Carbonic anhydrase iii s-glutathionylation is necessary for anti-oxidant activity

Carbonic anhydrase isozyme CA3 protects cells against oxidative stress. Ectopic expression of murine Ca3, but not Ca2, protects proto-oncogene Evi1 expressing Rat1 fibroblast cells (ca3low) against hydrogen peroxide (H 2 O 2 ) induced stress. Ca3 is S-glutathionylated via glutathione adducts with cysteines 181 and 186. Substitution of both Ca3 cysteines with serine fails to protect cells from oxidative stress. Insertion of cysteine at 181 and 186 in Ca2 is insufficient for conferring efficient anti-oxidant activity. This shows for the first time that S-glutathionylation of cys181 and cys186 residues is required for Ca3 anti-oxidant activity but that additional factors are also required.

These ubiquitous enzymes are of significant importance for many physiological processes and are implicated in various pathological conditions including atherosclerosis [3], retinitis pigmentosa [4], myasthenia gravis [5] and cancer [6].
Carbonic anhydrase III (CA3) is unique in this family as it has low hydratase activity [7] but it is very abundant in liver, skeletal muscle and adipose tissue. However, surprisingly for such an abundant protein (e.g. 2% of wet weight in slow oxidative muscle (type 1) [8]) its function remains an enigma. Until recently Ca3KO mice were believed to lack any functional deficit [9] but have now been shown to display impaired mitochondrial ATP synthesis [10]. In addition, these mice show changes in expression of genes involved in oxidative stress [11].
This implies CA3 might participate in the cellular response to oxidative stress.
There is mounting evidence that strongly suggests CA3 is an anti-oxidant.
Tissue in which it is abundant is those whose metabolic activity results in considerable oxidative stress, including aerobic respiration in skeletal muscle and lipid metabolism in adipose tissue. CA3 expression is co-induced with established anti-oxidant genes such as superoxide dismutase following endurance training in elite athletes when skeletal muscle is exposed to increased oxidative stress [12] [13].
Functional molecular evidence, in addition to the associated expression studies above, shows that CA3 has anti-oxidant activity. Enforced CA3 expression in NIH3T3 cells protects them from hydrogen peroxide (H 2 O 2 )-induced apoptosis [14]. EVI1 proto-oncoprotein transformed Rat1 fibroblasts have repressed ca3 expression and either transgene mediated restoration of Ca3 in these cells or direct RNAi mediated Ca3 KD in parental Rat1 cells (high endogenous ca3) protects and sensitizes cells to H 2 O 2 induced apoptosis respectively [15]. Insight into a possible mechanism of CA3 anti-oxidant activity has been obtained from post-translational modification observed in cells during oxidative stress. CA3 is modified by S-glutathionylation in response to t-butylhydroperoxide or menadione in cultured hepatocytes [16] as well as in stressed skeletal muscle [11].
Crystal structure studies and site directed mutagenesis reveal two cysteine residues, cys181 and cys186, are available for the addition of glutathione adducts via transient formation of oxidised cysteine sulfenic acid intermediates [17] [18].
S-glutathionylation of CA3 is believed to help protect and aid recovery of cells from the damaging effects of oxidative agents. Previous studies showing that CA3 has anti-oxidant activity and that the protein is S-glutathionylated has led to speculation that the two processes are connected [18]. Direct functional evidence for this is lacking. In this study advantage is taken of our previous analysis in Rat1 fibroblast cells [15]. Derivative Evi1 proto-oncoprotein expressing Rat1 fibroblast cells, designated 5.61, have low ca3 and increased sensitivity to oxidative stress but become resistant upon restoration of ca3 levels by ectopic expression of Ca3. 5.61 cells are used here to explore the relationship between CA3 S-glutathionylation and CA3 anti-oxidant activity.

Statistical Analysis
Statistical significance was determined by two-way ANOVA using GraphPad PRISM ® 7.0c software. P ≤ 0.05 was considered significant.

Ca3 but not Ca2 protects cells from oxidative stress
Murine Ca2 and Ca3 proteins show 60% amino acid identity (Figure 1). In-    (Figure 3(b)). As before, there is a highly significant reduction in caspase 3 enzyme catalytic activity in cells with enforced expression of wt Ca3  pRC CMV vs Ca2S181CN186C ***P < 0.001).

Discussion
It is now widely accepted that CA3 undergoes S-glutathionylation [19] and that this post-translational modification is elevated by either chemical [16] [20], including H 2 O 2 [21] or exercise [11] mediated oxidative stress in cultured cells and intact tissues in various mammalian species examined. X-ray crystallography shows that both rat ca3 and S-glutathionylated ca3 are structurally very similar and that two surface exposed cysteine residues (C181 and C186) participate in adduct formation [17]. Substitution of both C181 and C186 for serine completely abolishes S-glutathionylation of Ca3 [18]. We show in this study that Ca3 C181 and C186 are both essential for in vivo Ca3 anti-oxidant activity and that S-glutathionylation is therefore an essential feature of the mechanism by which this protein participates in protecting cells from oxidative stress.
Mutation of either C181, C186 or both to serine residues are equally effective at inhibition of Ca3 anti-oxidant activity. Previous studies show that C186 is preferentially S-glutathionylated relative to C181, suggesting that this residue might be more important in anti-oxidant activity [18]. However, the same studies also show that the efficiency of C186 S-glutathionylation is significantly reduced in the C181 mutant Ca3 protein (70% reduction). These observations are consistent with our results that mutation of either one or both of these cysteine residues to encode a serine has a significant impact on Ca3 biological activity as an anti-oxidant.
Our results also confirm previous studies [14] that the closely related Ca3 The S-glutathionylated cysteine residues in Ca3 have a low pKa [18] and S-glutathionylation is affected positively and negatively by lysine 211 and glutamic acid 188/aspartic acid 212 respectively. The pKa of the cysteine residues in mutant Ca2 have not been determined, however the lysine, aspartic acid and glutamic acid amino acids are conserved in Ca2. The S-glutathionylation process can occur spontaneously (reviewed in [22]) but can be catalysed. Glutathione-S-Transferase π (GSTP) catalyses S-glutathionylation of both 1-CYS peroxiredoxin [23] and cardiac aldose reductase [24]. Homozygous KO mice depleted of GSTP show a general reduction in oxidative stress induced protein S-glutathionylation [25] suggesting this protein might be involved in conjugation of glutathione to other proteins, including Ca3. Therefore, there might be other molecular determinants of Ca3 S-glutathionylation besides C181 and C186 that are absent from Ca2.
The CA3 C181 and C186 residues are conserved in all mammalian species examined, including human, rat, murine and bovine, but only C181 is found in xenopus and neither residues are observed in chicken or zebrafish (data not shown). This suggests that only mammalian CA3 has evolved anti-oxidant activity, but this would need to be tested experimentally as other cysteine residues might be S-glutathionylated in non-mammalian species. For example, the CA3 isozyme CAVII is also S-glutathionylated but at cysteine residues C183 and C217 [26].