Role of SLC4 and SLC26 solute carriers during oxidative stress.

The SLC26 family of multifunctional ion transporters and channels comprises 11 genes ( SLC26A1‐11 ) of which 10 are protein coding and one ( SLC26A10 ) is a pseudogene. With the exception of SLC26A5 , which encodes for the outer hair cells motor protein prestin, these genes encode multifunctional anion exchangers, of which SLC26A7, A9 and A11 can also operate as uncoupled ion transporters in a channel‐like mode. The anion exchangers of this family can accept divalent as well as monovalent anions and generally show versatile substrate selectivity, being able to exchange chloride for inorganic anions such as bicarbonate, hydroxyl, sulphate and iodide, or small organic anions such as formate and oxalate. SLC26A3, A4, A6, A7, A9, and A11 can operate in Cl − /HCO 3 − exchange mode, while SLC26A1 and A2 are selective sulphate transporters. 115 , 116 , 117 The SLC26 family was discovered more recently and is less well characterized compared with SLC4. SLC26 members are highly hydrophobic, large (700‐1,000 amino acids) proteins sharing relatively low (21%‐43%) amino acid identity with the members of the same family. The structure of mammalian SLC26 transporters has long been elusive. Structures of the murine Slc26a9, 118 two bacterial homologues—a bicarbonate transporter from cyanobacteria (BicA) 119 and a proton‐coupled fumarate transporter from the bacterium Deinococcus geothermalis (SLC26Dg), 120 and a plant homologue of SLC26—the vacuolar H + /SO 4 2 symporter SULTR4;1 from Arabidopsis thaliana 121 were solved recently by X‐ray crystallography and cryo‐electron microscopy. These structures show a shared architecture consisting of 14 transmembrane α‐helices organized in two intertwined inverted repeats of seven transmembrane segments each and a large C‐terminal cytosolic domain referred to as the sulphate transporter and anti‐sigma factor antagonist (STAS) domain. It is widely accepted that mammalian SLC26 members and their lower organism counterparts act as functional homodimers, although the dimerization mechanism was not unequivocally determined and probably differs among family members. Two swapped STAS domains have been determined as major determinants of the homodimeric assembly of murine Slc26a9, 122 Arabidopsis thaliana SULTR4;1 121 and BicA from Synechocystis sp, 119 while the STAS domain is not a requisite for SLC26Dg dimerization, which indeed relies on contacts between the transmembrane domain and is centred on TM14. 123 Some of the members of the SLC26 family show restricted tissue distribution, whereas other isoforms are more broadly expressed and orchestrate the ion transport across epithelia such as the intestine and kidney. 124 , 125 Pathogenic sequence alterations in SLC26 genes cause inherited diseases, including diastrophic dysplasia (phenotype MIM number 222600) and other osteochondrodysplastic syndromes (SLC26A2), secretory chloride diarrhoea (SLC26A3, phenotype MIM number 214700), Pendred syndrome (SLC26A4, phenotype MIM number 274600) and non‐syndromic deafness (SLC26A4 and A5, phenotype MIM numbers 600791 and 613865, respectively), calcium oxalate nephrolithiasis (SLC26A1 and A6, phenotype MIM number 167030), and spermatogenic failure (SLC26A8, phenotype MIM number 606766). SLC26A3, A4, A6, A8 and A9 physically and/or functionally interact with the CFTR chloride channel and regulate the CFTR activity or are regulated by CFTR. 126 , 127 , 128 , 129 The information regarding the impact of OS on SLC26 function is sparse, and will be reviewed in the following and in Table  2 . Link between SLC26 members and oxidative stress (OS). Anion exchangers of this family are all Na + ‐independent SLC26A3/DRA was initially identified as a candidate tumour suppressor gene that was downregulated in adenoma (hence the alias DRA). 130 Later, SLC26A3 was identified by positional cloning as the gene mutated in recessive congenital chloride‐losing diarrhoea. 131 This transporter is expressed on the apical membrane of epithelial cells in the intestine and is particularly abundant in the colon and duodenum. SLC26A3 functions as a Cl − /HCO 3 −  exchanger in tandem with NHE3, which mediates a Na + /H +  exchange, to generate an electroneutral NaCl reabsorption across the intestinal mucosa, or gives rise to HCO 3 −  secretion in the absence of NHE3. 117 Oxidative stress plays an essential role in the pathogenesis and progression of inflammatory bowel disease (IBD). 132 It is well known that SLC26A3 expression is reduced in animal models of intestinal inflammation as well as in patients with ulcerative colitis (UC), an effect that was attributed to inhibition of gene transcription by proinflammatory cytokines. 133 , 134 Interestingly, H 2 O 2 inhibited the SLC26A3 and SLC26A6‐mediated Cl − /OH − (HCO 3 − ) exchange activity in Caco‐2 cells independently of the prostaglandin/COX‐dependent pathway, but occurring via a signalling cascade that involved the activation of the Src kinase Fyn, PI3K, PLC γ 1 and the Ca 2+ ‐dependent PKC α . 135 These authors concluded that phosphorylation of the anion exchangers or regulatory proteins, rather than modification of their plasma membrane trafficking, might be involved in modulation of their activity. In a recent study, the gene expression of SLC26A3 was decreased in the colonic mucosa from patients with active UC compared with patient in remission and non‐inflamed donors. In parallel, SOD2 was significantly upregulated in the colonic mucosa from patients with active UC compared with controls, thus denoting an increased OS, and SOD2 levels correlated with severe histological activity in the inflamed mucosa. 136 Together, these findings suggest that OS might contribute to inhibition of expression and function of SLC26A3 in the context of IBD, thus aggravating fluid loss and stool acidification consequent to lack of NaCl reabsorption and HCO 3 − secretion. SLC26A4/pendrin is an electroneutral Cl − /anion exchanger abundantly expressed on the apical membrane of distinct epithelial cells of the stria vascularis and endolymphatic duct and sac of the inner ear, as well as in the thyroid and kidney. In the inner ear, pendrin drives HCO 3 − secretion and Cl − reabsorption and controls the endolymphatic pH and volume. In the thyroid, pendrin participates in the iodide flux into the thyroid follicle, possibly via its I − /Cl − exchange activity. In addition, pendrin expression is upregulated in the airways and oesophageal mucosa by the pro‐inflammatory cytokines IL‐4/IL‐13 via a STAT6‐mediated pathway. 137 , 138 Loss or reduction of function of pendrin consequent to gene mutation causes autosomal recessive forms of non‐syndromic as well as syndromic sensorineural hearing loss, ie, DFNB4 and Pendred syndrome, of which a malformation of the inner ear called enlarged vestibular aqueduct (EVA) with or without cochlear incomplete partition type 2 is the main radiological finding. Vestibular dysfunction is also observed in a fraction of patients. In Pendred syndrome, deafness is associated with a partial iodide organification defect in the thyroid that may lead to subclinical or overt hypothyroidism and goiter. 139 , 140 In the cortical collecting duct and connecting tubule of the kidney nephron, pendrin is expressed on the apical membrane of β and non‐α, non‐β intercalated cells, mediates Cl − reabsorption and HCO 3 − excretion and controls the systemic electrolyte, vascular volume and acid‐base homoeostasis by working in concert with other ion absorbing transport systems, including the sodium chloride cotransporter NCC and the epithelial sodium channel ENaC. 141 A possible link between OS and pendrin expression and function in the inner ear was firstly reported by Wangemann and collaborators, who described hyperpigmentation of the stria vascularis in pendrin‐knockout mice. Hyperpigmentation of the stria vascularis was linked to an increased melanin production in the strial intermediate cells, which are in charge of detoxification of free radicals produced by the metabolically active strial marginal cells. The authors proposed a model where lack of bicarbonate secretion in the endolymph due to the absence of pendrin, with consequent increase of bicarbonate concentration and pH in the intra‐strial fluid, may have led to inhibition of cysteine uptake and glutathione synthesis in the intermediate cells and free radical damage of these cells. Oxidative stress in the intermediate cells was proposed to be responsible for the reduced protein expression of KCNJ10/Kir4.1, an inwardly rectifying K + channel essential for the generation and maintenance of the endochoclear potential and proper hearing function. 142 In this context, pH control via pendrin would exert a protective role against OS. Accordingly, elevated amounts of oxidized and nitrated proteins were found in the stria vascularis of pendrin‐knockout mice, and OS reduced Kcnj10 expression in a heterologous system. 143 These findings underscore the physio‐pathological relevance of OS, leading to failure of KCNJ10 expression, lack of endochoclear potential and deafness in mouse models and possibly in patients with pendrin‐related hearing loss. A recent study has evidenced a significant increase in the Slc26a4 transcript levels in the cochlea of three distinct mouse models of age‐related hearing loss linked to chronic OS, represented by mice subjected to intermitted hypoxia alone, high‐fat diet plus galactose injection and intermitted hypoxia combined with high‐fat diet and galactose injection for 12 weeks, respectively. These changes were paralleled by an increase in the hearing threshold and morphological damage of cochlear hair cells. 144 Although the pathophysiological significance of the altered Slc26a4 transcript levels in these mouse models remains unclear, this study suggests that pendrin transcriptional regulation might respond to OS with compensatory changes. In the thyroid, iodide reaches the follicular lumen via pendrin and is subsequently incorporated (organified) into thyroglobulin (Tg) by the thyroid peroxidase (TPO), which requires H 2 O 2 as cofactor. H 2 O 2 is generated by dual oxidases (DUOXs). In a normal thyroid, TPO and DUOX expression is restricted to the apical membrane of thyrocytes or confined within Caveolin‐1‐positive intracellular vesicles. Similar to what was observed in the inner ear of pendrin‐knockout mice, OS was augmented in the thyroid of a patient with Pendred syndrome, as evidenced by greatly increased lipid peroxidation as well as catalase and peroxiredoxin PRDX5 expression. In parallel, TPO and DUOX showed greatly enhanced expression, lost their normal apical distribution and were mislocalized within the cytosol, while the expression of Caveolin‐1 was absent. These authors linked the increased OS in the thyroid tissue with aberrant, intracellular thyroid hormone biosynthesis, a process that would require H 2 O 2 production within the cytosol. 145 Although the mechanism by which TPO and DUOX were mislocalized remained unexplained, these findings underscore the importance of proper pendrin activity in protecting from dramatic OS‐related pathological changes in the thyroid. Pendrin transcript was found downregulated (−4.6 and −6.0 fold after 52 and 100 weeks, respectively) in the kidney from F344 male rats exposed to 400 ppm (high) potassium bromate in drinking water. Potassium bromate is a nephrotoxic compound and a kidney and thyroid carcinogen, can induce deafness and promotes OS, as evidenced by de‐regulation of OS‐related genes. 146 Although the mechanism leading to decreased pendrin mRNA levels remained unexplained, this study suggests that pro‐oxidant compounds might suppress pendrin transcription in the kidney and perhaps other organs, including the inner ear. Further studies are needed to verify this hypothesis. Similarly to thyroid, the airway and lung epithelium has the ability to generate H 2 O 2 at the apical membrane through DUOX1 and DUOX2. H 2 O 2 serves to oxidize luminal SCN − to OSCN − (hypothiocyanite) in a reaction catalyzed by lactoperoxidase and other peroxidases. Both anions are pro‐oxidant species, have an antimicrobial role and play an important function in the innate mucosal defense. Pendrin is a major determinant of the IL‐4/IL‐13‐stimulated SCN − secretion at the luminal surface of human bronchial epithelial cells and lung. 147 , 148 , 149 When upregulated by ILs in the context of airway inflammation and allergic asthma, SCN − secretion may lead to increased OSCN − production. OSCN − in low doses can be sensed by the OS‐sensitive PKA, which dimerizes and activates NF‐κB, a transcription factor critical for inflammatory responses. High doses of OSCN − can induce necrosis of epithelial cells, further triggering inflammation. 149 The implications of these findings are that (i) a defective SCN − secretion via pendrin might be linked to increased bacterial colonization of the airway epithelium 147 and (ii) inhibition of pendrin activity 150 or peroxidase‐dependent conversion of SCN − to OSCN − 151 might be beneficial in treating asthma and other chronic inflammatory airway conditions. These studies highlight how ion transport, OS and inflammation are tightly interconnected and their dysregulation might profoundly affect mucosal function and health. SLC26A5/prestin is the transmembrane sensor‐motor protein of the outer air cells (OHC) of the mammalian organ of Corti and enables changes in length of these cells depending on the frequency of a sound stimulus, which is the basis of the mechanism of cochlear amplification of sounds. 152 Although prestin shares high homology with other members of the SLC26 family of proteins including pendrin and DRA, it does not transport ions across the plasma membrane, but instead changes its structure by voltage‐dependent translocation of anions within the molecule itself. 153 Targeted deletion of prestin exons 3‐7 in mice resulted in loss of OHC electromotility in vitro and a 40‐60 dB loss of cochlear sensitivity in vivo. 154 A limited number of SLC26A5 mutations has been identified in patients with hearing loss; one of these mutations was linked to autosomal recessive deafness DFNB61 but later was found non‐pathogenic. 9 Due to the limited number of patients, whether any of the identified SLC26A5 mutations are true causes of hearing loss remains unclear. Katbamna et al have shown that prenatal smoke exposure (≥10 cigarettes/d) leads to significant reductions in the cochlear response amplitudes and auditory brainstem recording (ABR) wave latencies in newborns. These alterations in the auditory function were paralleled by a dysregulation of the expression of placental genes, including SLC26A5, which was upregulated 8.95 fold, again suggesting an adaptive change. Also, several OS pathway genes were found dysregulated. The authors suggested that placental gene expression might be a good surrogate of foetal gene expression. 155 The study raises the possibility that prenatal exposure to OS might damage the hearing function in the newborn by targeting genes involved in regulating the hearing function. Oxidative stress following production of ROS is a hallmark of noise‐induced hearing loss and mainly targets OHC. 156 Prestin was used to selectively target ROS‐scavenging nanoparticles to OHC in a guinea pig model of noise‐induced hearing loss. This innovative strategy allowed for preservation of the morphological integrity of OHC and led to significant improvement of the hearing function, denoting that protection of OHC and their delicate molecular machinery from OS is crucial in the prevention and treatment of noise‐induced hearing loss. 157 SLC26A6 putative anion transporter 1 (PAT1) and chloride‐formate exchanger (CFEX) was cloned based on the homology to the genes encoding SLC26A3 and SLC26A4, was found most abundantly expressed in the kidney and pancreas, 158 but was also detected in other tissues, including the intestine. SLC26A6 can work in Cl − /HCO 3 − , Cl − /OH − , chloride/oxalate and chloride/formate exchange modes. 159 Mutant mice lacking Slc26a6 develop calcium oxalate urolithiasis, have significant hyperoxaluria and elevation in plasma oxalate concentration, both resulting from a defect in intestinal oxalate excretion, which leads to an enhanced net absorption of oxalate. 160 In the human pancreas and intestine, SLC26A6 is co‐expressed with SLC26A3 and CFTR and participates in chloride‐dependent HCO 3 − secretion by acting synergistically with CFTR; in addition, SLC26A6 mediates oxalate excretion in the intestine and kidney proximal tubule and Cl − /HCO 3 − exchange in the myocardium. These findings imply a role of SLC26A6 in the development of intestinal and pancreatic diseases, nephrolithiasis and arrhythmia. 161 In normal rat proximal tubular epithelial (NRK‐52E) cells, SLC26A6 expression correlated with oxalate‐induced cell injury, apoptosis, crystal adhesion, ROS formation and lipid peroxidation. Accordingly, selective attenuation of SLC26A6 expression in the kidney of rats with lentivirus‐transfected siRNAs decreased SOD generation, cell apoptosis and crystal formation. 162 This study highlights that SLC26A6‐mediated oxalate excretion in the kidney might be a determinant of stone formation as well as OS‐induced cell injury and identifies SLC26A6 as a target of potential protective therapeutic approaches.

All authors contributed to performing the literature search and writing the manuscript. All authors have read and agreed to the final version of the manuscript.

Several studies have investigated in detail the impact of OS on some transporters of the SLC4 family in different experimental conditions (Table  1 ). In particular, great attention has been reserved to SLC4A1. In this regard, human erythrocytes exposed to various oxidizing agents or extracellular pH variations have shown both a reduction in transport efficiency and changes in the structural state of SLC4A1. Instead, in ex vivo or in vitro models of inflammation, diabetes mellitus and ageing, the SLC4A1 ion transport was found accelerated. While the reasons of this apparent discrepancy are only partly elucidated, it becomes increasingly obvious that the activity of SLC4A1 is sensitive to OS and increases or decreases in ion transport efficiency most likely reflects the formation and/or the amount of glycated haemoglobin and methaemoglobin, as well as the integrity of the lipid bilayer ad the intricate intracellular network of interacting proteins (Figure  3 ). The potential impact of OS on the expression and activity of SLC26 family members was less extensively studied compared with SLC4A1. For several SLC26 members, there is no information available in this context. The current knowledge denotes a dual role of these transporters in being potential targets as well as determinants of OS (Figure  3 ). For example, function and possibly expression of SLC26A3 and SLC26A6 can be reduced during OS. Similarly, SLC26A4 and SLC26A5 gene expression appeared dysregulated in OS. In turn, lack of expression of SLC26A4 was associated with increased OS in the human thyroid and mouse cochlea. At the same time, SLC26A4 upregulation in the context of inflammation and SLC26A6 appeared aggravating or mediating OS (Table  2 ). More studies are needed to explain how SLC26 members sense, respond to or determine OS. Particularly intriguing is the hypothesis of a possible dysfunction of SLC26A4 and SLC26A5 mediated by OS in noise‐induced and/or age‐related hearing loss. Further research should establish whether and how members of the SLC4 and SLC26 family of proteins can be safely and effectively targeted by antioxidants in the prevention and treatment of OS‐related conditions, including inflammation, metabolic dysfunctions, hearing loss and ageing.

Oxidative stress (OS) is frequently described as an imbalance between the production of reactive oxygen species (ROS) in biological systems and the ability of the latter to defend themselves through their sophisticated antioxidant machinery. 1 , 2 The term ROS is applied to a collection of highly reactive chemicals, namely free radicals, including superoxide anion (O 2 ·− ), hydroxyl radical (OH · ) and singlet oxygen ( 1 O 2 ), or alternatively their non‐radical intermediates, such as hydrogen peroxide (H 2 O 2 ), nitric oxide (NO) and hypochlorous acid (HOCl), which are typically less reactive. 3 , 4 In eukaryotes, endogenous ROS represent a normal byproduct of the cell metabolism and mainly arise from the incomplete reduction of O 2 during the process of oxidative phosphorylation in the mitochondria. 5 In addition, endogenous ROS are produced by various cellular enzymes, including NADPH oxidases (NOXs) and nitric oxide synthase (NOS), peroxisomes, ionizing and UV radiation, as well as by the metabolism of drugs and xenobiotics. 6 , 7 However, the redox state in the cell is normally regulated by a complex endogenous antioxidant system, which is composed of proteins with enzymatic activities, like glutathione peroxidase, catalase (CAT), and superoxide dismutase and non‐enzymatic small‐molecule compounds, like glutathione, which are able to quickly neutralize ROS and ensure a low production of reactive species. In addition, antioxidant molecules, such as vitamin C, vitamin E, some minerals, carotenoids, and polyphenols, which can be supplied through the diet, can help the activity of endogenous antioxidants, thus promoting the redox homoeostasis of the cell. 8 Nevertheless, some oxidants in controlled amount possess important cell defense and signalling functions within the cell. Specifically, cells can generate ROS with function of second messengers, use them for intracellular signalling and for stimulating redox‐sensitive pathways in order to modify the cellular levels of cytoprotective regulatory proteins. 9 , 10 For example, the massive ROS production by activated macrophages not only represents a first‐line defense against environmental pathogens, but can also stimulate T‐cell function, including the production of cytokines. 11 However, when oxidants are produced in excess, or when the antioxidant defenses that regulate them are ineffective, the balance between antioxidant and pro‐oxidant capacity can be perturbed, thus resulting in OS and ultimately cell death through apoptosis or necrosis. In these conditions, biomolecules such as nucleic acids, membrane lipids, enzymes, and structural proteins can be altered through oxidation to an extent that exceeds repair capacity. 12 , 13 Abnormal ROS levels can influence several cell signalling pathways. In this regard, the role of OS in the pathogenesis of disease is widely acknowledged. A perturbed redox homoeostasis may be the common denominator underlying ageing and different chronic diseases, 14 , 15 , 16 , 17 , 18 , 19 although specific mechanisms contributing to OS‐induced damage are poorly investigated. By representing the boundary between the cell interior and the extracellular environment, the cell membrane is most vulnerable to free radical attack. 20 The plasma membrane contains a wide range of proteins and lipids that elicit distinct cellular reactions in response to extracellular stimuli and stressors, and each of these can be target of OS. The solute carrier (SLC) superfamily of transporters comprises integral membrane proteins known as the gatekeepers for all cells, as they control the transmembrane flux of inorganic ions, sugars, amino acids, nucleotides, fatty acids, neurotransmitters, and drugs. 21 Currently, the human SLC superfamily includes over 458 members grouped into 65 families. 22 This organization has been established by the Gene Nomenclature Committee (HGNC) of the Human Genome Organization (HUGO), and arranged such that member proteins within each family share at least 20%‐25% sequence similarity with at least one other member of the family. 23 The SLC transporters are grouped into four main types according to the transport mode, which are cotransporters, exchangers, facilitated transporters and orphan transporters (with no known substrate), thus playing a central role in a plethora of physiological and pathological functions in almost all cells and tissues. 22 These membrane transporters are widely expressed throughout the body, most notably in the epithelia of major organs, such as the liver, intestine, kidney and organs with barrier function, such as the blood‐brain barrier, testes and placenta. Many transporters are also expressed in an organ‐specific manner, thus facilitating the entry and elimination of endogenous and xenobiotic compounds. 24 The substrate specificity of these transporters can be determined by interactions between the amino acid backbone and/or side chains and the substrate, as well as by intramolecular interactions that regulate gating and/or selectivity elements. 25 The modulation of the activity of membrane transport systems is part of the cellular response to OS. 26 In particular, the link between OS and membrane transport systems during ageing as well as in OS‐related diseases is incompletely understood, and is essential for a deeper understanding of mechanisms through which such processes and diseases develop. Indeed, the relationship between ion transport and cellular redox balance is complex. The activity of ion transporters and channels can be stimulated or blocked during OS and, in turn, these molecular entities can even be involved in determining or attenuating OS. To name just a few examples, volume regulated anion channels LRRC8/VRAC can be either activated or inhibited by OS depending on their subunit composition, while LRRC8/VRAC inhibitors seem to lower OS. 27 The expression as well as the function of the cystic fibrosis transmembrane conductance regulator (CFTR) chloride channel is impaired during OS, thus contributing to the progression of airway dysfunctions, including chronic obstructive pulmonary disease (COPD), consequent to alteration of the mucociliary transport in airway epithelial cells. 28 On the other hand, CFTR is permeable to not only chloride ions, but also organic anions such as reduced glutathione, and consequently CFTR dysfunction potently contributes to the OS burden at the airway surface in cystic fibrosis. 29 Recent studies have established that the Na + /K + ‐ATPase can cause OS with mechanisms distinct from its well‐understood function of ion pump but rather dependent on the scaffolding properties of the alpha1 subunit, which is targeted by OS and post‐translationally modified, thus leading to the activation of a downstream signalling cascade eventually amplifying ROS production. 30 In this review, we will focus on the impact of OS on plasma membrane SLC transporters, and specifically on the members of SLC4 and SLC26 families of chloride/bicarbonate (Cl − /HCO 3 − ) exchangers, which are essential for maintaining crucial homoeostatic functions, such as the regulation of systemic and intracellular pH and ion composition, and the regulation of cell and extracellular fluid volume. 31 , 32 We will emphasize their pathophysiological relevance and their role in OS‐related conditions, including ageing. In this regard, only isoforms involved in OS events will be considered. Also, we will discuss their potential as targets of antioxidant therapies.

The authors declare no conflict of interest.