Structure-function relationship of Gossypium hirsutum NAC transcription factor, GhNAC4 with regard to ABA and abiotic stress responses

Our previous study demonstrated that the expression of GhNAC4, a NAC transcription factor from cotton, was induced by abiotic stresses and abscisic acid (ABA) treatment. In the present study, we investigated the molecular mechanisms underlying ABA and stress response of GhNAC4. Over-expression of GhNAC4 in transgenic tobacco conferred tolerance to salinity and drought treatments with associated enhanced expression of several stress-responsive marker genes. GhNAC4 is a nuclear protein that exhibits transcriptional activation property, and the C-terminal transcriptional regulatory (TR) domain is responsible for it. GhNAC4 also forms homo-dimers and the N-terminal NAC-domain is essential for this activity. The structure-function relationship of NAC transcription factors, particularly with respect to abiotic stress tolerance remains largely unclear. In this study, we investigated the domains essential for the biochemical functions of GhNAC4. We developed transgenic tobacco plants overexpressing the GhNAC4 NAC-domain and the TR-domain separately. NAC-domain transgenics showed hypersensitivity to exogenous ABA while TR-domain transgenics exhibited reduced sensitivity. Abiotic stress assays indicated that transgenic plants expressing both the domains separately were more tolerant than wild-type plants with the NAC-domain transgenics showing increased tolerance as compared to TR-domain transgenics. Expression analysis revealed that various stress-responsive genes were upregulated in both NAC-domain and TR-domain transgenics as compared to wild-type under salinity and drought treatments. These results suggest that the stress tolerance ability of GhNAC4 is associated with both the component domains while the ABA responsiveness is largely associated with N-terminal NAC-domain. Key Message NAC and transcriptional regulatory domains are responsible for the abiotic stress tolerance ability of the cotton NAC transcription factor GhNAC4 while the ABA-responsiveness is largely associated with the NAC-domain.

(4 mg/mL in Z-buffer, HiMedia) was added and vortexed briefly before incubating at 30 o C for 2 min 1 4 5 to 4 h. The reaction was terminated by the addition of 400 µL of 1 M Na 2 CO 3 . The sample was 1 4 6 centrifuged for 2 min at 16000 x g, and the absorbance of the supernatant read at 420 nm and 550 nm.

8 4
Total chlorophyll content: The 100 mg samples of treated leaf discs were used for extracting total 1 8 5 chlorophyll using 80 % acetone according to Arnon (1949). Total chlorophyll content was expressed 1 8 6 as mg/g FW. This experiment was repeated three times.

8 7
The extent of lipid peroxidation: The treated leaf discs (100 mg) were homogenized in 1 ml of 0.1 1 8 8 % trichloroacetic acid (TCA) and centrifuged at 10,000 x g for 10 min. To the supernatant, 5 ml of 1 8 9 thiobarbituric acid (0.5 % in 20 % TCA solution) was added and the mixture was boiled for 25 min at 1 9 0 100 o C. The reaction was terminated by quick cooling on ice, followed by centrifugation at 10,000 x g 1 9 1 for 5 min. The supernatant was used to measure malondialdehyde (MDA) content according to Heath 1 9 2 and Packer (1968). The MDA content was expressed as nmoles/g FW. The experiment was repeated 1 9 3 three times.

9 4
Proline content: The treated leaf discs (100 mg) were homogenized in 0.5 ml of 3 % sulfosalicylic 1 9 5 acid and centrifuged at 10,000 x g for 10 min. To 100 μl of supernatant, a mixture of 100 μl of 3 % 1 9 6 sulfosalicylic acid, 200 μl of glacial acetic acid and 200 μl of acid ninhydrin were added. The mixture 1 9 7 was boiled for 60 min at 100 o C and the reaction was terminated by quick cooling on ice. The samples 1 9 8 were extracted with toluene and the upper organic phase was used for measuring the proline content 1 9 9 according to Bates (1973). The proline content was expressed as μmoles/g FW. The experiment was 2 0 0 repeated three times.

0 1
Expression analysis of stress marker genes in tobacco:

0 2
To assess the transcript levels of ABA and stress-responsive genes in tobacco seedlings, quantitative 2 0 3 real-time PCR (qPCR) was employed. For this, two weeks old seedlings of wild-type, GhNAC4, 2 0 4 GhNAC4-N, GhNAC4-C genotypes were treated with 200 mM NaCl and 15% PEG 8000 for 24 h.

0 5
Total RNA was extracted using Trizol reagent (Invitrogen, USA) and DNA contamination was 2 0 6 removed by treating it with RNase-free DNase I (Takara Bio, China and two independent biological replicates were used for the analyses.

3
In our previous study, we identified that GhNAC4 is highly induced in PEG, NaCl and ABA

6
GhNAC4 is a nuclear protein 2 2 7 The recombinant construct carrying the pEGAD:GhNAC4 translational fusion was agroinfiltrated in 2 2 8 onion epidermal cells and visualized through laser scanning confocal microscopy. The

9
GhNAC4:EGFP fusion protein has been identified to localize in the nucleus (Fig. 1). The EGFP of the 2 3 0 empty pEGAD vector, which was used as a control was observed throughout the whole cell. This 2 3 1 result suggests that GhNAC4 is a nuclear-localized protein. activity as compared to the full-length protein.

4 3
To determine the minimum region essential for the transcription activation property, a 2 4 4 progressive series of deletion variants of GhNAC4 were analyzed in yeast. The minimal truncated 2 4 5 region that showed transcription activation was from amino acid 140-199 (GhNAC4∆1-199-C5).

4 7
The β -gal assay demonstrated that progressive deletion of 30 amino acids reduced the activity by half, 2 4 8 as indicated in Fig. 3.

4 9
approximately 90%, 67% and 40% respectively whereas the wild-type seedlings root length reduced 3 0 3 to 19% (Fig. 6). This result clearly indicated that GhNAC4 overexpressing plants are tolerant to 3 0 4 salinity and drought stress and this property is conferred by both NAC-and TR-domains.

0 5
Next, we considered the possible effect of ABA on the root growth. As in the case of seed 3 0 6 germination, the GhNAC4-N seedlings showed hypersensitivity to ABA (10 µM) with significantly 3 0 7 lower primary root growth (approximately 20%) and whereas the GhNAC4-C seedlings were 3 0 8 insensitive (Fig. 6). This further confirms the ABA receptivity role of NAC-domain in GhNAC4.

1 8
This corroborated the observation that GhNAC4-N genotype shows hypersensitivity to ABA while 3 1 9 GhNAC4-C genotype is nearly insensitive and GhNAC4 is a positive regulator of ABA-mediated 3 2 0 stomatal closure.

2 1
The modulation of the NAC and TR-domains in ABA-mediated stomatal closure in the 3 2 2 transgenic plants has again prompted us to examine whether they show differences in the regulation 3 2 3 of dehydration responses. To investigate this, we excised leaves of wild-type, GhNAC4, GhNAC4-N 3 2 4 and GhNAC4-C genotypes and water loss was examined over a period of time (Fig 8). After 3 2 5 dehydration for 9 h, the water loss from excised leaves was higher for wild-type plants (~ 75%) as 3 2 6 compared to the GhNAC4 plants (~ 25%). The excised leaves of GhNAC4-N genotype lost water 3 2 7 more slowly (~ 35%) as compared to GhNAC4-C genotype (~ 55%) (Fig. 8). An increase in water 3 2 8 loss of GhNAC4-C plants could be partially due to the insensitivity to ABA-mediated stomatal 3 2 9 closure as compared to GhNAC4-N plants.

0
Both NAC and TR-domains are required for stress tolerance during vegetative growth 3 3 1 To assess the stress tolerance capabilities of wild-type, GhNAC4, GhNAC4-N and GhNAC4-C 3 3 2 genotypes, leaf discs from fully expanded leaves were floated on 200 mM NaCl and 15 % PEG compared to wild-type and GhNAC4-C (Fig. 9b).

4 6
Under abiotic stress conditions, plants produce osmolytes such as proline to maintain the

5 5
Both NAC and TR-domains regulate the expression of stress-responsive genes 3 5 6 Given the differential role NAC and TR-domains in ABA and abiotic stress responsiveness, it is 3 5 7 pertinent to question whether they alter the expression of abiotic stress and ABA-responsive genes.

6 2
and NtSUSY (sucrose synthase) under NaCl and PEG treatments in all the genotypes. Under control 3 6 3 conditions, the expression levels of all the 9 genes were very low and no significant differences were 3 6 4 observed between wild-type, GhNAC4, GhNAC4-N, GhNAC4-C genotypes. However, under NaCl 3 6 5 and PEG treatment, the transcript levels of all the nine genes were up-regulated (Fig. 10).

6 6
The transcriptional regulatory role of GhNAC4 was supported by the altered expression of 3 6 7 many ABA and stress-responsive genes. In the GhNAC4 genotype, the transcript levels of all the 9 3 6 8 3 9 8 Transgenic rice over-expressing SNAC1 showed enhanced ABA sensitivity and had more rapid  (NtNCED3,NtDREB3,NtSOS1,NtERD10C,NtERF5,NtAPX,NtCAT1,NtMnSOD,and NtSUSY) 4 1 8 were highly upregulated by GhNAC4 under salt and drought stress conditions. Upregulation of stress-  Table   4 2 4 3). This suggests that GhNAC4 may act as a possible direct upstream regulator of abiotic stress 4 2 5 responsive genes.

2 6
Interestingly, a very high upregulation of NtNCED3 and NtDREB3 transcripts was observed

1
We demonstrated that the stress tolerance ability of GhNAC4 is associated with both the 4 4 2 domains, NAC-and TR. Tobacco transgenics expressing both GhNAC4-N and GhNAC4-C 4 4 3 constructs independently were tolerant to salt and drought treatments as compared to wild-type.

4
However, GhNAC4-N genotype plants were more tolerant compared to GhNAC4-C expressing Sensitive 1) encodes a plasma membrane Na + /H + antiporter and plays a critical role in controlling the 4 5 6 transport of Na + from root to shoot, thus being essential for maintaining Na + and K + homeostasis.

5 7
Increased expression of SOS1 is reported to improve salt tolerance in Arabidopsis (Shi et al. 2003). A

6 3
Again, a level of higher induction of ROS-scavenging genes such as catalase 1 (NtCAT1), ascorbate  and guard cell cytosolic free calcium in response to oxidative stress. Plant Physiol 111:1031-