Organelle Signaling: How Stressed Chloroplasts Communicate with the Nucleus

Abstract

Plastids are able to relay information to the nucleus to regulate stress responses. A new genetic screen has identified an isoprenoid intermediate that accumulates in stressed plastids and acts as a novel retrograde signal.

Over one billion years ago, an early eukaryote engulfed a cyanobacterium-like ancestor and the plastid was born. Most of the bacterial genes were either lost or transferred to the nucleus, but today plastid genomes still encode a set of rRNAs, tRNAs, and about 80 proteins. On average, over 95% of the ~3,000 proteins in the plastid are encoded in the nucleus, translated in the cytoplasm, and then imported into the organelle [1,2]. This genetic heterogeneity is also reflected at the protein level, as many multi-subunit complexes, including the transcription, translation, and photosynthetic machinery, contain subunits that are encoded by both genomes. Thus, at a very basic level, communication between the spatially separated nucleus and organelle is necessary to regulate the plastid proteome and protein complex stoichiometry. Although the nucleus is in control of organelle function, plastids can signal back to the nucleus to regulate gene expression using retrograde signals [3,4]. Such communication likely involves metabolites produced within plastids, but the identities of these molecules, what regulates their production, and the signaling pathways are mostly unclear. Although retrograde signaling was once believed to be controlled by a single ‘plastid factor,’ it is now overwhelmingly clear that multiple signals exist. A new study, recently published in Cell, sheds new light on these signaling mechanisms, and identifies a novel retrograde signaling mechanism between the plastid and the nucleus.

One type of retrograde signal, a ‘biogenic’ signal, is used by plastids as they develop into chloroplasts (plastids specialized for photosynthesis) in young photosynthetic tissue. Such signaling can be easily demonstrated in the laboratory using drugs that inhibit organelle function and development, leading to photobleaching [5]. When chloroplast development is blocked in this way, the nucleus responds by greatly reducing the expression of hundreds of genes involved in photosynthesis and chloroplast biogenesis [6]. Genetic screens have identified the Arabidopsis genomes uncoupled (gun) mutants that still express these nuclear genes even when photobleached [7]. These studies have implicated the branch point of the heme/chlorophyll biosynthesis pathway in the plastid as a source of a retrograde signal(s) [8]. Although heme is known to exit the plastid [9] and is a mitochondrial retrograde signal in yeast [10], how it or other chlorophyll intermediates are shuttled to the cytoplasm to interact with signaling factors is not known. A downstream target of this pathway may be the ABI4 transcription factor, which binds a conserved promoter motif found in many of the regulated genes [6].

Although not always as easy to demonstrate empirically, plastids also emit ‘operational’ signals in response to external cues and stress (Figure 1) [11]. Unlike gun signals, operational signals control nuclear genes to limit and repair damage from reactive oxygen species generated by stresses such as excess light and drought. As such, these signals are used to ensure photosynthetic efficiency and cell viability in an ever-fluctuating environment. In the new manuscript by Xiao et al. [12], a novel operational retrograde signaling pathway is described where methylerythritol cyclodiphosphate (MEcPP), a precursor of isoprenoids produced by the plastid methylerythritol phosphate (MEP) pathway, accumulates during different stresses to regulate the expression of nuclear stress-responsive genes. This work clearly describes how plastids can sense external cues by altering levels of a specific metabolite that is then capable of relaying information to the nucleus.

Figure 1.

Three mechanisms for plastid retrograde signaling during stress.

High light and drought stress can lead to the production of several metabolites within plastids that act as messengers to regulate nuclear gene expression. Singlet oxygen (¹O₂) production occurs in high light when excess electrons are transferred from chlorophyll (Chl) molecules in photosystem II (PSII) to molecular oxygen. This leads to the oxidation of β-carotene to produce β-cyclocitral, which can then regulate ¹O₂-responsive genes in the nucleus. Although its role is unknown, the plastid Executor1 (EXE1) protein is required for normal ¹O₂ signaling [19]. High light or drought stress also leads to the accumulation of 3′-phosphoadenosine 5′-phosphate (PAP) and methylerythritol cyclodiphosphate (MEcPP). PAP is able to travel through the cytoplasm to the nucleus where it interacts with 5′ to 3′ exoribonucleases (XRNs) to regulate stress responsive genes. SAL1 regulates PAP levels in the plastid by converting PAP to AMP. MEcPP is an intermediate of the plastid methylerythritol phosphate (MEP) pathway and its accumulation during stress is specifically responsible for the regulation of nuclear stress genes. In bacteria, MEcPP destabilizes histone-like-DNA complexes, suggesting a possible model for gene regulation. Figure adapted from [12,14].

The authors begin their study with a relatively simple genetic screen to isolate mutants with a constitutive stress response. Using a transgenic line of Arabidopsis that expresses firefly luciferase from the promoter of the stress-inducible marker gene HPL, they isolate mutants that constitutively express their marker gene without additional stresses. One such mutant, named ceh1 for ‘constitutively expressing HPL’, was mapped to a gene encoding the previously characterized plastid-localized HDS, which catalyzes the conversion of MEcPP to hydroxymethylbutenyl diphosphate, the bottleneck step in the MEP pathway.

In addition to high HPL expression, the ceh1 mutant also has elevated levels of salicylic acid (SA) and transcript levels of ICS1, which encodes a key enzyme in the SA biosynthesis pathway. The ceh1 mutant is also resistant to the biotrophic pathogen Pseudomonas syringae, which is likely due to the increased stress responses of the mutant. However, these stress responses in ceh1 appear to be specific. Another stress-induced signal molecule, jasmonic acid, is not induced in ceh1, nor is ceh1 more resistant to the necrotrophic pathogen Botrytis cinerea. The ceh1 mutation does not appear to affect genes encoding photosynthesis-related proteins, making it distinct from the classic gun mutants. Unfortunately, the authors did not perform a genome-wide expression analysis of ceh1. Thus, it is unclear what other genes and pathways may be regulated by the ceh1 retrograde signal, or if there is any overlap with other known retrograde signaling pathways. HPL is also induced in the alx8 mutant, which accumulates the retrograde signal 3′-phosphoadenosine 5′-phosphate (PAP) [13], suggesting that these two signals may converge or interact.

Next, the authors set out to determine how a disruption in the MEP pathway could lead to a retrograde signal. By systematically lowering the expression of most of the genes in the MEP pathway, they were able to show that only lines with reduced expression of HDS exhibited the retrograde phenotype of increased HPL transcripts and SA levels. Moreover, only the lines with reduced HDS had elevated levels of MEcPP, which was found to be the MEP intermediate with the greatest levels of fluctuation (up to a 700-fold increase). Finally, as confirmation, the authors showed that direct exogenous application of MEcPP to wild-type plants was able to induce HPL expression. Together, these results established that the retrograde signal in ceh1 was not due to a general disruption in the MEP pathway, but was a result of the specific accumulation of MEcPP due to a bottleneck step in the biosynthetic pathway. To show that this new retrograde signal was activated under physiologically relevant conditions, the authors subjected plants to two basic stresses: wounding and high light. In both cases, these stresses increased both MEcPP levels and HPL expression.

Together, these results clearly demonstrate how an abiotic stress can lead to a metabolic change within the plastid that then acts as a retrograde signal to regulate nuclear gene expression. Yet, one large question still remains: is MEcPP the actual signal molecule that exits the plastid, travels to the cytoplasm or nucleus, and interacts with protein factors to regulate HPL and other genes? Unlike the study by Estavillo et al. [14] where PAP was shown to move from the plastid to the nucleus, these authors mention no attempt to track MEcPP migration. The observation that MEcPP affects chromatin remodeling in chlamydial cultures by disrupting the association of histone-like proteins with DNA offers the possibility that MEcPP has a function in the nucleus [15,16]. Surely future studies will aim to test this possibility and to elucidate the mechanism of MEcPP in signaling.

Other recent work has suggested that MEcPP is probably not the only retrograde signal used by stressed plastids (Figure 1). The metabolite PAP also accumulates in Arabidopsis plastids in response to high light or drought stress and is able to regulate stress-responsive genes in the nucleus [14]. PAP levels are regulated by the chloroplast/mitochondrial-localized protein SAL1, which is a phosphatase that converts PAP to AMP. PAP appears to be capable of traveling to the nucleus, because a nuclear-localized SAL1 protein is able to decrease total PAP levels and turn off PAP-induced nuclear gene expression. Once in the nucleus, PAP may regulate gene expression by inhibiting 5′ to 3′ exoribonucleases (XRNs). As such, PAP is the first plastid retrograde signaling candidate that has been convincingly shown to fit the classical definition of a signaling molecule — one that is generated in one compartment and travels to another to elicit a response. A third molecule, the volatile compound β-cyclocitral, was also shown to accumulate during high light stress [17]. β-Cyclocitral is formed though the oxidation of β-carotene by singlet oxygen, another plastid retrograde signal candidate that accumulates due to excess light [18]. The genomic response of Arabidopsis to an exogenous application of β-cyclocitral was startlingly similar to the classic singlet oxygen signal (over 80% overlap in regulated genes) [18], suggesting that β-cyclocitral, rather than short-lived singlet oxygen, could be the signal molecule that exits the plastid to regulate nuclear gene expression.

The research mentioned herein should help to significantly advance our understanding of plastid–nuclear communication. Several new retrograde signaling molecules have been identified, and the regulation of their production has at least been partially explained. Still, we are still almost completely blind about how these signals are transmitted to the nucleus, their mechanism of action, and how the cell integrates them with one another and other cell/organelle functions. Answering these questions will likely be a challenge considering that many of the biosynthetic pathways involved are essential, which limits our ability to generate loss-of-function mutants. Furthermore, the essential and integrated nature of the plastid leads to pleiotropic affects when its function is compromised, which can lead to unclear or ambiguous results. Nonetheless, it is now clear that plastid retrograde signals control important and specific events in cell physiology and these studies will provide useful tools to unravel their mechanisms.