An ethylene-induced NAC transcription factor acts as a multiple abiotic stress responsor in conifer

Abstract The proper response to various abiotic stresses is essential for plants' survival to overcome their sessile nature, especially for perennial trees with very long-life cycles. However, in conifers, the molecular mechanisms that coordinate multiple abiotic stress responses remain elusive. Here, the transcriptome response to various abiotic stresses like salt, cold, drought, heat shock and osmotic were systematically detected in Pinus tabuliformis (P. tabuliformis) seedlings. We found that four transcription factors were commonly induced by all tested stress treatments, while PtNAC3 and PtZFP30 were highly up-regulated and co-expressed. Unexpectedly, the exogenous hormone treatment assays and the content of the endogenous hormone indicates that the upregulation of PtNAC3 and PtZFP30 are mediated by ethylene. Time-course assay showed that the treatment by ethylene immediate precursor, 1-aminocyclopropane-1-carboxylic acid (ACC), activated the expression of PtNAC3 and PtZFP30 within 8 hours. We further confirm that the PtNAC3 can directly bind to the PtZFP30 promoter region and form a cascade. Overexpression of PtNAC3 enhanced unified abiotic stress tolerance without growth penalty in transgenic Arabidopsis and promoted reproductive success under abiotic stress by shortening the lifespan, suggesting it has great potential as a biological tool applied to plant breeding for abiotic stress tolerance. This study provides novel insights into the hub nodes of the abiotic stresses response network as well as the environmental adaptation mechanism in conifers, and provides a potential biofortification tool to enhance plant unified abiotic stress tolerance.


Introduction
Global warming and extreme weather make abiotic stress as an important factor affecting world food security [1][2][3]. Plants grown in natural environments always face different survival stress. In addition, biotic stress like herbivorous animal attacks, human activities and diseases caused by microorganisms have certain effects on plants [4,5]. In the context of global warming, the effects of abiotic stresses on plant growth and development become more pronounced [6][7][8][9].
Abiotic stress normally refers to adversity growth conditions like drought, salt, heat, and cold, as well as ultraviolet (UV) which can affect plant growth [10]. Since the Green Revolution, great efforts had been taken by scientists to enhance plant ability to resist abiotic stress and shorten plant lifespan without affecting yield [11]. It is a great challenge to enhance plant abiotic stress resistance without affecting its growth in plant biotechnology by the reason of the trade-off between defense and yield [12,13].
To realize this ideal, the regulatory mechanisms of response to abiotic stress in model plants have been extensively studied. In angiosperm, CPKs and MAPKs family genes were regarded as responders that responded to multiple abiotic stress by the abscisic acid (ABA) signaling pathway [5]. To understand the abiotic stress regulatory mechanisms, the receptors and signaling pathways of ABA have been identified [14]. However, what the scene is in conifers remains elusive. Although some clues indicate ABA participates in some abiotic stress responses in conifer, some abiotic stress related genes had been identified [15][16][17]. The integration mechanism of multiple abiotic stress response pathways in conifer is still unknown.
Conifer serves as one of the most widely distributed trees in the northern hemisphere and has crucial ecological and economic value. As a gymnosperm, conifer accounts for 39% of the global forests and possesses 615 extant species due to its super adaptability [18]. For conifers, the ability to respond to abiotic stress not only predetermined potential distribution but also the length of the life-span. There was much evidence to prove conifers have a stronger ability against unfavorable environmental conditions [10,16,19]. Interestingly, the whole genome-wide analysis of Pinus tabuliformis (P. tabuliformis) found C-repeat/DREB binding factors (CBFs), which played an important role in abiotic stress in angiosperm, were lacking in conifer [19], indicating they probably have a different abiotic stress response pathway in conifer.
NAC transcription factors (TFs) especially the ATAF subfamily have been widely reported as stress-related TFs [20][21][22]. In rice, stress-related NAC genes, such as OsNAC5, OsNAC6, OsNAC9, and OsNAC10 responded to drought, high salinity, and ABA [23]. Overexpressed OsNAC6, OsNAC9, and OsNAC10 enhanced drought stress resistance and overexpressed OsNAC5, OsNAC9, and OsNAC10 increased grain yield under both normal and drought conditions [24,25]. Compared to the other three TFs, OsNAC5 responded to almost all abiotic stress, overexpressed OsNAC5 increased plants tolerance to multiple abiotic stress [26], and owing to its powerful function in abiotic stress response, OsNAC5 was regarded as a potential tool for biofortification strategies in rice [27]. The NAC genes RhNAC3 and JUNGBRUNNEN1 (JUB1) were reported to enhance dehydration tolerance in rose [28] and drought tolerance in tomatoes [29]. Ectopic expression of HaNAC1, VvNAC17, ZmSNAC13, PeNAC034, PeNAC045, and PeNAC036 in Arabidopsis enhanced multiple abiotic stresses tolerance in an ABA pathway-dependent manner [30][31][32][33]. Taken together, NAC TFs played a critical role in enhancing abiotic stress tolerance, and the function was conserved in angiosperm.
In gymnosperm, SNAC family members participated in abiotic stress response in many cases. In recent years, PpNAC2 and PpNAC3 were isolated in Pinus pinaster and were found in response to multiple abiotic stress like salt, cold, and wounding treatment [34]. In Norway spruce (Picea abies), PaNAC03 was found induced by stress and participated in stress defense by secondary metabolite production [35]. In Picea wilsonii, PwNAC2 was reported to enhance plant tolerance under abiotic stress [36]. In P. tabuliformis, PtNAC3 was found to be related to cold stress [37]. Although part of the SNAC family gene was found related to abiotic stress, there were 26 members in P. tabuliformis, and their role in adversity requires more research.
In conifers, the molecular mechanisms that coordinate multiple abiotic stress responses remain largely unknown. Here, we systematically detected the transcriptomic response to various abiotic stresses like salt, cold, drought, osmotic, and heat shock in P. tabuliformis seedlings. The core TFs that respond to multiple abiotic stresses were selected, and a NAC family member was highly induced by all test treatments. Furthermore, overexpression of the PtNAC3 enhanced unified abiotic stress tolerance without growth penalty in transgenic Arabidopsis. Interestingly, we show that the activation of PtNAC3 was mediated by ethylene (ET) rather than ABA. These results provide novel insights into the hub nodes of the abiotic stresses response network as well as the environmental adaptation mechanism in conifers, and provide a potential genetic tool to enhance plant unified abiotic stress tolerance.

The transcriptomic landscapes and hub genes response to diverse abiotic stress in P. tabuliformis
Conifers are adaptive to harsh environments, such as extreme cold and drought, particularly many species in the pine family have large distribution areas covering multiple climate zones in the northern hemisphere [2]. To investigate the core genes response to different abiotic stress in conifer, the seedlings of P. tabuliformis were treated with cold (4 • C), heat (40 • C), drought, salt (2 M NaCl), and osmotic (300 mM mannitol). RNAseq analysis showed that different stress treatments tended to induce specific transcriptome response profiles, such as, 3222/3137 genes were specifically induced/repressed by salt, 1760/1824 genes were specifically induced/repressed by heat, 1244/510 genes were specifically induced/repressed by osmotic, 1209/341 genes were specifically induced/repressed by cold, 532/692 genes were specifically induced/repressed by progressive drought. Meanwhile, the most significant differential gene induced/repressed by different abiotic stress shared less similarity ( Fig. 1A; Fig. S1A, see online supplementary material). Strikingly, 32/16 induced/repressed genes were shared by all abiotic stress treatments ( Fig. 1B and C). The 32 abiotic stress induced genes (ASIG) include four TFs: PtNAC3, PtNAC5, PtMYB175 and PtZFP30 (Table S2, see online supplementary material). The expression profiling showed that NAC (NAM/ATAF/CUC) transcription factors PtNAC3 and PtNAC5 have higher expression levels and more obvious differences than the other two TFs when suffering abiotic stress (Fig. S1B, see online supplementary material), indicating NAC TFs may play an important role in the abiotic stress response of conifers.
PtNAC3 acts as a core responsor of abiotic stresses response in P. tabuliformis We further analysed the detailed expression profiles of these multiple stress response TFs under moderate unfavorable growing conditions ( Fig. 2; Fig. S2, see online supplementary material). All four abiotic stress-induced TFs show a similar response profile to these stress conditions; however, compared with other ASIGs, PtNAC3 showed a more sensitive and vigorous expression response, indicating that it probably acts as a core transcription activator in abiotic stress response. The result showed that PtNAC3 was up-regulated under 4 • C, 10 • C, 30 • C, 40 • C temperatures compared with 20 • C, and induced by 8 days of moderate drought treatment and slightly up-regulated for 23 days of progressive drought treatment, while its expression level was recovered to control level after rewatering one day [16]. We found that the response of these genes should be a very rapid process. For example, one day of UVB treatment is enough to induce the expression of PtNAC3 hundreds of times, and 8 hours of injury treatment can activate its response [10]. The PtZFP30 (PtJG40760), PtNAC5 (Pt2G26090), and PtMYB175 (Pt3G25430) also show a similar response profile albeit to a lesser extent. Compared to other treatments, including osmotic stress induced by mannitol, NaCl treatment induced the expression of these genes to the greatest extent ( Fig. 2A; Fig. S2, see online supplementary material).
As stress response genes, the four TFs were expressed at low levels during normal growth seasons in the field during the annual cycle, but only highly accumulate in the winter (Fig. 2B, Fig. S2, Fig. S6, see online supplementary material). In the diurnal cycle, these four TFs were also only showing high accumulation in winter and did not show obvious circadian rhythm (Fig. S2, see online supplementary material). Notably, the expression patterns of these four TFs were very similar, and the correlation coefficient between the expression of PtZFP30 and PtNAC3 was R 2 = 0.76 (Fig. 2C), indicating that they may be in the same response pathway.

PtNAC3 is a stress-related NAC (SNAC) transcription factor
The NAC genes belong to a large TF family in P. tabuliformis, which includes 123 members in the genome (Fig. S3, Table S3, see online supplementary material). By the phylogenetic analysis of the NAC family in P. tabuliformis, both PtNAC3 and PtNAC5 belong to the SNAC/ATAF subfamily (Fig. S3, see online supplementary material), which has been widely reported to participate in abiotic stress in plants [20,21]. PtNAC3 has a typically conserved Nterminal NAC domain (Fig. 3A), in which the amino acid residues that are critical for DNA binding ability are identical with the ATAF1 and AtNAC3 in Arabidopsis (Fig. 3B). Consistent with its transcription factor function, the PtNAC3 protein was exclusively located in the nucleus of Arabidopsis protoplasts (Fig. 3C), and can activate the GAL4 reporter gene transcription in yeast (Fig. 3D). , salt (2-month-old seedlings were treated under 2 M NaCl for 3 days) and osmotic (2-month-old seedlings were treated under 300 mM mannitol for 3 days); needles were harvested for RNA-seq. The abiotic stresses driving TFs were harvested after differential expression analysis toward RNA-seq via edge R. B The Venn diagram analysis was conducted to determine the differential genes expressed in P. tabuliformis needles under five different abiotic stresses. C The histogram displays the number of abiotic stress induced/repressed genes in response to abiotic stress. The dates were visualized by GraphPad.

PtNAC3 positively regulates PtZFP30 by directly binding its promoter
PtZFP30 belongs to the C2H2 family, which also has been reported to participate in abiotic stress response in plants [38] (Fig. S5, Fig. S7, see online supplementary material). By co-expression analysis, we found the PtNAC3 co-expression TF PtZFP30 has a putative AtNAC3 binding motif in its promoter region [39], indicating PtNAC3 may directly target the PtZFP30 and form a cascade in the stress response pathway. To investigate this possibility, we used the yeast one-hybrid (Y1H) assay [40], dualluciferase reporter assay [41], electrophoretic mobility shift assay (EMSA) and CUT & Tag assay to test the role of PtNAC3 in the regulation of PtZFP30. The results showed that PtNAC3 binds to the PtZFP30 promoter in yeast and digested X-gal and let yeast turn blue ( Fig. 4A and B). To confirm whether PtNAC3 regulated the expression of the PtZFP30, we use Nicotiana benthamiana leaves to run a transient expression assay ( Fig. 4C and D). The result confirmed that PtNAC3 could activate the PtZFP30 promoter activity (Fig. 4E). Moreover, EMSA confirmed that PtNAC3 binds to the promoter of PtZFP30 containing the ACACGTAA motifs (Fig. 4F). An in vivo study was performed by a transient transformation system to transiently express 35S::PtNAC3-HA-GUS and 35S:: HA-GUS in the hypocotyl of P. tabuliformis. With CUT & Tag assay, the results of PCR and droplet digital PCR showed that PtNAC3 binds the promoter of PtZFP30 occurred in P. tabuliformis (G-I).
Taken together, these results indicate that PtNAC3 serves as the core responsor in the upstream of PtZFP30 in the abiotic stresses response pathway in P. tabuliformis.

The induction of PtNAC3 was associated with the ethylene pathway under abiotic stress
Phytohormones, especially for ABA-signaling pathways, have been well studied for their involvement in abiotic stress responses.
To determine whether there was any connection between phytohormones and PtNAC3-PtZFP30 abiotic stress response pathway, 2-month-old P. tabuliformis seedlings were treated with ABA, 1-amino cyclopropane carboxylic acid (ACC), methyl jasmonate (MeJA), salicylic acid (SA), strigolactones (GR24), indole-3-acetic acid (IAA), trans-zeatin (TZ), brassinolide (BR), gibberellin A3 (GA 3 ), gibberellin A4 and A7 (GA 4 + 7 ), and paclobutrazol (PAC). Based on RNA-seq analysis, the expression level of both PtNAC3 and PtZFP30 was significantly upregulated under ACC and MeJA treatment, but not response to ABA (Fig. 5A). As the PtNAC3 had the highest expression response under the salt treatment condition ( Fig. 2A), we then measured the endogenous hormone levels in P. tabuliformis seedlings which were treated with 2 M NaCl for 3 days. Compared with the control condition, ACC was significantly accumulated, but there were no significant differences in ABA, ABA-GE, SA, and SAG, the JA content was also not changed; however, the JA active form JA-ILE level was significantly decreased (Fig. 5B). To confirm the ACC response and determine the response time of PtNAC3 and PtZFP30, the timecourse assay was conducted and showed that both PtNAC3 and PtZFP30 were significantly induced by ACC within 8 hours and the effect lasted until 48 hours (Fig. 5C). Under ACC treatment, PtNAC3 and PtZFP30 also tightly co-expressed with the R 2 = 0.82 (Fig. 5D).

PtNAC3 confers tolerance to multiple abiotic stresses and promotes reproductive success by shortening the lifespan in Arabidopsis
As the PtNAC3 acts as a multiple abiotic stress response responder in P. tabuliformis, we were interested in whether it could be used as a biological tool that applied to plant breeding for unified abiotic stress tolerance. To investigate this possibility, we overexpressed PtNAC3 in Arabidopsis thaliana (Fig. S4, see online supplementary material). Unexpectedly, 35S:PtNAC3 overexpression lines (OEs) did not exhibit any typical negative traits like growth retardation and lower reproductive yields phenotype of overexpressing stress response TFs under normal conditions. On the contrary, compared with wild-type (WT), OEs were more vigorous, such as growing faster and higher (Fig. 6A), having more and longer siliques ( Fig. 6A and B) and higher yield (Fig. 6C). The OEs also showed early maturity traits by shortening lifespan (Fig. 6A, D and E), both leaf (Fig. 6B) and siliques (Fig. 6D) senescence faster in OEs than  Table S2 (see online supplementary material) under abiotic stress treatment were used to analyse correlation coefficients. GraphPad was used to calculate the correlation coefficients with default parameters.
WT. Moreover, OEs grow faster and stronger than WT under both normal and abiotic stress conditions (Fig. 6F-J), and promote reproductive success under all tested stress treatment including drought (Fig. 6H), cold (Fig. 6G), salt (Fig. 6I), and hot (Fig. 6J). After 14 days without water treatment, OEs grow faster and stronger than WT, and OEs siliques are less affected by drought while the yields of OEs are higher than WT (Fig. 6H). After 14 days of 4 • C treatment, both OEs and WT growth were affected, but OEs f lowered earlier and had more siliques compared with WT (Fig. 6G). After 14 days of 200 mM NaCl treatment, the siliques of WT were about to fall and seeds fail to develop normally, while OEs were less affected and harvested more seeds (Fig. 6I). After 7 days of 40 • C treatment, the siliques of WT were about to fall and seeds fail to develop normally, while OEs were less affected (Fig. 6J). These results suggest that PtNAC3 has great potential as a biological tool in transgenic plants and confers tolerance to various abiotic stresses without growth penalty.

Discussion
Since the Green Revolution, many efforts had been taken to improve the crop's tolerance to stress [42]. With the increase of the greenhouse effect, abiotic stress had received more attention [42,43]. The core regulatory pathways of plants' abiotic stress response in angiosperm were elucidated [4,44]. As widely distributed species in the northern hemisphere, however, the molecular mechanism of conifer response to abiotic stress is largely unknown. In the present study, we identified 32/16 genes induced/repressed by all five common abiotic stress which includes four/two TFs ( Fig. 1B and C). In the study of model plants, it was found that CPKs and MAPKs family genes play a crucial role in response to multiple abiotic stress [5,45], but we observed it is TFs instead of protein kinases that participate in abiotic stress response in P. tabuliformis. NAC, MYB as well as C2H2 transcription factors were widely reported in responses to environmental pressure in plants [38,43,46]; in this study we found PtNAC3 induced by all common abiotic stress more obviously than other TFs (Fig. S1, see online supplementary material), implying PtNAC3 is important in abiotic stress response. In recent years, the C2H2type zinc finger protein family was also identified in response to abiotic stress, PeSTZ1 was reported to enhance freezing tolerance in Populus euphratica [38]; SlZF3 has been reported to enhance salt tolerance in tomato [47]; ZFP183 was reported involved in abscisic acid induced abiotic stress defense [48]. PtZFP30, a member of the ZFP transcription factor family, may play an important role in abiotic stress response. Previous studies found ASITFs responded to abiotic stress sensitively [21,49,50]. In the study, we found PtNAC3 up-regulates under drought treatment and down-regulates to a common level after rewatering a day. This interesting pattern also disappeared under UVB treatment: PtNAC3 sharply up-regulates and gradually down-regulates ( Fig. 2A). In the annual period, PtNAC3 and PtZFP30 up-regulate significantly in winter, because the winter in Beijing is extremely cold with drought all day long (Fig. 2B). The results suggested that PtNAC3 and PtZFP30 are important for P. tabuliformis to adapt to winter. However, they can also respond to multiple abiotic stresses. Combined with the data of daily periodicity, we found PtNAC3 responded to abiotic stress very rapidly, but when abiotic stress disappeared, the expression profile of PtNAC3 will immediately recover normally (Fig. S2, see online supplementary material). Although the expression level of PtZFP30 is much lower than PtNAC3 when plants suffer abiotic stress, PtZFP30 showed a strong correlation with PtNAC3 under abiotic stress by co-expression analysis (Fig. 2C). All the features mentioned above illustrate PtNAC3 was sensitive to abiotic stress, and was suitable for being a bio-marker to check whether P. tabuliformis is under stress. Although there are many homologous genes in SNAC group TFs in P. tabuliformis (Fig. 3A), only PtNAC3 and PtNAC5 were found to respond to multiple abiotic stress. The function of other TFs needs further research.
In the past few years, genes belonging to the SNAC sub-family have been discovered, most of them were nucleic localization and acted as transcription activators [28,31,51]. In this study, we found PtNAC3 is nucleic localization (Fig. 3C), implying the NAC3-ZFP30 module may work in the nucleus; our result also showed PtNAC3 activates the GAL4 reporter gene in yeast (Fig. 3D), indicating PtNAC3 may activate PtZFP30. SNAC TFs had been reported to respond to abiotic stress in many cases; however, the downstream genes of SNAC TFs were less reported. In the study, Y1H assay, dual-luciferase assay, CUT & Tag and EMSA were performed to prove PtNAC3 can bind the promoter of PtZFP30 (Fig. 4), suggesting PtZFP30 is one of the downstream genes of PtNAC3.
Hormones, especially ABA and JA, participate in many abiotic stress response pathways, and attention had been paid to  Table S1 (see online supplementary material). H-I Droplet digital PCR was used to further confirm that PtNAC3 can bind the PtZFP30 promoter in vivo.
clarifying the regulatory mechanism of ABA and JA pathways in abiotic stress [52][53][54]. In angiosperm, SNAC group TFs participated in abiotic stress signaling pathway dependent on ABA [55]; however, since being treated with various hormones, we found that PtNAC3 and PtZFP30 responded only to ACC and MeJA treatment, while PtNAC3 and PtZFP30 showed the most obvious response under salt stress. Therefore, we chose salt stress as a representative to verify the hormone content in P. tabuliformis under abiotic stress. Under salt treatment, the content of ABA in P. tabuliformis was stable, but ACC accumulated significantly. After ACC treatment for 8 hours, PtNAC3 and PtZFP30 began to be up-regulated until 48 hours (Fig. 5). These results showed that ACC, but not ABA, was involved in the NAC3-ZFP30-mediated abiotic stress signaling pathway in P. tabuliformis. However, whether such hormonal changes are similar under other abiotic stresses deserves further investigation. Our data provide valuable information for understanding the molecular mechanism of abiotic stress response in conifers. The results suggested that the NAC3-ZFP30 module acting as a responder response to multiple abiotic stress was ethylene-induced and provided new insight into the study of ancient plants' response to abiotic stress.  Table S1 (see online supplementary material). Error bars are SD (n = 3). Statistics, t tests ( * * P < 0.01, * * * P < 0.001). D The co-expression relativity between PtNAC3 and PtZFP30 in P. tabuliformis phytohormone treatment. The transcript date of PtNAC3 and PtZFP30 under phytohormone treatment shown in Table S2 (see online supplementary material) were used to analyse correlation coefficients. GraphPad was used to calculate the correlation coefficients with default parameters.
In the study, we also found PtNAC3 has great potential to be a biological tool in crop breeding. Previous studies found many SNAC group TFs overexpressed in plants not only enhanced plants' abiotic stress tolerance but also improved yield [26,56,57]. In some studies, SNAC group TFs also reported related to leaf senescence [40,43,58]. Our study suggested overexpressed PtNAC3 in Arabidopsis enhanced abiotic stress tolerance, improved yield as well as promoted leaf senescence (Fig. 6). PtNAC3 overexpressed plants seem to gather all the advantages SNAC group TFs had reported, which is suitable for being a biological tool.

Plant materials and treatments
The P. tabuliformis seedlings were planted in 6 cm pots with turfy soil (PINDSTRUP) and grew in the greenhouse (16 h light/8 h night; 21 ± 1 • C) for two months. The ecotype Columbia Arabidopsis (WT) and 35S::PtNAC3 transgenic lines were planted in pots with the turfy soil mixed with the nutrient soil Fangjie (2:1 v/v). N. benthamiana was planted in pots containing turfy soil mixed with vermiculite (1.2:1 v/v).
For abiotic stress treatment of Arabidopsis, the WT and 35S::PtNAC3 grew in the greenhouse for 4 weeks before f lowering, and then treatments were started; for salt treatment, 1 L 200 mM NaCl was watered in a tray twice a week for two weeks; for drought, treatment kept the tray without water for two weeks; for cold treatment, transfer the Arabidopsis to a cold treatment greenhouse (16 h light/8 h night; 4 ± 1 • C) for two weeks; for hot treatment, transfer the Arabidopsis to a cold treatment greenhouse (16 h light/8 h night; 40 + 1 • C) for a week. G Four-week-old Arabidopsis grew in the greenhouse and were transformed to cold treatment (4 • C), 35S::PtNAC3 grew faster, had more siliques and higher yield than WT after cold treatment for two weeks. H Four-week-old Arabidopsis grown in the greenhouse were transformed to drought treatment (without water), 35S::PtNAC3 grew faster, had more siliques and higher yield than WT after drought treatment for two weeks. I Four-week-old Arabidopsis grew in the greenhouse and were transformed to salt treatment (200 mM NaCl), 35S::PtNAC3 grew faster, had more siliques and higher yield than WT after salt treatment for two weeks. J Four-week-old Arabidopsis grew in the greenhouse and was transformed to hot treatment (40 • C), 35S::PtNAC3 grew faster, and had more siliques than WT after hot treatment for a week.

DNA/RNA extractions and cDNA synthesis
P. tabuliformis genomic DNA was extracted from the needles following the protocol of the Plant Genomic DNA kit (TIANGEN, Beijing, China). Total RNA from different tissues of P. tabuliformis or Arabidopsis was extracted following the protocol of the Vazyme Plant Total RNA Isolation Kit (Vazyme, Nanjing, China) and stored at −80 • C. 1 μg total RNA was used for first-strand cDNA samples generated by the Vazyme kit (Vazyme, Nanjing, China).

RNA-sequencing, qRT-PCR, detection of phytohormones, and bioinformatics analysis
The RNA-seq analysis followed the protocol of [37,59]. The clean reads were mapped to the P. tabuliformis reference genome [19] and the transcript abundances were calculated by Kallisto software [60]. The RNA-seq data used in this study were listed in Table S2 (see online supplementary  material).
The qRT-PCR was performed by QuantStudio 6 Rex (Thermo Fisher Scientific) with the protocol of TSINGKE qPCR Mix. Primer Primer5 (www.PremierBiosoft.com) was used to develop the primers listed in Table S1 (see online supplementary material).
The phytohormones contents were detected by LC-MS/MS. After 2 M NaCl was irrigated daily for 3 days, 2 g needles were collected from 2-month-old seedlings. Rapid grinding of samples into powder was performed by liquid nitrogen freezing, 50 mg samples were used and extractant added containing 75% methanol and 5% formic acid. After 12 000 g, 5 min centrifugation at 4 • C, the supernatant was removed. The precipitate was re-dissolved by 100 μL 80% methanol and passed through the 0.  [63][64][65]. For MS/MS, the source temperature of ESI was 550 • C, the mass spectrometry voltage was set to 5500 V/ -4500 V as positive/negative ion mode, and the CUR was set to 35 psi [66][67][68].
Upset figure, heat map and gene structure analysis used the TBtools software (www.tbtools.come). SNAC and C2H2-B subfamily proteins conserved domains were identified by NCBI CCD Tools, and their motifs were predicted by MEME which selected classic mode to find five motifs, multiple alignments used DNAMAN v8 (https://www.lynnon.com/), phylogenetic analysis using MEGA software with ML tree which was based on the JTT model, the bootstrap values were 200 bootstrap replicates. Microsoft Office (https://www.microsoft.com/zh-cn), R program and GraphPad Prism 8 were used to analyse the experimental data.

Subcellular localization analysis
The recombined plasmid pHBT-NAC3-mCherry was extracted using the CsCl gradient method to get purified plasmid, and the protocol was from the Sheen lab website. The protoplast transient expression assay followed the protocol as described previously [69,70]. The 10 μL 20 mg/mL of DAPI was added into 1 mL W5 solution with transfected protoplast for 10 minutes to indicate nuclei. The confocal microscope (Leica SP8) was used to detect f luorescence signals (mCherry, excitation 587 nm, emission 600-620 nm; GFP, excitation 480 nm, emission 510-550 nm; DAPI, excitation 405 nm, emission 430-480 nm).
The recombined plasmid pB42AD-PtNAC3 and pLacZ2μ-PtZFP30p were co-transformed into yeast EGY48 strains (Clontech; CAT#: YC1030) and grew on the SD-Trp/-Ura medium for 4 days at 28 • C and transfer to SD/−Trp/-Ura + Gal+Raf + X-gal for 3 days at 28 • C. The pLacZ2μ and the pB42AD, pLacZ2μ and the pB42AD-PtNAC3, the pLacz2μ-PtZFP30p and the pB42AD were used as the negative control.
The recombined plasmid pGBKT7-PtNAC3 and pGADT7-T were co-transformed into yeast Y2HGold strains (Clontech; CAT#: YC1002) and grew on the SD-Leu/−Trp medium for 2 days at 30 • C. The single colonies were transferred to SD-Trp-Leu-His-Ade medium for 5 days at 30 • C.

The electrophoretic mobility shift assay (EMSA)
The GST and PtNAC3-GST proteins were expressed by BL21(DE3) cells following the protocols [71]. The PtZFP30 promoter probes containing ACACGTAA motifs (Table S1, see online supplementary material) which were 5 biotin-labeled or unlabeled as control were synthesized by Beijing Tsingke. The EMSA luminous kit (GS009, Beyotime) was used to experiment. The images were taken by the BIORAD ChemiDocTM MP imaging system.

CUT & Tag assay
The transient transformation assay in the hypocotyl of P. tabuliformis was carried out according to the method of Liu [72]. The cell nucleus extraction followed the step of CelLytic™ PN Isolation/Extraction Kit (CELLYTPN1; Sigma). The CUT & Tag assay was performed by NovoNGS CUT&Tag 3.0 High-Sensitivity Kit (Cat. No: N259-YH01) to construct the libraries. The droplet digital PCR (ddPCR) was performed with a 1 μl template in which CUT & Tag libraries were diluted five times. The ddPCR reagents were used from BioRad.

The generation of transgenic lines
The recombined plasmid pBI121-PtNAC3 was transformed into Agrobacterium strain EHA105 (Shanghai Weidi Biotechnology Co., Ltd; CAT#: AC1010), and 35S::PtNAC3 overexpressed lines were generated by the f loral-dip method [68,73], and homozygous lines were screened by 50 mg/L Kana. MS mediums. Three 35S::PtNAC3 lines were used for further research, and 30 events were used to measure WT and transgenic lines' lifespan.