Autophagy-mediated compartmental cytoplasmic deletion is essential for tobacco pollen germination and male fertility

Abstract

In plants, macroautophagy/autophagy has mainly been associated with stress-related processes but how it impacts normal physiological and developmental processes remains largely unexplored. Pollen germination is the critical first step toward fertilization in flowering plants. It is metabolically demanding and relies on high levels of cytoplasmic reorganization activities to support a dramatic morphological transformation that underlies the development of a pollen tube as the conduit to deliver sperm for fertilization. The role of autophagy in this process remains unclear. Here we provide evidence that pollen germination is accompanied by elevated autophagic activity and successful pollen tube emergence depends on autophagy-mediated cytoplasmic deletion. Genetic and cytological experiments demonstrate that inhibition of autophagy prevents pollen germination while induces the persistence of a layer of undegraded cytoplasm at the germination aperture. Together, these results unveil a novel compartmentalized autophagy. Furthermore, high-throughput comparative lipidomic analyses show that suppressed autophagy-induced inhibition of pollen germination is accompanied by altered profiles of stored and signaling lipids. Proteomic analyses reveal that autophagy likely exert its role in pollen germination via downstream mitochondria-related pathways. These findings reveal a critical role for autophagy in initiating pollen germination and provide evidences for compartmental cytoplasmic deletion being crucial for male fertility. Abbreviations: 3-MA: 3-methyladenine; ATG: autophagy-related gene; Cer: ceramide; CL: cardiolipin; Con A: concanamycin A; DAG: diradylglycerol; GO: gene ontology; HAG: hour after germination; LC-MS: liquid chromatography-mass spectrometry; MAG: min after germination; MDC: monodansylcadaverine; PE: phosphatidylethanolamine; PI: phosphatidylinositol; PLD: phospholipase D; PtdIns3K: phosphatidylinositol 3-kinase; RT-qPCR: quantitative real-time reverse transcription PCR; TAG: triradylglycerol; TEM: transmission electron microscopy; TMT: tandem mass tagging.

Keywords: Autophagy; autophagy-related genes; lipidomics; male sterility; pollen germination; tobacco.

Figure 1.

Dynamic involvement of autophagosomes during the process of pollen germination. (A) Immuno-fluorescence of pollen tube at different time points after germination using the ATG8 (red) antibody; bar: 5 μm. (B) Number of autophagosomes in the germination aperture labeled by ATG8 immuno-fluorescence. Data are from three independent experiments, with 40 pollen grains or pollen tubes analyzed in each experiment. Data at all later time points were significantly different from that at 0.25 h (t-test, p < 0.01)

Figure 2.

Deficiency of autophagy through silencing ATGs expression inhibits pollen germination. (A) Expression level of ATG2 or ATG5 was reduced in different RNAi transgenic lines compared with the WT. Number refer to relative mRNA levels, with WT set as 1. Each datum represents the mean of three biological replicates. (B, C) Pollen germination was reduced in ATG2 and ATG5 RNAi plants (bar: 200 μm). The frequencies of pollen germination of different ATG2 or ATG5 RNAi lines were reduced compared with the WT. The frequencies of pollen germination were determined by scoring 300 pollen grains with four independent replicates for every RNAi line at 2 HAG. (n = 1200). (t-test, ** indicates p < 0.01) (D) ATG protein levels in the WT, ATG2, ATG5 and ATG7 RNAi plants were evaluated through western blot using antibodies specific for individual ATGs. Three tobacco ATG13a isoforms are predicted to be about 67 kDa, thus appeared as a single band. β-actin was used to confirm comparable protein loading. (E) Relative protein levels of ATGs in WT, ATG2, ATG5 and ATG7 RNAi pollen. Each data bar represents the mean ± SD from three independent experiments. For quantification, the band intensity for each protein was first normalized to the band intensity of actin and then to the WT intensity (set as 1). ATG5 and ATG6 were used as positive and negative controls, respectively. ** indicates statistically significant difference compared to WT (t-test, p < 0.01)

Figure 3.

Downregulation of ATGs blocks autophagosome formation and leads to the persistence of cytoplasm in the inside aperture. (A–C) Immuno-fluorescence of germination aperture of pollen from WT and ATG-silencing lines using the ATG8 (red) antibody (bar: 5 μm). (D) Number of autophagosomes in the germination aperture labeled by immuno-fluorescence. Data are from three independent experiments, with 40 pollen grains analyzed in each experiment (n = 120). (E) Lipidation of ATG8 was confirmed by Phospholipase D treatment. Mem, membrane fraction collected after centrifugation and solubilized in 0.5% Triton X-100; PLD, phospholipase D. (F) ATG8–PE adduct was decreased in pollen from ATG-silencing plants. Membrane fractions from WT and ATG-silencing plants were used for immunoblot analysis. Ponceau staining was used to determine the loading control. (G) Relative ATG8–PE levels in WT, ATG2 and ATG5 RNAi pollen. Each data bar represents the mean ± SD from five independent experiments. (H) A layer of undegraded material (arrows) was observed in the aperture of germination-aborted pollen from ATG-silencing plants (bar: 2 μm, n = 16–19). Arrowheads, autophagosomes; in, intine; ex, exine. (I) Area of undegraded layer at the germination aperture in ATGs RNAi pollen (n = 16–19). ** indicate statistical difference compared to the WT (t-test, p < 0.01)

Figure 4.

Altered levels of autophagy-related lipids in ATG-silencing pollen. (A) OPLS-DA of lipid levels. The ellipse represents the Hotelling T2 with 95% confidence. X-axis indicates predictive principal components, whereas Y-axis indicates orthogonal principal components. (B) Levels of phosphatidylinositol (PI) decreased in ATGs-silencing pollen. (C) Heat map shows the relative PE and LysoPE (LPE) levels in ATGs-silencing pollen compared with the WT. (D–F) Inhibition of autophagy leads to the increase of Cer (D), CL (E), and TAG (F) in ATGs-silencing pollen, respectively. (t-test, * and ** indicate p < 0.05 or 0.01, respectively)

Figure 5.

Proteomic analysis of autophagy-deficient pollen. (A) Venn diagrams showing the numbers of significantly differentially expressed proteins in ATG2-silencing pollen or ATG5-silencing pollen compared with the WT and the overlaps of differentially expressed proteins between ATG2 and ATG5 group. Up, upregulated proteins; Down, downregulated proteins. (B, C) Volcano plot showing statistical significance of the differences in protein abundances due to autophagy deficiency. Yellow dots indicated upregulated proteins in autophagy-deficient pollen, while blue dots indicated downregulated proteins in autophagy-deficient pollen (t-test, p < 0.05). (D) GO analysis at cellular component level of common differentially expressed proteins in ATG2-silencing pollen and ATG5-silencing pollen. (E) GO analysis at biological process level of common differentially expressed proteins in ATG2-silencing pollen and ATG5-silencing pollen. (F) Enriched pathways in common differentially expressed proteins of ATG2-silencing pollen and ATG5-silencing pollen