ORIGINAL ARTICLES
Autophagy-like Processes are Induced during Tepal Senescence in Lily
2025 Volume 94 Issue 1 Pages 81-91
Details
2025 Volume 94 Issue 1 Pages 81-91
Autophagy is a conserved system responsible for the degradation of intracellular components and remobilization of nutrients in eukaryotes. Autophagy-like processes have been observed in senescing petals of several plant species. Further, significant expression levels of autophagy-related genes are induced along with an increase in ethylene production in flowers that show ethylene-dependent senescence. However, knowledge of autophagy-like processes in ethylene-independent flower senescence remains limited. In this study, we isolated autophagy-related genes (LhATG5, LhATG6, five LhATG8 homologues, and LhATG10), and analyzed their expression levels during tepal senescence in the Oriental hybrid lily, which shows ethylene-independent senescence. The transcript levels of LhATG5, LhATG6, and LhATG8s, but not LhATG8e, increased as tepals senesced. Furthermore, we observed cellular structures stained with monodansylcadaverine (MDC) dye, which stains acidified vesicular compartments, in lily tepals. MDC-stained structures were observed more frequently in the mesophyll cells of senescing tepals than in pre-senescent tepals. These results suggest that autophagy-like processes are induced as lily tepals senesce. Moreover, the amino acid content gradually increased in tepal mesophyll as the flowers senesced. Such increased amino-acid content was accompanied by a decrease in protein content and an increase in the transcript levels of autophagy-related genes, suggesting the involvement of autophagy-like processes in the production of nitrogenous components resulting from bulk protein degradation. Overall, our data indicated that autophagy-like processes were induced during lily tepal senescence. Further studies on the regulatory mechanisms of autophagy-like processes in ethylene-independent petal senescence may contribute to improvements in flower quality.
Flower longevity is one of the most important horticultural traits and it has a large economic impact on the flower industry. Petal senescence is defined as a type of programmed cell death (PCD; van Doorn and Woltering, 2008), a process during which, nucleic acids, proteins, and other macromolecules are degraded (Shibuya et al., 2016; van Doorn and Woltering, 2008), and presumably remobilized from senescing petals to other organs, such as developing ovaries (Bieleski, 1995; Jones, 2013; Shibuya et al., 2013).
Based on the involvement of ethylene in the process, flower senescence can be classified into two major types: ethylene-dependent and ethylene-independent (Shibuya, 2018; Woltering and van Doorn, 1988). Increased endogenous ethylene production triggers petal senescence in ethylene-dependent flowers and, in general, ethylene action inhibitors such as silver thiosulfate (STS) and 1-methylcyclopropene (1-MCP) prolong the vase life of such flowers, but have little effect on the vase life of flowers in which senescence is ethylene-independent (Shibuya, 2018; Woltering and van Doorn, 1988). Therefore, an understanding of the mechanisms of ethylene-independent flower senescence is required to develop effective techniques to improve flower longevity.
Lily (Lilium spp.) is one of the most important ornamental flowers worldwide and it is generally classified as an ethylene-independent flower with respect to senescence (Elgar et al., 1999; Han and Miller, 2003; Woltering and van Doorn, 1988). PCD-related genes, including the KDEL-tailed cysteine protease gene (CYP), the S1/P1 type nuclease gene (NUC), the vacuolar processing enzyme gene (VPE), and the senescence-associated gene (SAG), have been identified in Lilium spp. and L. longiflorum (Battelli, 2010; Battelli et al., 2014; Luo et al., 2021; Mochizuki-Kawai et al., 2015). These genes are highly expressed as tepals senesce, implying their involvement in PCD events such as protein and nucleotide degradation. For example, by means of transcriptome analysis, Luo et al. (2021) observed that the transcription factor gene LoNAC29 is highly upregulated during lily tepal senescence. LoNAC29, which is referred to as LbNAP, has been proposed as a key regulator of lily tepal senescence since the downregulation of LbNAP by virus-induced gene silencing delays tepal senescence (Liu et al., 2023).
Autophagy, so-called “self-eating”, is an intracellular system highly conserved in eukaryotes and it is responsible for degradation of cytoplasmic components such as proteins, their aggregates, organelles, and nutrient remobilization (Magen et al., 2022). In particular, remobilized nitrogenous components are amino acids and small peptides produced by protein degradation and transported to other organs (Havé et al., 2017). Plant autophagy is commonly divided into two types, macroautophagy and microautophagy; the former being usually referred to as autophagy (Liu and Bassham, 2012), in which case sequestered membranes (phagophores) arising in the cytoplasm elongate to form a characteristic lipid-bilayer organelle known as double-membrane autophagosome, which encapsulates the cellular components targeted for degradation. The outer membrane of autophagosomes fuses with the vacuolar membrane. Subsequently, a single-membrane structure, namely an autophagic body, is delivered into the vacuole, and both the inner membrane and encapsulated cargo are degraded by vacuolar lysis. Finally, the degraded components in the vacuole are returned to the cytoplasm for reuse.
The core of the autophagy machinery is regulated by several proteins encoding autophagy-related genes (ATGs) originally described in yeast (Saccharomyces cerevisiae). Subsequently, multiple AtATG genes involved in autophagy have been identified as homologues in Arabidopsis thaliana (Doelling et al., 2002; Hanaoka et al., 2002). In particular, ATG8 is the most well-studied molecular marker of autophagosome formation among ATG proteins. Firstly, ATG8 conjugates with the membrane lipid phosphatidylethanolamine (PE) via the conserved C-terminal glycine residue of ATG8 (Magen et al., 2022). Autophagosome formation requires two ubiquitination-like systems, ATG8-PE lipid adduct, and ATG12-ATG5 conjugation, which are important for the elongation and enclosure of the phagophore. Autophagosome formation is regulated by multiple ATG proteins, their complexes, and their interactions with other proteins. For instance, ATG6 is a core regulator that interacts with phosphoinositide 3-kinase (non-ATG protein, PI3K), and the PI3K complex, including ATG6, involves vesicle nucleation, which is required for autophagosome initiation (Zhuang et al., 2018). In turn, ATG10 is essential for the conjugation of ATG12 and ATG5 (Phillips et al., 2008). Further, the formation of autophagic bodies is induced concomitantly with the upregulation of ATG genes as leaves senesce (Sobieszczuk-Nowicka et al., 2018; Wojciechowska et al., 2018), suggesting that autophagy is involved in leaf senescence.
Autophagosome-like structures have been observed by electron microscopy in the senescing petals of several ornamental flowers (Kamdee et al., 2015; Matile and Winkenbach, 1971; Shibuya et al., 2013; Smith et al., 1992). The acidotropic dye monodansylcadaverine (MDC) labels autophagosome-like structures including late-stage autophagosomes and acidified vesicular compartments (Klionsky et al., 2021). Increased numbers of MDC-stained structures have been observed in the senescent petals of petunias (Petunia hybrida; Shibuya et al., 2013). Although the exact nature of the MDC-stained structures in senescing petals has not been identified, and interpretation requires caution, these microscopic analyses suggest the occurrence of an autophagy-like process during petal senescence. In carnation (Dianthus caryophyllus), Japanese morning glory (Ipomea nil), petunia, and rose (Rosa hybrida), ATG genes (ATG1, ATG4, ATG5, ATG6, ATG7, ATG8, and ATG13) are highly induced in senescing petals (Broderick et al., 2014; Kondo et al., 2020; Pillajo et al., 2018; Shibuya et al., 2011, 2013; Wang et al., 2023; Yamada et al., 2009), suggesting that autophagic processes are induced during petal senescence. Moreover, exogenous ethylene treatment or pollination-induced endogenous ethylene production accelerates petunia petal senescence, accompanied by increased expression of ATG8s and MDC-stained structures (Shibuya et al., 2013). In carnations exhibiting ethylene-dependent senescence, the ATG8 expression pattern corresponds to the timing of endogenous ethylene production during petal senescence (Kondo et al., 2020), thereby supporting the presumed involvement of ethylene in autophagy induction. Therefore, ethylene is considered to be a key regulator of autophagy in ethylene-induced petal senescence. In contrast, in Japanese morning glory that exhibits ethylene-independent senescence, a significant increase in the transcript level of InATG4 following ethylene treatment has been observed, while the expression of InATG8a is not induced by exogenous ethylene (Yamada et al., 2009). In ethylene-independent flowers, studies on autophagy during petal senescence are limited, and the link between petal senescence and autophagy is not well understood.
In this study, we investigated the expression of autophagy-related genes and performed a microscopic analysis of lily tepals. Additionally, changes in protein and amino acid contents during tepal senescence were analyzed. Flowers that exhibit ethylene-independent senescence include commercially important plants such as lilies, tulips, and irises, but no effective vase life extension techniques have been developed for these plants. Understanding the autophagic processes involved in lily tepal senescence would be helpful for developing strategies to improve the postharvest quality of flowers displaying ethylene-independent senescence.
Oriental hybrid lily ‘Siberia’ (Lilium spp.) bulbs were cultivated in a greenhouse at the Institute of Vegetable and Floriculture Science (NARO), in Tsukuba, Japan, and harvested between May 31, 2023, and June 6, 2023. The Oriental hybrid lily is categorized as a flower that displays ethylene-independent senescence (Elgar et al., 1999). ‘Siberia’ flowers used in this study also displayed ethylene-independent senescence as described in the Results section (Fig. S1; Table S1). When the flower buds at the lowest stem position of the inflorescence turned white, the stems were harvested and trimmed to a length of 80 cm. Then, the cut flower stems were placed in distilled water in an evaluation room at 23°C, 70% relative humidity, 10–15 μmol·m−2·s−1, and under a photoperiod regime of 12 h/12 h (light/dark) for the duration of the experiment. Individual flowers that had tepal tips partially or horizontally opened were sequentially sampled from the bottom part of the inflorescence and placed in distilled water. Individual flowers with partially opened tepal tips developed horizontally the following day. The horizontally opened flowers were designated as day 0 flowers. The central parts of the inner tepals were used in subsequent analyses as described in Figure S2A. A foliar section (3 mm × 4 cm) was sampled from the middle of the leaf at three different senescence stages; non-senescent maturity (MT, entire leaf is deep green), early senescence (ES, base of the leaf is yellow), and late senescence (LS, entire leaf is yellow), for gene expression analysis (Fig. S2B). The inner tepals and leaves sampled were immediately frozen in liquid nitrogen and stored at −80°C until use.
Ethylene, 1-MCP and STS treatmentFor ethylene treatment, individual lily flowers on day 0 were sealed in a plastic chamber with 10 μL·L−1 of ethylene for 24 h. For 1-MCP (an ethylene action inhibitor) treatment, individual flowers were sealed in a plastic chamber with 2 μL·L−1 of 1-MCP (EthylBlocTM Sachat; Floralife Walterboro, SC, USA) for 24 h. For STS (an ethylene action inhibitor) treatment, individual flowers on day 0 were placed in plastic tubes containing distilled water or 0.2 mM STS solution for 24 h. After STS treatment, the individual flowers were placed in distilled water. The treatments with ethylene, 1-MCP, and STS were performed in the evaluation room under the same conditions as described above. After the chemical treatments, flowers were placed in the evaluation room and the visible senescence symptoms of individual flowers were evaluated daily based on the degree of both browning and wilting. The symptoms of the individual flowers were defined by the following eight scores: 0 = no browning and wilting; 1 = cream-colored whole flower and/or partial water-sink; 2 = less than 5% browning and/or wilting; 3 = 5–20%; 4 = 20–40%; 5 = 40–60%; 6 = 60–80%; 7 = more than 80%. The vase life of individual flowers ends when the score reaches 3.
Ethylene measurementIndividual flowers were placed in 2.95 L glass jars (DURAN DN150; SCHOTT, Mainz, Germany), sealed and incubated at 23°C for 24 h. Headspace gas samples (1 mL) were taken at 7 h and 24 h after incubation, and injected into a gas chromatograph (GC-14B; Shimadzu, Kyoto, Japan) equipped with an alumina column and a flame ionization detector. Ethylene measurements were performed on three different individual flowers per senescence stage.
Assessment of visual appearance during tepal senescenceVisual senescence of the flowers was investigated between days 0 (horizontally opened flowers) and 8 (completely wilted flowers). To investigate changes in the visual appearance of lily flowers during senescence, the colorimetric components of the Commission Internationale de l’Eclairage (CIE) L*a*b* of the left and right center parts across the midrib on the adaxial surface of the inner tepal were measured using a colorimeter (NF 555; Nippon Denshoku Industries Co., Ltd., Tokyo, Japan; Fig. S2). The colorimetric values of six parts of the three inner tepals (two parts per inner tepal) collected from each individual flower were measured and averaged. Measurements were performed on five individual flowers.
Determination of protein contentsSoluble protein was extracted from the inner tepal segment (1 cm × 2 cm) with 200 μL of extraction buffer (50 mM Tris-HCl pH 7.5, 1 mM EDTA pH 8.0, 0.1% TritonX-100, 1 mM DTT in dH2O) according to Battelli et al. (2011). The extracted solution was vortexed and centrifuged at 12,000 g, for 5 min, at 4°C. The soluble protein content of the supernatant was quantified using Bradford dye reagent (Takara Bio Inc., Shiga, Japan), and the BSA standard curve was diluted with 0.9% NaCl according to the manufacturer’s instructions. Tepal soluble-protein contents per tepal area were calculated in μg·cm−2.
Gene isolationTotal RNA was extracted from frozen samples using an RNeasy Plant Mini Kit (Qiagen, Hilden, Germany) according to manufacturer’s instructions with an RNase-Free DNase Set (Qiagen) to avoid genomic DNA contamination. The RNA concentration was estimated using a NanoDrop spectrophotometer (Thermo Fisher Scientific Inc., MA, USA). cDNA was synthesized from 100 ng of the DNase-treated total RNA with a PrimeScriptTM RT Master Mix (Perfect Real Time, Takara Bio Inc.) and diluted with 15 μL of 1 × TE. The cDNA was used for gene isolation and qRT-PCR analysis. The primers were designed based on RNA-seq data of the lily flower ‘Tiara’ (Shibuya et al., unpublished). Gene-specific primers used are listed in Table S2. All genes, including the open reading frame, were amplified with PrimeSTAR GXL DNA polymerase (Takara Bio Inc.) and gene-specific primers, and cloned into pGEM-T easy vectors (Promega, Madison, WI, USA). The cloned gene sequences were read using an ABI 3730XL Genetic Analyzer (Thermo Fisher Scientific Inc., Tokyo, Japan). Alignment and phylogenetic analyses of the isolated genes were performed using MEGA11 software (ver. 11.0.13) and ClustalW (https://www.genome.jp/tools-bin/clustalw). The amino acid sequences of the isolated ATG genes from lily tepals were compared with those of A. thaliana ATG genes (AtATG5, At5g17290; AtATG6, At3g61710; AtATG8a, At4g21980; AtATG8b, At4g04620; AtATG8c, At1g62040; AtATG8d, At2g05630; AtATG8e, At2g45170; AtATG8f, At4g16520; AtATG8g, At3g60640; AtATG8h, At3g06420; AtATG8i, At3g15580; AtATG10, At3g07525); D. caryophyllus (ATG8a, LC505515); Hordeum vulgare (HvATG6, AK362923); I. nil (InATG8a, AB544067; InATG8b, AB544068; InATG8c, AB544069; InATG8d, AB544070; InATG8e, AB544071; InATG8f, AB544072); Oryza sativa (OsATG5, NP_001388682; OsATG6a, BAS73712; OsATG8a, AK059939; OsATG8b, AK121268; OsATG8c, AK062573; OsATG8d, LOC4349525; OsATG8e, LOC107276566; OsATG10a, AK241831); P. hybrida (PhATG8a, AB721297; PhATG8b, AB721298; PhATG8c, AB721299; PhATG8d, AB721300); Zea mays (ZmATG8a, EU958456; ZmATG8b, EU956425; ZmATG8d, LOC100240697; ZmATG8e, EU970911) and S. cerevisiae (ScATG5, NP_015176; ScATG6, NP_015205; ScATG8, AY692870; ScATG10, AAS56516). The amino acid sequences of LhNAP, LhSAG39, and LhCYP obtained in this study were compared with those of known genes isolated from Lilium spp. and L. longiflorum (LoNAC29, MW548579; LoSAG39, MW520943; LlCYP, HF968474).
qRT-PCR analysisAnalysis by quantitative real-time RT-PCR (qRT-PCR) was performed using 1 μL of the diluted cDNA, a TB Green Premix Ex Taq II (Takara Bio Inc.), gene specific primers, and a TaKaRa Thermal Cycler Dice Real-Time System TP-800 (Takara Bio Inc.), after checking primer specificity. The primer sequences used for qRT-PCR are listed in Table S3. The qRT-PCR reaction was performed under the following conditions: 95°C for 30 s; 40 cycles of 95°C for 5 s and 60°C for 30 s; and a cycle of 95°C for 15 s, 60°C for 30 s, and 95°C for 15 s. The qRT-PCR for LhCYP was performed under the following conditions: 95°C for 30 s; 40 cycles of 95°C for 5 s and 60°C for 40 s; and a cycle of 95°C for 15 s, 60°C for 40 s, and 95°C for 15 s. The expression levels of PCD- and autophagy-related genes were normalized with those of the elongation factor gene (LhEF1α, accession no. LC806228). The qRT-PCR analysis was performed using four biological replicates.
Microscopic observation of MDC-stained structuresStaining with MDC was performed as previously described by Shibuya et al. (2013), after modifications to the concentration of the dye reagent. Briefly, the epidermis was removed from the inner tepals with forceps and the remaining mesophyll was rinsed in 500 μM MDC (Sigma-Aldrich, St. Louis, USA) and 0.4 M mannitol solution in PBS buffer (pH 7.4) for 30 min at room temperature, then washed with PBS containing 0.4 M mannitol. MDC-stained structures were observed with a fluorescence microscope (AX70; Olympus, Tokyo, Japan) with a narrow band pass filter for 4',6-diamino-2-phenylindole (DAPI; excited by a wavelength of 360–370 nm and detected at 430–460 nm through U-MNUA2; Olympus). To visualize the nucleus, isolated mesophyll of lily inner tepals on day 3 was stained with DAPI solution (CyStain UV Precise P; Sysmex Corporation, Kobe, Japan), and then observed using a fluorescence microscope (AX70; Olympus) with a U-MNUA2 filter (Olympus).
Protoplasts were prepared by digesting inner tepals on days 3 and 5 with 0.5 M MES buffer (20 mM MES, 0.5 M sorbitol, 2 mM CaCl2, 2 mM MgCl2; pH 5.8) containing 1% (w/v) cellulase (Yakult Pharmaceutical Industry Co., Ltd, Tokyo, Japan) and 0.2% (w/v) macerozyme (Yakult Pharmaceutical Industry) at 30°C for 3 h. Protoplasts were stained with 50 μM MDC (Sigma-Aldrich) in the MES buffer for protoplast preparation for 20 min in the dark. After centrifuging at 100 g for 2 min at 15°C, MDC-stained structures in the precipitated protoplasts were observed using a fluorescence microscope (AX70; Olympus) with a U-MNUA2 filter (Olympus).
Free amino-acid contentFree amino acids were determined according to the method described by Magné and Larher (1992), after minor modifications. The epidermis of the inner tepals was removed using forceps, and the mesophyll was isolated, homogenized in liquid nitrogen, and deproteinized with 500 μL of 3% sulfosalicylic acid. After centrifugation, 40 μL of the supernatant of the sample was mixed with 20 μL of citrate buffer (pH 4.6) and 40 μL of 1% ninhydrin solution (ninhydrin and 0.03% ascorbic acid in 2-methoxyethanol). The mixture was boiled at 95°C for 3 min and immediately cooled on ice. Then, 100 μL of chilled 60% ethanol was added to the mixture and it was read at 570 nm on a microplate reader, Synergy HT (Biotek Instruments, Inc., Vermont, USA). Free amino-acid analysis was performed using four biological replicates, and the content was determined from a standard curve of glycine.
First, we analyzed the involvement of ethylene in tepal senescence of the lily ‘Siberia’ used in this study. Exogenous ethylene (10 μL·L−1) treatments did not accelerate tepal senescence, and 1-MCP (2 μL·L−1) and STS (0.2 mM) treatments, which are ethylene action inhibitors, did not significantly delay tepal senescence (Fig. S1A–C). Ethylene production rates in ‘Siberia’ flowers were very low during flower senescence (Table S1). These results indicated that ‘Siberia’ flowers exhibited ethylene-independent senescence similar to other Oriental hybrid lily cultivars.
The flower buds of lily ‘Siberia’ underwent a clear senescence process (Fig. 1A). The visible signs of senescence were not observed until day 4 (Fig. 1A), whereas the marginal areas of the inner tepals were partially softened and water-soaked on day 5. Brown-colored areas of the inner and outer tepals partially appeared on day 6 and expanded as senescence progressed. The b* values did not change until day 4, but gradually increased from days 5 to 8 (Fig. 1B).
Visual changes and protein degradation during tepal senescence of lily flower ‘Siberia’. (A) Photographs of flowers were taken daily until complete wilting (day 8) starting with horizontally opened flowers (day 0). (B) Changes in flower color parameter, b* value, during tepal senescence. Positive b* value indicate yellowish coloration. Asterisks indicate significant differences compared with day 0 by Dunnett’s multiple comparison test (*** P < 0.001, n = 5). (C) Changes in protein content of the inner tepals during senescence. Asterisks indicate significant differences compared with day 0 by Dunnett’s multiple comparison test (* P < 0.05, *** P < 0.001, n = 4).
The protein content of the central parts of the inner tepals tended to gradually decrease until day 4 and then increased transiently on day 5 (Fig. 1C). However, protein content was significantly lower on days 6 and 7 than that on day 0.
Isolation of autophagy-related genes from lily tepalsFive LhATG8 homologues were isolated from the inner tepals of the Oriental hybrid lily ‘Siberia’. The amino acid sequences of LhATG8s contained the C-terminal glycine-binding PE and the residues forming the W- and L-pockets that interact with other proteins possessed an ATG8 interaction motif (Kellner et al., 2017; Noda et al., 2010; Ohsumi, 2001; Fig. S3). Furthermore, LhATG8a-e shared 76–89% identity with AtATG8a-i, and LhATG5 (accession no. LC806225), LhATG6 (LC806226), and LhATG10 (LC806227) also contained conserved sites and residues among their orthologs (Noda et al., 2012; Phillips et al., 2008; Shemi et al., 2015; Figs. S4–6) and showed identity homology levels of 56.6%, 64.9%, and 40.0% with those of A. thaliana, respectively.
The AtATG8 family in A. thaliana was divided into subgroups a-d, e-g, and h-i (Seo et al., 2016). Our phylogenetic analysis indicated that the five LhATG8s were clustered into the three subgroups, with one exception (Fig. 2). LhATG8a (LC806220) was in subgroup a–d, including AtATG8a and AtATG8d. Meanwhile, LhATG8b (LC806221) was similar to InATG8e in the e-g subgroup. LhATG8c (LC806222) was present in the h-i subgroup, including AtATG8h and AtATG8i. However, the proteins encoding LhATG8d (LC806223) and LhATG8e (LC806224) differed from those in the other subgroups.
Phylogenic trees of ATG8 homologues in the ‘Siberia’ lily and other species. MEGA software and ClustalW were used to construct the phylogenic trees of ATG8 proteins in Lilium spp. Oriental hybrid lily ‘Siberia’ (Lh), A. thaliana (At), D. caryophyllus (Dc), I. nil (In), O. sativa (Os), P. hybrida (Ph), Z. mays (Zm), and S. cerevisiae (Sc). The length of each branch is proportional to the genetic distance calculated using the neighbor-joining method. The three groups categorized by Seo et al. (2016) are shown beside the branches.
The transcript levels of LhNAP (LC806229) were low from days 0 to 2 (Fig. 3A), gradually increased from days 3 to 5, and then remained high until day 7. The transcript levels of LhSAG39 (LC806230) were low from days 0 to 3, gradually increased from day 4, peaked on day 5, and then decreased. Meanwhile, the transcript levels of LhCYP (LC806231) were low from days 0 to 2, increased from day 3, peaked on day 4, and subsequently decreased with senescence.
Time-course of the expression levels of PCD-related and LhATG genes during senescence of tepals and leaves. Gene expression levels of PCD-related genes (A) and LhATG genes (B) were normalized with LhEF1α as an internal control. Days 0–7 represent flower senescence stages, corresponding to the photographs in Figure 1A. MT, ES, and LS represent non-senescent maturity, early senescence, and late senescence, respectively (Fig. S2). Mean ± SE values are shown from four biological replicates.
LhATG8s except for LhATG8e were expressed at low levels between days 0 and 3, similar to LhNAP, LhSAG39, and LhCYP (Fig. 3B). Subsequently, the expression levels of LhATG8c and d gradually increased as tepal senescence continued to completion, while the transcript levels of LhATG8a and b showed no difference or a decrease between days 6 and 7, respectively. The expression pattern of LhATG8e differed from those of other LhATG8 genes. Particularly, LhATG8e was highly expressed in pre-senescent tepals in which visual senescence signs were not yet observed. The transcript level tended to decrease with senescence progression. The incremental rates of LhATG5 and LhATG6 transcripts during tepal senescence were lower than those of LhATG8a-d from days 3 to 5, and then there was no clear difference in transcript levels until day 7. The transcript level of LhATG10 increased from days 0 to 2, then decreased until day 5, rose again on day 6 and decreased thereafter on day 7.
Except for LhATG10, all other PCD- and autophagy-related genes were highly expressed in senescent leaves, whereas their transcript levels were low in non-senescent mature, and/or early senescent leaves (Fig. 3A, B). LhATG10 was expressed in non-senescent mature and early senescent leaves, and its expression level decreased in senescent leaves.
Microscopic observation of MDC-stained structures in tepals throughout the senescence processThe mesophyll of the lily tepals was composed of tetrapod-like cells connected in a mesh-like pattern, and the individual cells possessed DAPI-stained nuclei (Fig. 4A). In mesophyll cells, MDC-stained structures of various sizes (approximately 0.5–10 μm) were observed (Figs. 4B, C and 5A–C; Fig. S7). MDC-stained structures, regardless of their size, were frequently observed from days 5 to 7, especially on day 6, but were rare from days 0 to 4 (Fig. 4C; Fig. S7). Larger MDC-stained structures were more frequently observed from days 5 to 7 (Fig. 4B, C; Fig. S7).
Microscopic observation of MDC-stained structures during lily tepal senescence. (A) Brightfield image (left) and DAPI staining (right) of the mesophyll of the lily inner tepals on day 4. Arrowheads indicate cell-to-cell boundaries. Asterisks indicate nuclei. Scale bar represents 90 μm. (B) MDC staining in the mesophyll of the inner tepals on day 7. Arrows indicate representative examples of MDC-stained structures of various sizes in the mesophyll cells. Scale bar represents 40 μm. (C) Micrographs of MDC staining in the mesophyll of the inner tepals on days 0, 2, 4, and 6. Arrows in the micrograph on day 6 indicate presentative examples of MDC-stained structures. Scale bars represent 90 μm.
Micrographs of MDC-stained structures in protoplasts. Protoplasts prepared from the mesophyll of the inner tepals on day 3 (A) and day 5 (B and C). Protoplasts observed in a bright field (left) and stained with MDC (right) are presented. The protoplast shown in (B) is different from the protoplast shown in (C). Scale bars represent 20 μm.
In protoplasts prepared from tepal mesophyll on day 3, a few small MDC-stained structures were observed (Fig. 5A). In protoplasts prepared from mesophyll on day 5, larger MDC-stained structures were observed in addition to smaller ones (Fig. 5B, C)
Changes in amino acid contents during tepal senescenceThe amino acid contents of the mesophyll of the inner tepals tended to increase from days 1 to 2 (Fig. 6). The amino acid contents increased further from days 4 to 6 and tended to decrease on day 7, when most parts of the tepals appeared wilted.
Changes in free amino acid contents during tepal senescence. Different letters indicate significant differences in Tukey’s multiple comparison test (P < 0.05, n = 4).
Autophagy is a conserved degradation system of cellular components involved in nutrient recycling (Magen et al., 2022). Autophagy-like structures have been observed in the petals of several flowers, including carnation (Smith et al., 1992), petunia (Shibuya et al., 2013), Dendrobium (Kamdee et al., 2015), rose (Wang et al., 2023), morning glory (Matile and Winkenbach, 1971; Shibuya et al., 2009), and Iris (Kamdee et al., 2015). An increase in transcript levels of ATG genes in senescing petals was reported in petunia (Broderick et al., 2014; Pillajo et al., 2018; Shibuya et al., 2013), carnation (Kondo et al., 2020), and rose (Wang et al., 2023). Consistently, transcript levels of ATG8 increased concomitantly with ethylene production in petunia and carnation, which show ethylene-dependent petal senescence. In contrast, to date, information on autophagy at the molecular level in flowers that show ethylene-independent senescence has been limited to a model plant, Japanese morning glory (Shibuya et al., 2009, 2011; Yamada et al., 2009). In this study, we investigated autophagy-like processes during tepal senescence in the Oriental hybrid lily to better understand the mechanisms underlying ethylene-independent flower senescence. In flowers that show ethylene-dependent senescence such as carnation, ethylene action inhibitors markedly extend vase life (Woltering and van Doorn, 1988). In contrast, effective vase life extension techniques have not been developed for ethylene-independent flowers, even though these flowers include many commercially important plants such as lilies, tulips, and irises. Understanding the autophagic processes involved in lily tepal senescence could give us clues for developing strategies to improve postharvest quality of flowers displaying ethylene-independent senescence.
The number of ATG8 homologues varies among plant species, and most plants have redundant genes (Seo et al., 2016). Herein, we isolated five LhATG8 homologues from lily. The phylogenetic analysis showed that, except for LhATG8d-e, LhATG8 genes could be classified into three subgroups, as determined by Seo et al. (2016). This finding implies different evolutionary processes for LhATG8d-e. LhATG8d shared 92.4% sequence identity with LhATG8e. ATG8s belonging to the h-i subgroup have an exposed catalytic glycine that does not require cleavage by the proteolysis of ATG4 (Boycheva Woltering and Isono, 2020; Seo et al., 2016), whereas some have additional residues in addition to glycine (Seo et al., 2016). We also found that LhATG8c possessed no additional residues after the catalytic glycine at the C-terminus (Fig. S3). These accurate reflections of ATG features suggested that LhATGs are as functional as their counterparts in other plants. Monocots such as O. sativa and Z. mays have lost the e-g subgroup owing to extinction and expansion in specific lineages (Seo et al., 2016). However, LhATG8b clustered with the e-g subgroup even though lily belongs to the monocot group, suggesting that lily maintains LhATG8b from the e-g group that has avoided extinction.
Several ATG genes are highly expressed in the senescent petals of several species (Broderick et al., 2014; Kondo et al., 2020; Pillajo et al., 2018; Shibuya et al., 2011, 2013; Wang et al., 2023; Yamada et al., 2009). The transcript levels of LhATG5, LhATG6, and some LhATG8a-d except for LhATG8e also increased as the lily tepals senesced. Thus, the upregulation of LhATG suggests the induction of an autophagy-like process during ethylene-independent tepal senescence. Additionally, except for LhATG10, all other LhATG genes were also highly expressed in senescent leaves, whereas LhATG10 expression levels decreased in senescent leaves, as they did in senescent tepals, suggesting that the induction of autophagy may be at least partially the same in tepal and leaf senescence.
In the expression analysis, LhATG8e and LhATG10 displayed different expression patterns from those of other ATG genes during tepal senescence. We considered the following hypotheses to explain these results: the phylogenetic analysis indicated that LhATG8e is very similar to LhATG8d (Fig. 2). Additionally, the expression of LhATG8e increased in senescing leaves, but not in tepals, to the same extent as that of LhATG8d (Fig. 3B). Therefore, the different expression patterns of LhATG8d and LhATG8e during tepal senescence may be attributed to neofunctionalization or pseudogenization following the duplication event described by Lynch and Conery (2000). The transcript level of LhATG10 increased before the appearance of any visual signs of senescence (before day 5) and was low in senescent tepals (days 5–7) and leaves (Fig. 3B). ATG10 is essential for the formation of ATG12-ATG5 conjugates in leaves, but its role in petal senescence in ornamental flowers has not yet been reported. Consistently, the expression levels of the two OsATG10 homologues in older rice leaves were lower than those in younger leaves (Shin et al., 2009). Based on these results, the authors argued that higher OsATG10 expression in young foliar tissues seemingly allows autophagy to be readily induced. Other sets of LhATG genes such as LhATG10 may be highly expressed before tepal senescence to readily induce autophagy during petal senescence. Further investigation of the expression patterns of other ATG genes will provide deeper insights into the autophagic processes during flower senescence.
LhNAP is a transcription factor that regulates the transcription of LhSAG39 in lily flowers (Luo et al., 2021). In our experiment, the expression pattern of LhNAP was similar to those of LhATG8a, c, and d. Thus, LhNAP may also be involved in the regulation of these LhATG8 genes during tepal senescence, as well as other PCD-related genes. The transcript levels of LhCYP and LhSAG39 encoding proteinase were expressed maximally before and at the beginning of senescence signs (on days 4 and 5, respectively) and then decreased as the tepals senesced (Fig. 3A). However, the transcript levels of LhATG8a-d increased from before the onset of visual senescence signs (day 4), and then their transcript levels remained high until the tepals almost wilted (day 7). The protein content of the lily tepals decreased slightly before any signs of senescence began to appear (from days 0 to 4), and the subsequent protein content largely decreased in the more senescent tepals (on days 6 and 7), accompanied by increased amino acid contents. However, the amino acid contents tended to increase prior to visible senescence signs (from days 1 to 2; Fig. 6), even though no significant decrease in protein content occurred. This may be due to de novo synthesis of amino acids, including asparagine and glutamine, in pre-senescent tepals as reported in Sandersonia (Eason et al., 2000). On day 7, when most parts of the tepals had wilted, and free amino acid contents tended to decrease from day 6. In petunias and lilies, nitrogen has been reported to translocate from senescing petals (Mochizuki-Kawai et al., 2015; Shibuya et al., 2013). This decreased tendency of amino acids on day 7 may reflect nutrient translocation. Considering these patterns of change in LhATG gene expression, protein, and amino acid contents throughout tepal senescence, autophagy-like processes are likely involved in the bulk protein degradation associated with lily tepal senescence.
In petunia, the number of MDC-stained structures, referred to as autophagosome-like structures, increased in senescing petals, concomitantly with increased ethylene production after pollination and increased ATG8 gene expression (Shibuya et al., 2013). Our microscopic observations also indicated a large number of MDC-stained structures in the mesophyll of senescing lily tepals (days 5–7; Figs. 4B, C and 5B, C; Fig. S7). Furthermore, the increase in the abundance of MDC-stained structures appeared to be accompanied by an increase in the transcript levels of some LhATG genes. Thus, the microscopic analysis supported the idea that autophagy-like processes are induced during tepal senescence in lily. However, cautious interpretation is required, because the MDC dye labels late-stage autophagosomes and other acidic components (Klionsky et al., 2021). Further studies using electron microscopy, GFP-ATG8 fusion proteins and immunostaining with antibodies specific to ATG proteins will reveal the detailed mechanisms underlying autophagy in lily tepal senescence.
In this study, we identified multiple ATG genes that are induced in the senescing tepals of ethylene-independent lily flowers, suggesting that autophagy-like processes are induced in lily tepal senescence. This was supported by the results of MDC-staining, which visualizes autophagosome-like structures. Our data also suggest that autophagy-like processes involve the production of nitrogenous components during tepal senescence in lilies. Further understanding of the autophagic machinery in ethylene-independent senescence may lead to important clues for prolonging flower longevity.
The ‘Siberia’ flowers were cultivated in a greenhouse by Dr. Katsuhiko Inamoto (Institute of Vegetable and Floriculture Science, NARO), and gifted from him.
