Introduction

Plant growth and development are often severely affected by abiotic stresses. Abiotic stresses are known to induce osmotic imbalance in plants and disturb the plant water homeostasis. Aquaporins, the membrane intrinsic proteins facilitating and regulating the passive movement of water molecules down a water potential gradient, play a central role in maintaining a turgor and water transport in plants (Maurel et al. 2015). Above all, the plasma membrane intrinsic protein (PIP) aquaporins have been proven to participate in responses to abiotic stresses, such as drought, salinity, or chilling (Lian et al. 2004; Liu et al. 2007). Moreover, PIP gene expression can be regulated by these abiotic stresses (Aroca et al. 2012) as well. In addition, there is an opinion that PIP1-type aquaporins can function as sensors of blue light (BL) because of their localization in the plasma membrane and ability to bind flavins (the cofactors of BL photoreceptors) (Kaldenhoff and Eckert 1999; Lorenz et al. 2003). Furthermore, the expression of PIP1;2 aquaporin was showed to be enhanced by BL in Arabidopsis thaliana cell tissue cultures (Kaldenhoff et al. 1995, 1996).

In plant seeds, aquaporins play a central role in the physiology of water economy (Maurel et al. 2015; Obroucheva 2013; Obroucheva et al. 2017). Embryo growth during seed germination is driven by water uptake which is necessary for embryo cell elongation and expansion. In general, seed germination follows three phases. In phase I, imbibition of water, reactivation of metabolism and at the end of this phase, cell elongation, and radicle emergence occurs. In phases II and III, the embryonic axis grows further as cells elongate and divide to establish the seedling (Toole et al. 1956). Aquaporins are not involved in the early imbibition of dry seed by water, but they play a role in embryo growth (Obroucheva 2012; Willigen et al. 2006). It was reported that the expression of aquaporin genes is low during seed imbibition and starts when a seed hydration level of 50–55% (fresh weight) is achieved (Obroucheva 2012). The expression of aquaporin genes is strongly activated at radicle emergence and just after it. Despite dry seeds contain the transcripts of aquaporin genes and aquaporin themselves, it seems that the aquaporins function only after radicle protrusion to provide enhanced water inflow to growing cells (Obroucheva 2013). In addition, Shiota et al. (2006) showed that plasma membrane aquaporins are expressed during tomato seed development. Apparently, they participate in the delivery of water and possibly some other compounds into the developing seeds (Obroucheva 2013).

Seed germination is one of the physiological processes which are influenced by light. Although very little is known about perception of blue light (BL) by seeds (Goggin and Steadman 2012), it was shown that BL mostly reduces seed germination in tomato and that tomato seeds under osmotic stress germinate better in the dark than in BL (Fellner and Sawhney 2002; Piterková et al. 2012). Besides, BL can modulate transcription of many genes including plant aquaporins (Baaziz et al. 2012; Kaldenhoff et al. 1995, 1996) and water stress acts on aquaporin function at the level of transcription and gating through direct effects on osmotic gradient (Luu and Maurel 2005).

Tomato (Solanum lycopersicum L.) is one of the most important crop species. The 7B-1 mutant displays male sterility under long days, but under short days, it produces fertile flowers (Sawhney 1997; Sheoran et al. 2009). Compared with the corresponding WT (cv. Rutgers), the 7B-1 shows reduced de-etiolation of hypocotyl growth associated with hypersensitivity to, and high endogenous levels of, ABA (Fellner et al. 2001). The 7B-1 mutation impairs various BL responses, such as BL-induced stomata opening (Hlavinka et al. 2013; Ježilová et al. 2012), hypocotyl phototropism (Bergougnoux and Fellner, unpublished results) and chloroplast movement (Špundová and Savara, unpublished results). It suggests that 7B-1 mutation may somehow affect phototropin signalling, while the possibility that 7B-1 is a CRY mutant is not likely (Sheoran et al. 2006). In long-day conditions, 7B-1 seed germination was found to be less responsive than corresponding WT to various salts (including NaCl, Na2SO4, KCl, and K2SO4), osmotic stress and to low-temperature stress (Fellner and Sawhney 2001). Interestingly, Fellner and Sawhney (2002) also reported that seed germination in 7B-1 is more tolerant to exogenous ABA and osmotic stress specifically under BL. Moreover, we reported that BL amplifies the inhibitory effect of osmotic stress on tomato seed germination (Fellner and Sawhney 2002; Piterková et al. 2012). Recently, we identified number of potentially novel miRNAs, which are associated with enhanced tolerance of 7B-1 to abiotic stress under BL (Omidvar et al. 2015). Our data also showed that response to different lights and stresses in 7B-1 and WT involves remodelling DNA methylation, highlighting the differences in epigenetic and transcriptional regulation of light and stress responses between 7B-1 and WT (Omidvar and Fellner 2015). Specifically in BL conditions, 7B-1 is less sensitive to biotic stress as well (Bergougnoux et al. 2009). Recently, a genetic characterization of the 7B-1 mutant was published and SlGLO2 gene was proposed as a candidate gene underlying the 7B-1 mutation (Pucci et al. 2017). SlGLO2 belongs to class B MADS-box genes which are involved in stamen development (Pucci et al. 2017). However, a role of SlGLO2 in BL- and/or stress signalling is not known and it is under investigation.

Number of reports describing a direct contribution of BL in the ability of plants to tolerate abiotic stress is very limited. Therefore, 7B-1 mutant seems to be a suitable model to study a possible interaction between BL and osmotic stress. Various environmental factors regulate aquaporins mostly at the transcriptional level (Liu et al. 2013). For this reason, we hypothesized that the tolerance of 7B-1 mutant seeds to BL and osmotic stress may involve differential expression of aquaporin genes. Shiota et al. (2006) showed that genes of PIP aquaporin family (namely PIP1;2, PIP1;3, PIP1;4, PIP1;5, PIP2;1, PIP2;2, and PIP2;3) participate in tomato seeds development, while expression of some PIPs can be regulated by abiotic stresses (Aroca et al. 2012). Thus, we focused on this aquaporin family and we studied the expression of these seven genes in 7B-1 seeds as a function of mannitol-induced osmotic stress and BL conditions. We revealed that the gene expression patterns of PIP1;2, PIP1;3, PIP2;1, and PIP2;3 were altered in response to BL, osmotic stress, and/or by 7B-1 mutation. Interestingly, both 7B-1 mutation and osmotic stress decreased expression of some of the studied PIPs, most probably to preserve enough water for radicle elongation under stress conditions. As 7B-1 mutation causes defects in BL responses, and expression of PIP1;2 and PIP2;1 is regulated by BL, we hypothesized that these genes may play roles in BL sensing or signalling.

Materials and methods

Plant material and stress treatment

Solanum lycopersicum L. (tomato) seeds of cultivar Rutgers (WT) and recessive single gene 7B-1 mutant (Sawhney 1997) were used. 7B-1 seeds are less responsive than WT to BL-induced inhibition of seed germination and to various abiotic stresses specifically under BL (Fellner and Sawhney 2001, 2002; Piterková et al. 2012).

All experiments were performed in in vitro conditions. Sterilized seeds (2.8% sodium hypochlorite solution, Bochemie, Czech Republic) were incubated in squared Petri dishes filled with basal MS medium (Murashige and Skoog 1962) or with MS medium supplemented with 20 or 70 mM mannitol. The edge of each plate was twice sealed with an air permeable tape (Batist, Czech Republic). The dishes with seeds were placed in vertical position into controlled growth chambers (Microclima 1000E, Snijders Scientific B. V., The Netherlands) in continuous BL conditions or they were wrapped in tinfoil (to simulate the darkness) and placed into the same chamber. BL was provided by blue tubes TLD-36W/18-Blue (Philips, USA) with a maximum irradiance at 460 nm. Total photon fluence rate of light was 10 µmol m−2 s−1. The light spectrum was measured using a portable spectroradiometer (model LI-1800; Li-COR, NE, USA).

Number of germinated seeds was counted the seventh day after sowing and the average percentage of seed germination was calculated. At least 11 experiments were done for each treatment. To show a trend in course of seed germination, the number of germinated seeds was scored from second to seventh day as well. A control on MS was sown for each experiment. For each treatment, there were at least four independent experiments per each time point.

Gene expression analyses

The WT and 7B-1 mutant seeds germinated on basal MS medium or MS medium supplemented with 70 mM mannitol were collected (96 h after sowing, radicle 1–10 mm long) into liquid nitrogen. Seeds from BL were harvested under BL and seeds from darkness under green safelight. Harvested seeds were ground under liquid nitrogen and total RNA was isolated with the Isolate II RNA Plant Kit (Bioline) using RLS buffer according to a manufacturer´s instruction. At the last step, RNA was eluted with 40 µl of RNase-free water. As soon as isolation was done, residual DNA was removed from samples by recombinant DNaseI treatment (Takara Bio Inc., Japan) and recombinant RNase inhibitor (Takara Bio Inc., Japan) for 60 min at 37 °C. DNaseI was inactivated by heat treatment according to manufacturer´s instruction. Following this, first-strand cDNA mixtures were prepared from 0.7 µg of total RNA using PrimeScript™ 1st strand cDNA Synthesis Kit (Takara Bio). The gene expression was analysed using SensiFAST™ SYBR Lo-ROX Kit (Bioline) on the CFX96 Touch™ (BioRad, USA). The PCR reaction mixture was prepared according to manufacturer’s protocol—the reaction mixture contained 10 µl of SensiFast SYBRR Lo-ROX Mix, 0.8 µl of each primer (10 µM), 4.4 µl of sterile RNase-free water, and 4 µl of cDNA template (diluted 1/50). The final reaction volume was 20 µl. Each sample was measured in triplicate. The following PCR running conditions were used—an initial denaturation at 95 °C for 2 min, followed by 40 cycles of 95 °C for 5 s and 60 °C for 15 s. Melting curve was analysed at the end of PCR reaction. The experiments were repeated five times.

PP2Acs (protein phosphatase 2A catalytic subunit) (Lovdal and Lillo 2009) and TIP41-like (Dekkers et al. 2012) genes were used as reference genes. The sequences of PIP primers used are given in Table 1. The primers were designed using PrimerQuest® program (IDT, Coralville, USA). The primer efficiencies were calculated from the slope of the dilution curve and they ranged from 0.9 to 1.0.

Table 1 Sequences of PIP primers used in RT-qPCR analyses
Full size table

Statistical analysis

The significant differences between control (seeds on basal MS medium) and mannitol treatments were tested using Kruskal–Wallis test (a distribution-independent test) with multiple comparisons at the 0.05 significance level, since the data were not normally distributed. The analysis was done separately for dark- and BL-grown samples. For each treatment, the significance of differences between the dark and BL was assessed using Mann–Whitney U test (a non-parametric test) at the 0.05 significance level. Statistical analyses were performed using the STATISTICA 12 software (StatSoft, OK, USA). The results of germination assay are presented as box plots, where the lower and upper ends of the box are the first and third quartiles, respectively. The horizontal line inside the box is the median value (the square inside the box represents a mean) and the whiskers indicate the adjacent values (1.5 times interquartile range).

The relative gene expression of target gene was determined against the expression level of both reference genes with taking into account the different PCR efficiencies of genes (Pfaffl 2001). Then, a geomean of the calculated relative expressions was used as a dependent variable to design a matrix for multiple linear regression analysis. The independent variables (predictors) designed as a mutation, light, and stress could take on one of two values, 1 or 0, to indicate the presence or absence, respectively, of the effect of the predictor (the binary system). To find the significance of the effect of predictors and to quantify it, the multiple linear regression analysis was done using the QC.Expert 2.9 software (TriloByte Statistical Software, Czech Republic).

Results

The main goal of this work was to investigate how osmotic stress (induced by 70 mM mannitol) and BL influence the expression profile of PIP genes in 7B-1 mutant seeds and its corresponding WT and how their gene expression patterns associate with reduced 7B-1 sensitivity to osmotic stress and BL. We focused on the effect of two different mannitol treatments and BL on seed germination in both genotypes. Then, we studied changes in the expression of PIP genes in seeds induced by 70 mM mannitol, BL and 7B-1 mutation.

Seed germination assays

In the dark as well as in BL, WT seed germination was significantly affected by treatment with 70 mM mannitol, but not by 20 mM mannitol (Fig. 1a). Similarly, seed germination in 7B-1 mutant was reduced significantly by 70 mM mannitol in the dark as well as in BL (Fig. 1b). However, in the dark, 70 mM mannitol tended to reduce 7B-1 mutant seed germination less than WT seed germination, but this difference did not reach statistical significance (at the 0.05 significance level). In both the dark and BL, 20 mM mannitol had not significant effect on 7B-1 mutant seed germination (Fig. 1b).

Fig. 1
figure 1

Effect of mannitol on germination of tomato cv. Rutgers (WT) (A) and 7B-1 mutant (B) seeds incubated for 7 days in the dark (D) or under continuous blue light (BL). Seeds germinated on the basal MS medium (MS) serve as a control. The significant difference of mannitol treatments compared to corresponding control on medium without mannitol was tested using Kruskal–Wallis test at 0.05 significance level, a: in the dark, significantly different from D-control, b: in BL, significantly different from BL control. For each treatment, at least 11 experiments were done

Full size image

For each treatment, the effect of BL on WT and 7B-1 mutant seed germination compared to D was assessed using Mann–Whitney test (at the 0.05 significance level). The effect of BL was always significant (not marked in Fig. 1).

The trends of kinetics of seed germination shown in Fig. 2 indicate that WT and 7B-1 mutant differed just very slightly in the kinetics of seed germination on the basal MS medium in the dark. On the other hand, BL reduced markedly kinetics of WT seed germination (Fig. 2a), whereas in 7B-1 mutant, its effect was not so pronounced (Fig. 2b).

Fig. 2
figure 2

Kinetics of WT and 7B-1 mutant seed germination on MS medium (control) and on MS medium supplemented with 70 mM mannitol during 7-day cultivation in the dark (A) or blue light (B). The data show medians and the error bars represent the first and third quartiles. For each treatment, there were at least four independent experiments per each timepoint. Thirty seeds per each dish were sown for each treatment

Full size image

Changes in aquaporin gene expression under blue light and osmotic stress

In this study, we focused on investigation into the effect of BL, osmotic stress induced by 70 mM mannitol and 7B-1 mutation on RNA levels of plasma membrane aquaporins (namely, PIP1;2, PIP1;3, PIP1;4, PIP1;5, PIP2;1, PIP2;2, and PIP2;3) in germinated seeds (radicle up to 10 mm) cultivated for 96 h.

The multiple linear regression analysis showed that none of studied predictors (blue light, osmotic stress, and 7B-1 mutation) reached statistical significance (at 0.05 level) for PIP1;4, PIP1;5, and PIP2;2 genes (Table 2). In other words, expression of none of these genes was affected significantly by BL, osmotic stress, or by 7B-1 mutation. With the predictors mutation and stress held fixed, BL affected significantly the expression of PIP1;2 (upregulation), PIP1;3 (downregulation), and PIP2;1 (downregulation) genes (Table 3). As Table 3 further shows, 7B-1 mutation decreased significantly the expressions of PIP1;2 and PIP2;1 genes when all the other predictor variables (BL and osmotic stress) were held fixed. Finally, held all the other predictors constant, the mannitol-induced osmotic stress reduced significantly the expression levels of PIP1;3 and PIP2;3 genes (Table 3).

Table 2 Results of a multiple linear regression analysis for genes PIP1;4, PIP1;5, and PIP2;2, whose gene expression was not affected significantly by blue light (BL), 7B-1 mutation, or mannitol-induced osmotic stress
Full size table
Table 3 Results of a multiple linear regression analysis showing the effect of blue light (BL), 7B-1 mutation and mannitol-induced osmotic stress on the expression of PIP genes
Full size table

In other words, the results of multiple linear regression analysis can also be presented as the equation of multiple linear regression involving the predictors with significant impact on the expression of each gene (Table 4).

Table 4 Equations of a multiple linear regression for PIP1;2, PIP1;3, PIP2;1, and PIP2;3 genes
Full size table

Discussion

Expression of the aquaporins is divergently regulated by various environmental factors as a part of an adaptation mechanism by which plants reduce water losses through efflux to cope with stress conditions. Since the specific function of plant aquaporins under different stress and light conditions is still unclear, the comprehensive studies of aquaporin gene expression in divergent plant species are crucial to improve our understanding of their function in responses to various challenges. We wanted to assess statistically the potential impact of 70 mM mannitol-induced osmotic stress, BL and 7B-1 mutation on PIPs gene expression. The expression patterns could provide useful indications about the functions of the studied aquaporins. Tomato 7B-1 mutant seed germination, showing BL-specific reduced responsiveness to osmotic stress (Fellner and Sawhney 2002), is a valuable model system to study the possible involvement of aquaporins in light-regulated stress signalling.

Various responses of aquaporins to osmotic stress suggest that each aquaporin isoform could contribute differently to water transport and regulation of water homeostasis in germinated seeds. Several studies demonstrated that plants downregulate the aquaporin expressions to avoid excessive loss of water in challenging environment (Alexandersson et al. 2005; Jang et al. 2004; Lian et al. 2006). Here, we found that the expressions of PIP1;3 and PIP2;3 genes were downregulated by osmotic stress induced by 70 mM mannitol and it was associated with reduction of seed germination. All these literature data suggest that some aquaporins may help to adapt or tolerate the stress by reducing their expression. On the other hand, other aquaporins might help in maintaining the normal physiological processes in seeds, and thus, their expressions remained stable under the stress conditions. In our study, this was the case of PIP1;2, PIP1;4, PIP1;5, PIP2;1, and PIP2;2 aquaporin genes. On the contrary, some reports show that the expression of PIPs can also be upregulated by abiotic stresses, suggesting that these aquaporins can facilitate water transport under stress conditions (Jang et al. 2004; Liang et al. 2013; Liu et al. 2013).

Although the difference in percentage of WT and 7B-1 mutant seed germination under 70 mM mannitol was not so distinct as reported for higher mannitol concentrations (Fellner and Sawhney 2002), 7B-1 mutation had a statistically significant negative effect on expression of aquaporin genes PIP1;2 and PIP2;1. This indicates the tendency of 7B-1 mutant seeds to preserve water for radicle growth during seed germination. Thus, the mutation acted on PIP gene expressions in the similar way as osmotic stress (conserve the water under the stress) making 7B-1 mutant pre-adapted to osmotic stress. This is further supported by the elevated level of ABA in 7B-1 mutant (Fellner and Sawhney 2001, 2002).

Blue light tended to modulate PIP gene expression levels in WT and 7B-1 mutant seeds (Balarynová and Fellner, unpublished results). Here, we confirmed statistically this trend and found that BL enhanced PIP1;2 and decreased PIP2;1 and PIP1;3 gene expressions in tomato seeds. It is known that BL reduces tomato seeds germination (radicle protrusion) (Piterková et al. 2012 and here Fig. 2). On the other hand, BL stimulates tomato root elongation (Balarynová and Fellner, unpublished results); thus, the same stimulatory effect could be expected for radicle. Therefore, the effect of BL on expression of PIPs genes involved in radicle elongation during seed germination could be dual. PIPs expression should be sufficient to ensure the radicle elongation but not so high to prevent immoderate loses of water which is important for cell elongation.

Literature reporting participation of aquaporins in seed germination or radicle growth is very limited (Obroucheva 2012; Obroucheva et al. 2017, Willigen et al. 2006). Interaction between light and aquaporins was studied especially in plant leaves because of their putative involvement in regulation of water transport and turgor in leaf cells. It was showed that plasma membrane aquaporins play a role in response of the leaf hydraulic conductance to light which stimulates leaf conductance to water. The light stimulation of leaf hydraulic conductance was not associated with stomatal opening. For example, Juglans regia aquaporin JrPIP2;1 and JrPIP2;2 transcript abundances are related to the dynamics of water transport in leaves (Cochard et al. 2007). Interestingly, BL and green light seem to stimulate leaf hydraulic conductance the most effectively (Baaziz et al. 2012). In our experiments, BL reduced significantly the percentage of seed germination and it was associated with downregulation of PIP1;3 and PIP2;1 genes. These results suggest that these PIPs might play an important role in water transport during seed germination. On the other hand, expression of PIP1;2 gene was induced by BL. Aquaporin PIP1;2 seems to be important element of a whole plant water status in A. thaliana. Its expression and activity was reported to be control by multiple factors including BL (Kaldenhoff et al. 1993; Postaire et al. 2010). Besides, PIP1;2 was one of the most transcribed aquaporin in germinated tomato seeds (Tables 3, 4). Thus, we can expect its significant contribution to maintain appropriate water status in tomato seeds under BL conditions. Moreover, PIP1;2 may compensate for the reduced expression of PIP1;3 and PIP2;1 genes. Finally, PIP1;2 gene could perform another roles in seeds (see below), because it is believed that PIP2 aquaporins are better water channels than PIP1 aquaporins (Maurel et al. 2015).

Aquaporins transport not only water but also other solutes of great physiological importance such as carbon dioxide, hydrogen peroxide or oxygen. For example, AtPIP1;2 seems to be a CO2 transporter (Uehlein et al. 2003, 2012) and Nicotiana tabacum aquaporin PIP1;3 is thought to be involved in oxygen transport. Thus, PIPs appear to play a dual function. They are important for preserving water homeostasis and also participate in regulation of photosynthesis or cell signalling. AtPIP1;4 is able to translocate extracellular hydrogen peroxide into the cytoplasm and active immune responses induced by bacteria (Tian et al. 2016).

There is a hypothesis that PIP1 aquaporins function as sensors of BL, which is supported by their localization in the plasma membrane and some of their structural features (Kaldenhoff and Eckert 1999; Lorenz et al. 2003). Interestingly, in our experiments, we showed that expression of PIP1;2 is significantly stimulated by BL but significantly reduced in 7B-1 mutant (Table 3). Then, the reduction of PIP1;2 expression in the mutant is in agreement with the fact that 7B-1 mutation confers reduced plant and/or seed responsiveness to BL (Fellner et al. 2001; Fellner and Sawhney 2002; Hlavinka et al. 2013; Ježilová et al. 2012). Therefore, we agree that PIP aquaporins could be more than water transporters in seeds.

Altogether, we showed that both osmotic stress and 7B-1 mutation decreased the expression levels of four tested PIP genes, probably to protect seed against excessive water loss during the seed germination under the osmotic stress. Moreover, BL affected PIP transcription levels indicating the connection between BL signalling and aquaporins. Thus, the inhibitory effect of BL and osmotic stress on tomato seed germination could be, at least partly, related to changes in aquaporin transcripts abundance. As 7B-1 mutation causes defect in phototropin responses (Hlavinka et al. 2013; Bergougnoux and Fellner, unpublished data), we hypothesize that PIP1;2 and PIP2;1 may play a role in BL sensing or signalling. This is supported by the fact that the expression of both genes is regulated by BL. Other experiments based on analysis of tomato phototropin mutants would support or disprove this hypothesis.

Author contribution statement

JB performed and designed all experiments and wrote the manuscript. JD made data analyses and edited the manuscript. MF designed the experiments and edited the manuscript.