Transcriptional or post‐transcriptional regulation – how does a plant know when to senesce?

Plant senescence is familiar to most people in the form of autumn leaf senescence, a process that results in the yellow and red coloration of trees. This process is important for the recycling of nutrients, which would be lost if the old leaves were dropped without prior senescence. One of the best studied individual trees is probably an aspen (Populus tremula) tree on the Umeå University campus (e.g. Keskitalo et al., 2005). In this tree, leaf senescence starts on the same day in September every year. So how does the tree know that it is autumn? This question was addressed by Stefan Jansson (Umeå Plant Science Centre, Sweden) in his introductory lecture of the Third European Workshop on Plant Senescence, organized by Karin Krupinska (University of Kiel, Germany) and Klaus Humbeck (University of Halle, Germany). How autumn leaf senescence and senescence of annual plants are regulated was one of the main topics discussed during the workshop. Other important areas of senescence research that were addressed at the workshop included the processes involved in nitrogen remobilization from senescing leaves and programmed cell death in plants. ‘... initiation of senescence may not be regulated at the transcriptional level and ... induction of SAGs is likely to be a consequence rather than the cause of senescence’ The molecular processes that initiate senescence have occupied scientists for some time. As early as 1980, Thomas and Stoddart (Thomas & Stoddart, 1980) described a ‘temporal discontinuity between the transcription of senescence-specific genes and their expression as the senescence syndrome’, raising the question of whether transcriptional regulation can be responsible for the initiation of senescence. While temperature affects the rate at which aspen leaves senesce, the initiation of senescence is under tight photoperiodic control, preparing the tree for a possible drop in temperature once the days get shorter. Microarray studies conducted by Stefan Jansson's group (Umeå Plant Science Centre, Sweden) revealed that, although the senescence-related decline in chlorophyll starts on the same day each year, changes in the expression of senescence-associated genes (SAGs) do not coincide when different years are compared. This suggests that the initiation of senescence may not be regulated at the transcriptional level and that induction of SAGs is likely to be a consequence rather than the cause of senescence. In contrast to autumn leaf senescence in trees, senescence of annual plants, such as Arabidopsis, is related to reproduction. Microarray studies (e.g. Buchanan-Wollaston et al., 2005; van der Graaff et al., 2006) have revealed a large number of Arabidopsis SAGs, but it is still unclear which of them have a regulatory function. Various approaches are being used to pinpoint regulatory genes, for example Vicky Buchanan-Wollaston's group (University of Warwick, UK) are using systems biology to investigate transcriptional networks in order to identify hubs in the regulation of gene expression. Another group which has focused on identifying transcriptional regulators of senescence is that led by Bernd Müller-Röber (University of Potsdam, Germany). They have analysed the expression of over 2000 transcription factors using quantitative reverse transcription–polymerase chain reaction (RT-PCR). One member of the group, Salma Balazadeh, presented analysis of a mutant in a senescence-induced transcription factor confirming a role for this transcription factor in senescence regulation. Similarly, WRKY53 has been identified as a senescence regulator (Miao et al., 2004) by Ulrike Zentgraf's group (University of Tübingen, Germany). Transcription of genes during senescence could also be under epigenetic control. While the role of epigenetic control of flowering has been analysed in detail, research on the epigenetic regulation of other life history traits that affect plant fitness (Blödner et al., 2007) is still in its infancy. Klaus Humbeck (University of Halle, Germany) presented exciting results on the importance of epigenetic changes in chromatin structure for the regulation of senescence. Humbeck showed that senescence can be delayed or accelerated, respectively, by overexpression or knock out of a histone methyltransferase. These recent experimental findings using Arabidopsis confirm that transcriptional events play a role in senescence regulation, but to what extent they determine the initiation or only the rate of senescence, once it has been initiated, remains to be shown. Analysis of the proteome of senescing leaves by Shimon Gepstein's group (Technion, Haifa, Israel) has revealed that only half of the genes whose transcript levels are increased during senescence also show up-regulation at the protein level, suggesting a role for post-transcriptional regulation. Perhaps the most intriguing presentation of the workshop discussed a key role for a rare post-translational modification in the regulation of senescence: John Thompson (University of Waterloo, Canada) explained how hypusination of eukaryotic translation initiation factor SA (eIF5A), a protein that recruits mRNAs from the nucleus for protein synthesis, can determine senescence and growth in plants (Duguay et al., 2007). While the hypusinated form probably promotes senescence, the unmodified lysine form of eIF5A promotes plant growth. Manipulation of this post-translational switch provides new opportunities for crop production, for example in extending the shelf life of fruits and enhancing growth. A possible role of protein targeting was described by Karin Krupinska (University of Kiel, Germany). The single-stranded DNA-binding protein Whirly1 from barley, which binds to the promoter of a barley SAG, localizes to the plastids in young flag leaves, but to the nucleus in old flag leaves. The Arabidopsis Whirly protein (AtWhy1) also localizes to the plastid (Krause et al., 2005), but it modulates telomere length in the nucleus (Yoo et al., 2007). In mammalian cells, telomere length is related to senescence, whereas in senescing leaves telomere length is not affected. Nonetheless, modifications in telomere structure may play a role in leaf senescence (Zentgraf et al., 2000). Whereas mutants in AtWhy1 do not show a developmental phenotype, knocking out the Arabidopsis gene for a mitochondrial Whirly protein, AtWhy2, results in delayed senescence, suggesting that protein targeting may be important for senescence regulation. As discussed at the workshop, senescence can be triggered by biotic and abiotic stress. In addition to these environmental factors, internal, age-dependent factors influence senescence by determining the response of plants to their environment. In the Mediterranean perennial plant Cistus clusii, plant age determines the response to drought and high light stress. Sergi Munné-Bosch (University of Barcelona, Spain) presented data showing that, when leaves of the same age originating from plants of different age are compared, the leaves from old plants are found to suffer increased oxidative stress (Munné-Bosch & Lalueza, 2007). Age is also an important component in the senescence response of annual plants. In Arabidopsis, as demonstrated by Paul Dijkwel's group (University of Groningen, the Netherlands), ethylene only induces senescence in plants that have reached a defined age. This treatment has been used to isolate onset of leaf death (old) mutants. The old5 mutant (Jing et al., 2005), which responds early to ethylene treatment and also senesces early in air, was described by Jos Schippers. This mutant is affected in quinolinate synthase, an enzyme that is involved in the synthesis of nicotine adenine dinucleotide (NAD). Metabolic profiling of the old5 mutant revealed changes in primary metabolism. In yeast, NAD has been proposed to play a role in extending lifespan in response to calorie restriction, for example during decreased glucose availability (Lin et al., 2000). In this process, an NAD-dependent histone deacetylase, silent information regulator (Sir2), is required for epigenetic gene silencing. Similar to calorie restriction in heterotrophic organisms, changes in primary metabolism, especially sugar signalling, also play a role in the regulation of leaf senescence (Wingler et al., 2006). At the workshop, this subject was addressed by Thomas Roitsch (University of Würzburg, Germany), who showed how the regulation of cell wall invertase by cytokinin can control senescence. The old5 mutant could now provide a link between changes in metabolism, the epigenetic regulation of senescence and plant age. During the workshop it became clear that plant senescence is regulated at several levels. However, trying to separate the causes and consequences of senescence raises several ‘chicken and egg’ problems. For example, what comes first – transcriptional or post-transcriptional regulation, age or stress, metabolites or gene expression? Latest research makes it possible to suggest a model in which age-dependent changes to chromatin and possibly also telomere structure could make a leaf or a plant competent to react to senescence-inducing environmental stimuli, such as day length or stress (Fig. 1). These stimuli could regulate gene expression either directly or, for example, via plant hormones, metabolites or reactive oxygen species. Changes in metabolism are likely to be the cause as well as the consequence of transcriptional and post-transcriptional changes. Furthermore, a possibly age-related switch in post-translational modification, such as hypusination of eIF5A, could regulate protein synthesis. While downstream processes, such as the degradation of chlorophyll, are highly conserved (Armstead et al., 2007), there are many different ways in which senescence can be induced. Unravelling the causal relationships in the regulatory network therefore remains a challenging task. Hypothetical model for regulatory mechanisms that control senescence in plants.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.