Stressed plants might flower as an emergency response to produce the next generation. In this way, plants can preserve their species, even in unfavorable environments. In order for this to be a biologically advantageous response, plants induced to flower by stresses must produce fertile seeds, and the progeny must develop normally.
P. nil Violet was grown in a 1/10-strength nutrient solution or tap water throughout its life. The plants that were induced to flower by poor-nutrition stress conditions reached anthesis, fruited and produced seeds (Wada et al. 2010a). The seeds produced by the stressed plants were the same size as or slightly smaller than the control seeds produced by plants that flowered by short-day treatment. All of these seeds germinated, and the progeny developed normally. The progeny responded to short-day treatment and formed floral buds. Furthermore, a normal second generation was produced from the stress progeny.
Red-leaved P. frutescens plants were grown under long-day conditions with low-intensity light beginning at the stage in which the cotyledons expanded. Plants were then continuously grown under the same conditions. The plants induced to flower by the low-intensity light stress conditions reached anthesis and formed seeds (Wada et al. 2010b). There were four seeds per flower as in the normal plants. The seeds produced under low-intensity light were heavier than the control seeds produced under usual short-day conditions. The seeds produced under stress conditions germinated, and the progeny grew normally and were induced to flower in response to short-day treatments.
These results in P. nil and P. frutescens indicate that the stressed plants do not need to await the arrival of a season when photoperiodic conditions are suitable for flowering, and such precocious flowering might assist in species preservation. Therefore, stress-induced flowering might have a biological benefit, and it should be considered to be as important as photoperiodic flowering and vernalization.
Transmissible flowering stimulus produced by stress
The presence of cotyledons is necessary for the long-day flowering of P. nil in response to poor nutrition or low temperature (Shinozaki and Takimoto 1982; Shinozaki 1985). This suggests that a flowering stimulus like florigen, which is involved in photoperiodic flowering, is involved in stress-induced flowering and is produced in cotyledons. If the stress-induced flowering stimulus is transmissible, defoliated scions may flower when grafted onto rootstocks with cotyledons and grown under stress conditions.
P. nil Violet and Tendan were grafted in several combinations, and the grafted plants were grown in tap water under long-day conditions for 20 days (Wada et al. 2010a). The Violet scions grafted onto the Violet rootstocks flowered (Fig. 1). The flowering may have been caused by the influence of the rootstocks because all the leaves had been removed from the scions and the cotyledons had been maintained on the rootstocks. This suggests that a transmissible flowering stimulus is involved in the stress-induced flowering of P. nil.
Violet scions flowered even when grafted onto Tendan rootstocks, although Tendan plants themselves were not induced to flower by the stress treatment. On the other hand, Tendan scions did not flower when grafted onto Violet rootstocks. It was predicted that Tendan would not produce such a flowering stimulus because Tendan did not flower in response to the poor-nutrition stress conditions. If this were the case, Violet would not be expected to flower when grafted onto Tendan rootstocks. However, defoliated Violet scions grafted onto Tendan rootstocks with cotyledons were induced to flower. The difference in flowering response between the scions grafted onto Tendan and those grafted onto Violet was not statistically significant. Therefore, Tendan may produce almost the same amount of the flowering stimulus as does Violet. Conversely, the Tendan scions grafted onto Violet rootstocks were not induced to flower. These results indicate that Tendan produces a transmissible flowering stimulus but does not respond to it.
Endogenous substances involved in stress-induced flowering
CGA and some other phenylpropanoids were found to accumulate in cotyledons during the treatments by poor nutrition, low temperature or high-intensity light in P. nil (Shinozaki et al. 1988a, b, 1994; Hirai et al. 1993, 1994). Phenylpropanoid synthesis is involved in the stress response (Dixon and Paiva 1995). Stress promotes the metabolism of t-cinnamic acid to SA via benzoic acid (Gidrol et al. 1996; Mauch-Mani and Slusarenko 1996). The flowering induced by these conditions is accompanied by an increase in PAL activity (Hirai et al. 1995), and AOA inhibited flowering in P. nil (Shinozaki et al. 1988a, 1994, Hatayama and Takeno 2003). Some compound(s) in the metabolic pathway regulated by PAL might act as flowering stimuli. Phenylpropanoids, such as CGA, were prominent candidates for this in earlier studies (Shinozaki et al. 1988a, b, 1994; Hirai et al. 1993, 1994). However, exogenously applied CGA failed to induce flowering (Shinozaki et al. 1988a, 1994; Hatayama and Takeno 2003). No flower-inducing activity was detected in other phenylpropanoids, including 4-O-p-coumaroylquinic acid, 3-O-feruloylquinic acid, dehydrodiconiferylalcohol-13-O-β-D-glucoside, and (+)-pinoresinol-β-D-glucoside (unpublished data). Therefore, CGA and related phenylpropanoids are not involved in the stress-induced flowering of P. nil. The close positive correlation between CGA content and flowering response was merely coincidence.
In addition to CGA, several compounds including SA and anthocyanin are derived from t-cinnamic acid of which conversion from phenylalanine is catalyzed by PAL (Dixon and Paiva 1995). Dihydrokaempferol-7-O-D-glucoside derived from the pathway from t-cinnamic acid to anthocyanin via p-coumaric acid has been reported to promote the flowering of P. nil (Nakanishi et al. 1995). Furthermore, AOA inhibits 1-aminocyclopropane-1-carboxylic acid (ACC) synthase. ACC synthase catalyzes the conversion of S-adenosylmethionine to ACC, which is converted to ethylene. Such components or the other substances in the metabolic pathways derived from t-cinnamic acid might be involved in stress-induced flowering. Accordingly, the flowering of P. nil was induced by low-temperature or poor-nutrition stress, and AOA treatment was used to inhibit the flowering. Several metabolic intermediates in the pathways were applied together with AOA (Hatayama and Takeno 2003; Wada et al. 2010a). Among the intermediates tested, t-cinnamic acid, benzoic acid and SA were shown to counteract the inhibitory effect of AOA (Fig. 2), whereas p-coumaric and caffeic acids did not. These results suggest that SA is involved in the stress-induced flowering of P. nil and that the pathways to CGA and anthocyanin are not involved. Flowering was completely inhibited in the presence of ACC (Hatayama and Takeno 2003). Thus, the ACC route is not involved. This is consistent with the observation that ethylene derived from ACC inhibits the photoperiodic flowering of P. nil (Suge 1972).
The leaves of red-leaved P. frutescens were deep green when induced to flower under low-intensity light (Wada et al. 2010b). The greening of the leaves was due to a decrease in anthocyanin content. There was a negative correlation between anthocyanin content and percentage flowering. Therefore, the metabolic pathway related to anthocyanin synthesis may be involved in the regulation of flowering. It is possible that some substances such as SA which are synthesized by the common metabolic pathway for anthocyanin synthesis are involved in flowering as mentioned above for P. nil. Low-intensity light may induce the flowering of P. frutescens by influencing the endogenous level of SA through suppression of PAL activity. However, this conflicts with previous reports. Stress generally increases PAL activity and promotes anthocyanin biosynthesis (Christie et al. 1994; Dixon and Paiva 1995; Chalker-Scott 1999). Actually, PAL activity increases in the stress-induced flowering of P. nil as mentioned above. Therefore, it was examined whether the PAL inhibitor could promote or inhibit the low-intensity light stress-induced flowering in P. frutescens (Wada et al. 2010b). The PAL inhibitor AOPP did not induce flowering when applied under non-inductive normal-intensity light conditions and inhibited flowering in a dose-dependent manner when applied under inductive low-intensity light stress conditions (Fig. 3). The treatment with another PAL inhibitor, AOA, gave the same results. These results suggest that the same mechanism is involved in flowering that is induced by low-intensity light in P. frutescens and the flowering that is induced by several stress factors in P. nil. That PAL inhibitors inhibited stress-induced flowering suggests that the stress increased PAL activity. However, in P. frutescens, the decrease in anthocyanin content under low-intensity light suggests that stress limited the activity of PAL. These contradictory results must be explained in future.
Involvement of SA in stress-induced flowering
When plants are stressed, they generate stress substances that regulate gene expression to adapt to the stress conditions. The stress substances include reactive oxygen species, nitric acid, jasmonic acid, SA, ethylene and abscisic acid (Xiong et al. 2002; Moreau et al. 2010; Liu and Zhang 2004; Hey et al. 2010; Jaspers and Kangasjarvi 2010). Among these stress substances, SA and ethylene have been reported to induce flowering. Ethylene induces flowering in the Bromeliaceae, including pineapple. However, this is an exceptional case, and ethylene generally inhibits flowering in many plant species. The most likely stress substance involved in stress-induced flowering may be SA.
UV-C light stress promotes flowering in wild-type A. thaliana, but does not in SA-deficient nahG transgenic plants (Martínez et al. 2004). UV-C irradiation increased the expression of the SA-responsive PR1 gene in Col but not in nahG plants. The transcript of the SA induction deficient 2/isochorismate synthase 1 (SID2/ICS1) gene encoding the SA biosynthetic enzyme increased under UV-C irradiation in Col but not in nahG plants. These results suggest the involvement of SA in the UV-C stress-induced flowering of A. thaliana. Exogenous application of SA at 100 μM accelerated flowering of Col, but the nahG plants were not responsive to the SA treatment (Fig. 3). SA also regulates flowering time in non-stressed plants. SA-deficient nahG is late flowering (Martínez et al. 2004). The siz1 mutant that has elevated SA level is early flowering under short-days, and this phenotype is suppressed by expression of nahG (Jin et al. 2008).
When L. paucicostata 6746 was induced to flower by poor-nutrition stress, a larger amount of SA was detected in the flowered plants than in the control plants (Shimakawa 2011). This result suggests the involvement of SA in the stress-induced flowering of L. paucicostata. It is well known that exogenously applied SA induces flowering in L. paucicostata, L. gibba and the other Lemanceous plants (Cleland and Ajami 1974; Cleland and Tanaka 1979; Cleland et al. 1982). However, SA is not considered to be an endogenous flower-regulating factor in Lemna because the endogenous SA level is not altered by photoperiodic conditions (Fujioka et al. 1983). SA may be the endogenous flower-regulating factor in stress-induced flowering but not in the photoperiodic flowering of Lemna.
The treatment of P. nil with SA and benzoic acid, a precursor of SA, or some benzoic acid derivatives prior to low-temperature treatment enhances the flower-inducing effect of low temperature (Shinozaki 1985; Shinozaki et al. 1982, 1985). In addition to these effects of exogenous application, the flower-inhibiting effects of PAL inhibitors, which may have decreased the endogenous SA level in P. nil and P. frutescens, provided new evidence to suggest that SA acts as an endogenous regulator of stress-induced flowering (Wada et al. 2010a, b). The flowering response of cultured plumules excised from short-day treated P. nil seedlings was enhanced by benzoic acid (Ishioka et al. 1990). Amagasa et al. (1992) reported that AOA inhibited the photoperiodic flowering of P. nil. These observations suggest that SA is also involved in photoperiodic flowering. However, SA did not induce flowering at any concentrations in P. nil and P. frutescens under non-stress conditions (Wada et al 2010a, b). SA did not enhance the flowering response under the weak stress conditions. SA may be necessary but is not sufficient for the induction of flowering. Stress conditions may induce not only SA biosynthesis but also other essential factors to induce flowering.
The genes involved in stress-induced flowering
Expression of the CO, FT and SOC1 genes that promote flowering was analyzed in A. thaliana under UV-C stress conditions (Martínez et al. 2004). UV-C induced expression of FT, moderately induced expression of CO, and did not induce SOC1 expression in wild type (Fig. 4). Exogenous SA treatment reduced expression levels of the flower-inhibiting gene FLC. Thus, flowering promoted by UV-C requires the enhanced expression of FT and the reduced expression of FLC. SA application induced expression of the sunflower FT homolog, HAFT, in sunflower (Dezar et al. 2010). The flowering of A. thaliana is induced by long-day conditions, vernalization, autonomous cues and gibberellins, and these factors operate through a common pathway integrated by FT (Boss et al. 2004). It was shown that FT is also involved in stress-induced flowering.
Genome-wide analyses of transcriptomes detected the down-regulation of Pathogen and Circadian Controlled 1 (PCC1) in SA-deficient plants of A. thaliana (Segarra et al. 2010). PCC1 was initially characterized as a circadian clock-regulated gene that is rapidly up-regulated after pathogen inoculation. The expression of PCC1 was strongly activated by UV-C light irradiation in Col but not in nahG plants. SA application also activated PCC1 expression. The activation of PCC1 expression required CO. RNAi transgenic plants contained lower levels of FT transcript. The over-expression of PCC1 did not accelerate flowering, but suppression of its expression by RNAi delayed flowering. UV-C light irradiation of plants accelerates flowering through a SA-dependent process in wild-type but not in RNAi transgenic plants with reduced expression of PCC1, suggesting that neither SA nor PCC1 alone is sufficient to accelerate flowering in A. thaliana.
The flowering of A. thaliana is induced by four previously known factors and stress, and these factors function through the activation of FT expression. This suggests that the FT homolog could be involved in stress-induced flowering in other plants. Two orthologs of FT, PnFT1 and PnFT2, have been identified in P. nil, and these genes are expressed under inductive short-day conditions to promote flowering (Hayama et al. 2007). Therefore, the expression of PnFT genes in response to poor-nutrition stress conditions was examined. P. nil Violet was induced to flower by growth in tap water, the cotyledons and true leaves of these plants were collected, and the expression of PnFT1 and PnFT2 was examined by RT-PCR (Wada et al. 2010a; Yamada 2011). The expression of PnFT1 and PnFT2 was induced in cotyledons by a single short-day treatment, but neither gene was expressed without the short-day treatment. The expression of PnFT2 was induced in the cotyledons and true leaves of plants grown under the poor-nutrition conditions for two weeks or longer. The level of mRNA expression was closely correlated with the flowering response. Only weak PnFT2 expression was detected in the true leaves of plants grown under non-stress conditions for three weeks. On the other hand, PnFT1 was not expressed in the cotyledons or true leaves regardless of nutritional conditions. These results suggest that PnFT2, but not PnFT1, is involved in the poor-nutrition stress-induced flowering of P. nil. PnFT2 is involved in both photoperiodic flowering and stress-induced flowering, whereas PnFT1 is involved only in photoperiodic flowering. The two PnFT genes might have different roles in the regulation of flowering depending on the inductive cue. It is also possible that the essential gene for flowering is PnFT2 and that PnFT1 expression is induced only by short-day treatment and redundantly enhances the activity of PnFT2. SA might induce the expression of PnFT2, or the product of PnFT2 might induce the expression of genes involved in the biosynthesis of, response to or signal transduction of SA.
Concluding remarks
　It is apparent that plants can flower in response to several stress conditions. Constantly exposed to stresses that have negative effects on growth and development, plants establish protection and adaptation strategies to minimize stress influences. However, the protection or adaptation mechanism may not be sufficient if the stress is too severe. Precocious flowering may assist in species preservation under such conditions. Thus, stress-induced flowering can be considered an ultimate adaptation to stress and should be considered a central component, along with tolerance, resistance and avoidance, of stress physiology.
