Exosomes are present in almost all biological fluids and have been isolated from a variety of these fluids as well as from cell culture media (28–39). Exosome isolation is an extensive area of research and can be performed by various methods, including differential centrifugation, density gradient centrifugation, size exclusion chromatography, filtration, polymer-based precipitation, immunological separation, and isolation by sieving (40, 41). Each method has inherent advantages and disadvantages depending on the downstream applications of the isolated exosomes (42–44).

Exosomes have been described as having a “cup-shaped” morphology in electron microscopy. In addition, exosomes equilibrate at densities between 1.13 and 1.19 g/ml on continuous sucrose gradients (39). Identification of exosome specific markers has a vital role in characterizing exosomes and differentiating them from other EVs. These markers are proteins that are specific to the endosomal pathway. These include proteins related to MVB biogenesis, such as Tsg101, Alix, and tetraspanins (CD-63, CD-9, and CD-81); membrane fusion proteins, such as RAB GTPases and Annexins; and signaling molecules, such as cell adhesion molecules, growth factor receptors, and heat shock protein (HSP)-70 and HSP-90 (45–47). The endosomal sorting complex required for the transport (ESCRT) pathway facilitates membrane remodeling and has been implicated in the formation of intraluminal vesicles (48). An ESCRT-independent pathway has also been described as MVBs can be produced in the absence of all four ESCRT complex subunits (49, 50). Finally, the release of exosomes to the extracellular milieu occurs by the fusion of the matured MVB with the plasma membrane, mediated by Rab GTPases (51, 52). Exosomes are enclosed by the phospholipid bilayer of their parent cell and contain a small fraction of cytoplasm taken up from their cell of origin. Hence, exosomes are loaded with a wide variety of molecules, including proteins, RNAs, lipids, and fragments of genomic DNA (53–55) that are present in the parent cell. Exosomes, when released into the extracellular space, can act proximally but can also enter the circulation and cross physiological barriers, eliciting their actions at distal locations (30, 56, 57). The biological function of exosomes relies primarily on the interaction between the exosome and its target cell.

In order to exert their biological functions, exosomes must be taken up and release their contents into the new host cells. Understanding of the mechanisms by which the signals are processed by target cells is still at its infancy. However, a number of key discoveries have been made that aid the understanding of exosome uptake and signaling in the target cells.

Endocytosis of exosomes is via the exosomal trafficking pathway. The endocytosis process can occur via phagocytosis (58) or receptor and raft-mediated endocytosis (59, 60). The phagocytosis mechanism occurs mainly in phagocytic cells. Feng et al. (58) demonstrated that RAW 264.7 macrophages cells effectively internalized exosomes derived from K562 and MT4 cell lines. The internalization was actin-mediated and dependent on phosphatidylinositol 3-kinase (PI3K) and dynamin2. Similarly, Tian et al. (61) showed that pancreatic cancer cells internalized exosomes and the engulfed exosomes were shown to merge with endosomes of the recipient cell and potentially transported to neighboring cells (62).

By contrast, receptor-mediated endocytosis can occur via the classical or non-classical pathway. The former occurs via caveolin or clathrin membrane proteins. The exosomes derived from virus-infected cells were demonstrated to be internalized by target cells via caveolin-dependent endocytosis. Knockdown of the CAV1 gene lead to significantly reduced exosome uptake, proving caveolin-mediated endocytosis (63). Bone marrow-derived mesenchymal stromal cells were shown to take up PC12 cell-derived exosomes via clathrin-mediated endocytosis and contributed to alterations in gene expression through the transfer of miR-21 (64). Similarly, an investigation of uptake of macrophage-derived exosomes by the BeWo cell line and human trophoblast cells showed that uptake is an endocytic process mediated by clathrin (62). In addition, the uptake of exosomes induced secretion of pro-inflammatory cytokines by the placental cells. This study demonstrates a change in placental phenotype induced by exosomes.

On the other hand, the non-classical endocytic uptake of exosomes can occur independent of membrane proteins. It has been reported that exosome uptake by glioblastoma cells occur via lipid raft-mediated endocytosis and is dependent on extracellular signal-regulated kinase-1/2 and HSP27 (60).

Another form of exosome–cell interaction is the adhesion of exosomes to a potential docking site found on target cells. This mode of interaction is facilitated by the presence of transmembrane proteins on the surface of the exosomes. Dendritic cell-derived exosomes express intercellular adhesion molecule-1, major histocompatibility complex, and co-stimulatory molecules which enable the exosomes to interact with target cells via their respective signaling receptors (65–67).

By interacting with the recipient cells, exosomes potentially transfer their cargo which is capable of regulating the biological function of the recipient cells. This then orchestrates diverse signaling pathways and mediates a broad range of physiological and pathological conditions. Cellular responses to the microenvironment have a decisive role in determining the concentration and content of exosomes. This has opened up new avenues for biomarker discovery and therapeutic interventions (68–70).

All cell types in the human body secrete exosomes, including adipose tissue, liver, pancreas, skeletal muscle and placenta during pregnancy. Exosomes released from metabolically active cells could effectively coordinate communication between tissues and initiate metabolic reprogramming in the end target organs. This represents a potential platform for the progression of metabolic disease.

Co-incubation of differentiated C2C12 (muscle cells) with exosomes isolated from C2C12 pre-treated with fatty acid (FA) induced alteration in the gene and proteins expressions in the muscle cells. This indicates that exosomes transfer the effects of FA between the muscle cells and this could disrupt homeostasis and lead to IR in muscle cells. In the same study, C2C12-derived exosomes were injected into mice and were found distributed in various tissues, including metabolic tissues (71).

By utilizing pancreatic cancer-derived exosomes, Wang et al. (72) demonstrated that the exosomes entered skeletal muscle cells, initiated lipidosis, and inhibited glucose uptake. In addition, the exosomes downregulated the insulin and PI3K/Akt signaling pathway and impaired the activity of their downstream target, glucose transporter (GLUT)4. In a reciprocal experiment, it was shown that exosomes isolated from skeletal muscle of high fat diet fed mice were taken up by MIN6B1 cells and mouse islets. The release of the exosomal miRNA changed the expression of mRNAs and genes of the MIN6B1 cells as well as inducing the proliferation of MIN6B1 and islets (73). This suggests that skeletal muscle-derived exosomes could potentially provoke IR in distant cells via exosomes.

Similarly, IR in muscle cells was observed after co-incubation with macrophages treated with adipose tissue-derived exosomes (74). This suggests that adipose tissue-derived exosomes could act as a mediator for the onset of metabolic disease. The studies reviewed here suggest that exosomes secreted by cells from metabolic tissues can coordinate metabolism among tissues and be an effective initiator of the onset of metabolic disease, including diabetes and GDM during pregnancy.

Although exosomes contained a wide variety of molecules, miRNAs has been the center of attention mainly due to its role in regulating gene expression. The exosomal miRNAs are trafficked from their parent cells and the exosomal profile varies according to the physiological conditions of their parent cells. The chromosome 19 miRNA cluster (C19MC) is a unique group of 58 miRNAs exclusively expressed in the human placenta and in undifferentiated cells (75, 76). Growing evidence highlights the presence of these placental-specific miRNAs in exosomes (77, 78). Luo et al. (79) demonstrated that release of C19MC miRNAs is via exosomes and one of the C19MC-encoded miRNA is involved in tumor necrosis factor (TNF)-α signal transduction. miRNA profiling of whole blood and blood-derived exosomes obtained from patients with metabolic syndrome detected similar expression of miR-17, miR-197, miR-509-5p, miR-92a, and miR-320a (80). However, the proportion of exosomal miRNAs is higher than that in their parent cells (81). Interestingly, the exosomal miRNA profile can differ from those of their parent cells (15). The analysis of liver tissue and exosomes (and MV) isolated from a non-alcoholic fatty liver disease (NAFLD) animal model showed enrichment of miR-122 and miR-192 in the vesicles and relative deficiency in the tissue (82).

Hence, the shuttling of miRNAs from parent cells to exosomes involves selective mechanisms. However, there is a paucity of data defining the selective compartmentalization of miRNAs into exosomes.

Although exosomes are produced from cells in a constitutive manner, pathophysiological conditions and stress can modulate exosome biogenesis and release. Recent research provides insight into the selective sorting of proteins and miRNAs into exosomes in conditions of physiological change or pathological stimuli, leading to modification of exosome proteome and RNA profile and, thus, mirroring the microenvironment in the parent cell (83–85).

Hypoxia or low oxygen tension is a stress-induced physiological condition and a classical phenotype in several diseases, such as ischemic CVD, malignancies of diverse origins, obesity, preeclampsia, and physiological challenges such as pregnancy. Hypoxia induces the activation of hypoxia-inducible factor (HIF) which is a key mediator in the cellular adaptation to low oxygen concentrations. HIF, a major modulator of exosome biogenesis and HIF-mediated intercellular exosome signaling, has been identified in a vast array of physiological and pathological conditions (86, 87). Increased endothelial cell migration and angiogenesis is central to the cellular hypoxic response. Increasing evidence suggests the potential relevance of exosomes in mediating these vascular changes. Angiogenic ability has been attributed to exosomes derived from aggressive tumors.

The crucial role of exosomes in remodeling the hypoxia-induced tumor microenvironment has been well elucidated (85, 88–91). Hypoxic tumor exosomes are loaded with unique proteins and have an enhanced capacity for invasiveness, stemness, and tumor progression (87, 92). Hypoxia-induced endothelial dysfunction, a major driver of cardiac disease, is mediated by exosomes (93, 94). During pregnancy, hypoxia triggered exosome signaling increases placental vasculogenesis and augments cytotrophoblastic invasiveness and proliferation as adaptive mechanisms to protect the fetus from oxidative stress (95, 96). In addition, in metabolic disorders such as obesity, exosomes derived from hypoxic adipocytes show an enrichment of lipogenic proteins modulating lipogenic pathways in neighboring adipocytes and pre-adipocytes, thereby transferring characteristics of adipocyte dysfunction (97).

In addition to oxygen tension, the biogenesis and release of exosomes is also affected by glucose concentration. Investigation of the effects of glucose on exosome release showed elevated number of exosomes from trophoblast cells cultured under both high and low glucose concentration (98, 99). Furthermore, the released exosomes induced secretion of pro-inflammatory cytokines from endothelial cells (99). This mechanism potentially mediates the maternal pro-inflammatory profile seen in pregnancies with glucose intolerance. Comparison analysis of plasma exosomal miRNA showed upregulation of miR-326 in diabetic patients compared to controls and this increase negatively correlated with its target, adiponectin (100).

However, the exact mechanism of these alterations in exosome biogenesis and of exosomal miRNA profile under different extracellular glucose concentration is not completely understood. The current body of data suggests that changes in intracellular Ca²⁺ concentration may play a vital role in membrane trafficking, fusion, and retrieval and has intriguing roles in modulating exosome release in response to extracellular glucose (101–103).

Cytokines and hormones play major roles in the initiation and preservation of pregnancy. However, the endocrine functions of placenta are greatly affected by maternal obesity. Maternal metainflammation produces signals opposing the normal regulatory functions of the placenta and contributes to the adverse outcomes observed in obese pregnant mothers. The increase in maternal BMI has been positively correlated with an increase in placental weight (111). A population-based study showed that obese pregnant women had higher placental weight with higher plasma glucose and leptin than their non-obese counterparts at term (112).

Obese pregnancies have a dysregulated maternal cytokine profile with a significant rise in pro-inflammatory cytokines (113, 114). In addition to changes in the plasma, changes to the inflammatory profile of the placenta are also observed in obese pregnancies. An increase in TNF-α turnover in obesity is a well-known phenomenon. Similarly, reports of a significant elevation of TNF-α in the circulation and placenta of obese mothers are consistent (115–119). The placental production of leptin leads to maternal hyperleptinemia with downregulation of placental leptin receptors and resultant leptin resistance in obese mothers (120–122). The analysis of placentae from obese mothers also showed increases in other pro-inflammatory cytokines, such as interleukin (IL)-1β and IL-6 (115, 117). A sequencing study of placental RNA highlighted that levels of IL-12Rβ2, IL-21R, and CX3CR1 were increased while IL-R1, IL-1RAP, CXCR1, CXCR2, CCR3, and ADIPOR1 gene were decreased in placentae of obese women (123).

As a whole, obesity in pregnancy has profound effects, causing systemic inflammation. The increase in circulating pro-inflammatory cytokines from adipose tissue may provoke increased inflammatory cytokines secretion by the placenta and alter placental function. The obesity associated with GDM may have similar or enhanced negative consequences for the placenta.

Pathologically, GDM is characterized by the onset of glucose intolerance of variable severity that is first recognized during pregnancy (124) and a fasting glycemia level ≥92 mg/ml (125). An increase in IR is commonly due to changes in pregnancy-related hormones that occur during early gestation (126). The mother’s inability to secrete sufficient insulin to counteract the IR induced by the gluconeogenic placental hormones may cause the development of GDM (127).

The human placenta is at the materno–fetal interface. Due to its position, the placenta is greatly exposed to various adverse intrauterine conditions and can easily be affected by any changes in its milleu. Glucose is the primary placental energy substrate. Materno–fetal glucose exchange is vital for fetal survival and is observed throughout pregnancy. The gestational changes in maternal glucose metabolism and increased blood glucose level reflect the maternal metabolic adaptations to fulfill the nutrition requirements of the developing fetus. However, this phenomenon is exacerbated in GDM.

The hyperglycemic condition affects trans-placental glucose transport and dysregulation of GLUT activity. In GDM pregnancies, the expression of GLUT1 at the basal membrane was increased twofold with a 40% increase in glucose uptake (128). GLUT1 and mTOR signaling were significantly increased in placentae from GDM pregnancies when compared to normal pregnancies. Interestingly, these changes were associated with a 50% reduction in mitochondrial respiration in trophoblast cells isolated from GDM placentae when compared to the control (i.e., cells from normal placentae) (129). Similarly, utilizing GDM placental explants, a study demonstrated a twofold to threefold increase in glucose uptake (130).

Interestingly, the overexpression of pro-inflammatory cytokines seen in obesity is also observable in GDM placenta. The prominent increase in TNF-α seen in obese pregnancies has also been observed in the maternal circulation and placenta in GDM. The overexpression of TNF-α in GDM placenta is associated with increased fetal adiposity (131, 132). Similarly, Kuzmicki et al. (133) and Lepercq et al. (131) reported an increased IL-8 and leptin expression in GDM placenta, respectively.

The current body of literature suggests that maternal inflammation leads to the over-production of inflammatory cytokines by the placenta that would normally be expressed at significantly lower levels in healthy pregnancies. It is proposed that this enhanced inflammation is associated with the metabolic changes seen in GDM pregnancies. Although these data demonstrate an interaction between maternal obesity and the development of GDM, strikingly, the underlying mechanism that could explain why obesity-associated inflammation is transferred or enhanced in obese-GDM placenta is not understood. Therefore, it can be postulated that other factors mediate the development of GDM by influencing placental function.

Besides secreting hormones and cytokines, the placenta extrudes large quantities of EVs (Table 1) constitutively throughout gestation originating mainly from the syncytiotrophoblastic layer (134, 135). EVs, especially exosomes, are packed with a vast repertoire of proteins, miRNAs and phospholipids that play crucial roles in maintaining feto–maternal communication for healthy pregnancy outcomes (136). These exosomes can be identified through their molecular features. In particular, human placental alkaline phosphatase (PLAP) is an allosteric enzyme synthesized in the placenta. Exosomes isolated from the circulation of pregnant women carried PLAP on their membranes; hence, a PLAP⁺ phenotype can be used to identify placental origin (137, 138).

Maternal plasma is an excellent source for placenta-derived exosomes with their appearance reported as early as 6 weeks of gestation (138, 139) with concentrations varying in accordance with the stage of gestation (135, 137, 138). The presence of immune molecules such as HLA-G and B7 family in PLAP⁺ exosomes demonstrates their role in maternal immunomodulation. This counteracts allograft rejection of the fetus and sustains cellular adaptation in the face of the physiological changes associated with pregnancy (136, 140). Placenta-derived exosomes carry Synctin-1 that mediates trophoblastic syncytialisation (141) and regulates endothelial cell migration, thereby sculpting the maternal–fetal circulation (137). Thus, the involvement of PLAP⁺ exosomes in various processes, such as immunomodulation and vascular changes, can explain their several fold increase in the early stages of pregnancy (137). In general, the concentration of EVs is higher in pregnancy compared to non-pregnant states (142). Furthermore, the concentration of EVs varies in the presence of pathophysiological conditions such as preeclampsia (143) and GDM (144). Recent evidence suggests that individuals with GDM, in particular, may have a distinct exosomal profile when compared to those from healthy pregnancies. The total number of exosomes in maternal plasma between 11 and 14 weeks of gestation is up to twofold greater in women who later developed GDM (diagnosis at 22–28 weeks) compared to those who had a normoglycaemic pregnancy (144).

In GDM, an environment of hyperglycemic and oxidative stress induces exosome release (99). Interestingly, the elevation in total exosome concentration in maternal plasma significantly correlates with maternal BMI, whereas the ratio of PLAP⁺ to total exosome number decreases with higher maternal BMI across gestation (145). In GDM, the augmentation in exosome numbers is due to an increase in total exosomes other than PLAP⁺ exosomes (144). However, the source of these extra circulating exosomes present in obese and GDM mothers remains unknown.

Hypoxia and elevated glucose concentrations are the hallmarks of GDM, and this alters the exosome profile and bioactivity. Cytotrophoblasts cultured under different oxygen tensions (1, 3, and 8%) showed an increased production of exosomes under low oxygen tension (1%), which promoted increased invasion and proliferation of the cells (95). Co-incubation of exosomes with endothelial cells in vitro upregulated the cellular secretion of cytokines. Plasma exosomes isolated from obese and GDM subjects induced the secretion of pro-inflammatory cytokines from endothelial cells from normal and lean pregnancies (144, 145). These findings provide some interesting insights into the role of exosomes in the inflammatory phenomena typically associated with GDM.

Exosomal-mediated miRNA signaling is another fascinating scenario of feto-maternal communication, absolutely essential to maintain the physiological and metabolic harmony between the mother and fetus (79). The dysregulated expression of placental-specific C19MC miRNAs is associated with pathological pregnancies including GDM (146–148). Consistent with this, an increase in the expression of C19MC miRNAs in placental exosomes in the presence of high extracellular glucose was reported (78). Therefore, exosomal miRNA may potentially be involved in placental–maternal signaling.

Maternal body fat mass increases throughout the pregnancy, with accumulation of fat observed on the trunk (188, 189). During pregnancy, appropriate expansion of adipose tissue is vital in order to support nutrient supply to the fetus. However, the hypertrophic growth of adipose tissue is closely associated with metabolic abnormalities and IR (190–192). The ectonucleotide pyrophosphate phosphodiesterase-1 (ENPP-1) is a protein known to induce adipocyte IR. In a recent study, it was demonstrated that adipose tissue from obese patients with GDM expresses high level of ENPP-1 that correlates with the expression of GLUT4 and with insulin receptor substrate-1 serine phosphorylation (193).

Hypertrophy of adipocytes in adipose tissue can impair the functions of adipose tissue, overall. Hypertrophic adipose tissue is associated with excess amount of adiposity and results in a dysregulated secretory profile (194). A higher level of pro-inflammatory cytokines, especially TNF-α and IL-6 has been reported in obese pregnancies (195, 196). The abnormal secretion of adipocytokines is implicated as an essential factor in the development of GDM (197, 198).

Studies to date are suggesting that the relationship between hypertrophic growth of adipose tissue and inflammation is a pivotal factor that causes IR. However, the underlying mechanism by which these adipocytokines affect GDM is not fully understood. While our current understanding of GDM is limited to inflammation induced by adipocytokines, a wide variety of adipose tissue functions may be regulated by adipose tissue-derived exosomes. Therefore, the involvement of adipose tissue-derived exosomes in the development in GDM is possible and understanding of this mechanism is essential.

In addition to soluble factors, exosomes are also involved in various functions of adipose tissue (Table 2). Adipose tissue-derived exosomes have been isolated from culture medium of adipose tissue, adipocytes, and adipose tissue-derived stem cells (ADSC) (74, 199–202). A recent study demonstrated that both 3T3-L1 adipocytes and primary adipocytes secrete large proportions of exosomes (203). In addition, exosomes secreted by adipocytes were reported to be more abundant compared to exosomes secreted by melanoma cells (204). This suggests the probable participation of adipose tissue/adipocyte-derived exosomes in various biological functions.

Although most studies report adipose tissue-derived exosomes within the proposed size range of exosomes (203, 205), Katsuda et al. (206) reported ADSC-derived exosomes that were larger. This indicates that the size range of the exosomes may differ based on the cellular source of isolation. In addition to the identification of exosomal markers, adipose tissue-derived exosomes can be characterized based on the presence of adipose tissue-specific markers, such as fatty acid binding protein 4 (FABP4; adipocyte differentiation marker) and adiponectin (205, 207, 208).

Interestingly, the characterization of exosomes released pre- and post-adipogenesis showed differences in the protein content. Pref-1 and FABP4 were decreased while adiponectin was increased in the post-adipogenesis exosomes. However, there were no changes in the exosomal markers, such as CD9, CD63, TSG101, and Alix (13). This shows that proteins, which are commonly used for bio-marking exosomes, can be used to identify the adipose tissue-derived exosomes.

The release of exosomes has been reported to vary depending on body weight. The concentration of exosomes differs between adipose tissue from lean and obese individuals. The quantification of exosomes isolated from subcutaneous and omental ADSC of lean and obese donors showed that ADSC from obese individuals secretes higher concentrations of exosomes (202). Similarly, the primary adipocytes of obese animals secreted more exosomes when compared to lean animals (204). Surprisingly, exosomes isolated from adipose tissue explants derived from lean and obese individuals showed a higher number of exosomes released by lean adipose tissue (205). The secretion of a higher amount of exosomes by lean adipose tissue may be attributed to the fact that there is a higher number of adipocytes in lean adipose tissue. However, a larger proportion of adipose tissue in obese individuals may explain the isolation of higher number of exosomes from obese individuals (205). In addition, the release of adipose tissue-derived exosomes is influenced by the extracellular milieu. Adipose tissue hypoxia is one of the dysfunctional processes seen in adipose hypertrophy which can cause dysregulated secretion of adipocytokines (209). 3T3-L1 adipocytes cultured under hypoxic conditions released a higher amount of exosomes (97). Thus, the nature or condition of their parent cell determines exosome secretion.