Sucrose Starvation Induces Microautophagy in Plant Root Cells

Abstract

Autophagy is an essential system for degrading and recycling cellular components for survival during starvation conditions. Under sucrose starvation, application of a papain protease inhibitor E-64d to the Arabidopsis root and tobacco BY-2 cells induced the accumulation of vesicles, labeled with a fluorescent membrane marker FM4-64. The E-64d-induced vesicle accumulation was reduced in the mutant defective in autophagy-related genes ATG2, ATG5, and ATG7, suggesting autophagy is involved in the formation of these vesicles. To clarify the formation of these vesicles in detail, we monitored time-dependent changes of tonoplast, and vesicle accumulation in sucrose-starved cells. We found that these vesicles were derived from the tonoplast and produced by microautophagic process. The tonoplast proteins were excluded from the vesicles, suggesting that the vesicles are generated from specific membrane domains. Concanamycin A treatment in GFP-ATG8a transgenic plants showed that not all FM4-64-labeled vesicles, which were derived from the tonoplast, contained the ATG8a-containing structure. These results suggest that ATG8a may not always be necessary for microautophagy.

Keywords: E-64d; FM4-64; autophagy-related genes; microautophagy; microdomain; sucrose starvation; tonoplast; vacuole.

Figure 1

The phenotypes of E-64d vesicles in peup mutants. (A)Arabidopsis roots were stained with FM4-64 with or without E-64d for 24 h in the wild-type (WT). TZ, transition zone; EZ, elongation zone. Arrowheads indicate the aggregates. (B) Magnified and inverted image of the vesicle aggregates stained with FM4-64. Bar = 2 µm. (C) The phenotypes of the accumulation of vesicles after the 24 h treatment of E-64d and FM4-64 in the WT and peup mutants. Bar = 50 µm.

Figure 2

Identification of PEUP17 and PEUP22 genes. (A) A workflow of the identification of the causative genes in the peup17 and peup22 mutants. The information about (I) the number of polymorphisms detected with whole-genome sequencing, (II) the mapped region from map-based cloning, and (III) the number of candidate polymorphisms after filtering with the criteria of possibility to change the amino acid sequence are shown in Table 1. (B) Allelism test between peup17 and atg5-1 and peup22 and atg7-2. The phenotype of E-64d vesicle formation was assessed in the root of the WT and F1 seedlings. The vesicles were stained with FM4-64 with E-64d for 24 h in the 6-day-old seedlings. Bar = 50 µm. (C) Schematic structure of PEUP17/ATG5 and PEUP22/ATG7 genes. The black and grey boxes indicate untranslated regions and exons, respectively. The peup17 mutation substitutes the guanine to adenine at the splice-donor site in the third intron, and the peup22 mutation substitutes the cytosine to thymine in the eighth exon and causes the substitution of glutamine at position 522 with a stop codon. The peup4 is an allele of peup22 (Shibata et al., 2013). The T-DNA insertion lines, atg5-1 and atg7-2, are indicated by triangles. Arrows and numbers indicate the position of the primers used in following RT-PCR. (D–F) RT-PCR with cDNAs synthesized from the mRNAs of 7-day-old seedling of peup17 and peup22. The positions of primers are indicated in (C). (F) Indicates UBQ10 gene expressions. The sequence of the primers is in Supplementary Table 1 . Arrowheads, expected size of bands in WT; asterisks, the unusual length of PCR products.

Figure 3

autophagy-related phenotypes in the peup mutants. (A) Immunoblot analysis of the peup and atg mutants. Crude protein extracts of the WT, atg5-1, peup17, atg7-2, and peup22 from 7-day-old seedlings were subjected to SDS-PAGE and immunoblot analysis with anti-ATG5, and anti-ATG7, antibodies. Equal protein loading was confirmed by immunoblot analysis with Coomassie Brilliant Blue (CBB) staining. An asterisk indicates unknown bands specific to the anti-ATG7 antibody. (B) Senescence phenotypes of the peup and atg mutants. Arabidopsis seedlings were grown on sucrose deprivation media for 7 days and transferred to the darkness for 8 days. Photos were taken before and after 8 days of dark treatment and carbon deprivation. (C) Chlorophyll content in the WT and mutants. Chlorophylls were extracted from five seedlings, and six biological repeats were prepared for each plant lines. Chlorophyll content in WT was set as 1.0. Bar = ± SE, n = 6.

Figure 4

The phenotypes in E-64d vesicle formation in WT and atg mutants. (A) Inverted confocal micrographs of FM4-64 signals from 6-day-old seedlings treated with or without E-64d for 24 h under starvation. Images of the wide area from three replicates were shown in Supplementary Figure 5 . Bar = 20 µm. (B) Quantification analysis of the area of aggregates in the WT and mutants upon E-64d treatment for 24 h. The area size per cell (µm²/cell) was obtained from ImageJ software. For making the graph, 8–10 cells from the transition zone were used in each plant, and three plants were used for each line. Bar = ± SE.

Figure 5

E-64d vesicles in sucrose-starved tobacco cells. (A) E-64d vesicles are generated from the tonoplast under starvation. The tonoplast was visualized with FM4-64 before the induction of starvation and E-64d treatment (left). Arrows indicate the invagination sites of tonoplast observed after 5 h of starvation and the treatment with E-64d (middle). Arrowheads indicate E-64d vesicles observed after 20 h of the treatment (right). Bar = 10 µm; n, nucleus; t, tonoplast. (B) Cytosolic acid granules are captured into E-64d vesicles. FM4-64 and BCECF staining show tonoplast and acidic compartments, respectively. Arrows indicate the cytosolic acid granules. Arrowheads indicate E-64d vesicles. Bar = 10 µm; n, nucleus. (C) High-magnification images showing that the cytosolic acid granules (arrows) are captured by tubular invagination of the tonoplast to form E-64d vesicles (arrowheads) after 5 h of treatment with E-64d. Bar = 5 µm. (D) Concanamycin A (ConA) treatment did not interfere with E-64d vesicle formation. Cells were treated with (upper panel) or without (lower panel) ConA. Magenta and green colors show FM4-64 and quinacrine fluorescence, respectively. Arrowheads indicate E-64d vesicles. Bar = 10 µm. n, nucleus.

Figure 6

Simultaneous visualization of tonoplast protein and E-64d vesicles. (A) Confocal micrographs of root cells from Venus-VAM3 transgenic plants treated with E-64d under sucrose starvation. The tonoplast was visualized with FM4-64 before the induction of starvation and applying E-64d, and observed after 1, 5, and 10 h of the treatments. Magenta and green indicate the signals of FM4-64 and Venus, respectively. Bar = 10 µm. (B) Three-dimensional (3-D) image of the Venus-VAM3 root cell after 5 h of the treatments, which was reconstructed from a z-stack confocal microscopic images (0.3830 µm interval) with the ImageJ 3D Viewer plugin (https://imagej.nih.gov/ij/plugins/3d-viewer/). (C) An FM4-64–labeled vesicle is released from the tonoplast. The images were extracted from Supplementary Movie 2 . The tonoplast in the Venus-VAM3 cell was pre-stained with FM4-64, and then cells were starved with E-64d treatment for 6 h. The elapsed time is indicated at the lower left corner of each image. White and blue arrows indicate the vesicles stained with both FM4-64 and Venus-VAM3 and only with FM4-64, respectively. Bar = 2 µm.

Figure 7

Subcellular localization of GFP-ATG8a during starvation and E-64d treatment. (A) Confocal micrographs of root cells from GFP-ATG8a transgenic plants treated with E-64d under sucrose starvation. The tonoplast was visualized with FM4-64 before the induction of starvation. Cells were starved and treated with E-64d or ConA for 3 h. Magenta and green indicate the signals of FM4-64 and GFP, respectively. Bar = 10 µm. (B) GFP-ATG8a–labeled cytosolic autophagosomes were trapped to the tonoplast and taken into the vacuole. The tonoplast in the GFP-ATG8a cell was pre-stained with FM4-64, and then cells were starved with E-64d treatment for 5 h. The images were extracted from Supplementary Movie 6 . The elapsed time is indicated at the lower left corner of each image. White arrowheads indicate a captured autophagosome, and a black arrowhead indicates a vesicle which started random motion in the vacuole. Snapshot images of the longer time period are shown in Supplementary Figure 10 . Bar = 2 µm. (C) Magnified image of the area indicated in the bottom image of (B) Bar = 2 µm. (D) Magnified images of a GFP-ATG8a cell treated with ConA under starvation for 3 h. Bar = 2 µm. v, vacuole; c, cytosol. (E) Histogram plots showing the profiles of the relative intensity of GFP-ATG8a of the cytosol (grey), FM4-64–labeled vesicles in the vacuole (red), and autophagosomes in the cytosol (blue) from the GFP-ATG8a root cells treated with ConA under starvation. Fluorescent intensities in the vesicles and the cytosol were measured with ImageJ and normalized with the mean of three to five points of cytosolic fluorescence in each cell. Frequency is shown in a percentage of each total of counts. Error bars show ± SD in the profiles of “cytosol” and “autophagosomes in the cytosol.” Total 23 cells from more than six plants were analyzed.