Abstract
Mesenchymal stem cells (MSCs) have been demonstrated as promising cell sources for tissue regeneration due to their capability of self-regeneration, differentiation, and immunomodulation. MSCs also exert extensive paracrine effects through release of trophic factors and extracellular vesicles (EVs). However, despite extended exploration of MSCs in preclinical studies, the results are far from satisfactory due to the poor engraftment and low level of survival after implantation. Hypoxia preconditioning has been proposed as an engineering approach to improve the therapeutic potential of MSCs. During in vitro culture, hypoxic conditions can promote MSC proliferation, survival, and migration through various cellular responses to the reduction of oxygen tension. The multilineage differentiation potential of MSCs is altered under hypoxia, with consistent reports of enhanced chondrogenesis. Hypoxia also stimulates the paracrine activities of MSCs and increases the production of secretome both in terms of soluble factors as well as EVs. The secretome from hypoxia-preconditioned MSCs play important roles in promoting cell proliferation and migration, enhancing angiogenesis while inhibiting apoptosis and inflammation. In this review, we summarize current knowledge of hypoxia-induced changes in MSCs and discuss the application of hypoxia-preconditioned MSCs as well as hypoxic secretome in different kinds of disease models.
Impact statement
Mesenchymal stem cells (MSCs) have been applied in numerous cell-based and secretome-based therapies for tissue regeneration. Hypoxic conditions enhance the function of MSCs by increasing proliferation, survival, homing, differentiation, and paracrine activities. A timely up-to-date comprehensive overview of the effect of low oxygen tension to MSC, with emphasis on the influence and molecular mechanism of hypoxia preconditioning toward MSC's functionality is provided, including the therapeutic use of hypoxia-preconditioned MSC as well as hypoxic secretome in various prove-of-concept disease models. This knowledge would contribute to future engineering of MSC culture conditions for improved translational application.
Introduction
Mesenchymal stem cells (MSCs) are multipotent stem cells that can be harvested from various tissues such as adipose tissue (AT), umbilical cord blood (UCB), peripheral blood, and bone marrow (BM). 1 The capability of self-regeneration, differentiation, and immunomodulation makes MSCs an attractive cell source in regenerative medicines. Besides directly differentiating to tissue-specific cells at the site of regeneration, MSCs also exert extensive paracrine effects through release of trophic factors and extracellular vesicles (EVs).2,3 Reported studies have indicated limited survival of MSCs 4 and few differentiated cells 5 after transplantation, implicating that the therapeutic effect of MSCs are predominantly through their paracrine function.
For clinical application, large-scale production of MSCs is often required. However, in vitro expansion of MSCs usually suffers from limitations, such as cell death, senescence, and loss of multipotency. 6 Optimal culture conditions to improve MSC quality have been extensively explored. MSCs are commonly expanded in laboratories under normoxic conditions at 21% oxygen tension, which subject MSCs to abnormally high oxygen stress compared with their physiologic niche. The oxygen concentration in the BM is 1–9%,7,8 AT is 5–9%, 9 and UCB is 1–6%. 10 High oxygen concentrations lead to early senescence and genetic instability. 11 Numerous in vitro studies have thus explored the role of hypoxia on MSC survival, proliferation, migration, and differentiation, and addressed the underlying molecular mechanisms. Although some results are still debatable due to the variability in hypoxic conditions and sources of MSCs, the overall beneficial effects of hypoxic culture conditions support the potential use of hypoxia-preconditioned MSCs in tissue regeneration. The enhanced therapeutic effects of hypoxia cultured MSCs have been validated in various disease models. Furthermore, investigations have been conducted in recent years on the influence of hypoxia on MSC secretome, with the new generation of microarray and sequencing techniques. Secretome from hypoxia-preconditioned MSCs, in the form of either conditioned mediums or isolated exosomes, have been applied in different disease models and shown enhanced repair efficacy.
This review will provide an up-to-date comprehensive overview of the effect of low oxygen tension to MSCs, with emphasis on the influence and molecular mechanisms of hypoxia preconditioning toward multiple MSC functions, including their secretion of trophic factors, and the elicited biological responses in recipient cells. A summary of the therapeutic use of hypoxia-preconditioned MSCs in various proof-of-concept disease models is provided. This knowledge may help to better engineer culture conditions for MSCs to achieve the desired regenerative outcomes.
Molecular Response of MSCs to Hypoxia
Regulation of hydroxylation by prolyl hydroxylase domain proteins (PHDs) is key to hypoxia-induced molecular response in MSCs. Under hypoxic conditions, PHDs, in which hydroxylating function dependent on molecular oxygen, 12 are incapable to hydroxylate various signaling molecules, including hypoxia-inducible factors (HIFs), nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), and protein kinase Akt.
HIFs are a family of DNA-binding transcription factors that regulate a series of hypoxia-related genes when cells respond to low oxygen condition. The dimer of HIF consists of an oxygen-regulated alpha subunit and an intrinsically expressed beta subunit in nuclear. 13 Under normoxic conditions, HIF-α subunit is hydroxylated by PHDs, complexed with Von Hippel-Lindau protein and channeled to degradation in proteasome. 14 The absence of hydroxylation by PHDs under hypoxia allows accumulation and nuclear translocation of HIF-α to form heterodimer with HIF-β, which can then associate with the coactivators p300/CBP to regulate the transcription of multiple hypoxia-related genes. 15
NF-κB is sequestered by PHD-regulated inhibitor of κB kinase (IKK) complexes in normoxia. In hypoxia, nonhydroxylated IKK complex promotes the ubiquitination and degradation of inhibitor of NF-κB (IκB). 16 Consequently, NF-κB is released and translocated to the nucleus to activate the transcription of target genes. 16 Protein kinase Akt also responds to oxygen tension through hydroxylation process. Prolyl-hydroxylation of Akt under normoxia by the Egl-9 family HIF 1 (a PHD2 homolog) inhibits its activity. In hypoxia, Akt is activated and plays a critical role in cell growth and survival. 17
Effect of Hypoxic Culture Conditions on MSC Proliferation
Enhanced proliferation potential of MSCs cultured in low oxygen tension from 1% to 5% has been reported in most studies and is consistent for both human and rodent MSCs derived from different tissue sources. Hypoxia can extend the proliferation lifespan of human UCB and BM-MSCs through upregulation of genes involved in mitotic cell cycle, such as cyclin A2, cyclin B1, and cyclin D1, along with decreased expression of cyclin-dependent kinase inhibitor p27.18,19 Phosphorylation of Akt, a downstream protein kinase of phosphoinositide 3-kinase (PI3K) signaling, is a crucial step in mediating increased proliferation of rat BM-MSCs induced by hypoxia. 20 Upstream mediators of PI3K/Akt signaling, such as Apelin, 21 angiotensin II type 1 (AT1) 22 were reported to be upregulated by HIF-1α in hypoxic rat and mouse BM-MSCs. In hypoxic human AT-MSCs, PI3K/Akt axis can be activated by HIF-regulated glucose-regulated protein (GRP78) and further promote the expression of cell cycle-associated proteins, such as cyclin D1, cyclin E, CDK2, and CDK4, resulting in enhanced proliferation. 23 HIF-2α was also reported to be upregulated under hypoxia, which can directly bind to the mitogen-activated protein kinase (MAPK) promoter and activate the extracellular signal-regulated kinase (ERK) pathway to enhance the proliferation of human placenta-derived MSCs. 24
However, decreased or unaffected proliferation was reported when freshly isolated human BM-MSCs were expanded under 1–5% oxygen tension, in the presence of human platelet lysate supplementation.25–27 These contrasting results could be due to donor variability in cell source and platelet lysate supplement. In addition, the abundant growth factors in human platelet lysate could have surpassed the effect from hypoxia. Further study should be conducted before the adoption of platelet lysate supplement to hypoxic MSC culture. Overall, hypoxic conditioning during MSC expansion can greatly benefit MSC-based therapeutic application, since large numbers of MSCs are often required in clinical application.
Effect of Hypoxic Culture Conditions on MSC Differentiation
Studies exploring the effect of hypoxia during MSC differentiation have reported varying results, from having no effect, to enhancement or inhibition. The variation of differentiation could be caused by the difference in study designs, donor species, and levels of oxygen tension (Table 1). In most reported studies of human MSCs, a hypoxic environment could decrease adipogenesis and osteogenesis, and increase chondrogenesis.
Effect of Hypoxic Conditions on the Multipotency of Human Mesenchymal Stem Cells
Bullets in Table denotes “Yes”.
Adipogenesis/osteogenesis/chondrogenesis were conducted with standard adipogenic/osteogenic/chondrogenic media; in scaffold-free culture conditions.
AT, adipose tissue; BM, bone marrow; MSC, mesenchymal stem cell; UCB, umbilical cord blood.
For decreased adipogenesis under hypoxia, two different mechanisms have been implicated. Wagegg et al. showed that hypoxia downregulated adipogenesis of human BM-MSCs through the activation of HIF signaling, validated by singular knockdown of HIF-1α expression. 28 However, Tamama et al. reported that knockdown of HIF-1α, HIF-2α, and HIF-1β failed to reverse the decrease in adipogenic differentiation induced by hypoxia. 29 Instead, activation of the unfolded protein response (UPR) signaling was responsible since inhibition of UPR in hypoxic MSCs reversed the decrease in adipogenic differentiation. 29 On the other hand, some studies reported increased adipogenesis in human UCB-MSCs and BM-MSCs differentiated under hypoxia.30,31 Jiang et al. proposed that extreme hypoxic condition at 0.2% oxygen tension induced a HIF-1α-regulated lipogenic effect on BM-MSCs with increase in adipogenic genes, such as LPL, CFD, leptin, HIG2, and PGAR. 30
Studies on the effect of hypoxia during MSC osteogenesis have shown conflicting results and mechanisms. Hypoxic condition was reported to decrease human BM-MSC osteogenic differentiation through activation of HIF signaling pathway, in which knockdown of HIF-1 family genes could reverse the decrease of osteogenesis. 29 The role of HIF-1α in hypoxia inhibition of human BM-MSC's osteogenesis has been further implicated by studies using cobalt chloride, a HIF-1α stabilizer, 32 and the activation of TWIST, a downstream target of HIF-1α, 33 both recapitulated the effect of hypoxia inhibition. In contrast, HIF-1α-dependent increase in osteogenic differentiation of human BM-MSCs has been validated by either knockdown expression of HIF-1α, resulting in decreased osteogenesis under hypoxia, 28 or by stable expression of HIF-1α, which significantly enhanced osteogenic differentiation. 34 Human BM-MSCs treated with external HIF-1 has been reported with increased osteogenic ability, which is associated with the downstream RANKL/RANK/OPG pathway. 35
The role of hypoxia during MSC chondrogenic differentiation has been widely investigated due to the regenerative potential of MSCs in the extremely low oxygen tension environment of cartilage. Studies have consistently reported enhanced chondrogenesis of MSCs under hypoxia, with HIF factors implicated as the key mediators. Stabilization of HIF-1α in human BM-MSCs can reproduce hypoxia effects with increased type II collagen and aggrecan expression, 36 while blocking HIF-1α function, abolish the enhancement effect of hypoxia. 37 In hypoxic human AT-MSCs, increased expression of HIF-2α rather than HIF-1α was reported and corroborated with expression of chondrogenic transcription factors, SOX9, SOX5, and SOX6. 38 Activation of PI3K/Akt pathway also contributes to the increased chondrogenic differentiation in human BM-MSCs under 1% hypoxic condition with Akt acting upstream of both HIF-1α and HIF-2α. 39 The proclivity of MSC-derived cartilage to undergo hypertrophic differentiation or form bone in vivo remains a clinical concern for the application of MSCs in articular cartilage regeneration. To this end, hypoxic differentiating conditions, concomitant with the promotion of MSC chondrogenesis, has the added value of inhibiting hypertrophic differentiation,37,40 as well as reducing calcification. 40
Effect of Hypoxia Preconditioning to MSCs
Apart from directly influencing proliferation and differentiation of MSCs, hypoxia preconditioning has significantly enhanced effects on the postexpansion differentiation, survival, and migration ability of MSCs, which has profound implications toward MSC-based therapeutic applications.
Differentiation
In transplantation application of MSCs, the ability to differentiate to target tissue is key for their therapeutic effect. To explore the effect of hypoxia preconditioning on MSC multipotency in vitro, MSCs are expanded under hypoxic conditions for a period of days to weeks before trilineage differentiation under normoxia (Table 1). Adipogenic ability of the hypoxia-preconditioned MSCs is either not affected,25,29 decreased, 41 or increased.27,42,43 The effect of hypoxia preconditioning on MSC osteogenesis is also inconsistent, with most studies showing no effect,25,27,41 while others either increase42,43 or decrease. 29 In comparison, increased MSC chondrogenesis after expansion under 1–5% oxygen tension has been widely confirmed.40,43,44 The consistent reporting of increased chondrogenic ability of hypoxia-preconditioned MSC has huge implications in applying MSC therapy for cartilage regeneration.
Survival
One of the urgent problems in clinical application of MSCs is their limited survival after transplantation,5,45–47 as most of MSCs are perishing within several days due to senescence or apoptosis. For tissue regeneration, it is important for MSCs to maintain their viability to elicit their regenerative effect in situ. Increased senescence has been reported in late passages of human BM-MSCs expanded in normoxic conditions, while hypoxic cultures can significantly reduce the number of senescent cells. 48 Jin et al. showed that hypoxic condition retained human MSCs in a senescence-free state by inhibiting the expression of prosenescence gene p16, through downregulation of ERK pathway. 49 Hypoxia can prevent proliferation-induced senescence of human BM-MSCs through HIF-1α-induced TWIST expression to directly inhibit the expression of E2A and p21, 43 which are related with cell cycle arrest and senescence. 50
The microenvironment of damaged tissue is often both hypoxic and nutrition poor, leading to cell apoptosis and death. Many studies have shown that hypoxic culture condition, on the contrary, could enhance MSC's survivability and inhibit nutrition deprivation-induced cell apoptosis through the autophagy process.51–53 Autophagy is an important catabolic process for cell survival under stress by degradation of unnecessary or dysfunctional cellular components. 54 LC3 II, a component of the autophagosome, is increased in human placental chorionic plate-derived MSCs under hypoxia, together with phosphorylation inhibition of mechanistic target of rapamycin (mTOR), 51 a molecule that suppressed autophagy. 55 Hypoxia preconditioning was also shown to increase leptin that promoted autophagy of mouse BM-MSCs and protected them from apoptosis through modulation of both AMPK and mTOR pathway. 52 In addition, Wu et al. found that hypoxia caused the activation of ERK1/2, inhibition of which, suppressed hypoxia-induced autophagic responses in mouse BM-MSCs. 53
Homing
Cell homing is a directional migration modulated by chemotactic gradient generated from the injury site, recognized by target receptors on the MSC membrane. The ability of MSCs to migrate to the site of injured or inflamed tissue is critical for their efficacy in tissue repair. Stromal cell-derived factor 1 (SDF-1), through interaction with its receptor, CXC chemokine receptors (CXCRs), are crucial for the migration of stem cells. 56 However, human BM-MSCs, expanded in vitro under normoxic condition, were reported to lose CXCR4 expression markedly, 57 which can reduce the response to chemotactic signals and lose homing ability after MSC implantation. 58 Hypoxia-preconditioned MSCs were found to upregulate the expression of SDF-1 receptors, such as CXCR4,59–61 CX3CR1, 60 and CXCR761 on MSCs, with enhanced migration. In addition, upregulated vimentin, fibronectin, and N-cadherin has been reported in hypoxia-conditioned human BM-MSCs, which play important roles in cell flexibility and motility. 19
Application of Hypoxia-Preconditioned MSCs in Tissue Regeneration
Ischemia
Ischemia is caused by a restriction of blood flow to tissues inducing a local shortage of oxygen supply, ultimately leading to the injury of various organs, such as brain, heart, kidney, and limbs. 62 It has been widely reported that, relative to normoxia, hypoxia preconditioning can promote survival of transplanted MSCs and enhance angiogenesis in ischemic tissue, contributing to superior recovery in various ischemic models (Table 2).
Therapeutic Effect Of Delivering Hypoxia-Preconditioned Mesenchymal Stem Cells
MSCs transplanted with collagen scaffold.
αMEM, alpha-Minimum Essential Medium; bFGF, basic fibroblast growth factor; CXCR, CXC chemokine receptors; DMEM, Dulbecco's modified Eagle's medium; FBS, fetal bovine serum; GRP78, glucose-regulated protein; HGF, hepatocyte growth factor; MMP-9, metalloproteinase; NI, not indicated; SDF-1, stromal cell-derived factor 1; TGF-β1, transforming growth factor beta1; VEGF, vascular endothelial growth factor.
Clinically, limb ischemia is commonly caused by atherosclerosis that may lead to tissue necrosis. In a mouse hindlimb ischemia model, significantly increased expression of angiogenic genes (IGF, Ang-1, COX-1, and MCP-1) in the muscle were reported after injection with hypoxia-preconditioned human UCD-MSCs, whereas proinflammatory cytokines, such as interleukin (IL)-1 and IL-20, were decreased. 63 Hypoxia preconditioning was shown to facilitate the proliferation and survival of transplanted human AT-MSCs in the ischemic tissue by increasing the expression of normal cellular prion protein (PrPC). 64 Upregulated secretion of angiogenic cytokines was observed in tissue transplanted with hypoxic AT-MSCs, resulting in increased neovascularization and improved functional recovery. 64 Furthermore, increased engraftment of hypoxic mouse BM-MSCs was shown in mouse ischemic hindlimb model, together with inhibited accumulation of natural killer cells in allogeneic recipients and decreased cytotoxicity. 65
Ischemia/reperfusion in kidney can induce renal fibrosis and inflammation, which are common features in both chronic and acute kidney diseases.66,67 The administration of both rat BM-MSCs and human BM-MSCs preconditioned with 1% oxygen was shown to significantly ameliorate renal fibrosis and inflammation compared with MSCs cultured with 21% oxygen 68 and effectively rescue renal function in a dose-dependent manner after acute kidney injury induced in rat model of renal ischemia/reperfusion injury. 69 Zhang et al. noted that implantation of hypoxia-preconditioned human AT-MSCs significantly improved renal function through increased vascularization, enhanced antioxidation, and inhibited apoptosis. 70 Hypoxia-preconditioned mouse BM-MSCs, but not normoxia-cultured MSCs, were reported to selectively home to ischemic kidneys through SDF-1/CXCR4 pathway that was associated with accelerated functional recovery and reduced cell apoptosis. 61
Cerebral ischemia-induced stroke is the most common central nervous system disease and frequently results in disability and death. 71 Intravenous injection of hypoxia-preconditioned rat BM-MSCs into a rat cerebral ischemia model was shown to inhibit microglia activity in the brain and promote better stroke recovery than normoxic MSCs. 72 Hypoxia-preconditioned rat BM-MSCs was also reported to decrease apoptosis, increase migration ability, and have better integration at ischemic area, resulting in decreased neuronal death and inflammation in the ischemic cortex. 73 Improved neural function and enhanced regeneration of neurons was observed in rats transplanted with hypoxia-preconditioned MSCs, which was dependent on the SDF-1/CXCR4 signaling pathway. 74 Similarly, in a murine ischemic stroke model, enhanced neuroprotective effect with reduced infarct volume and better performance in sensorimotor functional assay were shown in mice transplanted with hypoxia-preconditioned MSCs compared with normoxic MSCs. 75
Musculoskeletal tissue injury
Hypoxia-preconditioned MSCs have been applied to the regeneration of various musculoskeletal tissues, such as cartilage, bone, and tendon after traumatic injury. For cartilage repair, delivery of hypoxia-preconditioned rat AT-MSCs was shown to improve cartilage regeneration, compared with normoxic MSCs, in a rat osteoarthritis model 76 and rabbit articular cartilage defect model. 77 Similarly, in an intervertebral disc degeneration model, significantly increased extracellular matrix deposition in type II and XI collagen was reported in discs treated with hypoxic rabbit BM-MSCs. 78 Despite controversial in vitro osteogenic differentiation effects, delivery of hypoxia-preconditioned BM-MSCs with collagen scaffold in rat and rabbit calvarial defect model was shown to increase bone formation.43,79 Improved tendon regeneration effect was observed in rat Achilles tendon injury model 80 and rabbit patellar tendon injury model, 81 treated with hypoxia-preconditioned MSCs.
Effect of Hypoxic Conditions on MSC Secretome
Given the limited survival of MSCs and very few differentiated cells being observed after transplantation, the therapeutic effect of MSC secretome is gaining more attention in recent years. The secretome refers to the repertoire of factors secreted from MSCs. These factors are released into the surrounding extracellular space in the form of soluble factors or packed into EVs as exosomes and microvesicles. 82
Soluble factors
Hypoxia preconditioning has been applied to influence the secreted protein profile of MSCs and enhance its therapeutic potential for tissue regeneration (Table 3). Hypoxia has been reported to enhance the expression and secretion of trophic factors, including growth factors such as vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), hepatocyte growth factor (HGF), transforming growth factor beta1 (TGF-β1), insulin-like growth factor 1 (IGF-1), fibroblast growth factor 10 (FGF10), and epidermal growth factor (EGF); cytokines such as IL-6, IL-8, chemokine ligand 20 (CCL-20), and monocyte chemoattractant protein-1 (MCP-1); as well as matrix modulators such as angiogenin, tissue inhibitors of metalloproteinases-1 (TIMP-1), and metalloproteinase (MMP).
Biological Functions of Secretome of Hypoxia-Preconditioned Mesenchymal Stem Cells
CCL-20, chemokine ligand 20; EGF, epidermal growth factor; FGF10, fibroblast growth factor 10; IGF-1, insulin-like growth factor 1; IL-6, interleukin 6; MCP-1, monocyte chemoattractant protein-1; miR, micro RNAs; TIMP-1, tissue inhibitors of metalloproteinases-1.
Increased expression of bFGF, VEGF, TGF-β, 83 together with modulation of matrix degradation by angiogenin, 84 TIMP-1, and CCL-2085 can be attributed to the enhanced neovascularization activity of the MSC secretome by regulating vascular permeability, promoting endothelial cell migration, proliferation, survival, and formation of new vessels. 86 The conditioned media (CM) from hypoxia-conditioned MSCs applied in gastric mucosal injury model was reported to result in better repair compared with normoxic CM, with enhancement of angiogenesis and re-epithelization. 85 This therapeutic effect is associated with the phosphorylation of ErK1/2-MAPK pathway following with the activation of COX2-PGE2 axis, in which the neutralization of either TIMP-1 or CCL-20 significantly inhibits the hypoxic CM-induced phosphorylation of ErK. 85 In addition, graft vascularization was also improved with hypoxic CM, attributed to the increased secretion of VEGF, IL-6, MCP-1, and MMP-9. 87
Apart from angiogenic activity, hypoxia preconditioning was also reported to promote the wound healing paracrine ability of MSCs with increased secretion of VEGF and TGF-β1 that can enhance proliferation and migration of human dermal fibroblasts in vitro, as well as promote wound closure in a skin injury model. 88 VEGF and TGF-β1 regulate PI3K/Akt and TGF-β/SMAD2 pathways,89,90 while inhibition of either of these pathways can reduce the migration of fibroblasts. 89 Hypoxia-increased secretion of HGF and VEGF was also shown to stimulate neurogenesis and improve cognitive function in a rat traumatic brain injury model. 91 Furthermore, the significant increase in HGF, VEGF, FGF10, and EGF production of hypoxia-conditioned MSCs was reported to promote tissue remodeling of irradiation-induced damaged salivary gland cells, in which the antiapoptotic effect is attributed to the FGF10-FGFR/PI3K–AKT pathway. 92 Increased secretion of IGF-1 from hypoxic MSCs was also reported to inhibit the apoptosis of hydrogen peroxide-treated intestinal epithelial cell. 93 On liver regeneration, the beneficial effect of hypoxic CM is attributed to the pleiotropic roles of IL-6 in inflammation and metabolism through IL-6–JAK/STAT3 signaling. 94
Extracellular vesicles
EVs play a fundamental role in eliciting paracrine effect of hypoxia-conditioned MSCs. Hypoxia influences both the biogenesis and release of EVs, as well as the content of the secreted EVs. In a HIF-dependent manner, hypoxia promotes microvesicles shedding by regulating small GTPase RAB22A, 95 which is involved in EVs biogenesis. 96 Microvesicles from hypoxia-preconditioned human UCB-MSCs were reported to promote the proliferation and capillary network formation of endothelial cell in vitro and improve blood flow recovery in a hindlimb ischemia rat model. 97 In addition, microvesicles derived from hypoxia-primed human BM-MSCs was shown to enhance osteosarcoma cell viability and migration in vitro, which is partially related with the PI3K/AKT and HIF-1α pathway. 98
Apart from microvesicles, emerging studies have focused on the exosomes, in particular, the abundant nucleic acid contents with critical biological functions (Table 3). MicroRNAs (miRNAs) profiling by quantitative polymerase chain reaction (qPCR) and sequencing has shown that hypoxia significantly increased the expression of various exosomal miRNAs with validated functions, including miR-210,99,100 miR-125b, 101 miR-216a-5p,102,103 miR-126, 104 miR-130a, 105 miR-21, 106 miR-26a, 107 and miR22. 108 Other miRNAs without determined functions are also altered, such as miR-34a-3p, 109 miR-98-3p, 110 miR-181, miR-23a, miR-17, and miR-146a. 111
miR-210 is one of the most consistently unregulated exosomal miRNA in response to hypoxia and is directly regulated by HIF-1a signaling. 112 Augmented secretion of miR-21099,100 was reported to increase vascular density, reduce fibrosis and apoptosis, and increase recruitment of cardiac progenitor cells in the infarcted heart mice model. 100 With gain- and loss-of-function approaches, Cheng et al. validated that exosomal-miR-210 improved the survival of cardiomyocytes in vitro and improved heart function in vivo, and this effect was associated with miR-210-targeted genes, PI3K/Akt and p53. 99 Significant enrichment of miR-125b-5p in hypoxic exosomes was also implicated in the reduction of myocardial infarction, as administration of hypoxic exosomes with miR-125b knockdown significantly increased the infarction area and suppressed postmyocardial infarction cardiomyocyte survival. 101 Feng et al. reported that miR-22-enriched exosomes were secreted by MSCs following hypoxia preconditioning, delivery of which significantly reduced cardiac fibrosis, in which the antiapoptotic effect of miR-22 was mediated by direct targeting of methyl CpG-binding protein. 108 Park et al. observed that miR-26a was significantly increased in hypoxic exosomes, which showed significant cardioprotective effects and reduced ischemia/reperfusion injury by suppressing GSK3β expression. 107
MiR-126, together with miR-130a, and miR-210, have been identified as proangiogenic exosomal miRNAs that can enhance tube formation ability of human umbilical vein endothelial cells. 105 HIF-1α-dependent miR-126 in hypoxic exosomes was shown to promote bone fracture healing. 104 Exosomal miR-216a-5p was reported to exert a protective effect in chondrocytes by promoting their proliferation and migration while inhibiting apoptosis through the miR-216a-5p/JAK2/STAT3 signaling pathway. 103 MiR-216a-5p was also found to involve in hypoxic exosome-mediated microglial polarization through TLR4/NF-kB/PI3K/Akt signaling pathway. 102 Ren et al. reported the antiapoptotic and prometastatic effect of hypoxic exosomes mediated by miR-21-5p, knockdown of which significantly abrogated the macrophage M2 polarization. 106
Further Perspectives and Conclusion
Optimizing the culture environment of MSCs has been actively investigated to maximize the therapeutic outcome of stem cell therapy. Increasing numbers of studies have implicated the beneficial effect of hypoxic conditions to MSC proliferation, survival, homing, differentiation, and paracrine activities. Specifically, the therapeutic advantage of delivering hypoxic MSCs has been validated in numerous ischemia disease models and some musculoskeletal defect models, with the majority of these studies employing MSCs derived from various sources, such as BM, AT, and UCB, thus implicating the importance of hypoxic preconditioning in future strategies for MSC-based regenerative medicine.
Compared with other approaches such as the provision of 3D culture, manipulation of microenvironment through scaffold material engineering, mechanical stimulation and growth factors, and adoption of a hypoxic culture environment is relatively convenient to achieve in vitro, and is available for scale-up production. However, a standardized protocol of hypoxia preconditioning is still controversial due to the discrepant oxygen tension and hypoxia exposure times across studies. To achieve this, an accurate measurement of oxygen level is essential to remove ambiguity with regard to the optimized oxygen levels, and a stable and defined oxygen tension during standard culture procedures should be provided to avoid large fluctuations in oxygen tension. In particular, the hypoxia exposure period in in vitro studies as well as preconditioning before in vivo delivery has ranged from 24 h to as long as a few passages. A comparison of short-term hypoxia exposure (24–48 h) versus that of a long period of hypoxia culture (several passages) are warranted to optimize the culturing conditions for future clinical application.
Although the therapeutic efficacy of hypoxia-preconditioned MSCs has been demonstrated in various animal models, to date, very few clinical trials have been conducted. Only four clinical trials have applied hypoxia-preconditioned MSCs for pulmonary emphysema, 113 acute myocardial infarction,114,115 and ischemic lime disease, 116 and no conclusive outcomes have been reported yet. With the emerging interest in the secretome of hypoxia-treated MSCs, it is interesting to note the recent launch of two clinical trials using secretome of hypoxia-conditioned MSCs for Posterior Cruciate Ligament Injury 117 and Severe COVID-19. 118 We envisage that with the standardization and establishment of the optimum conditions for the generation of hypoxia-conditioned MSCs, or MSC-derived secretome, future work could focus on developing more defined strategies of using hypoxia-treated MSCs in clinical applications.
Footnotes
Acknowledgment
The authors gratefully acknowledge support from the National University of Singapore Tissue Engineering Program.
Authors' Contributions
Y.Y. performed literature research and wrote the article. E.H.L. and Z.Y. edited the article and acquired funding. All authors who participated in the creation of this article are listed.
Disclosure Statement
No competing financial interests exist.
Funding Information
This work was supported by the National Medical Research Council of Singapore (MOH-CIRG19may-0002). Y.Y. was supported by NUS Research Scholarship.
