ATHB2 is a negative regulator of germination in Arabidopsis thaliana seeds

The germination timing of seeds is of the utmost adaptive importance for plant populations. Light is one of the best characterized factors promoting seed germination in several species. The germination is also finely regulated by changes in hormones levels, mainly those of gibberellin (GA) and abscisic acid (ABA). Here, we performed physiological, pharmacological, and molecular analyses to uncover the role of ATHB2, an HD-ZIP II transcription factor, in germination of Arabidopsis seeds. Our study demonstrated that ATHB2 is a negative regulator and sustains the expression of transcription factors to block germination promoted by light. Besides, we found that ATHB2 increases ABA sensitivity. Moreover, ABA and auxin content in athb2-2 mutant is higher than wild-type in dry seeds, but the differences disappeared during the imbibition in darkness and the first hours of exposition to light, respectively. Some ABA and light transcription factors are up-regulated by ATHB2, such as ABI5, ABI3, XERICO, SOMNUS and PIL5/PIF1. In opposition, PIN7, an auxin transport, is down-regulated. The role of ATHB2 as a repressor of germination induced by light affecting the gemination timing, could have differential effects on the establishment of seedlings altering the competitiveness between crops and weeds in the field.

Seed dormancy is crucial for the adjustment of plant populations to their surrounding environment. Environmental signals that regulate dormancy depth and its alleviation define the germination timing of a seed population and thus are of the outmost adaptive importance 1 . Relevant environmental factors such as temperature, light, and nitrates regulate seed dormancy relief to promote seed germination 2 .
Light has been one of the most characterized factors that regulates seed germination. Phytochromes are well-known photoreceptors that perceive Red (R) and Far-Red (FR) light. They are synthesized in its inactive form, known as Pr (maximum absorption in R), and are photo-converted into its active form, known as Pfr (maximum absorption in FR). Arabidopsis thaliana has five phytochromes, known as phyA to phyE. When PHYTOCHROME INTERACTING FACTOR 1 (PIF1), also known as PHYTOCHROME INTERACTING FAC-TOR 3-LIKE 5 (PIL5), interacts with the Pfr forms of phyA and phyB a massive alteration in the pattern of gene expression in several seed tissues is triggered, influencing the metabolism or sensitivity of several hormones and cell wall proteins, in a direct or indirect manner [3][4][5] . Gibberellins (GA) promote seed germination and abscisic acid (ABA) represses it. The balance between the content and the hormone sensitivity in the seed tissues, finely regulates the growth of the embryo during the germination process. Environmental signals can regulate this balance by modifying the expression of different metabolic enzymes and the expression of positive and negative regulators of GA and ABA, many of which are feedback regulated 6 .
The Arabidopsis thaliana genome encodes around 1500 transcription factors, 40% of which are specific to plants 7 . Homeodomain-Leucine zipper (HD-ZIP) transcription factors are unique to plants and characterized by a HD closely linked to a leucine zipper motif 8 . HD-ZIP proteins constitute a large class of 48 members classified in four distinct families in Arabidopsis. Members of the HD-ZIP II protein family are mostly known for their roles in shade avoidance responses 9 , as well as in carpel margin development 10 and organ polarity 11,12  www.nature.com/scientificreports/ HD-ZIP II protein members control embryonic apical patterning and SAM (shoot apical meristem) function, at least in part, through the interaction with HD-ZIP III proteins 12 . Interestingly, ARABIDOPSIS THALIANA HOMEOBOX2 (ATHB2, also known as HAT4), a member of the HD-ZIP II family, is a direct target of PIF4 and PIF5 13 . In young seedlings and mature plants, the expression of ATHB2 transcripts is repressed in high R:FR ratios but is rapidly and strongly induced by low R:FR ratios typically found in plant canopy environments through the phytochrome system 14,15 . While seedling hypocotyl growth is activated by ATHB2, primary root growth and lateral root formation are both inhibited by ATHB2, and this is rescued by exogenous auxin [16][17][18] . Some previous studies demonstrated that HD-ZIP members can be involved in the regulation of seed germination. For example, ATHB20, as a member of the HD-ZIP I family, was also found to be induced by light in seeds, whereas ABA, NaCl or cold treatments had no effect 19 . Moreover, it has been shown that ATHB20 is strongly expressed in the micropylar endosperm and in the root cap of embryo, acting as positive regulator of seed germination 20 . Another member of this family, ATHB23, functions as a positive regulator of the R light-dependent seed germination and mediates cotyledon expansion 21 . Although ATHB2, the most extensively studied HD-ZIP protein, regulates many biological processes controlled by light in plants, its role in seed germination has not been explored. Here, we demonstrated that ATHB2 delays light-induced germination in Arabidopsis by sustaining the expression of regulatory genes that increases ABA sensitivity and altering hormonal contents in the seeds.

Results
The expression of HD-ZIP II transcription factor genes is altered in light-induced seed germination. We have previously performed RNAseq analyses using imbibed Col-0 seeds in response to a R and FR pulses mediating the promotion and inhibition of germination (85% vs ~ 3%, respectively; 22 ). We identified a total of 5785 genes significantly affected by light: 3099 genes were up-regulated whilst 2686 genes were down-regulated ( Supplementary Fig. 1). Based on this experiment, we reanalyzed the data for the presence of transcription factor genes that might be involved in the regulation of seed germination; we found that the R light significantly affected the expression of 492 transcription factors genes (~ 8.50% of our transcriptome): 251 genes were up-regulated and 241 genes were down-regulated (Supplementary Fig. 1 Table 1).
Among the HD-ZIP II transcription factors, ATHB2 and ATHB4 were down-regulated whereas HAT3 was up-regulated by R light, suggesting that HD-ZIP II transcription factors act as negative (ATHB2 and ATHB4) or positive (HAT3) regulators of seed germination induced by light, (Fig. 1 and Supplementary Table 1). Surprisingly, ATHB2 appeared among the first 20 genes with the highest fold change in our transcriptome (logFC = 4.57). These results are in accordance with previous studies showing that the expression of the ATHB2 gene and of other members of the HD-ZIP II family is regulated by light during germination through the activity of the phyB 5,22,23 and the phyA 24 .
Since several transcriptomic studies showed that the expression of some members of the HD-ZIP II family can be modulated by white light, we further analyzed the expression of HAT1, HAT2, HAT3, ATHB2 and ATHB4 under our experimental conditions by qRT-PCR. RNA was extracted from imbibed Col-0 seeds stratified for www.nature.com/scientificreports/ 3 days at 5 °C and then exposed to continuous white light (WLc) for 12 h (defined as T 1 ). We also harvested a darkness control prior to the irradiation (defined as T 0 ). Consistently, white light slightly induced the expression of HAT2 and HAT3 and strongly repressed the expression of ATHB4 and ATHB2 (Fig. 2), confirming that the expression of different members of HD-ZIP II family can be either induced or repressed during light-mediated seed germination.

ATHB2 is down-regulated in endosperm cells upon germination under continuous white light.
As ATHB2 appeared among the first 20 genes with the highest fold change under our experimental conditions, we focused on the role played by ATHB2 in the regulation of seed germination. To this end, we firstly analyzed the localization and distribution of ATHB2:GUS fusion protein in ATHB2:ATHB2:GUS seeds exposed to light. Imbibed seeds were stratified for 3 days at 5 °C and then exposed to continuous white light (WLc) for 12 h. ATHB2:GUS expression was analyzed at four time points between T = 0 (after the stratification and before light irradiation) and T = 12 h (12 h upon light irradiation; Fig. 3). We analyzed separately endosperm sacs and embryos, since discrete roles in the photocontrol of seed germination have been previously shown 25 . At T = 0 h, ATHB2:GUS fusion protein was detected as sharp nuclear staining in endosperm cells (Fig. 3). At the same time, in the embryo ATHB2:GUS expression is also nuclear and is visible in the root tip, mainly in the lateral root cap (LRC), and weakly in the vasculature. At T = 2 h after continuous white light, the signal in the endosperm nuclei is less sharp; in the embryo, the GUS signal is still high in the LRC and expression starts in vascular initial cells. At T = 4 h, the GUS signal in the endosperm is delocalized within the whole cells and the nuclei are hardly recognizable; in the embryo, the signal in LRC decreased while the vascular staining becomes predominant. From T = 8 h to T = 12 h there is a drastic drop of staining in the endosperm cells, as well as in the number of stained endosperm sacs. On the contrary, embryos showed a well-defined change in the ATHB2 pattern, lower in the LRC and higher in the vasculature, with the number of stained embryos remaining constant within 12 h (Fig. 3).
We also harvested a darkness control at the end the irradiation, that was similar to the darkness control at T = 0 even at a longer time ( Supplementary Fig. 2). Overall, these data suggest that the decrease of ATHB2 expression caused by light during germination is mainly happening in the endosperm.

ATHB2 delays seed germination under continuous white light.
We then evaluated the germination of athb2 mutants under WLc. In particular, we studied the ability to germinate of Col-0, two loss-of-function mutants (athb2-1 and athb2-3) and two of gain-of-function lines (35S::3HA:ATHB2 and athb2-2). To this end, seeds were incubated in darkness at 5 °C for three days prior to WLc irradiation at 22 °C. Seed germination time-course was evaluated upon light treatment for 7 days (Fig. 4). Col-0 seeds started to germinate around 24 h reaching maximum germination at the 7th day (i.e., at 168 h ~ 90%; Fig. 4a). Loss-of-function athb2 seeds fully germinated (athb2-1) or accelerated the process germinating ~ 70% (athb2-3) at 36 h. On the contrary, seeds of the athb2-2 gain-of-function mutant delayed the germination (Fig. 4b) and the over-expressing 35S::3HA:ATHB2 line also reduced the final germination compared to wild-type (Fig. 4a). Dark germination at the end of the experiment was ~ 0% for all genotypes (Fig. 4, right pannels).
ATHB2 controls the concentration of ABA and auxin during seed germination. We quantified GA 1 , GA 4 , ABA and IAA levels in Col-0 and athb2-1 seeds incubated in darkness at 5 °C for 3 days days prior to irradiation with WLc at 22 °C. We measured hormone levels at four time points: (a) mature dry seeds, (b) 72 h stratified imbibed seeds in darkness (prior to light irradiation), (c) 72 h stratified imbibed seeds in darkness and then irradiated for 9 h with WLc and (d) 72 h stratified imbibed seeds in darkness and then irradiated for 18 h with WLc. GA 4 is slightly increased in stratified seeds irradiated for 9 h with white light, both in Col-0 and athb2-1 seeds (Fig. 5) whilst GA 1 levels were not affected neither by the time point nor by the genotype (Supplementary Fig. 3), supporting the idea that GA synthesis is not clearly related to the ATHB2 function during seed germination. ABA levels were higher in athb2-1 compared to Col-0 in dry seeds (Fig. 5), and then similar ABA levels were found between both genotypes in stratified seeds in darkness (Fig. 5). IAA content was higher in athb2-1 compared to Col-0 in dry seeds (Fig. 5) as well as in imbibed seeds stratified for 72 h in darkness (Fig. 5) or irradiated for 9 h with WLc after stratification (Fig. 5). These results suggest that ATHB2 can affect the accumulation of ABA and auxin levels in the dry seeds and during the imbibition in darkness at low temperatures as well as after the first hours of light irradiation in the case of IAA.
ATHB2 positively affects ABA signaling during seed germination. As GA and ABA are the most relevant hormones regulating seed germination, we analyzed dose-dependent germination curves responses of Col-0 and athb2-1 seeds incubated with inhibitors and/or synthetic hormones. We evaluated the sensitivity of the seeds to GA by adding increasing doses of GA 4 in a medium supplemented with 50 µM paclobutrazol, which inhibits endogenous GA biosynthesis. Col-0 and athb2-1 seeds germinated at the same rate in all the doses evaluated, ranging from 0 µM to 1 µM. The germination was promoted between ~ 20% and ~ 95% (Fig. 6a). We also evaluated the sensitivity of the seeds to ABA in a medium containing increasing doses of ABA supplemented with 100 µM fluridone, which inhibits endogenous ABA biosynthesis. Whereas the addition of fluridone in a medium without ABA did not affect the germination of Col-0 and athb2-1 seeds (~ 92% and 100%, respectively; Fig. 6b), increasing the concentration of ABA in the medium reduced the germination of the seeds. In fact, the addition of 0.1 µM ABA inhibits by ~ 40% the germination of Col-0 seeds, while it did not significantly affect the germination of athb2-1 seeds. Interestingly, the germination of Col-0 seeds was significantly lower than athb2-1 seeds (40% vs. 80%, respectively) at ABA concentration equal to 0.3 µM. Seeds of both genotypes did not germi- www.nature.com/scientificreports/ . RNA was extracted from imbibed Col-0 seeds stratified for 3 days at 5 °C and then exposed to WLc for 12 h (defined as T 1 ) or kept in darkness prior to the irradiation (defined as T 0 ). Gene expression levels of (a) HAT1, (b) HAT2, (c) HAT3, (d) ATHB4 and (e) ATHB2 were normalized to T 0 . Each bar represents mean ± ES (n = 3). Significant differences between means are shown with different letters (p < 0.05 by ANOVA followed by Fisher post test). www.nature.com/scientificreports/  www.nature.com/scientificreports/ nate with 1 µM ABA (Fig. 6b). We conclude that ATHB2 affects ABA sensitivity without altering GA signaling in imbibed seeds during germination. Finally, we evaluated the germination of athb2-1 and Col-0 seeds imbibed in a medium with the addition of Picloram (PIC), a synthetic auxin, or 1-NOA, an auxin influx and efflux inhibitor. In both cases, increasing doses of PIC or 1-NOA reduced the germination of both genotypes at similar levels. Doses of 1 mM PIC reduced the germination ~ 20% and 10 mM PIC led to complete inhibition of germination in both genotypes ( Supplementary  Fig. 5). Adding 1 µM 1-NOA did not change the germination of Col-0 nor athb2-1 seeds; while adding 200 µM 1-NOA reduced the germination of Col-0 and athb2-1 in similar proportion. Higher doses of 1-NOA equal to 500 µM led to complete inhibition of germination in both genotypes ( Supplementary Fig. 5).
ATHB2 promotes gene expression of negative regulators of seed germination. In order to have a better understanding of the molecular mechanisms underlying ATHB2-related seed germination, we analyzed www.nature.com/scientificreports/ the expression of genes belonging to different signaling pathways (auxin, GA and ABA). We collected Col-0 and athb2-1 samples for RNA isolation after 12 h in WLc (Fig. 7a). Since some members of the HD-ZIP II family can regulate plant development through the auxin signaling 12,16 , we evaluated the expression of some auxin-related components in the seeds. AUX1 expression, an influx auxin carrier, did not change in athb2-1 when compared to Col-0 seeds, while the expression of PIN7, an efflux auxin carrier, was significantly induced (Fig. 7b). These results suggest that ATHB2 down-regulates PIN7 expression. The expression of SPATULA, a light-stable repressor of the thermo-response in seeds, and GA3ox1, an enzyme involved in the synthesis of GA, did not change in athb2-1 compared to Col-0 seeds (Fig. 7b). Finally, we observed a generalized pattern of repression of ABA and light-related genes in athb2-1 compared to Col-0 seeds (Fig. 7c). ABI3 and ABI5, regulators of ABA signaling, and NCED9, a key enzyme in the biosynthesis of ABA, were significantly less expressed in athb2-1 seeds than in Col-0 seeds (Fig. 7c). PIF1/PIL5, XERICO and SOMNUS expression were also significantly decreased in athb2-1 compared to Col-0 seeds (Fig. 7c). We conclude that an accelerated germination of athb2-1 seeds correlates well with a decrease in the gene expression of transcription factor related with ABA and light signaling that act as repressors of germination in Arabidopsis.

Discussion
Seed germination is a process of extreme relevance for a successful seedling establishment in the field. The information concerning transcriptional and gene expression regulation of light-induced seed germination is abundant 4,5,22,[26][27][28][29] . By comparative transcriptomic analysis, we identified ATHB2 as a negative regulator of lightinduced seed germination. We showed that a pulse of R light can affect the expression of ~ 20% of Arabidopsis transcriptome in Col-0 imbibed seeds (5785 genes, Supplementary Fig. 1 Fig. 1, 22 ). Interestingly, ATHB2 appeared among the first 20 down-regulated genes with the highest fold change 22 . Previous studies suggested that both phyA and phyB www.nature.com/scientificreports/ regulate the expression of ATHB2 in imbibed Arabidopsis seeds ( Fig. 1; 5,22,24 ). Interestingly, different members of the HD-ZIP II class show opposite expression patterns. Some HD-ZIP II transcription factors are promoted (HAT1, HAT2 and HAT3), whilst others are repressed (ATHB2 and ATHB4) in Arabidopsis seeds (Fig. 1). In particular, white light represses ATHB2 expression within 12 h after induction of seed germination (Fig. 2). In these conditions, we also showed that the ATHB2 expression decreases specifically in the endosperm and in the embryonic root cap cells, whereas it increases in the vasculature of the embryo (Fig. 3). Since the ABA produced in the endosperm is critical for seed dormancy establishment and germination, we can also not discard differentially expression of ATHB2-regulated gene expression in embryo and endosperm tissues. The endosperm has been previously shown to be responsible for the inhibition of embryo germination in Arabidopsis imbibed dormant seeds, actively producing and releasing ABA 30 . Furthermore, phytochromes control of seed germination also acts in separate times and compartments, with FR light initially repressing germination of the embryo through phyB contained in the endosperm, and later sustaining a phyA-dependent germination from the embryo. This involves ABA release from the endosperm and distinct spatial activities of phytochrome signaling components 25 .
The observed higher expression of ATHB2 in the endosperm of stratified seeds, just before white light-induced seed germination, and its rapid downregulation within the first hours of light exposure, is therefore suggestive that the negative role exerted by the endosperm on seed germination requires ATHB2 (Fig. 3). However, ATHB2 is also expressed in the vasculature during the germination, where it appears to be positively regulated by light (Fig. 3), suggesting that ATHB2 could have a different role in this tissue, likely more linked to auxin signaling preparing for post-germination events associated with seedling de-etiolation (see below). In the future, it will be interesting to study in detail the specific roles of ATHB2 in the endosperm and in the embryo, respectively. ABA accumulation in dry seeds has been observed in both dormant and non-dormant Arabidopsis seeds, and it rapidly decreases upon imbibition in both cases 31,32 . The difference between dormant and non-dormant seeds has been shown to lie primarily in their ability to synthesize de novo ABA after imbibition 31 . However, the higher concentration of ABA observed in athb2-1 mutant dry seeds has no effect on germination after imbibition (Fig. 5). Conversely, ABA levels in athb2-1 mutant decrease to amounts comparable to those in the wild type after imbibition (Fig. 5). Therefore, the reduced levels of ABA signaling regulators observed after 12 h of imbibition under WLc suggest a role of ATHB2 in this process downstream or independent of ABA. In this plausible scenario, ATHB2 might be responsible -directly or indirectly-for the higher levels of ABA observed in athb2-1 dry seeds (Supplementary Fig. 6). Accordingly, Stamm et al. (2012) have shown that ATHB2 appears to be highly expressed in seed tissues 33 . Moreover, analyses of different transcriptomes during seed development in Arabidopsis also show that ATHB2 is expressed in embryos 34 , dry seeds 35 and in seeds across different timecourses during germination 36 . It has been reported that HAT1 and its homolog HAT3 redundantly and negatively regulate both ABA biosynthesis and signaling genes in a complex transcriptional regulation 37 . A similar regulatory circuit might be operating during seed maturation through ATHB2. Indeed, we have shown that ATHB2 exerts its function in several processes in combination with at least HAT3 and ATHB4 12,38 . Further work needs to be done to understand the role of ATHB2 during seed development and/or after-ripening.
Recently, it has been shown that pif1/pif5 and the quadruple pif1/pif3/pif4/pif5 are insensitive to ABA whereas phyB mutant displayed hypersensitivity to ABA in seed germination, respectively, reinforcing the hypothesis that the PIFs inhibit germination by modulating ABA signaling 39 . Interestingly, here we show that the lack of ATHB2 results in a lower expression of PIF1/PIL5 and SOM genes respect to the control (Fig. 7). As PIF1/PIL5 positively regulates ATHB2 expression 5,23,24 , our results suggest that ATHB2 might be involved in a positive feedback loop with PIF1/PIL5, which in turn control the expression of SOM 40 , thus enhancing the repressor activities in seed germination mediated by ABA. Accordingly, the lack of ATHB2 result in a lower expression of genes encoding relevant components of the ABA signaling such as those of ABI3, ABI5, NCED9, XERICO and SOM (Fig. 7). XERICO promotes ABI3 expression 41 , a transcription factor that is involved in the ABA signaling; and SOM promotes the expression of NCED9 42 , a gene involved in the ABA synthesis. In agreement with that, the ABA sensitivity (Fig. 6) and the ABA signaling components (Fig. 7) of the seeds are also reduced in athb2-1 mutant. Therefore, our results suggest that ATHB2 acts as a negative regulator of the light-dependent germination, exerting its function by affecting ABA signaling. Therefore, we suggest that ATHB2 acts as a negative regulator of the light-dependent germination, exerting its function by affecting ABA signaling. Interestingly, similar roles for other HD-ZIP members during seed germination have also been established. Barrero et al. (2010) have shown that a member of the HD-ZIP I family, ATHB20, whose expression is also regulated by light, promotes germination in Arabidopsis seeds. ATHB20 is expressed in the micropylar endosperm and in the root cap, and reduces ABA sensitivity 20 . ATHB23, another member of the same family as ATHB20, has also been associated with the promotion of seed germination 21 . Based on our results and those on ATHB20, we can speculate that ATHB2 and ATHB20 play different and antithetical roles at least in the root cap where they are oppositely regulated by light.
Links between HD-ZIP II proteins and auxin have been established 43,44 . However, the molecular mechanisms involved between HD-ZIP II transcription factors and the auxin signaling is still poorly understood. Our analysis found that ATHB2 controls the IAA production in dry seeds and during a short period of seed imbibition (up to 9 h upon light irradiation, Fig. 5). We also demonstrated that ATHB2 negatively affects at least the expression of the auxin transporter PIN7 in light-induced germination (Fig. 7). The transcriptome induced by phytochromes in seeds showed that auxin transport proteins are a pre-requisite for germination 5,[22][23][24] . PIN7 has been involved in the pattern specification during root development, being expressed at lateral and basal membranes of provascular cells in the meristem and elongation zone 45 . Then, the higher expression of PIN7 in the athb2-1 mutant might produce positive effects on proliferation and cellular elongation of the embryonic axis during germination, thus accelerating the emergence of the embryonic radicle.
In conclusion, our work suggests that ATHB2 is a negative regulator of light-mediated germination that control a myriad of ABA and light-related transcription factors by reducing ABA sensitivity. The knowledge generated in this work using Arabidopsis seeds could be potentially expanded to other species in the future, www.nature.com/scientificreports/ especially crop plants, to decipher the molecular networks for the relief of dormancy by light. Furthermore, the identification of ATHB2 as a repressor that delays the timing of seed germination could have effects on the establishment of seedlings changing the competitiveness between crops and weeds in the field.

Materials and methods
Plant material and growth conditions. Arabidopsis thaliana plants were grown as previously described in 22 . Plants were grown under long day conditions (16 h light/8 h dark, PAR = 100 µmol m −2 s −1 ) with an average temperature of 21 ± 2 °C. Wild-type and mutant plants were grown together and their mature seeds were harvested at the same time to avoid differences in post-maturation that can affect seed germination. Seeds of each genotype were harvested as a single bulk consisted of at least 5 plants. Seeds were stored in open tubes inside a closed box and maintained in darkness with silica gel at 4 °C until the experiments were performed. Arabidopsis thaliana wild-type Columbia-0 (Col-0) seeds were obtained from the Arabidopsis Biological Resource Centre. In this study, we used insertional knock-out lines: athb2-1 (Salk_106790; 46 ), and athb2-3 12 ; an over-expressing line: athb2-2 (Salk_006502, 12 ) and two over-expressing tagged lines: ATHB2::ATHB2:GUS 12  Gene expression analysis by quantitative RT-PCR. We followed our protocol described in 22 . Seed samples (~ 10 mg seed dry weight) were sown in clear plastic boxes, each containing 10 mL of 0.8% (w/v) agar in de-mineralized water and incubated for three days in darkness at 5 °C. Seeds were then exposed to white light at 22 °C and samples were collected 12 h upon light irradiation. After sampling, seeds were immediately frozen in liquid nitrogen and stored at − 80 °C. RNA was extracted using the Spectrum Plant Total RNA Kit (Sigma-Aldrich, Steinheim, Germany) according to manufactures protocol. cDNA was synthesized using MMLV High Performance Reverse Transcriptase (Epicentre, Madison, USA) and oligo-dT primers. The synthesized cDNAs were amplified with FastStart Universal SYBR Green Master (Roche, Madison, USA) using the 7500 Real Time PCR System cycler (Applied Biosystems, Foster City, CA, USA). PP2A gene was used as a normalization control. The primers used are described in Supplementary Table 2.
Quantification of hormone levels. The thoroughly ground tissue (~ 30-50 mg seed dry weight) was suspended in 80% methanol − 1% acetic acid containing internal standards and mixed by shaking during one hour at 4 °C. The extract was kept a − 20 °C overnight and then centrifuged and the supernatant dried in a vacuum evaporator. The dry residue was dissolved in 1% acetic acid and passed through a reverse phase column (HLB Oasis 30 mg, Waters), as described in 47 . The final residues were dried and dissolved in 5% acetonitrile − 1% acetic acid and the hormones were separated by UHPLC with a reverse Accucore C18 column (2.6 µm, 100 mm length; Thermo Fisher Scientific) with a 2 to 55% acetonitrile gradient containing 0.05% acetic acid, at 400 µL/min over 21 min. The hormones were analyzed with a Q-Exactive mass spectrometer (Orbitrap detector; Thermo Fisher Scientific) by targeted Selected Ion Monitoring (tSIM; capillary temperature 300 °C, S-lens RF level 70, resolution 70.000) and electrospray ionization (spray voltage 3.0 kV, heater temperature 150 °C, sheath gas flow rate 40 µL/min, auxiliary gas flow rate 10 µL/min) in negative mode. The concentrations of hormones in the extracts were determined using embedded calibration curves and the Xcalibur 4.0 and TraceFinder 4.1 SP1 programs. The internal standards for quantification of each of the different plant hormones were the deuteriumlabelled hormones (purchased from OlChemim Ltd, Olomouc, Czech Republic). The hormone quantification analysis was carried out at the Plant Hormone Quantification Service, IBMCP, Valencia, Spain. www.nature.com/scientificreports/ GUS analysis. GUS analyses were performed as described 48 . In brief, seeds were fixed in cold 90% acetone for 1 h. During fixation, the testa was peeled off the seed and the endosperm was separated from embryo under a stereomicroscope. Endosperms and embryos were washed twice in 100 mM phosphate buffer. The material from the same time point was then stained in the same staining solution at 37 °C for 24 h. Staining was stopped in 80% EtOH, and then samples were mounted in chloral hydrate for DIC analysis under an Axioscop 2 Plus (Zeiss, Germany) binocular microscope. Experiments were performed four times with two independent lines.
Experimental design and statistical analysis. Physiological and hormonal experiments were done using at least three different populations of seeds harvested for at least 6 plants. To test for significant differences in the response of the seeds, we conducted two-way ANOVAs for each wild-type and mutant group, using the angular transformation of the percentage of germination and the InfoStat Software version 2017 (Grupo InfoStat, FCA, Universidad Nacional de Córdoba, Argentina). Fisher post-test was used to test differences when significant interactions were observed.
All the experiments and collection are done in accordance with the relevant institutional, national, and international guidelines and legislation. We have also obtained the corresponding permission to collect Arabidopsis thaliana seeds and plant material.