Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Evolution and Geographic Distribution of Desiccation-Tolerant Plants

1.1 A Window into Past Research of Desiccation Tolerance

Three centuries ago science of desiccation tolerance began with a lengthy period of discovery and doubt. In 1743, Henry Baker announced to the Royal Society in London that some animals could tolerate desiccation, which means that they could dry to equilibrium with air and resume normal functions upon rehydration (Keilin 1959). Henry Baker’s example of desiccation tolerance was the larva of the nematode Anguillulina tritici. The discovery that a nematode could lose virtually all its free internal water without dying was remarkable, because most animals and plants die instantly, if their cells equilibrate with even moderately dry air, because water maintains the structure of intracellular macromolecules and membranes. The next step was to identify more organisms that tolerate desiccation. Subsequently, it was also observed in rotifers by van Leeuwenhoek in 1702 (Keilin 1959). Eventually desiccation tolerance was reported in four other phyla of animals, in some algae, fungi and bacteria, in ferns, in most bryophytes, lichens, seeds of flowering plants, and in about 300 unique angiosperm species termed resurrection plants (Alpert 2006). The vegetative tissues of resurrection plants are able to survive protoplastic dehydration of less than 2% relative water content. The majority of flowering plants and also gymnosperms have desiccation-tolerant seeds or pollen. This trait is strictly developmentally regulated, it is acquired during embryogenesis (Bartels et al. 1988) or pollen development, but lost during germination.

Among plants, desiccation tolerance is common in primitive plants such as lichens and bryophytes. This form of anhydrobiosis seems to comprise constitutively expressed cell protection mechanisms associated with inducible repair systems that are activated after rehydration (Oliver and Bewley 1997 and covered in detail in Chap. 4). Also among algae, it is possible to find species able to tolerate desiccation, either in terrestrial or in marine intertidal algae (Trainor and Gladych 1995; Abe et al. 2001).

The taxonomic representation of desiccation tolerance in plants is now fairly well established. Although desiccation tolerance in plants is very uncommon, it occurs in many different taxa except gymnosperms (Alpert 2000). Desiccation-tolerant species are found on all continents and among species of all growth forms except trees, but it remains a mystery why desiccation tolerance is not more widespread.

There have been only few systematic surveys for desiccation tolerance within taxa or habitats. The list of tolerant vascular plants from different regions published by Gaff and co-workers (Gaff 1977, 1986; Gaff and Latz 1978), from rock outcrops (Porembski and Barthlott 2000) and reports by Fischer (1992, 1995) (Table 16.1) are probably the closest approaches to survey higher plants (see also Tables 8.1 and 9.1). The overall pattern one sees is that taxonomic and geographic breadth is contrasted with ecological narrowness. In the 1960s, researchers started to investigate the physiological ecology of desiccation tolerance in plants, and since the 1980s, emphasis has shifted to the biochemistry and molecular biology of desiccation tolerance.

Table 16.1 Desiccation-tolerant species in addition to those listed by Proctor and Pence (2002) (see also Tables 8.1 and 9.1)
Full size table

1.2 Evolution of Desiccation Tolerance

Many primitive autotrophs, such as cyanobacteria, are desiccation tolerant. This indicates that desiccation tolerance is a very early trait in evolution. According to Oliver et al. (2000), phylogenetic evidence exists that among land plants tolerance to vegetative desiccation was present in the basal clade of bryophytes. This adaptive trait was likely to be critical for colonization by primitive terrestrial forms of the relatively dry land environment (Kappen and Valladares 1999). With the evolution of more complex vascular plants, desiccation tolerance was lost in vegetative tissues but was retained in reproductive tissues (Oliver et al. 2000). The acquisition of desiccation tolerance is part of a maturation programme during seed development in most higher plants (Angelovici et al. 2010). The majority of terrestrial plants are capable of producing desiccation-tolerant structures such as seeds and pollen, which can remain viable in the desiccated state for long periods as demonstrated in the case of the ancient Nelumbo nucifera (sacred lotus) seed from China (Shen-Miller et al. 1995). It is hypothesized that desiccation tolerance was crucial for ancestral freshwater autotrophs to live on land. Independent evolution or re-evolution of desiccation tolerance has happened in the Selaginellales, leptosporangiate ferns, and in at least ten families of angiosperms. Evolutionary progress may be evident from the fact that pteridophytes, like bryophytes, are able to synthesize rehydrin proteins but can also synthesize dehydrin proteins, which are typical for angiosperms (Oliver et al. 2000).

The genes involved in desiccation tolerance were not lost, but were instead recruited for activation of drying responses in reproductive structures through inductive mechanisms during developmental programs of the plant. It seems that resurrection plants have re-directed gene expression of seed-specific genes to be expressed in vegetative tissues (Illing et al. 2005). Recent phylogenetic evidence suggests that vascular plants gained the ability to withstand desiccation of their vegetative tissues from a mechanism present first in spore-bearing plants, and that this evolution (or re-evolution) has occurred on at least ten independent occasions within the angiosperms (Oliver et al. 2005). The presence of this information in the genomes of vascular plants permitted many species to spread over regions with seasonal rain fall and to re-evolve desiccation tolerance in vegetative tissues based on inducible responses triggered by environmental cues (Rascio and La Rocca 2005). The loss of desiccation tolerance in vegetative tissues can be interpreted as the result of suppression of genes in vegetative tissues (Dickie and Pritchard 2002). There may have been a selection against desiccation tolerance in vegetative tissues in favour of other traits. It can be concluded that the ability to survive dehydration must be basic and ubiquitous. On the other hand, desiccation tolerance must have been acquired at various advanced stages of organismic complexity by evolving metabolic or structural protective mechanisms. These must be very specific, either because of a distinct combination of quantitatively selected ubiquitous “house-keeping” mechanisms (Illing et al. 2005) or because of inventing genetic structures that act species-specific, like the CDT-1 gene in Craterostigma plantagineum (Furini et al. 1997; Bartels and Salamini 2001).

1.3 Geographic Distribution and Ecology

Resurrection plants are found in places where rainfall is seasonal and sporadic. They often grow on rocky outcrops at low to moderate elevations in tropical and subtropical climates (Porembski and Barthlott 2000). Under these conditions, they are subjected to frequent cycles of drying and rehydration throughout the year and thus tolerate being dry in a broad range of temperatures (Mundree et al. 2002; Moore et al. 2005, 2007). Most resurrection plants have been reported from the southern hemisphere of Africa, India, Australia, and South America.

Resurrection plants have been identified within the angiosperms among both monocotyledonous and dicotyledonous plants. The dicotyledonous plants are represented mainly in the Linderniaceae, Scrophulariaceae, and Myrothamnaceae families, whereas the monocotyledonous are more scattered among different families. A first phylogenetic analysis among the Scrophulariaceae suggests a clustering of desiccation-tolerant plants represented by the genera Craterostigma and Lindernia (Rahmanzadeh et al. 2005).

It has been suggested that desiccation tolerance is connected with a size limitation, since all examples of desiccation-tolerant flowering plants do not exceed a certain height (Bewley and Krochko 1982), the largest known resurrection plant is the small woody shrub Myrothamnus flabellifolius, which can be between 0.5 m and 1.5 m tall (Moore et al. 2007). Most resurrection plants are herbaceous plants.

A list of desiccation-tolerant tracheophytes was compiled by Proctor and Pence (2002) and by Fischer (1992, 1995) (Table 16.1). It shows that there are more than 300 species of vascular plants. This number indicates that less than 0.15% of the total species possess vegetative desiccation tolerance.

The majority of resurrection plants were originally described in the 1970s (Gaff 1971; Gaff and Ellis 1974; Gaff and Churchill 1976; Gaff and Latz 1978). Most of the desiccation-tolerant angiosperms such as M. flabellifolia (Child 1960), Xerophyta spp., or Craterostigma spp. (Gaff 1977) are native to Southern Africa (especially South Africa, Namibia, and Zimbabwe) or to Australia, as it is the case for Borya nitida (Gaff and Churchill 1976). Physiological and anatomical characterizations of resurrection plants have been well investigated in species such as B. nitida (Gaff et al. 1976; Hetherington et al. 1982), Craterostigma wilmsii (Sherwin and Farrant 1996, 1998; Vicré et al. 1999; Cooper and Farrant 2002), Eragrostis nindensis (Vander Willigen et al. 2001, 2004), M. flabellifolius (Sherwin and Farrant 1996; Farrant and Kruger 2001), or Reaumuria soongorica (Liu et al. 2007) and most of the molecular aspects of desiccation tolerance have been studied in C. plantagineum (Bartels et al. 1990; Bartels and Salamini 2001; Hilbricht et al. 2002; Phillips et al. 2002), Xerophyta viscosa (Mundree et al. 2000; Mowla et al. 2002; Garwe et al. 2003) and Xerophyta humilis (Collett et al. 2003) and Sporobolus (Neale et al. 2000). A list of the most widely studied resurrection plants is compiled in Table 16.2.

Table 16.2 Best-studied resurrection plants
Full size table

1.4 Diversity Within Linderniaceae

Phylogenetic analysis showed that desiccation-tolerant Craterostigma species cluster with Lindernia species, and the close phylogenetic relationship is supported by microsynteny between desiccation-related genes (Phillips et al. 2008). Interestingly, some members of the Linderniaceae are desiccation tolerant and some desiccation sensitive. This observation is being utilized to understand the evolution of desiccation tolerance. Lindernia brevidens, a close relative of C. plantagineum, was surprisingly identified as desiccation tolerant. This is remarkable, as L. brevidens is endemic to montane rainforests of East Africa, where it never experiences dry periods. L. brevidens occurs exclusively in two areas of the ancient Eastern Arc Mountains, which were protected from Pleistocene droughts by the mild temperatures of the Indian Ocean. High levels of species endemism and species diversity have developed in this area because of its protection and isolation. Although L. brevidens evolved in rainforest areas, it did not lose desiccation tolerance. Molecular analysis of desiccation-related genes suggests a high level of conservation between L. brevidens and C. plantagineum including the octulose sucrose conversion during desiccation (Phillips et al. 2008). It appears that L. brevidens is a neoendemic species, which has retained desiccation tolerance through genome stability. The question why L. brevidens did not lose desiccation tolerance cannot be answered; one hypothesis is that desiccation tolerance is linked to another trait, which is advantageous for L brevidens in its ecological niche. Comparative analysis between the closely related species with contrasting desiccation tolerance phenotypes could be a useful approach to understand the acquisition of desiccation tolerance.

2 Cellular Aspects

2.1 Morphological Adaptations

Leaf folding is one of the most obvious morphological changes in desiccation-tolerant vascular plants (Gaff 1989; Scott 2000; Farrant et al. 2003; Vander Willigen et al. 2003; Farrant et al. 2007). Leaves of C. wilmsii or C. plantagineum, which are fully expanded when watered, progressively curl inward during drying and become tightly folded so that only the abaxial surfaces of the older leaves in the outer whorl are exposed to the sun (Fig. 16.1 and Sherwin and Farrant 1998). Leaf folding is thought to limit oxidative stress damage from UV radiation and is thus an important morphological adaptation for surviving desiccation. Indeed, C. wilmsii plants do not survive desiccation in sunlight, if the leaves are mechanically prevented from folding (Farrant et al. 2003).

Fig. 16.1
figure 1_16

Presentation of a dehydration and rehydration cycle in the resurrection plant Craterostigma plantagineum. A single plant was followed throughout the dehydration–rehydration cycle. This plant dehydrates to 2–5% RWC within 15 days and rapidly rehydrates within 24 h and starts photosynthesis. It has been estimated that leaves of this plant shrink to around 15% of the original leaf area. The plant is able to flower, as shown in the photograph in the middle; the photograph was taken 15 days after rehydration

Full size image

The leaf blades of X. humilis fold in half along the midrib upon dehydration, leaving only the abaxial surface exposed to the light (Sherwin and Farrant 1998). In Sporobulus stapfianus, the leaf adaxial side, which is most exposed to sun radiation, is very rich in epicuticular waxes, whose function is, besides decreasing transpiration, to reflect light and to limit irradiation and temperature increase (Dalla Vecchia et al. 1998). Leaf movements occurring during dehydration have been suggested to reduce the effective transpiring surface during early stage dehydration and/or to prevent excessive irradiation of air-dry younger tissue (Gaff 1989; Farrant 2000).

2.2 Mechanical Stress: Cell Wall Changes, Vacuole Fragmentation and Water Substitution

One problem that resurrection plants face is the reduction of cell volume that occurs during desiccation. The examples below show that resurrection plants developed different mechanisms to avoid the potential mechanical stress during dehydration. A loss of water from cells leads to shrinkage of the central vacuole and to a drawing inward of the cell contents. This causes tension between the plasmalemma and the cell wall, which generally exhibits limited elasticity. If the cell membrane detaches from the cell wall, plasmodesmatal connections are ruptured, which can be lethal (Cosgrove 2000). Adaptations have acquired to reduce mechanical stress. In some resurrection plants, mainly dicotyledons, the protoplast shrinkage occurring upon dehydration is accompanied by an extensive folding of cell walls, which results in a contraction of the entire cell and avoids the tearing of the plasmalemma from the cell wall (Farrant and Sherwin 1997; Thomson and Platt 1997; Farrant 2000; Vicré et al. 1999, 2004). This phenomenon prevents the development of negative turgor pressure and reduces the potential for irreversible mechanical damage (Vander Willigen et al. 2001).

Recently, a relatively high degree of cell wall flexibility was shown for Craterostigma and M. flabellifolius. This feature can be attributed to the specific composition of leaf cell walls. In C. wilmsii, a significant increase in xyloglucans and unesterified pectins is observed in the cell wall during drying (Vicré et al. 1999). Dehydration also induces a considerable reduction of glucose in the hemicellulose fraction of C. wilmsii cell walls (Vicré et al. 2004). These changes are thought to enhance the tensile strength of the cell wall, allowing it to contract and to fold without collapsing in the dried tissue. An increase in expansin activity during desiccation is associated with a rise in cell wall flexibility and folding. This increase in cell wall flexibility is correlated with an increase in expansin transcript levels and activity (Jones and McQueen-Mason 2004; Moore et al. 2008). Expansins are cell wall loosening factors and are thought to act by disrupting the hydrogen bonds between cellulose and hemicellulose polymers in the cell wall (McQueen-Mason and Cosgrove 1995). Thus, the ability of cell walls to fold by changing their chemistry and texture is an adaptive strategy of resurrection plants, but the molecular basis of this alteration is hardly known. Recently, a glycine-rich protein has been correlated with cell wall flexibility in Boea hygrometrica (Wang et al. 2009). These exceptional features are remarkable when compared to the rigidification of the cell wall in response to drought in non-resurrection plants (Lu and Neumann 1998; Munns et al. 2000).

Some species, especially monocotyledons, utilize a strategy of water substitution and try to maintain the original cell volume by replacing water with non-aqueous substances such as amino acids, small proteins, and sugars (Farrant 2000; Marais et al. 2004). At the same time, the central vacuole is fragmented into a number of smaller units (Quartacci et al. 1997; Dalla Vecchia et al. 1998; Vander Willigen et al. 2004). This is of advantage during desiccation, as the formation of several small vacuoles increases the surface to volume ratio and thus presents a better mechanical stability (Iljin 1957).

2.3 Membrane Fluidity

Membrane fluidity may be another parameter that contributes to cell and tissue flexibility during desiccation. A higher degree in polyunsaturation in membrane phospholipids results in better membrane fluidity. This was demonstrated by changing lipid unsaturation in chloroplasts (Moon et al. 1995). Therefore, membrane composition was analysed in some resurrection plants, but the general picture is still patchy and no general conclusions can be drawn.

Hoekstra (2005) reports a negative correlation between the longevity of desiccation-tolerant tissues and the number of double bonds in the polar lipids of membranes. In desiccation-tolerant vascular plants, dehydration generally causes a decrease in total lipids as well the unsaturation level of individual phospholipids (Quartacci et al. 2002). However, the opposite trend is observed in the resurrection plant Boea hygroscopica where increased unsaturation of fatty acids was observed in all lipid classes upon dehydration (Navari-Izzo et al. 1995). In S. stapfianus phospholipid content and the level of polyunsaturated lipids within the plasma membrane increased in desiccation-tolerant, dried, attached leaves, but decreased in non-desiccation-tolerant detached leaves during desiccation (Quartacci et al. 1997; Neale et al. 2000).

3 Physiology

3.1 Photosynthesis

During desiccation, angiosperm resurrection plants shut down photosynthesis. This is correlated with down-regulation of photosynthesis-related gene expression (Bockel et al. 1998; Bernacchia et al. 1996) and with degradation of photosynthetic structures. Two different strategies are distinguished according to which plants are classified as homoiochlorophyllous and poikilochlorophyllous species (see Table 16.2). Homoiochlorophyllous species retain the chlorophyll and thylakoid membranes, although changes in photosynthetic pigment distribution are observed (Alamillo and Bartels 2001). In homoiochlorophyllous plants, chloroplasts do, however, undergo changes during desiccation, becoming roundish, with altered inner membrane organization and stacking. Also changes in the ratios between lipids and proteins as well as between the different lipid classes can occur in thylakoid membranes (Thomson and Platt 1997; Farrant 2000; Navari-Izzo et al. 2000; Quartacci et al. 1997).

In poikilochlorophyllous species, chlorophyll and the photosystem complexes are broken down (Tuba et al. 1998). The degradation of chlorophyll is advantageous, because the plants avoid the accumulation of toxic reactive oxygen species (ROS). Photosynthetic capacities are recovered during the rehydration process. Homoiochlorophyllous plants resume photosynthesis faster than poikilochlorophyllous species, which have to synthesize all components de novo. Ingle et al. (2008) demonstrated that the chloroplast biogenesis during rehydration in X. humilis resembles etioplast–chloroplast transition and uses similar molecular mechanisms involved in photomorphogenesis. The authors conclude from their observations that chloroplast degradation and regeneration during desiccation and rehydration do not involve novel genes, but require altered gene regulation of genes existing probably in most higher plants.

As the photosynthetic structures in the homoiochlorophyllous species need to be protected, it is not surprising that different classes of desiccation-induced proteins seem to be involved in the maintenance of chloroplast stability (Schneider et al. 1993; Alamillo and Bartels 1996; Neale et al. 2000). One of these is a chloroplast-localized, desiccation-related protein found in the leaves of C. plantagineum (Bartels et al. 1992; Alamillo and Bartels 2001) and S. stapfianus (Neale et al. 2000) shows high similarities to early light-inducible proteins (ELIPs). The ELIPs are thylakoid chlorophyll binding proteins that are transiently expressed during greening of etiolated plants (Adamska 1997). However, they are also synthesized in mature leaves exposed to excess of light or other environmental stresses causing the enhancement of free radicals (Montané et al. 1997; Zeng et al. 2002; Provart et al. 2003; Savenstrand et al. 2004). These proteins, similar to the HLIPs (high light-induced proteins) of cyanobacteria and the LHC (light harvesting complex) proteins of photosystems, have a protective function against photo-oxidative damage of the photosynthetic apparatus (Hutin et al. 2003). ELIPs might also bind chlorophylls, thus keeping free pigments at low levels in conditions of high light stress (Hutin et al. 2003). Earlier reports suggested that ELIPs might bind zeaxanthin, taking part in the non-photochemical quenching of light energy (Król et al. 1999).

A novel nuclear gene family expressed in response to dehydration has been identified in the leaves of C. plantagineum (Phillips et al. 2002). The genes encode plastid-targeted proteins (CpPTP), which have the ability to interact with plastid DNA, suggesting a role in the down-regulation of chloroplast-encoded genes or a role in protecting DNA during dehydration. In addition, several classes of Lea (=late embryogenesis abundant)-like proteins and detoxifying enzymes such as aldehyde dehydrogenase were shown to accumulate in chloroplasts of C. plantagineum during dehydration (Schneider et al. 1993; Kirch et al. 2001).

3.2 Antioxidant Systems

Recovery of a resurrection plant correlates with its capacity to establish a number of antioxidant protective mechanisms during dehydration and to maintain these systems upon rehydration (Kranner et al. 2002; Kranner and Birtić 2005). The level of reactive oxgen species (ROS) undergoes a significant rise in water-stressed cells that is dictated largely by chloroplasts as they are the most aerobic compartment in plant cells (Moran et al. 1994; Kranner and Lutzoni 1999). As respiration declines early during dehydration, mitochondria may be less involved in ROS production. In conditions of water deficit, stomata close limiting the supply of CO2 to chloroplasts. This inhibits carbon fixation and causes overexcitation of chlorophylls, which then transfer their energy to oxygen, giving rise to oxygen singlets. M. flabellifolius retains high concentrations of chlorophyll during desiccation. When desiccated M. flabellifolius was rehydrated, antioxidants such as ascorbate, glutathione, and α-tocopherol accumulated in different tissues. When, however, the antioxidant pathways were broken down during long exposures to light, the plant did not recover anymore from desiccation (Kranner et al. 2002). This demonstrates that antioxidant systems are essential components of the recovery pathway.

In leaves of X. viscosa, a new desiccation-inducible antioxidant enzyme, corresponding to a form of 1-cys peroxiredoxin (Prxs), has been identified (Ndima et al. 2001; Mowla et al. 2002). Peroxiredoxins are a class of conserved thiol-specific antioxidant enzymes (Chae et al. 1994), active on substrates such as hydroperoxides, and are common to all organisms from archaebacteria to plants and mammals (Kozak 1999). The Prx found in leaves of X. viscosa (XvPer1) shows more than 70% sequence identity to seed-specific 1-cys peroxiredoxins (Aalen et al. 1994; Haslekas et al. 1998; Lewis et al. 2000), and it is the only 1-cys Prx that has been noticed in vegetative tissues. Its transcript is absent in fully hydrated leaves, but it accumulates upon dehydration and upon other stresses that lead to increased levels of ROS (Mowla et al. 2002). The nuclear location of XvPer1 points to an involvement of this protein in protecting the nuclear compartment from free oxygen radicals (Mowla et al. 2002).

3.3 Abscisic Acid Regulates Desiccation Tolerance Pathways

The plant hormone abscisic acid (ABA) is known to be involved in embryo maturation and in the response to osmotic stress in vegetative tissues. The key role of ABA was demonstrated by physiological and genetic experiments as well as measurements of ABA levels. In most resurrection plants, including the aquatic species Chamaegigas intrepidus, ABA is involved in attaining desiccation tolerance and in stimulating the synthesis of dehydration-induced proteins (Gaff 1980, 1989; Gaff and Loveys 1984; Bartels et al. 1990; Reynolds and Bewley 1993; Hellwege et al. 1994; Schiller et al. 1997, Chap. 9andChap. 12). Leaves of M. flabellifolia and B. nitida did not survive dehydration, if they were dried so rapidly that ABA could not be accumulated (Gaff and Loveys 1984). A role for ABA in desiccation tolerance has also been demonstrated in C. plantagineum. Undifferentiated callus tissue of C. plantagineum, although intrinsically not desiccation tolerant, acquires tolerance after it has been cultured on medium containing ABA (Bartels et al. 1990). Molecular analysis suggests that ABA in resurrection plants has a similar function in signalling cascades as has ABA in dehydration responses in Arabidopsis. The role of ABA in signalling responses to dehydration in Arabidopsis has been intensively analysed through molecular genetics (Seki et al. 2007).

4 Gene Expression

Only a few resurrection plants have been studied on the molecular level, the best-studied species include C. plantagineum, X. viscosa, X. humilis, S. stapfianus, and B. hygrometrica (Table 16.2). Desiccation involves sequential gene expression programmes. Moderate tissue dehydration creates a drought condition also requiring the activation of drought tolerance genes in resurrection plants. More substantial drying, leading to a relative water content of 3% or less, entails the utilization of additional strategies and corresponding gene expression programmes. Thus, initially general mechanisms of water-deficit responses shared with drought-tolerant species are implemented; successively, specific programmes of desiccation tolerance are activated (Neale et al. 2000). It is assumed that gene expression in resurrection plants is geared towards the synthesis of protective molecules. Desiccation-induced genes can be grouped into genes encoding regulatory factors and proteins, which may exert a protective function directly or which may represent enzymes catalysing the synthesis of other protective molecules. Examples for different groups are described below.

To investigate the molecular basis of desiccation tolerance in a vascular resurrection species, Iturriaga et al. (2006) characterized 1,046 ESTs from a cDNA library constructed from S. lepidophylla undergoing desiccation for 2.5 h, which represented 873 unique transcripts. Putative functions were assigned to 653 (62.4%) of these clones after comparison with protein databases, whereas 212 (20.2%) sequences having significant similarity to known sequences whose functions are unclear and 181 (17.3%) sequences having no similarity to known sequences. For those ESTs for which functional assignments were made, S. lepidophylla had a higher percentage of ESTs within categories of molecular chaperons, i.e., heat shock proteins or late embryogenesis abundant (LEA) proteins. Gene discovery efforts, such as EST collections, constitute an important first step in understanding which genes need to be expressed for desiccation tolerance.

4.1 Regulatory Molecules

In view of the fact that upon dehydration resurrection plants synthesize transcripts in vegetative tissues that are similar or even identical to the seed maturation progamme of non-tolerant plants, it seems that evolution has found a way to express genes from other pathways for desiccation tolerance in vegetative tissues. This may also be true for other stress pathways, as exemplified by a heat shock transcription factor, which is expressed in response to dehydration in C. plantagineum but in Arabidopsis thaliana only in response to high temperature (C. Bockel and D. Bartels, University of Bonn unpublished). Another example for utilizing genes of existing pathways is the formation of chloroplasts during rehydration in X. viscosa as discussed above. Therefore, the search for regulatory switches is important for understanding the molecular mechanisms of desiccation tolerance. However, little work on regulatory genes has been reported in resurrection plants. The largest group of characterized genes represents members of different classes of transcription factors, which seem to be very similar in sequence and in function to the corresponding counterparts from non-tolerant genetic model plants. When the transcription factors, isolated from resurrection plants, were tested in e.g., Arabidopsis, they seem to function like the endogenous homlogues (for examples see below). Like it was shown for non-tolerant plants, the expression of dehydration-responsive genes in resurrection plants is mediated by ABA independent as well as ABA-dependent signal transduction pathways (Bartels 2005). It seems that the transcription factors have been optimized towards target gene activation, and it can be speculated that adaptation towards a special function in the desiccation pathway may modulate their role in gene activation. This hypothesis is in accordance with the observation that many transcription factors have been evolutionary conserved within a broad range of organisms.

Examples of the desiccation-induced transcription factors isolated thus far from resurrection plants have mainly been reported from C. plantagineum. Many of them belong to the Myb, the homeodomain-leucine zipper (HD-Zip), or the bZip (basic leucine zipper) families (Iturriaga et al. 1996; Frank et al. 1998; Ditzer and Bartels 2006). Plant myb-related genes comprise a large family and are likely to participate in a variety of functions (Meissner et al. 1999). Two Myb-related genes, cpm10 and cpm7, show differential expression and regulation in response to desiccation and ABA in C. plantagineum (Iturriaga et al. 1996). Cpm10 is expressed in undifferentiated callus tissue and is up-regulated by ABA, while cpm7 is induced by dehydration in roots. Transgenic Arabidopsis plants overexpressing cpm10 displayed increased tolerance to drought and salt stress (Villalobos et al. 2004). These plants also showed ABA hypersensitivity and glucose insensitivity, suggesting that cpm10 is involved in mediating ABA and glucose signalling responses in Arabidopsis as well as in the response to drought stress. HD-ZIP genes encode proteins that have only been identified in plants so far and are thought to regulate development and responses to environmental cues (Ramanjulu and Bartels 2002). Two HD-Zip genes, CPHB-1 and CPHB-2, are induced by dehydration in leaves and roots of C. plantagineum, but show different responses to exogenously applied ABA (Frank et al. 1998). ABA treatment induces the transcription of CPHB-2, but not that of CPHB-1. Five other HD-Zip genes have been isolated from C. plantagineum, two of which were induced by ABA in undifferentiated callus, while the other three were not. This suggests that these HD-Zip genes act in different pathways of the dehydration response; some mediated by ABA, while others are independent of ABA (Deng et al. 2002).

Research related to dehydration- and cold-regulated genes in Arabidopsis led to the discovery of the dehydration-responsive elements (DRE) (Yamaguchi-Shinozaki and Shinozaki 1993). Therefore, many genes that are responsive to cold and drought stress contain DRE/CRT (dehydration-responsive element/C-repeat) elements within their 5′ regulatory regions and are regulated by DREB/CBF (DRE-binding protein/C-repeat binding factor) transcription factors (Stockinger et al. 1997; Liu et al. 1998). The DREB/CBF proteins belong to the AP2/EREBP (ethylene-responsive element binding protein) family of transcription factors which are important regulators of flower development and plant responses to abiotic stresses (Riechmann and Meyerowitz 1998; Shigyo et al. 2006). One such AP2/EREBP transcription factor has been reported as a regulator of the drought response pathway in S. stapfianus (Le 2005).

Other characterized genes that are involved in the regulatory network are phospholipases, VP1 homologues, or the SAP domain transcription factor CpR18 (Bartels et al. 2006). Like for the nature of the regulatory genes, the recognition promoter elements seem to be conserved, and no desiccation-specific elements have been identified so far (Bartels et al. 2006). This supports the hypothesis that no new mechanisms were invented, but that new combinations of regulatory elements are sufficient for expressing genes in desiccation tolerance.

Mutants would be extremely valuable for the unravelling of the basis of desiccation tolerance, but no genetic system has been developed yet to generate mutants in resurrection plants. The reason for this is that most resurrection plants appear not to be diploid. Over 10 years ago an activation tagging approach, which generates dominant mutations led to the discovery of the unusual CDT-1 gene in C. plantagineum (Furini et al. 1997). No homologues of this gene have so far been identified in species apart from Craterostigma. Constitutive overexpression of cdt-1 leads to desiccation tolerance in C. plantagineum callus tissue activating the expression of desiccation-induced transcripts. CDT-1 transcripts accumulate in response to desiccation, but no corresponding polypeptides are synthesized; thus, CDT-1 functions as a regulatory, non-protein coding RNA (Hilbricht et al. 2008 (see also Chap. 13.3)). CDT-1 is a member of a large gene family, of which all members have features of retrotransposons (Smith-Espinoza et al. 2005). Retrotransposon elements may be required for expression of desiccation tolerance, as they are able to activate genes via small RNAs such as CDT-1. Small RNAs as regulators of gene expression are emerging players for essential environmental and developmental switches (Phillips et al. 2007).

4.2 Aquaporins

Aquaporin-mediated water transport across membranes provides a fundamental contribution to plant–water relations (Maurel and Chrispeels 2001 and Chap. 10). This makes it obvious to assume that aquaporins are involved in desiccation tolerance. In leaves of C. plantagineum, one TIP (tonoplast intrinsic protein) and several PIP (plasma membrane intrinsic protein) have been found, whose genes are differentially expressed during dehydration (Mariaux et al. 1998). A dehydration-responsive TIP has also been isolated from leaves of S. stapfianus (Neale et al. 2000). One particularly relevant finding is the identification of a tonoplast aquaporin, namely TIP 3;1, in dried leaves of E. nindensis (Vander Willigen et al. 2004). The interest in this desiccation-dependent channel protein is due to the fact that a TIP 3;1 had never been noticed in vegetative tissues. TIP 3;1 proteins (formerly called α-TIPs) are considered to be unique to orthodox seeds (Maurel et al. 1995). It participates in modulating the water permeability of the tonoplast during rehydration and may aid in mobilizing sugars, proteins, and other solutes that have accumulated inside the small vacuoles of bundle sheath cells with which it is associated (Vander Willigen et al. 2004).

4.3 Carbohydrates

A common observation in the desiccation process is the accumulation of soluble sugars, and the importance of sugars has been implicated in several studies. During desiccation, starch is rapidly converted into glucose (Crowe et al. 1998). Apart from a role in osmotic adjustment during the dehydration phase, the protective effects of sugars may also include protein stabilization in desiccated cells (Ramanjulu and Bartels 2002). Conversion of starch into sucrose is attributed primarily through activation of sucrose phosphate synthase by reversible protein phosphorylation upon perception of osmotic stress. Sucrose and trehalose have been proposed to prevent protein denaturation and membrane fusions (Crowe et al. 1998; Peters et al. 2007). Both sugars are capable of forming biological glasses (a process called vitrification) within the dried cell. Vitrification of the cytoplasm may not be due to the effects of sugars only, but probably results from the interaction of sugars with other molecules, most likely proteins (Hoekstra 2005). Cells of many desiccation-tolerant plants and animals undergo vitrification upon drying to protect organelles from damage (Buitink and Leprince 2004), and conformational changes of proteins and membrane fusion are prevented (Crowe et al. 1998). Vitrification is thought to limit the production of free radicals by slowing down chemical reactions and molecular diffusion in the cytoplasm (Hoekstra 2005). Thus, vitrification may be an important protective mechanism for resurrection plants.

In fully hydrated S. stapfianus, leaves, glucose, fructose, and galactose are present in large amounts. During dehydration, sucrose increases to high levels in air-dry leaves to become the predominant sugar (Ghasempour et al. 1998). Other sugars such as raffinose and trehalose are also detected, and they may complement the role of sucrose as osmotic protectants during drying (Rascio and La Rocca 2005). The changes in sugar composition observed in S. stapfianus leaves were quantitatively similar to those reported for other resurrection plants where the common trend was the conversion of monosaccharides in hydrated leaves into sucrose in dried leaves and vice versa in rehydrated leaves (Murelli et al. 1996). In C. plantagineum, the unusual eight-carbon sugar, 2-octulose, is the predominant sugar in hydrated leaves, which is converted to sucrose, with the reverse process being observed during rehydration (Bianchi et al. 1991). Sucrose also accumulates in the roots of drying C. plantagineum, although the most prevalent sugar stachyose does not undergo big changes (Bianchi et al. 1991; Norwood et al. 2003).

Sugar conversions during the desiccation process have been correlated with corresponding gene expression. Again, it seems that no new genes have evolved, but the conversion processes utilize existing genes of the sugar metabolism. Sucrose biosynthetic genes have been shown to be induced by both desiccation and ABA in the resurrection plant C. plantagineum (Kleines et al. 1999). Perhaps gene duplications may happen during acquisition of desiccation tolerance to accommodate desiccation-related metabolism. A possible example for this is the presence of a transketolase localized in the cytoplasm of C. plantagineum. It has been suggested that this transketolase is involved in the octulose sucrose conversions during the dehydration/rehydration cycle and has changed its substrate specificity (Willige et al. 2009).

4.4 Compatible Solutes

A common phenomenon in drought stress is the accumulation of organic compatible solutes because they are supposed to stabilize proteins and membranes (Levitt 1980; Crowe and Crowe 1992). Upon dehydration, many plants accumulate non-toxic solutes such as proline, mannitol, and glycine betaine (Chen and Murata 2002). The compatible solute accumulation results in an increase in cellular osmolarity, which in turn leads to an influx of water into, or at least a reduced efflux from cells (Hare et al. 1998). The main function of compatible solutes is, however, not so much to increase the water holding capacity of cells, but compatible solutes have been proposed to protect cells through the stabilization of cytoplasmic constituents and ion sequestration (Hare et al. 1998). This is due to the preferential exclusion of these compounds from the surfaces of proteins and membranes, which thus remain preferentially hydrated (Hoekstra et al. 2001).

Several resurrection plants have been analysed for the presence of potential compatible solutes. A diverse spectrum of molecules, which may serve as compatible solutes, has been identified, but no general mechanism is identified. Inositol, present in X. viscosa, may be an effective osmoprotectant (Mayee et al. 2005). Application of proline had no effect on detached leaves of B. nitida and M. flabellifolius (Gaff 1980). In the latter species, polyphenols have been identified that might be relevant to desiccation tolerance, as provenances from Namibia subjected to greater drought stress were genetically different from those in South Africa and contained more and different polyphenols (e.g., 3,4,5-tri-O-galloquinic acid) (Moore et al. 2005). M. flabellifolius contains a broad spectrum of potential osmotica in particular glucose–glycerol, which has been directly correlated with osmotic tolerance in cyanobacteria (Bianchi et al. 1993).

4.5 Protective Proteins: LEA Proteins and Heat Shock Proteins

The abundant accumulation of proteins during desiccation is the most obvious change on the molecular level. The majority of the proteins are LEA proteins. LEA proteins accumulate during seed maturation and during dehydration of vegetative tissues. A protective role for the LEA proteins has been suggested, and it is assumed that they are a major contribution to desiccation tolerance in resurrection plants (Ingram and Bartels 1996; Mtwisha et al. 2006). Despite this fact LEA proteins are not discussed further in this chapter, as there have been excellent recent reviews discussing LEA proteins in detail (Battaglia et al. 2008; Tunnacliffe and Wise 2007 and references within). Besides LEA proteins also heat shock proteins are associated with desiccation. The role of heat shock proteins in this context is summarized.

Proteins commonly indicated as heat shock proteins, being up-regulated by heat stress, also seem to be implicated in desiccation tolerance (Goyal et al. 2005). They are induced by the same stresses as LEA proteins and their synthesis coincides with the acquisition of desiccation tolerance (Wehmeyer et al. 1996; Mtwisha et al. 2006). These proteins function as molecular chaperones and play primary roles in protein biosynthesis, folding, assembly, intracellular localization, secretion, and degradation of other proteins (Feder and Hofmann 1999). Intracellular proteins such as heat shock proteins can compensate the loss of hydrogen bonds to water molecules by hydrogen’s bonding to other molecules. Generally, HSPs are able to maintain partner proteins in a folding competent, folded or unfolded, state to minimize the aggregation of non-native proteins or to target non-native or aggregated proteins for degradation or removal from the cell (Feder and Hofmann 1999). This function could be important in maintaining cells in resurrection plants viable.

Small heat shock proteins are expressed in the vegetative tissue of C. plantagineum during water stress and are correlated with desiccation-tolerant callus (Alamillo et al. 1995) and in the cysts of the desiccation-tolerant brine shrimp Artemia franciscana under desiccation stress (Clegg 2005).

5 Rehydration

Most research on resurrection plants has examined strategies that counteract desiccation damage and promote survival in the dry state. Less attention has been paid to the rehydration process. The events occurring during rehydration involve gradual recovery of correct cellular organization, which is accompanied by reactivation of normal metabolic functions. The processes leading to the reestablishment of vital functions and metabolic activities are an integral part of the entire phenomenon of desiccation tolerance. It seems that there are differences in the rehydration process in lower plants such as mosses and in higher vascular plants. It has been discussed whether desiccation tolerance in mosses such as Tortula ruralis is constitutive, with recovery essentially a matter of reassembly of components conserved intact through a drying–rehydration cycle, or non-constitutive repair processes. The idea of a repair-based mechanism of desiccation tolerance in bryophytes was introduced by Bewley (Bewley 1979; Bewley and Krochko 1982). It was extended by Oliver and Bewley (1984) to take into account the fine-structural change, and it has since been reiterated in a succession of reviews (Bewley and Oliver 1992; Oliver 1996; Oliver and Bewley 1997; Oliver et al. 1998).

The successful rehydration in C. plantagineum requires a discrete period of time (Bartels et al. 1990). Dried Craterostigma plants complete the water uptake in 12–15 h when placed in a water-saturated environment. The comparison of protein patterns from dried and rehydrated leaves of C. plantagineum revealed that no new mRNAs were detectable during early phases of rehydration (Bernacchia et al. 1996). Within the first 6–12 h of rewatering, desiccation-related transcripts and their corresponding proteins disappear. Only after 12–15 h, when the leaf tissues reach about 80–90% RWC, changes occur in translatable mRNA populations and protein profiles, mainly related to enhanced expression of genes that were down-regulated upon drying. New photosynthetic pigments are synthesized, and respiration, photosynthesis and other metabolic pathways are reactivated, showing that mRNAs or enzymes necessary to restore cell activities had been accumulated in a stable form during dehydration and were sufficiently protected in the dry state.

Even though the desiccation tolerance of resurrection plants largely depends on strategies of cell protection during drying, these plants also have mechanisms for preventing and/or repairing cell damage upon rehydration (Cooper and Farrant 2002). In rehydrated leaves of several species, the de novo synthesis of some proteins, also produced upon drying, has been observed. Proteins involved in both dehydration and rehydration include the previously mentioned expansins (Cooper and Farrant 2002). Moreover, aquaporins, whose mRNAs are produced and stored during the late stage of drying, also seem to play an important role in rehydration by regulating the rate of water flow into cells (Mariaux et al. 1998).

6 Conclusions and Outlook

In this chapter, we have tried to summarize what is known about the physiology and molecular analysis of desiccation tolerance in higher plants. The phenomenon of desiccation tolerance has been well characterized for a few species. However, the mechanism is not understood yet. The obstacles are the complexity of desiccation tolerance and the lack of genetic approaches connected with the complex genomes. The genomes known so far are relatively small in estimated size, but they are not diploid. Recently, DNA sequencing technologies have become available at an economically affordable level. This will allow transcriptome analysis for plant species for which no genome sequence is available and thus a comparative genomic analysis could provide more insight into desiccation tolerance (Rodriguez et al. 2010). It can be expected from such approaches to identify genes that are unique to desiccation-tolerant species and thus could give a clue for the requirements of desiccation tolerance.

In addition to understanding which genes are expressed and which proteins are active, epigenetic mechanisms turn out to be a new dimension of modifying adaptations to responses to environmental stresses. This has not been studied at all in any resurrection plants but needs to be considered for explaining desiccation tolerance mechanisms. Epigenetic mechanisms can be seen as enhancing the complexity of regulation of gene expression and may enhance the flexibility of plant genomes to adapt to diverse environments.