same methodologies, another group was able to identify another DA pathway that involved downregulation of lncRNA UCA1 which led to upregulation of miR-582-5p and ultimately pathogenesis (Yang and Han, 2020). Interestingly, the findings from both these studies found levels of the studied RNAs and enzymes to be modified both within serum exosomes and within VSMCs themselves. This suggests that pathway modification due to exosomal signaling may occur before, during, or after exosomal transport (Wu G. et al., 2020; Yang and Han, 2020).

The studies above highlight exosomes in microangiopathic DA. Similarly, the other complications from T2DM listed in the remaining sections are associated with microangiopathic disease. Complications related to macroangiopathy, such as stroke and atherosclerosis, were omitted from this review because the underlying contributors are generally considered not to be isolated to T2DM alone (i.e., overlap with hypertension) (Verhagen and Visseren, 2011; Frostegård, 2013; Murphy and Werring, 2020). However, it should be noted that the role of exosomes in these pathologies is an active area of research (Venkat et al., 2018; Wang et al., 2018; Guo et al., 2022).

DCM is the occurrence of heart failure in diabetic patients that cannot be attributed to hypertension or coronary artery disease (Gilbert and Krum, 2015). Nearly 50% of T2DM patients with no cardiac symptoms have been reported to have diastolic dysfunction, suggesting that DCM may be severely underdiagnosed (Patil et al., 2011). DCM pathology is multifactorial, with groups reporting the involvement of factors such as oxidative stress, inflammation, the lack of myocardial angiogenesis, lipid accumulation, and fibrosis (Bugger and Abel, 2014). Despite this, the true pathophysiological underpinnings of DCM are yet to be ascertained. For a while, however, it has been known that exosomes play an important role in the processes of myocardial repair and remodeling (Ibrahim et al., 2014; Sahoo and Losordo, 2014; Waldenström and Ronquist, 2014). And more recently, greater focus has since been provided to their role, specifically in models of DCM.

Building on their group’s earlier findings that genetic deletion of matrix metallopeptidase 9 (MMP9) improves cardiac contractility and repair in diabetic patients, the Tyagi group aimed to assess whether cardiac exosomes were regulating MMP9 (Moshal et al., 2008; Mishra et al., 2012; Chaturvedi et al., 2015). Using a db/db diabetic mice model (i.e., downregulation of the leptin receptor), the group screened for four miRNAs that were sequentially similar to the 3′ region of the MMP9. They hypothesized that exercise would upregulate the production of miRNAs which would suggest inhibitory regulation of MMP9 expression. They ultimately found that one of those miRNAs, miR-455, was significantly upregulated in exercise and that its presence significantly downregulated MMP9 (Chaturvedi et al., 2015). Taken together with the group’s earlier findings, this provided the first evidence that the benefits of exercise in DCM could be measured through miRNA biomarkers (Moshal et al., 2008; Mishra et al., 2012; Chaturvedi et al., 2015).

Since then, other groups have discovered additional biomarkers for this pathology. De Gonzalo-Calvo et al. tested a few miRNAs that were known to previously have been signs of myocardial ischemia to assess their correlation with DCM (Bostjancic et al., 2010; de Gonzalo-Calvo et al., 2017). The group assessed miRNAs not only in-vitro and in-vivo but also in humans with DCM without other diabetic complications. Their findings suggest that serum miR-1 and miR-133a are strongly correlated with the presence of myocardial steatosis, as well as some functional heart indicators such as E/Ea ratio. Furthermore, the group was able to create a multivariable logistic regression model that used miR-1 and miR-133a as well as five demographic and basic lab values to generate a predictive c-statistic of 0.88 (95%CI: 0.79–0.97). This group was the first to assess miRNA biomarkers in humans for DCM, and their findings suggest realistic clinical applicability of this biomarker. However, their findings are not without limitations, most notably that 14 to 22 percent of the patients had serum levels that were below the limit of detection, and that only male patients were studied (de Gonzalo-Calvo et al., 2017). Veitch et al. also looked into miRNAs that could serve as biomarkers. However, this group’s findings looked at EVs as a whole and not just exosomes. Still, it was notable that they isolated miR-30d-5p and miR-30e-5p as serum markers in both mice and rats that correlated with cardiac microvascular endothelial dysfunction and preceded diastolic dysfunction. Interestingly, these miRNAs were noted only to be upregulated in the left ventricle when compared with tissue from six other organs. This finding suggests that highly specific miRNA biomarkers for DCM may exist that precede symptomology. Unsurprisingly, the group also discovered this miRNA family to be involved in key in-vitro and in-vivo pathways that have been suggested to be involved in the DCM pathogenesis (Veitch et al., 2022).

Beyond the findings from these two in-depth studies, miRNA families 1, 34, 143, 145, 150, 181, 214, 223, 320, 373, 374, 449, and 761 have been hypothesized to be involved in the pathogenesis of DCM. Most of these families, however, either have not had rigorous studies as to their potential as biomarkers or have not been confirmed to be found in exosomes (Nandi and Mishra, 2018). Several other exosomal components, however, have also been noted to be generally cardio-toxic. For example, miR-21 and miR-146a, which are released by cardiac fibroblasts; miR-320, which are released by diabetic cardiomyocytes; and NADPH, nitro oxid synthetases, and protein disulfide isomerase, which are released by platelets have all been shown to be toxic to cardiomyocytes and/or promote cardiomyocyte hypertrophy (Janiszewski et al., 2004; Azevedo et al., 2007; Halkein et al., 2013; Bang et al., 2014; Indolfi and Curcio, 2014; Wang et al., 2014). This suggests a myriad of possible therapeutic targets for DCM (Ailawadi et al., 2015; Nandi and Mishra, 2018).

One of the groups assessing these deleterious effects noticed that diabetic cardiomyocytes produced exosomes with downregulated heat shock protein 20 (Hsp-20) (Wang et al., 2014; Ailawadi et al., 2015). The group subsequently hypothesized that upregulation of exosomal Hsp-20 could mitigate the deterioration of cardiac function in DCM. The group first confirmed that upregulation of Hsp-20 conferred cardioprotective effects using a streptozotocin-injected Hsp-20 transgenic mouse model. Importantly, the group demonstrated colocalization of Hsp-20 with TSG101, suggesting that Hsp-20 is housed in exosomes released from cardiomyocytes. Further in-vivo and in-vitro analysis demonstrated that these exosomes also included protective proteins (i.e., p-Akt, superoxide dismutase 1, and survivin). That same analysis also confirmed that upregulation of these exosomes leads to reduced oxidative stress, decreased fibrosis, and increased angiogenesis, all of which suggest mitigation of DCM (Wang et al., 2016). Another group identified a separate cardioprotective exosomal heat shock protein, Hsp-70 (Vicencio et al., 2015). Interestingly, the pathway the group identified to be activated by Hsp-70 (i.e., ERK1/2) was no longer activated in a diabetic rat model in spite of them containing exosomal Hsp-70. However, when exosomes were isolated control rats and injected into diabetic rats, cardioprotective pathways were again activated. These findings suggest that exosomal Hsp-70 is likely glycated during diabetic onset and loses its potential to activate cardioprotective ERK1/2 (Davidson et al., 2018).

Beyond Wang et al., a few other groups have also made key findings about exosomal therapeutic targets for DCM. Hu et al., having previously proposed mammalian sterile 20-like kinase 1 (Mst1) to be a key regulatory protein of autophagy and apoptosis in DCM, attempted to characterize whether or not the protein was involved in the pathology as an exosomal component (Lin et al., 2016; Zhang et al., 2016; Zhang M. et al., 2017; Hu et al., 2018). In a mouse model, the group was able to show that Mst1 is created in cardiac endothelial cells and subsequently transported to cardiomyocytes through exosomes. Further in-vitro work suggested that endothelial Mst1 inhibits physiologic autophagy, promotes apoptosis, and suppresses glucose metabolism in cardiomyocytes (Hu et al., 2018). Zhang et al. attempted to assess the therapeutic effects of exosomes as a whole by extracting them from cultured human umbilical cord mesenchymal cells. Shockingly and excitingly, injecting these exosomes into diabetic rats with established DCM was able to reverse the pathology, as measured by histology and functional heart metrics. The group further demonstrated that the effects were mediated through the upregulation of the AMPK-ULK1 signaling pathway, which in turn inhibits pathological autophagy (Zhang et al., 2022).

Ultimately, the pathogenesis of DCM is complicated and not fully understood. However, several groups have demonstrated that its pathogenesis intimately involves a myriad of exosomes and exosomal components (Table 2). Although much of the research done thus far is limited to in vitro models and animal studies, exosomes and exosomal components are appearing promising as biomarkers and therapeutics for DCM.

DN is a complex microvascular complication that generally presents in patients ten to 20 years after diagnosis of T2DM (Young et al., 2003). More than 40% of patients with T2DM develop DN, making it the leading cause of end-stage renal disease in the Western world (Gnudi and Gentile, no date). It is characterized by glomerular basement membrane thickening, mesangial sclerosis, and endothelial dysfunction (Haraldsson et al., 2008). In addition, it is suggested that epithelial-mesenchymal transition may play a role in the development of diabetic nephropathy through the advancement of renal fibrosis (Yamaguchi et al., 2009). Although diabetic nephropathy itself is not the most common cause of death in T2DM, it increases the risk of cardiovascular complications, which are the leading cause of death in T2DM (Gross et al., 2005). Hence, early recognition and subsequent mitigation of diabetic nephropathy are crucial, and exosomes offer that possibility.

Before studies specifically focused on exosomes, one study focused on the potential of proteins within microvesicles. In this observational study, the group found microvesicle-bound dipeptidyl peptidase-4 (DPP-4) activity to be decreased in the serum. Conversely, it was found to be increased in the urine of T2DM patients with DN when compared to non-diabetic healthy controls (Sun et al., 2012). This positive correlation further confirms previous findings that suggested the direct relationship between DPP-4 activity and worsening DN (Ahrén, 2007; Sun et al., 2012). Hence, microvesicular DPP-4 could serve as a biomarker for the diagnosis and progression of DN.

After this, a few other studies identified exosomal proteins associated with DN (Raimondo et al., 2013; Zubiri et al., 2014). Raimondo et al. conducted a proteomic analysis of 286 proteins found in urinary exosomes from Zucker diabetic fatty rats and compared them to non-diabetic rats (Raimondo et al., 2013). The levels of only two enzymes were found to be significantly different: Xaa-Pro dipeptidase and major urinary protein-1 (MUP1). Xaa-Pro dipeptidase, an enzyme involved in the metabolism of proline-containing peptides and proteins, was increased, which could indicate higher metabolic activity or elevated renal protein turnover (i.e., the rate that proteins are synthesized, degraded, and replaced). Conversely, MUP1, a protein mainly produced in the liver and secreted into the urine, was decreased. This may suggest impaired renal function or reduction in the synthesis and secretion of MUP1 in DN. Another study by Zubiri et al. utilized the more reliable ExoQuick reagent-based isolation technique to isolate exosomes from urine (Zubiri et al., 2014). Here, three exosomal proteins in urine were found to be possible biomarkers in DN patients: upregulated ɑ-1-microglobulin/bikunin precursor (AMBP), upregulated lysine-specific methyltransferase 2C (MLL3) and downregulated voltage-dependent anion channel 1 (VDAC1) (Zubiri et al., 2014). AMBP is a glycoprotein involved in inflammation by inhibiting proteases, exerting antioxidant effects, modulating cytokine activity, and influencing immune cell function (Grewal et al., 2005). AMBP has also been reported to be increased in T2DM patients’ urine, in comparison to healthy individuals (Pisitkun et al., 2004; Gonzales et al., 2009; Li et al., 2011). MLL3 is a histone methyltransferase involved in gene regulation and chromatin remodeling (Hosseini and Minucci, 2017). MLL3 has also been linked to the activation of PPAR𝛾, which in turn controls glucose by degrading SOCS5, a negative regulator of JAK/STAT pathway, consequently regulating glucose (Corrales et al., 2018). Finally, VDAC1 is a protein that is primarily located in the outer mitochondrial membrane, where it plays a critical role in regulating the transport of ions and metabolites, particularly the exchange of ATP and ADP. Since mitochondrial dysfunction has been observed in DN, VDAC1 expression or activity could contribute to the onset and progression of this pathology (Pinti et al., 2019). Zubiri et al.’s findings of decreased urinary VDAC1 reinforce the association of DN with increased rate of cell death and renal fibrosis (Zubiri et al., 2014). More recent work by Wen et al. identified that mice with diabetes had reduced expression of exosomes at large by the tubular interstitial cells (Wen et al., 2020). More specifically, their in-vitro Nephroseq analysis revealed a potential role for the enolase 1 protein in the pathogenesis of the disease as it was differentially expressed in various areas within the diabetic kidney and between non-high glucose and high glucose treated renal cells (Martini et al., 2008; Wen et al., 2020). Tao et al. also conducted similar genetic network analysis but focused biomarker study in humans. The group’s work identified adipocyte enhancer binding protein 1 to have a strong ability to distinguish DN patients from healthy controls (AUC = 0.880) and from T2DM patients (AUC = 0.742) (Tao et al., 2021).

Other exosomal cargo, such as miRNA, have also shown potential as biomarkers. Early work by Lv et al. identified reduced exosomal miR-29 and miR-200 families as biomarkers for chronic kidney disease. The group was able to create various models that predicted the extent of renal fibrosis with high AUCs (i.e., 0.9 and above) (Lv et al., 2013). More recent work focused on DN, was able to create a multivariable model using urine exosomal miR-126, miR-146, miR-155, as well as other paraclinical data, to predict between patients with T2DM with and without DN while maintaining high sensitivity (0.963) and specificity (0.935) (González-Palomo et al., 2022). These findings, as well as those of other studies that identified exosomal miRNA biomarkers for DN are summarized in Table 3 and Figure 2 (Eissa et al., 2016a; Eissa et al., 2016b; Delić et al., 2016; Mohan et al., 2016; Xie et al., 2017; Li et al., 2018; Zang et al., 2019).

Beyond their role as biomarkers, exosomes could also be therapeutic for diabetic nephropathy. At the time of the first studies on exosomes as therapies in DN, it was already known that mesenchymal stem cells (i.e., those present in the bone marrow and in mesenchymal tissues; MSCs) function as effective therapeutics for diabetic nephropathy in murine models (Ezquer et al., 2009; Zhang et al., 2013; Abdel Aziz et al., 2014; Ezquer et al., 2015). The mechanism was unknown, however, a leading theory suggested that the therapeutic mechanism of action was via paracrine effects of trophic factors that were secreted by MSCs (Kuroda et al., 2011). Nagaishi et al. attempted to build on this theory, as well as the growing evidence of exosomes from MSCs showing therapeutic potential in studies unrelated to DN, by assessing exosomal effects on diabetic nephropathy (Katsuda et al., 2013; Yu et al., 2014; Nagaishi et al., 2016). Using both streptozotocin-injected mice and high-fat diet mice as diabetic animal models, the group first confirmed the therapeutic effect of MSCs. It tested the therapeutic effect of MSC-conditioned medium. Interestingly, the MSC-conditioned medium mice performed similarly or better than those treated with MSCs as measured by albumin to creatinine ratio, histopathological analysis, and blood glucose levels (Nagaishi et al., 2016). The group further assessed these effects to show that both therapeutics inhibited the infiltration of bone-marrow-derived cells into kidneys, a mechanism that the group had previously shown to be nephrotoxic in the setting of diabetic nephropathy (Yamashita et al., 2012; Nagaishi et al., 2016). Further experiments revealed both MSCs and MSC-condition medium downregulated expression of intercellular adhesion molecule-1, TNF-α, and phosphorylated p38 mitogen-activated protein kinase, which, taken together, suggest a reduction in inflammation. Finally, the improvement of nephrological structural integrity was confirmed by morphological assessment of the proximal tubules, downregulation of transforming growth factor beta (TGF-β) in tubular epithelial cells, indicating less epithelial to mesenchymal transition, and upregulation of zona occludens protein-1, a key marker of tight junction function (Rajasekaran et al., 1996; Fragiadaki and Mason, 2011; Nagaishi et al., 2016). Throughout these experiments, the question remained as to whether the effects seen in the MSC-conditioned medium group were due to exosomes in the medium. Two experiments attempted to address that question. First, the group purified exosomes from MSC-conditioned medium and unilaterally administered them into the subcapsular space of streptozotocin-induced rats. After sacrificing the rats, the two kidneys were compared. The exosome-treated kidneys abated the expansion of renal tubules, atrophy of tubular epithelial cells, and infiltration of inflammatory cells. In addition, downregulation of TGF-β in tubular epithelial cells and upregulation of zona occludens protein-1 were once again noted (Nagaishi et al., 2016). These findings suggest that exosomes are the likely paracrine factor enabling MSC’s therapeutic potential in diabetic nephropathy.

Building on this critical work, other groups identified therapeutic possibilities in exosomes. For example, Jin et al. furthered the therapeutic potential of MSCs (i.e., specifically adipose-derived MSCs) by identifying miR-486 as a likely driver of the benefit seen in in-vivo DN mice models. The group identified that miR-486 downregulated renal Smad1 expression, which inhibited mTOR activation, ultimately leading to the promotion of autophagy and the reduction of podocyte apoptosis (Jin et al., 2019). Hao et al. used similar methods to identify miR-125a as a likely driver of the benefit of exosomal therapy from adipose-derived MSCs seen in in-vivo DN rat models (Hao et al., 2021). The group suggested that miR-125a bound to histone deacetylase 1 which further downregulated endothelin-1 and provided rescue of glomerular tissue integrity and renal function (Tung et al., 2018; Hao et al., 2021). Tsai et al. harvested proximal tubular epithelial cells and mesangial cells to see if exosomes from the former would drive pathological changes in the latter. Their in-vitro exosomal analysis identified miR-92a-1-5p uptake by glomerular mesangial cells and further identified this miRNA to be the driver of myofibroblast transdifferentiation. This suggests miR-92a-1-5p to be involved in the pathogenesis of DN. The group engineered an inhibitor of miR-92-1-5p and showed rescue of nephropathic features both in-vitro and in a DN mice model (Tsai et al., 2023).

DN is a complex microvascular complication that significantly impacts the lives of individuals with T2DM. Identifying and characterizing exosomes in DN have contributed to our understanding of the disease pathology and opened up new possibilities for diagnosis and treatment. In particular, several exosomal proteins and miRNAs have emerged as potential biomarkers for DN. Additionally, exosomes derived from MSCs have shown therapeutic potential in preclinical models. The exploration of exosomes in DN represents a promising avenue for the development of non-invasive diagnostic tools and targeted therapeutic interventions. Continued research in this field will further unravel the intricate interplay between exosomes and DN, ultimately leading to improved patient outcomes and a deeper understanding of this condition. A complete list of exosomal involvement in in-vivo DN studies is presented in Table 3.

DPN is a debilitating complication of T2DM that affects 50% of individuals with T2DM and estimated to have a worldwide prevalence of 425 million people (Nascimento et al., 2016). DPN is most commonly a symmetric, length-dependent sensorimotor polyneuropathy that manifests as sensory disturbances, neuropathic pain, and impaired motor function (Tesfaye et al., 2010; Feldman et al., 2019). The mechanism by which T2DM induces DPN is not fully understood, but is said to involve microvascular disease, neuroinflammation, and oxidative stress (Feldman et al., 2019). Over the recent years, there has been a growing interest in exploring the role of exosomes as a therapeutic intervention for DPN. Jia et al. were amongst the first to explore the effects of exosomes on DPN and focused on those that were derived from high-glucose (v. non high glucose) stimulated Schwann cells. Their findings identified levels of miR-28, miR-31a, and miR-130a to be higher in exosomes derived from high glucose conditions. The group hypothesized that these exosomes, and likely through the action of these three miRNAs, would accelerate DPN (Jia et al., 2018). In-vitro studies confirmed this hypothesis and identified that each of the miRNAs modulate four proteins, an endocytic adaptor protein (NUMB), synaptosome-associated protein 25 (SNAP25), DNA methyltransferase 3a (DNMT3A), and growth-associated protein 43 (GAP43), which are collectively involved in neuroplasticity, axonal growth, and nerve sprouting (Wakamatsu et al., 1999; Johnson, 2003; Frassoni et al., 2005; Nguyen et al., 2007; Feng et al., 2010; Jia et al., 2018). Specifically, miR-28 targeted NUMB, miR-31a targeted SNAP25, and miR-130a targeted both DNMT3A and GAP43; such targeting led to proteomic downregulation. The group’s in-vivo work in diabetic mice confirmed an inverse relationship between the three miRNAs and their target proteins in the sciatic nerves. However, interestingly, this correlation was not seen in the dorsal root ganglions, suggesting this exosomal regulation may be focused to distal nerve fibers (Jia et al., 2018).

In an effort to study exosomes from another cell type, Fan et al. conducted a study using MSC-derived exosomes to understand their effects on DPN. The results of the study revealed that treatment with MSC-exosomes significantly improved neurological outcomes in diabetic mice with DPN. The treatment group had a decrease in the threshold for thermal and mechanical stimuli and an increase in the nerve conduction velocity. Histopathological analysis showed enhanced density of blood vessels and increased numbers of intraepidermal nerve fibers, myelin thickness, and axonal diameters in the sciatic nerves. In addition, Western blot analysis demonstrated a shift from proinflammatory M1 macrophage phenotype markers to anti-inflammatory M2 markers. Notably, although these results do suggest a significant improvement, the rescue did not yield return of nerve function to baseline (i.e., non-DPN group) (Fan et al., 2020). Network analysis identified miR-17, miR-23a and miR-125b as exosomal components that targeted genes involved in the TLR4/NF-κB and receptor for advanced glycation end product signaling, which are known to be crucial in DPN pathogenesis (Toth et al., 2007; Lawrence and Fong, 2010; Fan et al., 2020). Interestingly, the group’s subsequent work focused on miR-146 enriched MSC-derived exosomes which showed accelerated in-vivo therapy through downregulation of the TLR4/NF-κB pathway (Fan et al., 2021). Identification of this pathway, and the multitude of exosomal components that interact with it, suggest a potential mechanism underlying the exosomes’ therapeutic effects (Lawrence and Fong, 2010; Fan et al., 2020; Fan et al., 2021).

Rather than simply injecting the exosomes into the serum, Singh et al. engineered a novel exosomal delivery mechanism using porous nerve guidance channels that could be implanted at the site of injury. The group also created unique DPN rat models as they introduced crush and cut injuries to the sciatic nerve after a 2-month induction with hyperglycemia. Although the results were collected separately, the findings were largely similar across both injury models. Exosomes derived from bone marrow mesenchymal stromal cells (BMSCs) normalized nerve conduction velocity and compound muscle action potential in DPN rats (i.e., returned function similar to healthy control animals). The treatment also led to the recovery of gastrocnemius muscle mass and morphology. Ultimately, for most of these findings, there was no significant improvement whether the porous nerve guidance channels were used or whether the exosomes were injected intramuscularly of note, the study did not conduct a thorough pathway analysis nor exosomal component analysis (Singh et al., 2021). Instead the authors pointed to the mechanism of action of the exosomes hypothetically, stating that previous studies had identified miR-133b, miR-17/92 cluster, brain-derived neurotrophic growth factor, nerve growth factor, insulin growth factor, and fibroblast growth factor as exosomal components of BMSCs that exhibit positive effects on neural regeneration (Xin et al., 2012; Zhang Y. et al., 2017; Qing et al., 2018; Dong et al., 2019; Singh et al., 2021). As these findings were not verified in the paper, they were omitted from Table 2.

More recently, the work conducted by Kasimu et al. focused on astrocyte-derived exosomes. In their study, the group identified the therapeutic component of these exosomes to be miR-125a-5p (Kasimu et al., 2022). Further pathway analysis found that miR-125a-5p functions by downregulating TRAF6, a key signaling molecule associated with inflammation (Cao et al., 1996; Kasimu et al., 2022) TRAF6 is known to activate NF-кB, which leads to rise of pro-inflammatory cytokines or pathogen-associated molecular patterns, and induce neuropathic pain by activation of JNK/MCP-1 pathway (Lu et al., 2014; Liu et al., 2018). The group’s in-vivo diabetic mice model analysis yielded findings that suggested reduced astrocytic activation and downregulation of several inflammatory proteins in the treatment group (i.e., miR-125a-5p tail vein injection) (Kasimu et al., 2022).

Interestingly, all the studies in this pathology have focused on therapeutic interventions (Table 4). In doing so, the studies have revealed important findings regarding the pathogenesis of DPN. As each of the studies has derived their therapeutic exosomes from different cell lines, this confirms previous studies that have mentioned the multitudes of modalities through which DPN develops (Feldman et al., 2019). Further investigation into the cargo and pathways involved in exosomal signaling will enhance our understanding of DPN pathophysiology and create new possibilities for therapeutic interventions.

DR is the most common complication of T2DM and is a leading cause of visual impairment and blindness worldwide (Cao et al., 2021). It is estimated that over 80% of patients with insulin-dependent T2DM and 50% of insulin-independent diabetic patients will develop DR (Stitt et al., 2016). The pathophysiology of DR is characterized by abnormal angiogenesis, loss of retinal pericytes, increased vascular permeability, and proliferation and migration of retinal endothelial cells (Tarr et al., 2013).

Retinal endothelial cells are largely implicated in the pathogenesis of DR as they are the first cells in the eye to sense and respond to changes in circulating blood glucose (Xing et al., 2017). High glucose conditions have been shown to induce proliferation, migration, and tube formation of human retinal endothelial cells secondary to hyperglycemia-induced retinal inflammation (Cao et al., 2021). Aberrations in pericyte-endothelial cell signaling contribute to abnormal angiogenesis, beginning with disruption of the stable association between pericytes and endothelial cells in normal vasculature. This dissociation leads to increased vascular permeability and initiates endothelial cell proliferation and migration into the surrounding tissue. The subsequent dysfunction of the retinal vasculature causes hypoxia and release of vascular endothelial growth factor (VEGF), which further contributes to endothelial cell proliferation, their dissociation from pericytes, vascular leakage, and neovascularization (Raza et al., 2010). Inflammation of the retinal microvasculature simultaneously accelerates the progression of DR. Finally, aberrant activation of the complement system is thought to be involved in the hyperglycemia-induced retinal inflammation and breakdown of the blood-retina barrier. Plasma exosomes have been implicated in this process and have been shown to increase inflammatory markers in T2DM. Huang et al. demonstrated that exosomes rich in IgG are able to activate the classical complement pathway. T2DM increases the levels of IgG-laden exosomes, resulting in a greater degree of exosome-induced complement activation and downstream retinal vascular damage in diabetic patients. Inhibition of exosome-induced complement activation may offer a potential therapeutic approach to prevent or slow the progression of DR (Huang et al., 2018).

Early work by Beltramo et al. identified the role of EVs in the disruption of pericyte-endothelial cell associations. The group’s seminal in-vitro work demonstrated that EVs derived from MSCs are able to enter pericytes, leading to subsequent detachment from endothelial cells, vessel destabilization, increased vascular permeability, and angiogenesis. By culturing MSCs in physiologic and diabetic-like conditions, the group demonstrated that high glucose worsened the retinopathic effects of MSC-derived EVs. Their findings suggested that diabetic-like conditions may contribute to DR pathogenesis through paracrine signaling of EVs (Beltramo et al., 2014). Further work was subsequently conducted by the same group to characterize the relative expression of miRNAs in circulating EVs in T2DM patients with and without retinopathy compared to healthy controls. After confirming that serum EVs isolated from both diabetic groups induced features of DR in vitro, the researchers assessed the expression of 754 miRNAs using qRT-PCR and found 3 that were differentially expressed in the DR group compared to the non-DR T2DM and healthy groups (Table 2). Although this early work focused on EVs, their findings identified a role for microvesicles, and specifically miRNA cargo, in the pathogenesis of DR (Mazzeo et al., 2018).

Other studies have similarly focused on miRNA. For example, Kamalden et al. demonstrated that exosomal miR-15a is secreted by pancreatic β-cells via exosomes in response to high-glucose conditions. Levels of exosomal miR-15a in the serum and retina of diabetic patients were significantly elevated, and the degree of elevation was found to correlate with DR severity. The group further conducted in-vitro analysis that demonstrated circulating exosomal miR-15a is taken up by retinal cells, inducing oxidative stress and contributing to retinal injury by activating apoptosis (Kamalden et al., 2017). Subsequently, Sangalli et al. aimed to see if exosomal miR-15a could be used as a biomarker to be utilized for screening ahead of the DR’s clinical presentation. The group measured ganglionic cell complex thickness as an indicator for early retinal injury in three groups of human subjects (T2DM, impaired glucose tolerance, and healthy controls). This study was the first to demonstrate that ganglionic cell complex thickness reduction preceded the onset of overt retinopathy or T2DM. Furthermore, this reduction was negatively correlated with the concentration of serum exosomal miR-15a in plasma (Sangalli et al., 2020). Taken together, these experiments suggest that exosomal miR-15a is not only a pathogenic contributor of DR, but it can also be used as an early and correlative marker for DR development and progress, even in patients with preclinical T2DM (Kamalden et al., 2017; Sangalli et al., 2020).

Rather than focus on the utilization of exosomal miRNA as a biomarker, Zhang et al. specifically looked at the therapeutic potential of exosomal miR-126. In a previous study, the group identified miR-126 as a factor that may improve DR by enhancing vascular repair and inhibiting hyperglycemia-induced retinal inflammation (Wang and Yan, 2016). In the present study, the group isolated MSC-derived exosomes from human umbilical cord-derived MSCs and cultured them with high levels of miR-126, leading to overexpression of the miRNA in the exosomes. These exosomes were then either injected intravitreally into diabetic rats or co-cultured with high glucose-treated human retinal endothelial cells. They demonstrated that high glucose levels in both the in-vitro and in-vivo models increased retinal inflammation (via upregulation of IL-1β, IL-8, and caspase-1) and induced expression of high-mobility group box 1 (HMGB1), a protein that regulates the inflammatory response. The administration of exosomes overexpressing miR-126 was effectively able to reverse the levels of inflammatory markers and inhibit the HMGB1 signaling pathway (Zhang et al., 2019).

Beyond miRNAs, several exosomal lncRNAs have also been implicated in DR. Ye et al. were the first to study the relative expression levels of various plasma exosomal lncRNAs between T2DM patients with and without DR. Using a logistic regression model to analyze the relationships between DR and lncRNAs, the group identified upregulation of lncRNA DLX6-AS1 in the DR group, suggesting it is likely a risk factor for the pathology. Opposingly, the relative expression of lncRNA PRINS and lncRNA FAM190A-3 was lower in the DR group, suggestive of a role as possible protective factors. After running a univariate analysis, the group identified a multivariate model that includes lncRNA DLX6-AS1 and lncRNA PRINS levels to be a significant predictor for DR with an area under the curve of 0.813 (95%CI: 0.740–0.886) (Ye et al., 2022). Furthermore, Cao et al. identified the possible role of exosomal lncRNA SNHG7 in the pathogenesis of DR by isolating them from MSCs. The group first cultured human retinal microvascular endothelial cells (HRMECs) under high glucose conditions and showed that endothelial damage from high glucose exposure is associated with endothelial-mesenchymal transition, downregulation of lncRNA SNHG7, and upregulation of miR-34a-5p. They then cultured HRMECs with lncRNA SNHG7-overexpressing exosomes, which led to a significant reduction in endothelial-mesenchymal transition and tube formation in HRMECs by downregulating miR-34a-5p expression. These findings suggest that exosomal lncRNA SNHG7 may be therapeutic for DR (Cao et al., 2021).

While the pathophysiology of DR is well understood, exosomes are an emerging field of study in the characterization of DR. They have been shown to have utility in predicting the onset and progression of DR and were recently implicated in the pathogenic mechanisms leading to DR. In-vivo studies have confirmed their potential application in changing the trajectory of DR by improving or worsening retinal damage depending on their miRNA or lncRNA content (Table 5). Targeting the exosomal contents to modify their downstream effects may serve as a therapeutic intervention to prevent the development or slow the progression of DR.

DWH is a well-characterized process that involves cell migration, proliferation, fibrogenesis, angiogenesis, and reepithelialization (Falanga, 2005). In T2DM, this process is significantly impaired, which leads to chronic non-healing ulcers and lower extremity amputation (Rosenberg, 1990). In conjunction with the impaired immune response in T2DM patients, diabetic wounds are a source of potentially life-threatening infections (Brem and Tomic-Canic, 2007). Despite the serious nature of diabetic wounds, treatment options consist primarily of conservative practices, such as dressings, pressure offloading, surgical debridement and revascularization, infection management, and glycemic control, which are supportive rather than curative and often ineffective (Everett and Mathioudakis, 2018). Exosomes offer the opportunity to address the limitations of current practice for DWH management.

The majority of studies on exosomal effects in DWH involve the utilization of MSCs. Zhao et al. investigated such a therapy using exosomes from human adipose-derived MSCs in a diabetic mouse model. The wound healing rate was significantly increased in the exosome-treated group. Further functional assays revealed that these exosomes enhanced diabetic wound healing via a multitude of mechanisms (e.g., reepithelialization, cell proliferation, collagen synthesis, tissue remodeling, skin barrier repair, angiogenesis, and inhibition of apoptosis and inflammation). Finally, the group conducted biochemical pathway analysis that revealed that the exosomes stimulated fibroblast proliferation and migration by negatively regulating MMP1 and MMP3 expression (Zhao et al., 2021).

Li et al. took a different approach to address DWH by first analyzing patient tissue from diabetic foot ulcers (DFUs). After identifying overexpression of miR-152-3p in these tissues, the group was able to subsequently show that exosomes from bone marrow derived MSCs contained lncRNA H19, which was the key inhibitor of miR-152-3p. Further in-vitro pathway analysis identified that miR-152-3p mediated its effects by downregulating PTEN, which is vital for preventing apoptosis and inflammation. Finally, the group confirmed exosomal effects by injecting exosomes with overexpression of lncRNA H19 into a DFU mouse model and demonstrating rescue of DFUs. Taken as a whole, this study suggests that MSC-derived exosomal lncRNA H19 stimulates repair of DFUs by promoting fibroblast proliferation and migration and suppressing apoptosis and inflammation via regulation of the lncRNA H19/miR-152-3p/PTEN axis (Li et al., 2020).

Two other groups studied DWH by focusing on the role of MSCs on senescent cells (Bian et al., 2020; Xiao et al., 2021). Bian et al. initially demonstrated that prolonged stimulation with high glucose accelerated senescence of in-vitro human dermal fibroblasts (HDFs) while simultaneously suppressing HDF proliferation and migration rates. However, when the group introduced exosomes from decidua-derived MSCs effectively, the effects of high glucose were rescued. Further analysis revealed these exosomes inhibited the generation of reactive oxygen species and protected against HDF senescence by suppressing the expression of advanced glycation end-product receptors. Local application of these exosomes in the wounds of diabetic mice enhanced wound healing suggesting that exosomes from decidua-derived MSCs may be a promising candidate for DWH (Bian et al., 2020). The second group, Xiao et al., focused on the effects of exosomes from adipose tissue derived MSCs and demonstrated that these exosomes were able to rescue senescence of in-vitro human umbilical vein endothelial cells induced by both hydrogen peroxide and high glucose. Introduction of these exosomes to a DWH mouse model demonstrated improved wound healing as quantified by scar width and percentage of re-epithelialization. The group went on further to conduct an in-depth genomic analysis of over 1,000 mature miRNAs that ultimately demonstrated the effects of the MSC-exosomes were driven by their miR-146a cargo (Xiao et al., 2021).

Two groups also showed that MSC-derived exosomes improve DWH via their circRNA cargo (Shi et al., 2020; Wang et al., 2023). Leveraging a previous finding that circRNA mmu_circ_0000250 expression is abnormally low in diabetic mice, Shi et al. conducted further in-vitro analysis. Their findings revealed that exosomes secreted by mmu_circ_0000250-overexpressing adipose-derived MSCs decreased high glucose-induced endothelial cell apoptosis and increased angiogenesis by promoting autophagy. mmu_circ_0000250 was found to activate autophagy through inhibition of miR-128-3p and subsequent upregulation of sirtuin 1 (SIRT1) (Shi et al., 2020), a protein known to be involved in regulating inflammation, cell migration, and wound healing (Qiang et al., 2017). In-vivo application of these exosomes into wounds of diabetic mice improved DWH, confirming the in-vitro therapeutic findings (Shi et al., 2020). Wang et al. applied similar methodologies and discovered a similar role for exosomal circRNA astrotactin 1 (circ-Astn1) from adipose-derived MSCs. Circ-Astn1 was found to inhibit miR-138-5p and subsequently upregulate SIRT1. Higher concentrations of circ-Astn1 accelerated full-thickness cutaneous wound healing in diabetic mice (Wang et al., 2023). Interestingly, both studies identified a common end-pathway protein for the effects of exosomal circRNA. These findings suggest even more potential therapeutic options (Shi et al., 2020; Wang et al., 2023).

Another therapeutic option that allows for the forgoing of the invasive biopsy required to generate adipose-derived MSCs is urine-derived stem cells (USCs). Chen et al. explored the therapeutic potential of exosomes from human USCs to address DWH. After demonstrating that these exosomes have therapeutic benefit in-vitro and in-vivo, the group identified that the major mechanism of DWH improvement was through an increase in the rates of reepithelialization, collagen deposition, and vessel formation. Proteomic analysis revealed that a pro-angiogenic protein, deleted in malignant brain tumors 1 (DMBT1), was highly expressed in exosomes from USCs. The group subsequently analyzed the effects of knocking out DMBT1, which yielded return of poor wound healing in both in-vitro and in-vivo studies. Findings from this study demonstrate that local transplantation of exosomes from USCs is able to accelerate diabetic wound healing in mice and suggests a promising therapeutic strategy in humans (Chen et al., 2018).

Rather than use stem cells, Huang et al. bioengineered exosomes to address wound healing. The group first used patient samples to identify miRNA expression levels in diabetic chronic wounds. After identifying miR-31-5p to be significantly underexpressed in such tissue, they showed that in-vitro expression of miR-31-5p promoted angiogenesis, fibrogenesis, and epithelialization. The group subsequently engineered exosomes with miR-31-5p and introduced them to a streptozotocin-induced diabetic rat model. These rats demonstrated a significantly faster rate of wound healing compared to two control groups—those that received non-engineered exosomes and those that received no exosomes. This finding suggests a novel technique towards exosomal therapeutic delivery that bypasses the utilization of stem cells to generate a “cell-free” therapy (Huang et al., 2021).

Taken together, the findings from these studies suggest that the effect of exosomes in DWH is complex and involves a multitude of both RNA and protein cargo. As a result, there are various therapeutic targets that are now available to explore in order to improve DWH management (Table 6). Given that current treatment protocols are largely supportive, implementation of these findings in humans may radically change the future management of this pathology.