Effects of Antioxidant Supplements on the Survival and Differentiation of Stem Cells

Although physiological levels of reactive oxygen species (ROS) are required to maintain the self-renewal capacity of stem cells, elevated ROS levels can induce chromosomal aberrations, mitochondrial DNA damage, and defective stem cell differentiation. Over the past decade, several studies have shown that antioxidants can not only mitigate oxidative stress and improve stem cell survival but also affect the potency and differentiation of these cells. Further beneficial effects of antioxidants include increasing genomic stability, improving the adhesion of stem cells to culture media, and enabling researchers to manipulate stem cell proliferation by using different doses of antioxidants. These findings can have several clinical implications, such as improving neurogenesis in patients with stroke and neurodegenerative diseases, as well as improving the regeneration of infarcted myocardial tissue and the banking of spermatogonial stem cells. This article reviews the cellular and molecular effects of antioxidant supplementation to cultured or transplanted stem cells and draws up recommendations for further research in this area.


Introduction
Stem cells are undifferentiated cells, characterized by selfrenewal and the ability to differentiate into several cell types (potency) [1]. They can be totipotent (differentiating into embryonic and extraembryonic cell types), pluripotent (differentiating into cells of the three germ layers), or multipotent (differentiating into cells of a closely related family) [2]. Stem cell research runs with an incredible speed and its applications are under investigation in different medical fields [3,4]. There are two main types of stem cells: embryonic stem cells (ESCs) (present in the inner cell mass of the blastocyst) and adult stem cells (present in different mature tissues to replace dead cells) [5,6].
Induced pluripotent stem cells (iPSCs) are adult cells, genetically reprogrammed to express genes and factors, required for maintaining the properties of ESCs. However, the reprogramming process itself results in oxidative stress by generating high levels of reactive oxygen species (ROS) [7,8], which cause damage to DNA, RNA, and cell proteins and may induce apoptosis [9][10][11]. However, ROS are required in physiological levels to maintain the self-renewal capacity of stem cells and to fight invading microbes [11][12][13][14].
Antioxidants are biochemical supplements that protect cellular constituents from oxidative stress by neutralizing free radicals and terminating the oxidative reaction chain in the mitochondrial membrane [15]. They can be classified into enzymatic and nonenzymatic, endogenous and exogenous [16], and water-soluble (reacting with oxidants in the cytosol or plasma) and lipid-soluble antioxidants (preventing lipid peroxidation of cell membranes) [17].
Over the past decade, several studies have shown that antioxidants can not only mitigate oxidative stress and improve stem cell survival but also affect the potency and differentiation of these cells. In our article, we reviewed the results of preclinical studies that investigated the effects of antioxidants on cultured or transplanted stem cells in an attempt to draw up recommendations for further research in this area.

Induced Pluripotent Stem Cells (iPSCs)
As highlighted earlier, the reprogramming of iPSCs is associated with generation of high ROS levels. Several reports showed that, in comparison to somatic precursor cells, iPSCs exhibit the following criteria: (1) marked protection against nuclear and mitochondrial DNA (mtDNA) damage and (2) significantly lower levels of ROS due to upregulation of intrinsic antioxidant enzymes [18,19]. Dannenmann et al. found a 10-fold decrease in ROS level and a fourfold increase of glutathione (GSH) and glutathione reductase (GR) levels in iPSCs, compared to fibroblasts [18]. In another study by the same authors, they showed that several glutathione Stransferases (GSTs), which act as antioxidant and detoxifying enzymes, were upregulated in iPSCs, compared to their somatic precursor cells [19].
Ji and colleagues reported that mitigation of oxidative stress during cellular reprogramming by antioxidant supplementation protects the genome of reprogramming cells against DNA damage and leads to iPSCs with fewer genomic aberrations [20]. In the same vein, Luo and colleagues [21] found that iPSCs grew well and "stemness" was preserved for up to two months after the addition of a lowdose antioxidant supplement. Moreover, using comparative genomic hybridization (CGH) analysis, they showed that antioxidant supplementation lowered the levels of genetic aberrations in cultured iPSCs [21].
Hämäläinen and colleagues showed that the reprogramming and self-renewal abilities of iPSCs were diminished after subtle increases in ROS levels, originating from mtDNA mutagenesis. However, the addition of two different antioxidants [N-acetyl-L-cysteine (NAC) and mitochondriatargeted ubiquinone (MitoQ)] efficiently rescued these abilities in mutator iPSCs [22]. N-acetyl-L-cysteine raises cellular GSH pool and promotes the processing of H 2 O 2 in the cytosol [23], whereas MitoQ acts upstream to prevent superoxide production within the mitochondria before H 2 O 2 generation [24]. Of note, Hämäläinen et al. highlighted that the therapeutic window of MitoQ for iPSCs is narrow, while high concentrations of NAC were not associated with toxic effects on iPSCs [22].
Interestingly, other reports showed no effect of antioxidant supplementation on the expression of 53BP1 and ATM proteins (two molecules involved in DNA repair pathways) [25][26][27]. Recently, it has been found that high-dose antioxidants downregulates DNA repair-related kinases, which conversely results in genomic instability of iPSCs [21]. Therefore, adjusting the dose of supplementary antioxidants is critical.

Bone Marrow-Derived Mesenchymal (BMSCs) and Hematopoietic Stem Cells (HSCs)
Several studies showed that the ex vivo expansion of ESCs and mesenchymal stem cells (MSCs) [28][29][30][31] and the in vitro expansion of HSCs [32] may cause genomic instability. Through a serial transplantation assay, Jang and colleagues showed that elevated ROS levels reduce the self-renewal ability of HSCs [33]. Therefore, decreasing O 2 concentrations to physiological levels or adding proper dosages of antioxidants can reduce in vitro, culturestimulated aneuploidy, providing potential methods to limit genomic alterations when expanding HSCs in vitro [32,34,35]. Hamid et al. conducted an in vitro study to evaluate the antioxidant effects of Hibiscus sabdariffa L.
(roselle) on bone marrow-derived HSCs. They showed that roselle supplementation increased superoxide dismutase (SOD) expression (at 125, 500, and 1000 ng/mL) and HSCs survival (at 500 and 1000 ng/mL) and protected against H 2 O 2 -induced DNA damage [36]. In another study by Halabian et al., treatment of BMSCs with Lipocalin-2 (Lcn2), a natural cytoprotective factor generated upon exposure to stressful conditions, increased cellular resistance against oxidative, hypoxic, and serum deprivation stresses. Moreover, Lcn2-treated cells showed SOD gene upregulation, increased proliferation, maintained pluripotency, and improved cellular adhesion to culture media upon H 2 O 2 exposure, in comparison to untreated cells [37]. Similarly, Fan and colleagues studied different methods for isolation of BMSCs, aiming at reducing the number of chromosomal abnormalities in isolated cells. They reported that culturing isolated BMSCs at a low O 2 concentration (2%) or with antioxidant (NAC) supplementation increased cellular proliferation and genomic stability, in comparison to cultured cells at normoxic concentrations (20% O 2 ) [38].
Another study by Choi et al. demonstrated that adding ascorbic acid 2-phosphate (AAP) at different concentrations can influence the fate of BMSCs, that is, AAP significantly increased osteogenic differentiation at 50 mM concentration, while a significant induction of adipogenic differentiation with oil droplet formation was noted at concentrations of 250 mM and higher [39].

Cardiomyoblasts and Vascular Progenitor Cells
According to Li and colleagues, culturing cardiac stem cells with antioxidant increased the number and severity of cytogenic abnormalities. This could be explained by the excessive decrease in ROS to subphysiological levels, which may downregulate DNA repair enzymes [40]. In another study by Rodriguez-Porcel et al., modulation of the microenvironment, using antioxidants, leads to a higher rate of cardiomyoblast survival, early after transplantation to the myocardium of small animals [41]. Therefore, oxidative stress blockade may provide a favorable microenvironment for stem cells' engraftment and survival in the heart [42].
Song and colleagues reported increased ROS production during differentiation of human ESCs into vascular progenitor cells (CD34+ cells) due to increased activity of NADPH oxidase-4 (Nox4) enzyme. They found that moderate ROS scavenging, using selenium, enhanced the vascular differentiation of human ESCs, while complete ROS scavenging, using NAC, totally inhibited the vascular differentiation of these cells. This confirms that a minimal level of ROS is required for vascular stem cell differentiation to occur [43].

Neural Stem Cells (NSCs)
Neural stem cells are multipotent stem cells that have been suggested as a therapeutic agent to enhance the recovery of injured tissues in neuroinflammatory diseases [44]. Park and colleagues tested the effects of GV1001, a novel antioxidant agent, derived from human telomerase reverse transcriptase, on in vitro-cultured mouse NSCs. They showed that GV1001 treatment attenuated the effects of H 2 O 2 exposure, reduced lipid peroxidation and mtDNA mutation, and induced the expression of survival-related proteins [45]. Hachem et al. reported that treatment of NSCs, isolated from the spinal cords of transgenic mice, with brain-derived neurotrophic factor improved cell viability by increasing the levels of GR and SOD enzymes; however, it had no effect on cellular proliferation [46].
Nitric oxide (NO) and nitric oxide synthase (NOS)dependent signaling pathways have been implicated in different neurodegenerative diseases [47,48]. Moreover, NO levels were linked to neural precursor cell (NPC) survival and cell fate determination [49], that is, elevated levels of NO suppress NSC proliferation and enhance differentiation of NPCs into astrocytes [50,51]. Melatonin is a hormone synthesized in the pineal gland [52] with indirect antioxidant abilities through induction of antioxidant enzymes [53] and inhibiting NO production in glial cultures through p38 inhibition [54]. It has been shown to protect NSCs against lipopolysaccharide-(LPS-) induced inflammation [52]. Moreover, Negi et al. demonstrated that melatonin mitigates neuroinflammation and oxidative stress via upregulating nuclear factor (erythroid-derived 2) (Nrf2) [55], a transcription factor which stimulates the PI3K-Akt survival signaling pathway [56,57] and increases the expression of the antioxidant enzyme heme oxygenase-1 (HO-1) [55].
To test the effects of in vitro antioxidant supplementation, Petro et al. divided male rats with experimentally induced thromboembolic stroke into four groups: normal rats, untreated rats with stroke, rats receiving tissue plasminogen activator (tPA) only, and rats receiving tPA + CAT/ SOD (loaded on nanoparticles) at three hours post stroke. Two days later, brain tissue samples were harvested for analysis. Brain sections from the untreated group showed evidence of NSC migration through the rostral migratory stream (through detection of NSCs markers, such as nestin, GFAP, and SOX2), confirming the occurrence of neurogenesis following stroke. However, brain tissue samples from the tPA-alone group showed reduction in NSCs migration, indicating that tPA treatment suppresses neurogenesis, either directly or through reperfusion-induced ROS generation injury. Interestingly, tPA + Nano-CAT/ SOD treatment restored and significantly increased NSCs migration [58].

Human Adipose-Derived Stem Cells (ADSCs)
Adipose-derived stem cells are multipotent stem cells that can be isolated from the human adipose tissue and are capable of in vitro expansion. Sun and colleagues reported that both hypoxia and antioxidants promoted ADSCs proliferation by raising the number of cells in the S phase, but the maximal increase in cell number was produced in the presence of antioxidants [59]. Hypoxia is believed to influence the secretion of several growth factors [60,61], such as insulin-like growth factor and hepatocyte growth factor [62], while antioxidants increase the expression of stemness genes (CDK2, CDK4, and CDC2) and the differentiation potential of ADSCs [59]. Another study by Higuchi et al. found that lentivirus-mediated NADPH oxidase-4 (Nox-4) overexpression did not increase ROS production in insulin, dexamethasone, indomethacin, and 3-isobutyl-1-methylxanthine (IDII)-stimulated ADSCs [63]. This finding was later explained by the increased expression of endogenous antioxidants, such as SOD and CAT during adipogenesis [63,64].
Yang et al. showed that treatment of ADSCs with fullerol (a polyhydroxylated fullerene) potentiated the expression of the transcription factor FoxO1 and its downstream genes, such as Runx2 and SOD2. Moreover, it enhanced the osteogenic activity of ADSCs, as evidenced by increased mineralization and expression of osteogenic markers (Runx2, OCN, and alkaline phosphatase) [65]. Wang and colleagues showed that pretreatment of ADSCs with NAC (3 mM) or AAP (0.2 mM) for 20 hours suppressed advanced glycosylation end product-(AGE-) induced apoptosis via a microRNAdependent mechanism by inhibiting AGE-induced overexpression of miRNA-223: a key modulator of intracellular apoptotic signaling [66].

Human Periodontal Ligament Cells (hPDLCs)
In a recent study, Chung and colleagues showed that treating hPDLCs with deferoxamine (DFO), an iron chelator, results in a dose-dependent elevation in ROS levels, 24 hours after treatment [67]. The same finding was reported in rabbit cardiomyocytes [68] and normal human hepatocytes [69]. However, DFO has the ability to act on Nrf2, increasing its nuclear translocation and the expression of its target genes, including GST and glutamate cysteine ligase (GCL) [67]. Therefore, DFO has both beneficial (Nrf2-mediated antioxidant effect) and cytotoxic (increased ROS levels) effects. GSH depletion, using buthionine sulfoximine (BSO) and diethyl maleate (DEM), was shown to inhibit DFO-stimulated hPDLC differentiation into osteoblasts [67]. Moreover, GSH depletion was also reported to repress myogenic differentiation of murine skeletal muscle (C2C12) cells [70] and phorbol-12-myristate-13-acetate-(PMA-) stimulated differentiation of human myeloid cell line (HL-60) [71].    (ii) There is an optimal level of ROS in stem cells, above which genomic instability occurs due to ROS-induced DNA damage and below which DNA repair enzymes are not activated to maintain the DNA stability.

Human embryonic stem cells (ESCs). (i) Ascorbic acid significantly increased
ESCs differentiation into cardiac myocytes in a dose-dependent manner, as evidenced by increased expression of the cardiac specific gene (myosin heavy chain (MHC)).
(ii) Ascorbic acid, independent of its antioxidative property, induced ESCs differentiation into cardiac myocytes.
This was evidenced by the absence of this effect with other antioxidants, such as NAC, Tiron, and vitamin E.
Ascorbic acid increased the expression of cardiac muscle genes, such as GATA4, Nkx2.5, α-MHC, β-MHC, and atrial natriuretic factor (ANF), with subsequent cardiac-specific protein production. (ii) At the molecular level, GV1001 induced the expression of survival-related proteins and diminished the expression of apoptosis-related proteins.

Neural stem cells (NSCs
The neuroprotective effect of BDNF is exerted through its ROS-scavenging activity and induction of antioxidant enzymes, such as GR and SOD. Moreover, significant reductions in apoptotic features were noted in BDNFtreated cells, compared to the control group. Song et al. [52] Melatonin (100 nM).
Neural stem cells (NSCs) from mice embryonic cortical tissue. (i) Melatonin significantly reduced LPSinduced toxicity and apoptosis of NSCs through reducing nitric oxide (NO) production and inducing antioxidant enzymes.
(i) Melatonin increased the expression of multiple transcriptional factors, involved in NSCs proliferation, self-renewal, and differentiation, such as orphan nuclear receptor TLX and fibroblast growth factor receptor-2.
Sun et al. [59] N-Acetyl-L cysteine (NAC) In antioxidant-supplemented media, PCR showed diminished levels of cyclin-dependent kinase inhibitors (CDK: important cell cycle regulators that control entering S1 phase), with enhanced expression of stemness-related genes, compared to the control group.
Lyublinskaya et al. (iii) Adding antioxidants to the medium after proliferation induction and before initiation of S1 phase blocked S1 transition. Antioxidant did not have the same effect when added after S1 initiation.
(i) Cells, treated with antioxidants, showed expression of the proliferative marker (Ki-67), which is absent in the nucleus of quiescent cells, indicating that the cell left the quiescent state and was arrested in G1 phase.
(ii) Antioxidants, through dosedependent reduction of ROS levels, can be used to control cellular proliferation during in vitro culturing.
Adipose-derived stem cells (ADSCs) from the subcutaneous adipose tissue from a female patient, undergoing abdominoplasty. (i) Ascorbic acid significantly increased ADSCs proliferation, preserved cellular stemness and increased the potentiality for adipogenic, hepatic, neural and osteogenic differentiation.
(ii) In AAP-induced cell sheet, there was a significant increase in the genetic expression and secretion of collagen, laminin, and fibronectin proteins.
(iii) The AAP-induced cell sheet improved wound healing in a murine wound model, which indicates the possibility of ADSCs differentiation in a non-mesenchymal lineage.
(i) Adding AAP to ADSCs increased the expression of stemness-related proteins.
(ii) Using other antioxidants, such as NAC did not show an increase of stemness markers. In contrast, adding a collagen synthesis inhibitor abolished AAP-induced overexpression of stemness proteins, indicating that the involved mechanism in AAP action is collagen synthesis, not ROS scavenging.
Wang et al. [66]  (ii) Interestingly, a similar increase in CD31+ endothelial cells in NACtreated and untreated MDSCs was observed.

Muscle-Derived Stem Cells (MDSCs)
According to Drowley and colleagues, injection of injured skeletal muscles with NAC-treated MDSCs significantly increased muscle regeneration, compared to muscles injected with untreated or DEM-treated MDSCs. The direction of scar tissue formation was opposite the direction of the host muscle regeneration [72]. Additionally, they showed an improved survival of NAC-treated MDSCs, probably due to stimulation of extracellular signal-regulated kinase (ERK) pathway, as evidenced by decreased survival of NAC treated cells after inhibition of the ERK pathway [72,73]. Moreover, they demonstrated that experimentally infarcted hearts, injected with NAC-treated MDSCs, showed a more significant reduction in the percentage area of collagenous scar tissue than hearts injected with either untreated, DEM-treated, or phosphate buffered saline-(PBS-) treated MDSCs. There was no difference in myocardial scar formation between hearts injected with DEM-treated MDSCs and those injected with PBS [72].

Spermatogonia Stem Cells (SSCs)
Cryopreservation of spermatogonial stem cells, in the presence of catalase (CAT) and α-tocopherol (α-TCP), promoted cell viability and suppressed apoptosis through inducing the expression of the antiapoptotic BcL-2 gene and inhibiting the expression of the proapoptotic BAX gene [74]. In other studies, cryopreservation with antioxidants could promote cell enrichment and increase the efficiency of colony formation in isolated SSCs [75,76]. Spermatogonia-derived colonies showed increased SSC marker activity, enhanced expression of self-renewal genes, such as promyelocytic leukemia zinc finger (Plzf) protein and DNA-binding protein inhibitor ID4, and suppressed expression of the protooncogene (c-kit) in both CAT and α-TCP treated groups [74]. This technique can increase the possibility of SSCs banking for men with malignant diseases and promote the resumption of spermatogenesis in SCCs recipients. A summary of the design and main findings of included studies is illustrated in Table 1.

Discussion
Our review highlights that antioxidants can influence stem cell activities by [1] mitigating oxidative stress through neutralization of free radicals and increasing the expression of antioxidant enzymes and [2] influencing the differentiation fate of precursor stem cells. Further beneficial effects of antioxidant treatment include increasing genomic stability, improving the adhesion of stem cells to culture media, and enabling researchers to manipulate stem cell proliferation by using different doses of antioxidants. Figure 1 summarizes the effects of antioxidants on different types of stem cells.  We also discussed that a physiological level of ROS (oxidative optimum) is needed for proper differentiation of stem cells, especially for proper cardiogenesis and vasculogenesis [40]. These findings can have several clinical applications, such as improving neurogenesis in patients with stroke and neurodegenerative diseases, as well as improving the regeneration of infarcted myocardial tissue and the banking of SCCs.
Antioxidants are prevalent supplements worldwide. However, little is known about their cell-type-specific actions. It has been shown that a therapeutic dose may vary between different cell types: a dose that rescues a pathology in one tissue may roughly challenge the function of another [22]. Therefore, there is a need for dose-effect studies on antioxidants to confirm their safety as nutritional supplements or therapeutic agents-particularly in the case of antioxidants accumulating in the mitochondria. Our review also showed the potential of some endogenous molecules, such as melatonin, BDNF, and the adipokine (lipocalin-2) in preserving stem cell viability and differentiation potential. Whether these compounds can be used in future clinical applications of stem cells and whether other endogenous molecules with proven antioxidant activities, such as adiponectin [84], can be useful in this regard require further investigation.

Recommendations
(1) Multiplicity of stem cell sources within the body (different home environments) and their variable ROS scavenging capacity make them susceptible to oxidative stress at different thresholds. Therefore, we tried to review each stem cell type as a separate entity and we believe that clearing those differences on the molecular and genetic levels will optimize the clinical application of stem cells in different medical fields.
(2) Most of stem cell characteristics are established within in vitro culturing environments. More in vivo studies are required to define their interactions within the body. Furthermore, few in vivo studies have focused on the long-term survival of transplanted stem cells; therefore, this should be the interest of future studies.
(3) The effect of ROS level and redox state on the longterm oncogenicity of stem cells should be further investigated prior to in vivo clinical trials.

Conclusion
Using antioxidants can improve the viability and self-renewal capacity of stem cells and affect their differentiation potential. More research is needed on the dose-effect association and cell-type-specific actions of antioxidant before applying these findings in human therapeutic trials.