PdEPF1 regulates water‐use efficiency and drought tolerance by modulating stomatal density in poplar

Summary

Water deficiency is a critical environmental condition that is seriously reducing global plant production. Improved water‐use efficiency (WUE) and drought tolerance are effective strategies to address this problem. In this study, PdEPF1, a member of the EPIDERMAL PATTERNING FACTOR ( EPF ) family, was isolated from the fast‐growing poplar clone NE‐19 [Populus nigra × (Populus deltoides × Populus nigra)]. Significantly, higher PdEPF1 levels were detected after induction by dehydration and abscisic acid. To explore the biological functions of PdEPF1, transgenic triploid white poplars (Populus tomentosa ‘YiXianCiZhu B385’) overexpressing PdEPF1 were constructed. PdEPF1 overexpression resulted in increased water deficit tolerance and greater WUE. We confirmed that the transgenic lines with greater instantaneous WUE had approximately 30% lower transpiration but equivalent CO ₂ assimilation. Lower transpiration was associated with a 28% reduction in abaxial stomatal density. PdEPF1 overexpression not only strongly enhanced WUE, but also greatly improved drought tolerance, as measured by the leaf relative water content and water potential, under limited water conditions. In addition, the growth of these oxPdEPF1 plants was less adversely affected by reduced water availability than plants with a higher stomatal density, indicating that plants with a low stomatal density may be well suited to grow in water‐scarce environments. Taken together, our data suggest that PdEPF1 improves WUE and confers drought tolerance in poplar; thus, it could be used to breed drought‐tolerant plants with increased production under conditions of water deficiency.

Introduction

Water is essential for plant growth because it provides a medium for many cellular functions. Biomass production and accumulation are negatively affected by water deficits because such conditions reduce leaf cell turgor, restrict cell expansion, and postpone development (Chaves et al., 2003). With global climate change, water resources will be a critical factor in plant productivity (Marshall et al., 2012). An efficient strategy for plant tolerance to drought is the reduction of water loss through transpiration. However, decreasing transpiration usually leads to lower biomass accumulation due to reduced carbon assimilation (Sinclair et al., 1984; Udayakumar et al., 1998). Land plants must balance CO₂ absorption and water loss to maximize fitness. The biomass produced per unit water consumption or the ratio of the CO₂ absorption rate to transpiration is referred to as water‐use efficiency (WUE) (Farquhar and Sharkey, 1982, Farquhar et al., 1989; Dawson et al., 2002).

Gas exchange occurs primarily through stomata, which are surrounded by two guard cells (Hetherington and Woodward, 2003). Plants regulate transpiration and water use by managing stomatal conductance. Stomatal conductance is determined by stomatal movement and density (Casson and Hetherington, 2010; Chaerle et al., 2005; Huang et al., 2009; Song and Matsuoka, 2009; Yoo et al., 2009). Stomatal movement is a rapid response induced by many environmental factors, including abscisic acid (ABA), blue light, Ca²⁺, CO₂, and water status (Chater et al., 2014; Kim et al., 2010; Shimazaki et al., 2007). Regulation of the number of stomata is a longer‐term response. Stomatal density is adjusted in developing leaves through the sensing of environmental conditions by mature leaves (Lake et al., 2001; Miyazawa et al., 2006). Under conditions of water stress, stomatal density responses differ depending on the drought intensity and species. Reduced stomatal density due to drought stress is present in wheat, squash cotyledons, and umbu trees (Quarrie and Jones, 1977; Sakurai et al., 1986; Silva et al., 2009). In rice and Leymus chinensis leaves, an increase in stomatal density was observed in response to moderate drought, but a decrease occurred in response to more severe drought (Xu and Zhou, 2008).

In Arabidopsis, peptides play important roles during stomatal development. EPIDERMAL PATTERNING FACTOR (EPF) and EPIDERMAL PATTERNING FACTOR‐LIKE (EPFL) encode a cysteine‐rich secreted peptide family whose members have six or eight conserved carboxy‐terminal cysteines that form disulphide bridges (Ohki et al., 2011). There are eleven members (EPF1, EPF2, and EPFL1–9) that participate in diverse developmental processes in Arabidopsis (Hara et al., 2007, 2009; Rowe and Bergmann, 2010; Rychel et al., 2010; Torii, 2012). Four EPFf members, EPF1 (At2g20875), EPF2 (At1g34245), CHALLAH (CHAL/EPFL6: At2g30370), and STOMAGEN/EPFL9 (At4g12970), could affect stomatal development; however, only EPF1 and EPF2 are specifically expressed in stomatal lineage cells. EPF1 is produced in late‐stage meristemoids, guard mother cells (GMCs), and young guard cells (Hara et al., 2007). EPF1 functions as an orientation signal to provide positional information about the location of stomatal precursors and stomata relative to one another. The epf1 mutation produces occasionally paired stomata that do not observe the one‐cell spacing rule. Overexpression of EPF1 suppresses the transition from meristemoid to GMC, resulting in the transdifferentiation of meristemoids into pavement cells. EPF1 conveys positional information to promote and enforce the one‐cell spacing rule (Hara et al., 2009; Lee et al., 2012). EPFf peptides interact with three leucine‐rich repeat receptor‐like kinases [ERECTA (At2g26330), ERECTA‐LIKE1 (At5g62230), and ERECTA‐LIKE2 (At5g07180)] and a leucine‐rich repeat receptor‐like protein called TOO MANY MOUTHS (At1g80080). These four receptors contain extracellular leucine‐rich repeat domains. Biochemical studies using a biosensor chip have shown ER‐EPF2 and ERL1‐EPF1 as the predominant ligand‐receptor pairs, where ER‐EPF2 regulates entry divisions and ERL1‐EPF1 regulates spacing divisions (Lee et al., 2012).

Forest trees are a major renewable resource that has long provided materials and energy. Trees also provide environmental benefits, including carbon sequestration and participation in the water cycle (Song et al., 2006). Poplars are significant trees with great economic and ecological importance in temperate regions. They show high productivity, but exhibit a strong demand for water (Monclus et al., 2006). Environmental abiotic stress, especially drought, could negatively affect poplar growth (Tschaplinski et al., 1994). In some dry or semi‐dry areas, including northern China, high WUE acts as a determining factor in breeding trees. Over the last few decades, a series of studies have indicated that changing the process of stomatal development impacts WUE and drought tolerance (Yoo et al., 2010). Most studies have focused on Arabidopsis, with few investigations of woody plants (Han et al., 2013; Xing et al., 2011).

In our study, a member of the EPF family, PdEPF1, which is involved in stomatal development in poplars, was cloned from the high‐WUE poplar genotype NE‐19 [Populus nigra × (Populus deltoides × Populus nigra)]. Using Agrobacterium tumefaciens‐mediated transformation, PdEPF1 was transferred into triploid white poplar (Populus tomentosa ‘YiXianCiZhu B385’) (Zhu et al., 1998) and transgenic seedlings were created. The purpose of our study was to develop transgenic plants with a reduced stomatal density and increased WUE. The drought tolerance of poplar trees could be enhanced through this method.

Results

Molecular characterization of PdEPF1

Based on our previous study, we selected the high‐WUE genotype NE‐19 from seven different Populus genotypes (data not shown) to clone PdEPF1 (GenBank accession number: KF530817) for further characterization. The PdEPF1 cDNA is 360 bp in length and encodes 119 amino acids. A comparison showed that PdEPF1 shares 54.6% identity with AtEPF1.

To identify proteins homologous to PdEPF1, a phylogenetic tree was constructed for poplar and Arabidopsis EPFL family members following an amino acid sequence alignment (Figure 1a). As shown in Figure 1a, PdEPF1 clustered with AtEPF1. A multiple sequence alignment revealed that PdEPF1 has six conserved cysteine residues at the C‐terminal end, similar to other members of the EPFL family (Figure 1b), and two additional ones found only in AtEPF1, AtEPF2, and AtEPFL7 (Hara et al., 2009; Shimada et al., 2011).

Figure 1.

The PdEPF1 gene of NE19. (a) Phylogenetic relationships between poplar and Arabidopsis EPF family members. (b) Multiple alignment of amino acid sequences of PdEPF1 and other plant EPFs.

PdEPF1 is involved in plant water stress responses

To investigate the response of PdEPF1 in Populus to water deficit, the expression level of PdEPF1 in response to dehydration was tested using RT‐PCR and qRT‐PCR. Our results show that the expression level of PdEPF1 gradually rose with stress duration and peaked after 6 h (Figure 2a). ABA is an important hormone in the response to various stresses, and it mediates some drought‐responsive genes. When plant roots are subjected to water stress, ABA accumulation may be initiated by a drought‐sensing mechanism located in the roots; from there, it can be exported to leaves (Pei and Kuchitsu, 2005). Thus, ABA treatment was used to test the response of PdEPF1 in poplar. PdEPF1 expression increased about twofold after 1 h and then decreased sharply after 6 h (Figure 2b).

Figure 2.

Expression analysis of PdEPF1 under dehydration and ABA in Populus. (a) qPCR and PCR assay of accumulation of PdEPF1 transcripts in response to dehydration. (b) qPCR and PCR assay of accumulation of PdEPF1 transcripts in response to ABA. (c) Cis elements in the 5’ flanking region of the PdEPF1 gene.

To assess the mechanism underlying the inducible expression of PdEPF1 under water stress, a 1090‐bp PdEPF1 promoter fragment was isolated by adaptor PCR from NE‐19 genomic DNA. A sequence analysis revealed that the PdEPF1 promoter contains CAAT and TATA motifs near the transcriptional start site (Figure 2c). A predictive analysis of the PdEPF1 promoter using the PLACE database (http://www.dna.affrc.go.jp/PLACE/) was performed to label a series of water‐related stress‐responsive and development‐associated elements, including one LTRECORE (an ABA‐responsive element), three TAAAGSTKST1 (expressed in guard cells), two AGCT (involved in the response to early drought), and eight MYC (a cis‐acting element involved in drought responsiveness) elements. An analysis of our RT‐PCR, qRT‐PCR, and predictive results indicated that PdEPF1 expression was induced in response to water deficit and ABA.

Analysis of transgenic poplar overexpressing PdEPF1

To assess the role of PdEPF1 in water stress tolerance, transgenic poplar plants overexpressing PdEPF1 under the control of the CaMV 35S promoter were generated. Nine transgenic lines were confirmed by PCR using specific primers, all of which exhibited the expected band (Figure 3a). The expression levels of PdEPF1 among the different transgenic lines were different; because oxPdEPF1‐28 and ‐39 expressed higher levels than those of the other lines (Figure 3b), we selected these two lines for physiological and drought stress experiments. DNA hybridization showed that PdEPF1 was integrated into the genomes of the transgenic plants as one (oxPdEPF1‐5, ‐16, ‐18, ‐25, ‐28, ‐30, ‐37, and ‐41) or two copies (oxPdEPF1‐39), whereas no hybridization was observed in wild type (Figure 3c). In the physiological studies, the wild‐type line derived from the same regeneration experiment as the transgenic lines was used as a control.

Figure 3.

Analysis of the transgenic poplar plants overexpressing PdEPF1. PCR analysis was carried out in 9 putatively mannose‐resistant poplars for the presence of PdEPF1 genes (a). qRT‐PCR was used for the analysis of PdEPF1 transcription (b). Thirty microgram of genomic DNA from each plant line was digested with Eco RI for DNA blot hybridization (c).

PdEPF1 reduces stomatal density and changes stomatal size

Overexpressing PdEPF1 in poplar decreased the stomatal density on the abaxial leaf surface without relation to age or leaf part (Figure 4). The stomatal number per unit area in oxPdEPF1 lines was 24.1–28% lower than in wild‐type immature leaves (Figure 4d). In the functional part of the leaf, the stomatal density in the oxPdEPF1 lines was decreased by 26.2–50% compared to wild‐type plants (Figure 4d). In addition to stomatal density, stomatal size was also changed in the transgenic lines (Figure 5). The stomata in the oxPdEPF1 lines were 27.3 and 36.8% longer than in wild type, while no visible difference in width was detected among the lines (Figure 5d).

Figure 4.

Leaf stomatal density in poplar. Scanning electron micrograph of the abaxial leaf epidermis of wild‐type (a), oxPdEPF1‐28 (b), and oxPdEPF1‐39 (c). Scale bars represent 100 μm. (d) The abaxial stomatal densities in wild‐type and transgenic plants. M is mean mature leaf. Data are mean±SE (n = 9). Asterisks denote significant differences: **P ≤ 0.01

Figure 5.

Leaf stomatal size in poplar. Scanning electron micrograph of the abaxial leaf epidermis of wild‐type (a), oxPdEPF1‐28 (b), and oxPdEPF1‐39 (c). Scale bars represent 10 μm. (d) The abaxial stomatal size in wild‐type and transgenic plants. M is mean mature leaf. Data are mean±SE (n = 9). Asterisks denote significant differences: **P ≤ 0.01

PdEPF1 overexpression improves WUE by reducing transpiration

To study the effects of stomatal density on poplar, leaf gas exchange (water and CO₂) was examined using an infrared gas analyser under a range of external CO₂ and light concentrations. Photosynthesis‐CO₂ and photosynthesis‐light curves indicate that the oxPdEPF1 plants resembled wild type in their photosynthetic responses (Figure 6a and b). The stomatal conductance data show that the oxPdEPF1 plants had altered stomatal conductance (G_s) responses to CO₂ and light compared to wild type. The G_s of the transgenic plants was lower than that of wild type under both conditions (Figure 6c and d). Leaf transpiration in the oxPdEPF1 plants was 24.8–29.8% lower than in wild‐type plants at saturating CO₂ levels (Figure 6e). At saturating light levels, leaf transpiration was 26.9–31.1% lower than in wild type (Figure 6f). Consequently, the oxPdEPF1 plants had higher instantaneous WUE (photosynthesis/transpiration) values (Figure 6g and h). This was attributable to the reduced transpiration in oxPdEPF1 plants (Figure 6e and f). The vapour pressure deficit (VPD) was similar under both conditions (Figure 6i and j), indicating that the reduced transpiration in the transgenic plants was not due to differences in VPD.

Figure 6.

OxPdEPF1 plants exhibit higher instantaneous WUE due to lower transpiration and stomatal conductance. (a) A‐C_i curve. (b) A‐light curve. (c) G_s‐Ci curve. (d) G_s‐light curve. (e) Transpiration‐C_i curve. (f) Transpiration‐light curve. (g) Instantaneous WUE‐C_i curve. (h) Instantaneous WUE‐light curve. (i) VPD‐C_i curve. (j) VPD‐light curve. Data are mean±SE (n = 6). Asterisks denote significant differences: *P ≤ 0.05; **P ≤ 0.01

PdEPF1‐overexpressing plants exhibited increased drought tolerance under short‐term drought stress

To observe the differences between oxPdEPF1 and wild‐type plants under water deficit, 12 days of drought stress were imposed by withholding water. On day 12, the leaves of the wild‐type plants were wilted, while those of the oxPdEPF1 plants remained turgid (Figure 7a). Figure 7b–d shows the variation in photosynthesis, G_s, and transpiration between the transgenic and wild‐type plants during the 12 days of drought stress. During the first 8 days, there were no significant differences between the plants in CO₂ assimilation rate (Figure 7b). From day 8 to 9, however, the wild‐type plants exhibited a more acute decline than the oxPdEPF1 plants. Stomatal conductance in the oxPdEPF1 plants began to decline slowly on day 8, while that in the wild‐type plants fell rapidly (Figure 7c). Transpiration in the transgenic plants also showed a gentle decline compared to the steep decline seen in wild type (Figure 7d). Therefore, the oxPdEPF1 plants exhibited better performance than wild type under short‐term drought stress.

Figure 7.

Overexpression of PdEPF1 confers drought tolerance in poplar during short‐time drought. (a) Morphological differences in short‐time drought experiments. (b) A‐drought time curve. (c) Gs‐drought time curve. (d) Transpiration‐drought time curve. Data are mean ± SE (n = 6). Asterisks denote significant differences: *P ≤ 0.05; **P ≤ 0.01

PdEPF1‐overexpressing plants showed increased drought tolerance under long‐term water deficit

To further explore the effects of decreased stomatal density on poplar during water deficit, a 40‐day drought stress test was applied to wild‐type and oxPdEPF1 plants by maintaining the soil relative water content (RWC) at a constant level (70, 45, and 20%). After 40 days, the condition of the wild‐type plants was poorer than that of the overexpressing plants at 45 and 20% soil RWC (Figure 8a). The better performance of the oxPdEPF1 plants was associated with the capacity to maintain a higher leaf RWC than wild type under mild to severe stress (Figure 8b). The leaf water potential also differed significantly between the transgenic lines and wild type under drought stress. The oxPdEPF1 plants had a higher leaf water potential compared with wild type under drought stress (Figure 8c).

Figure 8.

Overexpression of PdEPF1 confers drought tolerance in poplar during long‐term drought. (a) Morphological differences in long‐term drought experiments. (b) Leaf relative water content under three water statuses. (c) Leaf water potential under three water statuses. Data are mean ± SE (n = 6). Asterisks denote significant differences: *P ≤ 0.05; **P ≤ 0.01

An analysis of photosynthesis showed that the transgenic plants maintained a significantly higher photosynthetic rate than wild type at a low soil RWC (Figure 9a). Chlorophyll is an important component of thylakoid pigment–protein complexes. The function of pigment–protein complexes is affected if the chlorophyll content decreases, thereby reducing the absorption of light energy by chloroplasts. The chlorophyll content differed under changing water conditions (Figure 9b–d). At 70% soil RWC, the total chlorophyll content of the wild‐type plants was similar to that of the oxPdEPF1 plants. Under drought stress, the oxPdEPF1 plants had a relatively higher chlorophyll content than wild‐type plants (Figure 9b). This change was due to large differences in the chlorophyll a content between wild‐type and transgenic plants, with only subtle differences in chlorophyll b content (Figure 9c and d). Thus, decreases in stomatal density due to PdEPF1 overexpression led to strong drought tolerance in poplar.

Figure 9.

Physiological analysis of wild‐type and overexpressing PdEPF1 lines under long‐term drought experiments. The photosynthetic rate (a), content of chlorophyll a and b (b), chlorophyll a (c), and chlorophyll b (d) were measured. Data are mean±SE (n = 6). Asterisks denote significant differences: *P ≤ 0.05; **P ≤ 0.01

To observe the difference in growth between transgenic and wild‐type plants, we monitored WUE and shoot elongation (Figure 10). The oxPdEPF1 plants exhibited greater WUE than wild type at all water levels (Figure 10a). In accordance with the photosynthetic rate, the transgenic plants had a significantly higher overground biomass than that of wild type at a low soil RWC, while there was no difference among the lines under well‐watered conditions (Figure 10b). Under mild stress, the biomass of the transgenic lines was 37 and 30% higher, respectively, than that of wild type, while under severe stress, the advantage was 32.6 and 40%. Under well‐watered conditions, growth as measured by shoot height was similar in the oxPdEPF1 and wild‐type plants (Figure 10c). However, under mild stress, stem elongation in the transgenic plants was greater than that in wild type after 15–25 days of treatment (Figure 10d). Under severe stress, the oxPdEPF1 plants grew better than wild type (Figure 10e). Hence, PdEPF1 overexpression in poplar benefits growth under water stress conditions.

Figure 10.

Growth analysis of wild‐type and overexpressing PdEPF1 lines under long‐term drought experiments. (a) Aboveground plant integral WUE. (b) Aboveground plant biomass. (c–e) Stem height growth rate under no water stress (c), mild water stress (d), and severe water stress (e). Data are mean±SE (n = 6). Asterisks denote significant differences: *P ≤ 0.05; **P ≤ 0.01

Discussion

PdEPF1 was cloned from poplar. Multiple alignments revealed that PdEPF1 has high AA sequence homology to PtEPF1 and AtEPF1 (Figure 1). PdEPF1 contains eight conserved cysteine residues at the C‐terminal end, similar to AtEPF1. PdEPF1 also has a similar function, resulting in reduced stomatal density on the abaxial leaf surface without relation to leaf part or age (Figure 4). In addition, transgenic poplar overexpressing PdEPF1 had significantly enhanced WUE and drought tolerance due to the reduction in water loss from leaves.

The improved instantaneous WUE was attributed to lower transpiration but equivalent CO₂ assimilation. The modification of stomatal density directly resulted in a significant reduction in stomatal conductance (Figure 6c and d). Generally, stomatal conductance is positively related to transpiration. Transpiration in the transgenic plants declined as stomatal conductance fell (Figure 6e and f). Another transpiration factor, VPD, was similar for all lines under all conditions (Figure 6i and j), indicating that the reduced transpiration in the transgenic plants was not due to differences in VPD. Therefore, the reduction in transpiration was attributed to the reduction in stomatal density.

The reduced stomatal density was integral to the WUE in plants of the same biomass in the form of reduced water consumption. Total water loss depends on stomatal conductance, epidermal properties, and total leaf area. On the other hand, biomass is determined by the carbon gain in photosynthetic tissues, which is regulated by the net assimilation rate and functional leaf area. Eliminating the common factor, leaf area, the improved integral WUE was the result of changes in the other components. Similar rates of photosynthesis resulted in a similar biomass, while low stomatal conductance resulted in less water loss.

An interesting observation is that the reduced stomatal density did not lead to a decrease in photosynthesis (Figure 6a and b). A possible explanation is that the entry of CO₂ was less affected by stomatal conductance than was the release of water vapour. The photosynthetic rate of most plants is limited due to nonstomatal factors such as the regeneration of ribulose‐1,5‐bisphosphate (Farquhar and Sharkey, 1982), and a modest decrease in stomatal density does not affect photosynthesis (Bacso et al., 2008).

Previous studies on the alteration of SDD1 expression levels to increase or decrease stomatal density showed that stomatal conductance could be compensated for by adjusting the size of the stomatal aperture (Bussis et al., 2006). Longer stomas were observed on oxPdEPF1 poplar leaves (Figure 5). However, our results suggest that PdEPF1 was unable to completely compensate for the altered stomatal density through the adjustment of stomatal size. Under differing light and CO₂ gradients, oxPdEPF1 leaves had lower transpiration rates than wild‐type leaves (Figure 6e and f). Thus, our results indicate a correlation between stomatal density and transpiration rate.

In the current experiments, PdEPF1 was induced in response to dehydration stress and ABA treatment (Figure 2a and b), supporting its role in stress tolerance. An element analysis of the promoter indicated that several stress‐responsive elements are present in the PdEPF1 promoter region (Figure 2c). The role of PdEPF1 in the regulation of drought tolerance was validated in this study using transgenic poplar. Transgenic poplar overexpressing PdEPF1 exhibited decreased water loss and significantly increased tolerance to drought.

While drought stress resulted in decreased photosynthesis and reduced chlorophyll a content in all plants, the transgenic plants displayed significantly higher levels of photosynthesis and chlorophyll a (Figure 9a and c). Our results suggest that poplars with reduced stomatal density may have a growth advantage under water stress. The transgenic poplars had reduced transpiration, a greater biomass, higher WUE, and increased growth rate at 45 and 20% soil RWC (Figure 10). The improved growth rate of the transgenic plants may be attributed to the improved tissue water status (Figure 8b and c). The reduction in stomatal density in the oxPdEPF1 plants probably limited water loss and played a major role in maintaining higher water potential, even when the plants were severely stressed. This may be the reason why oxPdEPF1 plants are more tolerant to drought stress. The overexpression of PdEPF1 in poplar increases the capacity to preserve available water inside leaves under drought stress, resulting in less damage than observed in wild‐type plants under drought stress.

In general, reducing stomatal conductance is a water conservation strategy, but it may lead to decreases in cumulative photosynthetic activity and limit biomass production (Tardieu, 2012). However, our oxPdEPF1 plants grew and developed relatively normally under unstressed conditions (Figure 10b and c). This trait would be highly desirable for plant improvement. Unlike other stress tolerance genes such as those encoding DREB factors or AtHB‐7 (Kasuga et al., 1999; Mishra et al., 2012), the overexpression of PdEPF1 did not lead to growth retardation. In addition, PdEPF1 had a smaller coding sequence relative to other stress tolerance genes. We suggest that PdEPF1 is a good candidate gene to improve drought tolerance in transgenic poplar through the modulation of stomatal development.

PdEPF1 may play multiple roles in poplar, including drought responses, ABA regulation, and stomatal development. Our current experimental evidence provides information regarding the function of an uncharacterized gene within the poplar genome. Our data indicate that the overexpression of PdEPF1 reduces transpiration by decreasing the stomatal density without affecting CO₂ assimilation under well‐watered conditions. Our findings also demonstrate that PdEPF1 overexpression leads to better water maintenance in leaves. It also leads to a higher chlorophyll a content and photosynthesis and growth rates in transgenic plants under water stress. We suggest that the insertion of PdEPF1 into poplar, an important forestry species, could be useful in the improvement of WUE and drought tolerance.

Experimental procedures

Plant materials and growth conditions

The poplar genotype NE‐19 was used in this study. Woody stem cuttings (25 cm in length) were planted in April in the nursery at Beijing Forestry University, Beijing, China (40°00′N, 116°20′E; 49 m above sea level), in loamy sandy soil (pH = 6.0) for analysis in July. ABA treatment was performed by spraying the leaves with a 250 μmol/L ABA solution. Dehydration was induced by removing plants from the soil and exposing them to air at 50% relative humidity and 25 °C under dim light for 8 h. For both treatments, leaves from different plants were collected at different time points and frozen immediately in liquid nitrogen.

Plantlets of triploid white poplar were cultured in vitro on solid Murashige and Skoog (MS) medium. Leaves and stems were cultured on a medium (pH = 5.8) containing 0.02 mg/L thidiazuron (TDZ), 0.1 mg/L α‐naphthalene acetic acid (NAA), and 0.8% (w/v) agar for shoot induction. Adventitious buds were elongated on MS medium containing 0.5 mg/L 6‐benzylaminopurine (6‐BA) and 0.8% (w/v) agar. The regenerated shoots were individually removed from the callus and transferred to a rooting medium [1/2 MS medium containing 0.05 mg/L NAA and 0.8% (w/v) agar] (Li et al., 2012). Regenerated plantlets were acclimatized in pots and then transferred to a controlled environment growth room [temperature: 25 °C (light)/20 °C (dark); relative humidity: ~45%; photoperiod: 16 h of light/8 h of dark].

cDNA cloning of PdEPF1 from NE‐19

Total RNA was extracted from the collected leaves using the CTAB method described by Ma et al. (2010). First‐strand cDNA synthesis was performed using M‐MLV Reverse Transcriptase and an oligo (dT) primer (Promega, Madison, WI) according to the manufacturer's instructions. The cDNA sequence of PdEPF1 was amplified by PCR; the sequences of the primers used are shown in Table 1.

Table 1.

Primer sequences used for cloning of PdEPF1 cDNA and RT‐PCR

Gene	Forward primers	Reverse primers
For cloning of PdEPF1 cDNA	5'ATGAAGATTTTTGTTGCAACATTAGTC3’	5'TCATGGGACAGGATAAGACTTG3’
For RT‐PCR
PdEPF1	5'GCCATATTTAACAGCAAGGGAGGG3’	5'GTGACCAAACTCTCAGTAGT3’
PdActin	5'GGCCTCCAATCCAGACACTGTA3’	5'AACTGGGATGATATGGAGAA3’
For PCR identify transgenic lines	5'AGTGGATTGATGTGATATCTCCACTGA	5'TCATGGGACAGGATAAGACTTG3’

Reverse transcription and qRT‐PCR analysis

Total RNA from each sample was extracted by the CTAB method, and 1 μg RNA was used for reverse transcription. Subsequently, 20 ng cDNA was used as the template for RT‐PCR amplification. The PCR products were examined on a 2% agarose gel stained with ethidium bromide. The same cDNA samples, as the template, and primers were used for quantitative real‐time PCR in 20 μL reactions containing 200 ng cDNA, 200 nm each primer, and 10 μL SYBR GREEN Master Mix (TianGen Bio Inc., Beijing, China) (Chen et al., 2009). Quantitative real‐time PCR was conducted using the StepOne Plus Real‐Time PCR System (Applied Biosystems, Inc., Carlsbad, CA) according to the manufacturer's instructions. Four technical and three biological replicates were performed in each experiment. The 2^−ΔΔCT method (Schmittgen and Livak, 2008) was used to calculate the relative expression level of PdEPF1, with PdActin used as the internal control. The sequences of the primers used are shown in Table 1. The primers were designed using the software tool Primer Premier 5 (Premier Biosoft International, Palo Alto, CA). Primer specificity was validated by melting profiles showing a single product‐specific melting temperature.

Phylogenetic tree construction using PdEPF1

To understand the relationship between PdEPF1 and other EPFL family members from Arabidopsis, a phylogenetic analysis of PdEPF1 was performed using full‐length amino acid sequences from Arabidopsis and Populus trichocarpa using phylogeny (http://www.phylogeny.fr/). The sequences used were retrieved from PopGenIE (http://popgenie.org/) and TAIR (https://www.arabidopsis.org/).

Domain analysis of PdEPF1

To determine the conserved structures between PdEP1 and other EPFL family members in Arabidopsis, DNAMAN 5 (Lynnon Biosoft Inc., San Ramon, CA) was used to analyse the amino acid sequences.

Vector construction and transformation

To obtain 35S::PdEPF1 transgenic poplars, the coding region of PdEPF1 was cloned into the pCAMBIA‐1301 binary vector under the control of the CaMV 35S promoter. To determine the validity of the positive selection system of mannose in the transformation of poplar, the 6‐phosphomannose isomerase gene from E. coli replaced the hygromycin phosphotransferase of pCAMBIA1301 (Yang et al., 2009). The construct was introduced into A. tumefaciens strain GV3101 and transformed into wild‐type triploid white poplar by the leaf disc method (Li et al., 2012).

Leaves of triploid white poplar excised from plantlets cultured in vitro were cut into discs and precultured on MS medium (pH 5.8) containing 0.02 mg/L TDZ, 0.1 mg/L NAA, and 0.8% (w/v) agar for 3 days. An overnight culture of Agrobacterium (OD₆₀₀ 0.8–1.0) was pelleted by centrifugation and re‐suspended in 1/2 MS liquid medium (OD₆₀₀=0.6) supplemented with 1.8 g/L galactose, 150 mg/L 2‐(N‐morpholino)ethanesulfonic acid (MES), and 50 mg/L acetosyringone (pH 5.0). The precultured leaves were then dipped in the diluted Agrobacterium culture for approximately 30 min and cultured on precultivation medium for 4 days after the liquid on the surface of the leaves had been absorbed by sterilized paper. After co‐cultivation in the dark for 4 days, the leaves were washed with sterile water containing 300 mg/L carbenicillin, blotted with sterile filter paper, and transferred to selective MS medium (pH 5.8) containing 0.02 mg/L TDZ, 0.1 mg/L NAA, 300 mg/L carbenicillin, 8 g/L mannose, and 0.8% (w/v) agar for shoot induction and selection. Adventitious buds were elongated and selected on MS medium containing 0.5 mg/L 6‐BA, 300 mg/L carbenicillin, 8 g/L mannose, and 0.8% (w/v) agar. The regenerated shoots were individually removed from the callus and transferred to a selective rooting medium (1/2 MS medium containing 0.05 mg/L NAA, 300 mg/L carbenicillin, 8 g/L mannose, and 0.8% [w/v] agar).

Molecular verification

Genomic DNA was extracted from nontransgenic and transgenic lines using CTAB method. Transformants were identified through PCR using forward primer for CAMV 35S promoter and reverse primer for PdEPF1 (Table 1).

The PdEPF1 transcript levels in transgenic lines were determined by qRT‐PCR. Total RNA was extracted by the CTAB method from leaves of wild‐type and transgenic plants. cDNA synthesis was performed using M‐MLV Reverse Transcriptase and an oligo (dT) primer (Promega) according to the manufacturer's instructions. The PdActin gene was an internal control. The primers used were the same as those used for the qRT‐PCR assay.

Southern blot analysis was performed to demonstrate transgene integration and gene copy number. Following digestion with EcoRI overnight, DNA samples (30 μg) were separated by electrophoresis on a 1% agarose gel and then transferred to a nylon membrane. DNA probes specific for PdEPF1 coding sequence were labelled with digoxigenin (DIG). Southern blot hybridization was performed according to the protocol of the DIG High Primer DNA Labeling and Detection Starter Kit I (Roche Diagnostics, Mannheim, Germany). After hybridization, the DNA filter was washed sequentially as follows: twice with 2 × SSC and 0.1% SDS for 5 min at room temperature and twice with 0.5 × SSC and 0.1% SDS for 15 min each at 65 °C.

Stomatal density and size determination

Leaves were collected from 27 independent plants (nine each of the wild‐type, oxPdEPF1‐28, and oxPdEPF1‐39 plants) after 2 months of growth in soil. The samples were fixed as described by Cao et al. (2007). The samples were first fixed in 25% glutaraldehyde for 24 h. The leaves were then dehydrated in 30%, 50%, 70%, 85%, and 95% ethanol (15 min each) and then twice in 100% ethanol (15 min each). The samples were then treated with isoamyl acetate: ethanol (1:1) and isoamyl acetate (100%) (15 min each). The number and size of stomata were examined under a scanning electron microscope (Hitachi S‐3400N; Tokyo, Japan).

Short‐term drought experiment

Six wild‐type and twelve oxPdEPF1 plants (six oxPdEPF1‐28 and six oxPdEPF1‐39) grown in soil for 2 months were used as the starting materials in this experiment. Plants of each genotype underwent drought stress induced by withholding water for 12 days. Net CO₂ assimilation, stomatal conductance, and transpiration were measured daily during the experiment.

Long‐term drought experiment

Water deficit stress was imposed on wild‐type and transgenic plants grown in soil for 1 month by keeping the soil RWC at 70% (no stress), 45% (mild stress), or 20% (severe stress) for 40 days. Fifty‐four plants were used as the starting materials in this experiment (six plants per line under various stresses). We weighed the containers daily and supplemented lost water. Three containers without plants were used to estimate the evaporation from soil based on the soil RWC. Plant water consumption was calculated as (supplemented water – soil evaporation): the daily value was recorded. The soil RWC was calculated as (fresh weight – dry weight)/(saturated water weight – dry weight) × 100. The height of each plant was measured every 5 days. After 40 days, the photosynthetic rates, leaf RWC, leaf water potential, chlorophyll content, and plant aboveground biomass were measured.

Physiological analysis

An infrared gas analysis system (LI‐COR 6400; Lincoln, NE) was used to measure net CO₂ assimilation, stomatal conductance, and transpiration in the seventh to ninth leaves of wild‐type and oxPdEPF1 plants. Light and CO₂ curves were obtained in fully expanded leaves of wild‐type and transgenic plants grown in soil for 2 months using the LI‐COR 6400 infrared gas analysis system. Eighteen plants were analysed (six per line). The index included net CO₂ assimilation (A), G_s, transpiration (E), and VPD. Instantaneous WUE was calculated as the ratio of A/E. Light curves were measured at photosynthetically active radiation (PAR) levels of 1500, 1200, 1000, 800, 600, 400, 200, 150, 100, 80, 50, 20, and 0 μmol/m²/s with 450 μmol/mol external CO₂. CO₂ curves were measured at the following external CO₂ values in the order shown: 2000, 1800, 1500, 1250, 1000, 800, 600, 400, 200, 150, 100, 80, 50, and 0 μmol/mol with 600 μmol/m²/s PAR.

The leaf RWC was calculated as (FW – DW)/(TW – DW) × 100. Leaves were removed and immediately weighed to obtain the leaf fresh weight (FW). The leaves were then placed in bottles filled with water for 24 h and then weighed to obtain the leaf turgid weight (TW). The leaves were then dried to a constant weight at 65 °C and reweighed to obtain the leaf dry weight (DW).

The leaf water potential was measured nondestructively in situ at the leaf surface using psychrometers (L‐51A; Wescor, Logan, UT) connected to the PSYPRO Water Potential System (Wescor).

Leaf chlorophyll was extracted with 80% acetone. The absorbance at 663 and 645 nm was obtained using an Ultrospec 3100 Pro UV/Visible Spectrophotometer (Amersham Biosciences, GE Healthcare, Little Chalfont, UK). The chlorophyll a content was calculated using the equation C_a = 12.7 × A663 – 2.69 ×  A645. The chlorophyll b content was calculated using the equation C_b = 22.9 ×  A645 – 4.64 ×  A663. Total chlorophyll was defined as the sum of C_a + C_b.

Water‐use efficiency was calculated as the aboveground plant dry weight/use water. Plant materials above ground were collected and dried to a constant weight at 65 °C to obtain the biomass. The water used was the supplemented water minus the water evaporated from the soil.

Statistical analysis

All data were subjected to analysis of variance according to a completely randomized design model using SPSS software (IBM. Armonk, NY). Differences among means for treatments or plant lines were evaluated by Duncan's test at 0.05 and 0.01 probability levels.