Fundamentals of redox regulation in biology

Abstract

Oxidation-reduction (redox) reactions are central to the existence of life. Reactive species of oxygen, nitrogen and sulfur mediate redox control of a wide range of essential cellular processes. Yet, excessive levels of oxidants are associated with ageing and many diseases, including cardiological and neurodegenerative diseases, and cancer. Hence, maintaining the fine-tuned steady-state balance of reactive species production and removal is essential. Here, we discuss new insights into the dynamic maintenance of redox homeostasis (that is, redox homeodynamics) and the principles underlying biological redox organization, termed the 'redox code'. We survey how redox changes result in stress responses by hormesis mechanisms, and how the lifelong cumulative exposure to environmental agents, termed the 'exposome', is communicated to cells through redox signals. Better understanding of the molecular and cellular basis of redox biology will guide novel redox medicine approaches aimed at preventing and treating diseases associated with disturbed redox regulation.

Fig. 1 |. Principles of redox regulation.

The process of redox regulation begins with initial causal input cues (for example, endogenous cues such as hormonal stimuli or exogenous input from nutrients). Endogenous and exogenous redox signal generation causes the production of reactive species of oxygen (ROS), nitrogen (RNS) and sulfur (RSS). These redox signals can be propagated by superoxide (O₂^•−), hydrogen peroxide (H₂O₂), nitric oxide (•NO) and hydrogen sulfide (H₂S), which are produced by NADPH oxidase (NOX), the electron transport chain (ETC) and other FAD-dependent mitochondrial dehydrogenases, nitric oxide synthase (NOS) and cystathionine γ-lyase (CSE), respectively. Redox signals can also be mediated through electrophiles, lipid hydroperoxides and nitroalkenes. Transduction of the redox signal leads to the modulation of the redox proteome and metabolome. The redox signal is propagated through the site-specific oxidation and reduction of cysteine switches, also called redox switches. As discussed in detail in Box 2, the redox switches are subject to reversible oxidation and reduction, either through S-sulfenylation, S-glutathionylation, S-nitrosylation or other redox-sensitive modifications, which transmit the redox signal. Redox signal propagation is also mediated through the redoxin family of enzymes (thioredoxins, glutaredoxins and peroxiredoxins), which are activated and/or deactivated by oxidation and reduction of catalytic thiols. The stress-sensing transcription factors such as nuclear factor erythroid 2-related factor (NRF2), nuclear factor-κB (NF-κB), hypoxia-inducible factor 1 (HIF-1) or heat shock factor 1 (HSF-1) are also redox sensors and are activated or deactivated by cysteine thiol oxidation and reduction, which affect their ability to activate the expression of genes required to invoke cell stress responses. Cumulatively, the presence of redox signals affects the operation of cellular networks, from genome and proteome to other omes, leading to the integration of the redox signal. An additional mode of redox regulation is provided by the multifaceted interaction of various types of regulatory non-coding RNAs (such as microRNAs (miRNAs), long non-coding RNAs (lncRNAs) or circular RNAs (circRNAs)). The hormetic outcome of activation of various stress responses such as the oxidative distress response (ODR), unfolded protein response (UPR), heat shock response (HSR), integrated stress response (ISR) and the DNA damage response (DDR) results in the decision among adaptation, maintenance, repair and health or cell destruction, removal or death. Continuously operating feedback loops serve to maintain redox stability by iterative coordinated fluctuations and oscillations, including gene expression feedback loops or transcription–translation feedback loops, which then reinforce or desensitize the redox signal.

Fig. 2 |. Redox eustress and transition to oxidative and reductive distress.

The spectrum of oxidant exposures ranges from minimal (reductive distress) to physiological (redox eustress) to very high levels (oxidative distress). The hormetic range includes transition phases from physiological (blue) to non-physiological (red) exposure levels, that is, the hormetic dose–response, which results in the activation of redox-sensing transcription factors that promote cell adaptation. The borderline thresholds from physiological to subphysiological or supraphysiological levels are denoted as redox stress signalling thresholds (RSTs). The hypoxia-inducible factor 1 (HIF-1) is activated predominantly when oxidant levels are low (towards the subphysiological side), whereas the nuclear factor-κB (NF-kB) comes into action towards the supraphysiological side, and the nuclear factor erythroid 2-related factor (NRF2) predominantly covers the physiological range. The bottom of the figure lists some of the major causes of redox eustress and of the two forms of redox distress with the resulting effects mentioned subsequently as detailed in the main text.

Fig. 3 |. Schematic overview of redox communication in and between cells.

This simplified scheme focuses on H₂O₂ as a major cellular signalling oxidant. Plasma membrane-located NADPH oxidase (NOX) generates O₂^•− extracellularly, which is then dismutated to H₂O₂ by superoxide dismutase (SOD3). Entry of H₂O₂ into cells occurs either by transfer through peroxiporins, which are specialized aquaporins (AQPs), or by endocytosis to form redox-active endosomes, which have been named ‘redoxosomes’, formed predominantly at caveolae. Intracellularly, the organelles such as mitochondria, peroxisomes and the endoplasmic reticulum (ER) as well as around 40 oxidases produce H₂O₂ for signalling (see ref. 9). The three organelles are in close contact, facilitated by tethering of their membranes at membrane contact sites (MCSs). Intercellular redox communication occurs by gap junctional communication through connexons, formed from two connexin hemichannels between directly adjacent cells, or by AQPs or endocytosis of extracellular vesicles (EVs), the redoxosomes mentioned earlier. The desmosomes and tight junctions serve to connect cells to one another, bringing the adjacent cell receiving the redox signal near the cell producing the H₂O₂ messenger. The protein desmin has a single cysteine (Cys333), which upon oxidation causes desmosome disruption. For distant cells, H₂O₂ signalling can occur by endocytosis of redoxosomes or through AQPs. It should be noted that other modes of redox communication exist, for example, the release of redox-active enzymes for paracrine signalling between cells, and the release of nitric oxide from nitric oxide synthase. The full potential of the redoxosome and other EVs in cell-to-cell redox signalling needs to be explored. EGFR, epidermal growth factor receptor.

Fig. 4 |. Major cellular processes under redox regulation.

a, Metabolism: as a prototypical example for redox control of metabolic pathways, glyceraldehyde 3-phosphate dehydrogenase (GAPDH), a central enzyme of glycolysis, contains a redox-sensitive cysteine at the active site. Its oxidative modification causes loss of the dehydrogenase function (GAPDH^ox) and blocks glycolysis. Glucose is shunted to the pentose phosphate pathway to form NADPH, which is needed to restore the reduced glutathione (GSH) pool. GAPDH^ox can also shuttle into the nucleus, where it oxidatively modifies and inactivates the NAD-dependent protein deacetylase sirtuin-1 (SIRT1). Failure to deacetylate p53 can lead to apoptotic cell death. b, Epigenetics: the epigenetic landscape is fine-tuned by the prevailing redox environment of the cell. Illustrated here are two redox-sensitive histone modifiers: the histone-lysine N-methyltransferase 2A (MLL1, also known as KMT2A), a member of the COMPASS complex, contains a redox-sensitive zinc centre, whose oxidation leads to a decrease in global H3K4me3 levels. Oxidation of SIRT1 results in the inactivation of its deacetylase function and hence in the accumulation of acetylated lysines on histones. This contributes to alterations in gene expression. c, Gene expression: oxidants and electrophiles directly activate select transcription factors, resulting in changes in gene expression (Table 1). A prominent example is the Kelch-like ECH-associated protein 1 (KEAP1), which binds the nuclear factor erythroid 2-related factor (NRF2) and targets it for degradation under non-stress conditions. Oxidation of redox-sensitive cysteines in KEAP1 leads to the release of NRF2, which subsequently relocalizes to the nucleus, heterodimerizes with a member of the small MAF (sMAF) family of proteins and activates genes encoding proteins of the antioxidant defence systems. d, Protein translation, folding and degradation: these processes are either directly or indirectly controlled through redox-regulated proteins. Oxidative distress causes a global restriction of protein synthesis by causing the phosphorylation of initiation factors (for example, eIF2α), changes in transfer RNA dynamics and ribosome function. The ATP-dependent folding chaperone (that is, foldase) heat shock protein 70 (HSP70) or the protein targeting factor GET3 serves as examples of oxidative stress-activated chaperones. Oxidative modification of key thiols leads to their conversion into ATP-independent holding chaperones (that is, holdases), which bind unfolding proteins and prevent their aggregation. This stress-specific activation compensates for the oxidative-stress-mediated decrease in cellular ATP levels and the concomitant functional downregulation of ATP-dependent chaperones. Protein ubiquitination and 26S proteasome function are inhibited by redox-mediated changes, whereas 20S proteasome activity increases to degrade oxidatively damaged proteins. e, Proliferation, differentiation and migration: many kinases and transcription factors that induce proliferation and differentiation are under redox control. Activation occurs either through the direct oxidation of cysteine switches or indirectly through inhibition of upstream inhibitors of the pathway. Activation of phosphatidyl inositol-3-kinase (PI3K) or extracellular signal-regulated kinases (ERK) is extended through the oxidative deactivation of redox-sensitive upstream phosphatases such as phosphatase and tensin homologue (PTEN) and protein tyrosine phosphatase 1B (PTP1B, not shown). Nuclear factor κ-light-chain-enhancer (NF-κB) is activated directly through cysteine oxidation and indirectly through inhibition of NF-κB inhibitors. Similarly, the stress-activated mitogen-activated protein kinase (MAPK) p38 is both indirectly and directly activated by the presence of oxidants. f, Regulated cell death (RCD) modes: parthanatos is an RCD pathway primarily regulated by oxidant-mediated DNA damage. Overactivation of poly[ADP-ribose] polymerase 1 (PARP1) overproduces polyA-ribose (PAR), which causes the release of the apoptosis-inducing factor from mitochondria causing cell death. Ferroptosis, which is a non-apoptotic form of RCD, is triggered by Fe²⁺-mediated peroxidation of unsaturated fatty acids. Necroptosis is mediated by the tumour necrosis factor (TNF), which activates a multiprotein complex called the necrosome. One redox-sensitive player of the necrosome is receptor-interacting serine/threonine-protein kinase 1 (RIPK1), which, upon oxidation, undergoes autophosphorylation, interacts with RIPK3 and mixed lineage kinase domain-like protein (MLKL) and forms an active necrosome. Apoptosis is mediated by a proteolytic cascade of (pro) caspases, which in their activated form degrade cellular components and cause cell death. Oxidative activation of the apoptotic regulator BAX or oxidative inactivation of the anti-apoptotic factor BFL1 (also known as BCL2A1) triggers apoptosis. Pyroptosis is an inflammatory form of cell death, involving the multiprotein complex inflammasome, whose member caspase 1 activates the pore complex gasdermin D. Inflammasomes containing the nucleotide-binding domain (NOD)-like receptor protein 3 (NLRP3) as their sensor module appear to be especially sensitive to oxidants. Oxeiptosis is an RCD pathway directly controlled by the redox-sensitive player KEAP1 (ref. 206). Upon oxidation of KEAP1, the mitochondrial serine–threonine phosphatase PGAM5 is released and subsequently dephosphorylates and activates the death factor, apoptosis-inducing factor mitochondrial 1 (AIFM1).