Portal de Periódicos da CAPES

Steering the solar panel: plastids influence development

2009; Wiley; Volume: 182; Issue: 2 Linguagem: Inglês

10.1111/j.1469-8137.2009.02808.x

1469-8137

Enrique López‐Juez,

Algal biology and biofuel production

As if they were industrial factories, plant cells reveal a multitude of interactions between compartments, with many activities requiring the traffic of components and products between them. Thirty years ago researchers made what seemed, at first, a surprising observation: a mutant of barley that, through maternal inheritance of ribosome-deficient plastids, possessed contiguous stripes of green and white tissue, showed, in the white tissue, not only very low levels of plastid-based enzymatic activities but also very low levels of synthesis template for them in the cytoplasm (Bradbeer et al., 1979). The deficient state of the plastids was providing information to the nucleus of those cells, leading to low mRNA levels of the nuclear-encoded genes for those enzymes in the cytoplasm where they were synthesized. The compartments were sharing information, including commands from seemingly subordinate organelles (namely the chloroplasts), to control expression of genes in the nucleus. This specific flow of interorganellar information is known as plastid–nuclear communication or plastid retrograde signalling. In this issue of New Phytologist, Ruckle & Larkin (pp. 367–379) provide evidence showing that this retrograde signalling controls not just the nuclear processes involved in the biogenesis and function of the organelle, but also aspects of the differentiation of cells and the development of the plant overall in its response to light, at least at its crucial seedling stage. Light is a key environmental cue without whose presence neither chloroplast biogenesis nor normal (photomorphogenic) seedling development take place. Ruckle et al. (2007) recently showed that plastid signals are capable of ‘rewiring’ the light signalling network controlling a gene (Lhcb1) for a major chloroplast protein; the cryptochrome1 (cry1*) photoreceptor and a bZIP downstream transcription factor, HY5, seemed to reverse their roles in the expression of this gene if plastids were damaged. In this issue, Ruckle & Larkin show that this reversal, as judged by the consequences of the absence of cry1, applies also to several other aspects of seedling photomorphogenesis: for example, the expansion of the first photosynthetic organs (the cotyledons) and, under some conditions, the elongation of the seedling stem (the hypocotyl), although it does not affect responses related to the production of sunscreens in nonphotosynthetic cells. ‘Thirty years after this intracellular conversation was identified in plant cells, a bewildering number of facts remain shrouded in mystery.’ Retrograde signalling, as observed by Bradbeer et al. (1979) (Fig. 1), is part of a wider set of mechanisms that ensure appropriate development and performance of chloroplasts within photosynthetic cells (recently reviewed by Pogson et al., 2008). When photobleaching of plastids occurs, or when plastid protein translation is impaired, the synthesis of nuclear-encoded polypeptides for many plastid proteins ceases. The generation of reactive oxygen species in the plastid also has impacts on nuclear gene expression, which are diverse, depending on the reactive species, for plastidic as well as for detoxifying, cytoplasmic proteins, and even initiates cell death; this is not simply a passive toxicity effect because it requires specific proteins (suitably named EXECUTERS, Lee et al., 2007). Meanwhile, at least in algae, the synthesis of plastid subunits of complexes is co-ordinated with that of their partners encoded in the nucleus by means of ‘epistasy of synthesis’, by which plastid translation is sensitive to the presence of unassembled polypeptides. Finally, even after initial development, redox imbalances resulting from unequal light excitation of photosystems are sensed and initiate, via kinase cascades, homeostatic gene-expression changes in both the plastid (Puthiyaveetil et al., 2008) and the nucleus (Bonardi et al., 2005). The signalling cascades of retrograde signalling proper are incompletely understood, but important insights began with the identification of ‘genomes uncoupled’ (gun) mutations, which partly uncouple the expression of nuclear genes, such as Lhcb1, from plastid dysfunction (Susek et al., 1993). A major breakthrough has been the identification of GUN1, a protein associated with plastid DNA (nucleoids) and that belongs to a family, several of whose members regulate organelle translation; evidence shows that it integrates signals from a variety of stimuli and initiates the repression of nuclear photosynthetic genes (Koussevitzky et al., 2007). The extent to which photo-oxidation and plastid translation defects share a signalling pathway, as well as the nature of the ‘photo-oxidation’ signalling molecule, have been a matter of conflicting evidence in the literature (Cottage et al., 2007; Koussevitzky et al., 2007; Mochizuki et al., 2008; Moulin et al., 2008). Evidence for plastid retrograde signalling. Arabidopsis seedlings carrying a nuclear reporter construct containing an Lhcb1 promoter driving the expression of β-glucuronidase (GUS) were grown for 10 d in white light in standard tissue culture medium (−lin) or for 2 d in standard medium and then for 8 d in medium supplemented with the plastid translation inhibitor lincomycin (+lin). The GUS activity appears as dark blue staining. The reporter is equivalent to that used as part of the gun mutant screening (Susek et al., 1993). One approach used in an attempt to unravel the plastid retrograde signalling pathway was identification of the promoter regions it targets. A key outcome of this approach was that it was found to be impossible to separate the target elements that mediate plastid and light regulation (Kusnetsov et al., 1996). Koussevitzky et al. (2007) demonstrated that ABI4, a transcriptional repressor that mediates much of the response dependent on GUN1, binds a CCAC element that overlaps the ACGT G-box element, mediating the light response for many photosynthesis-associated genes, including Lhcb1. In this manner, plastid dysfunction would prevent the light response. However other minimal elements have been identified that do not share this architecture, yet they can recapitulate, in gain-of-function experiments, both the light response and the plastid dependency (Acevedo-Hernández et al., 2005). It is evident that these two signalling pathways are central regulators of chloroplast development and are very closely related. The extent to which they are so became dramatically evident with the identification, by Ruckle et al. (2007), of cry1 as a gun mutant. In its wild-type form, cry1 is a nuclear protein and the main photoreceptor responsible for perceiving blue light. Although cry1 was a weak gun mutant on its own, double cry1 gun1 mutants showed almost complete Lhcb1 independence of plastid status. It is interesting to note that the presence of cry1 had also been shown to be essential for a form of programmed cell death induced by singlet oxygen generation in plastids in the light (Danon et al., 2006). Importantly, under normal conditions of normal chloroplast function, cry1 is an activator of photosynthetic gene expression and chloroplast development, whereas under conditions of plastid stress it was the absence of cry1 that stimulated Lhcb1 expression. This could be interpreted as the recruitment of different partners converting a transcriptional activator (the downstream HY5) into a transcriptional repressor. Larkin & Ruckle (2008) described this broadly as a ‘gas and break’ system that adjusts the synthesis of at least some chloroplast proteins to the prevailing light and plastid functional status. Ruckle & Larkin now show that the expansion of the seedling's first photosynthetic organs, the cotyledons, is under similar control: it is promoted by cry1 in the light when plastids are functional, but repressed by cry1 when plastids are dysfunctional. In their scrupulously carried out experiments, Ruckle & Larkin showed that a similar, ‘switchable’ control by cry1 also applies to the hypocotyl, although only under certain conditions. The plastid functional status therefore appears to have a relatively broad developmental influence, and plastid dysfunction in general represses photomorphogenesis, at least for photosynthetic organ development. Some degree of repression of photomorphogenesis had also been observed in mutants with impaired plastid development (Vinti et al., 2005). In retrospect, this developmental influence is not surprising. It has been observed repeatedly that mutations which affect chloroplast differentiation also prevent the normal differentiation specifically of leaf palisade cells (Bellaoui & Gruissem, 2004) and that the developmental response of such cells to high light is severely impaired, in a cell-autonomous manner, by plastid dysfunction (Fig. 2). Although not highlighted as such, one of the most spectacular observations of Ruckle & Larkin is the fact that the differentiation of guard cells in the epidermis of cotyledons is influenced by plastid signals: in both gun1 and cry1 mutants plastid dysfunction greatly reduces the expansion of epidermal cells, but only in the gun1 mutant does this cause a large increase in the density of stomata, implying that under certain conditions, complete cell-fate decisions can be under organellar control. Developmental, cell-autonomous effects of the plastid status. A leaf of the variegated chm1 mutant, fully expanded under high fluence rate white light (600 µmol m−2 s−1), was fixed and sectioned for microscopic analysis. Palisade cells on the green sector (left of the section) appear elongated, whereas those in the white sector are less so, and the number of cell layers is greater in the green sector, implying a plastid function requirement for cell elongation and cell division. Adapted from Tan et al. (2008). The concept of ‘rewiring’ provides a beautifully simple evolutionary design principle for the action of plastid signals. The impact of the signals dependent on GUN1 would feed into a pre-existing network that utilized HY5, a positive regulator of many light-responsive genes. Conversely, it is worth noting that a form of plastid retrograde signalling, one in which the nuclear gene for a heat shock protein depends on plastid stress signals (von Gromoff et al., 2006), occurs in Chlamydomonas, a single-celled green alga in which chloroplasts do not require light to develop. The plastid stress signals are both chlorophyll precursors and haem in this case. It is difficult to gauge whether this pathway is orthologous to any of those that control Lhcb1 in flowering plants. It has, however, been shown that plastid retrograde signalling is functional in the absence of COP1, a ubiquitin ligase that targets HY5 for degradation and thus acts as a central repressor of photomorphogenesis in the dark (Sullivan & Gray, 1999). COP1 belongs to an ancient class of conserved eukaryotic regulators of basic cellular processes. The control of chloroplast development by light, meanwhile, is likely to be relatively recent, because many gymnosperms, like pine, while showing photomorphogenic responses, exhibit greening in the dark and light-independent expression of Lhcb1 (Yamamoto et al., 1991). It would be interesting to explore which of these two signalling processes controlling chloroplast development preceded which, in evolutionary terms, and this would also help to reconstruct the intricacies of their interaction. The existence of gun mutants provides genetic evidence most-parsimoniously consistent with plastid signals being active repressors (i.e. ‘distressed’ plastids produce negative signals that repress photosynthetic nuclear genes) and loss of the GUN genes relieves this repression. This is the conceptual framework for the ‘break’ metaphor of Larkin & Ruckle (2008). One could argue that the gun mutants are, to plastid retrograde signalling, equivalent to what the de-etiolated or constitutively-photomorphogenic mutants are in relation to light signalling. Of course, the existence of the latter does not imply the absence of active photoreceptors. This is an interesting issue, but one very difficult to tackle experimentally. Do distressed plastids produce repressive signals, or do functional plastids produce positive signals that are lost when plastids malfunction? In fact, the answer could be affirmative for both questions. While chloroplasts with accumulated, photodynamic chlorophyll precursors, like those of the flu mutant (Lee et al., 2007), can be easily conceptualized as sources of repressive signals, the lack of photosynthetic gene expression when plastid development is severely reduced, for example when aminoacyl-tRNA synthetases are lacking and plastids barely differentiate (Barkan & Goldschmidt-Clermont, 2000), is less easily so. Would such plastids contain highly active, repressive GUN1? While activity would be difficult to measure, the levels of GUN1 in such plastids could easily be. It is hard, however, to imagine how to address experimentally this ultimate nature of plastid signalling except by examining the gene expression status of cells devoid of plastids, functional or otherwise – certainly a challenging task. Thirty years after this intracellular conversation was identified in plant cells, a bewildering number of facts remain shrouded in mystery. Clearly, a cell is more than the sum of its parts, and the complexity of the exchange of information between those parts does not cease to amaze us.