Genetic engineering of AtAOX1a in Saccharomyces cerevisiae prevents oxidative damage and maintains redox homeostasis

This study aimed to validate the physiological importance of Arabidopsis thaliana alternative oxidase 1a (AtAOX1a) in alleviating oxidative stress using Saccharomyces cerevisiae as a model organism. The AOX1a transformant (pYES2AtAOX1a) showed cyanide resistant and salicylhydroxamic acid (SHAM)‐sensitive respiration, indicating functional expression of AtAOX1a in S. cerevisiae. After exposure to oxidative stress, pYES2AtAOX1a showed better survival and a decrease in reactive oxygen species (ROS) when compared to S. cerevisiae with empty vector (pYES2). Furthermore, pYES2AtAOX1a sustained growth by regulating GPX2 and/or TSA2, and cellular NAD +/NADH ratio. Thus, the expression of AtAOX1a in S. cerevisiae enhances its respiratory tolerance which, in turn, maintains cellular redox homeostasis and protects from oxidative damage.

This study aimed to validate the physiological importance of Arabidopsis thaliana alternative oxidase 1a (AtAOX1a) in alleviating oxidative stress using Saccharomyces cerevisiae as a model organism. The AOX1a transformant (pYES2AtAOX1a) showed cyanide resistant and salicylhydroxamic acid (SHAM)-sensitive respiration, indicating functional expression of AtAOX1a in S. cerevisiae. After exposure to oxidative stress, pYES2-AtAOX1a showed better survival and a decrease in reactive oxygen species (ROS) when compared to S. cerevisiae with empty vector (pYES2). Furthermore, pYES2AtAOX1a sustained growth by regulating GPX2 and/or TSA2, and cellular NAD + /NADH ratio. Thus, the expression of AtAOX1a in S. cerevisiae enhances its respiratory tolerance which, in turn, maintains cellular redox homeostasis and protects from oxidative damage.
Alternative oxidase (AOX) is a nonproton pumping ubiquinol oxidase localized in the inner mitochondrial membrane of higher plants, fungi, some protists and was recently identified in 28 animal species [1]. In contrast to cytochrome c oxidase (COX), it is cyanide resistant and branches from the 'standard' mitochondrial respiratory chain at the level of ubiquinone (UQ). It is considered as a sink for excess electrons as it reduces the molecular oxygen to water, bypassing the oxidative phosphorylation at both complex III and IV. Thus, AOX plays an important role in maintaining the cellular energy balance [2][3][4]. A crystal structure of AOX from Trypanosome brucei revealed that it is a homodimer, which exists as an integral interfacial membrane protein with a nonhaem diiron carboxylate active site buried within a four helix bundle. The active site is ligated by four glutamate residues and a highly conserved Tyr220, which mediates its catalytic activity. Furthermore, the two hydrophobic cavities occur per monomer which bind to ubiquinol and Tyr220 for catalytic cycle and O 2 reduction [5][6][7].
AOX was first identified in thermogenic plants to provide favorable temperature during floral development to attract pollinators [8][9][10][11]. In nonthermogenic plants, AOX is known to prevent over-reduction of UQ and generation of reactive oxygen species (ROS) while allowing continued operation of the tricarboxylic acid (TCA) cycle [12][13][14]. On exposure to abiotic stress, AOX-deficient plants showed an increase in intracellular ROS and a decrease in photosynthetic performance as compared to wild-type plants [15][16][17][18][19]. On the other hand, AOX overexpression lines showed an enhanced photosynthetic efficiency with lower levels of cellular ROS when compared with wild-type plants during abiotic stress conditions [20][21][22]. In Arabidopsis, overexpression of AOX1a alleviated the Al-induced programmed cell death (PCD) by decreasing the ROS production due to efficient mitochondrial electron flux and caspase-3-like activation [23]. Also, the role of AOX has been studied extensively in lower organisms since last two decades. Kumar [43][44][45], which indicate that genetic engineering of AOX1a might be a promising tool to combat oxidative stress in AOX deficient strains or organisms. In the present study, AtAOX1a was heterologously expressed in S. cerevisiae (an eukaryotic organism devoid of AOX) to characterize its role in response to oxidative stress. To create an oxidative environment inside the cells, S. cerevisiae were incubated with H 2 O 2 and tertiary-butyl hydroperoxide (t-BOOH). The functional expression of AtAOX1a and its characterization have been studied by monitoring the changes in respiration, growth, viability, ROS, antioxidant system, and redox state of S. cerevisiae under oxidizing conditions. yeast nitrogen base without amino acids, 2% w/v glucose as carbon source) and amino acids.

Cloning of AtAOX1a and plasmid construction
Total RNA was isolated from A. thaliana wild-type leaves using TRI reagent (Sigma-Aldrich, St. Louis, MO, USA). One microgram of total RNA was used for the first-strand cDNA synthesis using iScript TM cDNA synthesis kit (Bio-Rad, Hercules, CA, USA). AtAOX1a encoding a mature protein was amplified by Phusion DNA polymerase (Clontech, CA, USA) using the following primers: F-GAGAATTCGCTAGCACGATCACTCTGG and R-GGCTCGAGTCAATGATACCCAATTGGAG, and cloned into a pET28a(+) TM expression vector. In contrast, AtAOX1a encoding a mature protein along with its leader sequence was amplified by using the primers: F-GG GAATTCTGATGATGATAACTCGCGGTGG and R-G GCTCGAGTCAATGATACCCAATTGGAG, and cloned into a pYES2/NT expression vector. Clones were confirmed by DNA sequencing. The recombinant plasmids were transformed into their respective host strains, i.e., BL21(DE3) pLysS and INVSc1.

Protein expression, purification, and antibody generation
The expression of AtAOX1a in E. coli BL21(DE3)pLysS was induced by 0.1 mM isopropyl-b-D-thiogalactopyranoside (IPTG) at 28°C for 4 h. The recombinant protein was purified under denaturing conditions with Ni-NTA agarose column using standard protocols and the purified protein from the gel slice was subjected to a matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF/TOF) analysis as described in ref. [46] for confirmation as AtAOX1a. The purified protein was used to generate a polyclonal antibody in rabbit using standard protocols (Animal ethics approval number is UH/IAEC/KPMS/2014-1/24).

Oxidative stress
Oxidative stress analyses were performed as described earlier [48]. Treatment duration was different for each set of experiments depending on their feasibility. The duration of oxidative stress treatment was fixed at 10 min for ROS estimation, 4 h for survival rate and growth recovery assay, and 75 min for pyridine nucleotides and transcript level analyses.

Measurement of O 2 uptake and cell survival rate
The respiratory O 2 uptake measurements (10 min) were performed using Clark-type O 2 electrode [49,50]. The viability of cells was examined with fluctuation assay as reported by Dalal et al. [50].

Measurement of ROS
The intracellular ROS level was measured following Jang et al. [51]. The cells were incubated with 100 lM 2 0 ,7 0dichlorodihydrofluorescein diacetate (H 2 DCF-DA; Sigma) for 5 min in the dark at 25°C, and the change in DCF fluorescence was imaged under a laser-scanning confocal fluorescence microscope (LSM 710 NLO ConfoCor 3; Carl Zeiss, Jena, Germany).

Measurement of pyridine nucleotide content
The extraction and estimation of NAD + and NADH were done as per Queval and Noctor [52]. The assay involves phenazine methosulfate (PMS) catalyzed reduction of dichlorophenolindophenol (DCPIP) in the presence of alcohol dehydrogenase (ADH) and ethanol. The NAD + and NADH content were calculated using the relevant standard (0-40 pmole).

RNA isolation and expression analysis
Total RNA was isolated using the acid-phenol method [53]. First strand cDNA was synthesized with 2 lg of total RNA using SuperScript Ò III (Invitrogen) according to manufacturer's instructions. Primers used for real-time PCR analysis are listed in Table 1 [42]. Comparative C T method was used to analyze the relative gene expression levels [54].

Statistical analysis
All values are presented as means AE standard errors of the means (SEM). The statistical evaluation of the data was

Expression of AtAOX1a in Escherichia coli and mass analysis
The expression of AtAOX1a protein induced in the presence of 0.1 mM IPTG in E. coli was visualized on SDS/PAGE as a~36 kDa band as it includes AtAOX1a sequence encoding a mature protein (32.34 kDa) and pET28a(+) vector sequence (3.83 kDa) (Fig. 1A). Four major peptide fragments obtained during MALDI-TOF-TOF analysis of a trypsin-digested protein showed the following sequences in Biotools: WPTDLFFQR (1209.81 Da), DVNHFASDIHYQGR (1658.04 Da), GNIENVPAPAIAIDYWR (1898.27 Da), and ELDKGNIENVPAPAIAIDYWR (2383.58 Da). As the sequences from these peptides showed 100% matching with Arabidopsis AOX1a (Fig. 1B, and Figs S1, S2A-D), the purified protein was injected into a rabbit and the polyclonal antibody was obtained. Table 1. List of primers used in real-time PCR study. ACT1 was used as housekeeping gene.

Functional characterization of AtAOX1a in Saccharomyces cerevisiae
The protein expression of AtAOX1a in S. cerevisiae was confirmed through western blot analysis ( Fig. 2A).
To ascertain the function of AtAOX1a, cyanide-sensitive respiration was monitored using 1 mM KCN, an inhibitor of complex IV in COX pathway, while cyanide-insensitive respiration was monitored in the presence of 2 mM salicylhydroxamic acid (SHAM) or 100 lM propyl gallate (PG), inhibitors of AOX in the alternative pathway. In the absence of metabolic inhibitors, the respiratory rates of pYES2AtAOX1a (8.6 AE 0.11 nmol O 2 s À1 ) were similar to pYES2 (8.45 AE 0.09 nmol O 2 s À1 ). But, in the presence of KCN, pYES2 showed a pronounced decrease in respiratory rates when compared with pYES2AtAOX1a. In contrast, addition of SHAM or PG significantly decreased the respiratory rates of pYES2AtAOX1a but not of pYES2 (Fig. 2B).
The exponential growth pattern of both pYES2 and pYES2AtAOX1a were found to be similar (OD 600 = 2.1) up to 6 h. But, treatment with KCN remarkably decreased the exponential growth in yeast cells (Fig. 2C). However, the decrease in exponential growth of pYES2 was more significant when compared with pYES2AtAOX1a. Furthermore, in the presence of KCN, growth recovery was found to be higher in pYES2AtAOX1a than pYES2 (Fig. 2D). Taken together, these results indicate that AtAOX1a was successfully expressed and functional in S. cerevisiae.

Changes in cellular ROS during oxidative stress
Under control conditions, the cellular ROS was minimal in both pYES2 and pYES2AtAOX1a as indicated by DCF fluorescence. However, upon treatment with KCN, H 2 O 2 , or t-BOOH, the fluorescence increased significantly in pYES2. In contrast, pYES2AtAOX1a restricted the increase in fluorescence during oxidative stress indicating the importance of AOX1a in preventing and/or regulating the ROS generation (Fig. 3).

Changes in cell survival rate and growth recovery during oxidative stress
Among the two oxidants, H 2 O 2 was found to be more lethal than t-BOOH. Upon treatment with these oxidants, the survival rate of pYES2 decreased drastically as compared to pYES2AtAOX1a (Fig. 4A). Also, recovery assays clearly indicated an enhanced colony number in pYES2AtAOX1a than in pYES2 under oxidizing conditions with a clear visible difference at 1 9 10 À2.5 and 1 9 10 À3 dilutions (Fig. 4B). It appears that AOX1a plays a critical role in decreasing the rates of cell death and improving their growth recovery under oxidizing conditions.

Differential antioxidant gene expression profile during oxidative stress
The ROS scavenging efficiency of pYES2 and pYE-S2AtAOX1a was measured by monitoring the changes in transcript levels of antioxidant genes viz., Superoxide dismutase 1 (SOD1), Superoxide dismutase 2 (SOD2), Glutathione peroxidase 2 (GPX2), and Thioredoxin peroxidase 2 (TSA2) during oxidative stress ( Fig. 5A-D). Under control conditions, the expression of these antioxidant genes was approximately similar in both pYES2 and pYES2AtAOX1a. Upon treatment with H 2 O 2 or t-BOOH, the expression of SOD1 (> 8fold), SOD2 (> 6-fold), GPX2 (> 52-fold), and TSA2 (> 157-fold) increased significantly by several fold in both pYES2 and pYES2AtAOX1a (Fig. 5A-D). But, the expression of GPX2 was down-regulated significantly  in pYES2AtAOX1a when compared with pYES2 in the presence of both H 2 O 2 and t-BOOH (Fig. 5C). In contrast, the expression of TSA2 was down-regulated significantly in pYES2AtAOX1a when treated with t-BOOH, while remained unchanged in the presence of H 2 O 2 (Fig. 5D).

Changes in cellular redox during oxidative stress
The role of AtAOX1a in maintaining the cellular redox balance during oxidative stress was revealed by monitoring the changes in pyridine nucleotide (NAD + and NADH) redox couple. In control, the cellular levels of NAD + , NADH, and the redox ratio of NAD + /NADH were similar in both pYES2 and pYE-S2AtAOX1a. Upon treatment with H 2 O 2 , the cellular NAD + levels decreased significantly in both pYES2 and pYES2AtAOX1a (Fig. 6A). In contrast, the decrease in cellular NADH levels was significant in pYES2AtAOX1a alone (Fig. 6B). Consequently, the cellular redox ratio of NAD + /NADH was maintained at much higher levels in pYES2AtAOX1a when compared with pYES2 in the presence of H 2 O 2 (Fig. 6C).
The responses of NAD + , NADH, and consequently NAD + /NADH were quite different in t-BOOH-treated samples as compared to H 2 O 2 treatment. In the presence of t-BOOH, both NAD + and NADH levels increased significantly, while the redox ratio of NAD + /NADH decreased drastically in pYES2A-tAOX1a when compared with pYES2 ( Fig. 6A-C).

Discussion
In higher plants, AOX is known to perform several mitochondrial and extramitochondrial functions, viz: (a) alleviation of reactive oxygen and nitrogen species, and cell death [16, [55][56][57], (b) preventing over-reduction of chloroplastic/mitochondrial electron transport carriers, particularly plastoquinone or UQ [13], (c) maintenance of cellular redox and carbon balance [18,19,58], (d) modulation of cellular energy level [59], and (e) optimization of photosynthesis during a wide range of biotic and abiotic stresses [18,19,60,61]. The role of AOX in alleviating ROS levels and oxidative stress is not only confined to plants but was also revealed in several nonphotosynthetic organisms including fungi, protists, bacteria, and human cells [29,39,62,63]. These observations suggest that engineering of AOX into such species which are deficient in AOX may help them to cope up against various biotic and abiotic stresses.
Saccharomyces cerevisiae lacks an AOX homolog [64]. Therefore, AtAOX1a was expressed in S. cerevisiae to validate its physiological function during oxidative stress ( Fig. 2A). It is well known that any restriction of electron flow through the COX pathway or exposure to oxidative stress leads to an induction of AOX in plants and fungi [21,27,60,65]. Corroborating with these studies, restriction of electron transport through the COX pathway by KCN caused a significant reduction in the total respiratory rates of pYES2 and pYES2AtAOX1a. However, due to an AOX catalyzed respiration, pYES2AtAOX1a showed higher respiratory rates compared to pYES2. While the SHAM-insensitive respiration in pYES2 indicates the absence of AOX-catalyzed respiration, SHAM or PGsensitive respiration in pYES2AtAOX1a confirms the functional expression of AtAOX1a in yeast (Fig. 2B) [29,39]. Any increase in the respiratory activity is known to increase the chronological and replicative lifespan of yeast [66]. Also, the recovery in the growth curve assays and a rise in the total respiratory rates of pYES2AtAOX1a in the presence of KCN reveal the significance of AOX-catalyzed respiration in the maintenance of yeast cell growth (Fig. 2B-D).
ROS production is a common phenomenon in cells, which occurs during aerobic respiration or in response to several biotic or abiotic stresses. But, excessive ROS production leads to oxidative stress [67][68][69]. Yeast cells show a range of responses depending on the concentration of cellular ROS. At very low levels of ROS, the cells try to adapt themselves, while at higher levels of ROS, the cells activate their antioxidant defense system mediated by Yap1p and Msn2,4p transcription factors [70]. Beyond this, ROS might arrest the cell cycle leading to apoptosis [71,72]. In the present study, the higher levels of cellular ROS induced by KCN, H 2 O 2 , or t-BOOH in pYES2 were positively correlated with cell death and negatively correlated with growth recovery. In contrast, the lower levels of ROS, better survival rate, and growth recovery recorded under oxidizing environment in pYES2AtAOX1a indicate the importance of AOX catalyzed respiration in mitigating the cellular ROS production (Figs 3 and 4A,B).
Redox homeostasis is a basic requirement to maintain the cellular metabolism and ROS, particularly during aging [72,73]. Accumulation of NADH decreases the Sir2 activity, which is essential for chromatin silencing and extension of life span. Thus, any increase in the redox ratio of NAD + /NADH extended the chronological as well as replicating life span of yeast cells [74,75]. The pYES2AtAOX1a showed an increase in the NAD + /NADH ratio when compared with pYES2 upon treatment with H 2 O 2 . In contrast, pYES2AtAOX1a maintained the cellular redox homeostasis by minimizing the redox ratio of NAD + / NADH raised by t-BOOH. These results elucidate the importance of AtAOX1a in the maintenance of cellular redox homeostasis to increase the life span as evident by cell survival rate of yeast (Figs 4A,B and 6C).
Furthermore, the sulphydryl (-SH) group plays a critical role in proper functioning of several of the enzymes, transcription factors, and membrane proteins, which in turn play a significant role in maintaining the cellular redox homeostasis [73]. During oxidative stress, cysteine sulfhydryl residues are oxidized to disulfide bonds, thereby leading to a loss in protein activity. Small heat-stable oxidoreductases, glutaredoxins, and thioredoxins catalyze the reduction of disulfides to thiols using thiolated cysteine residues present in the active sites [73,76,77]. A few studies reported the role of glutaredoxins and thioredoxins in supplying reducing equivalents to the regulatory sulfhydryl/disulfide system of AOX to activate it, which in turn play a role in preventing the over-reduction of mitochondrial electron transport carriers and thereby ROS generation [78][79][80]. In the present study, a several fold increase in the transcript levels of GPX2 and TSA2 in pYES2 and their down-regulation in pYES2AtAOX1a in the presence of t-BOOH and/or H 2 O 2 suggests the role of AOX1a in regulating the expression of these antioxidant enzymes, which play an important role in the detoxification of ROS and the maintenance of cellular redox balance (Figs 3, 5C,D and 6C).  The results from the present study suggest that transformation of AtAOX1a introduced AOXcatalyzed respiration in S. cerevisiae, which in turn mitigated ROS generation by regulating GPX2 and TSA2 to maintain cellular redox homeostasis and better cell survival rate during oxidative stress.

Supporting information
Additional supporting information may be found in the online version of this article at the publisher's web site: Fig. S1 MALDI-TOF-TOF mass spectrum of trypsin digested purified AtAOX1a protein between 500 and 5000 m/z.