JXB Advance Access originally published online on November 15, 2006
Journal of Experimental Botany 2007 58(1):11-26; doi:10.1093/jxb/erl196
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
RESEARCH PAPER |
Plastid biogenesis, between light and shadows
School of Biological Sciences, Royal Holloway, University of London, Egham, Surrey TW20 0EX, UK
* To whom correspondence should be addressed. E-mail: e.lopez{at}rhul.ac.uk
Received 13 June 2006; Accepted 11 September 2006
 |
Abstract |
|---|
 Top Abstract IntroductionPlastid biogenesis: the...Plastid biogenesis: the...References |
|---|
Plastids are cellular organelles which originated when a photosynthetic prokaryote was engulfed by the eukaryotic ancestor of green and red algae and land plants. Plastids have diversified in plants from their original function as chloroplasts to fulfil a variety of other roles in metabolite biosynthesis and in storage, or purely to facilitate their own transmission, according to the cell type that harbours them. Therefore cellular development and plastid biogenesis pathways must be closely intertwined. Cell biological, biochemical, and genetic approaches have generated a large body of knowledge on a variety of plastid biogenesis processes. A brief overview of the components and functions of the plastid genetic machinery, the plastid division apparatus, and protein import to and targeting inside the organelle is presented here. However, key areas in which our knowledge is still surprisingly limited remain, and these are also discussed. Chloroplast-defective mutants suggest that a substantial number of important plastid biogenesis proteins are still unknown. Very little is known about how different plastid types differentiate, or about what mechanisms co-ordinate cell growth with plastid growth and division, in order to achieve what is, in photosynthetic cells, a largely constant cellular plastid complement. Further, it seems likely that major, separate plastid and chloroplast ‘master switches’ exist, as indicated by the co-ordinated gene expression of plastid or chloroplast-specific proteins. Recent insights into each of these developing areas are reviewed. Ultimately, this information should allow us to gain a systems-level understanding of the plastid-related elements of the networks of plant cellular development.
Key words: Chloroplast, light, photosynthesis, phytochrome, plastid
|
Introduction |
|---|
|
TopAbstractIntroductionPlastid biogenesis: the...Plastid biogenesis: the...References |
|---|
Chloroplasts, arguably, define plant life. The overall vegetative structure of a plant can be ultimately conceptualized as a set of organs (the leaves) occupied by cells filled with chloroplasts and exposed to light and the atmosphere, and accompanied by their ‘ancillary’ organs, the roots and stem. In other words, plants are, effectively, self-standing solar panels, with chloroplasts being the solar ‘cells’. Chloroplasts, however, do not just carry out photosynthesis (photoreduction of carbon, nitrogen, and sulphur), but are central hubs in plant metabolism (Neuhaus and Emes, 2000). They manufacture fatty acids, aromatic and non-aromatic amino acids (essential for protein synthesis, but also for a vast array of plant secondary metabolites), purine and pyrimidine bases, isoprenoids (like carotenoids and sterols) and tetrapyrroles (like haem and chlorophyll). Most of these functions are essential for every cell type, and chloroplasts have integrated into cellular development pathways by differentiating into a variety of other, interconvertible, non-photosynthetic plastid types (Whatley, 1978; Waters and Pyke, 2004). In parallel with cellular differentiation, the range of plastid types even includes a slimmed-down, ‘meristematic cell’ equivalent, the undifferentiated proplastid.
Like mitochondria, plastids are double-membrane organelles derived from an engulfed endosymbiont, in their case a photosynthetic cyanobacterium. Its closest, well-characterized, known living relatives belong to the genus Nostoc (Dyall et al., 2004; Martin et al., 2002), whose genome encodes in excess of 5000 proteins (depending on strain). Plastids have retained a semi-autonomous character, a minimal genetic machinery, and genes for a small number of polypeptides, the expression of which needs to be directed by the nucleus at appropriate times. However, the majority of plastid proteins are encoded in the nucleus, translated in the cytosol, imported into the organelle and further targeted to one of its suborganellar compartments. Plastids also need to grow and multiply to keep pace with their ‘host’ cells, and increase their number by binary fission. Furthermore, plastids ‘report’ on their physiological status to the nucleus of the cell, to ensure co-ordination between the two genomes (Nott et al., 2006). This review will attempt to summarize the recent, and dramatic, progress in these areas. However, it intends to argue that some very important aspects of plastid biology remain very poorly understood, to summarize recent insights into some of the less-well-explored areas, and to suggest those areas in which increased effort of inquiry could be justified.
A recent comprehensive update on plastid biology is available (Møller, 2004). Excellent reviews on plastid genetics (Sugiura et al., 1998; Mache and Lerbs-Mache, 2001; Wakasugi et al., 2001), plastid protein import (Jarvis and Robinson, 2004; Bedard and Jarvis, 2005), division (Osteryoung and Nunnari, 2003), interorganellar communication (Nott et al., 2006), and the role of the ancestral organellar genome in the host (Timmis et al., 2004) have been published. A more extensive overview of plastid developmental processes, and of the evidence for roles of plastids in plant development, has been attempted elsewhere (López-Juez and Pyke, 2005).
|
Plastid biogenesis: the achievements |
|---|
|
TopAbstractIntroductionPlastid biogenesis: the...Plastid biogenesis: the...References |
|---|
Endosymbiosis and the origin of plastid proteins
The evidence for an endosymbiotic origin of chloroplasts is overwhelming. Initial cell biological observations noted not only the morphological similarity of chloroplasts and free-living cyanobacteria but also the apparent relative autonomy of behaviour of chloroplasts within plant cells (Martin and Kowallik, 1999). This included physical intracellular movements, and even the production and subsequent retraction from the body of chloroplasts and other plastids, of transient projections, now called stroma tubules or stromules and the subject of renewed interest (Köhler and Hanson, 2000).
The endosymbiotic process has gained great new insights in recent years. The availability of entire nuclear and plastid genome sequences has made it possible to compare those of photosynthetic eukaryotes with cyanobacteria and non-photosynthetic eukaryotes (Martin et al., 2002; Richly and Leister, 2004; Timmis et al., 2004). Cyanobacterial genomes, of Synechocystis and of two species of Nostoc, are estimated to contain from 3000 to over 7000 genes. The chloroplast proteome has been estimated for Arabidopsis and rice, based on the presence of transit peptides in nuclear-encoded genes, predicted through a combination of computing algorithms (see below). The number of chloroplast proteins thus predicted, with an obviously large degree of uncertainty (Richly and Leister, 2004) ranges from 2100 (Arabidopsis) to 4800 (rice). Meanwhile chloroplast genomes encode around 130 genes in total, of which around 80 code for proteins (Martin et al., 2002; López-Juez and Pyke, 2005). It is obvious that (a) the majority of the chloroplast ancestral genome is now located in the nucleus of plant cells, and (b) the majority of plastid proteins are encoded in the nucleus. A major outcome of these studies, however, is the observation that an extensive amount of reshuffling has taken place (genes in these ‘a’ and ‘b’ sets overlap only by about 50%), with both pre-existing eukaryotic genes having been recruited into functions located in the plastids, and cyanobacterial genes having acquired novel plastid-unrelated functions in the nucleus of plant cells (Martin et al., 2002; Timmis et al., 2004). An example of the former is the carbon-reducing Calvin cycle, which is composed of a number of enzymes of both cyanobacterial and host origin. An example of the latter is probably the range of bacterial sensory histidine kinases now present in plants, which are absent in metazoans and very rare in fungi. A note of caution has been introduced by the analysis of the full genome of the unicellular red alga Cyanidioschyzon merolae (Matsuzaki et al., 2004). This tiny red alga (cell of 2 µm diameter, with a single chloroplast and a single mitochondrion) is predicted to have a minimal photosynthetic eukaryote genome, with fewer genes (5300) than Nostoc. Cyanidioschyzon does not, for example, contain a gene for phytochrome, the light sensor in plants derived from bacterial histidine kinases, and has no nuclear-encoded histidine-kinase response regulators. Whether this is a result of gene loss or reflects a smaller contribution of the chloroplast ancestor's genome than otherwise thought, remains an open question. Its chloroplast, however, appears to be derived from the same, single endosymbiotic event that gave rise to all extant chloroplasts.
Plastid genetics
Plastids possess a genome (plastome) of between 120 and 160 kbp that encodes between 120 and 135 genes (see the Organelle Genome Megasequencing Program, http://megasun.bch.umontreal.ca/ogmp, for a complete set of available genomes). Besides polypeptides, the genome encodes eubacterial-type ribosomal and transfer RNAs. Many plastome genes are organized in operons. Plastome-encoded proteins include subunits of the eubacterial-type RNA polymerase (rpo), further genetic machinery (splicing and ribosomal proteins), photosynthetic polypeptides, including several for the four main thylakoid complexes, NADH dehydrogenase genes, and a few polypeptides of other functions (Sugiura et al., 1998; De las Rivas et al., 2002). Plastids are highly polyploid, with plastomes probably arranged as concatenated, long, linear molecules rather than small circles, and, in general, physically linked to the inner plastid envelope through Plastid Envelope DNA-binding (PEND) and other proteins (Sato et al., 2003; Wycliffe et al., 2005). Plastome copy number, at least in leaf cells, is very high (several thousand per nuclear genome), undergoes a phase of proliferation in young cells (Baumgartner et al., 1989) and in spite of reports to the contrary, remains constant thereafter (Li et al., 2006). One possible explanation for the highly polyploid nature of chloroplasts is the need to synthesize very large amounts of photosynthetic proteins. Another (not mutually-exclusive) reason is that it confers protection against mutations that, in asexually reproducing organisms (as these organelles are), accumulate without a possibility of repair through recombination during sexual reproduction. The occurrence of this protective gene conversion mechanism has recently been proven to occur in chloroplasts (Khakhlova and Bock, 2006).
Nuclear control of plastid gene expression
The expression of plastid-encoded genes is carried out by two RNA polymerases of different origin (Shiina et al., 2005). One is the plastid-encoded polymerase (PEP), which has been retained from the ancestral endosymbiont, is composed of three subunits (encoded by rpoA, B, and C genes) and recognizes E. coli-like promoters (Suzuki et al., 2004). The other is nuclear-encoded (NEP), phage-type, made of a single subunit, and probably derived from its homologue, the mitochondrial RNA polymerase (Hedtke et al., 1997; Sato et al., 2003). In fact, three phage-type polymerases are encoded in the nucleus of Arabidopsis, one targeted to mitochondria, one to chloroplasts, and one to both organelles (Hedtke et al., 2000). A mutation in the dual-targeted NEP polymerase, RpoT;2, results in defects only in plastid gene expression, and delayed greening and leaf and root growth (Baba et al., 2004). A mutation in the plastid-only NEP causes more severe defects in leaf development and gene expression (Hricová et al., 2006), and the combination of both of these mutations results in very early seedling lethality (Hricová et al., 2006). Although genes transcribed by both NEP and PEP polymerases exist, in general, a sequential action occurs: genes with NEP promoters are transcribed early in chloroplast development and are involved in housekeeping functions, primarily constituting the plastid genetic machinery, including the subunits of PEP. Thereafter PEP is involved in the expression of photosynthesis-related genes (Hajdukiewicz et al., 1997). Interestingly, a developmental changeover seems to take place, with glutamyl-tRNA, a product of PEP transcription that is also a precursor for chlorophyll biosynthesis, actively binding and repressing NEP, and therefore ensuring that once a commitment to photosynthetic development has taken place, minimal resources are ‘diverted’ (Hanaoka et al., 2005).
Although PEP is plastid encoded, the transcription of PEP-transcribed genes is also under nuclear control. This is a consequence of sigma-70 factors, which in eubacteria determine promoter specificity of the RNA polymerase, being encoded in the nucleus of plant cells (Isono et al., 1997). In Arabidopsis, six sigma factors exist with partially discrete functions. For example, SIG6 is a sigma factor that acts early and generically in chloroplast development (Ishizaki et al., 2005). SIG6 function is probably taken over by SIG3 and SIG1. SIG1 is a general sigma factor but is expressed later that SIG6 (Ishizaki et al., 2005). SIG3 is constitutively expressed, but the protein associates with plastid internal membranes which are not well developed early on (Privat et al., 2003). SIG2, meanwhile, is involved in transcription of tRNAs, including the precursor of chlorophyll, Glu-tRNA (Kanamaru et al., 2001). One example of a regulatory circuit has recently been uncovered: a number of plastid-encoded photosynthetic genes, including psbD, are induced by intense blue and red light (Mochizuki et al., 2004), probably as an adaptive response to the fact that the polypeptides they encoded are damaged by high light, and need to be turned over. The photoreceptors are the nucleo/cytoplasmic phytochrome A and cryptochromes, their primary action being the induction of AtSIG5 gene expression. The SIG5 protein then activates the psbD blue light-responsive promoter (Nagashima et al., 2004; Tsunoyama et al., 2004).
The specific role of SIG5 beautifully illustrates how the nucleus holds control over plastid gene expression. As indicated earlier, this is one direction in a two-way traffic of control processes between the nucleus and the plastid. The physiological status of the plastids sets in motion signalling processes which, in turn, control the expression of nuclear-encoded, plastid-related genes (reviewed by Nott et al., 2006).
Protein import and targeting
The endosymbiotic event, followed by transfer of genes to the nucleus, posed obvious challenges for the cell. One was the need to import proteins synthesized in the nucleo/cytoplasm into the organelle, i.e. across the double envelope. Second was the routing of these imported polypeptides, either to remain in the stroma, be targeted into the thylakoid membrane or the thylakoid lumen, or indeed end as components of either chloroplast envelope or its intermembrane space (a total of six possible destinations; Jarvis and Robinson, 2004). While the import stage, the uptake of proteins, is a novel process, the re-routing is equivalent to an ancient, pre-existing prokaryotic process, protein targeting or secretion, and indeed some pathways used by chloroplasts are shared with eubacteria.
The translocation of polypeptides across the envelopes is carried out by the Toc (translocon of the outer envelope of chloroplasts) and Tic (translocon of the inner envelope of chloroplasts) complexes (Fig. 1). The structure and evolution of these have been the subject of excellent reviews (Soll, 2002; Jarvis and Robinson, 2004; Bedard and Jarvis, 2005). The complexes recognize cytosolic proteins that carry a plastid transit peptide, an N-terminal sequence of between 20 and 100 amino acids, with surprisingly little conservation, except for a high content of hydroxylated and small amino acids and a low content of large and acidic amino acids (i.e. having a positive charge overall). Toc is composed of five transmembrane proteins. One, Toc75, which is likely to form the pore complex, has a transmembrane ß-barrel structure and is part of an ancient family of outer membrane proteins (OMPs) present in eubacteria and organelles (Moslavac et al., 2005). Two other core polypeptides, Toc 159 and Toc34, jointly contribute to the recognition of plastid transit peptides. These proteins function through a cycle of GTP hydrolysis, are of eukaryotic origin, and their mode of action has been extensively explored, although uncertainties remain as to whether, for example, the receptor Toc 159, in soluble form, directs plastid protein precursors from the cytosol towards the pore, or awaits them, as a membrane protein at the pore. The Tic complex is less fully understood, with conflicting evidence as to whether Tic20 or Tic110 form the pore (Bedard and Jarvis, 2005). The driving force for translocation is probably provided by chaperones and cochaperones part of, or associated with, Tic (Hsp93 and Tic40; Chou et al., 2003). After import, the N-terminal transit peptide is removed by a stromal processing peptidase.
|
pH (Tat) systems import proteins into the lumen. Adapted from Not all plastid proteins are imported through Toc/Tic complexes. Envelope proteins have unique properties. Outer envelope proteins often do not contain transit peptides (Hofmann and Theg, 2005), and some inner envelope proteins have been shown to be routed by novel, Toc-independent pathways (Nada and Soll, 2004). One outcome of the systematic identification of plastid proteins is the large number of experimentally-determined envelope-associated proteins that, instead of containing transit peptides, are computationally predicted to be targeted to the endomembrane, secretory pathway (Kleffmann et al., 2004). This prediction reflects the fact that they contain a nearly-N-terminal transmembrane domain, that resembles a cleavable signal peptide, but that is in fact used to target directly to the chloroplast envelope, and is never cleaved (Lee et al., 2001; see below).
Diversity of import pathways
The most important recent insight into plastid protein import pathways has probably been that they are, to some extent, substrate-specific (Kessler and Schnell, 2006). This prevents low-abundance but essential proteins from being out-competed by the much more abundant photosynthesis-related ones, and is possible thanks to the existence of small gene families for most of the translocon components. For example, for the four homologues of Toc159 encoded in the Arabidopsis genome, a deficiency in AtToc159 causes a loss of photosynthetic proteins, and a phenotype which is most clear in leaves, while a double defect in AtToc132 and AtToc120 leads to low import of housekeeping proteins and defects in non-photosynthetic tissues (Kubis et al., 2004). Similarly, loss of AtToc33 has a clear photosynthetic impact and reduces import of photosynthesis-associated proteins, while loss of the highly homologous AtToc34 does not reduce import of photosynthetic proteins, but instead affects root growth (Constan et al., 2004). Although the study of this specialization of import receptors has been carried out primarily in Arabidopsis, it is likely to be a general phenomenon, at least in angiosperms (Voigt et al., 2005). Even the outer envelope pore itself, Toc75, exists in two versions with different specificities (Baldwin et al., 2005).
Once in the stroma, routing of polypeptides destined to cross into the thylakoid lumen requires a second transit peptide, and uses two possible routes, both homologous to secretion/export pathways in bacteria. The Sec pathway translocates unfolded proteins using ATP hydrolysis (Schuenemann et al., 1999). The alternative, Tat pathway translocates fully folded proteins and uses the photosynthetic pH gradient as the source of energy (Jarvis and Robinson, 2004). Finally, among thylakoid membrane proteins, a few use components of the bacterial ‘signal recognition particle’ pathway for integration, but the majority are inserted without the assistance of any other known proteins, and so are thought to possess intrinsic biophysical properties that allow ‘spontaneous’ membrane insertion (Jarvis and Robinson, 2004).
Plastid division
Plastids cannot arise by de novo biogenesis due to the presence of their independent genome. Instead new plastids arise through the division of a pre-existing organelle. As with bacterial division, plastid division is easily observed under the microscope (López-Juez and Pyke, 2005). Recent years have seen dramatic progress in our understanding of plastid division as a result of successful forward and reverse genetic approaches (Osteryoung and Nunnari, 2003; Margolin, 2005; Fig. 2). A classical genetic approach (the isolation of plastid division mutants in Arabidopsis, and identification of the corresponding genes) was pioneered by Pyke and Leech (1992) and has yielded Arc3 (Shimada et al., 2004), Arc5 (Miyagishima et al., 2003), Arc6 (Vitha et al., 2003), and Crumpled Leaf (Asano et al., 2004). Meanwhile a reverse genetics approach (the identification of homologues in plants of bacterial division genes) was pioneered by Osteryoung and Vierling (1995) and has yielded FtsZ1, FtsZ2 (Vitha et al., 2001), and MinE (Itoh et al., 2001). A combined approach, in which a mutation was isolated and the mutant gene was identified by homology to prokaryotic division proteins, has yielded, arc11/MinD (Fujiwara et al., 2004), and GC1/AtSuIA (Maple et al., 2004; Raynaud et al., 2004).
|
Two plastid division rings have been visualized under the microscope and shown to be made of constituents other than those mentioned in the paragraph above (Miyagishima et al., 2001). The overall division process involves the sequential assembly at the envelopes of FtsZ1/2 (the Z ring) and the plastid division rings, in the mid-organelle plane of a plastid. FtsZ1 and 2 are GTPases able to form polymers, and are related to tubulin. The central assembly location of these rings is determined by the MinD and MinE proteins, which act to prevent ring formation at the tips of the organelle. These proteins act together (Aldridge and Møller, 2005). The assembly of the rings is aided by Arc6 and Arc3 (Shimada et al., 2004; Maple et al., 2005). Constriction of the three rings begins, in the case of the Z ring, by removal of subunits. Once a substantial constriction has taken place, Arc5 assembles into a further ring. Arc5 is a dynamin-related protein. Dynamins are involved in vesicle formation and generally membrane ‘pinching’ in eukaryotes, including mitochondrial division. In keeping with this proposed function, the ring of Arc5, on the outer envelope of chloroplasts, forms at the stage of membrane constriction (Gao et al., 2003; Miyagishima et al., 2003; Fig. 2).
The membrane constrictions necessary during plastid division are likely to generate increased hydrostatic pressure inside the plastids. By the laws of dimensionality, two daughter plastids carrying the same amount of envelope membrane as their progenitor will always contain a lower combined volume. While this had not been given consideration before, Haswell and Meyerowitz (2006) have recently examined the function of two Arabidopsis proteins, MSL2 and MSL3, that are homologous to bacterial mechanosensing ion channels and have chloroplast transit peptides. Both proteins localize in patches on the chloroplast envelope associated with the division protein MinE. Simultaneous disruption of MSL2 and MSL3 results in heterogenous plastid populations, with some plastids becoming enlarged and spherical-shaped. The authors interpreted their observations as consistent with a role for MSL2 and MSL3 in the relief of turgor associated with division.
Are the remaining plastid components shared between daughter organelles simply by a process of random segregation brought about by physical separation? In the case of chloroplasts, specific processes dividing the thylakoid complex into two halves during organelle division seemed to be revealed by mutants in the ARTEMIS gene. In such mutants, chloroplasts showed separate and distinct thylakoid systems within incompletely separated organelles (Fulgosi et al., 2002). The original ARTEMIS gene has recently been shown to be a mistaken merger of two separate open reading frames; one of them, Alb4 (Gerdes et al., 2006), encodes a member of the Alb3/Oxa1/YidC family of proteins, involved in the insertion of other proteins in membranes of chloroplasts, mitochondria, and bacteria. The reason for the phenotype of the original ARTEMIS mutant is not fully understood. In terms of the segregation of the nucleoids, in which the plastid genome copies are contained, they arrange themselves before division into a reticulum associated with the envelope. This is assisted by the PEND protein, which probably allows their equitable partition (Terasawa and Sato, 2005).
|
Plastid biogenesis: the challenges |
|---|
|
TopAbstractIntroductionPlastid biogenesis: the...Plastid biogenesis: the...References |
|---|
Do we know the majority of plastid proteins?
In spite of tremendous progress in recent years, and the availability of whole genome sequences, it is clear that we are nowhere near having a complete catalogue of the proteins that function in plastid biogenesis and, therefore, are far from a full list of plastid functions or developmental processes (Leister, 2003). There are two large areas of uncertainty: the list of plastid-localized proteins, and the catalogue of proteins necessary for plastid biogenesis, whether or not they are plastid localized.
Plastid-localized proteins have been catalogued on the basis of our understanding of plastid protein import. A number of algorithms predicting the presence of plastid transit peptides in user-supplied sequences have been developed, of which TargetP (Emanuelsson et al., 2000), which also predicts other subcellular destinations, appears to show the lowest rates of both false positive and false negative predictions. Use of TargetP on the full Arabidopsis genome led to a prediction of 3100 plastid-targeted proteins. Simultaneous use of a combination of algorithms (which decreases the number of false positive predictions, although it obviously increases the false negatives) predicts 2100 plastid-targeted proteins in Arabidopsis and, surprisingly, 4800 in rice, which has a similar total number of genes (Richly and Leister, 2004). Besides the uncertainty in bioinformatics predictions, a major caveat to the completeness of these lists has emerged from proteomics programmes, in which total proteins of diverse plastid types, or plastid fractions, are being detected by mass spectrometry methods (van Wijk, 2004, and references therein; Kleffmann et al., 2006). A major outcome is that a substantial number of proteins present in the outer chloroplast envelope, that often do not contain a plastid transit peptide, are predicted to be targeted to the secretory pathway. In the case of the homologous proteins OEP14 (pea) and OEP7 (Arabidopsis), it has been shown that the domain that TargetP wrongly identifies as the hydrophobic core of a signal peptide for the endomembrane system, in fact targets the protein to the outer chloroplast envelope and remains present, presumably as a transmembrane span, in the mature protein (Lee et al., 2001; Hofmann and Theg, 2005). The number of such proteins (predicted as targeted to the secretory pathway, but actually present in the chloroplast envelope) is larger than previously thought (Kleffmann et al., 2004). One further caveat to the use of bioinformatics to predict the final list of plastid proteins has been raised by the case of the Arabidopsis CAH1 carbonic anhydrase (Villarejo et al., 2005). This protein is correctly predicted to be targeted, through a cleavable signal peptide, to the secretory pathway, but only transiently while it is N-glycosylated on its way towards the chloroplast stroma, its final destination.
Do we know the majority of proteins/processes involved in plastid biogenesis?
The eventual catalogue of proteins required for plastid biogenesis, regardless of their location, shows an even greater uncertainty at present. A number of mutants detected on the basis of defects in chloroplast development have been analysed (López-Juez and Pyke, 2005). These include mutants defective in chloroplast and leaf mesophyll development, and variegated mutants (Sakamoto et al., 2003). A direct screen for defective chloroplast biogenesis (clb) mutants has been carried out by Gutiérrez-Nava et al. (2004), who searched among existing stock-centre collections for albino seedlings that do not result from photo-oxidative damage, and identified six such loci. The loci identified through these various mutant classes are involved in the assembly or function of the plastid genetic machinery (for example Pesaresi et al., 2006; Albrecht et al., 2006), including pentatricopeptide repeat proteins with RNA-binding roles (Lurin et al., 2004; Cushing et al., 2005; Gothandam et al., 2005), and the assembly or repair of protein complexes requiring proteolytic steps (Chen et al., 2005; Sakamoto, 2006). Novel roles for such classes of proteins probably remain to be uncovered, as shown by the role of a tetratricopeptide repeat protein in the regulation of transcription (Weber et al., 2006) or the possible signalling-related role of an intermembrane protease (Bolter et al., 2006). One class of mutations uncovered surprisingly frequently in screens for mutants with obvious chloroplast defects, affects genes involved in the methyl-erythritol 4-phosphate pathway of isoprenoid biosynthesis in plastids (Estévez et al., 2000; Rodríguez-Concepción and Boronat, 2002). Among the six CLB genes, the two thus far identified at the molecular level encode two enzymes in this pathway (Gutiérrez-Nava et al., 2004; Guevara-García et al., 2005). These mutations cause developing chloroplasts to arrest at the proplastid stage. The reason why defects in this pathway should be a frequent cause of albinism is unclear and, as mentioned, unrelated to loss of carotenoids and consequent photo-oxidative damage. Since both abscisic acid and gibberellins are isoprenoid derivatives, one possibility is the existence of yet-undiscovered hormonal control of chloroplast development.
Will these classes of mutations encompass the majority of the processes required for plastid biogenesis? This is probably far from the case. Evidence for this is provided by a large-scale mutant screen carried out by Budziszewski et al. (2001). They identified, in a population of 38 000 insertional mutant lines, over 500 seedling-lethal mutants. Interestingly, 
85% of the mutants exhibited albino, yellow or pale cotyledons, indicative of defective chloroplasts. For 23 mutants, 18 of which showed an albino or pale phenotype, the mutated gene was identified. Seven of those 18 genes encoded proteins whose function was either unknown or of unsuspected plastid relevance (like pyrimidine permease or Zn-finger protein). A simple extrapolation, ignoring the possibility of multiple alleles at single loci, would estimate 165 genes of unsuspected plastid-related function to be detected by this mutant screen alone. This gives an indication of the scale to which our understanding of plastid biogenesis is incomplete. A much smaller screen, for defects in expression of the nuclear photosynthetic gene Lhcb (cue mutants; Li et al., 1995; López-Juez et al., 1998; Vinti et al., 2005), has identified seven mutant loci causing defective plastids. While CUE1 encodes a plastid envelope transporter (Streatfield et al., 1999; Voll et al., 2003), progress towards the cloning of CUE8 has narrowed it to a region of chromosome 5 containing 20 genes, none of which was previously suspected to play a role in plastid biogenesis or function (D Maffei, JR Bowyer, E López-Juez, unpublished results).
The evidence above suggests that, as has been proposed (Leister, 2003), large-scale screens for plastid biogenesis-defective mutants are still justified. One systematic existing approach is the maize Photosynthetic Mutant Library (http://chloroplast.uoregon.edu). Such genetic approaches, however, would still only uncover genes of non-redundant function. Clearly we are still a long way from a comprehensive catalogue of plastid biogenesis genes.
How are plastids integrated into the plant developmental programmes?
In spite of many years of observations, our knowledge about the make-up and development of different types of plastids, other than chloroplasts, is still limited (Neuhaus and Emes, 2000; Waters and Pyke, 2004; Kleffmann et al., 2006). Figure 3 shows plastids visualized using a RecA transit peptide-targeted red fluorescent protein (Haswell and Meyerowitz, 2006). An Arabidopsis seedling probably contains plastids with many different functions (photosynthetic, large chloroplasts in leaf and cotyledon mesophyll and in hypocotyl cortex, smaller plastids in epidermis and in roots), but also different cells of an Arabidopsis seedling are occupied by plastids to very different extents. For example, in cotyledons, epidermal cells have a very small plastid complement compared with mesophyll cells, and in roots the pericycle cells have a larger complement than cortical or epidermal cells (Fig. 3). Interestingly, a gene expression atlas of the Arabidopsis root has identified a substantial level of expression of photosynthesis-related genes in pericycle cells (Birnbaum et al., 2003). Clearly both ‘plastid type’ and ‘cellular plastid compartment size’ are characteristic of the cell type, and its stage of differentiation.
|
Very little work is being carried out to understand the mechanisms and possible signalling pathways presumably underlying these plastid differentiation responses, or the relative contributions to them of the nucleus and the plastid. One recent insight has come from the analysis of plastid protein import mechanisms. As described earlier, different isoforms of components of the Toc complex have different specificities, and as a result it is likely that separate Toc complexes primarily import either photosynthetic or non-photosynthetic, housekeeping proteins (Kessler and Schnell, 2006). This has been confirmed directly through an analysis of knock-out mutants of individual isoforms. One unexpected observation was the correlative change in protein import and in expression of the corresponding gene. For example AtToc33 or AtToc159 mutants show defects both in the rate of import of photosynthetic proteins and in the expression of photosynthesis-related nuclear genes (Kubis et al., 2003, 2004). By contrast AtToc34 or AtToc132/120 mutants display defects in the import of non-photosynthetic proteins and in the expression of their genes (Constan et al., 2004; Kubis et al., 2004). This brings about a positive feedback mechanism that could potentially play an important role in plastid differentiation. For example, a cell type expressing low levels of Toc33 and high levels of Toc34, would import photosynthetic proteins with low efficiency and eventually, as a consequence, would express low levels of photosynthesis-related genes, which would in turn lead to the development of non-photosynthetic plastids.
The extent of the integration of plastids into the biology of the cell is manifest in the roles played by plastids in whole plant development and phenotype. Many lines of evidence support these roles, but the underlying reasons are very poorly understood. Inhibition of plastid translation in sectors of chimeric plants (Ahlert et al., 2003) causes developmental arrest, including an inhibition of cell division, in the sectors affected. The same consequences follow the targeted knockout of essential plastome genes, clpP1 and AccD (Kuroda and Maliga, 2003; Kode et al., 2005). Global blockage of plastid translation can be achieved through mutations in non-redundant aminoacyl-tRNA synthetases, and this (when observed in segregating populations of heterozygous plants) leads to arrests at the middle, transition stage of embryogenesis (Berg et al., 2005). The import of nuclear-encoded proteins, including the RNA polymerase, is essential even earlier for embryo survival, because a defect in the outer envelope main pore component, AtToc75-III, causes arrest of embryogenesis at the two-cell stage (Baldwin et al., 2005).
Besides being essential for cellular development, organellar genomes (plastids and mitochondria) can, in spite of their high degree of conservation and slow evolution, contribute substantially to plant phenotype later in development. This is shown by lines of maize bearing cytoplasms of distantly-related teosinte species (Allen, 2005). How such influence takes place, and earlier in development what the nature of the essential plastid function is (it is unlikely to be simply the loss of essential metabolites; Ahlert et al., 2003; Gutiérrez-Nava et al., 2004) remains unknown.
How do plastids respond to their cell's circumstances?
Two plastid types of major crop and biotechnological importance are starch-storing amyloplasts (Andon et al., 2002; Balmer et al., 2006) and carotenoid-storing chromoplasts (Bramley, 2002). In both cases their differentiation is under endogenous, hormonal control. For example, chromoplasts in tomato fruit differentiate from chloroplasts during fruit ripening, which is partially ethylene-dependent (Alba et al., 2005). Ethylene must play a role in chromoplast differentiation, as judged from the phenotype of the Never Ripe ethylene receptor mutant (Wilkinson et al., 1995). However, there is also evidence that plastid-derived signals are involved in the intriguing plastid conversion circuitry, perhaps in a feedback loop. For example, a chloroplast-localized heat shock protein is required for the transition from chloroplasts to chromoplasts, and elevating its expression accelerates the conversion (Neta-Sharir et al., 2005). A role for the carotenoid accumulated in chromoplasts, lycopene, as a signal has been postulated (Barr et al., 2004).
One further aspect of the integration of plastids within their host cells is the ability to respond to environmental signals that the cell responds to. Perhaps the clearest example of this is the biogenesis of chloroplasts in photosynthetic cells only in the light, and the continued fine-tuning of chloroplast composition to the prevailing conditions. Evidence described earlier demonstrated the way in which cytosolic photoreceptors can control the expression of plastid genes, by regulating the nuclear expression of a plastid-targeted sigma-70 factor. The rapid biogenesis of chloroplasts is a major response during seedling establishment, and this is reflected by the fact that the first dark-to-light transition of seedlings (the ‘de-etiolation’ process) brings about massive gene expression reprogramming (Ma et al., 2001; Tepperman et al., 2001), and around 50% of the genes eventually affected encode plastid-targeted proteins. Some evidence supports the notion that the subset of chloroplast development responses to light is, to an extent, separable from other light responses. For example, defects or mutations in plastid biogenesis affect the light response of photosynthesis-related genes specifically (López-Juez et al., 1996, 1998; Vinti et al., 2005). However, a screen to identify components of the pathway of light control of chloroplast development has been performed based on reporting Lhcb expression in a light perception-deficient background (Hills, 2002; A Hills, T Shindo, Y Niwa, E López-Juez, unpublished results) and the two mutants that have been identified to date have defects in both photosynthesis-related genes and de-etiolation in general.
When/how are plastid growth and division controlled?
Given that a specific plastid subtype (the proplastid) has evolved to facilitate inheritance of plastids, it is reasonable to presume that instead of plastid inheritance being left to chance, mechanisms to co-ordinate plastid and cell division would also have evolved. This, indeed, appears the case, but the mechanisms remain very poorly understood. In Cyanidioschyzon, the ‘minimal’ red alga, the connection between cell and organelle division is tight and inescapable, since each cell has a single mitochondrion and a single plastid (Nishida et al., 2005), although inhibition of DNA synthesis, surprisingly, acts as a signal for multiple chloroplast divisions. In plants the situation is necessarily more complex, as cell growth and division are not obligately coupled, and because different cells carry different plastid complements (Fig. 3). It has long been known that in cereal leaves (Hashimoto and Possingham, 1989; Baumgartner et al., 1989) plastid proliferation and plastome replication take place rapidly in post-meristematic cells. In young dicot leaves, light that causes chloroplast differentiation but not full cell enlargement, activates plastid DNA replication, but not an increase in plastid numbers (DuBell and Mullet, 1995). This suggests separate links between differentiation and plastid growth, and between cell enlargement and plastid division.
Mesophyll cells observed under a microscope show a remarkably constant degree of occupancy by chloroplasts, even when cell size changes markedly, for example, as a response to the environment (Weston et al., 2000). A strict correlation appears to exist between the size of the cell and that of its plastid compartment (Pyke, 1997), even when comparing mesophyll cells of multiple species (Pyke, 1999) or examining cells vastly enlarged as a result of a prevention of mitosis (Jasinski et al., 2003). Obviously mechanisms for sensing plastid density, controlling both plastid growth and division, would explain this behaviour, but the nature and action of such mechanisms remain poorly understood. One suggestion put forward by Pyke (2006) is that MscS-like (MSL) proteins, the mechanosensing plastid envelope ion channels (Haswell and Meyerowitz, 2006), play a role in such density-sensing, at least in mesophyll cells, where chloroplasts occupy the cytoplasm in one layer until they can touch, even compress each other. Such mechanosensors could play a similar role in other cells with a low density of plastids, if they acted in stromules, the tubules that protrude from plastids more frequently in such cell types (Pyke, 2006). If this model were to hold true, it would need to link plastid/cell size ratios to plastid growth, because the constancy of the chloroplast compartment can be seen even when plastid division is prevented, for example, in the arc6 mutant (Pyke, 1999). As to the processes underlying chloroplast growth, photosynthetic components obviously accumulate, but the increase in plastid genome copy number seems to precede this accumulation (Baumgartner et al., 1989). Importantly, and perhaps unsurprisingly, an increase in plastid ribosomes and the capacity for protein synthesis also precedes the build-up of the photosynthetic apparatus (Harrak et al., 1995). This may be of importance, since genes involved in plastid protein synthesis may play regulatory and rate-limiting roles in photosynthetic development (Pesaresi et al., 2006). It is interesting that genes for cytoplasmic ribosomal and translation-related proteins seem to play important, equivalent roles in cellular growth (Li et al., 2005). Perhaps a greater understanding of the regulation of the synthesis of the plastid translation machinery would help to uncover fundamental mechanisms of plastid growth control.
Links between plastid and cell division also exist, even though uncoupling of these is obviously possible (Jasinski et al., 2003). Multiplication of plastids can be triggered by cytokinin (but not auxin) in dark-grown cotyledons, and under those conditions a simultaneous increase in FtsZ expression can be observed (Ullanat and Jayabaskaran, 2002). FtsZ expression also mimics that of a cell cycle-associated gene in cell cultures (El-Shami et al., 2002). AtCDT1a and AtCDT1b are genes encoding related forms of a key component of the complex of proteins needed to initiate nuclear DNA replication (the pre-replication complex). Interestingly AtCDT1a contains sequence for a predicted chloroplast transit peptide, fusion proteins with a green fluorescent protein-reporter are targeted to both nuclei and plastids, and a double mutant for both CDT1 genes shows, besides cell cycle-related defects, pale leaves whose mesophyll cells contain large, unevenly divided chloroplasts (Raynaud et al., 2005). Thus AtCDT1 proteins are strong candidate constituents of the link between plastid and cell division.
Can the size of the plastid compartment be altered?
Given the roles of plastids as subcellular factories, the ability to manipulate the size of their cellular complement could have important implications for the biology of the plant and for its biotechnological exploitation. Mesophyll cells appear to have reached close to a maximum in chloroplast occupancy, but in other cell types increases in the plastid complement are physically possible. It has been seen earlier that both chloroplast biogenesis and leaf development are tightly regulated by light signalling pathways. A number of proteins are known to play positive and negative regulatory roles in light signalling, as revealed primarily by the phenotype of defective mutants. While genetic studies of light signalling (Schäfer and Bowler, 2002) have focused on seedling morphology in Arabidopsis, mutant tomato plants have been particularly useful in uncovering plastid-related phenotypes. In tomato, a high pigment-2 mutant (hp-2), with fruit enriched in the chromoplast pigment lycopene, is defective in LsDET1, the orthologue of Arabidopsis light-signalling negative regulator DET1 (Mustilli et al., 1999). Another mutant with a similar fruit-pigment phenotype, hp-1, shows an elevated plastid DNA copy number (Yen et al., 1997), and displays fruit pericarp cells with a small but significant increase in plastid cell index (the ratio between plastid and cell projected areas) of between 13% and 29% (Cookson et al., 2003). A greater increase in plastid cell index in hp-1 occurs in leaf palisade mesophyll cells, whose morphology resembles that observed in high light-grown leaves (Weston et al., 2000). hp-1 is defective in the tomato UV-damaged DNA-binding protein 1, LsDDB1, gene (Liu et al., 2004). DDB1a and DET1 physically interact and act together in Arabidopsis (Schroeder et al., 2002). It appears, therefore, that exaggerated light signalling causes increased fruit plastid pigment content (Liu et al., 2004) and increases in the cellular plastid complement which, although small, are significant as they demonstrate that such plastid complement is not developmentally fixed, since it can be influenced by environmental signals.
Are there chloroplast master controllers?
The radically different phenotype of plastids in photosynthetic compared with non-photosynthetic tissues could most easily be explained if mechanisms existed that simultaneously controlled a large number of genes encoding chloroplast proteins. Arguably, however, plastids play central roles, other than in photosynthesis, in the biology of specific cell types. A biologically economical way of building such plastids would be to have co-ordinated control of genes encoding plastid proteins, irrespective of whether they are directly related to photosynthesis. In other words, the existence of both ‘chloroplast-specific’ and ‘plastid-generic’ master switches could be anticipated. Evidence has emerged that is consistent with the existence of both kinds of centralized regulation of gene expression.
A common phenotype of the Arabidopsis det1 and the tomato hp-2 mutants is the partial development of chloroplasts in cells or under conditions where they would not normally appear. This includes semi-developed chloroplasts instead of etioplasts in dark-grown cotyledon cells (Mustilli et al., 1999) but also, crucially, instead of proplastids in root cells (Chory and Peto, 1990; Mustilli et al., 1999). One possible interpretation is that control elements for the transcription of genes involved in chloroplast biogenesis are common to both light and tissue-specific signals. Several photosynthesis-related gene promoters have been extensively analysed for the presence of light-regulated elements. Different versions of one such element, called G-box, are common and have been shown to, in combination with other elements, confer both light responsiveness and plastid-dependent expression (Puente et al., 1996). A short, single element (CMA5) in the tobacco RbcS8B promoter is sufficient to confer light-dependent expression to a reporter gene in Arabidopsis (Martínez-Hernández et al., 2002). Importantly, the reporter gene is expressed in leaf, but not in root tissue, and disappears from leaves when chloroplasts are photodamaged by treatment of seedlings with norflurazon. Again this suggests the existence of photosynthetic gene-regulatory mechanisms shared by light and tissue-specific signals, and by signals of plastid-to-nucleus communication. Similar commonality has been found in the regulatory signals of other photosynthesis-related genes, even in cases where they act at the post-transcriptional rather than transcriptional level (Helliwell et al., 1997; Sullivan and Gray, 2002).
Large-scale gene expression profiling experiments have also provided evidence in this direction. Richly et al. (2003) used a custom-made array of gene-sequence tags to examine the expression of 3292 genes, greatly enriched for those encoding chloroplast-targeted proteins. They monitored changes in gene expression caused by 35 different environmental or genetic conditions. They observed three different broad classes of regulation: in two (their classes one and three) the majority of genes were up- or down-regulated, respectively. Conditions that caused a class one response included high light, the gun1 and gun5 mutations that lead to expression of photosynthesis-related nuclear genes in the absence of functional plastids (Susek et al., 1993; Mochizuki et al., 2001), and the ppi1 mutation (causing reduced plastid protein import; Kubis et al., 2003). Conditions that caused a class three response included the cue1 mutation, which causes defective plastid biogenesis, and loss of Lhcb1 expression (Streatfield et al., 1999). Richly and collaborators interpreted their data as evidence for the existence of a ‘master switch’. This acted in a binary mode (on in class one/off in class three), and controlled genes for plastid-targeted proteins (regardless of their relation to photosynthesis, as proved by the ‘on’ condition in the ppi1 mutant). An extension of this analysis to a total of 101 conditions (Biehl et al., 2005) provided further insights: a total of 23 distinct clusters of co-expressed genes, known as regulons, could be identified. Two of those clusters showed the tightest co-regulation, and contained primarily genes for either the photosynthetic light reactions or for plastid protein synthesis (the largest number being chloroplast ribosomal proteins). These two clusters escaped the previous ‘plastid master switch’ model, but they themselves, in turn, on the basis of their tight co-regulation, uncovered a second master switch, this one specific for ‘chloroplast’ function.
Evidence suggesting the existence of large-scale co-regulation of genes for chloroplast biogenesis has also been obtained. Global gene expression changes taking place specifically in shoot apical meristematic regions, and separately in cotyledons when etiolated Arabidopsis seedlings are first exposed to light, have been examined (E Dillon, G Beemster, L Bögre, E. López-Juez, unpublished data). Broad clusters of genes showing co-ordinated expression were identified, and in a small number of these clusters genes associated with ‘plastid’, ‘chloroplast’, or ‘thylakoid’ descriptions (gene ontology terms) were statistically over-represented. The small number of clusters and relatedness of expression pattern within each of them was such that just a few upstream regulatory factors could, in theory, be responsible for the co-ordinated induction of a large number of chloroplast function-related genes.
The search for plastid master switches
What could the nature of such master switches be? Genetic approaches have uncovered a number of genes worth exploring as candidates. Not surprisingly, studies of light signalling provide several of them. The Arabidopsis hy5 mutant shows limited and delayed responses to light, including reduced expression of photosynthesis-related genes (McCormac and Terry, 2002). While the reduction is not dramatic, the fact that HY5 encodes a G-box binding bZIP transcription factor makes it of significance. Maize Golden 2 was first identified as a bundle-sheath specific chloroplast defective mutant, but it was later found that both maize and Arabidopsis contain two Golden 2-like (GLK) genes, and when both are simultaneously mutated, defective chloroplasts, showing a severe reduction in thylakoids, are seen in every photosynthetic cell (Fitter et al., 2002). GLK genes are attractive candidates for being regulatory switches for ‘chloroplasts’ in photosynthetic cells because they encode transcription factors conserved in all lineages of land plants, including bryophytes, but not in photosynthetic single-cell organisms (Yasumura et al., 2005). However, their role cannot be fundamental since the chloroplast development defect in double mutants is, again, not strong. Transcription co-activators may be as important as transcription factors in controlling or co-ordinating states of activity of transcriptional complexes. The Arabidopsis HAF2 gene encodes a protein homologous to TATA box-binding factors in yeast, and has histone acetyl-transferase activity (Bertrand et al., 2005). Interestingly, HAF2 knockouts show an obvious ‘slow greening’ phenotype, similar to the one in virescent cue mutants (see above), and reduced/delayed expression of photosynthesis-related genes.
Overall, then, there are some leads towards the uncovering of ‘chloroplast master switches’, but they are far from having been identified. Recently, an elegant genetic screen was designed in an attempt to identify chloroplast regulatory factors directly (Niwa et al., 2006). Kobayashi, Niwa and collaborators generated activation-tagged mutations on callus tissue that had been previously transformed to express a herbicide detoxifying activity under the control of a photosynthetic gene promoter. Normally such promoters are silent in colourless callus, but the activation tagging procedure can, potentially, turn them active and cause herbicide tolerance. In this way the ces (callus expression of RbcS) 101 mutant was successfully identified. The nature of the CES101 gene, a receptor-like kinase, raised the possibility of proteins involved in the sensing of extracellular (potentially intercellular-communication) signals being critical activators of chloroplast biogenesis. What such signals could be is currently in the realm of speculation, but both the results of this work and its experimental approach seem highly promising.
Towards plastid systems biology
Successfully uncovering such master regulatory genes would not be the end of the road. Such ‘switches’ would need to be integrated into overall networks that together underlie the growth and development of plants. It is known that such integration is necessary, and phenomena that have been described earlier, such as the existence of plastid ‘developmental signals’, are either mechanisms or manifestations of the fact that the integration does take place. Already over 15 years ago it was postulated that chloroplast biogenesis could be nothing but a manifestation of a ‘leaf mesophyll cell’ initiation programme, with chloroplast and leaf development activities tightly intertwined (Chory and Peto, 1990).
A powerful method to uncover co-ordination of wide biological processes has become available with the ability to monitor, through microarray experiments, global gene expression. Such a technique is being used to address many separate environmental or developmental situations, in many different laboratories, but through only a small number of technology platforms, making it possible to compare results from hundreds, even thousands of experiments, at least in the Arabidopsis model. Bioinformatic tools to monitor such data are being developed and refined (Zimmerman et al., 2004; Toufighi et al., 2005). Figure 4 shows a comparison of the expression profiles of a number of key genes playing a role or potentially regulating aspects of plastid biology referred to earlier, across Arabidopsis individual tissues or developmental stages. The comparison uses Expression Browser (Toufighi et al., 2005). It is clear that genes for processes biochemically very distinct, such as plastid division, plastid gene expression, and expression of nuclear photosynthetic genes, show very tight co-regulation patterns, while within a single process, such as plastid protein import, homologous components clearly associate with distinct process (see Toc33 or Toc159 association with FtsZ1, MinD and Arc6, a subset of chloroplast division genes, while Toc34 or Toc 132 associate much more closely with mitotic and DNA-synthesis cyclins and Histone 2A, cell division genes of high activity in both shoot and root apical meristems). The associations may reflect the sharing of transcriptional regulatory mechanisms. While many key transcription factors playing roles in embryo development in Xenopus have been known for a number of years, it is only now that networks that link them together into biological processes are being drawn (Longabaugh et al., 2005). Only such interlinking networks can achieve full descriptive and predictive power, the ultimate aim of ‘systems biology’.
|
We are clearly in the infancy of a ‘plastid systems biology’ era. Our understanding of how diverse cells differentiate distinct plastid types, particularly chloroplasts, of the correct type, to an extent that constitutes an appropriately-sized cellular plastid complement, is only beginning. Such an understanding could bring obvious rewards (for example, to manipulate the size of the plant's ‘plastid factory’), and can only be expected to expand dramatically in the next few years.
|
Acknowledgements |
|---|
I am grateful to Edyta Dillon, Gerrit Beemster, Laszlo Bögre, Douglas Maffei, and John R Bowyer for allowing the inclusion of interpretations of data prior to publication, to International Journal of Development Biology for permission to reproduce, with modifications, Figs 1 and 2, and to Elizabeth Haswell for kindly providing the RecARed fluorescent plastid reporter line used to generate Fig. 3. Work in the author's laboratory has been or is funded by BBSRC grants GO7782, PO7790, C19322 [GenBank] , and studentship C125.
This manuscript is dedicated to the memory of Professor JR Bowyer (1955–2006), exceptional scientist and outstanding human being, whose discussions and insights were a continuous source of help and inspiration.
|
Abbreviations |
|---|
NEP, nuclear-encoded RNA polymerase; PEP, plastid-encoded RNA polymerase; PS, photosystem; Tic, translocon of the inner membrane of chloroplast; Toc, translocon of the outer envelope of chloroplast.
|
References |
|---|
|
TopAbstractIntroductionPlastid biogenesis: the...Plastid biogenesis: the...References |
|---|
Ahlert D, Ruf S, Bock R. (2003) Plastid protein synthesis is required for plant development in tobacco. Proceedings of the National Academy of Sciences, USA 100 15730–15735.
Alba R, Payton P, Fei Z, McQuinn R, Debbie P, Martin GB, Tanksley SD, Giovannoni JJ. (2005) Transcriptome and selected metabolite analyses reveal multiple points of ethylene control during tomato fruit development. The Plant Cell 17 2954–2965.
Albrecht V, Ingenfeld A, Apel K. (2006) Characterization of the Snowy cotyledon 1 mutant of Arabidopsis thaliana: the impact of chloroplast elongation factor G on chloroplast development and plant vitality. Plant Molecular Biology 60 507–518.[CrossRef][ISI][Medline]
Aldridge C and Møller SG. (2005) The plastid division protein AtMinD1 is a Ca2+-ATPase stimulated by AtMinE1. Journal of Biological Chemistry 280 31673–31678.
Allen JO. (2005) Effect of teosinte cytoplasmic genomes on maize phenotype. Genetics 169 863–880.
Andon NL, Hollingworth S, Koller A, Greenland AJ, Yates JR 3rd, Haynes PA. (2002) Proteomic characterization of wheat amyloplasts using identification of proteins by tandem mass spectrometry. Proteomics 2 1156–1168.[CrossRef][ISI][Medline]
Asano T, Yoshioka Y, Kurei S, Sakamoto W, Sodmergen W, Machida Y. (2004) A mutation of the CRUMPLED LEAF gene that encodes a protein localized in the outer envelope membrane of plastids affects the pattern of cell division, cell differentiation, and plastid division in Arabidopsis. The Plant Journal 38 448–459.[CrossRef][ISI][Medline]
Baba K, Schmidt J, Espinoso-Ruiz A, Villarejo A, Shiina T, Gardestrom P, Sane AP, Bahlerao RP. (2004) Organellar gene transcription and early seedling development are affected in the rpoT;2 mutant of Arabidopsis. The Plant Journal 38 38–48.[CrossRef][ISI][Medline]
Baldwin A, Wardle A, Patel R, Dudley P, Park SK, Twell D, Inoue K, Jarvis P. (2005) A molecular-genetic study of the Arabidopsis Toc75 family. Plant Physiology 138 715–733.
Balmer Y, Vensel WH, Dupont FM, Buchanan BB, Hurkman WJ. (2006) Proteome of amyloplasts isolated from developing wheat endosperm presents evidence of broad metabolic capability. Journal of Experimental Botany 57 1591–1602.
Barr J, White WS, Chel L, Bae H, Rodermel S. (2004) The GHOST terminal oxidase regulates developmental programming in tomato fruit. Plant, Cell and Environment 27 840–852.[Medline]
Baumgartner BJ, Rapp JC, Mullet JE. (1989) Plastid transcription activity and DNA copy number increase early in barley chloroplast development. Plant Physiology 89 1011–1018.
Bedard J and Jarvis P. (2005) Recognition and envelope translocation of chloroplast preproteins. Journal of Experimental Botany 56 2287–2320.
Berg M, Rogers R, Muralla R, Meinke D. (2005) Requirement of aminoacyl-tRNA synthetases for gametogenesis and embryo development in Arabidopsis. The Plant Journal 44 866–878.[CrossRef][ISI][Medline]
Bertrand C, Benhamed M, Li YF, Ayadi M, Lemonnier G, Renou JP, Delarue M, Zhou DX. (2005) Arabidopsis HAF2 gene encoding TATA-binding protein (TBP)-associated factor TAF1, is required to integrate light signals to regulate gene expression and growth. Journal of Biological Chemistry 280 1465–1473.
Biehl A, Richly E, Noutsos C, Salamini F, Leister D. (2005) Analysis of 101 nuclear transcriptomes reveals 23 distinct regulons and their relationship to metabolism, chromosomal gene distribution and co-ordination of nuclear and plastid gene expression. Gene 344 33–41.[CrossRef][ISI][Medline]
Birnbaum K, Shasha DE, Wang JY, Jung JW, Lambert GM, Galbraith DW, Benfey PN. (2003) A gene expression map of the Arabidopsis root. Science 302 1956–1960.
Bolter B, Nada A, Fulgosi H, Soll J. (2006) A chloroplastic inner envelope membrane protease is essential for plant development. FEBS Letters 580 789–794.[CrossRef][ISI][Medline]
Bramley PM. (2002) Regulation of carotenoid formation during tomato fruit ripening and development. Journal of Experimental Botany 53 2107–2113.
Budziszewski GJ, Lewis SP, Glover LW, et al. (2001) Arabidopsis genes essential for seedling viability: isolation of insertional mutants and molecular cloning. Genetics 159 1765–1778.
Chen G, Bi YR, Li N. (2005) EGY1 encodes a membrane-associated and ATP-independent metalloprotease that is required for chloroplast development. The Plant Journal 41 364–375.[CrossRef][ISI][Medline]
Chory J and Peto CA. (1990) Mutations in the DET1 gene affect cell-type-specific expression of light-regulated genes and chloroplast development in Arabidopsis. Proceedings of the National Academy of Sciences, USA 87 8776–8880.
Chou ML, Fitzpatrick LM, Tu SL, Budziszewski G, Potter-Lewis S, Akita M, Levin JZ, Keegstra K, Li H-M. (2003) Tic40, a membrane-anchored co-chaperone homolog in the chloroplast protein translocon. The EMBO Journal 22 2970–2980.[CrossRef][ISI][Medline]
Constan D, Patel R, Keegstra K, Jarvis P. (2004) An outer envelope membrane component of the plastid protein import apparatus plays an essential role in Arabidopsis. The Plant Journal 38 93–106.[CrossRef][ISI][Medline]
Cookson PJ, Kiano JW, Fraser PD, Romer S, Schuch W, Bramley PM, Pyke KA. (2003) Increases in cell elongation, plastid compartment size and phytoene synthase activity underlie the phenotype of the high pigment-1 mutant of tomato. Planta 217 896–903.[CrossRef][ISI][Medline]
Cushing DA, Forsthoefel NR, Gestaut DR, Vernon DM. (2005) Arabidopsis emb175 and other ppr knockout mutants reveal essential roles for pentatricopeptide repeat (PPR) proteins in plant embryogenesis. Planta 221 424–436.[CrossRef][ISI][Medline]
De las Rivas J, Lozano JJ, Ortiz AR. (2002) Comparative analysis of chloroplast genomes: functional annotation, genome-based phylogeny and deduced evolutionary patterns. Genome Research 12 567–583.
DuBell AN and Mullet JE. (1995) Continuous far-red light activates plastid DNA-synthesis in pea leaves but not full cell enlargement or an increase in plastid number per cell. Plant Physiology 109 95–103.
Dyall SD, Brown MT, Johnson PJ. (2004) Ancient invasions: from