The EAR Motif Controls the Early Flowering and Senescence Phenotype Mediated by Over-Expression of SlERF36 and Is Partly Responsible for Changes in Stomatal Density and Photosynthesis

The EAR motif is a small seven amino acid motif associated with active repression of several target genes. We had previously identified SlERF36 as an EAR motif containing gene from tomato and shown that its over-expression results in early flowering and senescence and a 25–35% reduction of stomatal density, photosynthesis and stomatal conductance in transgenic tobacco. In order to understand the role of the EAR motif in governing the phenotypes, we have expressed the full-length SlERF36 and a truncated form, lacking the EAR motif under the CaMV35S promoter, in transgenic Arabidopsis. Plants over-expressing the full-length SlERF36 show prominent early flowering under long day as well as short day conditions. The early flowering leads to an earlier onset of senescence in these transgenic plants which in turn reduces vegetative growth, affecting rosette, flower and silique sizes. Stomatal number is reduced by 38–39% while photosynthesis and stomatal conductance decrease by about 30–40%. Transgenic plants over-expressing the truncated version of SlERF36 (lacking the C-terminal EAR motif), show phenotypes largely matching the control with normal flowering and senescence indicating that the early flowering and senescence is governed by the EAR motif. On the other hand, photosynthetic rates and stomatal number were also reduced in plants expressing SlERF36ΔEAR although to a lesser degree compared to the full- length version indicating that these are partly controlled by the EAR motif. These studies show that the major phenotypic changes in plant growth caused by over-expression of SlERF36 are actually mediated by the EAR motif.


Introduction
The AP2-ERF (APETALA2-ETHYLENE RESPONSE FAC-TOR) domain family of transcription factors is one of the largest families of transcription factors comprising about 140-280 members in various plants [1]. The family governs plant responses to various biotic and abiotic stresses largely by controlling responses to different plant hormones such as ABA, ethylene and jasmonic acid (JA). A small sub set of this family is characterized by the presence of a seven amino acid repression motif designated as the ERF-associated amphiphilic repression (EAR) motif that is present at the C-terminal end of the protein.
The EAR motif with an L/FDLNL/FxP sequence functions in concert with the AP2 domain but is not restricted to the ERF family. It has been found to be associated with about 21 different types of transcriptional regulator families that also include the AUX/IAA family (with the similar LxLxL motif), C2H2 Zn finger family and the JAZ family [2,3]. The domain can confer the ability to repress transcription even when it is hooked on to proteins that otherwise function as transcriptional activators [4,5]. The EAR motif and the related LxLxL motif actively repress target genes by recruitment of co-repressors such as AtSAP18/SIN3 [6,7] or those of the TOPLESS family such as TPL/TPR [8]. These corepressors in turn interact with histone deacetylases and remodel chromatin to repress transcription [9,3]. The presence of the EAR motif is essential for the repression activity with mutations in the D and L residues of the motif affecting the repressor function [4,10,11].
At least eight EAR motif containing ERF genes are known in Arabidopsis and rice [1] and at least seven are present in tomato [12,13]. They show diverse roles such as in herbivory and wounding [14,15], cold and drought stresses [16], salt stress signalling [17], ABA responses [6] ethylene response [13,18,19], JA responses [20] and senescence [21].While the EAR motif has been shown to function as a repressor motif in vitro, its role and the extent of its contribution in governing the processes mediated by its expression are still not clear. We had previously identified SlERF36/SlERF.F.1 (Accession No. SGN-U564952) as an EAR motif containing AP2 domain gene from tomato, the expression of which accelerated flowering and senescence, and reduced stomatal density by 25-35% in transgenic tobacco plants with direct or indirect effects on photosynthesis, stomatal conductance and transpiration [22]. In this paper, we show that the phenotypic effects of early flowering and senescence imparted by SlERF36 over-expression are largely governed by the EAR motif while the reduction in stomatal density appears to be partly dependent on the EAR motif.

Development of constructs and transgenic Arabidopsis plants
The SlERF36 gene was identified in a previous study as an EAR motif containing AP2 domain gene encoding a protein of 221 amino acids [22]. Constructs containing the SlERF36 gene under the CaMV35S promoter in a pBI121 background (used for transgenic tobacco plants [22]) were also used to transform the Col-0 ecotype of Arabidopsis. In order to study the effects of lack of EAR repressor motif, a reverse primer SlERF36DEAR-R GTCAAGGCAGTGGATTTCTGAGAGATGA was designed just upstream of the EAR motif that incorporated a termination codon. This primer was used in combination with the forward primer ATGGATCCTATGAGAAGAGGCAGAGC (containing the initiation codon) to amplify a fragment of 598 nucleotides that contained an ORF of 585 nt. This ORF encoded a protein of 195 amino acids and lacked the last 26 amino acids that also included the EAR motif. The fragment was cloned in pBI121 under the CaMV35S promoter. Arabidopsis plants were transformed with these constructs using the floral dip method [23].
Of the various transformants generated, two each (lines 1-8 and 2-1 for SlERF36 and lines 6-1 and 9-1 for SlERF36DEAR, lacking the EAR motif) were taken to the third generation to obtain homozygous lines. Progeny of these homozygous lines was used for all phenotypic and physiological studies as well as for gene expression. The plants were grown in culture racks in cups in soilrite mix at a light intensity of 150 mmoles and a photoperiod of 16 h light/8 h dark at 22uC unless otherwise mentioned. For short day treatments plants were grown in a 10 h light/14 h dark cycle. The plants were monitored and compared for various growth parameters that included leaf number, rosette diameter, time of flowering, time of senescence, chlorophyll content, plant height, flower and silique sizes as well as photosynthetic parameters and stomatal density.

Gene expression
Gene expression studies were carried out using RNA isolated from seedlings of control (Col-0) and transgenic plants (with and without the EAR motif). RNA was isolated using the Spectrum Plant total RNA isolation kit (Sigma) and cDNA prepared using the REVERTAID MMLV kit (Fermentas). Expression of SlERF36 in the transgenic lines was carried out using the primers SlERF36-FRT:  TTCCGTCGTTGACCACGGCG  and  SlERF36-RRT: TGCAGTGAAGA TCGTCGGCGC and normalized against actin amplified using the primers Act-F:  ATGACATGGAGAAGATCTGGCATCA and Act-R: AGCCTGGATGGCAACATACAT AGC. Analysis of FT transcript levels was carried out using the primers FT-F: TGTTGGA-GACGTTCTTGATCC and FT-R: AGCCACTCTCCCTCT-GACAA [24] while analysis of SEN4 was carried out using the primers SEN4-F: TCTTCTTCACGACTCTTCTC and SEN4-R: TTGCCCAATCGTCTGCGTTC [25].

Physiological studies
Homozygous Arabidopsis plants expressing SlERF36 and SlERF36 DEAR under the CaMV35S promoter were grown in pots in the culture room under white light. Progeny of homozygous lines was used for each experiment with 5-6 plants per line. Gas exchange parameters were determined on plants enclosed in an Arabidopsis chamber (3010-A) under controlled conditions using the GFS-3000 (Heinz Walz Gmbh, Effeltrich, Germany) instrument attached with a fluorescence module (LEDarray/PAM fluorometer 3055-FL, Walz, Germany). Measurements were made on 20-day-old plants. Measurements of steadystate photosynthesis rate (A), stomatal conductance (g s ) and transpiration rate (E) were carried out at 70% humidity, 25uC leaf temperature, CO 2 concentration of 400 mmol mol 21 and PPFD adjusted to 400 mmol photons m 22 s 21 as described before [22].
Chlorophyll content was measured by isolating chlorophyll from leaves of 12-day-old plants (4 th leaf from bottom) and calculated according to Arnon [26].

Estimation of stomatal density
Leaf epidermis (about 1 cm 2 ) from the abaxial surface of fully expanded leaves (7 th leaf from bottom of 30-day-old plants) was peeled off with a pair of forceps and placed immediately in water and later mounted in 10% glycerol and observed under a light microscope (Nikon Eclipse TE300 Inverted microscope). Stomata

Statistical Analysis
Stastitical analysis for all growth parameters and stomatal numbers was carried out using Student's t-test with P values of , 0.05 considered statistically significant. For physiological parameters, the significance of correlations was tested by using linear regression, with P values of ,0.05 considered statistically significant. Means were compared by using one-way analysis of variance and post hoc means comparison (Scheffé Test). Data analysis and plotting were performed with SigmaPlot version 8 0.

Results
Over-expression of SlERF36/SlERF.F.1 was previously shown to affect flowering time and senescence and was responsible for a 25-35% reduction in stomatal density that affected several photosynthetic parameters in tobacco [22]. In order to perform a more detailed study regarding the role of the EAR motif in these effects, constructs containing the full length SlERF36 and one lacking the EAR motif (SlERF36DEAR) were generated for expression under the CaMV35S promoter (Fig. 1A). Both constructs were used to transform Arabidopsis. Of the various independent lines generated, two designated as SlERF36-1-8 and SlERF36-2-1 (over-expressing the full length SlERF36) and SlERF36DEAR-6-1 and SlERF36DEAR-9-1 (over-expressing SlERF36DEAR) were selected for detailed analysis in the third generation. Progenies of homozygous lines were first checked by semi-quantitative RT-PCR (using actin as internal control) with primers specific to a region common to SlERF36 and SlERF36-DEAR to confirm that the genes were expressed in the respective transgenic lines (Fig. 1B). The plants were then monitored for various visible growth parameters such as time of flowering, leaf shape and size, senescence, height, and physiological characters.

SlERF36 over-expressing plants show early photoperiod independent flowering
One of the most prominent features of transgenic SlERF36 over-expressing plants was the early flowering phenotype ( Fig. 2A). In comparison to control plants which produced inflorescence bolts at 28.3361.52 days, transgenic SlERF36 over-expressing plants produced inflorescence bolts at about 21.6660.57 days (SlERF36-1-8) and 21.061.0 days (SlERF36-2-1) indicating a decrease in flowering time by about 7 days compared to the control (Fig. 2B). In contrast, transgenic SlERF36DEAR overexpressing plants lacking the EAR motif, produced inflorescence bolts at 27.061.0 days (SlERF36DEAR-6-1) and 28.061.0 days (SlERF36DEAR-9-1). Flowering in transgenic SlERF36 plants was initiated with fewer leaves as compared to control and SlERF36DEAR (Fig. 2C). Control plants had an average number of 13.6660.57 leaves at the time of flowering while transgenic SlERF36 over-expressing plants flowered at 9.061.00 leaves/plant (SlERF36-1-8) to 9.3360.57 leaves/plant (SlERF36-2-1). This early flowering phenotype of SlERF36 over-expressing plants was no longer observed after deletion of the C-terminal region containing the repressor domain. Transgenic SlERF36DEAR over-expressing plants flowered at 13.061.00 leaves/plant (SlERF36DEAR-6-1) to 14.3360.57 leaves/plant (SlERF36DEAR-9-1).
We next checked for expression levels of FT (FLOWERING LOCUS T) transcript in 20-day-old plants. As shown in Fig. 2D, a higher transcript level of FT was observed in SlERF36 plants compared to Col-0 and SlERF36DEAR plants indicating that the presence of the C-terminal region containing the EAR motif was essential for the increase in FT levels.
We then tested whether the early flowering phenotype was dependent on photoperiod. For this, plants were grown under 24 h light as well as under short day conditions. Under all conditions, flowering was earlier in SlERF36 over-expressing plants as compared to control and SlERF36DEAR over-expressing plants. Under a 24 h light period, flowering was advanced by about 10 days in transgenic SlERF36 plants while under short day conditions it was advanced by one and half months compared to control and SlERF36DEAR plants (Fig. 2E) indicating that the early flowering conferred by SlERF36 was independent of photoperiod.

SlERF36 over-expressing plants show early senescence
The initiation of flowering was followed by rapid senescence and leaf death in transgenic SlERF36 over-expressing lines as shown in Fig. 3A. The early onset of senescence manifested itself in the form of reduced chlorophyll content, early yellowing and early death in lower leaves of transgenic SlERF36 over-expressing lines as compared to the controls ( Fig. 3A and B). At the 12-day-stage, control plants (fourth leaf from bottom) had a chlorophyll content of 4.5560.13 mg/g FW. In contrast, the leaves of transgenic plants of SlERF36 over-expressing lines had a chlorophyll content of 3.5160.07 mg/g FW (SlERF36-1-8) and 2.6060.05 mg/g FW (SlERF36-2-1) indicating a decrease in chlorophyll content by 25-50% of the control (Fig. 3B). In contrast, transgenic SlERF36-DEAR over-expressing plants showed normal wild type senescence with chlorophyll content ranging from a minimum of 4.1760.23 mg/g FW (SlERF36DEAR-9-1) to 4.4260.02 mg/g FW (SlERF36DEAR-6-1). The early senescence in SlERF36 expressing plants was also evident at the molecular level by higher expression of the SEN4 gene (a marker gene for senescence) [25] (Fig. 3C).
The early flowering phenotype markedly affected plant growth, height and flower and silique length. In general, transgenic SlERF36 over-expressing plants had a smaller rosette diameter as compared to control Col-0 and SlERF36DEAR over-expressing plants (Fig. 4A). Rosettes measured in 28-day-old plants had a diameter of 7.060.5 cm for control plants. In contrast the rosette diameter of transgenic plants of SlERF36 over-expressing lines ranged from a minimum of 4.160.85 cm (SlERF36 -2-1) to 5.061.0 cm (SlERF36 -1-8) indicating a decrease by 2-3 cm in diameter compared to the control. Deletion of the repressor domain led to formation of normal rosettes with size ranges from 6.960.45 cm (SlERF36DEAR-6-1) to 7.3360.28 cm (SlERF36-DEAR-9-1).
Transgenic SlERF36 over-expressing plants also had smaller flowers and siliques as compared to control Col-0 and SlERF36-DEAR over-expressing plants (Fig. 4B). As compared to a silique length of 11.660.89 cm in controls, transgenic SlERF36 overexpressing plants showed siliques of sizes 7.460.54 cm (SlERF36-1-8) and 7.060.0 cm (SlERF36-2-1). In contrast, transgenic SlERF36DEAR over-expressing plants had siliques ranging in length from 9.860.83 cm (SlERF36DEAR-9-1) to 10.660.89 cm (SlERF36DEAR-6-1) (Fig. 4C). Full grown plants over-expressing SlERF36 were shorter as compared to control Col-0 and SlERF36DEAR over-expressing plants (Fig. 4D) with a final To study if reduced growth and early senescence was due to an effect on photosynthesis we carried out a detailed analysis of the various photosynthetic parameters of the transgenic lines using a GFS-3000 system. As shown in Fig. 5, photosynthetic rates showed a marked reduction in transgenic lines. Compared to control plants that showed photosynthetic rates of 8.7561.24 mmol CO 2 fixed m 22 s 21 plants over-expressing SlERF36 showed reduced photosynthetic rates ranging from 4.861.1 (SlERF36-1-8) to 5.960.54 mmol CO 2 fixed m 22 s 21 (SlERF36-2-1). This indicated a reduction of 33-45% compared to the control. Interestingly, the reduction in photosynthetic rates brought about by over-expression of SlERF36 was partially affected by the removal of the EAR motif with transgenic SlERF36DEAR over-expressing lines Transgenic SlERF36 over-expressing lines show reduced stomatal density Our previous results had shown that over-expression of SlERF36 affects stomatal density and that many of the defects in photosynthetic parameters were most likely associated with adaptive responses due to reduced stomatal number [22]. To test whether the reduction in photosynthetic parameters was associated with stomatal number, we measured the stomatal density on the abaxial leaf surface (7 th leaf from bottom of 30-day-old plants) of control and transgenic lines. As shown in Fig. 6 (A and B) control Arabidopsis plants showed a stomatal number of about 8.9460.76/240 mm 2 . In contrast transgenic SlERF36 overexpressing lines showed a stomatal number of 5.5560.61/ 240 mm 2 (SlERF36-1-8) and 5.6160.61/240 mm 2 (SlERF36-2-1) in the same field. This indicated a decrease in stomatal number to just 61-62% of the controls in transgenic SlERF36 overexpressing lines. The removal of the EAR motif in transgenic SlERF36DEAR over-expressing lines led to a comparative increase in stomatal numbers with values of 6.6660.6/240 mm 2 and 6.7760.64/240 mm 2 in the lines SlERF36DEAR-6-1 and SlERF36DEAR-9-1 respectively. Thus over-expression of the truncated gene lacking the EAR motif affected the stomatal number to a lesser degree with a decrease of 25-26% compared to the decrease of 38-39% in lines expressing the full length SlERF36 gene. The difference between the stomatal numbers of the SlERF36 and SlERF36DEAR over-expressing lines was significant at P,0.01. Interestingly, the decrease in stomatal number appeared to be partly due to an increase in cell size. Cells of transgenic lines appeared to be larger in size with fewer cells in the same relative area. The decrease in non-stomatal cell number was about 25% in transgenic SlERF36 over-expressing lines and those over-expressing SlERF36DEAR (Fig. 6C).

Discussion
SlERF36 is a repressor type of ERF containing a C-terminal EAR motif that was shown to affect flowering time and senescence in transgenic tobacco plants [22]. Another prominent phenotype in tobacco was a 25-35% reduction in stomatal density that in turn reduced stomatal conductance, CO 2 uptake and utilization and photosynthesis thereby affecting development. These could either be independent effects or they could be related. In order to elucidate the role of the EAR motif in these different changes, we generated transgenic Arabidopsis over-expressing the complete SlERF36 and a truncated version of the gene lacking the EAR motif. Interestingly, all the effects observed in transgenic tobacco namely early flowering, early senescence, reduced stomatal number and reduced photosynthesis were replicated in transgenic Arabidopsis upon over-expression of the full-length SlERF36. The observations have two major implications: 1. The targets of SlERF36 action for early flowering, early senescence and stomatal number are most likely conserved between tomato/tobacco and Arabidopsis suggesting functional conservation of these pathways. 2. The sites for SlERF36 binding, upstream within the promoter of the target genes, are also most likely conserved between Arabidopsis and tomato/tobacco for it to bind and produce similar effects in these plants.
The EAR motif has been associated with active repression of transcription by several repressor genes that function in different processes [27] and its presence is necessary for repression since mutations in the EAR motif reduce or abolish the repression activity and gene function [4,10]. For instance, plants expressing the Arabidopsis RAP2.1 show enhanced sensitivity to drought while mutation of the EAR motif reduces drought sensitivity [16]. Similarly, plants expressing SlERF3 show severe growth suppression while plants expressing the gene without the EAR motif have no suppressive effect on growth [17]. Likewise, in tobacco, induction of hypersensitive cell death expression by NtERF3 is dependent on the EAR motif [28]. In another study, transgenic rice expressing a mutated version of the EAR motif of the OsERF3 gene, showed increased ethylene biosynthesis and greater drought tolerance compared to the non-mutated OsERF3 expressing lines that showed suppression of ethylene biosynthesis genes [19]. In some cases however, complete removal of the EAR motif does not affect all aspects of the gene function. For example, expression of ZAT7 reduces growth and imparts salt tolerance but deletion of the EAR motif from ZAT7 does not reduce growth suppression although plants lose their salinity tolerance [29].
Of the changes that we observed, the most prominent visible change was the early flowering phenotype. Interestingly, early flowering was seen not only under long day conditions (both 16 h light and 24 h light) but also under short day conditions (10 h light) indicating that the effects on flowering were largely photoperiod independent. Nevertheless, it should be noted that transgenic plants grown under short day conditions did show a delay in flowering compared to those grown under long day conditions indicating that photoperiod did to some extent influence the timing of flowering in transgenic lines. The early flowering was associated with higher levels of the FT (FLOWER-ING LOCUS T) transcript, the gene involved in initiating flowering [30]. Expression of the truncated form of SlERF36 (lacking the EAR motif) abrogated the early flowering phenotype of full length SlERF36 expression under both short day and long day conditions. Its expression did not affect (increase) FT transcript levels (Fig. 2D) in spite of the presence of the AP2 domain. This indicated that the presence of the EAR motif was essential for the higher FT transcript levels and the early flowering phenotype although an effect of other deleted C-terminal residues cannot be ruled out. The fact that SlERF6 over-expression accelerates flowering regardless of photoperiod and in plants as different as Arabidopsis and tobacco suggests that SlERF36 (and the EAR motif) might interact in some way with the general flowering machinery and regulate a component that is common to both photoperiods. Considering that EAR motif containing proteins function as active repressors of transcription, and that SlERF36 expression leads to increased FT transcript levels, one could envisage a possibility where the direct or indirect repression of a floral inhibitor by SlERF36 could activate FT and thereby flowering in the transgenic lines. An interesting possibility that would require further studies is whether SlERF36 affects expression of homologues of TEMPRANILLO (TEM1 and TEM2) that are known to directly repress FT expression [31] or whether it in some way controls TOE1/TOE2 or miRNA172, the regulation of which affects flowering in both short and long day conditions [32]. Incidentally, both TEM and TOE members belong to the AP2/ERF/RAV domain family of transcription factors.
Of the other changes, those related to early senescence appeared to be a consequence of the early flowering phenotype and therefore developmental in nature. This is based on the observations that although senescence was early in transgenic SlERF36 plants (as seen through reduced chlorophyll and higher expression of the SEN4 gene), it was dependent on the photoperiodic flowering and was delayed when flowering was delayed in short day conditions. Under these conditions, rosette Figure 6. Reduction in stomatal density in transgenic SlERF36 and SlERF36DEAR plants. A. Stomatal density on the leaf abaxial surface in control (C) and transgenic Arabidopsis plants from two independent lines (lines 1-8 and 2-1 over-expressing SlERF36 and lines 6-1 and 9-1 over-expressing SlERF36DEAR). Stomatal density from leaf epidermal peels was estimated in the leaf sections in three different regions of three different leaves (7 th leaf from bottom from 30-day-old plants) under a light microscope (Nikon Eclipse TE300 Inverted microscope). The small black bar at the base of each picture on the left hand side represents a length of 10 mm. B. Graphical estimation of the stomatal density of the lower leaf epidermis of control (Col-0) and transgenic SlERF36 and SlERF36DEAR over-expressing lines from Fig. 6A. Values represent the average stomatal density 6 SD in an area of 240 mm 2 of three independent leaves (from the same position). ** P,0.01; ***P,0.001, ****P,0.0001. Figure 6C. Graphical estimation of the nonstomatal cell number of the lower leaf epidermis of control (Col-0) and transgenic SlERF36 and SlERF36DEAR over-expressing lines from Fig. 6A. Values represent the average cell number6 SD in an area of 240 mm 2 of three independent leaves (from the same position). * P,0.05. doi:10.1371/journal.pone.0101995.g006 sizes were larger than under long day conditions and plants took a longer time to senesce (Fig. 2E). The reduced sizes of the various organs (rosettes, flowers and siliques) in transgenic SlERF36 overexpressing plants under long day conditions was most likely an effect due to reduced vegetative growth of these plants and the fewer number of leaves that were present at the time of flowering, leading to fewer photosynthates being synthesized and translocated. In transgenic SlERF36DEAR plants, where the timing of flowering was not affected, senescence was also normal.
In contrast to the flowering and senescence phenotypes, the other major phenotype namely stomatal number was affected in both transgenic SlERF36 plants and those lacking the EAR motif (albeit to a lesser extent). At least in plants over-expressing SlERF36DEAR the apparent reduction in stomatal number by 25% could be attributed to an increase in cell size which increased by about 25%. Stomatal density is known to be tightly controlled by a large number of negative regulators such as ERECTA, ERL1, ERL2, EPF1, EPF2, CHALLAH, YODA, TMM, and SDD1 that are responsible for determining the spacing between stomata [33]. The fact that SlERF36 reduces stomatal density by 25-35% in tobacco and 38-39% in Arabidopsis would indicate that a stomatal development regulator common to both tobacco and Arabidopsis might be controlled by the repressor SlERF36. Stomatal density is also controlled by environmental factors such as CO 2 , humidity, light intensity and water availability with CO 2 levels being by far the most important determinants of stomatal density and conductance in angiospermic plants. An inverse relationship between CO 2 levels and stomatal density and conductance has been noted under experimental conditions as well as in fossil studies [34,35,36] with CO 2 doubling leading to an average decrease of almost 22-29% in stomatal density in Arabidopsis and other plants [37,38]. Changes in stomatal density can affect photosynthesis and growth [39] particularly under conditions where there is no corresponding increase in CO 2 levels. Indeed the reduction in stomatal number and density does affect photosynthesis with growth being affected in SlERF36 overexpressing plants but not in SlERF36DEAR over-expressing plants (where the reduction in stomatal number is lower). Several studies in Arabidopsis have shown that reduced stomatal densities in the range of 20-25% such as in mutants like gtl1, edt1, gpa1 do not affect plant growth [40,41,42] unlike in SlERF36 expressing plants where a much a larger decrease is observed. It is likely that beyond a certain point the reduced density may limit CO 2 availability affecting plant growth. Whether the effects on stomatal density/photosynthesis and flowering are related or independent effects is not yet known. Photosynthetic rates control starch reserves and these in turn could affect C/N ratios and thereby flowering. SlERF36 has recently been shown to actively repress the ethylene responsive GCC box in vitro [13]. The net effect of such a function would be to reduce ethylene responses. In this context, it is interesting to note that recent microarray studies in SlERF36 expressing plants, although not conclusive, showed reduction in transcript levels of AtERF1, AtERF2 (involved in ethylene responses) and AtMKK9 (involved in ethylene biosynthesis), suggesting a reduction in ethylene responses (data not shown). The net effect of a reduction in ethylene responses would be an increase in cell size since ethylene is known to repress cell elongation. Ethylene is also known to delay flowering. Both these effects of increased cell size and early flowering, possibly indicative of reduced ethylene responses, are seen in transgenic SlERF36 over-expressing lines. However, more detailed studies especially through loss of SlERF36 function lines are required to get at a causal relationship between SlERF36, ethylene responses and the early flowering.
In conclusion, we demonstrate that the EAR motif of SIERF36 is most likely responsible for the strong early flowering phenotype and a reduction in stomatal density and photosynthesis that is common to both Arabidopsis and tobacco when SlERF36 is overexpressed. The indication that this motif may also directly or indirectly control the expression of FT, although not studied in detail as yet by us, adds a new dimension to the complex pathways by which flowering is controlled in plants.