- Top of page
- Methods and Materials
- Supporting Information
The green plants constitute one of the major classes of eukaryotes on the planet, comprising approx. 500 000 species. Phylogenetic and palaeobotanic evidence suggests that the present variety of land plants diversified from a single colonization of the land during the mid-Ordovician period (c. 480 million yr ago; Kenrick & Crane, 1997). Fundamental to the success of plants in colonizing a terrestrial habitat is the acquisition of adaptations to an uncertain supply of water. The currently dominant taxa among the land plants – the tracheophytes – display numerous anatomical adaptations to the terrestrial environment, including ramifying root systems to scavenge water from the substratum, extensive vascular tissues for its delivery throughout the plant, and stomates, cuticles and lignin that restrict evaporative loss, whilst facilitating gas exchange, and providing mechanical strength. By contrast, the first plants to assume a terrestrial lifestyle lacked these adaptations and must, necessarily, have exhibited a variety of biochemical and physiological mechanisms to ensure their survival during times of drought (Oliver et al., 2005).
Most bryophytes retain the property of vegetative desiccation tolerance, in the form of anhydrobiotic survival, and occupy niches characterized by frequent cycles of dehydration and rehydration (Dilks & Proctor, 1974). In the evolution of the vascular plants, this property has been lost, in favour of the adaptations associated with added complexity and increased diversity. True desiccation tolerance, in the form of anhydrobiotic survival, has become restricted to the protection of metabolically quiescent reproductive propagules: spores and pollen (sometimes) and seeds (usually) (Oliver et al., 2000). In only a few remarkable species (termed ‘resurrection plants’) has vegetative desiccation tolerance re-evolved within the tracheophyte phylogeny (Ingram & Bartels, 1996; Bartels, 2005).
Nevertheless, many processes characteristic of the early stages of embryonic desiccation tolerance may be recognized within water-stressed vegetative tissues. These include the early accumulation of compatible osmolytes, and of potentially stabilizing compounds such as proline, glycine-betaine, polyhydric alcohols and disaccharides (Bianchi et al., 1991; Ishitani et al., 1995; Yoshiba et al., 1995; McKue & Hanson, 1996). ABA coordinates these stress responses, mediating physiological processes such as stomatal closure (an immediate response to restrict evaporative water loss), osmolyte accumulation, and also the synthesis of stress-related proteins, including late embryogenesis abundant (LEA) and heat shock proteins (HSPs), as well as compounds associated with the scavenging of reactive oxygen species that are implicated in desiccation-related membrane damage (Leopold et al., 1991; Ingram & Bartels, 1996; Hoekstra et al., 2001).
The molecular responses to dehydration in higher plants have been studied by genome-wide microarray analysis of gene expression processes in the model angiosperm, A. thaliana, resulting in the identification of several hundred individual genes whose expression is induced by dehydration stress, salinity stress and cold stress, as well as an additional subset of c. 150 genes expressed during the recovery from these stresses. The identification of groups of coregulated genes has enabled the identification of common sequence motifs among the promoters of these genes, and the identification of the transcription factors that regulate their expression (Seki et al., 2002; Oono et al., 2003; Shinozaki et al., 2003).
Since dehydration tolerance is an ancient evolutionary adaptation within the plant kingdom, it is of clear interest to determine the extent to which common genetic mechanisms leading to dehydration stress and desiccation tolerance have been conserved, and to identify the processes by which the phenomenon of desiccation tolerance has become developmentally restricted within the higher plant lineages. For the desiccation-tolerant moss Tortula ruralis, the ability to survive rapid and complete desiccation appears to rely on an ABA-independent, constitutive mechanism supported by the induction of a repair-associated gene set upon rehydration (Oliver, 1991; Oliver et al., 2004, 2005). Such a mechanism is not universal among bryophytes, however, and in the dehydration-resistant, but desiccation-sensitive, moss Physcomitrella patens, ABA-related induction of stress-related gene products, clearly homologous with higher-plant stress-related genes, is associated with the response to, and survival of, both dehydration stress (Knight et al., 1995; Machuka et al., 1999; Frank et al., 2005; Kamisugi & Cuming, 2005; Oldenhof et al., 2006), and of freezing tolerance (Minami et al., 2003; Takezawa & Minami, 2004; Oldenhof et al., 2006).
The developing genomic resources available for Physcomitrella have been exploited, using an oligonucleotide microarray to identify a larger set of genes expressed in response to ABA and water stress in a near relative of the earliest land plants. A number of genes whose stress-related expression is conserved between moss and angiosperms were identified. Analysis of the putative promoter regions of these genes identifies a number of potential cis-acting elements similar to those identified in stress-induced genes of higher plants. A number of GC-rich sequence elements were also identified that appear significantly over-represented in the promoters of the induced gene set, and that represent candidates for further functional characterization.
- Top of page
- Methods and Materials
- Supporting Information
During dehydration, substantial accumulation of novel transcripts was observed, many of which are also induced by the growth regulator ABA, and which encode conserved members of gene families associated with the acquisition of desiccation tolerance. A significant number of these encode LEA proteins: a class of protein originally identified in seeds during the later stages of embryogenesis, and since found to be highly conserved in other anhydrobiotic taxa, including green algae (Honjoh et al., 1995), nematodes (Browne et al., 2001) and rotifers (Tunnacliffe et al., 2005). LEA proteins are believed to relieve the consequences of desiccation through sequestration of water or ions (Cuming, 1999; Hoekstra et al., 2001), acting as chaperone-like ‘molecular shields’ (Goyal et al., 2005) or through structurally reinforcing the cell following desiccation-induced structural changes (Goyal et al., 2003; Wise & Tunnacliffe, 2004). Thus, there appears to be a close correspondence between the responses of ABA- and dehydration-stressed Physcomitrella protonemal tissue and the preparations made by the seeds of higher plants for the desiccation that is both an inherent part of their development and a requirement for the longevity that ensures their dispersal in time as well as in space.
Another significant group of Physcomitrella genes induced both by ABA and by dehydration encode proteins with a probable function in osmoregulation. These include membrane transport functions such as aquaporins and sugar transporters (Table 3), but also metabolic enzymes such as a stress-associated aldehyde dehydrogenase of the moss-specific ALDH21 subfamily (Chen et al., 2002; Kirch et al., 2004) (Table 2).
There are more genes whose transcripts accumulate in response to dehydration, than in response to other treatments, and also a significant number of genes that are down-regulated by dehydration but not by the other treatments. This may simply reflect the longer period over which dehydration stress was applied, relative to other treatments (24 vs 2 h). Alternatively, it may indicate the existence of ABA-dependent and ABA-independent response pathways. In flowering plants, ABA-specific induction of stress-related gene expression typically operates via the binding of bZip transcription factors to ABREs, a mechanism that also occurs in Physcomitrella (Knight et al., 1995; Kamisugi & Cuming, 2005). The angiosperm ABA-independent dehydration-specific response is mediated by the ‘DREB’ class of Apetala2-type transcription factors binding to a GC-rich DRE (Shinozaki & Yamaguchi-Shinozaki, 1997; Dubouzet et al., 2003; Shinozaki et al., 2003). The existence of a comparable pathway in mosses has yet to be experimentally demonstrated. However, our observation that a DREB-like transcription factor is up-regulated during dehydration stress, together with the identification of DRE-like elements within the 5′-flanking sequences of some dehydration-induced genes, suggests that this mechanism may also be conserved between bryophytes and angiosperms.
Transcripts exhibiting a substantial reduction in abundance included a preponderance of chloroplast-specific gene products. This was also observed in Arabidopsis (Seki et al., 2002) and likely accounts for the generally inhibitory effect of drought stress on photosynthetic activity. However, mosses typically exhibit morphogenetic changes following prolonged exposure to dehydration stress, or to ABA, resulting in the formation of brachycytes, or ‘brood cells’ (Bopp & Werner, 1993): these are typically thick-walled, lipid-rich vegetative spores (Schnepf & Reinhard, 1997), with substantially altered chloroplast morphology. Thus, widespread changes in chloroplast gene products may reflect the early stages of brachycyte differentiation. Similarly, proteomic studies of Physcomitrella brachycyte formation have identified specific changes in the extracellular protein spectrum in response to ABA (S. Tintelnot, pers. comm.), including changes in several gene products identified in this study (pectin methylesterase, proline-rich wall proteins, LRR-containing proteins, a plant-specific fasciclin and germin-like proteins) with probable wall-modifying functions.
Desiccation tolerance is widespread among bryophytes, yet the underlying mechanism of tolerance and its relationship to the mechanisms operating in the desiccation-tolerant stages of angiosperm development remain unclear. In the desiccation-tolerant moss Tortula ruralis, desiccation-associated changes in gene expression occur principally during the rehydration phase, rather than in the period during which water loss is occurring, leading to the suggestion that such species are constitutively prepared for desiccation, and that novel gene expression is required for the rapid repair of desiccation-induced cellular damage (Oliver, 1991; Oliver et al., 2004, 2005). It is clear that Physcomitrella differs in this respect, and interesting that many of the genes induced before dehydration encode proteins similar to those identified in the rehydration transcriptome of Tortula (Oliver et al., 2004, 2005).
It is noteworthy that the genes up-regulated by ABA and stress treatment of Physcomitrella are generally fewer in number than those identified in similar experiments undertaken with Arabidopsis thaliana. Physcomitrella is intermediate in its degree of dehydration stress tolerance in comparison with Tortula and Arabidopsis; although protonemal tissue does not survive desiccation, it is nevertheless highly tolerant of dehydration (Frank et al., 2005). Moreover, plants comprising both protonemata and gametophores will tolerate complete desiccation following slow drying, if first pretreated with ABA, a process that is associated with a substantial increase in intracellular concentrations of sucrose (Oldenhof et al., 2006), as is cold-acclimation leading to freezing tolerance (Nagao et al., 2005). This is similar to the ABA-mediated induction of desiccation tolerance in cultured tissue of the resurrection plant, Craterostigma plantagineum, and to the ABA-mediated acquisition of desiccation tolerance during angiosperm embryogenesis (Bartels et al., 1988, 1990; Bianchi et al., 1991; Ooms et al., 1993; Bartels, 2005; Smith-Espinoza et al., 2005). It may be that Physcomitrella retains a residual population of constitutively expressed genes with a stress-protective function, whereas the corresponding genes in angiosperms have undergone ‘evolutionary capture’ by an inducible mechanism.
It is clear that genes that confer desiccation tolerance have not been lost during the evolution of the land plants. Instead, their expression has become developmentally sequestered within the reproductive stages of the life cycle (typically during seed development). Moreover, the frequency with which vegetative desiccation tolerance has independently re-evolved throughout the land plant phylogeny (Oliver et al., 2005) implies that mutations in a relatively small number of regulatory genes may account for the this developmental sequestration.
Within the angiosperms, a subset of the genes expressed during late embryogenesis is restricted to this stage of development owing to a requirement for seed-specific transcription factors. Principal among these is the ABI3 class of transcriptional activator that mediates ABA-induced gene expression specifically in developing seeds through an interaction with the ABI5-type bZip factors (Ezcurra et al., 2000; Nakamura et al., 2001). In Physcomitrella protonemata, a similar mechanism activates the ABA-mediated expression of the group 1 LEA genes that in angiosperms are seed-specific (Knight et al., 1995; Kamisugi & Cuming, 2005; Marella et al., 2006). Interestingly, whereas all fully characterized angiosperm genomes contain only a single ABI3 gene, the Physcomitrella genome contains at least three such genes (Marella et al., 2006). If this multiplicity of the ABI3 family within the Physcomitrella genome reflects the situation more widely among the bryophytes, then it could be hypothesized that the evolutionary loss of the additional copies during the tracheophyte divergences, coupled with the developmental ‘capture’ of this transcription factor by the embryogenic developmental programme, would have resulted in the wholesale capture of its subordinate genes to this stage of the life cycle.