Chloroplasts autophagy during senescence of individually darkened leaves

Abstract

We recently reported that autophagy plays a role in chloroplasts degradation in individually-darkened senescing leaves. Chloroplasts contain approximately 80% of total leaf nitrogen, mainly as photosynthetic proteins, predominantly ribulose 1, 5-bisphosphate carboxylase/oxygenase (Rubisco). During leaf senescence, chloroplast proteins are degraded as a major source of nitrogen for new growth. Concomitantly, while decreasing in size, chloroplasts undergo transformation to non-photosynthetic gerontoplasts. Likewise, over time the population of chloroplasts (gerontoplasts) in mesophyll cells also decreases. While bulk degradation of the cytosol and organelles is mediated by autophagy, the role of chloroplast degradation is still unclear. In our latest study, we darkened individual leaves to observe chloroplast autophagy during accelerated senescence. At the end of the treatment period chloroplasts were much smaller in wild-type than in the autophagy defective mutant, atg4a4b-1, with the number of chloroplasts decreasing only in wild-type. Visualizing the chloroplast fractions accumulated in the vacuole, we concluded that chloroplasts were degraded by two different pathways, one was partial degradation by small vesicles containing only stromal-component (Rubisco containing bodies; RCBs) and the other was whole chloroplast degradation. Together, these pathways may explain the morphological attenuation of chloroplasts during leaf senescence and describe the fate of chloroplasts.

This article refers to:

The most abundant chloroplast protein is Rubisco, comprising approximately 50% of the soluble protein.1 The amount of Rubisco decreases rapidly in the early phase of leaf senescence, and more slowly in the later phase. During senescence, chloroplasts gradually shrink and their numbers gradually decrease in mesophyll cells.2,3 During leaf senescence, leaves lose approximately 75% of their Rubisco, while chloroplast numbers decrease by only about 15%.4 Previous studies showed chloroplasts localized within the central vacuole by electron microscopy, indicating chloroplast degradation in the highly hydrolytic vacuole.5 However, there was no direct evidence showing translocation of chloroplasts from the cytosol to the vacuole, and the mechanism of transportation was also unclear.

Recent reverse genetic approaches are helping to elucidate the autophagy system in plants, which has a similar molecular mechanism as in yeast.6–11 In Arabidopsis (Arabidopsis thaliana), atg mutants have phenotypically accelerated leaf senescence, insufficient root elongation in nutrient starvation condition and reduced seeds yields, therefore, autophagy is considered to be important for nutrient recycling especially nutrient starvation and senescence in plants.12

In Arabidopsis, individually darkened rosette leaves (IDLs) exhibit enhanced senescence.13 Appling IDLs treatment as an experimental model of leaf senescence, we recently demonstrated that chloroplasts are degraded in two different pathways by autophagy, one for RCBs,14,15 and one for whole chloroplast.16 Darkened leaves became pale in 3 to 5 days treatment, while illuminated parts normally grow in both wild-type and autophagy defective mutant, atg4a4b-1. Furthermore, genes specifically expressed during senescence, SAG12 and SEN1, were rapidly upregulated, meanwhile, photosynthetic genes, such as RBCS2B and CAB2B, were gradually downregulated. All analyzed ATG genes were also upregulated under IDL treatment, which suggests that autophagy is important in IDL senescence. It has been reported that approximately three quarter genes of upregulated in IDL were also upregulated in naturally senescing leaves, including the ATG genes.17 This suggests that the autophagy pathways used in IDLs are also used in naturally senescing leaves.

Over the 5 day treatment period, chloroplasts of wild-type IDL shrink to approximately one third their original size. In atg4a4b-1, by contrast, chloroplasts shrinkage occurred immediately after the start of IDL treatment after which no further shrinkage was noted. While the shrunk chloroplasts in fixed cells of wild-type were still smooth and round, while wrinkly chloroplasts were observed in atg4a4b-1. At same time, in the living mesophyll cells of wild-type IDL, RCBs accumulated in the vacuole (Fig 1B). The shrinkage of chloroplasts may be due to the consumption of the chloroplast envelope by RCB formation. Immunological quantification of inner and outer envelope proteins might confirm this hypothesis. The chloroplast number was also gradually decreased in IDL of wild-type plants, but no decline in chloroplast number was noted in atg4a4b-1. Chloroplasts exhibiting chlorophyll auto-fluorescence were found in the vacuole of wild-type IDLs, but not in atg4a4b-1 IDLs. These results show that whole chloroplast degradation is also performed by autophagy. However, the transport pathway of whole chloroplasts into the vacuole remains unclear. The chloroplast, even in its shrunken state, is a large organelle, and the autophagosome, the carrier bodies of autophagy, which usually target small spherical organelles like mitochondria and peroxisomes, may be incapable of isolating large organelles. In the yeast autophagy system, specific cellular organelles and fractions are also transported via vacuolar membrane invagination using the microautophagy system.18 RCB uptake into the vacuole is termed macroautophagy, while larger organelles, such as chloroplasts, are engulfed in a process known as microautophagy. Whether there exists a molecular difference between these processes, or whether this is an arbitrary division based solely on the size of the consumed body is unclear.

Whole darkened plants exhibit retarded leaf aging, in contrast to the accelerated senescence in IDLs.13 Whole darkened plants suppress leaf senescence with the leaves retaining green color. After 5 days, in the mesophyll cells of whole darkened plants, any translocation of chloroplast components, stroma-targeted DsRed, RCBs, and whole chloroplasts, into the vacuole could hardly be detected (Fig. 1C). This suggests that autophagy is not induced by darkness alone, and is associated closely with senescence. ATG genes were downregulated in the whole darkened wild-type plants less than control plants during the treatment. Previous studies have shown that following about 5 day period of whole plant darkening, atg mutants lose their ability to protect themselves against photo-damage.7 Upon return to the light, these plant quickly undergo terminal photo-bleaching.

Concentrations of chlorophyll, soluble protein, leaf nitrogen and Rubisco rapidly declined under IDL condition of both wild-type and atg4a4b-1. Considering the accumulated fluorescence of stroma-targeted Ds-Red in the vacuole and autophagy dependent size shrinkage of chloroplasts in IDL, in wild-type plants RCB autophagy appear to be responsible for a sizable proportion of chloroplast protein degradation. In atg4a4b-1 which cannot form RCBs, alternative degradation pathways must be upregulated, with chloroplast proteases the most likely candidates. Intriguingly, the decrease in Rubisco concentration proceeds at the almost identical rates in both wild-type and atg4a4b-1 plants, despite the different degradation pathways. It seems likely that the rate of Rubisco degradation may be regulated at an early step in the degradation pathway, by some, as yet unknown, factors.

Chloroplasts appear to have the ability to control their volume during cell division, dividing and increasing their density up to the certain level,19 and transferring their cellular components between them via stromules.20 How chloroplasts are able to regulate their volume remains unclear, but it seems likely that chloroplasts grow and divide, like any other bacteria, as long as sufficient resources remain in the environment, in this case the cell. Total chloroplast volume, therefore, may be limited by the availability of carbon, nitrogen, or other nutrients in the cell during leaf emergence. Chloroplasts may be also able to reduce and control their volumes during leaf senescence via multiple degradation pathways. Our next goal is to estimate the contribution of both RCBs and whole chloroplasts autophagy in chloroplast protein degradation during natural leaf senescence. Further investigations are required for understanding the specific molecular mechanisms of RCB production and whole chloroplast degradation.

Figures and Tables

Figure 1 Visualization of stroma-targeted DsRed and chlorophyll autofluorescence in living mesophyll cells of wild-type plants by laser-scanning confocal microscopy. A excised control leaf (A, Light) and an individually darkened leaf (B, IDL) from plants grown under 14 h-photoperiod condition and a leaf from whole-plant darkened condition (WD, C) for 5days were incubated with 1 µM concanamycin A in 10 mM MES-NaOH (pH 5.5) at 23C° for 20 h in darkness. Stroma-targeted DsRed appears green and chlorophyll fluorescence appears red. In merged images, overlap of DsRed and chlorophyll fluorescence appears yellow. Small vesicles with stromal-targeted DsRed, i.e. RCBs, can be found in the vacuole (A, B). In IDL (B), massive accumulation of stroma-targeted DsRed is entirely seen in the vacuolar lumen and chloroplasts losing DsRed fluorescence are found in some cells. Bars = 50 µm.

Acknowledgements

We thank Dr. Louis Irving for critical reading of the manuscript. This research was supported by KAKENHI (nos. 19039004, 20200061 and 20780044) to H.I.

Addendum to: