A novel zinc‐binding alcohol dehydrogenase 2 from Arachis diogoi, expressed in resistance responses against late leaf spot pathogen, induces cell death when transexpressed in tobacco

A novel zinc‐binding alcohol dehydrogenase 2 (AdZADH2) was significantly upregulated in a wild peanut, Arachis diogoi treated with conidia of late leaf spot (LLS) pathogen, Phaeoisariopsis personata. This upregulation was not observed in a comparative analysis of cultivated peanut, which is highly susceptible to LLS. This zinc‐binding alcohol dehydrogenase possessed a Rossmann fold containing NADB domain in addition to the MDR domain present in all previously characterized plant ADH genes/proteins. Transient over‐expression of AdZADH2 under an estradiol inducible promoter (XVE) resulted in hypersensitive response (HR)‐like cell death in tobacco leaf. However, the same level of cell death was not observed when the domains were transiently expressed individually. Cell death observed in tobacco was associated with overexpression of cell death related proteins, antioxidative enzymes such as SOD, CAT and APX and pathogenesis‐related (PR) proteins. In A. diogoi, AdZADH2 expression was significantly upregulated in response to the plant signaling hormones salicylic acid, methyl jasmonate, and sodium nitroprusside.

A novel zinc-binding alcohol dehydrogenase 2 (AdZADH2) was significantly upregulated in a wild peanut, Arachis diogoi treated with conidia of late leaf spot (LLS) pathogen, Phaeoisariopsis personata. This upregulation was not observed in a comparative analysis of cultivated peanut, which is highly susceptible to LLS. This zinc-binding alcohol dehydrogenase possessed a Rossmann fold containing NADB domain in addition to the MDR domain present in all previously characterized plant ADH genes/ proteins. Transient over-expression of AdZADH2 under an estradiol inducible promoter (XVE) resulted in hypersensitive response (HR)-like cell death in tobacco leaf. However, the same level of cell death was not observed when the domains were transiently expressed individually. Cell death observed in tobacco was associated with overexpression of cell death related proteins, antioxidative enzymes such as SOD, CAT and APX and pathogenesis-related (PR) proteins. In A. diogoi, AdZADH2 expression was significantly upregulated in response to the plant signaling hormones salicylic acid, methyl jasmonate, and sodium nitroprusside.
Zinc-binding alcohol dehydrogenases are NAD(P)dependent oxidoreductases involved in hydride ion transfer from alcohols to NAD + catalyzing reversible oxidation of alcohols to aldehydes or ketones. They are found in all plants, playing important roles in plant growth, pollen, and seedling development and fruit ripening [1,2]. They are well-studied members of the medium-length dehydrogenase/reductase (MDR) protein superfamily [3].
Expression of zinc-binding alcohol dehydrogenase was induced in Cyanobacterium PCC 6803 upon exposure to different environmental stresses [4] and in citrus upon interaction with Citrus leprosis virus [5].
Cinnamyl alcohol dehydrogenase (CAD), a zinc-binding MDR enzyme, is involved in lignin biosynthesis exhibiting defense-related activity and strengthening of cell walls [6]. A zinc-binding alcohol dehydrogenase is induced in Oomycete plant pathogenic fungus, Phytophthora infestans during the infection process [7]. Recently, another member of the MDR protein superfamily, an alcohol dehydrogenase gene of the fungus Metarhizium anisopliae, was found to be expressed during insect colonization and invasion [8].
In maize, ADH1 transcripts increase significantly under anaerobic conditions [9]. Chung and Ferl [10] demonstrated that transgenic Arabidopsis plants expressing ADH promoter-Gus fusions exhibited high Gus expression when grown under conditions simulating anoxia. When these plants are grown in conditions that mimic normal soil growth, the Gus expression was low, showing the relative levels of ADH under anoxia and normal soil conditions. In line with these observations, Ismond et al. [11] showed that the Arabidopsis-ADH1-null mutants exhibited decreased hypoxic survival. Dolferus et al. [12,13] studied the expression of the ADH gene in Arabidopsis and showed that the gene is induced by dehydration, cold and hypoxia and other environmental stresses. An alcohol dehydrogenase of Lotus japonicus (LjADH1) conferred tolerance to H 2 O 2 induced stress in E. coli and yeast cells along with tolerance to heavy metal salts in the latter [14]. Myint et al. [15] analyzed the expression of alcohol dehydrogenase genes in Arabidopsis under polyethylene glycol mediated stress and observed upregulation of ADH1 and other ADH genes in the treatments indicating that the operation of drought tolerance probably involved ethanolic fermentation.
Several wild relatives of peanut including Arachis diogoi are highly resistant to many foliar diseases and biotic stresses [16]. In an earlier investigation, we have identified a Zinc-binding alcohol dehydrogenase 2 (AdZADH2) gene in Arachis diogoi in its interaction with Phaeoisariopsis personata that causes the serious late leaf spot disease in the cultivated peanut, Arachis hypogaea [17]. In this communication, we have focused on the cloning of the novel AdZADH2 and its involvement in HR-mediated cell death in tobacco. Rapid amplification of cDNA ends (RACE) was performed to clone the full length cDNA of AdZADH2 by using SMART TM RACE cDNA Amplification kit (Clontech, California, USA) using gene specific primers designed from the sequence of a partial fragment identified in an earlier study using cDNA-AFLP [17]. The ORF was amplified with primers AdZADH2-ApaI-F and AdZADH2-SpeI-R containing ApaI and SpeI restriction enzymes, respectively, by using Phusion TM DNA polymerase (Finnzymes, Loughborough, UK, NEB, Table S1), cloned in pTZ57R/T and sequenced. The open reading frame (ORF) of the AdZADH2 with flanking ApaI and SpeI sites was digested with appropriate restriction enzymes and cloned into pER8 vector (Gifted by N.-H. Chua, Rockefeller University, USA) to obtain the recombinant pER8:AdZADH2. We further cloned both the domains separately in pER8 vector as NADB Rossmann domain D1 (pER8:AdZADH2-D1) and MDR domain D2 (pER8:AdZADH2-D2) by using specific primers (AdZADH2-ApaI-F and AdZADH2-D1-SpeI-R for D1 and AdZADH2-D2-ApaI-F and AdZADH2-SpeI-R for D2) to obtain recombinant vectors pER8:AdZADH2-D1 and pER8:AdZADH2-D2 respectively (Fig. S5). Primer sequences used in this study are listed in Table S1.

Materials and methods
Agro-infiltration, chemical treatment, and cell death assessment The vectors were mobilized into Agrobacterium strain LBA4404 using freeze thaw method. Agrobacterium strains harboring different binary vector constructs were grown in Luria Broth in the presence of respective antibiotics, pelleted, resuspended and infiltrated as described earlier [18][19][20]. After the observation of cell death phenotype, the samples were collected, quick-frozen in liquid nitrogen and stored at À80°C for further analysis.

Cell death quantification and ion leakage experiment
The extent of cell death was quantified by Evans Blue staining following Baker and Mock [21] method. For electrolyte leakage analysis, leaf discs (8 mm diameter) were cut off from agro-infiltrated regions after 48 h post infiltration using a cork borer and transferred to distilled water containing 30 lM 17b-estradiol in Petri dishes under bright light with 14/ 10 h photoperiod [17,22,23]. Electrolyte leakage values were determined by using an electro-conductivity meter (Digisun Electronics, Hyderabad, India) at different time intervals and data were plotted from three biological replicates.

Quantitative RT-PCR analysis post estradiol treatment
Leaves were infiltrated with agrobacterium strain LBA4404 harboring pER8, pER8:AdZADH2, pER8:AdZADH2-D1, and pER8:AdZADH2-D2 vectors, respectively. Total RNA was isolated from leaf samples with and without estradiol treatment collected at 0, 24, and 48 h time points. First strand cDNA was synthesized and used in the analysis of the transcript levels for several defense-associated genes along with HR marker genes using gene specific primers (Table S2). Actin was used as internal control for calculating relative gene expression. Relative fold change in RNA expression was estimated using DDC T method [24].

Hormonal and SNP treatments
Twigs from the field-grown plants were cut and kept in a tray for 2 weeks on moist cotton saturated with sterile distilled water at the base and the tray was covered with a polythene sheet to maintain humidity for adventitious root formation and recovery. For various chemical treatments, the rooted twigs were kept in the corresponding solution. The treatments given were 100 lM salicylic acid (SA), 100 lM methyl jasmonate (MeJA), 100 lM abscisic acid (ABA), 250 lM ethephon, 100 lM sodium nitroprusside (SNP), a combination of 100 lM each of salicylic acid and methyl jasmonate. A mock treatment with water served as control. The treatments were carried out at different time intervals up to 24 h and incubated in a growth room at 27 AE 1°C under 14/10 h photoperiod provided by light intensity of 30 lmolÁm À2 Ás À1 . Samples were collected at regular intervals, quickly frozen in liquid nitrogen, and stored at À80°C until use.
Expression analysis of AdZADH2 using qRT-PCR Total RNA, isolated from control-and hormone-treated peanut samples was treated with RNase free DNase1 (Sigma-Aldrich, Carlsbad, CA, USA) to eliminate any DNA contamination and was reverse transcribed with oligo-dT primer using SMART TM MMLV Reverse Transcriptase (Clontech, Becton Dickinson, Palo Alto, CA, USA). cDNA samples were diluted fivefold and 0.5 lL of the diluted reaction mixture was taken as qRT-PCR template in a 20 lL total reaction volume containing 0.4 lM gene specific primers with10 lL SYBR Premix Ex Taq with ROX (Takara Bio Inc., Kusatsu, Shiga, Japan) and the samples were appraised in three technical replicates. PCR analysis was carried out in Realplex amplifier (Eppendorf, Hamburg, Germany) with the following cycle parameters: 95°C for 5 min; 40 cycle of 95°C for 20 s, 58°C for 20 s, 72°C for 20 s followed by melting curve. Gene specific primers were designed from 3 0 UTR of AdZADH2 (Table S1). Polyubiquitin (UBI1) and alcohol dehydrogenase III (ADH3) were used as internal controls for calculating relative quantification of gene expression [25,26]. Relative fold change in RNA expression was estimated using DDC T method [24].

Results and discussion
AdZADH2 sequence analysis A partial fragment of a zinc-binding alcohol dehydrogenase 2 was differentially upregulated in a cDNA-AFLP analysis of an interaction between Arachis diogoi and the late leaf spot pathogen P. personata [17]. It was made full length using 3 0 /5 0 RACE. The cDNA of AdZADH2 comprised an ORF of 1902 bp (Fig. S1) and codes for a polypeptide of 633 amino acids (GenBank Accession number KT321126). Its upregulation under fungal treatment has been revalidated in a comparative real time PCR analysis of the resistant wild peanut and its susceptible cultivated peanut counterpart (Fig. 1C); its upregulation under pathogen challenge was observed only in the wild peanut.
Multiple sequence alignment and phylogenetic analyses displayed its close relation in plants to soybean predicted zinc-binding ADH2-like protein (different from ADH2 studied [27-29]) with 85% sequence similarity (Fig. S2a). AdZADH2 shows BLASTP specific hit of Mgc45594_like in MDR domain region (Table S3) with an undetermined function. Psort (http://psort.hgc.jp/form.html) identifies AdZADH2 as a resident of peroxisomes (probability of 85.2%); moreover, PredPlantPTS1 (http://ppp.gobics.de) predicted the presence of a Cterminal PTS1 domain (tripeptide AKL) for peroxisome targeting (probability 100.0%). This C-terminal signal for peroxisome translocation is absent in previously studied ADH proteins. Moreover, we made multiple sequence alignment between AdZADH2 with some previously studied ADH proteins using network protein sequence analysis (NPS, https://npsa-prabi.ibcp.fr/cgibin/align_multalin.pl, [30]. Interestingly, the NADB Rossmann domain was absent in other plant ADH proteins studied so far (Fig. S4) showing that it is a novel ADH2. The predicted molecular weight and pI of AdZADH2 protein were found to be totally different from ADH proteins studied earlier (Table S3).
A protein BLAST analysis of AdZADH2 revealed two conserved domains, a NADB Rossmann domain and MDR domain (Fig. S2A,B). These domains are reportedly involved in various activities such as oxidoreductases, catalysis, and enzymatic reactions providing catalytic and structural stability to the protein [3,27]. Rossmann fold in NAD-dependent dehydrogenases found in many protein superfamilies (mostly oxidoreductases). It decides the specificity of hydride transfer thereby providing NAD or NADP as a cofactor to MDR domain for interconversion of ethanol and acetaldehyde. Interestingly NADB Rossmann domain of AdZADH2 was not much conserved as it showed just 83%, 72%, and 64% homology with NADB domains soybean (Glycine max; XP_003518845.1), tobacco (Nicotiana tomentosiformis; XP_009612756.1) and Arabidopsis (Arabidopsis thaliana; NP_175390.2), respectively; however, MDR domain was found to be highly conserved (Fig. S2).
Plant alcohol dehydrogenases (ADH) have been well characterized for their role in cell survival under hypoxic and anaerobic conditions. During anaerobic glycolysis i.e., fermentation, ADH diverts the formation of lactic acid from pyruvate to less toxic and more diffusible ethanol, thereby helping cells to survive in the absence of oxygen required for normal respiration. Available evidences suggest that plant ADH gene family comprises survivors of multiple rounds of gene expansion and contraction resulting in gains in roles like production of characteristic scents that act to attract animals, pollinators, or agents of seed dispersal and to protect plants against herbivores besides anaerobic fermentation [1,2].  Fig. 2A,B). Induced expres-sion of AdZADH2 was observed to be strong at 24 and 48 h postinfection (hpi). This strong expression of AdZADH2 was found to be associated with enhanced expression of defense-related genes (NtPR1a, NtPR1b, and NtPR5-TLP). Along with the expression of genes for pathogenesis-related proteins, marker genes for cell death such as NtPAT3, HMG-CoA R, HSR203J, and HIN1 and antioxidant enzymes such as catalase (CAT), ascorbate peroxidase (APX), and superoxide dismutase (Mn-SOD) also got upregulated in the infiltrated regions analyzed by quantitative RT-PCR up to 48 h postestradiol application (Fig. 3). HMG-CoA R, HSR203J, and HIN1 were found to be co-upregulated at 24 and 48 h of treatment. HIN1 and HSR203J have been shown to be closely associated with cell death in incompatible interactions in tobacco and were used as markers for HR [32][33][34]. Along with HR marker genes, we have Differentially expressed TDF of Zinc binding alcohol dehydrogenase 2 also observed the upregulation of PAT3 gene; transient expression of a patatin like phospholipase in tobacco resulted in the accumulation of fluorescent phenolics, pathogenesis-related proteins CaPR1, CaSAR82A, and ROS leading to cell death in pepper leaves [35]. Expression profiling postfungal treatment in Populus trichocarpa showed possible involvement of CAD/CAD-like genes in plant development and defense against various pests and pathogens [36]. The accelerated cell death gene, acd11 of Arabidopsis constitutively expresses the defense-related genes that are associated with hypersensitive response normally triggered by avirulent pathogens [37]. There is evidence of the involvement of ADH in plant fungal interactions through the activation of fermentative pathway. Pathogen infection, particularly biotrophic pathogens, leads to a condition of reduced oxygen pressure that leads to a fermentation pathway and the development of reactive oxygen species. Under such conditions, upregulation of ADH has been reported in biotrophic pathogen infections. Proels et al., [38] have observed that the barley seedlings   Fig. 3. Relative expression of defense-associated genes and cell death related genes in leaves transiently expressing pER8:AdZADH2, pER8:AdZADH2-D1 and pER8:AdADH2-D2 in relation to control (pER8) after estradiol application at 0, 24, and 48 h. Quantitative RT-PCR analysis was performed using total RNA from samples collected at various time points. Relative expression was calculated by comparative DDC T method. Statistical analysis was performed with student's t-test, asterisks indicate upregulated expression relative to control samples (*P < 0.05, **P < 0.001, ***P < 0.0001). Gene specific primers were used for amplification of different genes with Actin serving as internal reference gene. The primers sequence used in the study were given in Table S2. treated with the biotrophic pathogen, Blumeria graminis f.sp hordei upregulated ADH1 and ADH2, but not ADH3. It has been surmised that the pathogen infection might reduce photosynthetic activity in the infected tissues leading to reduced oxygen pressure resulting in a condition similar to anoxia and inducing the fermentation pathway to compensate the reduced energy levels. Moreover, ethanolic fermentation (a major means of ATP production under anaerobic condition in plants) leads to the accumulation of high levels of acetaldehyde, which is a strong cell toxin. It binds to nucleic acids and proteins forming stable acetaldehyde-protein adducts, eventually leading the cell toward PCD [39,40]. Uehara et al., [41] analyzed the differentially upregulated ADH gene in resistant and susceptible cultivars of tomato in response to the challenge from the cyst nematode and observed that there was differential upregulation of the ADH transcripts; incompatible interaction showed upregulation and compatible interaction evidencing downregulation of ADH. Hren et al., [42] observed the induced expression of sucrose synthase and alcohol dehydrogenase in resistant varieties of grapevine infected by phy-toplasma with the conversion of sucrose to alcohol through alcohol fermentation.
Cell death phenomenon was associated with the enhanced expression of defense-related genes, cell death marker genes and antioxidant enzymes (Fig. 3). Cell death was quantified by using Evans Blue and an ion leakage experiment [43]. Cell death with enhanced electrolyte leakage was significantly high in pER8: AdZADH2 regions in comparison to the empty vector infiltrated areas (Fig. 2D,E). The AdZADH2 protein carries two domains, a NADB Rossmann superfamily domain and a MDR Superfamily domain. Furthermore, to see if the cell death could be specifically associated with one of the two domains, we have cloned these domains [as both of them carry separate start (Met) codons] separately under the estradiol inducible promoter in pER8 as NADB Rossmann domain D1 (pER8:AdZADH2-D1) and MDR domain D2 (pER8: AdZADH2-D2). We found that after chemical induction, infiltrated area expressing MDR domain (pER8: AdZADH2-D2) showed significantly enhanced cell death in comparison to NADB Rossmann domain (pER8:AdZADH2-D1) (Fig. 2C). In addition, elec- Data were normalized to the Polyubiquitin (UBI1) and Alcohol dehydrogenase-3 (ADH-3) level as internal reference genes and plotted from three independent experiments. Statistical analysis was performed with student's t-test, asterisks indicates significant up-accumulation in relative to mock sample (*P < 0.05, **P < 0.001).
trolyte leakage analysis and cell death quantified by Evans Blue also supported this observation, where enhanced ion leakage and cell death have been observed with MDR domain in comparison to NADB Rossmann domain (Fig. 2D,E). However, the level of cell death induced by individual domains was less when compared with the complete protein, AdZADH2. Taken together, these results imply that transient expression of AdZADH2 induces cell death and MDR domain played a major role in this phenomenon probably in an interaction with the NADB domain (the expression of which also induced cell death, though not that extent as the MDR domain). Rossmann fold being a NAD-dependent dehydrogenase dictates the specificity of hydride transfer thereby providing NAD or NADP as a cofactor to MDR domain for interconversion of ethanol and acetaldehyde. Therefore, in the absence of Rossmann fold, MDR domain is probably not able to produce the same level of HR-like cell death as produced when both domains are present as in AdZADH2 owing to the reduced availability of NAD or NADP as a cofactor. Quantitative expression analysis shows overexpression of defense related genes, HR marker genes and ROS generated enzymes upon transient overexpression of pER8:AdZADH2 (Fig. 3); however, the relative fold changes were significantly less when these two domains were overexpressed individually. Moreover, changes in the expression level of HR marker genes viz. HIN1, HSR203J, and HMG-CoA R were not significant when Rossmann fold was overexpressed alone. In case of ethephon treatment, we found a trivial expression pattern of AdZADH2 at all-time points indicating that its expression is also not controlled by ethephon (Fig. 4). Thus, plants try to defend themselves from the pathogen attack and herbivory by activating a number of signaling molecules such as JA, SA and SNP, which share some common signaling components [46]. Though, SA has been initially thought to attenuate the JA pathway, subsequent analyses have conclusively proved the coordination between both the signaling molecules, which are interconnected in the plant to have economization of energy to combat various stresses that attack it [47]. Nitric oxide controls a signaling cascade that regulates plant responses to developmental processes, biotic and abiotic stress [48][49][50][51]. We have found induced expression in the AdZADH2 transcripts at late stage of SNP treatment. An interaction between hydrogen peroxide and nitric oxide is shown to be associated with several forms of cell death [52,53]. AdZADH2 appears to support nitric oxide (NO) mediated HR cell death; NO production leads to the production of H 2 O 2 (instead of ROS), which in turn results in HRmediated PCD [54]. Several lines of evidence have shown that the early nitric oxide (NO) burst in host plant cells after pathogen attack functions as a major defense response associated with resistance in plant pathogen interaction [55]. NO is an interesting molecule that is used by microorganisms in pathogenesis and host cells to activate the immune system to counter the pathogen challenge [56,57]. Thus, the significant upregulation of AdZADH2 by the resistant host against the LLS pathogen attack under SNP treatment could be justified as a resistance response.

Conclusion
Peroxisomal resident novel AdZADH2 protein was upregulated in A. diogoi in treatment with P. personata. Conditional overexpression of AdZADH2 in tobacco leaf under an estradiol inducible promoter (XVE) resulted in HR-like cell death associated with an enhanced electrolyte leakage and enhanced expression of defense-related genes (NtPAT3, NtPR1a, NtPR1b, and NtPR5), marker genes for cell death such as HMGR, HSR203J, and HIN1 and antioxidant enzymes such as CAT, APX, and SOD. The expression of AdZADH2 was upregulated post-SA, JA, and SNP treatments in A. diogoi. This report shows involvement of Zn containing ADH protein having NADB Rossmann fold and MDR domain in HR-like cell death in plants. Cell death activities are not domain additive, but are probably a result of interaction between these two domains; cell death was significantly decreased when the two domains are individually overexpressed. 20

Supporting information
Additional supporting information may be found in the online version of this article at the publisher's web site: Table S1. Oligonucleotides sequence used in the study. Table S2. Primers and their sequences used in study for semi-quantitative and quantitative PCR Analysis. Table S3. A list of previously studied ADH genes. Figure S1. Cloning of AdZADH2 from Arachis diogoi. Figure S2. Multiple sequence alignment of the deduced amino acid sequence of AdZADH2 with closely related uncharacterized Zinc-binding alcohol dehydrogenase sequences. Figure S3. Phylogenetic relationship of AdZADH2 with other zinc-binding alcohol dehydrogenase family members. Figure S4. Multiple sequence alignment of the deduced amino acid sequence of AdZADH2 with some previously studied zinc-binding alcohol dehydrogenase sequences from other organisms was done using default parameters of network protein sequence analysis (NPS). NADB Rossmann fold is absent in previously studied ADH proteins; moreover MDR domain of AdZADH2 shows no significant homology with other ADH proteins in the genes used in the alignment were from Zea mays (ZmADH1, ZmADH2), Oryza sativa (OsADH1), Solanum lycoperisicum (SlADH1), Arabidopsis thaliana (AtADH1, AtADH2), Glycine max (GmADH1, GmADH2), and Lotus japonicus (LjADH1), respectively. Accession numbers are given in Table S3. Figure S5. Cloning of NADB Rossmann domain and MDR superfamily domain. Figure S6. Transient constitutive expression of AdZADH2 induced cell death in tobacco leaf upon constitutive expression under 35S promoter.