The Role of Hydrogen Peroxide and Peroxiredoxins throughout the Cell Cycle.

Hydrogen peroxide (H2O2) is an oxidizing agent that induces cellular damage at inappropriate concentrations and gives rise to an arrest during cell cycle progression, causing cell death. Recent evidence indicates that H2O2 also acts as a promoter for cell cycle progression by oxidizing specific thiol proteins. The intracellular concentration of H2O2 is regulated tightly, enabling its use as a cellular signaling molecule while minimizing its potential to cause cellular damage. Peroxiredoxins (Prxs) have peroxidase activity toward H2O2, organic hydroperoxides, and peroxynitrite for protecting cells from oxidative stress. They are suggested to work as signaling mediators, allowing the local accumulation of H2O2 by inactivating their peroxidase activity uniquely compared with other antioxidant proteins such as catalase and glutathione peroxidase. Given that Prxs are highly sensitive to oxidation by H2O2, they act as sensors and transducers of H2O2 signaling via transferring their oxidation state to effector proteins. The concentrations of intracellular H2O2 increase as the cell cycle progresses from G1 to mitosis. Here, we summarize the roles of Prxs with regard to the regulation of cell cycle-dependent kinase activity and anaphase-promoting complex/cyclosome in terms of changes in H2O2 levels. Protection of the cell from unwanted progression of the cell cycle is suggested to be a role of Prx. We discuss the possible roles of Prxs to control H2O2 levels.


Introduction
The role of H 2 O 2 as a signaling molecule that regulates various biological processes, such as cell proliferation, differentiation, and migration, is well established [1,2]. Relationships between the change in H 2 O 2 concentrations and cell cycle progression are less well understood because of limited information for specific effectors of H 2 O 2 at different stages throughout the cell cycle. Constitutively elevated levels of H 2 O 2 caused by Nox1 overexpression can transform cells and can make them tumorigenic [3], indicating that H 2 O 2 is a promoter of cell proliferation. Persistent production of H 2 O 2 by treatment with extracellular reactive nitrogen species via possible activation of epidermal growth factor receptors [4] and subsequently activated nicotinamide adenine dinucleotide phosphate (NADPH) oxidases [5,6] or with glucose oxidase in media containing high concentrations of glucose induces cell cycle arrest at the boundary between the G 0 and G 1 phase [7,8]. Reactive oxygen species (ROS) are required for G 1 -S phase transition during cell cycle progression [9]. A study [10] in human melanoma cell lines reported the downregulation of ROS levels with treatment of diphenyleneiodonium or vitamin E and reported that Nox4 lockdown using siRNA oligos induces cell cycle arrest at the G 2 -M stage. However, the molecular mechanism by which ROS controls transition of specific stages during cell cycle progression remains to be elucidated.
Peroxiredoxins (Prxs) are abundant intracellular H 2 O 2 -removing enzymes that have the conserved peroxidatic cysteine (C P ) as the site of oxidation by H 2 O 2 , forming cysteine sulfenic acid (C P -SOH) and

Dynamic Change in Intracellular Morphology and Oscillation of Activities of Cdks and APC/C through the Cell Cycle
Regulation of the cell cycle has been attributed largely to oscillation of Cdk activities. In eukaryotes, about 20 different Cdk family proteins and four major cyclins (cyclin D, cyclin E, cyclin A, and cyclin B) exist. Cdk activities are regulated by the phosphorylation state of Cdks; by the concentrations of Cdk inhibitors, such as the Ink4 family (p16, p15, p18, and p19) and the Cip/Kip family (p21, p27, and p57) [22,23]; and by the amount of cyclins, which is increased in a transcription-dependent manner and is decreased by ubiquitin-proteasome pathway through the APC/C and the skp-cullin-F-box complex [24]. APC/C degrades two important cyclins (cyclin A, a cofactor for Cdk2 and Cdk1; cyclin B, a cofactor of Cdk1) at early mitosis and at late mitosis, respectively, to ensure an irreversible and directional cell cycle progression. Broadly, APC/C activity is inversely proportional to the activities of Cdk1 and Cdk2 throughout the cell cycle, except at the G 2 phase (Figure 1) [18,19]. At the G 2 stage, early mitotic inhibitor 1 inhibits APC/C through its binding with Cdc20 and Cdh1, two activators of APC/C [25,26]. Detection with 2',7'-dichlorodihydrofluorescein, which fluoresces on oxidation, showed that levels of ROS increase at G 1 -S phase transition, peak at mitosis, and decrease during mitotic exit to complete the redox cycle [9,16]. However, this probe shows several limitations, such as relatively low selectivity to react with various oxidants including peroxynitrite, nitric oxide, and hypochloride in addition to H 2 O 2 , and susceptibility to photooxidation and photobleaching [27][28][29]. At late G 1 -S phase transition, APC/C-Cdh1 activity is inhibited by increased endogenous ROS, but a direct target of ROS has not been investigated [9]. APC11, a subunit of APC/C, is sensitive to oxidation by H 2 O 2 [21] but is not known to be an effector of endogenous H 2 O 2 . The morphological change in subcellular compartments and the activities of cell cycle-dependent kinases (Cdks) and anaphase-promoting complex/cyclosome (APC/C) throughout the cell cycle in a mammalian cell: During the cell cycle, after DNA replication is finished at the S phase, a cell prepares to enter mitosis through the G2 phase. At the G2-M transition, the structure of subcellular organelles is changed dramatically as illustrated here. A nuclear envelop is broken and the Golgi (green), ER (yellow), and mitochondria (pink) are fragmented. After mitotic exit, all the organelles are reorganized and structured in filamentous states. Activity of APC/C is inversely propositioned to that of Cdks during cell cycle progression except at the G2 phase. APC/C activity is inhibited by early mitotic inhibitor1 at the G2 phase. In G0 (quiescent phase) cells, APC/C-Cdh1 remains active.
The structure of intracellular organelles is changed dramatically at G2-M phase transition ( Figure 1). The mammalian Golgi apparatus is structured in stacks of flattened membrane cisternae and plays a central role in protein and vesicular trafficking from the ER to the plasma membrane, the endosomes, or secretion outside of the cell [30]. Golgi fragmentation is required for mitotic entry [31] and induced through phosphorylation by polo-like kinase 1 (Plk1) [32], Mek1 [33], and subsequently Cdk1 [34]. In cells activated with growth factors, Golgi resident Duox1 and Duox2 proteins are responsible for production of H2O2 at the Golgi [35]. Activation of Duox1 and Duox2 proteins requires elevation of intracellular Ca 2+ concentration. Considering that 1) a transient increase in Ca 2+ concentration at mitotic spindle poles is observed [36][37][38] and 2) fragmented Golgi containing Duox1 and Duox2 are absorbed into the ER, a major reserve organelle for Ca 2+ , H2O2 molecules are produced by Ca 2+ -dependent activated Duoxs and can be used as signaling mediators during mitosis. Mitochondria are implicated in a variety of events, including energy metabolism through the production of ATP and the supply of metabolites, apoptosis by the release of cytochrome C and anti-inhibitor of apoptosis (anti-IAP), calcium homeostasis, and the production of ROS [39]. The morphology of mitochondria is changed dynamically throughout the cell cycle by fusion and fission. Mitochondria fuse to form a hyperactive long tubular network at the G1 phase and become the fragmented (granular) short-length state that facilitates their segregation at mitotic stage [40,41]. Mitochondria are a major cellular source of H2O2, which results from incomplete reduction of O2 during the steps of the electron-transfer chain [42,43]. Increased ROS from damaged mitochondria Figure 1. The morphological change in subcellular compartments and the activities of cell cycle-dependent kinases (Cdks) and anaphase-promoting complex/cyclosome (APC/C) throughout the cell cycle in a mammalian cell: During the cell cycle, after DNA replication is finished at the S phase, a cell prepares to enter mitosis through the G 2 phase. At the G 2 -M transition, the structure of subcellular organelles is changed dramatically as illustrated here. A nuclear envelop is broken and the Golgi (green), ER (yellow), and mitochondria (pink) are fragmented. After mitotic exit, all the organelles are reorganized and structured in filamentous states. Activity of APC/C is inversely propositioned to that of Cdks during cell cycle progression except at the G 2 phase. APC/C activity is inhibited by early mitotic inhibitor1 at the G 2 phase. In G 0 (quiescent phase) cells, APC/C-Cdh1 remains active.
The structure of intracellular organelles is changed dramatically at G 2 -M phase transition ( Figure 1). The mammalian Golgi apparatus is structured in stacks of flattened membrane cisternae and plays a central role in protein and vesicular trafficking from the ER to the plasma membrane, the endosomes, or secretion outside of the cell [30]. Golgi fragmentation is required for mitotic entry [31] and induced through phosphorylation by polo-like kinase 1 (Plk1) [32], Mek1 [33], and subsequently Cdk1 [34]. In cells activated with growth factors, Golgi resident Duox1 and Duox2 proteins are responsible for production of H 2 O 2 at the Golgi [35]. Activation of Duox1 and Duox2 proteins requires elevation of intracellular Ca 2+ concentration. Considering that 1) a transient increase in Ca 2+ concentration at mitotic spindle poles is observed [36][37][38] and 2) fragmented Golgi containing Duox1 and Duox2 are absorbed into the ER, a major reserve organelle for Ca 2+ , H 2 O 2 molecules are produced by Ca 2+ -dependent activated Duoxs and can be used as signaling mediators during mitosis. Mitochondria are implicated in a variety of events, including energy metabolism through the production of ATP and the supply of metabolites, apoptosis by the release of cytochrome C and anti-inhibitor of apoptosis (anti-IAP), calcium homeostasis, and the production of ROS [39]. The morphology of mitochondria is changed dynamically throughout the cell cycle by fusion and fission. Mitochondria fuse to form a hyperactive long tubular network at the G 1 phase and become the fragmented (granular) short-length state that facilitates their segregation at mitotic stage [40,41]. Mitochondria are a major cellular source of H 2 O 2 , which results from incomplete reduction of O 2 during the steps of the electron-transfer chain [42,43]. Increased ROS from damaged mitochondria enforce a G 1 -S cell cycle checkpoint as signaling mediators by activating mitogen-activated protein kinase signaling and thereby increasing an inhibitor of Cdk2-cyclin E in Drosophila melanogaster [44]. A hyperfused mitochondrial state is linked to the G 1 -S transition with proper accumulation of cyclin E [40], indicating that a signal from mitochondria controls cell cycle progression. The dynamin-related protein 1, a promoter of mitochondrial fission, is activated by cyclin B-Cdk1 complexes, thereby leading to mitochondrial fragmentation at the G 2 -M transition [45]. The dynamin-related protein 1 is degraded by APC/C-Cdh1 during mitotic exit [46]. The Cdks and APC/C modules regulate structural dynamics of mitochondria, Golgi, and ER. Retrograde signals from the organelles can regulate the activities of Cdks and APC/C. H 2 O 2 molecules are suggested to be signaling mediators in the mutual interaction between cell cycle controllers and change in organelle structure.

Localization of Peroxiredoxin Proteins Inside and Outside of the Cell
Peroxiredoxins are a major class of antioxidant enzymes that reduce H 2 O 2 with the use of electrons derived from NADPH [47]. Mammalian cells express six Prx isoforms (PrxI-PrxVI), which are usually present at high intracellular concentrations and have been implicated in a variety of cellular processes, including cell proliferation [47]. The structure of subcellular organelles is changed dynamically throughout the cell cycle, and therefore, the localizations of Prxs are reorganized depending on cell cycle phases.
Mammalian Prx isoforms localize to specific cellular compartments, including the cytosol, and to organelles such as mitochondria, peroxisomes, lysosomes, and the nucleus, as well as localizing outside the cell [13,[47][48][49] (Figure 2, Table 1). PrxI was localized in moderate amounts in the nucleus, cytosol, and plasma membrane [50,51]. In addition, a recent study described local regulation of H 2 O 2 around the centrosome through PrxI phosphorylation by Cdk1 during early mitosis [17]. PrxII was observed primarily in the cytosol, with distribution similar to PrxI, although some was found in the nucleus and involved in protection of cancer cell death from DNA damage [52]. PrxIII was found almost exclusively in the matrix of mitochondria, which is mainly responsible for reversible regulation of mitochondrial H 2 O 2 levels [53,54]. While PrxIV was found in both cytosol and extracellular fluid, it was the only isoform found in the ER [55,56]. Protein disulfide isomerase was oxidized by ER-localized PrxIV in the oxidative protein-folding pathway [57][58][59]. PrxV was found predominantly in the mitochondria, with some in the peroxisome, and low levels in cytosol and nucleus [60][61][62]. PrxV antioxidant enzyme protects against oxidant-generating peroxisome, containing several molecules that catalyze the oxidation of substrates and thereby produce H 2 O 2 [63]. PrxVI was located predominantly in the cytosol, but it was also expressed in lysosomal compartments with PrxIV [49,64]. Phosphorylation of PrxVI by the MAP kinase increases its PLA 2 (phospholipase A 2 ) activity [65][66][67].

Oscillation of the Intracellular Concentration of H2O2 during the Cell Cycle
ROS levels were shown previously to change during cell cycle progression when measured with the fluorescent probe chloromethyl derivative of 2′,7′-dichlorodihydrofluorescein diacetate [9,15]. This probe reacts with various oxidants including H2O2, the hydroxyl radical, NO, lipid peroxides, and peroxynitrite [9]. Furthermore, given that the oxidation reaction of the probe is irreversible, the change in fluorescence represents the sum of the oxidant molecules produced after

Oscillation of the Intracellular Concentration of H 2 O 2 during the Cell Cycle
ROS levels were shown previously to change during cell cycle progression when measured with the fluorescent probe chloromethyl derivative of 2 ,7 -dichlorodihydrofluorescein diacetate [9,15]. This probe reacts with various oxidants including H 2 O 2 , the hydroxyl radical, NO, lipid peroxides, and peroxynitrite [9]. Furthermore, given that the oxidation reaction of the probe is irreversible, the change in fluorescence represents the sum of the oxidant molecules produced after introduction of the probe-not transient changes in oxidant concentration. We measured intracellular H 2 O 2 levels in HeLa cells at various cell cycle stages with the use of a genetically encoded fluorescent probe, HyPer (Evrogen, Russia), that reacts reversibly and specifically with H 2 O 2 [68]. The relative fluorescence level in nocodazole-arrested prometaphase cells (~98% in the G 2 -M phase) was about three times that in asynchronous cells (~73% in G 1 ,~14% in S, and~13% in G 2 -M), whereas the level in G 1 -enriched cells (~52% in G 1 ,~1% in S, and~47% in G 2 -M) was similar to that in asynchronous cells ( Figure 3A,B). The fluorescence of HyPer is sensitive to changes in pH, and a HyPer mutant with Cys 199 replaced by Ser (HyPer-C199S) is sensitive to pH but not to H 2 O 2 . The fluorescence of HyPer-C199S in HeLa cells was not affected by nocodazole arrest ( Figure 3C). Release of cells from G 1 -S arrest also revealed that HyPer fluorescence intensity increased as cells progressed through S phase, reached a maximum when the G 2 -M population was largest, and then decreased as cells transited into G 1 phase ( Figure 3D-F). Linear regression analysis of the fluorescence intensities measured at six different time points after the release of cells from G 1 -S arrest yielded a ratio of 1:1.7:2.9 for the relative H 2 O 2 levels in G 1 , S, and G 2 -M phases, respectively (adjusted R 2 value = 0.982, p = 0.00145). These results indicate that endogenous H 2 O 2 levels oscillate in a cell cycle-dependent manner. Oscillations in thiols and their reductive capacity could lead to increased HyPer oxidation.
decreased as cells transited into G1 phase ( Figure 3D-F). Linear regression analysis of the fluorescence intensities measured at six different time points after the release of cells from G1-S arrest yielded a ratio of 1:1.7:2.9 for the relative H2O2 levels in G1, S, and G2-M phases, respectively (adjusted R 2 value = 0.982, p = 0.00145). These results indicate that endogenous H2O2 levels oscillate in a cell cycle-dependent manner. Oscillations in thiols and their reductive capacity could lead to increased HyPer oxidation.

Source for H 2 O 2 Generation throughout the Cell Cycle
Energy metabolism, which is accompanied by H 2 O 2 production, increases during cell proliferation; H 2 O 2 derived from NADPH oxidase has been shown to be required for G 2 -M progression [10]; cytosolic phospholipase A 2 activity, which generates arachidonic acid, is high in mitotic phase and decreases in G 1 phase [69]; and arachidonic acid-metabolizing enzymes such as cyclooxygenase, lipoxygenase, and cytochrome P450 generate H 2 O 2 [70]. There are thus multiple potential sources of H 2 O 2 production during cell cycle progression. The general flavoprotein inhibitor diphenyleneiodonium, the cytosolic phospholipase A 2 inhibitors arachidonyl trifluoromethyl ketone and methyl arachidonyl fluorophosphonate, and the lipoxygenase inhibitor nordihydroguaiaretic acid each inhibited entry into mitosis in HeLa cells [17]. This result suggests that NADPH oxidases and the cytosolic phospholipase A 2 -lipoxygenase system are responsible for the production of H 2 O 2 at G 2 -M transition.
Specific Cdks control structural change in cellular organelles such as the Golgi, ER, and mitochondria and thereby allow segregation between daughter cells of each organelle during mitosis [34,45]. Golgi H 2 O 2 is shown to be produced via Ca 2+ -dependent Duox in epidermal growth factor-stimulated cells [35]. Some portion of fragmented Golgi is redistributed into the ER containing a high concentration of Ca 2+ [71] and the production of H 2 O 2 via Duox is predicted at mitotic phase. Mild or severe oxidative stress induces mitochondrial fission and fragmentation [72]. Fragmented mitochondria can be degraded selectively by mitophagy to protect the cell from apoptosis. Granular (fragmented) mitochondria are linked to high levels in ROS, and tubular (filamentous) mitochondria are inversely proportional to ROS level in neurodegeneration [73]. During mitosis, the question of whether fragmented mitochondria produce H 2 O 2 more than filamentous mitochondria should be addressed.

Protective and Signaling Role of Peroxiredoxin around the Centrosomes
Local inactivation of PrxI is important for the transduction of H 2 O 2 signal in addition to local generation of H 2 O 2 by an activated NADPH oxidase during extracellular stimulation. PrxI is phosphorylated on Tyr 194 by a Src family kinase, and therefore, the inhibition of its peroxidase activity allows a transient increase in H 2 O 2 level in the plasma membrane microdomain in cells stimulated with growth factors such as platelet-derived growth factor and epidermal growth factor [50].
Centrosomes are microtubule-organizing centers that play key roles in spindle formation, chromosome segregation, cell cycle progression, and cell division throughout the cell cycle [74]. In a recent study, we proposed a role for PrxI associated with centrosomes for mitotic entry [17] ( Figure 4). Many proteins that participate in the mitotic entry network (Cdk1, cyclin B, Cdc25B, Cdc25C, Plk1, and Aurora A) or in the mitotic exit network (APC/C, Cdc14B, and Cdh1) are concentrated at the centrosome [19,[75][76][77][78][79][80]. Several feedback loops that underlie activation of Cdk1 and amplification of Cdk1 activity are initiated by cyclin B, Cdc25B, Cdc25C, Plk1, and Aurora A at the centrosome. Mitotic exit processes related to the activation of APC/C-Cdh1 by Cdc14B, which results in the degradation of mitotic activators (cyclin B, Plk1, and Aurora A), also have been suggested to be integrated at the centrosome [81].
The activity of Cdc14B must be controlled tightly to avoid premature degradation of mitotic activators before full activation of Cdk1. Our results indicate that the activity of centrosomal Cdc14B is controlled through regulation of the local H 2 O 2 concentration around the centrosome, which in turn depends on the H 2 O 2 -eliminating enzyme PrxI. PrxI is present at the centrosome, and the centrosome-associated PrxI, but not cytosolic PrxI, is specifically phosphorylated on Thr 90 , a consensus site for phosphorylation by Cdk1 [17,82]. Phospho-PrxI was detected at the centrosome of HeLa cells during early mitosis (prophase, prometaphase, and metaphase) but not during interphase (G 2 ) or late mitotic stages (anaphase, telophase, and cytokinesis). Phosphorylation of PrxI inactivates its peroxidase function, resulting in exposure of centrosomal Cdc14B to H 2 O 2 -dependent inactivation. This inactivation of Cdc14B likely blocks untimely dephosphorylation of Cdh1 during early mitosis and thereby prevents premature degradation of the Cdk1 activators cyclin B, Plk1, and Aurora A by APC/C-Cdh1. Whether centrosomal proteins are oxidized is difficult to investigate because the biochemical procedure to isolate the centrosome includes treatment with a thiol-reducing agent, such as 2-mercaptoethanol [80], which reduces the disulfide bond of oxidized proteins. Instead, the specific role of pericentrosomal H 2 O 2 was demonstrated by targeted expression of catalase at the centrosome, which resulted in downregulation of the centrosomal abundance of cyclin B1, Plk1, and Aurora A bỹ 30%-55% and delayed mitotic entry. Targeted expression of catalase also reduced the cellular level of Ser 40 -phosphorylated Cdh1 by~30%. The Ser 40 -phosphorylated form of Cdh1 was not detected reliably in isolated centrosomes, probably because phosphorylation reduces the affinity of Cdh1 for the organelle to a level insufficient to support the interaction during biochemical isolation. These Antioxidants 2020, 9, 280 8 of 16 results, together with the high sensitivity of Cdc14B to H 2 O 2 , thus support the notion that changes in pericentrosomal H 2 O 2 level associated with cell cycle progression markedly influence the stability of cyclin B1, Plk1, and Aurora A through the Cdc14B and Cdh1-APC/C axes. Centrosome-associated PrxI proteins shield the organelles from a high tide of H2O2 molecules from various intracellular sources at G2 phase. Anaphase-promoting complex/cyclosome (APC/C)-Cdh1 activity is high at the centrosome, and therefore, its substrates such as cyclin B, polo-like kinase1 (Plk1), and Aurora A are degraded. During mitosis, PrxI is inactivated by cell cycle-dependent kinase 1 (Cdk1) through phosphorylation and the large numbers of H2O2 molecules inhibit APC/C-Cdh1, which results in increased cyclin B, Plk1, and Aurora A proteins. Centrosome-associated PrxI can protect cells from other oxidative stress, including ultraviolet (UV) or ionizing radiation (IR).
The activity of Cdc14B must be controlled tightly to avoid premature degradation of mitotic activators before full activation of Cdk1. Our results indicate that the activity of centrosomal Cdc14B is controlled through regulation of the local H2O2 concentration around the centrosome, which in turn depends on the H2O2-eliminating enzyme PrxI. PrxI is present at the centrosome, and the centrosome-associated PrxI, but not cytosolic PrxI, is specifically phosphorylated on Thr 90 , a consensus site for phosphorylation by Cdk1 [17,82]. Phospho-PrxI was detected at the centrosome of HeLa cells during early mitosis (prophase, prometaphase, and metaphase) but not during interphase (G2) or late mitotic stages (anaphase, telophase, and cytokinesis). Phosphorylation of PrxI inactivates its peroxidase function, resulting in exposure of centrosomal Cdc14B to H2O2-dependent inactivation. This inactivation of Cdc14B likely blocks untimely dephosphorylation of Cdh1 during early mitosis and thereby prevents premature degradation of the Cdk1 activators cyclin B, Plk1, and Aurora A by APC/C-Cdh1. Whether centrosomal proteins are oxidized is difficult to investigate because the biochemical procedure to isolate the centrosome includes treatment with a thiol-reducing agent, such as 2-mercaptoethanol [80], which reduces the disulfide bond of oxidized proteins. Instead, the specific role of pericentrosomal H2O2 was demonstrated by targeted expression of catalase at the centrosome, which resulted in downregulation of the centrosomal abundance of cyclin B1, Plk1, and Aurora A by ~30%-55% and delayed mitotic entry. Targeted expression of catalase also reduced the cellular level of Ser 40 -phosphorylated Cdh1 by ~30%. The Ser 40 -phosphorylated form of Cdh1 was not detected reliably in isolated centrosomes, probably molecules from various intracellular sources at G 2 phase. Anaphase-promoting complex/cyclosome (APC/C)-Cdh1 activity is high at the centrosome, and therefore, its substrates such as cyclin B, polo-like kinase1 (Plk1), and Aurora A are degraded. During mitosis, PrxI is inactivated by cell cycle-dependent kinase 1 (Cdk1) through phosphorylation and the large numbers of H 2 O 2 molecules inhibit APC/C-Cdh1, which results in increased cyclin B, Plk1, and Aurora A proteins. Centrosome-associated PrxI can protect cells from other oxidative stress, including ultraviolet (UV) or ionizing radiation (IR).
Our proposal for sequential events determined by inactivation of PrxI via Cdk1 at the centrosome is shown in a schematic drawing ( Figure 5). Intracellular H 2 O 2 levels peak when the G 2 -M population is largest (see Figure 3D,E). In the G 2 phase, PrxI associated with the centrosome shields the organelles from the high tide of H 2 O 2 . As the abundance of cyclin B peaks in late G 2 phase and early mitosis, cyclin B binds to and activates Cdk1. The activity of Cdk1-cyclin B is then further increased through the operation of multiple feedback loops mediated by cyclin B, Cdc25, Plk1, and Aurora A. At the same time, Cdk1-cyclin B phosphorylates (inactivates) PrxI, which results in an increase in H 2 O 2 level in the centrosome and consequent inactivation of Cdc14B. In the absence of the H 2 O 2 -mediated inactivation of Cdc14B, which is highly enriched at the centrosome, Cdc14B is expected to dephosphorylate Cdk1-phosphorylated Cdh1 and to thereby activate APC/C-Cdh1. These events in turn are expected to result in degradation of the mitotic activators (cyclin B, Cdc25, Plk1, and Aurora A) and to retard mitotic entry. Given that Cdc25 is not oxidized by H 2 O 2 , the Cdk1 activation loop mediated by Cdc25B and Cdc25C is not affected by PrxI inactivation. In late mitosis, PrxI is dephosphorylated and H 2 O 2 is removed by active PrxI at the centrosome, which induces sequential reactivation of Cdc14B and activation of APC/C-Cdh1 by means of dephosphorylation of Cdh1. Plk1, and Aurora A) and to retard mitotic entry. Given that Cdc25 is not oxidized by H2O2, the Cdk1 activation loop mediated by Cdc25B and Cdc25C is not affected by PrxI inactivation. In late mitosis, PrxI is dephosphorylated and H2O2 is removed by active PrxI at the centrosome, which induces sequential reactivation of Cdc14B and activation of APC/C-Cdh1 by means of dephosphorylation of Cdh1. Protein phosphatase 1 (PP1) or PP2A (or both) is likely responsible for the dephosphorylation of PrxI during late mitosis, given that we observed that phosphorylated PrxI accumulates in HeLa cells Protein phosphatase 1 (PP1) or PP2A (or both) is likely responsible for the dephosphorylation of PrxI during late mitosis, given that we observed that phosphorylated PrxI accumulates in HeLa cells treated with okadaic acid, a specific inhibitor of these phosphatases [82]. This conclusion is consistent with previous observations that PP1 and PP2A are key mitotic exit phosphatases that counteract phosphorylation by Cdk1 [83]; that PP1 and PP2A, together with their regulatory proteins, are concentrated at the centrosome [84][85][86][87]; and that okadaic acid promotes mitotic entry through inhibition of these phosphatases [88].
While APC/C-Cdh1 remains active in G 1 and G 0 (quiescent phase) cells, Cdks must be inactive. Here, APC/C-Cdh1 blocks the premature accumulation of certain positive regulators of S-phase and mitotic progression. Since H 2 O 2 can trigger S-phase entry and mitotic signaling by inactivating Cdc14B and APC/C-Cdh1 at the centrosome, centrosome-associated PrxI can prevent unscheduled mitotic entry as a result of any accidental increase in H 2 O 2 (such as in response to ultraviolet or ionizing radiation) (Figure 4). This corresponds with previous reports that Cdc14B, Cdh1, and PrxI function as tumor suppressors [80,89,90].
In brief summary (Figures 4 and 5), Cdc14B is coupled to and inactivated by a high tide of cytosolic H 2 O 2 through Cdk1-dependent phosphorylation (inactivation) of PrxI at the centrosome during the early stage of mitosis, resulting in the formation of a previously unknown feedback loop in which Cdk1 activates itself by blocking the degradation of its positive regulators. During mitotic exit, however, Cdc14B is decoupled from cytosolic H 2 O 2 as a result of PrxI dephosphorylation and is reactivated to promote the degradation of the mitotic activators. The new finding reveals that H2O2 molecules from diverse cellular processes, including mitochondrial respiration, arachidonic acid metabolism, and activated NADPH oxidase in fragmented organelles (the Golgi or ER), control mitosis.
In the model (Figures 4 and 5), given that the peroxidase activity of phosphorylated PrxI by Cdk1 is about 20% of that of unphosphorylated PrxI [82], a local increase in H 2 O 2 levels at the centrosomes is reasonably expected to oxidize redox-sensitive proteins represented by Cdc14B. PrxI acts as an H 2 O 2 floodgate, preventing susceptible targets from oxidation by H 2 O 2 until the floodgate is opened [91]. However, if the residual activity of p-PrxI is enough to remove H 2 O 2 at the centrosomes or if other cytosolic Prxs replace inactivated PrxI, then Prxs-dependent redox relay for mitotic progression can be considered.

Possible Role of Peroxiredoxin around the Golgi, ER, and Mitochondria
The intracellular concentration of H 2 O 2 oscillates during cell cycle progression, and the sources of this H 2 O 2 include NADPH oxidase and arachidonic acid-metabolizing enzymes [17]. The best characterized H 2 O 2 effectors are members of the protein tyrosine phosphatase (PTP) family such as PTP1B and phosphatase and tensin homologue (PTEN) [2,92,93]. Indeed, in cells stimulated with various growth factors, the activation of protein tyrosine kinases or phosphatidylinositol 3-kinase is not sufficient to increase the steady-state level of tyrosine-phosphorylated proteins or 3-phosphorylated inositides; the concurrent inhibition of PTPs or of PTEN, respectively, is also required. Given that many kinases (such as Src family members) and transcription factors (such as p53 and AP1) are also direct targets of H 2 O 2 [2,92,94], oscillation of the intracellular H 2 O 2 concentration is expected to affect cell cycle progression at multiple stages.
Considering a dynamic change in cell morphology, including breakdown of the nuclear envelope and fragmentation of subcellular organelles such as the Golgi, ER, and mitochondria in mitosis, Prx isoforms are redistributed in the cell throughout the cell cycle. PrxI associated with the ER membrane functions in attenuating oxidative stress in the livers of ethanol-fed mice [95]. The cell at the G 2 -M transition and in mitotic phase requires an abundance of energy to drive dynamic change in cell morphology under oxidative stress (with high H 2 O 2 levels). To reduce the requirement of energy, mitotic cells likely perform H 2 O 2 -mediated signaling more efficiently than interphase cells, which needs to be verified in further investigations. Cytosolic Prxs, subcellular organelle-associated Prxs, and Prxs in the lumen of organelles can play key roles in controlling H 2 O 2 -mediated signaling, as in H 2 O 2 accumulation by the phosphorylation (inactivation) on Tyr 194 of PrxI at the plasma membrane [50] and by the phosphorylation (inactivation) on Thr 90 of PrxI at the centrosomes [17].

Conclusions
Abundant Prx proteins react with H 2 O 2 molecules and thereby act as sensors and transducers of H 2 O 2 -mediated signaling. The localization of Prxs is changed dynamically depending on structural reorganization of subcellular organelles throughout the cell cycle. The level of ROS, especially H 2 O 2 molecules, fluctuates throughout the cell cycle, is lowest in G 1 phase, and peaks in mitosis as shown in this review or other previous studies [9,15]. Given that Prxs protect cells from oxidative stress by removing H 2 O 2 molecules, Prxs ensure proper cell cycle progression in a time-dependent manner by inhibiting premature activation of Cdk1, as proposed for PrxI at the centrosome [17]. Prxs protect the cell from premature cell cycle progression under unwanted oxidative stress arising by UV or ionizing irradiation in interphase. Downregulation and overexpression of PrxII or PrxIII elicits a change in the distribution of cell cycle stages [8,96,97], indicating Prxs are critical for proper transition of each stage. Modulation of the expression of Prx isoforms has limitations for investigation of its role as the regulator of cell cycle. Identification of H 2 O 2 effectors in mitotic cells with high levels of H 2 O 2 and oxidatively damaged cells that are derived from cardiovascular diseases and neurodegenerative disorders sheds light on understanding the roles of Prxs during cell cycle progression.
Proper and ordered cell progression requires the accurate regulation of Cdks and APC/C activities. Among the regulators of Cdks-APC/C module, Cdc14B is known to be an H 2 O 2 effector [17]. The Ser/Thr phosphatases, PP1 and PP2, which play a critical role in mitotic entry and exit in mammalian cells [88,[98][99][100][101][102], are good candidates for H 2 O 2 effectors because these proteins have a redox-sensitive catalytic center [103][104][105]. The phosphorylation (inactivation) of PrxI by Cdk1 at the centrosomes induces APC/C inactivation through releasing an activator, Cdh1 ( Figure 5) [17]. Another example of the role of Prx in the regulation of Cdks-APC/C module is a phosphorylation (inactivation) on Thr 89 of PrxII by Cdk5 in Parkinson's disease (PD) [106]. Oxidative stress is a critical factor to induce cellular damage in PD [107]. Increased H 2 O 2 by Cdk5-mediated PrxII inactivation in PD occurs with APC/C-Cdh1 inhibition by Cdk5 [108]. It is suggested that the Cdk-Prx-APC/C axis is conserved in different cellular events.