Jarvis 2013

Chloroplasts contain the green pigment chlorophylland are responsible for the light-powered reactions of photo-synthesis, upon which essentially all life depends1,2. They are the prototypical members of a diverse family of orga-nelles, the plastids, which are found ubiquitously in plants and various algae3,4. In plants, other plastid family mem-bers are the starch-rich amyloplasts found in seeds, roots and tubers, which have important roles in energy storage and gravitropism, and chromoplasts, which accumulate carotenoid pigments and function as attractants in flow-ers and fruits5,6. Proplastids are small, undiffer entiated plastids that are found in meristem s and in reproductive tissues. Like amyloplasts and chromoplasts, proplastids have specific functions; these include organelle trans-mission between generations (usually through female gametes, but sometimes also via polle n) and to all cells of the organism. The more specialized plastids in plants are derived from proplastids through differentiation and all are bounded by a double-membran e system that is called the envelope5,6 (FIG.1).

Depending on the developmental context, plastids fulfil a diversity of important roles in addition to photo-synthesis. They are involved in the synthesis of amino acids, fatty acids, purine and pyrimidine bases, terpe-noids and various pigments and hormones, as well as in key aspects of nitrogen and sulphur assimilation5,7. The metabolic activities of plastids are fully integrated with those of other cellular compartments and to allow for this, the envelope contains numerous metabolite transporters8. Interorganellar cooperation that involves plastids is crucial for lipid synthesis, photorespiration and various other processes. The photosynthetic and

metabolic functions of plastids have been reviewed else-where1,7,8, and therefore we focus on aspects of organellar development and maintenance.

Similarly to mitochondria, plastids entered the eukary-otic lineage through endosymbiosis (~11.5billio n years ago). They are the descendants of an ancient photo synthetic prokaryote that is similar to extant cyano bacteria4,9. The contemporary organelle retains a functional genome (the plastome), but this genome has undergone substantial reduction during evolution and typically carries fewer than 100 protein-coding genes10. Most of the ~20003000 different proteins that are found in plastids are encoded in the nucleus and must be imported following synthesis in the cytosol6,11. In this Review, we present recent developments that have shed new light on how the nuclear and organellar genomes are coordinated to ensure the proper development and main-tenance of plastids during plant growth. We also discuss recent insights into organellar protein transport systems and their regulation, as well as the mechanisms under-lying the propagation of plastids through organellar divi-sion. The integration of plastid biogenesis into organismal developmental programmes is also discussed.

Synthesis of proteins in two compartments Chloroplasts are the ancestral (and arguably most important) plastid type, and their biogenesis has received the greatest attention. The dual localization of plastid polypeptide synthesis (that is, in the nucleo cytosolic compartment and the organelle itself) renders organel-lar assembly complex, necessitating the coordination of two genomes (FIG.2). As most plastid proteins are

1Department of Plant Sciences, University of Oxford, South Parks Road, Oxford OX1 3RB, UK.2School of Biological Sciences, Royal Holloway University of London, Egham, Surrey TW20 0EX, UK.Correspondence to P.J. e-mail: [email protected]:10.1038/nrm3702Corrected online 14 January 2014

MeristemsGroups of stem cell-like, undifferentiated cells, that are found particularly at the tips of shoots and roots. All other plant cells derive from the meristems.

Biogenesis and homeostasis of chloroplasts and other plastidsPaul Jarvis1 and Enrique Lpez-Juez2

Abstract | Chloroplasts are the organelles that define plants, and they are responsible for photosynthesis as well as numerous other functions. They are the ancestral members of a family of organelles known as plastids. Plastids are remarkably dynamic, existing in strikingly different forms that interconvert in response to developmental or environmental cues. Thegenetic system of this organelle and its coordination with the nucleocytosolic system, the import and routing of nucleus-encoded proteins, as well as organellar division all contribute to the biogenesis and homeostasis of plastids. They are controlled by the ubiquitinproteasome system, which is part of a network of regulatory mechanisms that integrate plastid development into broader programmes of cellular and organismal development.

R E V I E W S

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 14 | DECEMBER 2013 | 787

F O C U S O N O R g a N E l l E b I O g E N E S I S a N d h O m E O S ta S I S

2014 Macmillan Publishers Limited. All rights reserved

20 m

Nature Reviews | Molecular Cell Biology

1 m

1 cm

20 m

ba

c

1 cmStromaThylakoid

Plastoglobule

Plastoglobule

Envelope

Starchgrain

Chloroplast

Gerontoplast Chromoplast

Developingautophagosome

Plastoglobule

Amyloplast

Proplastid

Etioplast

Elaioplast

Carotenoidcrystal

PLB

Vesicle

Partiallydisassembledthylakoid

nucleus-encoded, the bulk of information flow during chloroplast biogenesis proceeds from the nucleus to the organelle.

The role of the nucleocytosolic compartment. Chloroplast assembly requires the nucleocytosolic synthesis of photo-synthetic complex components, the Calvin cycle and other primary metabolic enzymes, envelope transporters

(including those that mediate the import of nucleus-encoded polypeptides) and proteins involved in homeo-stasis of the above1 (FIG.2). Photosynthetic complexes assemble around plastid-encoded core components, such as the products of psaA and psaB for photo system I (PSI) and psbA and psbD for PSII, but even the expression of these components depends on nucleus-encoded factors. The proteomes of other plastid types12 are distinct, the

Figure 1 | Diversity of plastid forms and their interconversions. a | Plastids exist in different forms, and the identity andabundance of each are controlled by developmental and environmental cues. Different types interconvert (see the arrows) following reorganization of the organellar proteome, a process that is controlled by the differentially regulated import of nucleus-encoded proteins. Chloroplasts are photosynthetic plastids, amyloplasts are starch-storing plastids, chromoplasts are carotenoid pigment-accumulating plastids, and proplastids are undifferentiated plastids that can differentiate into the different types of plastids. Etioplasts are chloroplast progenitors that form in darkness and accumulate chlorophyll precursors (in paracrystalline membranous structures called prolamellar bodies (PLB)) that are ready for rapid differentiation upon illumination. Elaioplasts store lipids in lipid droplets known as plastoglobules and exist, for example, in tapetal cells during pollen development. Gerontoplasts form during senescence, owing to resource recycling through the disassembly of the photosynthetic machinery and autophagy. b | Sequential proplastid-to-chloro-plast development visualized by transmission electron microscopy in an Arabidopsis thaliana virescent mutant (in which this process is delayed). Small proplastids in leaf primordia (for a size comparison, see the mitochondrion indicated with a black arrow) possess very few internal thylakoid membranes. As leaf cells differentiate (from the leaf tip to basal margins), chloroplasts develop and the thylakoids form granal stacks. A fully developed A.thaliana mesophyll cell (viewed using Nomarski optics) is also shown (bottom right). c | Sequential proplastid-to-chloroplast development in wheat mesophyll cells. Along the length of the leaf, a dramatic increase in plastid number and size accompanies chloroplast differentiation. The image in part b is reproduced, with permission, from REF.5 (2005) UBC press.

R E V I E W S

788 | DECEMBER 2013 | VOLUME 14 www.nature.com/reviews/molcellbio

R E V I E W S



80S

Proplastid

?

HY5

HY5

PIF

Phytochromes

GLK

PEPNEP

NEP

SIG

PSILHC PSII

TIC

TIC

TOC

TOC

Chlorophyllsynthesis

Plastid division

Nucleus

Cytosol

Chloroplast

PSI, PSII, LHC,Calvin cyclePRPS, PRPL, PPRs, PDV,TOC159 and TOC33SIG5Chlorophyll synthesis

GLK2Others

NEP, SIG2, PRPS,PRPLTOC132 and TOC34 PEP

HousekeepingPSI, PSII

Etioplast Dark-to-lighttransition

Dierentiationin the light

Thylakoid

PPR

Calvincycle

70S

PDV

Reaction centresComplexes of proteins and photosynthetic pigments that are located at the centre of a photosystem. In these complexes, charge separation occurs to commence the photochemical reactions of photosynthesis.

Antenna complexesGroups of proteins and photosynthetic pigments that collect photons and deliver their excitation energy to the reaction centre of a photosystem.

ThylakoidThylakoids are flattened membranous sacks inside plastids that are responsible for photosynthetic light reactions. They are distinct from the envelope membranes and typically stack to form grana in mature chloroplasts.

composition of which is regulated at least in part bythe differential expression of nuclear genes in response to developmental (for example, the cellular context) and environmental (for example, light quality and quantity) cues, as discussed below12.

The photosynthetic pigments chlorophylls and carotenoids are important metabolites that are needed for chloroplast biogenesis. Although a eukaryotic carotenoid biosynthetic pathway exists in the cytosol of plant cells, plastid carotenoids are synthesized in the organelle by the methyl-d-erythritol phosphate (MEP) pathway, which is also used by cyanobacteria13. Like haem and other tetrapyrroles, chlorophylls are produced by a pathway that is localized in the plastid envelope, using Glu-tRNA (that is, a tRNA loaded with Glu at its 3 end) as starting material. Intermediates in

the pathway are dangerously phototoxic. Consequently, metabolites are channelled between enzymes, and the final assembly with reaction centres or antenna complexe s may occur rapidly at the envelope of proplastids or at the pro lamellar body (that is, a lattice membrane structure) in dark-grown chloroplast precursors called etioplasts14 (FIG.2). Thylakoid membranes (which are characteristically rich in galactolipids) may be formed from the inner envelope membrane by an internal vesicl e buddin g mechanism15.

Plastid genome content and function. Plastid DNA (or the plastome) is ~120160 kb in size in higher plants and highly conserved6. It is represented by a circular map but exists as linear or branched oligomers16. InArabidopsis thaliana, it encodes 54 core plastid proteins that are

Figure 2 | Light-mediated anterograde control of chloroplast development. Developmentally, the bulk of information flows from the nucleus to plastids. Non-photosynthetic plant cells require basal operation of plastid housekeeping proteins that have been synthesized in the cytosol (by 80S ribosomes) and that have been imported through a housekeeping translocator (TOC132TOC34, orange). Basal production of plastid ribosomal proteins (PRPS and PRPL, which are constituents of the 70S ribosome) also takes place. Factors driving and determining the extent of this initial plastid biogenesis are unknown (as indicated by the question mark). Plastid DNA expression is initiated by nucleus-encoded RNA polymerase (NEP); subsequently, plastid-encoded polymerase (PEP) becomes active. Proplastid-to-chloroplast differentiation during leaf development involves the expression of plastid genes that encode core components of the photosystems PSI and PSII, which is driven by PEP and involves regulators such as pentatricopeptide repeat proteins (PPRs) and abundant translational activity. Nucleus-encoded subunits, including those of the light-harvesting antenna complexes (LHC), assemble around these core components. Their import, in large quantities, uses a photosynthetic translocator ( TOC159TOC33; green), and is accompanied by extensive chloroplast growth and division (represented by the plastid division component PDV). Similar phenomena occur when chloroplasts develop from etioplasts following prolonged growth in darkness. Photosynthesis-related nuclear genes often possess G-box promoter elements. In the dark, the basic Leu zipper (bZIP) transcription factor HY5 is ubiquitylated and degraded, whereas basic helixloophelix (bHLH) PHYTOCHROME-INTERACTING FACTOR (PIF) regulators bind G-box elements and repress transcription. Photoreceptor activation causes PIF turnover and represses HY5 degradation; this enables its G-box binding and transcriptional activation. GLKs, which are Golden 2-like MYB transcription factors, are also major drivers of photosynthetic genes, with one (GLK2) being light-induced. Light- or circadian-mediated nuclear control of plastid DNA expression is exerted through -factors (SIG), which control PEP specificity (for example, SIG5).

R E V I E W S




MesophyllTissue in the interior of leaves, between the epidermal layers but not in contact with vascular bundle tissue, which contains most leaf chloroplasts and is the principal site of photosynthesis.

-factorA subunit of multimeric, prokaryotic-type RNA polymerase that is responsible for DNA binding and promoter recognition; it thus determines which genes are transcribed.

Pentatricopeptide repeat proteins(PPR proteins). Proteins that possess tandemly repeated copies of a degenerate, 35-residue motif that enables sequence-specific nucleic acid binding. These proteins are involved in RNA stability or editing in organelles.

Horizontal gene transferThe transfer of genes between two contemporaneous individuals of the same or different species or between two genetic systems of one individual.

involved primarily in photosynthesis, and a further 31 proteins that mediate plastid DNA expression; 45 addi-tional genes encode tRNAs or rRNAs5,6. Fluorescence staining of plastid DNA insitu revealed its presence in nucleoids, as in prokaryotes, and its attachment to the inner envelope or, in mature chloroplasts, the thylakoids6.

Meristematic cells contain fewer than 10 pro plastids, each of which has ~10 plastid DNA copies. Leaf mesophy ll cells can have up to 100 chloroplasts, and each of these can have ~50 plastid DNA copies (FIG.1b). Eventually, the number of copies of plastid DNA is two orders of magnitude greater than that of the nuclear genome in A.thalian a leaves, probably to accommodate the high demand for synthesis of abundant chloroplast proteins or the mutagenic load caused by photo synthetic reactions17. Very little is known about plastid DNA replication or about its transmission when plastids divide (although it has been suggested that envelop e attachmen t may have a role).

As in prokaryotes, plastid genes are organized in operons, with several genes being co-transcribed; the polycistronic psbB mRNA encodes four additional poly peptides18. Plastid transcription involves two types of RNA polymerase (FIG.2). One is homologous to bac-terial RNA polymerases and consists of five subunits: four invariable subunits that are encoded by plastid rpo genes; and a-factor that is encoded by a family of nuclear SIG genes (six in A.thaliana)19. The -subunit confers promoter specificity, thus giving transcriptional control to the nucleus. This multisubunit complex is called the plastid-encoded RNA polymerase (PEP). Plastids also contain a single-subunit, bacteriophage T7-type, nucleus-encoded RNA polymerase (NEP). InA.thalian a, two genes produce NEP20. The differential use of NEP or PEP by plastid genes has been analysed21, and the overall picture is that NEP initiates transcription of plastid housekeeping genes, including rpo, before the resulting PEP drives the expression of genes that encode photosynthetic proteins (FIG.2). Interestingly, defects in housekeeping plastid functions lead to the reduced expression of PEP-transcribed genes but also trigger the compensatory upregulation of NEP-dependent genes22.

The transcripts of several plastome genes have typeII introns or undergo editing (C-to-U single nucleotide conversions)23. This results in an unusually exten-sive need for RNA processing. In fact, many nucleus-encoded plastid proteins are devoted to the regulation of the plastid genetic machinery, with plastome-encoded mRNAs frequently being processed by several unique factors. Particularly noteworthy are the pentatricopeptid e repeat proteins (PPR proteins)24. The A.thaliana nucleus encodes 450 organellar PPR proteins, and half of them are targeted to chloroplasts. Each PPR protein is a sequence-specific binding factor that is involved in the splicing, turnover or editing of 13 plastid mRNAs. Less diverse but also important for plastid gene expression are mTERFs (MITOCHONDRIAL TRANSCRIPTION TERMINATION FACTORS), which were first iso-lated in animal mitochondria (several mTERFs are p lastid-localize d in plants)25,26.

Proteomic analyses of nucleoids identified factors that are involved in DNA replication, repair and anchor-ing, transcription, RNA splicing and editing, ribosome assembly, as well as translation, which strongly implies the combined occurrence of transcription and translation, as in prokaryotes26,27. Translation occurs in 70S ribosomes, which have a composition and antibiotic sensitivity remi-niscent of bacterial ribosomes. 70S ribosome-dependent translation is also heavily regulated and often the most important regulatory step in chloroplast gene expression28.

Stability of the plastid genome. Why has a plastid genome been retained? Some hypotheses suggest that the fol-lowing factors play a part: the extreme hydrophobicity of proteins encoded by the retained genes (which pre-cludes efficient protein transport); the metabolic role acquired by Glu-tRNA; and a requirement for very rapid modulation of expression29. Genes that are retained in chloroplasts encode core proteins of the photosynthetic machinery, around which the complexes assemble30. The location of these genes in the organelle enables the cell to timely regulate the synthesis of such components, which could be important to protect against photo-oxydative damage30. Accordingly, excessive excitation of one photosystem causes the compensatory rebalanc-ing of transcription of the reaction centres31, and such a readjustment is influenced by the kinases STN7 and CSK in the plastids32,33. Altered transcription of a gene (psaA) that encodes a core reaction centre component is trig-gered by the electron load at plastoquinone (an electron carrier that links PSII and PSI), which acts by changing the phosphorylation status and activity of SIG1, a process that is possibly mediated by the aforementioned kinases34.

The plastome is surprisingly promiscuous. Selectable markers inserted into plastid DNA enabled the observa-tion of the transfer of genes to the nucleus35. Moreover, horizontal gene transfer of plastid genes, and even com-plete plastid DNA or chloroplasts, between different individuals across grafts has been demonstrated36,37. The existence of multiple copies of plastid DNA in each orga-nelle perhaps aids repair by recombination, but conflict-ing evidence exists as to whether plastid DNA is retained as leaves mature16. Transmission of plastid DNA across generations generally occurs through the egg cytoplasm (plastid DNA is typically degraded in pollen), but excep-tions are not uncommon and therefore it has been argued that the frequently-observed uniparental inheritance simply minimizes damaged DNA repair costs16.

Plastid-to-nucleus signallingSeminal work showed that barley mutants with pri-mary defects in plastid ribosomes fail to synthesize nucleus-encoded plastid enzymes. This led to the notion of p lastid-to-nucleus information flow, which is also referred to as plastid retrograde signalling38. Since then, a complex picture of regulation has emerged39,40. Signals that operate during chloroplast biogenesis (biogenic control) differ from signals that are produced when, in response to a variable environment, chloroplasts relay their status to the nucleus for fine-tuning purposes (operational control).

R E V I E W S


R E V I E W S



Source of positive signal

Source of inhibitory signal

Source of stress signal,promotive of defence

Redox overload;ROS or their action

High light orother light stress

Proplastid

Chloroplast

Biogenesis Operation

PEPNEP SIG

PPR70S

FC1

FC1Chlorophyllsynthesis

CleavagePTM

PTM?

PTMPAP

ABI4

HY5GLK

PSI, PSII, LHCChlorophyll synthesisCalvin cycle

APXAntioxidantDefence

?

?

LHCPSI

FC1 haemsynthesis

PEPSIG

PSICalvincycle

Drought

PQ PSIPSILHC PSII

-cyclocitral

MEP MEcPP

PAPPAP

KIN

Nucleotide

X

SAL1

Haem

PSILHC PSII

Calvincycle

Thylakoid

Nucleus

Cytosol

-carotene

TOC

PhotoautotrophicAn organism that derives its energy from light and that uses CO2 as a carbon source. Theaerial tissues of plants are photoautotrophic when kept in the light.

Biogenic control. The control of the production of pro-teins that constitute the photosynthetic apparatus by biogenic signals enables the efficient establishment of photoautotrophic metabolism. The action of these signals was revealed by the analysis of plants that experience defective translation (induced genetically or chemically) or photooxidative damage (caused by carotenoid-defi-ciency, for example after treatment with norflurazon her-bicide)39. Although it has been argued that such stresses are unnaturally severe, similar effects also occur under

milder conditions, in response to partial defects in photo-synthetic protein import41 or in nucleus-encoded -factor (SIG2 or SIG6)-dependent plastid transcription42 (FIG.3).

Mechanistic insight emerged from a genetic screen in A.thaliana for genomes uncoupled (gun) mutants. In the wild type, norflurazon treatment (which causes chloro-plast photooxidative damage) leads to the repression of the nuclear LIGHT HARVESTING CHLOROPHYLL A/B BINDING PROTEIN B1 (LHCB1) gene, which encodes an abundant photosynthetic antenna protein;

Figure 3 | Plastid-to-nucleus or retrograde signalling pathways. Signals that operate during chloroplast biogenesis report the readiness of the organelle to receive photosynthetic proteins. Haem synthesized by ferrochelatase 1 (FC1) is a positive signal that promotes the expression of nuclear genes encoding components of the photosynthetic machinery (probably indirectly; dashed arrow). Meanwhile, failure of some final steps of chlorophyll synthesis causes the repression of the same nuclear genes, possibly through a negative signal. Plastid photooxidation, and defective plastid transcription, protein synthesis or protein import may all have an impact on either or both of these positive (haem) and negative signals. Information from the negative, photooxidative signal, and from stresses which may impair the FC1-derived haem signal, is relayed to the nucleus (directly; solid arrow) via cleavage of the envelope-associated PTM (PHD WITH TRANSMEMBRANE DOMAINS) transcription factor. These phenomena affect nuclear gene expression, in part by reducing the activity of Golden 2-like MYB transcription factors (GLK) or by promoting the activity of ABI4 (ABSCISIC ACID INSENSITIVE 4; a competitor of G-box-binding factors such as HY5). During chloroplast photosynthetic operation, signals are triggered by environmental changes. Imbalances in the excitation of the photosystems PSI and PSII can cause an excessive reduction of plastoquinone (PQ); this signals plastid kinases (KIN) to orchestrate short-term responses, including modulating plastid DNA transcription (for example, KIN-dependent -factor5 (SIG5) phosphorylation promotes PSI gene expression). Simultaneously, this regulates nuclear genes, which may include the genes that encode light-harvesting complex proteins (LHC). Nuclear genes that are coordinately repressed or activated by biogenic signals can be regulated in opposite directions by this mechanism. High light or loss of electron sinks in drought promote generation of reactive oxygen species (ROS), which leads to the production of cyclocitral, PAP (3 phosphoadenosine- 5 phosphate), MEcPP (methylerythritol cyclodiphosphate) and ABA (abscisic acid) (not shown). These factors activate antioxidant (including APX (ASCORBATE PEROXIDASE) and defence genes in the nucleus. Transcription factor targets of operational signals are largely unknown (represented by question marks). MEP, methyl-d-erythritol phosphate.

R E V I E W S




CryptochromeA family of plant photoreceptors that are sensitive to blue light and, together with phytochromes, control developmental responses.

thegunmutants were identified by their inability to enact such repression. The GUN1 gene encodes a PPR protein that was postulated to act as an integrator of multiple plastid-to-nucleus signals43. By contrast, all other gun mutations affect steps in the biosynthesis of tetrapyrroles (chlorophylls and haem), with GUN4 and GUN5 being involved in Mg-chelation of protoporphy-rin IX to produce Mg-protoporphyrin IX (Mg-proto), which is the first dedicated step of chlorophyllbio-synthesis (Fe-chelation by ferrochelatases produces haem)40,44. Accumulated Mgproto in photo-damaged seedlings has been proposed to mediate the repression of photosynthetic genes44. However, there has been much controversy, as quantitative analyses failed to detect the accumulation of Mgproto or any other tetrapyrrole in such plants45,46. Other observations also fit uneasily with the Mgproto negative signalling model47,48.

A breakthrough came after the gun mutant screen was redesigned, which enabled the identification of gain-of-function mutants and positive plastid signals, whereas the original screen could only identify nega-tive regulators49. The revised search identified GUN6, a gene that encodes ferrochelatase 1 (FC1), which produces a specific haem pool that promotes photo-synthetic gene expression. It has been argued that most observations to date can be explained by the existence of two signals40: a primary haem signal with biogenic function, reporting on the ability of plastids to take up and accommodate photosynthetic gene products; and a later, repressive stress signal which may occur when thylakoid assembly stalls, resulting from the short-lived presence of Mgproto and singlet oxygen (1O2) (FIG.3). What about the role of GUN1? A closely related PPR protein, pTAC2 (PLASTID TRANSCRIPTIONALLY ACTIVE CHROMOSOME 2), was identified in nucle-oids27,43, which suggests that GUN1, rather than being a signal integrator, could control plastid gene expres-sion in a way that affects tetrapyrrole synthesis and thus signalling40.

Mutants with defects in light signalling components (the photoreceptor cryptochrome1 and the transcription factor HY5) were identified as mild gun mutants, and this phenotype could be suppressed by second muta-tions that cause constitutive light responses47. A role for light (particularly blue light) in the production of the tetrapyrrole stress signal could explain such observa-tions47. Interestingly, gun mutants also display attenu-ated repression of plastid DNA-encoded genes (that is, not just nuclear genes) following norflurazon treat-ment48; this suggests that the mutants actually experi-ence reduced photooxidative stress, which could partly explain the gun phenotype.

How do haem or other tetrapyrroles relay plastid-status information to the nucleus and what are the subsequent nuclear events? Only FC1-generated haem is active, arguing against the simple possibility of free haem diffusion. Insights came from the identification of a chloroplast envelope-associated plant homeo-domain (PHD) transcription factor, PTM (PHD WITH TRANSMEMBRANE DOMAINS)50. Upon photo-oxidative exposure, proteolytic cleavage releases the

amino terminus of PTM, and this enables its nuclear localization and activation of the abscisic acid (ABA)-related transcription factor ABI4 (FIG.3). Acting down-stream of the GUN components, ABI4 may prevent the action of light-responsive transcription factors by occupying an overlapping DNA-binding site43. Another factor that may be involved is GLK1, which is a Golden2-like MYB transcription factor that activates various photosynthetic genes and is repressed following defective chloroplast protein import41,51.

Operational control. Chloroplasts host light-driven electron transport reactions that generate energized elec-trons for downstream processes. Imbalances in electron flow (for example, under varying light conditions or when electron sinks become limiting following stoma-tal closure) have potentially catastrophic consequences, as this leads to the generation of highly damaging 1O2 at PSII or superoxide (O2

) and eventually H2O2 at PSI. Thus, plants finely tune their cellular activities to envi-ronmental conditions by using c hloroplast-generate d signals 39 (FIG.3).

As mentioned earlier, the excitation balance between the photosystems controls the redox poise at plastoquinone. This results in the activation of STN7 and CSK, which may alter plastid gene expression to re balance photosystem stoichiometry32,33. This stimulus also causes broad changes in nuclear gene expression programmes and cellular metabolism52. For nucleus-encoded reaction centre genes, this requires STN7 and an unknown signal relay53 (FIG.3).

Another operational signal is 1O2, which is produced under high-light conditions and can induce stress-responsive genes (often antioxidants) or cause seedling death. The death response is not toxicity-related, but rather due to a genetic programme: in the absence of plastid EXECUTER 1 (EX1) and EX2, the accumula-tion of 1O2 causes neither death nor gene expression responses54. Exactly how EX1 and EX2 mediate 1O2 sig-nalling is unknown, but an active role for plastid tran-scription seems possible39. H2O2, which is also generated by chloroplasts under high-light stress, triggers the tran-scription of APX2 (a gene that encodes the antioxidant enzyme ascorbate peroxidase) and other genes, in vas-cular cells and systemically, by initiating the synthesis of ABA hormone55.

Three metabolites that have a role in p lastid-to-nucleus signalling were recently identified: 3 -phosphoadenosine-5-phosphate (PAP) 56, -cyclocitral57 and methylerythritol cyclodiphosphate (MEcPP)58. -cyclocitral, which is an 1O2-induced oxidation product of -carotene, causes a large tran-scriptional response in the nucleus. Because 1O2 is extremely short-lived and unlikely to leave chloro-plasts, and because transcriptional responses to 1O2 and -cyclocitral are highly similar, it seems likely that -cyclocitral mediates 1O2 responses as a plastid-to-nucleus signal57 (FIG.3). The phosphonucleotide PAP accumulates in response to high-light and drought stresses, and its levels are controlled by the chloroplast enzyme SAL1. PAP activates APX2 and other stress

R E V I E W S


R E V I E W S


TransloconA protein complex that is embedded within a membrane and acts in the translocation of client proteins into the membrane, or from one side of the membrane to the other.

barrelA structural arrangement found in proteins of the outer membranes of Gram-negative bacteria, mitochondria and plastids, in which a number of transmembrane -strands cooperate to form a cylindrical structure with a central pore.

StromaThe aqueous internal matrix of plastids enclosed by the inner envelope membrane, where, in chloroplasts, the Calvin cycle enzymes mediate photosynthetic carbon fixation.

genes. Crucially, this is prevented by SAL1 acting in chloroplasts or ectopically in the nucleus, which implies that PAP is a signal that reaches the nucleus56. Finally, MEcPP is the substrate of a bottleneck enzyme in the MEP isoprenoid synthesis pathway. Mutants of this par-ticular enzyme display constitutive nuclear stress gene expression, constitutive responses to abiotic stress and resistance to biotrophic pathogens. Moreover, MEcPP levels are increased by stresses in the wild type58. This shows that the MEP pathway is a global stress sensor and highlights the role of chloroplasts in the operational control not only of their own function but of the entire organism (FIG.3).

Overall, a picture emerges in which chloroplasts sig-nificantly influence the control of their own biogenesis. Moreover, as plastids experience and respond to various external stimuli, they have evolved central roles in the regulation of organellar and organismal functions.

Protein import, routing and quality controlAs already mentioned, most proteins in plastids are encoded in the nucleus and synthesized in the cytoso l. They initially possess an N-terminal targeting sig-nal which is called a transit peptide. Transit peptides engage cytosolic sorting factors59,60 and the trans-location machinery in the envelope membranes, thereby mediating post-translational transport into the orga-nelle11,6163. Despite their importance, transit peptides are remarkably variable in size (~20100 residues) and sequence64,65. Following import, transit peptide removal by the stromal processing peptidase (a metalloendo-peptidase of the M16 family)66 precedes protein folding or engagemen t of internal sorting pathways (FIG.4).

Envelope translocation. Translocation of precursor proteins (pre-proteins) across the envelope is mediated by multiprotein complexes termed TOC (translocon at the outer envelope membrane of chloroplasts) and TIC (translocon at the inner envelope membrane of chloro plasts)11,6163 (FIG.5). Pre-proteins initially inter-act with receptor components of the TOC complex, before deeply inserting to make contact with the TIC complex in an energy-dependent process. Completion of trans location requires much stromal ATP, but, in contrast with mitochondrial import, a transmembrane proton-motive force is not required67.

The TOC complex comprises three main proteins: TOC34, TOC75 and TOC159 (the numbers indicate the molecular weight)11,6163 (FIG.5). TOC75 is deeply embedded in the outer membrane and, like mitochon-drial TOM40 (TRANSLOCASE OF THE OUTER MITOCHONDRIAL MEMBRANE 40), it forms a -barrel-type translocation pore that accepts largely unfolded preproteins62. TOC34 and TOC159 project homologous GTPase domains into the cytosol, and it is generally thought that they mediate pre-protein recog-nition. This view is supported by the fact that GTP analogues block import invitro. However, analyses of GTP-binding or -hydrolysis receptor mutants in trans-genic plants reveal a complex picture of GTP-regulation that is not fully understood63,68.

TIC22 may facilitate pre-protein passage through the intermembrane space, acting as a chaperone, or aid in the formation of TOCTIC supercomplexes to enable the simultaneous transport across both membranes69 (FIG.5). The composition of the TIC machinery has been much debated, with different proteins proposed to fulfil the most basic function of channel formation70. However, the identification of a 1 MDa complex contain-ing TIC20 recently led to an important breakthrough71. This complex contains three additional nucleus-encoded components (TIC21, TIC56 and TIC100), as well as p lastome-encoded TIC214. The complex has chan-nel activity when reconstituted invitro and therefore has been proposed to be a general TIC translocon. Interestingly, the TIC56, TIC100 and TIC214 compo-nents are missing in grasses, which raises doubts about the broader relevance of the complex.

TIC110 exists in a smaller 200300 kDa complex which possibly acts later in the import process71 (FIG.5). This protein projects a rod-shaped helix-repeat domain into the stroma to recruit stromal chaperones72. By analog y with the heat shock protein 70 (HSP70)-based system that drives mitochondrial import, these chap-erones are putative components of an ATP-powered import motor. Remarkably, several chaperones (homo-logues of HSP60 (also known as CPN60), HSP70, HSP90 and HSP100) associate with the TIC machinery, but their functional interrelationships are unclear59,73. It is unlikely that all these chaperones are involved in import propulsion, and therefore some of them may function in other processes such as protein folding. An STI1 (HIP-HOP) domain-containing co-chaperone termed TIC40 control s TIC-associated chaperone activity74.

Additional proteins (TIC32, TIC55 and TIC62) may regulate import in response to plastid redox status75 (FIG.5). That import of some pre-proteins is influenced by light is consistent with this hypothesis, but mechanistic details underlying this putative redox-regulatory circuit remain elusive.

Other targeting mechanisms. Not all plastid proteins are targeted via canonical, transit peptide-mediated engagement of the TOCTIC machinery. Apart from outer membrane proteins, which are discussed below, as much as 10% of chloroplast proteins have been estimated to arrive via non-canonical routes76. For example, some proteins are sorted to the inner envelope by intrinsic (non-cleavable) targeting signals in a TOC-independent fashion77 (FIG.4). Another exciting development has been the discovery of a plastid-targeting pathway that involves the endomembrane system78. Clients of this pathway have signal peptides for co-translational endoplasmic reticulum (ER) transport and pass through the Golgi (where they may be glycosylated) to reach the plas-tids (FIG.4). Finally, mRNA targeting to the vicinity of p lastidsmay contribute to plastid protein sorting79.

Internal routing of plastid proteins. Chloroplasts are complex organelles with several distinct suborganellar compartments. Thus, internal sorting of chloroplast protein s is remarkably complex (FIG.4).

R E V I E W S





Golgi

TOC (GTP)

TIC (ATP)

Cytosolicfactors

Nucleus ER

70S

?

80S

Chloroplast

cpSEC2

Glycoprotein

cpSEC1ATP

cpSECA1cpSECY1/E1

cpSRP54cpSECY1/E1

cpTATpH

THA4, HCF106cpTATC

cpSRPGTP

cpSRP54/43cpFTSY, ALB3

Spontaneous

Co-translationaltransport

Co-translationaltransport

Non-canonicaltransport

Plastid DNA

TPP TPP

SPP

Stroma

Lumen

Lumen

Thylakoid

Thylakoid

Inner envelope

Outer envelope

AKR2, HSP17.8TOC75

Post-translational transport(nal destinations are indicated)

Thylakoid membrane

Lumen

Stroma

1

2

IEMOEM

Transit peptide (TOC and TIC)Signal peptide (ER)Intrinsic targeting signalLumenal targeting peptide (cpSEC1)Lumenal targeting peptide (cpTAT)Transmembrane domain

Targeting peptides and domains

?

Omp85A protein superfamily that comprises a range of -barrel proteins that are all involved in protein transport or assembly; BamA is a prototypical member of the family and is responsible for the biogenesis of bacterial -barrel proteins.

Most outer envelope proteins lack a cleavable sorting peptide80, and instead, the targeting information exists within hydrophobic transmembrane domains. Cytosolic sorting factors (AKR2 (ANKYRIN-REPEAT DOMAIN-CONTAINING 2) and HSP17.8) mediate the transport of such proteins to the organellar surface, whereas integration may require the TOC75 channel60. A notable exception

is TOC75 itself, which possesses a bipartite signal that comprises a standard transit peptide adjacent to an intra-organellar targeting peptide that enables translocon dis-engagement and membrane integration81. Topogenesis of -barrel-containing proteins such as TOC75 may depend on OEP80 (OUTER ENVELOPE PROTEIN 80), which like TOC75 is an Omp85-type protein82.

Figure 4 | The protein import and routing pathways of plastids. Most nucleus-encoded proteins enter the organelle post-translationally via the TOC (translocon at the outer envelope membrane of chloroplasts)TIC (translocon at the inner envelope membrane of chloroplasts) machinery, and have their transit peptides removed by the stromal processing peptidase (SPP) upon arrival. Intraorganellar sorting to the inner envelope or thylakoids may then ensue. Inner envelope membrane (IEM) targeting follows one of two pathways: the stop-transfer (1) and post-import (2) pathways. Thylakoidal targeting involves four different pathways: two pathways lead to the thylakoid lumen (cpSEC1 (CHLOROPLAST SEC1) and cpTAT (CHLOROPLAST TWIN ARG TRANSLOCASE)); and two pathways lead to the thylakoid membrane itself (cpSRP (CHLOROPLAST SIGNAL RECOGNITION PARTICLE) and the spontaneous pathway). The requirements and key mediators of each pathway are indicated. Lumenal targeting peptides are removed by the thylakoidal processing peptidase (TPP). Inaddition, plastid DNA-encoded proteins are co-translationally targeted to thylakoids by components of the cpSEC1 and cpSRP pathways. Unlike IEM proteins, outer envelope membrane (OEM) proteins typically do not have transit peptides, and instead are targeted to the membrane (via cytosolic sorting factors AKR2 (ANKYRIN-REPEAT DOMAIN-CONTAINING2) and HSP17.8 (HEAT SHOCK PROTEIN 17.8) and the TOC75 channel) by intrinsic targeting information within transmembrane regions. The existence of non-canonical inner envelope proteins that also use intrinsic targeting signals became recently apparent, but the transport mechanisms are unknown. Another unusual pathway involves the co-translational transport into the endoplasmic reticulum (ER) and passage through the Golgi (where glycosylation may occur). Lipids are delivered to developing thylakoids by vesicles derived from the inner envelope, and these perhaps also deliver proteins. ALB3, ALBINO 3; cpFTSY, CHLOROPLAST FILAMENTOUS TEMPERATURE SENSITIVE Y; HCF106, HIGH CHLOROPHYLL FLUORESCENCE 106; THA4, THYLAKOID ASSEMBLY 4.

R E V I E W S


R E V I E W S


YidC and Oxa1YidC and Oxa1 are homologous, polytopic membrane proteins found in bacteria and mitochondria, respectively. They mediate client insertion during membrane protein biogenesis.

PlastoglobulesLipid bodies that are found within plastids. They are responsible for the storage and biosynthesis of lipids, and in chloroplasts they are associated with the thylakoids.

Ubiquitin E3 ligaseOne of a large group of enzymes that catalyse the final step in the ubiquitylation cascade and that are responsible for determining target specificity.

Inner envelope proteins typically possess a transit peptide and are imported via the TOCTIC machiner y. Two different targeting routes are known: the stop-transfer and the post-import pathways83,84 (FIG.4). In the post-import pathway, complete translocation into the stroma precedes membrane insertion in a separate event. This is similar to conservative sorting to the mitochondrial inner membrane, so-called because it uses machinery of prokaryotic origin. Interestingly, components related to the bacterial Sec (secretion) system were recently identified in the plastid envelope, which implies that a cpSEC2 (CHLOROPLAST SEC2) system exists in addition to the well characterized thyla koidal cpSEC1 system (see below)85,86 (FIG.4). Stop-transfer insertion involves the lateral exit from the TIC translocon, which is mediated by client transmembrane regions, and may be important for aggregation-prone, hydrophobic proteins84.

Thylakoid biogenesis. Multiprotein photosynthetic complexes dominate the thylakoidal proteome, and these comprise plastid- and nucleus-encoded sub-units. Nucleus-encoded subunits are imported via theTOCTIC system and are subsequently sorted to the thylakoids by one of four pathways8689 (FIG.4). The relat-edness of these internal systems to those in bacteria exemplifies the conservative sorting concept, wherein mechanisms of the endosymbiont are parsimoniously retained.

Thylakoid lumen proteins use the cpSEC1 and cpTAT (CHLOROPLAST TWIN ARG TRANSLOCASE) pathways and possess bipartite targeting information: a standard transit peptide preceding a lumenal tar-geting peptide that is similar to bacterial signal pep-tides86,87,89 (FIG.4). The cpSEC1 pathway involves the cpSECA1 ATPase and the cpSECY1cpSECE1 trans-locon and accepts only unfolded proteins. By contrast, the cpTAT pathway (comprising THA4 (THYLAKOID ASSEMBLY 4), HCF106 (HIGH CHLOROPHYLL FLUORESCENCE 106) and cpTATC, homologues of bacterial TatA, TatB and TatC, respectively) is powered by the thylakoidal proton gradient and accommodates clients that acquire a folded structure in the stroma through cofactor binding or oligomerization. In both cases, the lumenal targeting peptide is removed upon arrival by the thylakoidal processing peptidase66.

Structurally simple thylakoid membrane proteins may insert spontaneously without the transport machin-ery, whereas others follow the cpSRP (CHLOROPLAST SIGNAL RECOGNITION PARTICLE) pathway86,87 (FIG.4). This pathway consumes GTP due to an interac-tion between cpSRP (comprising the cpSRP54 GTPase and cpSRP43, which is a chaperone component that is unique to plastids)90 and its membrane receptor, cpFTSY (CHLOROPLAST FILAMENTOUS TEMPERATURE SENSITIVE Y), and it mediates the insertion of light-harvesting components via an insertase that is related to YidC and Oxa1, ALB3 (ALBINO 3). In addi-tion, components of this pathway and the chloroplast SEC1 pathway enable the co-translationa l insertion of plastom e-encoded proteins.

Lipids are also essential thylakoidal constituents, and these are synthesized in the envelope (and elsewhere) before delivery by transport vesicles that bud off from the inner envelope membrane15,91 or in thylakoid-associated plastoglobules92. The aforementioned vesi-cles may also bring proteins to the thylakoids, perhaps following their arrival in cytosolic transport vesicles78. Acharacteristic feature of the thylakoid network in land plants is the stacking of lamellae to form grana (where PSII is concentrated)93, and this is enabled by CURT1 (CURVATURE THYLAKOID 1), which induces membran e curvature at granal margins94.

Client-specific protein import pathways. Most TOCTIC components were identified biochemically using pea chloroplasts. Genomic analysis in A.thaliana (and other plants) revealed that some components (particu-larly the receptors) are encoded by small gene families. For example, TOC159 is encoded by four A.thaliana genes (atTOC90, atTOC120, atTOC132 and atTOC159), and TOC34 by two genes (atTOC33 and atTOC34)95,96. The encoded receptor isoforms have distinct functions to enable client-specific import pathways: atTOC33 and atTOC159 cooperate in the import of highly-abundan t photosynthetic pre-proteins, whereas atTOC34, atTOC120 and atTOC132 act in the import of house-keeping pre-proteins (FIG.5). Accordingly, the plastid protein import1 (ppi1) and ppi2 mutants (which lack atTOC33 and atTOC159, respectively) exhibit defective chloroplast biogenesis, whereas loss of other isoforms affects non-photosynthetic development11,61,63. The operation of different import pathways may circum-vent damaging competition effects between precursors or contribute to the formation of different plastid types (FIG.1a). It may also enable age-dependent regulation of import65,97.

Proteolysis and quality-control. A RING-type ubiquiti n E3 ligase, SP1, was recently identified in the plastid outer envelope membrane98. SP1 mediates ubiquityla-tion of TOC components and their degradation by the cytosolic ubiquitinproteasome system (UPS) (FIG.5). This enables the reorganization of the TOC machin-ery to alter the balance of the aforementioned client-specific pathways, thereby influencing which proteins are imported, the composition of the organellar pro-teome and the developmental fate and functions of the organelle. Earlier studies had shown that the UPS removes pre-proteins in the cytosol that have not been imported60,99, but the study on SP1 demonstrated for the first time that the UPS acts directly on plastids, paral-leling findings in mitochondria100. Other work revealed a role for the UPS in regulating amyloplast function durin g gravitropism101.

Although the UPS directly regulates cytosolically exposed outer envelope proteins, proteolysis in the organellar interior is mediated by systems of prokary-otic origin. Such protease activity may control protein abundance or maturation, recycle cleaved targeting pep-tides or mitigate photooxidation effects by the removal of damaged components66,102,103. Numerous proteolytic

R E V I E W S





IMS

Cytosol

OEM

IEM

Stroma

atTO

C15

9GTP GTP

atTO

C33

TOC

75TI

C20

A

atTO

C13

2/12

0GTPGTP

atTO

C34

TOC

75

A

RNF

SP1

TIC214(YCF1)

TIC100

TIC56

TIC

21

TIC

110

TIC

40

STI1

TIC

55

HSP100 HSP70

HSP90HSP60

TIC62 TIC32

Putative redoxregulators

26SPGTP

ATP

GDP + Pi

ADP + Pi

Turnover ofTOC proteins

Highly-abundant,photosynthetic pre-proteins

Non-photosynthetic,housekeepingpre-proteins

?

TIC22

Transit peptideSPP

AAA+ family(ATPases associated with various cellular activities). A protein family comprising a diverse set of factors that mediate conformational changes in client proteins. They are characterized by the possession of Walker-type ATPase motifs and the ability to assemble into oligomeric rings.

systems operate inside plastids, and several proteins of these systems are encoded by multigene families, which enables the formation of hetero-oligomers. Stromal CLP proteases comprise separate chaperone and Ser protease subunits and possibly fulfil general house-keeping functions. By contrast, FTSH proteins are zinc-metalloendopeptidases that have both chaperone and proteolytic functions (within a single polypeptide), are thylakoid-associated and participate in the PSII repair cycle. Ser-proteases of the DEG family are also involved in this function, but how they cooperate with FTSH is

unclear. LON proteases are related to FTSH but lack transmembrane domains and are stromally located. Notably, CLP chaperones, FTSH and LON belong to the ubiquitous AAA+ family of ATPases, whereas DEG is ATP-independent.

Propagation and structural dynamicsIn parallel with organellar growth through the synthesi s and import of proteins, it is important that chloroplast numbers are maintained and enlarged as plant cells grow and divide. The propagation mechanism used is part of a broader range of structural dynamism exhibite d byplastids.

Chloroplast division. Reflecting their endosymbiotic origin, chloroplasts are propagated by binary fission of pre-existing organelles104107. The division machinery comprises components of prokaryotic and eukaryotic origin and is located both internally and externally at the organellar surface (FIG.6). Genetic analyses involving mutants with abnormalities in plastid size and number, as well as homology searches informed by well-characterize d bacterial division systems, have greatly advanced our understanding of plastid division in recentyears.

Multiple concentric rings form around the constric-tion site of a dividing organelle (FIG.6). An early event is the formation of the stromal Z-ring, which comprises two functionally-distinct homologues108 of the bacterial divi-sion component FtsZ a self-assembling, tubulin-like GTPase with mechanochemical function. The recruit-ment and assembly of the Z-ring is promoted by the inner membrane protein ARC6 (ACCUMULATION AND REPLICATION OF CHLOROPLASTS 6; which is homologous to the cyanobacterial division fa ctor Ftn2), a function that is dynamically regulated by the paralogous protein PARC6 and the stromal factor ARC3 (REFS109,110). ARC3 incorporates an incomplete FtsZ-like domain, which may enable it to antagonize FTSZassembly.

The correct equatorial placement of the division machinery is governed by additional factors that have been derived from the ancestral division machinery (FIG.6a). In Escherichia coli, pole-to-pole oscillations of MinD (minicell D) and MinE act to minimize the concentration of MinC (which inhibits FtsZ poly-merization) at the midpoint, ensuring centralized Z-ring formation. In A.thaliana chloroplasts, the system is dif-ferent: MINC is absent and ARC3 possibly fulfils its role109,111, and MIND cooperates with another eukary-otic addition, MCD1 (MULTIPLE CHLOROPLAST DIVISION SITE 1)112.Topological information is then conveyed to the outer membrane, through the recruit-ment of PDV1 (PLASTID DIVISION 1) and PDV2 by the intermembrane space domains of PARC6 and ARC6, respectively. The PDV proteins act to recruit cytosolic ARC5 (also known as DRP5B; which is a dynamin-related, m embrane-remodelling GTPase) to initiate constrictio n104107 (FIG.6b,c).

Plastid-dividing rings on either side of the envelope incorporate additional important factors (FIG.6b,c). Recent work in the red alga Cyanidioschyzon merolae

Figure 5 | The TOCTIC machinery mediates client-specific protein import and is controlled by the ubiquitinproteasome system. Most nucleus-encoded plastid proteins are imported by TOC (translocon at the outer envelope membrane of chloroplasts) and TIC (translocon at the inner envelope membrane of chloroplasts) complexes that are located in the outer envelope membrane (OEM) and inner envelope membrane (IEM), respectively. The core TOC machinery comprises two transit peptide receptors (TOC34 and TOC159) and channel-forming TOC75. The receptors are GTPases, accounting for the GTP requirement during the early stages of import. TOC159 additionally possesses an amino-terminal acidic (A) domain that may participate in transit peptide recognition. In Arabidopsis thaliana (and other plants), the receptors exist in multiple isoforms (TOC159: atTOC159, atTOC132 and atTOC120; TOC34: atTOC33 and atTOC34) with non-identical recognition specificities, which enables different, client-specific import pathways. SP1 is a RING finger (RNF) E3 ligase in the OEM that targets TOC proteins for ubiquitylation and turnover by the 26S proteasome (26SP), which results in the reorganization of the TOC machinery and balancing of client-specific pathways. A general TIC translocon of 1MDa, incorporating plastome-encoded TIC214 (also known as YCF1) and four nucleus-encoded components, may act upstream of a putative motor complex comprising TIC40 and TIC110. The latter complex recruits and regulates various stromal chaperones, which drive protein import and/or assist in protein folding at the expense of ATP hydrolysis. Additional IEM components may regulate import in response to internal redox signals (indicated by the question mark). HSP, heat shock protein. IMS, intermembrane space; SPP, stromal processing peptidase.

R E V I E W S


R E V I E W S



b

a

c

Stroma

Cytosol

IMSOEM

OEM

IEM

IEM PDV1

PARC6ARC6ARC3

PDV2

FTSZ1FTSZ2

OuterPD-ring

ARC5ARC5

OEMIEM

FTSZ1FTSZ2

OuterPD-ring

InnerPD-ring

ARC3MINDMINEMCD1

ARC3MINDMINEMCD1

Binary fissionThe chief mode of cellular division in prokaryotes, by which a cell divides symmetrically to form two equal daughter cells that each contain genetic material.

DynaminA large GTPase that assembles into helical arrays at membrane constriction zones in order to catalyse membrane fission during various membrane remodelling processes.

shed light on the composition of the outer plastid-dividing-ring: it is a bundle of 57 nm polyglucan fila-ments that incorporate a range of proteins, including the glucosyl transferase PDR1 (PLASTID-DIVIDING RING 1), which probably mediates filament synthe-sis113. It has been proposed that the polyglucan fila-ments slide over each other in an ARC5-dependent manner to generate the principal constrictive force during organellar division114. During late stages of con-striction, the Z-ring disassembles, and ARC5 migrates to a position beneath the plastid-dividing ring where it completes membrane fission using the pinchase activity that is characteristic of dynamins. Potential PDR1 orthologues exist in land plants, but the extent to which the aforementioned plastid-dividin g ring-related mechanisms are conserved remains to beseen104.

The division mechanisms above were elucidated in chloroplasts, which typically have a regular, defined shape. However, other plastid types (including pro-plastids, which divide extensively in meristematic tissues) are structurally more heterogeneous, and it is difficult to envisage how such a regimented, multi-ring division mechanism might operate efficiently in these cases. The existence of an alternative replica-tion mechanism could account for the effective par-titioning of plastids seen in even the most extreme canonical division mutants (which have 23 giant chloroplasts per mesophyll cell). Vesicle budding from the plastid body is one possibility, which has been shown to occur during chromoplast biogenesis in ripening tomato fruit and possibly also in endosperm amyloplasts106,115.

Structural dynamics. The shape of the plastid is flex-ible and partly governed by osmotic effects (which are linked to the naturally hypo-osmotic surround-ings of the organelle), and mechanosensitive ion channels related to bacterial MscS and envelope-localized macrostructures that contain VIPP1 (VESICLE-INDUCING PROTEIN IN PLASTIDS1; a homologue of PspA, which controls bacterial mem-brane integrity) have been shown to be involved116,117. VIPP1 is also implicated in the transfer of vesicles from the envelope to thylakoids and in photosynthetic comple x assembly15,118.

A further indication of the structural plasticity of plastids stems from the existence of stromules double-membrane-bound, stroma-filled tubules that are ~0.5 m in diameter and that extend from the orga-nellar surface106,119,120. Reflecting their greater structural heterogeneity, non-green plastids tend to have more and longer stromules than the generally more closely packed chloroplasts. Stromule frequency is also influ-enced by developmental and environmental factors, including biotic and abiotic stresses. Large macro-molecular complexes (including the 550 kDa RuBisCo (ribulose-1,5-bisphosphate carboxylaseoxygenase) holoenzyme) are able to pass through stromules, but nucleoids are excluded121. Stromules are highly dynamic structures that, in an actin- and myosin-dependent manner, can rapidly extend, contract, branch or even detach completely. Whether stromules are able to fuse and interconnect plastids is contro-versial122, but it has been suggested that they facilitate communication between plastids or even with other organelles. Alternatively, they may increase the plastid surface area to aid in transport processes or present organellar material in an accessible way for nutrient recycling through autophagy123.

Mitochondrial fusion occurs frequently and has been well characterized98,124. Chloroplast fusion occurs in green algae during syngamy125, but it has not been described in land plants. Thus, although the fusion of whole plastids seems unlikely, the general plasticity of these organelles and reports of possible stromule-mediated interconnections suggest that small, p lastid-derived structures may be able to fuse37,106.

Figure 6 | The plastid division machinery. a | Plastids typically divide by binary fission. This process is mediated by concentric rings of a division machinery on both sides of the envelope (at the stromal and cytosolic surfaces of organelle). Correct positioning of the division machinery at the constriction zone is mediated by the MIN (MINICELL) system (grey arrows), which acts to control Z-ring formation. b | Multiple rings surround the organelle at the division site to enable constriction. On the stromal side, the Z-ring (comprising functionally distinct FTSZ1 (FILAMENTOUS TEMPERATURE SENSITIVE Z1) and FTSZ2 homologues) and the inner PD-ring (of uncertain composition) are present. On the cytosolic side, an outer PD-ring that is composed of polyglucan filaments and a discontinuous ring of dynamin-related ARC5 (ACCUMULATION AND REPLICATION OF CHLOROPLASTS 5) operate. c | Positional information from the stromal Z-ring is conveyed to the cytosolic components through ARC6 and PARC6 in the inner membrane and PDV1 (PLASTID DIVISION 1) and PDV2 in the outer membrane through specific interactions in the intermembrane space (IMS). Correctly positioned PDV proteins enable the recruitment of ARC5 at the constriction zone. Z-ring formation is dynamically controlled by the action of PARC6 and ARC3. IEM, inner envelope membrane; MCD1, MULTIPLE CHLOROPLAST DIVISION SITE 1; OEM, outer envelope membrane. The images are adapted, with permission, from REF.107 (2013) Elsevier and REF.156 (2008) Elsevier.

R E V I E W S




RuBisCo(Ribulose-1,5-bisphosphate carboxylaseoxygenase). The enzyme of the photosynthetic Calvin cycle that is responsible for CO2 fixation and may be the most abundant protein on earth.

SyngamyThe complete fusion of two gametes to form a zygote, which develops into a new organism.

PhytochromesA family of plant and bacterial photoreceptors that are sensitive in particular to red and far-red light and are responsible for the control of developmental responses such as photomorphogenesis.

PhotomorphogenesisA developmental programme in higher plant seedlings that occurs in the light and is characterized by elongation inhibition, leaf expansion and chloroplast development. It is distinct from an alternative programme (skotomorphogenesis or etiolation) that occurs in the dark.

Bundle sheath cellsCells in the interior of leaves that are closely associated with the vasculature and have specialized functions in photosynthate loading of the vasculature. In C4 photosynthesis plants such as maize, they are the exclusive site of CO2 fixation by RuBisCo.

AuxinA plant growth hormone that, although necessary for leaf development, counteracts cytokinin action in aerial tissues and therefore represses greening and chloroplast development.

CytokininA plant hormone that, in aerial tissues, is associated with cell proliferation but also promotes greening and chloroplast development.

Integration of plastid and plant development In higher plants, all cells (including a diversity of non-photosynthetic types) possess plastids, and these all derive from embryonic proplastids. Thus, various plastid types exist, which necessitates the existence of several differentiation and interconversion pathways (FIG.1). These constitute one of the most important, yet least understood, areas of plastid biology.

Chloroplast assembly, light and leaf development. Proplastids in subepidermal meristematic cells (or etioplasts in dark-grown cotyledons) differentiate into mesophyll chloroplasts in the light126. Linear mono-cot leaves exhibit a developmental gradient, with young, proplastid-containing cells at the base near the meri stem and older, differentiated cells towards the tip. This enabled a view of unprecedented detail of cellular and organellar development127129: an early, heterotrophic phase of cellular proliferation and growth; a transition phase during which plastid bio-genesis proteins (for example, the plastid translation apparatus and enzymes involved in the synthesis of tetrapyrroles and isoprenoids) accumulate rapidly; and a maturation phase of large-scale photosynthetic protein accumulation and photosynthetic activity127,128 (FIG.1c). Transcription factors like GLK and HY5 (see below) are active during the transition and early mat-uration phases127. Similar transitions occur during dicot leaf development, although they are less spatially resolved126,130 (FIG.1b).

Light causes a dramatic change in the transcrip-tional programme, with the identity of 50% of expressed genes changing130,131. Light is perceived by two families of informational photoreceptors: phytochrome s and cryptochromes. Before light expo-sure, PIF (PHYTOCHROME-INTERACTING FACTOR) basic helixloophelix (bHLH) transcrip-tion factors repress photomorphogenesis genes, whereas basic Leu zipper (bZIP) transcription factors such as HY5 and HYH activate those genes upon illumina-tion (with both effects mediated through G-box promoter elements)2,131 (FIG.2). In the dark, HY5 is targeted for degradation by the ubiquitin ligase COP1 (CONSTITUTIVE PHOTOMORPHOGENESIS 1) and related repressors of photomorphogenesis, includ-ing DET1 (DEETIOLATED 1) (COP1 is inactivated in the light by phytochromes). Conversely, phyto-chromes phosphorylate PIFs in the light, promoting their UPS-mediated turnover.

Light signalling pathways are tightly intertwined with chloroplast biogenesis: PIF1 and PIF3 are nega-tive regulators of chloroplast development, particu-larly of tetrapyrrole biosynthesis genes132, whereas HY5 promotes chloroplast development133 (FIG.2). Remarkably, one protein, HEMERA (also known as pTAC12), is located in plastids, where it participates in gene expression, and in the nucleus, where it mediates phytochrome degradation upon light exposure134. The expression of plastid DNA-encoded genes is indirectly controlled by light and the circadian clock through the action of -factors (for example, SIG5)135,136.

Differential plastid development. Different cell types require different types of plastid, and we are just begin-ning to understand the basis for these differences12,126. Bundle sheath cells are more sparsely chloroplast-populate d than their mesophyll neighbours, and in C4 plants such bundle-sheath chloroplasts are proteomically distinct and differentiate earlier due to separate tran-scriptional regulatory networks128,137. Root cortex cells normally carry proplastids, but when roots are detached from shoots they show increased greening triggered by an altered balance of hormones (auxin and cytokinin), which acts through GLK2 and HY5 transcription fac-tors133. When chloroplasts differentiate into amylo-plasts (for example, in tubers), photosynthetic genes are repressed and a global decrease in plastid transcriptional and translational activity occurs138.

Chromoplast differentiation in tomato fruit is similarly accompanied by photosynthetic gene repression, as well as by the prevailing expression of carotenoid bio synthetic genes139. Molecular chaperones have an important regula-tory role in the chloroplast-to-chromoplast transition140. A positive regulator of chloroplast and chromoplast bio-genesis in fruit is APRR2-Like (ARABIDOPSIS PSEUDO RESPONSE REGULATOR2-LIKE), which is an intrigu-ing protein that contains pseudoresponse regulator and GLK-related MYB transcription factor domains (similar to cytokinin signalling targets)141. Feedback properties of client-specific protein import pathways, or direct UPS action on the import machinery, may also play a part in plastid conversions41,61,98.

Plastid compartment size determination. The size of the plastid population within a cell is regulated and cell type-specific. In mesophyll cells, a linear relationship exists between the cell plan area and the combined area of its chloroplasts (the cytoplasmic layer occupied by chloro-plasts is effectively a two-dimensional space) (FIG.1b). The chloroplast compartment set-point (that is, the constant value of the chloroplast-to-cell area ratio) can be seen across species and is achieved regardless of size changes that are linked to nuclear polyploidization or plastid divi-sion defects142. When the hormone gibberellin triggers cell expansion, chloroplast division genes and chloroplast growth are indirectly promoted to maintain the chloro-plast complement set-point143. The relationship is broken by mutations that affect photomorphogenesis regulators (for example, DET1)144, but its mechanistic basis is largely unknown. Defects in plastid DNA proliferation prevent normal plastid division and reduce leaf growth, which reveals a role for plastid DNA copy number as a check-point for plastid division and a relationship between plastid proliferation and leaf cell differentiation17. Cell and chloroplast division are also coordinated by a shared component with dual localization145. Our understanding of plastid compartment size determination is clearly in its infancy, despite its importance.

What drives chloroplast development? Various devel-opmental, hormonal and environmental cues influence chloroplast differentiation, but is there a mechanistic driver behind the entire programme? The notion of such

R E V I E W S


R E V I E W S



Nucleus Cytosol

Mitochondrion

80S

Cytochrome cI, III, IV, ATP synthaseTOMHaem synthesisTfam

FA oxidation

PGC1

NRF1

Energy deprivation Exercise Cold

IV

III

I

TfamNEP

Cytochrome c

ATP synthase

TOM

70S

Other TFs

a chloroplast driver is hypothetical, but, as the control of mitochondrial development in mammals reveals (BOX1), not entirely far-fetched. Dramatic examples of ectopic chloroplast differentiation suggest that the chloroplast driver may be an interrelated regulatory network of transcription factors; alternatively, these factors may act downstream of the hypothetical driver. For example, a GLK1 homologue induces greening in rice callus upon

overexpression146, whereas the combined action of HY5 and GLK2 (FIG.2) underlies hormone-dependent root greening133. The tomato uniform ripening phenotype, which results from uniformly pale-green immature fruit, is caused by a mutation in a GLK homologue147, whereas overexpression of a GLK-related gene produces dark-green fruit before ripening141. Other such regulators are the cytokinin-responsive GATA transcription factors GNC and CGA1, which induce ectopic chloroplasts in the hypocotyl epidermis upon overexpression148, and the cytokinin-responsive CRF2 transcription factor, which upregulates PDV proteins and promotes chloroplast division149 (FIG. 2).

Are chloroplast and leaf development linked? Variegated leaf phenotypes reveal a crucial stage in proplastid-to-chloroplast development during which irrecover-able damage may occur, which results in clones of white cells150. Such cells form leaf tissue with defective pali-sade differentiation5. A developmental role for plastids is supported by the crumpled leaf plastid division mutant, which produces cells that lack chloroplasts which leads to abnormal leaf lamina expansion151. However, this effect may be linked to the loss of essential metabolic functions (such as the synthesis of lipids and lipid hor-mones)25,152. The observation that chloroplast differen-tiation and photosynthetic gene expression coincide sharply with the transition of mesophyll cells from pro-liferation to expansion in A.thaliana suggests a direct developmental link153, although such transitions seem to be more gradual in maize129. Reminiscent of mitochon-drial involvement in mammalian apoptosis, a final link occurs during senescence or pathogen response, when the chlorophyll catabolite pheophorbide-a triggers cell death (even in darkness, thus excluding photooxidative effects)154.

The two-way relationship between plastid biogenesis and cellular differentiation, particularly between chlo-roplasts and mesophyll cells, is fascinating but poorly understood. Central to elucidating this relationship will be the integration of different aspects of plastid bio-genesis (including the functions of the genetic, protein import and division machineries) with each other and with the expression of relevant nuclear genes, as well as at appropriate stages of cellular differentiation. Indeed, these biogenic processes clearly coincide during devel-opment, as seen in the maize leaf 127,128. Moreover, they also coincide with the expression of transcription factors that are linked to light responses and plastid-to-nucleus communication.

ConclusionIn recent years, great strides have been made towards understanding the mechanisms that underlie the bio-genesis and homeostasis of plastids. Important break-throughs have included the realization that plastid development is directly regulated by the UPS, the iden-tification of a general TIC translocation complex com-prising novel components (including a component that is encoded by the plastome), and the molecular char-acterization of polyglucan-rich plastid division rings.

Box 1 | Control of organelle development: a mammalian paradigm

Like chloroplasts, mitochondria exist in differing numbers in different cell types (for example, they abound in skeletal muscle), where they assume distinct roles (such as, thermogenesis in brown adipose tissue). They form a cellular compartment which varies in size depending on the external stimuli (for example, exercise or cold) and are essential for all cells except shortlived erythrocytes. The mitochondrial DNA (mtDNA) encodes a small number of mitochondrial proteins (most mitochondrial proteins are nucleus encoded). Replication of mtDNA requires a nucleusencoded factor, Tfam (transcription factor A of mitochondria) (see the figure). Moreover, transcription involves a nucleus encoded RNA polymerase (NEP) and requires a specific mTFB (mitochondrial transcription factor B)-type transcriptional activator and a specificmTERF(mitochondrial transcriptiontermination factor)-type repressor (both are nucleus-encoded); other mTFBand mTERF factors promote translation. In the nucleus, NRF1 (nuclear respiratory transcription factor 1) and NRF2 (not shown) drive the expression of genes that encode components of respiratory chain complexes (I, III, IV), ATP synthase and their assembly factors, as well as Tfam, mTFB, subunits of the TOM(translocase of the outer mitochondrial membrane) import complex and haem biosynthetic enzymes. Thus, NRFs broadly regulate nuclear gene expression for mitochondrial biogenesis directly, and mtDNA replication and expression indirectly.Mouse embryos lacking NRF1 have lowmtDNA content and die before implantation.

Other transcription factors (TFs), such as ERR (oestrogenrelated receptor) and PPAR (peroxisome proliferatoractivated receptor), regulate specific mitochondrial functions, such as fatty acid (FA) oxidation. Importantly, a transcriptional co-activator called PGC1 (PPAR co-activator 1) was identified as a partner of PPAR (not shown), which mediates the differentiation of brown adipose tissue. PGC1 also binds to and activates the functions of NRFs, ERR and PPAR. Overexpression of PGC1 promotes mitochondrial biogenesis and energy metabolism. Cold, exercise and fasting all upregulate transcription of PGC1 or its homologues (or activate them through posttranslational modification), which leads to enhanced mitochondrial content and function. Thus, NRFs control basal mitochondrial biogenesis, and PGC1 functionsas a master switch that responds to upstream signals155. Haem is a biogenic signal for yeast mitochondria and a tetrapyrrole synthesis regulator in mammals40, but how this relates tothe above mechanisms is unknown.

R E V I E W S




We have also learned that the plastid genome can be shared horizontally between plants, identified several metabolites through which plastids exert operational control over nuclear gene expression, revealed that regulators of plastid activities additionally participate in environmental sensing or cell cycle control and found transcription factors that are capable of inducing ectopic chloroplasts. Nonetheless, some fundamental

aspects of plastid development are only just beginning to be understood, such as the mechanisms that regulate chloroplast build-up or the interconversion of different plastid types, as well as the nature of biogenic plastid-to-nucleus signals. We anticipate that future work, building on the exciting breakthroughs discussed in this Review, will shed light in these important areas of organelle biology.

1. Eberhard,S., Finazzi,G. & Wollman,F.A. The dynamics of photosynthesis. Annu. Rev. Genet. 42, 463515 (2008).

2. Waters,M.T. & Langdale,J.A. The making of a chloroplast. EMBO J. 28, 28612873 (2009).

3. Whatley,J.M. A suggested cycle of plastid developmental interrelationships. New Phytol. 80, 489502 (1978).

4. Keeling,P.J. The endosymbiotic origin, diversification and fate of plastids. Phil. Trans. R.Soc. B 365, 729748 (2010).

5. LpezJuez,E. & Pyke,K.A. Plastids unleashed: their development and their integration in plant development. Int. J.Dev. Biol. 49, 557577 (2005).

6. Sakamoto,W., Miyagishima,S.Y. & Jarvis,P. Chloroplast biogenesis: control of plastid development, protein import, division and inheritance. Arabidopsis Book 6, e0110 (2008).

7. Neuhaus,H.E. & Emes,M.J. Nonphotosynthetic metabolism in plastids. Annu. Rev. Plant Physiol. Plant Mol. Biol. 51, 111140 (2000).

8. Rolland,N. etal. The biosynthetic capacities of the plastids and integration between cytoplasmic and chloroplast processes. Annu. Rev. Genet. 46, 233264 (2012).

9. ReyesPrieto,A., Weber,A.P. & Bhattacharya,D. The origin and establishment of the plastid in algae and plants. Annu. Rev. Genet. 41, 147168 (2007).

10. Timmis,J.N., Ayliffe,M.A., Huang,C.Y. & Martin,W. Endosymbiotic gene transfer: organelle genomes forge eukaryotic chromosomes. Nature Rev. Genet. 5, 123135 (2004).

11. Li,H.M. & Chiu,C.C. Protein transport into chloroplasts. Annu. Rev. Plant Biol. 61, 157180 (2010).

12. van Wijk,K.J. & Baginsky,S. Plastid proteomics in higher plants: current state and future goals. Plant Physiol. 155, 15781588 (2011).

13. Phillips,M.A., Len,P., Boronat,A. & RodrguezConcepcin,M. The plastidial MEP pathway: unified nomenclature and resources. Trends Plant Sci. 13, 619623 (2008).

14. Tanaka,R., Kobayashi,K. & Masuda,T. Tetrapyrrole Metabolism in Arabidopsis thaliana. Arabidopsis Book 9, e0145 (2011).

15. Vothknecht,U.C., Otters,S., Hennig,R. & Schneider,D. Vipp1: a very important protein in plastids?! J.Exp. Bot. 63, 16991712 (2012).

16. Bendich,A.J. DNA abandonment and the mechanisms of uniparental inheritance of mitochondria and chloroplasts. Chromosome Res. 21, 287296 (2013).

17. Garton,S., Knight,H., Warren,G.J., Knight,M.R. & Thorlby,G. J. crinkled leaves 8 a mutation in the large subunit of ribonucleotide reductase leads to defects in leaf development and chloroplast division in Arabidopsis thaliana. Plant J. 50, 118127 (2007).

18. Meierhoff,K., Felder,S., Nakamura,T., Bechtold,N. & Schuster,G. HCF152, an Arabidopsis RNA binding pentatricopeptide repeat protein involved in the processing of chloroplast psbB-psbT-psbH-petB-petD RNAs. Plant Cell 15, 14801495 (2003).

19. Kanamaru,K. & Tanaka,K. Roles of chloroplast RNA polymerase factors in chloroplast development and stress response in higher plants. Biosci. Biotechnol. Biochem. 68, 22152223 (2004).

20. Hricova,A., Quesada,V. & Micol,J.L. The SCABRA3 nuclear gene encodes the plastid RpoTp RNA polymerase, which is required for chloroplast biogenesis and mesophyll cell proliferation in Arabidopsis. Plant Physiol. 141, 942956 (2006).

21. Hajdukiewicz,P.T., Allison,L.A. & Maliga,P. The two RNA polymerases encoded by the nuclear and the plastid compartments transcribe distinct groups of genes in tobacco plastids. EMBO J. 16, 40414048 (1997).

22. Qiao,J., Ma,C., Wimmelbacher,M., Bornke,F. & Luo,M. Two novel proteins, MRL7 and its paralog MRL7L, have essential but functionally distinct roles in chloroplast development and are involved in plastid gene expression regulation in Arabidopsis. Plant Cell Physiol. 52, 10171030 (2011).

23. Kotera,E., Tasaka,M. & Shikanai,T. A pentatricopeptide repeat protein is essential for RNA editing in chloroplasts. Nature 433, 326330 (2005).

24. SchmitzLinneweber,C. & Small,I. Pentatricopeptide repeat proteins: a socket set for organelle gene expression. Trends Plant Sci. 13, 663670 (2008).

25. Babiychuk,E. etal. Plastid gene expression and plant development require a plastidic protein of the mitochondrial transcription termination factor family. Proc. Natl Acad. Sci. USA 108, 66746679 (2011).

26. Majeran,W. etal. Nucleoidenriched proteomes in developing plastids and chloroplasts from maize leaves: a new conceptual framework for nucleoid functions. Plant Physiol. 158, 156189 (2012).

27. Pfalz,J., Liere,K., Kandlbinder,A., Dietz,K. J. & Oelmuller,R. pTAC2, 6, and 12 are components of the transcriptionally active plastid chromosome that are required for plastid gene expression. Plant Cell 18, 176197 (2006).

28. MarinNavarro,J., Manuell,A.L., Wu,J. & Mayfield,S.P. Chloroplast translation regulation. Photosynthesis Res. 94, 359374 (2007).

29. Barbrook,A.C., Howe,C.J. & Purton,S. Why are plastid genomes retained in nonphotosynthetic organisms? Trends Plant Sci. 11, 101108 (2006).

30. Allen,J.F., de Paula,W.B., Puthiyaveetil,S. & Nield,J. A structural phylogenetic map for chloroplast photosynthesis. Trends Plant Sci. 16, 645655 (2011).

31. Pfannschmidt,T., Nilsson,A. & Allen,J.F. Photosynthetic control of chloroplast gene expression. Nature 397, 625628 (1999).

32. Bonardi,V. etal. Photosystem II core phosphorylation and photosynthetic acclimation require two different protein kinases. Nature 437, 11791182 (2005).

33. Puthiyaveetil,S. etal. The ancestral symbiont sensor kinase CSK links photosynthesis with gene expression in chloroplasts. Proc. Natl Acad. Sci. USA 105, 1006110066 (2008).

34. Shimizu,M. etal. factor phosphorylation in the photosynthetic control of photosystem stoichiometry. Proc. Natl Acad. Sci. USA 107, 1076010764 (2010).Provides an elegant example of an adaptive response that involves dual control of a chloroplast-encoded PSI gene. Demonstrates that plastid transcription is regulated by a nucleus-encoded -factor, the activity of which is regulated by a kinase that is in turn controlled by the photosynthetic apparatus.

35. Huang,C.Y., Ayliffe,M.A. & Timmis,J.N. Direct measurement of the transfer rate of chloroplast DNA into the nucleus. Nature 422, 7276 (2003).

36. Stegemann,S. & Bock,R. Exchange of genetic material between cells in plant tissue grafts. Science 324, 649651 (2009).

37. Thyssen,G., Svab,Z. & Maliga,P. Celltocell movement of plastids in plants. Proc. Natl Acad. Sci. USA 109, 24392443 (2012).

38. Bradbeer,J.W., Atkinson,Y.E., Brner,T. & Hagemann,R. Cytoplasmic synthesis of plastid polypeptides may be controlled by plastidsynthesized RNA. Nature 279, 816817 (1979).

39. Inaba,T., Yazu,F., ItoInaba,Y., Kakizaki,T. & Nakayama,K. Retrograde signaling pathway from plastid to nucleus. Int. Rev. Cell. Mol. Biol. 290, 167204 (2011).

40. Terry,M.J. & Smith,A.G. A model for tetrapyrrole synthesis as the primary mechanism for plastidtonucleus signaling during chloroplast biogenesis. Front. Plant Sci. 4, 14 (2013).

41. Kakizaki,T. etal. Coordination of plastid protein import and nuclear gene expression by plastidtonucleus retrograde signaling. Plant Physiol. 151, 13391353 (2009).

42. Woodson,J.D., PerezRuiz,J.M., Schmitz,R.J., Ecker,J.R. & Chory,J. factormediated plastid retrograde signals control nuclear gene expression. Plant J. 73, 113 (2013).

43. Koussevitzky,S. etal. Signals from chloroplasts converge to regulate nuclear gene expression. Science 316, 715719 (2007).

44. Strand,A., Asami,T., Alonso,J., Ecker,J.R. & Chory,J. Chloroplast to nucleus communication triggered by accumulation of MgprotoporphyrinIX. Nature 421, 7983 (2003).

45. Mochizuki,N., Tanaka,R., Tanaka,A., Masuda,T. & Nagatani,A. The steadystate level of Mgprotoporphyrin IX is not a determinant of plastidtonucleus signaling in Arabidopsis. Proc. Natl Acad. Sci. USA 105, 1518415189 (2008).

46. Moulin,M., McCormac,A.C., Terry,M.J. & Smith,A.G. Tetrapyrrole profiling in Arabidopsis seedlings reveals that retrograde plastid nuclear signaling is not due to Mgprotoporphyrin IX accumulation. Proc. Natl Acad. Sci. USA 105, 1517815183 (2008).

47. Ruckle,M.E., DeMarco,S.M. & Larkin,R.M. Plastid signals remodel light signaling networks and are essential for efficient chloroplast biogenesis in Arabidopsis. Plant Cell 19, 39443960 (2007).

48. Voigt,C. etal. Indepth analysis of the distinctive effects of norflurazon implies that tetrapyrrole biosynthesis, organellar gene expression and ABA cooperate in the GUNtype of plastid signalling. Physiol. Plant 138, 503519 (2010).

49. Woodson,J.D., PerezRuiz,J.M. & Chory,J. Heme synthesis by plastid ferrochelatase I regulates nuclear gene expression in plants. Curr. Biol. 21, 897903 (2011).Demonstrates that the FC1-derived pool of haem (a metabolite with organelle-to-nucleus communication roles in other organisms) acts as a chloroplast-to-nucleus signal in plants and is a positive regulator of photosynthetic genes.

50. Sun,X. etal. A chloroplast envelopebound PHD transcription factor mediates chloroplast signals to the nucleus. Nature Commun. 2, 477 (2011).Identifies a mechanism for the relay of information on plastid status to the nucleus: the transcription factor PTM, which is initially anchored in the chloroplast envelope, is cleaved from its membrane-anchoring domain which enables it to travel to the nucleus and to (indirectly) elicit nuclear photosynthetic gene responses.

51. Waters,M.T. etal. GLK transcription factors coordinate expression of the photosynthetic apparatus in Arabidopsis. Plant Cell 21, 11091128 (2009).

52. Brutigam,K. etal. Dynamic plastid redox signals integrate gene expression and metabolism to induce distinct metabolic states in photosynthetic acclimation in Arabidopsis. Plant Cell 21, 27152732 (2009).

53. Pesaresi,P. etal. Arabidopsis STN7 kinase provides a link between sh

Jarvis 2013

Documents