Autophagy Is Involved in the Viability of Overexpressing Thioredoxin o1 Tobacco BY-2 Cells under Oxidative Conditions

Autophagy is an essential process for the degradation of non-useful components, although the mechanism involved in its regulation is less known in plants than in animal systems. Redox regulation of autophagy components is emerging as a possible key mechanism with thioredoxins (TRXs) proposed as involved candidates. In this work, using overexpressing PsTRXo1 tobacco cells (OEX), which present higher viability than non-overexpressing cells after H2O2 treatment, we examine the functional interaction of autophagy and PsTRXo1 in a collaborative response. OEX cells present higher gene expression of the ATG (Autophagy related) marker ATG4 and higher protein content of ATG4, ATG8, and lipidated ATG8 as well as higher ATG4 activity than control cells, supporting the involvement of autophagy in their response to H2O2. In this oxidative situation, autophagy occurs in OEX cells as is evident from an accumulation of autolysosomes and ATG8 immunolocalization when the E-64d autophagy inhibitor is used. Interestingly, cell viability decreases in the presence of the inhibitor, pointing to autophagy as being involved in cell survival. The in vitro interaction of ATG4 and PsTRXo1 proteins is confirmed by dot-blot and co-immunoprecipitation assays as well as the redox regulation of ATG4 activity by PsTRXo1. These findings extend the role of TRXs in mediating the redox regulation of the autophagy process in plant cells.


Introduction
Autophagy is a degradative process in which damaged or non-useful cellular components, including proteins, aggregates, and whole organelles, are degraded in a lytic cell compartment for recycling or elimination. This process has been observed as contributing to antioxidant defenses [1,2] and, similar to that described for other eucaryotes, it is activated and regulated by cells to cope with biotic and abiotic stress [3]. While this process has been less studied in plants than in animal systems, basal autophagy seems to be essential for growth and development in both cases [4,5]. More than thirty proteins participate in the autophagy process, which is regulated by different signaling and stress detection pathways, nutrient availability, metabolic activity, and environmental changes [2,6,7]. Three types of autophagy have been described in animals: Microautophagy, autophagy mediated by chaperones, and the most studied, macroautophagy [8]. In plants, micro-and macroautophagy have been described, as has mega-autophagy, an extreme process occurring in programmed cell death during development and pathogen attack, when the tonoplast releases hydrolases into the cytoplasm after its breakdown [9]. During macroautophagy, specialized double-membrane vesicles, termed autophagosomes, deliver the cytoplasmic content to lytic compartments such as lysosomes or vacuoles in animal and plant systems, respectively, for their hydrolytic degradation [10]. In plants, autophagosomes are produced de novo and the external membrane fuses to the tonoplast, delivering the content as a single-membrane autophagic vesicle in the lumen of the vacuole [3]. The autophagosomes can acquire hydrolytic enzymes from the vacuole to generate vesicles similar to the socalled autolysosomes in animals [5]. In both animal and plant cells, autophagy involves several conserved specific proteins termed ATG (Autophagy related), forming complexes and ultimately generating the double-membrane autophagosome [3]. Among core ATG proteins, ATG8 plays a key role and is associated with the membrane of the autophagosome by covalent binding to phosphatidylethanolamine (PE) at a glycine residue exposed at the C terminus of ATG8. Prior to the conjugation of ATG8 and PE, the cysteine protease ATG4 processes the C terminus of nascent ATG8 to expose the conserved glycine residue (priming activity). Then ATG8 is activated via the ubiquitin-conjugating E1-like enzyme, ATG7, and the E2-like enzyme, ATG3, to finally conjugate to PE. ATG4 participates again with a deconjugating activity that cleaves the PE residue from ATG8, thus allowing ATG8 recycling [11]. In silico analysis of 18 available plant genomes identified a large number of ATG4 and ATG8 genes that are conserved among plant lineages. While Chlamydomonas reinhardtii presents one ATG4 gene and one ATG8 gene, Arabidopsis thaliana presents two ATG4 and nine ATG8 and Nicotiana tabacum has one ATG4 and five ATG8 [12,13].
Thioredoxins are small oxidoreductase proteins involved in the redox regulation of specific target proteins, which they reduce from their oxidized forms, being able to regulate their structure and function [14][15][16][17]. Oxidative stress inhibits the protein degradation pathway and autolysosome maturation, provoking apoptosis in neuroblastoma cells, and it has been established that the thioredoxin system plays a central role in inhibiting apoptosis in human cancer cells [18]. Moreover, a more recent role for TRX and thioredoxin reductase (TR) in the initiation of autophagy has been suggested. It is known that this system plays a role in early protective autophagy while its deficiency promotes apoptosis, a situation that can be avoided using N-acetylcysteine, promoting cell viability [19]. TR is also involved in the stability of a key component in the autophagy process: Target of Rapamycin (TOR) kinase [20]. Regarding, ATG4 is the only ATG that has been shown to be redox regulated. Specifically, redox regulation of ATG4 protease by thioredoxin has been demonstrated in animal systems, Chlamydomonas and Sacharomyces cerevisiae [21,22], but not in plants. Among TRXs, PsTRXo1 is located in the plant mitochondria and nucleus [23] and is involved in the response to oxidative stress induced under adverse situations, such as salinity, high temperature, drought, and H 2 O 2 treatment, exerting its function on specific target proteins in both compartments [17,[24][25][26][27]. We have previously described the involvement of the protein PsTRXo1 in the response to oxidative stress induced by H 2 O 2 in Nicotiana tabacum L. cv. Bright Yellow 2 (BY-2) cell cultures overexpressing (OEX) this protein. Interestingly, OEX lines maintained a higher level of cellular viability than non-overexpressing cells, and in this scenario, delayed cell death, the antioxidant system and redox homeostasis collaborate with TRXo1 in the resistance to H 2 O 2 [26,28]. However, it is unclear whether the autophagy process also participates in this increase in cell viability and, if this is so, whether PsTRXo1 also plays a part as it has been described in animal systems. The role of autophagy in promoting plant viability during abiotic stress has been demonstrated [29], although this process is also involved in promoting programmed cell death (PCD) in barley during microspore embryogenesis [30]. This would indicate that, as has been shown during biotic stress, the dual function of autophagy depends on the type of stress, plant organ, and developmental stage [31]. In this work, we show that the overexpression of PsTRXo1 participates in the response of TBY-2 cells to H 2 O 2 treatment favoring an autophagy process, which plays a role in facilitating the survival of OEX cells.
Our data indicate that TRXo1 regulates the expression and activity of target proteins such as ATG4 protease and the expression and protein level of ATG8, both of which are involved in the autophagy process. In fact, in vitro assays demonstrate the redox regulation of ATG4 and its interaction with the TRXo1 protein, suggesting that similar redox regulation of ATG4 by TRX may occur in vivo in the overexpressing TRXo1 cells.

Transformation of Tobacco BY-2 Cells
Stable PsTRXo1 over-expressing (OEX) lines of Nicotiana tabacum 'Bright Yellow' (TBY-2) suspension cells were generated as previously reported [28]. Two of these independent OEX lines (OEX1 and OEX2, containing GFP as a transformation marker) and a control line (without TRXo1 overexpression but containing GFP as a transformation marker) were used. The suspension of tobacco cells was routinely propagated and cultured at 26 • C and a stationary culture of 7 days of growth was diluted 4:100 (v/v) in fresh medium.

Cell Viability
Cell viability was calculated as the percentage of cells that did not stain with trypan blue [28]. An aliquot of cell suspension was transferred to a test tube containing 50 % (v/v) trypan blue solution. After 5-10 min, the trypan blue cell suspension mixture was transferred to a microscope slide and viable (unstained) and non-viable (blue-stained) cells were counted. For each sample, 5000 cells were scored, and four independent experiments were performed. Cells were visualized using a Leica DM6 microscope equipped with a digital MC 190 HD camera.

H 2 O 2 Treatment
Cells from the three lines at day 5 of growth in the exponential phase were exposed to 35 mM H 2 O 2 (final concentration) as previously described [28]. At the indicated time points, cell aliquots were collected for analysis either by filtration on Whatman 3M paper or by centrifugation at 10,000× g for 10 min at 25 • C.

Incubation in the Presence of Autophagy Inhibitors
Two milliliters of tobacco culture at 5 days of growth were placed in ELISA plates and H 2 O 2 was added at a final concentration of 35 mM in the absence or presence of either 10 µM E-64d (Aloxistatin, a cysteine protease inhibitor, Sigma, Merck, Darmstadt, Germany) (dissolved in methanol) or 10 µM E-64d plus 5 mM 3-MA (3-methyl adenine, an inhibitor of class III phosphatidylinositol 3-kinase PI3K, Sigma, Merck, Darmstadt, Germany) (dissolved in MilliQ water), as described in Takasutka et al. [32]. As a control for E-64d treatment, 1% methanol (v/v) was added to the cells. The cells were further cultured on a rotary shaker at 118 rpm at 26 • C in darkness and cell viability and autophagy analysis were carried out at the indicated times.

Protein Extracts
Suspended cells were obtained with vacuum Kitasato flasks, weighed, and immediately frozen with liquid N 2 and stored at −80 • C until use. The samples (3 g) were homogenized with a pre-cooled mortar and pestle in 3 mL of extraction buffer containing 50 mM HEPES-KOH, 5 mM DTT (1,4-dithiothreitol), 0.5 mM PMSF (phenylmethanesulfonyl fluoride), 1 mM EDTA (ethylenedinitrilotetraacetic acid), and 0.1% Triton X-100 (v/v), pH 7.4. The mixture was sonicated with 30 pulses in a SONICLAB at 70% amplitude and finally centrifuged at 16,000× g at 4 • C for 15 min. The supernatant was obtained, and the protein content was estimated by the Bradford [33] method.

Quantitative Real-Time PCR
The cells were sedimented by centrifugation and, after removing the residual medium, the fresh weight was determined, and the cells were frozen in liquid N 2 and stored at −80 • C until analysis. RNA was isolated from cell sediment using the RNeasy Plant Mini Kit (Qiagen, Hilde, Germany) following the manufacturer's protocol. cDNA was obtained from RNA by using the QuantiTect ® Reverse Transcription Kit (Qiagen, Hilde, Germany) according to the manufacturer's instructions. RNA and cDNA were quantified in an ND-1000 spectrophotometer (NanoDrop, Wilmington, NC, USA). qPCR was performed on a QuantStudio 5 Real Time PCR System (Applied Biosystems by Thermo Fisher Scien-tific, Bleiswijk, The Netherlands) using Power SYBR Green V PCR Master Mix (Applied Biosystems by Thermo Fisher, Madrid Spain) in the CAID Service of the University of Murcia. Each reaction was performed in triplicate using the following conditions: 2 min at 50 • C, 10 min activation at 95 • C, and 40 cycles of amplification (15 s at 95 • C; 1 min at 60 • C), followed by the melting curve. The relative quantity was calculated using the comparative Ct method [34]. The specific primers were designed using Primer Express v 2.0 (Thermo Fisher, Madrid, Spain) using the sequence from the data bank NtATG4 (accession number KR336561), listed in Supplementary Table S1. The expression of the Actin 8 (ACT8 GQ339768) gene was used to normalize the data.

SDS-PAGE and Western Blot of Autophagy Markers
For Western blot analysis, SDS-PAGE was carried out on AnykD (BioRad, Madrid, Spain) polyacrylamide gels, and 40 µg of protein extracts were transferred to nitrocellulose membranes using a semi-dry Trans-Blot cell (BioRad, Madrid, Spain). Membranes were used for cross-recognition assays using polyclonal antibodies anti-CrATG4 (Agrisera, Vännäs, Sweden) or anti-CrATG8A (Agrisera, Vännäs, Sweden). For immunodetection in membranes, a goat anti-rabbit IgG-horseradish peroxidase conjugate (Santa Cruz Biotechnology, Heidelberg, Germany) was used as the secondary antibody, together with an enhanced chemiluminescence kit (Supersignal TM West Picoplus, Fisher Scientific, Madrid, Spain). Densitometry of the different bands was performed using an image analyzer (Amersham Imager 600, GE Healthcare, Barcelona, Spain) and the Amersham ImageQuanTL 8.1 Program (Cytiva, Barcelona, Spain). Ponceau staining was used to correct the loading control in each line.

ATG4 Cleavage Assay
ATG4 activity was measured by electrophoretic separation of the cleaved ATG8 substrate (processed pATG8) containing a C-terminal tag after the cleavage site. For this, a recombinant human GABARAP protein (Fc Chimera, ATG8-Fc, Abcam, Cambridge, UK) also containing an N-terminal 6His-tag was used, as previously described [35]. Cleaved ATG8 was detected by Western blot using an anti-His antibody as described below.
To analyze the effect of reducing and oxidizing conditions on the in vitro HsATG4 activity, a typical reaction mixture contained 0.22 µg of recombinant HsATG4B treated under reducing (15 µM DTT for 5 min at 25 • C) or oxidizing (1 mM H 2 O 2 or 3 mM H 2 O 2 for 5 min at 25 • C) conditions in TBSK buffer (50 mM Tris-HCl, 140 mM NaCl, 30 mM KCl, pH 8.0) and, where indicated, under reducing conditions followed by oxidizing ones. The reaction mixtures were then incubated with 0.48 µg ATG8-Fc for 45 min at 30 • C and stopped by the addition of reductant-free Laemmli sample buffer followed by incubation for 5 min at 95 • C. Proteins were resolved on TGX 4-20% SDS-PAGE gels (BioRad, Madrid, Spain). Two negative controls were performed using only HsATG4B in the absence of ATG8-Fc, and with ATG8-Fc in the absence of HsATG4B. Western blot analysis using the primary antibody monoclonal anti-6His (Agrisera, Vännäs, Sweden, 1:5000) and anti-mouse IgG HRP secondary antibody (Abcam, Cambridge, UK 1: 5000) followed by immunodetection using a chemiluminescence kit (Supersignal TM West Picoplus, Fisher Scientific, Madrid, Spain) according to the manufacturer's instructions allowed the detection of the reaction product of the ATG4 activity as the processed ATG8 (pATG8).
To analyze the HsATG4 in vitro activity in the presence of the thioredoxin system, HsATG4B previously reduced by 15 µM DTT was treated with 1 mM H 2 O 2 as described above. This oxidized HsATG4B was then reduced using 2.5 µg PsTRXo1 (or a mutated version mutPsTRXo1C37S [23], 1.5 µM AtNTRA (NADPH-dependent thioredoxin reductase A), and 0.04 mM NADPH in TBSK buffer [26]. Several control reactions were performed in the absence of any of these components or using 10 µM DTT instead of the NADPH/TRXo1/NTRA system. The different reaction mixtures were then incubated with 0.48 µg ATG8-Fc for 45 min at 30 ºC before adding Laemmli sample buffer and continuing SDS-PAGE as described above. The cleaved product of ATG8-Fc was identified after Western blot as described.
For the in vivo ATG4 activity, we followed a previously described method [35], whereby 50 µg of TBY-2 cellular protein extracts were incubated with recombinant ATG8-Fc (0.6 µg) in TBSK reaction buffer and 100 µM DTT at 30 • C for 45 min. The reaction wasstopped by adding sample buffer (reductant-free Laemmli buffer) and 5 min of boiling, after which the samples were resolved on AnykD (BioRad, Madrid, Spain) gels by SDS-PAGE and subsequently analyzed by Western blot using the anti-6His antibody as described above. A negative control was performed on a cellular extract (OEX1 line) without ATG8-Fc substrate to reveal non-specific signals.

Visualization of Autolysosomes
TBY-2 suspension tobacco cells (200 µL) were centrifuged at 4000× g for 5 min and the pellet was resuspended in 200 µL of 100 mM potassium phosphate buffer (PK) pH 6.5 containing 1 µM of Lysotracker Deep Red (Invitrogen, Thermo Fisher, Madrid, Spain) as previously described [36]. The mixture was incubated in the dark and at room temperature for 6 min. Subsequently, the preparation was centrifuged at 1000× g for 1 min, the supernatant was discarded, and the pellet was washed three times with PK buffer, pH 6.5 at 1000× g for one minute, and the final precipitate was resuspended in 150 µL of the same buffer. Fifty microliters were added to the slide and the structures were visualized in a Leica DM6 motorized fluorescence microscope equipped with Nomarski optic and a digital Hamamatsu C11440 camera. For the detection of autolysosomes, the preparations were observed using a band pass (BP) 545-580, DM 600, and a BA filter of 610. At least ten fields (approximately 20 cells per field) were evaluated in each of the three independent experiments carried out. Image analysis was performed using ImageJ software (Softonic, NIH, Bethesda, MD, USA).

Immunolocalization of ATG8
Fluorescence microscopy was used for the immunolocalization of ATG8. TBY-2 cells were fixed and stained as described by Voitsekhovskaja et al. [36] with some modifications. Cells were fixed with 1% glutaraldehyde in MS medium and treated with a mixture of cellulose (5 U/mL, Onozuka, Duchefa Biochemie, Haarlem, The Netherlands ), 10% hemicellulose (Macerozyme R-10, Duchefa Biochemie, Haarlem, The Netherlands), and 5 U/mL pectinase (Sigma, Merck, Darmstadt, Germany) in osmotic media. After washing and treatment with 0.3% Triton X-100 and blocking with 5% BSA, cells were incubated with the primary antibody anti-ATG8 (1:1000, Agrisera, Vännäs, Sweden) for 1 h and then with the Alexa-568-conjugated secondary antibody (1:100, Molecular Probes, Thermo Fisher, Madrid, Spain) for 1 h before washing four times. Cells were observed with a Leica DM6 motorized fluorescence microscope equipped with a digital camera with an excitation wavelength of 579 nm and emission of 603 nm, using a rhodamine (RHO) filter and a 100X/1.30 oil objective. At least ten fields (approximately 20 cells per field) were evaluated in each of the three independent experiments carried out. When indicated, images were treated with Huygens Essential Microscopy Analysis software (SVI, Stockton, CA, USA) at the University of Murcia (Scientific Image Treatment's Service) for better visualization of the vesicles. Image analysis was performed using ImageJ software (Softonic, NIH, Bethesda, MD, USA).

Protein Dot-Blot
Recombinant proteins PsTRXo1 (1.5 µg, [24], HsATG4B (Abcam, 2.3 µg), and bovine serum albumin (BSA) (1.5 µg, Sigma, Merck, Darmstadt, Germany) as negative control were spotted on 6 pieces of the membrane (2 membranes for each protein). The membranes were incubated in TBS for 5 min and then blocked in TBST buffer (25 mM Tris-HCl pH 7.5 containing 1% (w/v) BSA and 0.1% (v/v) Tween-20 (Sigma, Merck, Darmstadt, Germany) for one hour at room temperature. Then, the membranes were overlaid with 0.2 µg of recombinant PsTRXo1 (membranes A) or 0.1 µg of recombinant HsATG4B (membranes B) in TBST and incubated for 2 h at room temperature. The A membranes were incubated with the polyclonal anti-PsTRXo1 (1:1000) [24] and B membranes with monoclonal anti-CrATG4B (Agrisera, Vännäs, Sweden, 1:5000) diluted v/v in TBST, overnight at 4 • C with stirring. Subsequently, the membranes were washed five times with TBST for 5 min at room temperature and incubated with rabbit Santa Cruz anti-HRP 1:30,000 (v/v in TBST) for TRXo1 and anti-HRP mouse 1:2000 (v/v in TBST) for ATG4B. Finally, the membranes were washed five times with TBST for 5 min at room temperature and once again with TBS. Immunodetection was performed using a chemiluminescence kit (Supersignal TM West Picoplus, Fisher Scientific, Madrid, Spain) according to the manufacturer's instructions.

Co-Immunoprecipitation PsTRXo1/His-HsATG4
Co-immunoprecipitation of the PsTRXo1 and HisHsATG4 proteins was performed using the commercial µMACS Anti-His MicroBeads kit (MACS Miltenyi Biotec, Madrid, Spain). To this end, recombinant proteins capable of interacting (4 µg His-HsATG4B and 8 µg PsTRXo1) were diluted in 500 µL of "lysis buffer" provided by the kit. A 50 µL fraction was kept for use as a loading control containing both proteins (Input in the gel). Fifty microliters of Anti-His magnetic µbeads were added and the protein mixture was incubated for 2 h at 4 • C under mild stirring.
The columns provided by the kit were placed according to the manufacturer's instructions on the designed magnetic support and they were prepared by passing 200 µL of "lysis buffer" from the kit. Once the incubation time between the proteins and the Anti-His µbeads was over, the solution was deposited into the columns and washed 4 times with 200 µL of "wash buffer 1" and then rinsed with 100 µL of "wash buffer 2" (fraction named as Wash in the gel). Then, 20 µL of "elution buffer" preheated at 95 • C was added to the column and incubated for 5 min at room temperature Fifty microliters of preheated elution buffer at 95 • C was added to the column and the eluate was considered as coimmunoprecipitated (CoIP) PsTRXo1/His-HsATG4B proteins. Samples were analyzed by Western blot after SDS-PAGE using anti-His. Then the membrane was incubated in TBST at room temperature for 1 h and revealed for the presence of PsTRXo1 using anti-PsTRXo1.

Statistical Analysis
Each experiment was repeated at least three times with three biological replicates per treatment for each genotype. The figures show data from a typical experiment if not otherwise indicated. Values presented are mean ± standard error. Different letters indicate significant differences (p < 0.05) among genotypes in each condition according to Tukey's-b test using SPSS software (IBMVR SPSSVR Statistics 19, New York, USA). The asterisks indicate significant differences of each genotype after the H 2 O 2 treatment compared to the control condition using Student's t-test at p < 0.05 (*), p < 0.005 (**), or p < 0.001 (***).

Overexpressing PsTRXo1 Cells Present High Viability and Elevated Levels of Autophagy Markers
Overexpression of PsTRXo1 in tobacco BY-2 cells increases viability after oxidative treatment with 35 mM H 2 O 2, as previously reported by our group, due to increased antioxidant status and delayed cell death [28]. In the present work, we analyze whether autophagy might also be collaborating with those processes to increase cell viability. First of all, we confirmed that similar to that found in our previous work [28], a control line and two overexpressing Pstrxo1 lines (OEX1 and OEX2) presented similar viability (close to 100%) during culture growth up to 8 days ( Figure 1A). We also found that, as before, a 35 mM H 2 O 2 treatment at day 5 of growth provoked greater mortality in the nonoverexpressing cells (control line)-around 30% viability after 24 h ( Figure 1B) compared to the 70% observed in the two OEX lines. The fresh weight was similar in the three lines during growth in control conditions ( Figure 1C), and this parameter decreased sharply in non-overexpressing cells while OEX lines maintained a high fresh weight after the H 2 O 2 treatment ( Figure 1D). The difference observed in viability 24 h after oxidative treatment among cell lines led us to analyze how different autophagy markers were affected by the overexpression of PsTRXo1 and by the stress applied (day 6 of growth).
The analysis of ATG4 gene expression under control conditions ( Figure 2A) revealed that the transformation with PsTRXo1 increased the expression of ATG4. The H 2 O 2 treatment affected only the non-overexpressing cells (Cont-T), decreasing ATG4 gene expression, showing that the overexpression of PsTRXo1 allows the maintenance of high levels of ATG4 in OEX cells.
Western blot analysis of the ATG4 protein revealed the presence of two bands recognized by the antibody at around the described molecular weight of the isoforms X1 and X2 of the tobacco proteins, 53 and 57 kDa (Uniprot database: XP_016513796 and XP_016513797) ( Figure 2B). Western blot analysis of the ATG8 content in non-overexpressing cells showed two bands of different molecular weights ( Figure 2C) corresponding to the free cytosolic forms, namely ATG8 around 13 kDa and below it the lipidated form ATG8-PE (PE: Phosphatidylethanolamine) as previously described [37]. Overexpression of PsTRXo1 increased the level of both proteins, ATG8 (non-lipidated and lipidated forms) and ATG4 (left panels). In fact, ATG4 was detected only in OEX cells and not in the control line under non-oxidative conditions. The H 2 O 2 treatment increased ATG8 content particularly in the OEX lines (more than double) and especially the lipidated ATG8-PE form ( Figure 2C). The oxidative treatment induced the accumulation of both ATG4 isoforms in the Control (Cont-T) line but more intensively in the OEX-T lines, which presented a higher protein content at 53 kDa ( Figure 2B). These results show that both the overexpression of PsTRXo1 and H 2 O 2 treatment provoked an increase in ATG8 and ATG4 autophagy markers, with OEX-treated cells being the ones presenting higher levels than non-overexpressing cells.
To check whether changes in expression and protein levels correlated with activity levels, endogenous ATG4 activity was assayed in the three cell lines in the presence of DTT. In the absence of the H 2 O 2 treatment and parallel to the protein content, the control line presented low ATG4 activity while overexpressing Pstrxo1 cells showed high activity, as seen from the presence of a cleavage product at around 18 kDa (Figure 3, left pannel). The activity of recombinant HsATG4 was measured as a positive control. After the H 2 O 2 treatment, similar to that found for protein contents, OEX lines maintained higher ATG4 activity, being around double than that of the control line, which showed a 6-fold increase.
A non-specific band appeared around 24 kDa with higher intensity in the overexpressing lines (asterisk in Figure 3) as resulting from the incubation of cells without the ATG8-Fc substrate ( Figure S1).

Autophagy Is Related to Cell Viability
After demonstrating that the overexpression of PsTRXo1 induced an increase in cell viability after H 2 O 2 treatment with the cells presenting higher autophagy markers, such as ATG4 and ATG8, than non-overexpressing ones, we wondered whether this increased viability and higher levels of autophagy markers were accompanied by activation of the autophagy process. In this way, autophagy could act as a protective mechanism probably eliminating damaged components in response to the oxidative stress situation, thus increasing viability. In order to analyze whether autophagy was involved in the higher viability of the OEX lines under oxidative conditions, in addition to the treatments used in Figure 1, we measured viability in the presence of two autophagy inhibitors. E-64d, a Cys-protease inhibitor that functions by suppressing lysosomal proteases and results in the accumulation of autophagosome-and (auto)lysosome-like structures and 3-methyladenine (3-MA), which reverts the E-64d effect in tobacco cells subjected to induced autophagy, were investigated [5,32].
Similar to Figure 1, we first measured the viability of the control cells in the absence of, or 24 h after, H 2 O 2 oxidative treatment (+H 2 O 2 ). The H 2 O 2 treatment provoked a decrease in viability from 96% to 20% in the control line ( Figure 4A) and E-64d (+E-64d) alone did not affect the viability. When both H 2 O 2 and E-64d alone and in combination with 3-MA were applied to the cell culture, no effect was observed related to the + H 2 O 2 treatment. These results indicate that autophagy is not involved in the viability of control cells under oxidative conditions. A different effect was observed in OEX1 cells, which showed a decrease in viability from 99% to 65% with the oxidative treatment ( Figure 4B). Again, the inhibitor E-64d alone did not affect their viability. Interestingly, when H 2 O 2 plus E-64d were applied to the OEX line, a marked decrease in cell viability was observed (around 33%), which was reverted in the presence of 3-MA (last picture in Figure 4B), again leading to high cell viability (76%), indicating an autophagy-dependent cell survival situation. We then investigated whether the effect of the two inhibitors was observed in the OEX1 line 48 h after H 2 O 2 treatment, and similar results were obtained ( Figure S2), so the effect of inhibitors lasted for at least this time. Similar results were obtained with the OEX2 line (data not shown). We did not measure the effect of inhibitors in the control line 48 h after oxidative treatment because they were not alive at this time. We also determined that 3-MA did not affect OEX1 line viability when applied in the absence of E-64d ( Figure S3) in both the absence or presence of H 2 O 2 24 h after oxidative treatment, while it decreased viability around 5% 48 h after.
All these experiments point to a relationship between autophagy and viability in the cell response to the oxidative situation occurring 24 h/48 h after H 2 O 2 treatment, only when PsTRXo1 is over-expressed.   Values represent mean ± standard error of four independent experiments. Different letters indicate significant differences (p < 0.05) among treatments according to Tukey's test.
All these experiments point to a relationship between autophagy and viability in the cell response to the oxidative situation occurring 24 h/48 h after H2O2 treatment, only when PsTRXo1 is over-expressed.

Overexpressing Pstrxo1 Cells Present Lytic Vesicles 24 h and 48 h after Oxidative Treatment
To further our knowledge of the autophagy process taking place in the overexpressing cells under the oxidative H2O2 treatment, we decided to observe the presence of autophagic structures. Autophagy can be analyzed using the acidotropic fluorescent dye Lysotracker Deep Red (LDR), which labels autolysosomes. E-64d is an inhibitor that blocks the transfer of autolysosomes to the vacuole, allowing their accumulation, blocking the autophagy flux, and enabling the structures to be stained by LDR and visualized.
The analysis of the presence of vesicles was carried out in control and OEX1 lines ( Figure 5A,B, respectively) 24 h after the H2O2 treatment in the presence of the inhibitor E-64d using a fluorescence microscope with a Nomarski optic (DIC). An accumulation of vesicles visualized in both DIC-and LDR-labelled cells was observed mainly in OEX1 cells Values represent mean ± standard error of four independent experiments. Different letters indicate significant differences (p < 0.05) among treatments according to Tukey's test.

Overexpressing Pstrxo1 Cells Present Lytic Vesicles 24 h and 48 h after Oxidative Treatment
To further our knowledge of the autophagy process taking place in the overexpressing cells under the oxidative H 2 O 2 treatment, we decided to observe the presence of autophagic structures. Autophagy can be analyzed using the acidotropic fluorescent dye Lysotracker Deep Red (LDR), which labels autolysosomes. E-64d is an inhibitor that blocks the transfer of autolysosomes to the vacuole, allowing their accumulation, blocking the autophagy flux, and enabling the structures to be stained by LDR and visualized.
The analysis of the presence of vesicles was carried out in control and OEX1 lines ( Figure 5A,B, respectively) 24 h after the H 2 O 2 treatment in the presence of the inhibitor E-64d using a fluorescence microscope with a Nomarski optic (DIC). An accumulation of vesicles visualized in both DIC-and LDR-labelled cells was observed mainly in OEX1 cells when compared to control cells after the H 2 O 2 + E-64d treatment (third column in Figure 5). It seems that some autophagosomes are constitutively accumulated in OEX lines due to the detection of LDR signals in cells cultured in the absence or presence of H 2 O 2 without the inhibitor (the first and second columns in Figure 5) or when the compound 3-MA was also present (last column in Figure 5). This result indicated that the E-64d inhibitor principally affected OEX1 cells blocking the autophagy process and inducing the accumulation of vesicles. The accumulation was observed even 48 h after the treatments in OEX1 cells ( Figure S4). All these results demonstrated the occurrence of an autophagy flux evident by the accumulation of autolysosomes in the transformed PsTRXo1 cells subjected to the H 2 O 2 treatment.

Overexpressing Pstrxo1 Cells Accumulate ATG8 Protein Detected by Immunofluorescence 24 h after Oxidative Treatment
In order to corroborate the existence of an autophagy flux, the antiCrATG8 antibody was used to examine the presence of this protein by immunofluorescence (IF) in the OEX1 line treated with H 2 O 2 in the absence and presence of E-64d and in the presence of E-64d plus 3-MA. A high IF signal was detected in cells treated with H 2 O 2 plus E-64d compared with that seen in cells treated with the oxidant in the absence of the inhibitor (Figure 6), corroborating the existence of autophagosomes containing the ATG8 protein in these H 2 O 2treated OEX1 cells. This ATG8 accumulation was reverted by using 3-MA in combination with E-64d. Similar results were obtained in the OEX2 line (data not shown).  The punctate signals per cell were quantified in control and two OEX lines as the measurement of the autophagy flux, and the values are presented in Figure 7. After analyzing 20 cells per genotype and treatment, only the OEX lines presented a high accumulation of ATG8 24 after the H 2 O 2 + E-64d treatment, with the values being three times higher in these lines than in the H 2 O 2 -treated cells.

Autophagy Is Evident in OEX Cells and Not in Control Cells 14 h after Oxidative Treatment
Based on the observed low viability of the control non-overexpressing line 24 h after H 2 O 2 treatment and in order to assess the possible occurrence of an increase in autophagy as a short-term response in this control line, we attempted to ascertain the point these cells presented higher viability than at 24 h, choosing 14 h post H 2 O 2 treatment. At this time, the control line presented 40% viability and the OEX lines around 80% ( Figure 8A). Using E-64d, no significant effect on cell viability was observed in the control line while in the OEX lines, the viability decreased significantly to 60% ( Figure 8B). 3-MA did not affect viability in the control line after the oxidative treatment while it reverted the effect of E-64d in the OEX1 line. Similar results were obtained in the OEX2 line (data not shown). We also tested the effect of the inhibitors at 2 h after the H 2 O 2 treatment finding no effect on cell viability in any of the three lines ( Figure S5).
These results suggest the lack of autophagy in the control non-overexpressing Pstrxo1 line in response to the oxidative condition at 14 h. To corroborate this, we analyzed the presence of autolysosomes 14 h after H 2 O 2 treatment, in the absence and presence of the E-64d inhibitor. Similar to that observed for the effect of E-64d on cell viability, low-intensity fluorescence was accompanied by no autolysosome accumulation in the control line ( Figure 9A) while higher intensity was observed, and these structures mainly accumulated in E-64d-treated cells of the OEX line 14 h after H 2 O 2 treatment ( Figure 9B).

ATG4 and TRXo1 Interact In Vitro
One of the possible control mechanisms of the autophagy process may be the redox regulation by TRX of autophagy components such as certain ATG proteins, as previously reported for ATG4 in Chlamydomonas and yeast (described in the introduction section). To determine whether TRX could also regulate the ATG4 protein in plants, we first performed an experiment to analyze the possible interaction between both proteins using recombinant pea PsTRXo1 and human HsATG4B, and then studied any possible redox regulation in vitro of the ATG4 cleavage activity by the NTR/TRXo1 system.
To analyze the physical interaction of PsTRXo1 and HsATG4B, a protein dot blot experiment was carried out, spotting the proteins onto nitrocellulose membranes separately and also overlaying them separately. After extensive washing of the membrane, the presence of TRXo1 bound to spotted ATG4 was revealed by anti-PsTRXo1 (Figure 10 line A) and the presence of ATG4 bound to TRXo1 spotted on the filter was revealed by the anti-CreATG4 antibody (Figure 10 line B). As positive controls, the PsTRXo1 recombinant protein was spotted and revealed with anti-TRXo1 and HsATG4 was spotted and revealed with anti-ATG4 ( Figure 10 lines A, first spot and B, second spot). The interaction appeared to be specific, since no signal was detected in positions corresponding to equivalent amounts of BSA, spotted as negative controls (membranes in the last column). These results confirm the in vitro physical interaction of PsTRXo1 and HsATG4B proteins. To corroborate the result obtained in the dot-blot trap assay, a Co-immunoprecipitation (CoIP) of recombinant His-HsATG4B and PsTRXo1 proteins was performed followed by Western blot analysis of the presence of both proteins using Anti-His MicroBeads. The last wash of the magnetic particles before elution of the column was loaded on the first lane of the gel (membrane on the left in Figure 11) to check the absence of the proteins not retained before elution (Wash fraction). The third lane contained a fraction of the solution in which the proteins were incubated before passing through the column (Input), which demonstrated the presence of monomeric and dimeric His-HsATG4B protein (as revealed by anti-His). The last lane contained the proteins eluted from the microparticles (CoIP), which mainly revealed the presence of the monomeric HsATG4B protein. The same membrane was then subjected to mild stripping and incubated with anti-PsTRXo1, revealing the presence of the monomeric (around 12.5 kDa) form of this protein in the input fraction and also in the last lane containing the CoIP eluted from the magnetic microparticles (membrane on the right in Figure 11) together with several dimeric forms around 24-25 kDa. The presence of both proteins in the eluted CoIP fraction confirmed the interaction between HsATG4B and PsTRXo1 proteins. Due to the low amount of TRXo1 in the eluted CoIP fraction, the membrane was revealed with anti-PsTRXo1 over a long duration for the detection of the protein, which also allowed the detection of the dimeric HsATG4B in the input fraction (see whole membranes in Supplemental Figure S6).

ATG4 Activity Is Redox-Regulated
After confirming the interaction between HsATG4 and PsTRXo1 in vitro, we tried to ascertain whether the HsATG4 protein might be redox regulated by PsTRXo1 by monitoring the activity of the recombinant HsATG4B in the presence of PsTRXo1. Before this, we confirmed the redox regulation of the activity of the HsATG4 recombinant protein in the presence of oxidizing H 2 O 2 and reducing DTT treatments. The activity of the recombinant protein (HsATG4B His-tag of 47 kDa) was demonstrated by the 17 kDa peptide appearing as a result of the ATG8 cleavage ( Figure 12A, lane 2) confirming the functionality of the protein when reduced by DTT. In order to determine whether ATG4 activity is influenced by the redox environment, the protein was incubated in the presence of oxidizing (H 2 O 2 ) or reducing (DTT) compounds. Pretreatment of HsATG4B with 1 and 3 mM of H 2 O 2 ( Figure 12A, lanes 3 and 4) decreased the processing activity of ATG4 and resulted in the loss of most of the processed peptide, while DTT restored the activity ( Figure 12A, lanes 5 and 6). In fact, H 2 O 2 induced the oligomerization of the protein, while DTT disrupted it, allowing monomerization, which implied that ATG4 is inhibited by oxidation and activated by reduction both being fully reversible. A control experiment included the activity of individual recombinant His-tag proteins ATG4 and ATG8 after the DTT treatment ( Figure 12B). These results indicate that the activity is regulated by a redox post-translational modification, as previously described for yeast and Chlamydomonas ATG4.

ATG4 Activity Is Redox-Regulated by Thioredoxin o1 In Vitro
In order to test the ability to reduce ATG4 by the TRX system, including TRXo1, NADPH, and NTR, we used the recombinant proteins PsTRXo1 and AtNTRA, which were previously cloned, expressed in E. coli, and purified [26]. Reduced DTT-treated recombinant HsATG4B presented processing activity as seen from the 17 kDa cleavage product (Figure 13 lane 2). This activity decreased when DTT-treated HsATG4B was oxidized by 1 mM H 2 O 2 (lane 3), which was responsible for the oligomerization, as demonstrated by the appearance of a high-molecular-weight signal in Figure 13. Instead of DTT, we then used the TRX system to reduce H 2 O 2 -treated ATG4B. The activity was monitored with NADPH/NTR (without TRXo1) (lane 4), and the activity observed was clearly higher in the presence of TRXo1 (lane 5). Very low processing activity, similar to that observed with NADPH/NTR only, was observed when a catalytic cysteine mutant, TRXo1C37S (MutTRXo1), was used (lane 6). ATG4 activity was lower when TRX+DTT or DTT alone were used as reductants (lanes 7 and 8) instead of NADPH/NTR. These results point to TRX-dependent regulation of ATG4 activity possibly due to the reduction of a regulatory disulfide bond.

Discussion
In plants, the regulation of autophagy has acquired notable interest due to its favorable response to different stress conditions including oxidative stress, endoplasmic reticulum stress, nutrient deprivation, Cd treatment, drought, salinity, and pathogen response [7,[38][39][40][41][42]. In the present work, higher ATG4 gene expression and protein content of the autophagy markers ATG4, ATG8, and lipidated ATG8-PE proteins as well as high ATG4 activity support the involvement of autophagy process in the response of tobacco BY-2 cells subjected to an H 2 O 2 treatment when PsTRXo1 is overexpressed. In this oxidative situation, these changes in OEX cells may be important as a priming effect, probably influencing the response to a subsequent oxidative stress condition. In this scene, autophagy occurs as is also evident from an accumulation of autolysosomes and ATG8 inmunolocalization. Interestingly, this process correlates with increased cell viability demonstrated in the overexpressing PsTRXo1 in TBY-2 cells. Recent studies have shown that posttranslational modifications of some ATG proteins are able to control their activity, such as S-nitrosylation or phosphorylation of ATG4b and ATG1 [43,44] or persulfidation of ATG4 and ATG18a [7,45]. We have explored the role that the PsTRXo1 protein may play in the autophagy process by redox regulation of the ATG4 protein involved. In the following sections, we will discuss the main findings evidencing the role of autophagy in response to stress and its involvement in cell viability and redox regulation including the TRX system as a modulator of the activity of target proteins that are key components of the autophagy process. The discussed evidence will help support our findings related to the crosstalk between autophagy, cell viability, and redox signaling via TRX.

Autophagy, Oxidative Stress and TRX
Autophagy is slightly activated at a basal level to maintain normal human cellular homeostasis and is under several pathological conditions when this degradative process is induced [46]. In plants, a basal housekeeping level also exists under favorable conditions to eliminate the continuously generated damaged structures or cellular components [47,48]. It is under stress conditions that phenotypes are evident, and growth can be seriously compromised [49]. Hence, studying autophagy under these conditions may provide some clues concerning the relevance of the process in the plant response. Many stresses trigger ROS accumulation, which can induce signaling events involved in the plant response, and a key role of autophagy in the clearance of oxidized proteins has been reported. In this sense, Arabidopsis atg2 and atg7 mutants contained more oxidized proteins and were hypersensitive to both salt and osmotic stresses [50]. Other examples of the link between ROS and autophagy are the described induction of this process by salt stress and nutrient starvation, which can be blocked by NADPH oxidase inhibitors [39]. Moreover, AOXdependent ROS signaling is a key event triggering autophagy as evident from experiments in which the scavenging of H 2 O 2 in aox19 mutants induced autophagosome accumulation under drought stress [51]. More directly, autophagy has been reported to be induced by H 2 O 2 or methyl viologen treatments in the roots of Arabidopsis. Moreover, defective mutants in autophagosome formation in RNAi-AtATG18a transgenic plants are more sensitive to oxidative stress, presenting a lower degree of degradation of oxidized proteins as a result of the autophagic process [52]. All these experiments point to the involvement of autophagy in the oxidative stress response. In this context, we previously proposed that one of the PsTRXo1 functions in TBY-2 cells is to offer protection against ROS, mainly H 2 O 2 . In fact, a lower content of endogenous H 2 O 2 and oxidative parameters and a decrease in antioxidants after oxidative treatment were reported in PsTrxo1 overexpressing cells compared to non-overexpressing cells [28]. In this oxidative situation, autophagy may also occur, so it was analyzed whether TRX could play a role in the possible autophagy induction in response to H 2 O 2 treatment. The analysis of key autophagy markers revealed that ATG4 gene expression was downregulated in the control and maintained in OEXtreated cells (Figure 2), pointing to the possible establishment of the autophagy process as a result of overexpression. In this context, scarce information exists about the regulation of the ATG4 gene expression by redox components via the transcription factors involved. Hence, the reason for the observed effect of overexpression of TRXo1 on the increase in the ATG4 gene expression is unclear, although redox modulation cannot be ruled out. Moreover, in our study, the high and maintained ATG4 protein content was parallel to high ATG4 activity also maintained in H 2 O 2 -treated OEX cells (Figure 3), indicating that sufficient activity was present to sustain autophagy. In this sense, it has been reported that under pathological conditions, the ATG4 processing step is very fast, and only a small amount of ATG4B is enough to sustain the activity in yeast and mammalian cells [53]. On the other hand, an increase in the abundance of ATG8 or ATG8-PE protein levels, like the one we found in our cells, may not reflect a higher autophagic flux due to the fact that blockage of autophagosome formation may also result in the accumulation of these markers [54], so other assays are necessary to estimate this flux and thus ratify the occurrence of the autophagic process. Related to this, and in order to look more closely at this occurrence of autophagy, we monitored autolysosome and autophagosome accumulation using Lysotracker and an anti-ATG8 antibody as markers for autophagy visualization [55,56]. In our study, the accumulation of lytic vesicles was visualized 24 h after oxidative treatment in the OEX cells treated with the inhibitor E-64d, an effect that was reversed in the presence of 3-MA ( Figure 5), an inhibitor that has been shown to block starvation-induced autophagy in BY-2 cells thus avoiding autophagosome formation [5,32]. Adding 3-MA clearly inhibited the number of punctate signals, suggesting the presence of autophagic structures in these cells and not in the non-overexpressing ones, which showed no such accumulation at any of the times analyzed (Figures 6 and 9). These results corroborate the early reported effect of E-64d in tobacco cells, which were seen to accumulate autolysosomes outside the central vacuole [57]. In fact, the mode of action of both inhibitors is different: 3-MA has been reported as a selective inhibitor of the early stages of autophagy by inhibiting PI3K activity, which is required for the nucleation and thus the formation of the autophagosome, while E-64 d inhibits autophagy at the later stage mainly by inhibition of papain-like Cys proteases blocking the endocytotic pathway [32,58].
Our immunolocalization study also sustained the existence of autophagic structures due to the observed increase in the content of ATG8 in the H 2 O 2 -treated OEX cells in the presence of the inhibitor ( Figure 6). Among ATG proteins, ATG8 has been considered an ideal marker for ascertaining autophagy activity due to the fact that it is incorporated into the membranes of the autophagosomes before being directed to the lytic compartment to be degraded together with the cargo [54]. All the above-mentioned results point to an interesting relationship between oxidative stress and autophagy linked to the overexpression of TRXo1. Moreover, this situation is accompanied by an interesting effect on cell survival as discussed in the following section.

Autophagy, Cell Survival and TRX
The use of autophagy inhibitors allowed us to evidence the link between autophagy, oxidative stress, and cell viability when PsTRXo1 was overexpressed. In our TBY-2 cells treated with H 2 O 2 , autophagy was not evident in the first 2 h, at least considering the response to the inhibitor E-64d, which did not significantly affect the viability of any of the lines ( Figure S5). However, in an oxidative situation severe enough to provoke the death of most of the cells of the control line 24 h after H 2 O 2 treatment (only around 20% of cells survived, Figure 4), around 65% of the OEX1 line survived. The fact that the autophagy inhibitor E-64d induced a decrease in OEX1 line viability by up to 33% at this time, parallel to the maintenance of ATG4 gene expression presented by these cells, may sustain the existence of an autophagic process occurring with an interesting vital role in cell survival. This effect was still observed 48 h after oxidative treatment when the viability was maintained at 54% and decreased to 33% in the presence of the inhibitor, with the decreases again being reverted by 3-MA ( Figure S2). Moreover, at a shorter time after the oxidative treatment, at 14 h, when the control line presented higher viability than at 24 h, the autophagy inhibitor E-64d again did not affect cell viability (Figure 8), suggesting that at this time, basal autophagy may be minimal in non-overexpressing cells, a parallel situation to the reported induction of programmed cell death [28]. Related to the effect of 3-MA on viability, and similar to that found in our cells, Takatsuka et al. [32] demonstrated that 3-MA inhibited autophagy in TBY-2 cells although it did not provoke cell death for at least 2 days at 5 mM or 10 mM concentrations ( Figure S3), so cell death could not explain the effect of 3-MA on autophagy. In addition, they confirmed that the presence of both E-64c and 3-MA in culture media did not affect cell death as we have also found (Figures 4 and 8). Furthermore, Voitsekhovskaja et al. [36], using TBY-2 cells, observed a lack of effect of 3-MA on viability, although it was able to revert the effect of concanamycin in the degradation of peroxisomes by autophagy. They described the effect of 3-MA on viability at longer durations after inducing autophagy (after 7 days), considering that in the first 6 days, either starvation-induced peroxisome degradation was incompletely blocked or that it was due to another mechanism, different from autophagy. Recently, the contribution of autophagy to cell survival was confirmed in lace plants where a decrease in the cell death rate was observed when autophagy was enhanced with rapamycin, whereas the inhibition of autophagosome formation or blocking of the cargo degradation process had the opposite effect [59]. In this scene, autophagy mediated by TRX could act as a pro-survival mechanism in our cells, as it has been described in situations where recycling of cellular components is necessary, not only for development but mainly under stressful situations [60]. ROS regulation together with autophagy may act as cooperative mechanisms in the response of plants to environmental stresses in order to allow plant survival. Among the actors of ROS regulation, TRX appears to be a key protein in the regulation of the autophagy process, extending its role towards the regulation of cell survival similar to that described in animals. The signaling pathways for activation of the autophagy process, and how its function switches between cell survival and cell death, are just beginning to be elucidated, with several questions still to be solved. The relationship between autophagy and cell survival and the involvement of dysregulation of redox signaling have been identified in animal systems. For example, increasing autophagy by enhancing ATG8a expression in neurons of Drosophila increased lifespan [61] and the ubiquitous overexpression of peroxiredoxin dPrx5 in flies subjected to oxidative stress also extended survival [62] while the combined knockdown of dPrx5 and dPrx3 decreased protein thiol levels and the glutathione (GSH) redox state, provoking pro-apoptotic events and shortening lifespan [63]. In plants, where the process has been less studied, an association of autophagy and longevity has been described when autophagy mutants displayed an early leaf senescence phenotype [64]. However, the exact role of autophagy as being pro-death or pro-survival is unclear [65]. Related to this, in our system, the overexpression of PsTRXo1 in TBY-2 cells subjected to H 2 O 2 treatment allowed maintenance of fresh weight of the cultures parallel to a high cell viability and the occurrence of autophagy. In these H 2 O 2 -treated cells, we previously observed increased catalase activity and decreased H 2 O 2 and NO content parallel to the maintained GSH redox state as well as the level of PrxIIF. All these changes contributed to an observed delay in cell death, compared to control non-overexpressing cells [28]. It is also interesting to point out that the additional involvement of TRXo1 in the TBY-2 cell cycle progression, possibly through its link with the proliferating cellular nuclear antigen (PCNA) and GSH. was also reported. In fact, overexpressing Pstrxo1 cells presented a higher rate of cell proliferation than non-transformed cells, coinciding with upregulation of the target protein PCNA and an effect on cellular GSH content and localization [26]. The described interaction between PCNA and TRXo1 to a certain extent could be contributing to the greater viability in these cells. Now, the present work further adds the involvement of autophagy in the response to the oxidative treatment in overexpressing PsTRXo1 cells, especially in their increased viability and maintenance of fresh weight indicative of cell proliferation. Interestingly, similar to that described in our present work, a relationship between TRX, autophagy, and cell survival has been reported in human lens epithelial cells under oxidative treatment by H 2 O 2. In this situation , overexpression of thioredoxin binding protein 2 (TBP-2), which negatively regulates the thioredoxin antioxidative function, promoted cell injury and autophagy, worsening cell viability [66], with TRX then playing a key role as a pro-survival mechanism. In fact, the role for TRX may be to coordinate the response to oxidative stress, probably through the redox regulation of key components of the autophagy process, an aspect discussed in the following section.

TRX-Dependent Redox Regulation of Autophagy Target Proteins
HsATG4B is a target of PsTRXo1 as we demonstrate by dot-blot and co-immunoprecipitation assays using recombinant proteins, which sustained the redox regulation of ATG4 activity. In fact, PsTRXo1 but not its mutated version, together with NADPH and NTRA, is able to recover the lost processing activity of oxidized ATG4 on the ATG8 quimeric substrate, as shown in the activity assays ( Figure 12; Figure 13). ATG8 was processed not only by recombinant HsATG4B (described as the most-processing isoforms in animals) [67], but also by our TBY-2 cell extracts, which contained the endogenous ATG4 protein (Figure 3).
These results confirm the redox regulation of ATG4 activity by the plant TRX system as a result of the reduction of a regulatory disulfide bond. Interestingly, the participation of TRX in autophagy, particularly in ATG4 regulation, has only been demonstrated in Chlamydomonas and Sacharomyces [21,22], where TRXh and Trx1, respectively, efficiently reduced a disulfide bond between two Cys with low redox potential, corroborating the redox regulation of ATG4 by TRX.
Autophagy must be finely regulated by the cellular environment, and posttranslational modification of target proteins is emerging as a key mechanism of control. Redox regulation of the autophagy process has been described in animal, yeast, and plant systems as previously commented, but information concerning the specific targets and the effects of this regulation is scarce. Among ATGs, increasing interest is focused on the regulation of ATG4 in animal systems due to its involvement in multiple pathological diseases, recently being proposed as a potential anticancer target [68]. ATG4 is the sole protease among ATGs that regulates autophagy through the first very fast and efficient processing step that only needs a low amount of ATG4 [53], and a second deconjugating step on its substrate ATG8. This last activity is used to recycle ATG8 for the next round of the conjugation reaction, but it is also necessary to promote the elongation step of the autophagosome and its fusion to lysosomes as observed in yeast [11]. In our experiments, OEX cells presented a higher ATG4 protein content after the H 2 O 2 treatment than control cells, but this was not accompanied by a similar increase in the processing activity. This discrepancy may be explained by possible high delipidation activity of the protein (not measured in our work) or by the existence of a regulatory posttranslational modification such as reduction/oxidation that affects the processing activity, as we have demonstrated. In animal cells, during starvation, autophagy is regulated via superoxide production, while the activity of Arabidopsis ATG4a and ATG4b was shown to be reversibly inhibited by oxidation [69]. In yeast, ROS directly regulate the ATG4 protein, thus activating autophagy and nitrosoglutathione treatment of ATG4 and inhibiting the processing activity of the protein [21], implying that the regulation of critical Cys residues may play an important role in the function of this protein. It has been recently reported in Drosophila that oxidation of a key redox-regulatory cysteine residue in Atg4a increased autophagy and interestingly extended lifespan of the flies [70], although the authors described the redox regulation only for the de-lipidation activity of the ATG4 protein, with this then proposed as a potential pro-longevity target. Furthermore, the persulfidation of Arabidopsis ATG4a has been recently reported as regulating the proteolytic activity of this protease [45] indicating a negative regulation of autophagy by sulfide. Thus, one link between ROS and the induction of autophagy may be ATG4, although it does not seem to be the sole protein in the process that underlies oxidative regulation, since not only is the target of rapamycin (mTOR) redox regulated by Trx1 in the heart [71], but ATG3 and ATG7 have also recently been shown to be redox regulated in human embryonic kidney cells [72]. Interestingly, similar to that found in our work, these last authors described that an active cellular reducing system is required to maintain functional autophagy, in this case, through ATG3 and ATG7 redox regulation, maintaining the active thiols at a low level in order to interact with the ATG4 substrate. In this scene, the inhibition of TRX by overexpression of its inhibitor thioredoxin interacting protein (TXNIP) lowered LC3 lipidation through enhancing inhibitory ATG3 and ATG7 catalytic thiol oxidation, again pointing to the key role of the redox environment and TRX in autophagy regulation, as we have evidenced.

Conclusions
In this work, we have described that PsTrxo1 transformation protects TBY-2 cells from exogenous H 2 O 2 , with the over-expressing cells showing higher viability than nonoverexpressing cells. The increase in ATG4 and ATG8 proteins as well as the existence of an autophagy flux revealed by lytic autophagic structures and ATG8 accumulation using an autophagy inhibitor in H 2 O 2 -treated OEX cells point to a pro-survival autophagy process that occurs as a response to the oxidative situation. The high ATG4 activity in OEX PsTRXo1 cells, together with the in vitro interaction demonstrated between ATG4 and TRXo1 proteins, strongly suggest a possible role of TRX in the redox control of the ATG4 protein. In this sense, TRX may be a component of the autophagy process occurring during the response of TBY-2 cells to an oxidative situation, when it probably collaborates with other defense systems in the increased cell survival rate of the over-expressing cells.