Submergence research using Rumex palustris as a model; looking back and going forward

Abstract

Flooding is a phenomenon that destroys many crops worldwide. During evolution several plant species evolved specialized mechanisms to survive short‐ or long‐term waterlogging and even complete submergence. One of the plant species that evolved such a mechanism is Rumex palustris. When flooded, this plant species is capable to elongate its petioles to reach the surface of the water. Thereby it restores normal gas exchange which leads to a better survival rate. Enhanced levels of ethylene, due to physical entrapment, is the key signal for the plant that its environment has changed from air to water. Subsequently, a signal transduction cascade involving at least four (classical) plant hormones, ethylene, auxin, abscisic acid, and gibberellic acid, is activated. This results in hyponastic growth of the leaves accompanied by a strongly enhanced elongation rate of the petioles enabling them to reach the surface. Other factors, among them cell wall loosening enzymes have been shown to play a role as well.

Ethylene, phytohormones, Rumex, stress responses, submergence.

Introduction

Due to their sessile nature plants cannot physically escape biotic and abiotic stresses. As a consequence, they are forced to cope with their direct environment and have evolved acclimations and adaptations to counteract the long‐ and short‐term stresses they are exposed to. On the other hand certain environments will exclude plant species because of the specific adaptations that are required to survive. This selective mechanism accounts for biodiversity and, consequently, the ability of plants to invade almost all habitats on earth.

One of the abiotic stresses that has a dramatic impact on wild plants and crops worldwide is periodic flooding. Flooding can be defined as any situation with excess of water. In this review waterlogging is defined as the saturation of the soil with water around the roots, and submergence when the whole plant is completely covered by water. Different plant species vary widely in their degree to flooding tolerance, some species that inhabit floodplains, such as Rumex palustris are extremely tolerant, whereas others, such as many important crop species, are killed by just a few days of flooding (Voesenek et al., 1999). The adaptations these tolerant plants have evolved include anatomical, morphological (e.g. formation of aerenchyma, shoot elongation), and metabolic mechanisms (Colmer et al., 1998; Ricard et al., 1998; Jackson and Armstrong, 1999). These species are ideal for the study of flooding tolerance and to find the morphological, physiological and underlying molecular mechanisms they have evolved to cope with flooding stress. It is evident that a better understanding of mechanisms for flooding tolerance will lead to improved crops.

This review deals with the study of the terrestrial plant Rumex palustris, a submergence‐tolerant species which specifically inhabits environments that are periodically flooded. This plant species has been the subject of the study of flooding tolerance for more than 15 years (Blom and Voesenek, 1996), and the aim here is to review the progress that has been made to elucidate the signal transduction pathways that are involved in the tolerance towards submergence and to outline future research.

Description of the response to flooding

As plants are submerged, several changes in the evironment take place. Most importantly, the exchange of gases is severely limited, since gases diffuse by a factor of 10000 slower in water than in air (Armstrong, 1979). This induces a change in the endogenous concentration of O₂. A decrease from 21 to 3–10 kPa has been observed in situ in submerged petioles (Rijnders et al., 2000). The gaseous plant hormone ethylene increases due to physical entrapment. Assuming that the experiment is performed in a normal day–night rhythm, the CO₂ concentration will vary accordingly (Voesenek and Blom, 1999). The sensing of these changes and the subsequent signal transduction cascade ultimately set off the responses leading to the so‐called aerobic escape.

When Rumex palustris rosettes are subjected to complete submergence the orientation of petioles and leaf blades changes within a few h from rather horizontal to almost vertical (Voesenek and Blom, 1989). This phenomenon known as hyponastic growth results in an increase of the angle between the petiole and the axis of the gravitropic force. This response has been described in combination with shade avoidance (Ballaré, 1999; Smith, 2000). Preliminary results indicate that this is a prerequisite for the onset of stimulated elongation of (mainly) the petiole (MCH Cox and LACJ Voesenek, unpublished results). This differential growth is accompanied by enhanced elongation growth of the entire petiole. Microscopic analysis of the petiole and the leaf showed that, during the elongation phase, it is solely elongation of the cells that is causing the increase in length (Voesenek et al., 1990) and not a combination of cell division and elongation as described for deep water rice internodes (Kende et al., 1998; Sauter, 2000) and petioles of Ranunculus sceleratus and Nymphoides peltata (Ridge, 1987). The fast growth will last until contact with the air above the water is restored (Voesenek et al., 1993). The complete signal transduction pathway resulting in enhanced shoot elongation is not yet resolved, although certain components and their timing of action are known. The hormonal regulation of hyponastic growth is largely unknown, but it definitely requires ethylene (Voesenek and Blom, 1989), and probably also auxin (MCH Cox and LACJ Voesenek, unpublished results). The non‐differential elongation itself depends on ethylene and gibberellin (GA) signalling, whereas abscisic acid (ABA) signalling most likely has to be down‐regulated (Hoffmann‐Benning and Kende, 1992). Further downstream, there are indications that expansins, a family of cell wall proteins, that are involved in cell wall loosening (Cosgrove, 1999), are necessary to enable the cells to enlarge.

Cell elongation

The plant cell wall is a poorly understood fibrous structure whose properties, besides other factors, determine the form and function of plants. Cell elongation in general is a complex process that involves changes in pH caused by auxin, changes in turgor of cells, GA, and enzymes that are able to loosen the cell wall‐like expansins (McQueen‐Mason et al., 1992; Cosgrove, 1999), and more recently yieldins (Okamoto‐Nakazato et al., 2000) and an endo‐β‐1,4‐glucanase (Xu et al., 2000). The acid growth theory (Rayle and Cleland, 1992; Kutschera, 1994) states that within 10 min after auxin addition to cells, the walls are acidified and cell wall loosening and the onset of elongation occurs. Acid‐induced cell wall elongation involving expansins has been observed in submerged deep water rice, but whether this involves auxin as well is not clear (Cho and Kende, 1997). As mentioned before expansins may play a role during the stimulated elongation process in submerged Rumex palustris. Vriezen et al. isolated six distinct expansin genes (one cDNA and five PCR fragments from genomic DNA) from Rumex palustris (RpEXP1‐6, accession numbers AF167360, AF167361, AF167362, AF167363, AF167364, and AF167356, respectively) and five from Rumex acetosa; RT‐PCR generated fragments of RaEXP1 and RaEXP2, and PCR of genomic DNA generated fragments of RaEXP3‐5 (Vriezen et al., 2000, accession numbers for RaEXP1‐4, AF167365, AF167357, AF167358, and AF167359, respectively). The number of RpEXP clones was recently increased to 18 by isolating nine cDNAs (RpEXP7‐14 and RpEXP18) and three RT‐PCR generated clones (RpEXP15‐17) (N Wagemaker, WH Vriezen, TD Colmer, LACJ Voesenek, AJM Peeters, unpublished results). Studies examining the transcription dynamics of RpEXP1 have shown that the expression of this gene is up‐regulated in the petioles of Rumex palustris upon submergence or treatment with 5 μl l⁻¹ ethylene (Vriezen et al., 2000). The observed increase of this messenger, although accumulation of other members of the same size of this superfamily cannot be excluded, coincides with the increased elongation rate of the petioles during submergence. However, the same expression profile was observed in older petioles that did not show the enhanced elongation response (Vriezen et al., 2000). This may be attributed to a decreased responsiveness to this protein in more mature tissue (reviewed by Kende et al., 1998). Expression of RaEXP1 in Rumex acetosa under submerged conditions showed a constitutive pattern that was similar under aerated and ethylene (5 μl l⁻¹) conditions (Vriezen et al., 2000).

Role of oxygen (O₂) and carbon dioxide (CO₂)

One of the first changes that take place upon submergence are the changes in concentrations of O₂ and CO₂. Their internal concentrations are strongly influenced by photosynthesis and respiration. If submergence is prolonged over a period of days and nights the concentration of these gases will exhibit a strong diurnal pattern, with low oxygen and high carbon dioxide during the night and the reverse during the day (Stünzi and Kende, 1989), which makes them unreliable indicators for submergence (Voesenek and Blom, 1999). Although low oxygen (as low as 3 kPa) gives rise to an enhanced petiole elongation (Voesenek et al., 1997), low oxygen is not considered to be the key signal for the plant that its environment has changed from air to water, for the reason mentioned earlier. In situ measurements of O₂ concentrations in the petiole show that, under submerged conditions in the light, the levels of oxygen decline within 24 h to 10 kPa (Rijnders et al., 2000). Much lower levels (3 kPa), were measured during night‐time (Rijnders et al., 2000). Low oxygen‐induced petiole elongation was found to be ethylene‐dependent and the low level of oxygen is thought to sensitize petiole tissue in an unknown way to ethylene (Voesenek et al., 1997). High levels of CO₂ do not have an effect on petiole elongation in Rumex palustris (Voesenek et al., 1997), although it has been observed in rice that elevated levels of CO₂ (6%) could induce internodal elongation, but to a far lesser extent than ethylene (Raskin and Kende, 1984a).

Role of ethylene

Ethylene is very simple gaseous plant hormone. Over the past two decades the biosynthesis route (Kende, 1993), and thanks to mutant analysis in Arabidopsis thaliana, perception at the cellular level (Hua and Meyerowitz, 1998), and a large part of the signal transduction pathway have been elucidated (reviewed in Stepanova and Ecker, 2000). Ethylene is involved in many biological processes, like fruit ripening, flower and leaf abscission, senescence, many stress acclimations, and growth. Ethylene is generally regarded as a negative regulator of growth (Abeles et al., 1992).

In the 1970s, however, it was discovered that ethylene has a growth‐promoting activity on, respectively, rice coleoptiles and Callitriche platycarpa internodes (Ku et al., 1970; Musgrave et al., 1972). This growth‐promoting action of ethylene was subsequently found in a number of other semi‐aquatic species (Ridge, 1987). Ethylene was recognized to be one of the key hormones in flooding‐induced acclimations in deepwater rice (Kende et al., 1998).

Within 1 h after submergence of Rumex palustris a 20‐fold increase in the ethylene concentration, from 0.05 μl l⁻¹ to 1 μl l⁻¹ was observed (Voesenek et al., 1993; Banga et al., 1996), reaching its maximum value of about 8 μl l⁻¹ after 16 h (Banga et al., 1996). This increase can be explained by physical entrapment of the continuously produced gas since gases diffuse 10000 times slower through water than through air, provided that ethylene biosynthesis proceeds (Voesenek and Blom, 1999). Ethylene biosynthesis is regulated by two enzymes, ACC synthase and ACC oxidase (Kende, 1993; Fluhr and Mattoo, 1996). To study the transcription dynamics of the respective gene families in flooding‐tolerant and non‐tolerant species, one gene encoding each enzyme (for ACC oxidase from Rumex palustris two very homologous genes) have been cloned from both Rumex palustris (Vriezen et al., 1999; Vriezen, 2000) and from the flooding‐sensitive species Rumex acetosa (Vriezen, 2000). Upon submergence, the ACC concentration was strongly increased in both flooding‐tolerant and non‐tolerant Rumex species (in this case Rumex acetosella), whereas the conversion of ACC to ethylene was inhibited (Banga et al., 1996). Vriezen observed no increased accumulation of the transcripts of a RpACC synthase gene (RpACS1, accession number AF038945) upon submergence within this period (Vriezen, 2000). This discrepancy might be explained by the presence of additional, but differently regulated ACC‐synthase genes in the Rumex palustris genome, and/or by post‐transcriptional and post‐translational regulation of ACC synthase activity which has been observed in other species (Spanu et al., 1994; Oetiker et al., 1997; Vogel et al., 1998; Woeste et al., 1999). On the other hand, the activity of ACC oxidase might be limited due to hypoxic conditions during submergence, although it was found that the rate of ethylene production did not change with low oxygen conditions (Vriezen et al., 1999). The genes coding for RpACC oxidase (RpACO1 and RpACO2, accession numbers Y10034 and AF041479, respectively) were strongly induced upon submergence (Vriezen et al., 1999), especially in the petiole which undergoes increased elongation upon submergence, the level increased within 6 h. This correlates with an increase in enzyme activity in vitro (Vriezen et al., 1999), although the ethylene production rate did not alter during this period (Voesenek et al., 1993; Vriezen et al., 1997). The amount of RpACO transcript and enzyme increased when plants were subjected to 5 μl l⁻¹ ethylene (increase within 2 h) or 3% oxygen (increase within 6 h) (Vriezen et al., 1999). Positive regulation at the transcriptional level of ACC oxidase by ethylene is a well‐described phenomenon (Drory et al., 1993; Nadeau et al., 1993; Kim and Yang, 1994; Tang et al., 1994; Peck and Kende, 1995; Barry et al., 1996; Mekhedov and Kende, 1996). Low oxygen severely hampers the ACC oxidase activity (Imaseki, 1991) but by increasing the amount of ACC‐oxidase the ethylene production may be secured. The same results were obtained when the submergence non‐tolerant species Rumex acetosa was examined using a RaACC oxidase gene (RaO2; Vriezen, 2000). There, the same mechanism also seems to be present although it has no apparent function, since ethylene production does not have an effect on submergence‐induced petiole elongation in this species.

At the level of perception of ethylene two types of membrane receptor genes have been cloned in Arabidopsis thaliana (ETR and ERS; reviewed by Stepanova and Ecker, 2000). The ETR1 genes of Arabidopsis (Hua and Meyerowitz, 1998) and tomato (Tieman et al., 2000) are expressed constitutively (Chang et al., 1993; Zhou et al., 1996), whereas the ERS‐like genes are inducible (Wilkinson et al., 1995). Moreover, it was shown that the ETR1 and ERS1 genes in muskmelon showed marked changes in expression during ontogeny (Hall et al., 2001). By making use of the ETR1 gene of Arabidopsis, an ERS orthologue was cloned in Rumex palustris (Rp‐ERS1, accession number U63291; Vriezen et al., 1997). It was shown that during submergence the Rp‐ERS1 transcript accumulated, as is the case when plants are subjected to ethylene (5 μl l⁻¹) or low oxygen (3%). Whether the observed increase in transcripts leads to an increase in receptor concentration is unknown. Due to the negative regulatory nature of this receptor gene family it is difficult to speculate what the increase of the message of Rp‐ERS1 really means, possibly it will result in a decrease of the tissue sensitivity to ethylene (Hua and Meyerowitz, 1998; Tieman et al., 2000).

Role of auxins

Auxins are generally considered as growth promoting substances (Taiz and Zeiger, 1998). It is therefore obvious to suspect a role for these plant hormones in submergence‐induced petiole elongation. The putative role of auxins was studied during petiole elongation in Rumex palustris (Rijnders et al., 1996). Treatment with an auxin transport inhibitor (NPA) or removal of the leaf blade, which is considered a putative source of auxin (Horton and Samarakoon, 1982), did not alter the response in Rumex palustris, whereas a similar petiole response in Ranunculus sceleratus was severely inhibited (Rijnders et al., 1996). It has been suggested that auxin in the petioles of the latter species allows the tissue to respond to changes in ethylene levels (Horton and Samarakoon, 1982). Smulders and Horton, however, were able to separate the effect of ethylene, involved in cell elongation, and auxin, involved in radial expansion, during submergence of Ranunculus sceleratus petioles (Smulders and Horton, 1991). It was reported that during internodal elongation in rice no effect of auxin could be detected (Azuma et al., 1990). Given the information described above the role of auxin therefore remains obscure and, further, more precise methods, like the use of auxin‐responsive promoters elements fused to reporter genes (Sabatini et al., 2000), should be exploited to elucidate the role of auxin in submergence‐induced shoot elongation.

Role of abscisic acid (ABA)

ABA is a growth‐inhibiting plant hormone that is involved in many processes during plant life (Taiz and Zeiger, 1998). It is generally accepted that ABA and GAs act as antagonists in several growth processes, for example, seed germination (Koornneef et al., 1982; Gómez‐Cadenas et al., 2001). It was shown that high levels of exogenously applied ABA inhibited growth in petioles and stems in Ranunculus sceleratus (Smulders and Horton, 1991) and in stems in Potamogeton pectinatus (Summers and Jackson, 1996), respectively. It was found earlier in rice (Hoffmann‐Benning and Kende, 1992) that levels of ABA in elongating internodes decrease during submergence (or ethylene treatment) while the level of GA₁ increased 4‐fold suggesting that growth is determined by the ratio of ABA and GA. An ABA decrease using the same system has been reported (Azuma et al., 1995), but their results also suggest that, although ethylene may modulate the ABA concentration, ethylene decreasing ABA concentrations alone is not always enough to induce elongation. They observed no induced elongation at low relative humidity, indicating the involvement of other factors besides ethylene and ABA (Azuma et al., 1995). van der Straeten et al. showed, while comparing lowland and deepwater rice seedlings, that ABA levels decreased to the same extent in both cultivars upon submergence (van der Straeten et al., 2001). Whether this holds true for Rumex palustris is not known, but preliminary results using this species indicate that the ABA concentration in petioles decreases rapidly during the first h of submergence preceding the start of petiole elongation upon submergence (J Benschop and LACJ Voesenek, unpublished results). Moreover, external application of high concentrations of ABA inhibits the submergence‐induced petiole elongation (J Benschop and LACJ Voesenek, unpublished results), supporting an important role for ABA during petiole elongation in Rumex palustris upon submergence. To investigate the role of ABA at the molecular level further, a Rumex palustris orthologue of the 9‐cis‐epoxycarotenoid dioxygenase gene (cleavage enzyme) has recently been isolated (J Benschop, LACJ Voesenek, AJM Peeters, unpublished results), which is considered to catalyse the rate‐limiting step of ABA biosynthesis (Qin and Zeevaart, 1999).

Role of gibberellic acid (GA)

Studies of the involvement of GAs during elongation responses upon submergence have shown that this plant hormone plays an important role in the responses of Rumex palustris, deepwater rice and Ranunculus sceleratus. Application of GA biosynthesis inhibitors like paclobutrazol (Lever et al., 1982; Rademacher, 2000) decreases the submergence‐stimulated petiole elongation of Rumex palustris (Rijnders et al., 1997) similar to the response in Ranunculus sceleratus (Samarakoon and Horton, 1983), but not completely abolishes it as was found for deepwater rice (Raskin and Kende, 1984a). This indicated that GAs are probably not the sole factor inducing the growth response, since paclobutrazol is a very potent and specific GA biosynthesis inhibitor (Rademacher, 2000). The decrease of the response could be counteracted by application of GA₃ or GA₄. Upon submergence of Rumex palustris the endogenous levels of GAs rise (Rijnders et al., 1997) similar to the observation made in deepwater rice upon submergence (Suge, 1985; Hoffman‐Benning and Kende, 1992). It was shown that the increase in GAs was triggered by ethylene (Rijnders et al., 1997) as was suggested earlier (Suge, 1985; Hoffman‐Benning and Kende, 1992). Furthermore, it was shown in Rumex palustris that ethylene increases the sensitivity for GAs in petioles (Rijnders et al., 1997). These results were similar to those obtained for deepwater rice (Raskin and Kende, 1984b).

To investigate the role of GAs at the molecular level further, Rumex palustris orthologues of two distinct GA20‐ oxidases, a 2‐oxidase and a 3‐oxidase, have recently been isolated (J Bou, LACJ Voesenek, AJM Peeters, unpublished results).

Role of other factors

So far, the role of obvious and well‐studied factors involved in the submergence‐induced petiole elongation in Rumex palustris and other plant species have been discussed. There are some hormone or hormone‐like substances that also may contribute to this response. Polyamines are hormone‐like substances that were shown previously (Chang et al., 1999) to be involved in the regulation of ethylene‐ or auxin‐induced petiole growth upon submergence of Ranunculus sceleratus.

Recently, the sax1 mutant in Arabidopsis thaliana was characterized (Ephritikhine et al., 1999). This mutant exhibits a severe dwarf phenotype and altered responses to GAs, ABA, auxin, and ethylene, and could be restored to wild type by application of brassinolides. Together with the findings of others (Clouse, 1996; Kauschmann et al., 1996), who described the Arabidopsis dwarf mutants bri1 and cbb1–cbb3, respectively, it is suggested that brassinosteroids may play a role during growth. The cbb genes all show an altered expression of genes involved in cell wall synthesis; XETs (xyloglucan endotransglycosylases). These data suggest that brassinosteroids may play a role during the submergence response of Rumex palustris.

Conclusions and future research

The picture emerging from this review is that the response to submergence of Rumex palustris is multifactorial and involves at least four plant hormones and several other factors as is shown in the scheme in Fig. 1. It is intriguing that the plant hormone ethylene has contrasting effects; depending on the plant species and even organ type it stimulates or inhibits growth. The signal that the plant is submerged is being brought about by ethylene. Subsequently, most likely auxin, abscisic acid and gibberellic acid play a role, although the presence of ethylene is probably a prerequisite at all stages of the response (Voesenek and Blom, 1999). Two basic processess can be distinguished after the perception of the stress; hyponastic response and the accompanied petiole elongation.

One of the major unresolved challenges that remain is how all the factors that have been described in this review interact with each other. The described signal transduction pathways are far from complete and it will be a tremendous, but challenging, task to tie up all these loose ends in the near future. It is therefore not only important to study a plant species that shows this remarkable response, but it is evident that resolving the question why other members of the same genus (i.e. Rumex acetosa) and other species do not exhibit, or just show parts of this response can help to resolve the questions that remain.

Rumex sp. is a model system that can easily be studied at the physiological and the biochemical level. However, studies at the genetic and molecular levels are prevented by difficulties that lie within the species itself. Rumex palustris is probably tetraploid (Tischler, 1950), has a big genome and is very recalcitrant with respect to transformation using Agrobacterium tumefaciens (N Wagemaker, LACJ Voesenek, AJM Peeters, unpublished results). Molecular biology and genetics have had an enormous impact on plant physiological research over the last two decades (Koornneef et al., 1997). Using Arabidopsis thaliana in plant physiological research has many advantages, the easy generation of mutants, the recent completion of the genome sequence (The Arabidopsis Genome Initiative, 2000) and modern genetic techniques like QTL mapping, that makes use of natural variation to dissect polygenic traits, makes it an attractive model species. Furthermore, the availability of whole genome DNA micro‐arrays and the re‐birth of 2D‐PAGE, due to far better and more sensitive methods, and subsequent protein identification using MALDI‐TOF (the so‐called functional genomics and proteomics approach) will greatly enhance the knowledge of signal transduction cascades. Besides the ongoing research of Rumex palustris, studies of parts of the responses that Rumex palustris exhibits upon submergence have begun (and which are mimicked by ethylene addition) in Arabidopsis thaliana. Analysis of existing and newly isolated mutants (induced genetic variation) as well as natural genetic variation will be included in this research. Preliminary results show that there is considerable variation among Arabidopsis accessions for the hyponastic response (data not shown), and sensitivity towards ethylene (Fig. 2; W Huibers, LACJ Voesenek, unpublished results). Furthermore, tolerance to submergence in the dark among Arabidopsis accessions has been studied and showed that Kas‐1 and Ws are remarkably resistant whereas Shah is very sensitive to this treatment (D Tholen, LACJ Voesenek, unpublished results).

Moreover, a start has been made to involve modern genomic and proteomic techniques in Arabidopsis and Rumex palustris to unravel parts of the submergence signal transduction cascade with the ultimate goal of elucidating the mechanism of submergence‐induced petiole elongation in Rumex palustris.

Fig. 1.

Model of the signal transduction cascade of various components involved in submergence‐induced petiole elongation. (—) Proven interactions; (‐‐‐) speculative interactions; (→) positive/stimulating effect; (—|) negative/inhibitory effect.

Fig. 2.

Effect of ACC (20 μM) on hypocotyl growth of several Arabidopsis thaliana accessions and mutants.

1

To whom correspondence should be addressed. Fax: [phone]. E‐mail: [email protected]

The authors would like to thank Wim Huibers, Danny Tholen and Niels Wagemaker for excellent technical assistance. Furthermore, the authors greatly acknowledge the Dutch Science Foundation (NWO) for the PIONIER grant (800.84.470) to LACJV. JB is supported by an EU Training Network grant (HPRN‐CT‐2000‐00090). This review was written on the occasion of the Journal of Experimental Botany lecture held by LACJV at ComBio99, Gold Coast, Australia.