Lycopersicon esculentum seeds germinate after rehydration in complete darkness. This response was inhibited by a far-red light (FR) pulse, and the inhibition was reversed by a red light (R) pulse. Comparison of germination in phytochrome-deficient mutants (phyA, phyB1, phyB2, phyAB1, phyB1B2 and phyAB1B2) showed that phytochrome B2 (PhyB2) mediates both responses. The germination was inhibited by strong continuous R (38 µmol m−2 s−1), whereas weak R (28 nmol m−2 s−1) stimulated seed germination. Hourly applied R pulses of the same photon fluence partially replaced the effect of strong continuous R. This response was called ‘antagonistic’ because it counteracts the low fluence response (LFR) induced by a single R pulse. This antagonistic response might be an adaptation to a situation where the seeds sit on the soil surface in full sunlight (adverse for germination), while weak R might reflect that situation under a layer of soil. Unexpectedly, the effects of continuous R or repeated R pulses were mediated by phytochrome A (PhyA). We therefore suggest that low levels of PhyA in its FR-absorbing form (Pfr) cause inhibition of seed germination produced either by extended R irradiation (by degradation of PhyA-Pfr) or by extended FR irradiation [keeping a low Pfr/R-absorbing form (Pr) ratio].
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Similar with A. thaliana, several phytochrome-deficient mutants also exist in tomato (Kendrick et al. 1997; Emmanuel & Levy 2002). However, the phytochrome genes existing in tomato are different from those in A. thaliana; tomato having no phyC and phyD, but having phyF and a second phyB instead (Pratt et al. 1997). In contrast with A. thaliana, which requires phytochrome activation to induce effective germination, tomato seeds can germinate in total darkness. Initial studies on tomato showed that while continuous FR inhibits seed germination in wild type, the chromophore-deficient aurea mutant (Terry 1997) retains some capacity for germination under continuous FR (Georghiou & Kendrick 1991). At the same time, R promotes the germination of the tomato aurea mutant, indicating the participation of phytochrome(s) in regulating dark germination in tomato (Georghiou & Kendrick 1991). Subsequently, van Tuinen et al. (1995) demonstrated that the PhyA-deficient fri mutant does not show FR-mediated inhibition of seed germination, indicating the involvement of PhyA in the FR response. Shichijo et al. (2001) used the frihp-1w, trihp-1w and hp−1w mutants to study FR inhibition of seed germination in tomato. They showed that FR inhibition consisted of an LFR and assumed an HIR to describe the effect of continuous FR. While phytochrome B1 (PhyB1) and PhyA did not participate in the regulation of LFR, the effect of continuous FR was mediated by PhyA (Shichijo et al. 2001).
In the present study, we have used the phyA, phyB1 and phyB2 mutants and combinations of these mutants to examine the participation of different phytochrome species in the regulation of tomato seed germination. We introduce the term ‘antagonistic response’ to describe a novel phytochrome response generated by continuous R or repeated R pulses, which antagonizes the LFR.
MATERIALS AND METHODS
Seeds of wild-type Lycopersicon esculentum Mill., cv. Moneymaker (Koornneef et al. 1985) were either purchased from N.L. Chrestensen, Samen- und Pflanzenzucht GmbH (Erfurt, Germany) or received from R.E. Kendrick (Agricultural University of Wageningen, Wageningen, the Netherlands) as a gift. Several batches of seeds were produced in the Botanical Garden of the University of Jena, Jena, Germany. Harvests from 1994, 1998, 1999 and 2000 were used. Most of the phytochrome mutants used were harvested in 1999 and 2000. For the phyB1B2 mutant, only one harvest (2000) was available. Further experiments were also carried out with the phyA and phyB1 harvested 1994, phyAB1 harvested in 1996 and the aurea mutant harvested in 1988. The GT cultivar was also investigated, as some of the mutations were produced in this genetic background (kindly provided by M. Koornneef, Plant Science Group, Wageningen, the Netherlands). All experiments were carried out in 2002 and 2003, with the exception of the wild-type Moneymaker harvested in 2000 and investigated in 2001, and the GT harvested in 2000 and investigated in 2005. The genetic characteristics of the phytochrome-deficient mutants are summarized in Table 1. Several batches were investigated because the degree of dormancy as well as the dark response was very variable.
Table 1. Summary of the phytochrome mutants used in the present investigations
The tomato seeds were surface-sterilized for 4 min with 0.7% (w/v) NaOCl and washed five times with sterile water. Lots of 25 sterilized seeds were put in ultraviolet (UV)-sterilized transparent plastic boxes (10 × 10 × 6 cm) on a layer of 100 cm3 agar (0.7% w/v; Serva, Heidelberg, Germany) containing the following nutrients: 40 mm KNO3, 0.3 mm CaCl2, 0.15 mm MgSO4, 0.125 mm KH2PO4, 10 µm Na2EDTA, 10 µm FeSO4, 10 µm H3BO3, 10 µm MnSO4, 3 µm ZnSO4, 0.5 µm KI, 0.1 µm Na2MoO4, 10 pm CuSO4 and 10 pm CoCl2 adjusted to pH 5.7 (cf. Murashige & Skoog 1962). In most cases, an FR pulse (10 min) was applied 7 h after soaking the seeds and experimental light treatments were applied thereafter. Temperature was kept at 25.0 ± 0.2 °C during all experiments. Manipulations were carried out in dim green light (553 ± 8 nm, < 0.2 µmol m−2 s−1; cf. Appenroth & Gabrys 2001). Radicle protrusion was used as a criterion for judging seed germination.
Commercial slide projectors (S-slide projector Diafant 250 (Liesegang Optoelectrics, Duesseldorf, Germany) (24 V/250 W) were used as light sources. For continuous R (37 µmol m−2 s−1) and R pulses (490 µmol m−2 s−1), light passed through a filter combination of RG645 (3 mm thick; Schott, Jena, Germany) and FR cut-off filters IR7 (three times, 3 mm thick; OptoChem, Stromberg, Germany), resulting in a wavelength maximum of 658 ± 25 nm. Light for FR pulses (745 ± 57 nm, 65 µmol m−2 s−1) were filtered through a cut-off filter RG9 (3 mm; Schott) and an IR6 filter (3 mm; OptoChem). Seven hours after soaking, the seeds were either irradiated with light as indicated or immediately following a pre-treatment with FR for 10 min. Light pulses were applied for either 4.5 (R) or 10 min (FR). Repeated light pulses of R and FR were applied every hour for 4.5 min and 10 min, respectively. When necessary, photon fluence rates of continuous R were decreased by neutral density filters.
In most cases, eight germination boxes were loaded with seeds (50 each) and germination was investigated in two or three independent experiments (n = 8). Averages were calculated as mean values of the results from eight boxes and SEs were given. Significance was calculated using the two-sided Student's t-test (5% level).
The germination response of wild-type seeds was observed in complete darkness without any light treatment (Fig. 1). The rate and final level of response depended, to a large extent, on the storage period between harvest and experimentation. The seeds with the shortest period of storage (harvested in 2000) had practically no dormancy and germinated to ≈ 95% after 3 d in darkness. In seeds that were for 7 years (harvested in 1994) at room temperature (18 °C, 80% relative humidity), the maximal germination rate dropped to 25%. The effects of single R pulse and continuous R were also investigated in seeds of the wild types (Fig. 1). A single R pulse strongly induced germination in both wild types. This effect has been known for a long time (Mancinelli, Borthwick & Hendricks 1966). Unexpectedly, however, the application of a continuous R did not stimulate seed germination (Fig. 1). In the wild type with low dark germination (harvested in 1994), continuous R (38 µmol m−2 s−1) did not stimulate the response, in comparison with the dark response; and in the wild type with higher dark germination (harvested in 1999), there was even a strong effect of photoinhibition. In comparison with the effect of short R pulses followed by darkness, extended R irradiation inhibited germination in seeds from all harvests.
The germination responses of wild types in darkness could be decreased by the application of a saturating FR pulse (10 min) at different periods after soaking (Fig. 2). There was a strong inhibition when the FR pulse was applied 3 or 10 h after soaking, and even a pulse after 16 h was only slightly less effective. After 24 h, however, FR irradiation had little ability to inhibit germination response. The inhibiting effect of a single FR pulse on the rate of germination was likely caused by the photoreversion of pre-existing phytochrome from the FR-absorbing form Pfr into its R-absorbing form, Pr. Therefore, pre-irradiation by a single FR pulse 7 h after soaking (approximately the median point between 3 and 10 h) was used in several of the following experiments to investigate the effect of subsequent experimental light treatments.
The inhibition of seed germination by continuous R (cf. Fig. 1) was further investigated in the experiments that followed. Continuous R was applied and the influence of fluence rates was investigated. The lowest fluence rate used (28 nmol m−2 s−1) showed the strongest promotion of germination and was as effective as a single R pulse (Fig. 3). Increasing the fluence rate of extended R irradiation resulted in increased inhibition of germination, in comparison with a single R pulse. In some germination experiments, the cultivar GT (only 50 seeds per data point, instead of the Moneymaker, was used because some of the phytochrome-deficient mutants were produced in this genetic background (Table 1). No essential differences in the light responses of the two cultivars were detected (data not shown).
The effects of single R pulses and continuous R were compared with those of repeated, hourly applied light pulses (R, FR and R followed by FR; Fig. 4) to distinguish between the inducing (LFR) and non-inducing effects of light (HIR; Casal, Sanchez & Botto 1998). Whereas a single R pulse induced a strong germination response, repeated R pulses were significantly less effective (Student's t-test). In other words, the subsequently applied R pulses partially antagonized the effect of the first R pulse. Similar results were obtained using several other seed batches (harvested in 1998, 1999 and 2000; data not shown). The germination response in continuous R was even lower and was closer to the effect of a single FR pulse. The inhibiting effect of continuous R was stronger than that of repeated R pulses with the same hourly applied photon. Thus, the reciprocity law was not fulfilled. In the additional experiments, we intended to test whether the effect of hourly applied R pulses could be photoreverted by subsequent hourly applied FR pulses. However, hourly applied FR pulses alone (FRp h−1) were found to inhibit germination. The combined light treatment, i.e. hourly applied R pulses followed by hourly applied FR pulses (Rp + FRp h−1), decreased the germination response further (Fig. 4). Therefore, hourly applied FR pulses following hourly applied R pulses are not suitable for testing the photoreversibility of the response in tomato seed germination.
Seeds of phytochrome mutants
The germination response of several phytochrome mutants was determined 96 h after the seeds were soaked. In wild type, single FR pulses inhibited germination in relation to the dark response, resulting in positive values of the difference of germination percentage (D-FRp; Fig. 5). In phyA and phyB1 mutants, the D-FRp differences of germination percentage were comparable with those in wild-type plants, i.e. both were higher than zero. In the phyB2 mutant, however, there was no significant inhibition of germination by a single FR pulse as compared with the dark response. This was further confirmed using several batches of phyB2 mutant seeds harvested in 1999 and 2000 (data not shown). Therefore, phytochrome B2 (PhyB2) mediates this inhibiting effect of a single FR pulse. In accordance with this result, the photoinhibition by a single FR pulse was observed in the phyAB1 double mutant and was not detectable in the triple mutant phyAB1B2. The inhibiting effect in the phyB1B2 double mutant and in the aurea mutant was very small (no significant differences).
The inhibiting effect of repeated FR pulses was stronger than that of a single FR pulse (cf. Fig. 4), resulting in negative values of the difference of germination percentage, (FRp h−1-FRp; Fig. 5). This was shown for the wild-type seeds and for the two monogenic mutants, phyB1 and phyB2. In the phyA mutant, however, there was no significant difference between the effect of a single and of repeated FR pulses. It can be concluded that PhyA is responsible for the photoinhibition mediated by repeated FR pulses. In accordance with this conclusion, no photoinhibition in the germination response of the mutants phyAB1 and phyAB1B2 and in aurea was detectable. The absence of this effect in the phyB1B2 double mutant is difficult to understand.
A single R pulse stimulated the germination response, as shown in Fig. 4, for wild type. This effect was also investigated using seeds of the phytochrome mutants (Fig. 6). The effect, however, was not observed in the phyB2 mutant (i.e. no significant differences between responses in darkness and following a single R pulse), thus indicating PhyB2 as the operating photoreceptor. The results in the mutants phyAB1 (significant stimulation detectable) and phyB1B2, and phyAB1B2 and aurea (no significant stimulation detectable) confirmed this conclusion.
The effects of single R pulses, repeated R pulses and continuous R on seeds of wild type and phytochrome mutants were compared in Fig. 7. In most cases, the inhibiting effect of continuous R was stronger than that of repeated R pulses. In the phyA mutant, however, both light treatments slightly enhanced inhibition, in comparison with only a single R pulse. Thus, PhyA seems to be the photoreceptor involved in the inhibiting effect of repeated R pulses and continuous R. In the aurea mutant, both light treatments stimulate the germination response more, compared with only a single R pulse, which might be a consequence of the very low concentration of active phytochrome in this mutant. There was also a slight stimulation by repeated R pulses in the mutants phyB2 and phyB1B2, which might be caused by some unknown contribution of PhyB2 to this response.
In many plant species, such as A. thaliana (Shropshire, Klein & Elstad 1961), Lactuca sativa Grand Rapids (Hartmann 1966) and Oryzopsis miliacea (Negbi & Koller 1964), seed germination is induced or promoted by a single R pulse (cf. Casal & Sanchez 1998). Tomato, however, is an example among a few species, such as Cucumis sativus (Mancinelli & Tolkowski 1968) and varieties of L. sativa other than Grand Rapids (Small et al. 1979), where seeds require only rehydration in the dark in order to germinate. Pre-existing Pfr is regulating this response; thus, the response can be prevented by FR irradiation. The responsivity of tomato seeds to light (R or FR pulses) is restricted to a short time interval spanning from 3 to 16 h from the onset of rehydration of seeds.
In several systems, seed germination is not sufficiently stimulated by a single R pulse and requires either continuous light or repeated pulse exposure (Grubšić & Konjević 1990; Bewley & Black 1994), while in tomato, a single R pulse is sufficient to stimulate seed germination. However, continuous R or repeated R pulses decrease the germination response, in comparison with the effect of a single R pulse. Thus, extended R irradiation acts antagonistically to the effect of a single R pulse. We use the term ‘antagonistic response’ to describe this novel phytochrome response in tomato seed germination because an adequate light treatment (repeated R pulses or continuous R) antagonized the effect of the R pulse-induced LFR. The antagonistic response might be important in controlling the emergence of seedlings from seeds planted at or below the soil surface (Bliss & Smith 1985; Benvenuti 2003; Chung & Paek 2003). Such an antagonistic response is not seen in many plant species. To our knowledge, Phacelia tanacetifolia (Schulz & Klein 1963; Rollin & Maignan 1967) is the only species wherein germination is more strongly inhibited by continuous R rather than by continuous FR (cf. also Pirovano et al. 1999). The mechanism of photoinhibition in this species is not known. Hendricks, Tool & Borthwick (1968) reported the very weak inhibition of Amaranthus arenicola and Poa pratensis germination by continuous R, in comparison with that by continuous FR. The inhibition of seed germination in tomato by strong R has been shown here for the first time.
An analysis of phytochrome mutants in tomato shows that photoregulation of seed germination takes place in a fashion radically different from that in A. thaliana. In A. thaliana, the induction of seed germination by R is mediated by PhyB and PhyA, whereas induction by FR is mediated only by PhyA (Shinomura et al. 1996; Hennig et al. 2001). In the phyAphyB double mutant, a single R pulse weakly stimulates promotion of seed germination whereas repeated R pulses or continuous R stimulates a high percentage of germination (Poppe & Schäfer 1997). The contribution of phytochrome to tomato seed germination is evident in the aurea mutant, which shows reduced inhibition by continuous FR (Georghiou & Kendrick 1991). Because the aurea mutant is deficient in phytochrome chromophore biosynthesis, it may have an effect on all species of phytochromes and may serve as a control deficient in all phytochrome species (Sharma et al. 1993; Terry 1997). Using the available monogenic mutants lacking different phytochrome species (i.e. phyA, phyB1 and phyB2), as well as the double and triple mutants (Kendrick et al. 1997; Pratt et al. 1997), the role of specific phytochromes in the regulation of seed germination in tomato was ascertained in this study. As summarized in Table 2, the inhibition of the dark response by a single FR pulse is not observed in the phyB2 mutant. This result has been confirmed by an analysis of the response of the double and triple mutants. Induction of germination by a single pulse of R is therefore mediated by PhyB2. Neither PhyB1 nor PhyA play any detectable role in this LFR response. The dominant function of PhyB2 in seed germination of tomato has been shown here for the first time.
Table 2. Summary of the observed light responses in tomato seed germination
Effects on seed germination
R, red light; FR, far-red light.
Reversion of the effects of R pulse
Hourly FR pulses
Inhibition of the dark response
Hourly R pulses
In this paper, it was revealed that the inhibiting effect of repeated FR pulses is mediated by PhyA, as shown by the hourly applied FR pulses on the Moneymaker background and the six hourly applied light pulses on the hp-1w background (Shichijo et al. 2001). The same phytochrome species also mediates the effect of continuous FR (Shichijo et al. 2001).
Surprisingly, the antagonistic response initiated by R irradiation (continuous R or repeated R pulses) is also mediated by PhyA (Table 2). Assuming the existence of an R-HIR, one would expect the involvement of a PhyB-type photoreceptor (Smith 1995; Shinomura et al. 2000), which is evidently not the case. The establishment of a low PhyA-Pfr level above a threshold, whether by continuous R or FR or by repeated pulses of R or FR, causes the observed inhibition of the germination response. Similar levels of PhyA-Pfr may be produced under very different irradiation conditions, either by (1) extended FR irradiation as a direct consequence of the photo-equilibrium between Pr and Pfr or by (2) extended R irradiation indirectly by high degradation of the Pfr form of PhyA (cf. Zhou et al. 2002). Based on the results in P. pratensis and A. arenicola seeds, Hendricks et al. (1968) suggested an alternative explanation: inhibition by continuous irradiation could be the result of the same process, with highest effectiveness around 730 nm and low effectiveness in the range of 600–680 nm. In tomato, however, the effectiveness of R (this paper) is comparable with that of FR (Shichijo et al. 2001), and in P. tanacetifolia (Schulz & Klein 1963), R is even more effective. Moreover, it has been shown here that the photolabile PhyA mediates the antagonistic effect in tomato. Therefore, the two separate processes suggested in the present paper seem to be more adequate for describing the inhibitory effects of R and FR on seed germination.
Hourly applied R pulses only partially mimic the effect of continuous R irradiation on seed germination in tomato, demonstrating the existence of two distinct responses, i.e. an inducing reaction by repeated R pulses (LFR, cf. Casal et al. 1998) and a non-inducing response specifically promoted by continuous R (cf. Appenroth & Teller 2004). This is similar to the effects of FR irradiation. The different response modes as mediated by distinct phytochrome species in the germination of tomato seeds are summarized in Table 2. The fact that some reduction of germination is also seen in the phyB2 mutant background might indicate the participation of other phytochrome species by a co-action of more than one phytochrome.
We thank Prof. R.E. Kendrick, University of Wageningen, Wageningen, the Netherlands, for the generous gift of seeds of all the mutants investigated in this paper and for the helpful discussions. We also thank Prof. M. Koornneef, Plant Science Group, Wageningen, the Netherlands, for providing a batch of tomato GT seeds. Dr R. Wejnar, University of Jena, Jena, Germany, helped us produce seeds from several mutants.