doi:10.1093/plphys/kiaa015 PLANT PHYSIOLOGY 2020: 0: 1–10 Downloaded from https://academic.oup.com/plphys/advance-article/doi/10.1093/plphys/kiaa015/5991408 by Universitat Autònoma de Barcelona user on 27 January 2021 GENOMES UNCOUPLED1-independent retrograde signaling targets the ethylene pathway to repress photomorphogenesis Researh Report Charlotte M. M. Gommers ,1,2 Marı́a Águila Ruiz-Sola,1 Alba Ayats ,1 Lara Pereira ,3,† Marta Pujol 3,4 and Elena Monte 1,5,*,‡ 1 Plant Development and Signal Transduction Program, Center for Research in Agricultural Genomics (CSIC- IRTA-UAB-UB), Barcelona, Spain 2 Laboratory of Plant Physiology, Wageningen University & Research, Wageningen, The Netherlands 3 Plant and Animal Genomics Program, Center for Research in Agricultural Genomics (CSIC-IRTA-UAB-UB), Barcelona, Spain 4 Institut de Recerca i Tecnologia Agroalimentàries (IRTA), Barcelona, Spain 5 Consejo Superior de Investigaciones Cientı́ficas (CSIC), Barcelona, Spain *Author for communication: elena.monte@cragenomica.es (E.M.). † Present address: Center for Applied Genetic Technologies, University of Georgia, Athens, USA. ‡ Senior author. C.M.M.G. and E.M. conceived the project. C.M.M.G., M.A.R-S., A.A., L.P., and M.P. planned and performed the experiments. C.M.M.G., M.A.R-S., and A.A. analyzed the data. C.M.M.G. and E.M. wrote the manuscript. The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors is (https://academic.oup.com/plphys): E.M. (elena.monte@cragenomica.es). Abstract When germinating in the light, Arabidopsis (Arabidopsis thaliana) seedlings undergo photomorphogenic development, characterized by short hypocotyls, greening, and expanded cotyledons. Stressed chloroplasts emit retrograde signals to the nucleus that induce developmental responses and repress photomorphogenesis. The nuclear targets of these retrograde sig- nals are not yet fully known. Here, we show that lincomycin-treated seedlings (which lack developed chloroplasts) show strong phenotypic similarities to seedlings treated with ethylene (ET) precursor 1-aminocyclopropane-1-carboxylic acid, as both signals inhibit cotyledon separation in the light. We show that the lincomycin-induced phenotype partly requires a functioning ET signaling pathway, but could not detect increased ET emissions in response to the lincomycin treatment. The two treatments show overlap in upregulated gene transcripts, downstream of transcription factors ETHYLENE INSENSITIVE3 and EIN3-LIKE1. The induction of the ET signaling pathway is triggered by an unknown retrograde signal act- ing independently of GENOMES UNCOUPLED1. Our data show how two apparently different stress responses converge to optimize photomorphogenesis. Introduction nucleus is essential to build and maintain the photosynthe- As photoautotrophs, plants depend on light for growth and sis apparatus and to accurately adjust to environmental survival. Chloroplasts absorb photons to fuel the photosyn- stresses with potential damage to the photosystems. thesis reaction, which is the basis of all carbon sources on Retrograde signals from the plastid inform the nucleus earth. Most chloroplast proteins are encoded in the nuclear about disruptions in its function or development (reviewed genome, so tight communication between chloroplasts and in Chan et al., 2016; Crawford et al., 2018). Received September 2, 2020. Accepted September 22, 2020. Advance access publication 19 November 2020 C American Society of Plant Biologists 2020. All rights reserved. For permissions, please email: journals.permissions@oup.com V 2 | PLANT PHYSIOLOGY 2020: Page 2 of 10 Gommers et al. Quickly after germination, light triggers a developmental the kinase CONSTITUTIVE TRIPLE RESPONSE1 (CTR1), an program, which includes the opening and expansion of coty- inhibitor of the ER-localized, NRAMP-like protein ETHYLENE ledons, chloroplast maturation, reduced hypocotyl elonga- INSENSITIVE2 (EIN2). In the nucleus, EIN3-BINDING F-BOX1 Downloaded from https://academic.oup.com/plphys/advance-article/doi/10.1093/plphys/kiaa015/5991408 by Universitat Autònoma de Barcelona user on 27 January 2021 tion, and promoted root growth called photomorphogenesis (EBF1) and EBF2 bind and degrade the transcription factors (Gommers and Monte, 2018). The initiation of photomor- EIN3 and EIN3-LIKE1 (EIL1). In the presence of ET, the phogenesis depends on the activation of a set of photorecep- receptors and therefore CTR1 are inactive. This renders ac- tors, sensitive to red and far-red (phytochromes, phys), or tive EIN2 that inhibits the translation of EBF1 and EBF2 blue light (cryptochromes, crys). These repress a set of tran- mRNA. As a consequence, EIN3 and EIL1 accumulate and scription factors, including PHYTOCHROME INTERACTING target ET-regulated genes to optimize development to the FACTORS (PIFs) and ETHYLENE INSENSITIVE3 (EIN3), which stress situation that caused ET to accumulate (Binder, 2020). are repressors of photomorphogenesis in darkness (Leivar Even though lincomycin-induced RS and ET cause similar et al., 2008; Shi et al., 2018). phenotypes in light-grown seedlings, with MEcPP affecting When chloroplast development is interrupted, by chemi- ET-mediated hypocotyl elongation (Vogel et al., 2014; Jiang cals such as lincomycin (inhibitor of chloroplast translation) et al., 2020), and high light has been shown to induce a or norflurazon (inhibitor of carotenoid synthesis), or by high stress-pathway via ETHYLENE RESPONSE FACTORS (ERFs; light, this induces a retrograde signaling (RS) cascade and Vogel et al., 2014), very little is known about a possible over- inhibits cotyledon separation, possibly to minimize the lap between these stress signaling pathways. Here, we show stress-exposed area and protect the apical meristem (Ruckle that chloroplast RS promotes the expression of ET- and Larkin, 2009; Martı́n et al., 2016). This RS pathway acts responsive genes, which causes the delayed photomorpho- via the chloroplast localized, nuclear encoded, protein genesis in light-grown plants. We present evidence that this GENOMES UNCOUPLED1 (GUN1), a master integrator of RS-pathway acts independent of the well-studied GUN1-me- RS. GUN1 is a plastid-localized member of the pentatrico- diated signaling cascade. Our findings additionally suggest peptide repeat family involved in chloroplast synthesis as that various environmental stresses converge at a similar well as stress signaling, especially early during the leaf devel- core-set of nuclear genes to change plant development and opment (Koussevitzky et al., 2007; Jia et al., 2019; Pesaresi prevent further damage. and Kim, 2019). The lincomycin-induced, GUN1-mediated, RS pathway antagonizes phytochrome signaling and targets genes, which encode proteins that promote photomorpho- Results genesis and are repressed by PIFs in darkness. One of such ET affects photomorphogenesis in continuous light genes encodes the transcription factor GOLDEN2 LIKE1 Previous work has shown that ET can affect seedling photo- (GLK1). GLK1 transcriptional repression by GUN1-mediated morphogenesis. Dark-grown EIN3/pifqein3eil1 seedlings RS contributes to the closed cotyledon-phenotype in treated with ET precursor 1-aminocyclopropane-1-carboxylic lincomycin-treated seedlings (Martı́n et al., 2016). An addi- acid (ACC) had small, partly unseparated cotyledons but tional RS pathway with direct effect on light signaling is me- short hypocotyls compared to seedlings grown on regular diated by the plastid stress molecule methylerythritol MS medium (Shi et al., 2018). Dark-grown ein3eil1 seedlings cyclodiphosphate (MEcPP), which promotes phyB accumula- transferred to light had larger and more separated cotyle- tion and suppresses auxin and ethylene (ET)-mediated hypo- dons, while EIN3-ox displayed unopened and smaller cotyle- cotyl elongation in red light (Jiang et al., 2020). dons. In continuous red light, ACC triggered hypocotyl Another stress signal that affects photomorphogenesis is elongation, whilst inhibiting cotyledon expansion in wild the accumulation of the gaseous phytohormone ET. ET bio- type (WT) and EIN3-ox, but not in ein3eil1 (Shi et al., 2016). synthesis is induced during environmental stresses such as Similar phenotypes were found when short day-grown WT pathogen attack or vegetation shade, it accumulates in the seedlings were treated with ET gas (Das et al., 2016). To con- plant in situations where the gas-flow is limited (flooding, firm the effect of ET on several aspects of photomorpho- under a pressing soil layer, or in close canopies), and acts as genic growth in continuous white light-grown Arabidopsis a neighbor detection molecule (Dubois et al., 2018). thaliana seedlings, we germinated seeds on media with dif- Depending on the concentration, light availability, and spe- ferent concentrations of ACC and analyzed hypocotyl cies, ET can either suppress or induce plant growth (Pierik length, cotyledon unfolding, and apical hook opening. et al., 2006; Dubois et al., 2018). In young Arabidopsis seed- Compared to the control, low levels of ACC caused a slight lings, it represses several aspects of photomorphogenesis, increase of hypocotyl length in WT seedlings. This effect was such as hypocotyl growth arrest and cotyledon separation reversed by higher concentrations (4 5 mM), as visualized and expansion (Das et al., 2016; Shi et al., 2016, 2018). by the heatmap in Figure 1A, which represents relative ET is perceived by multiple ET receptors located at the values compared to WT without ACC (Supplemental golgi and ER membranes: ETHYLENE RESPONSE SENSOR 1 Figure S1). At these higher concentrations, other aspects of (ERS1), ERS2, ETHYLENE RESISTANCE 1 (ETR1), ETR2, and photomorphogenesis were repressed, resulting in partly ETHYLENE INSENSITIVE 4 (EIN4; Lacey and Binder, 2014). In unseparated cotyledons and apical hook formation the absence of ET, these receptors are active and activate (Figure 1, A–C, Supplemental Figure S1). ACC additionally Plant Physiology, 2020 PLANT PHYSIOLOGY 2020: Page 3 of 10 | 3 Downloaded from https://academic.oup.com/plphys/advance-article/doi/10.1093/plphys/kiaa015/5991408 by Universitat Autònoma de Barcelona user on 27 January 2021 Figure 1 ACC inhibits photomorphogenesis in the light via ETHYLENE INSENSITIVE2 (EIN2) and ETHYLENE INSENSITIVE3 (EIN3). A, Relative hy- pocotyl length, cotyledon separation (presented as the angle between cotyledons), and apical hook angle of 3-d-old WT Columbia-0 (Col-0), ein2, 35S::EIN2-GFP/ein2 (EIN2ox), ein3eil1, and 35S::EIN3-GFP/ein3eil1 (EIN3ox) seedlings grown in continuous low light (approx. 1.5 mmol m–2 s–1) on medium supplemented with different concentrations of ACC as compared to normal growth medium (complete dataset in Supplemental Figure S1). B, Representative seedlings grown in control medium or supplemented with 10 mM ACC. Scale bar = 1 mm. Images of plants were digitally extracted for comparison. C, Boxplots representing the absolute data for control and 10 mM ACC-treated seedlings. Same data as in (A). Different letters indicate significant differences, p 5 0.05, Kruskal–Wallis with post hoc Dunn test. In (A) and (C), biological replicates n = approx. 40 seedlings. strongly inhibited primary root growth (Figure 1B). All these Figure 2A, these ET mutants were partially insensitive to the phenotypes depended on the activation of the ET pathway lincomycin treatment and cotyledon separation was only via EIN2 and EIN3/EIL1, shown by the ACC-insensitivity of partially inhibited compared to WT. The WT-like phenotype ein2-5 and ein3eil1 mutants. For further experiments we fo- of the single ein3 and eil1 mutants as compared to the sig- cused on the cotyledon phenotype using 10 mM ACC, a nificantly distinct cotyledon phenotype of ein3eil1 in linco- concentration that strongly prevented cotyledon separation mycin, shows that EIN3 and EIL1 act redundant in the in continuous light without compromising on the overall repression of cotyledon opening (Supplemental Figure S2A). growth. Overexpression of EIN3 slightly, but significantly, enhanced the lincomycin-induced phenotype compared to the WT Suppression of photomorphogenesis by retrograde control. Contrastingly, the ctr1 mutant, with a constitutive signals requires the ET signaling pathway ET response, displayed constitutively closed cotyledons and The closed cotyledon phenotype caused by ACC resembles was insensitive to lincomycin (Figure 2A). Together, these the phenotype seen in light-grown Arabidopsis seedlings results suggest that lincomycin-induced inhibition of cotyle- treated with chloroplast inhibitor lincomycin (Figure 2A; don separation in light-grown seedlings requires ET signaling. Martı́n et al., 2016), suggesting that both might be regulated These findings were further confirmed by chemical inhibi- by a similar pathway. To test this possibility, we treated ein2 tion of ET perception by silver (AgNO3), which did not af- and ein3eil1 mutant seedlings with lincomycin. As shown in fect WT seedlings under normal light conditions but 4 | PLANT PHYSIOLOGY 2020: Page 4 of 10 Gommers et al. Downloaded from https://academic.oup.com/plphys/advance-article/doi/10.1093/plphys/kiaa015/5991408 by Universitat Autònoma de Barcelona user on 27 January 2021 Figure 2 Lincomycin-mediated repression of cotyledon separation acts via the ET signaling pathway without altering ET emission and ETHYLENE INSENSITIVE3 (EIN3) levels. A, Cotyledon separation (degrees) of 3-d-old, light-grown WT, ein2, ein3eil1, and ctr1 seedlings on regular growth me- dium or medium supplemented with 0.5 mM lincomycin (linc). Pictures are representative seedlings, scale bar = 1 mm. B, Cotyledon separation of 3-d-old, light-grown WT seedlings on medium supplemented with 5 mM AgNO3, 0.5 mM linc, or a combination of both. Pictures are represen- tative seedlings, scale bar = 1 mm. C, Cotyledon separation of WT and ET-deficient amiR(acs) mutants, grown as in (A). For (A–C), biological rep- licates n = approx. 40 seedlings. D, ET emission (in ppm/24 h) of 120 WT, amiR(acs) and gun1 seedlings grown in light, on regular growth medium, or supplemented with 1 mM (WT only) or 10 mM ACC, or 0.5 mM linc. Dots represent individual measurements (n = 4), bars are aver- ages. E, EIN3 protein accumulation in 3-d-old 35S::EIN3-GFP/ein3eil1 seedlings grown in low light on regular growth medium, or medium supple- mented with 10 mM ACC or 0.5 mM linc, detected by anti-GFP antibodies. Data are relative to light control and Coomassie staining, dots represent biological replicates (n = 3), bars are averages. Complete immunoblot image presented in Supplemental Figure S3A. In (A–E), different letters mark significant differences, p 5 0.05 (A and D: two-way ANOVA and post hoc Tukey, B and C: Kruskal–Wallis with post hoc Dunn, E: one-way ANOVA with post hoc Tukey). Images of plants in (A) and (B) were digitally extracted for comparison. partially blocked the lincomycin-induced inhibition of coty- levels of ET. We exposed this mutant to lincomycin and ledon separation (Figure 2B). Although less robustly than lin- found that the synthesis of ET is essential for the comycin, norflurazon treatment also inhibited cotyledon lincomycin-induced repression of cotyledon separation in separation significantly, an effect that was blocked by inhibi- the light (Figure 2C). Next, we used gas chromatography tion of ET perception using AgNO3 as well (Supplemental followed by mass spectrometry (GC–MS) to measure ET Figure S2B). emission by seedlings grown on low (1 mM) and higher Our findings above suggested that the suppression of (10 mM) levels of ACC and compared this to seedlings cotyledon separation in light induced by RS might be a treated with lincomycin (Figure 2D). We included consequence of ET accumulation. To test this end, we first amiR(acs) and gun1 mutants as controls for seedlings that examined the response of the ET-deficient mutant lack ACC-synthesis or retrograde signals, respectively. As amiR(acs; Tsuchisaka et al., 2009). The conversion of S- expected, 10 mM ACC strongly induced ET emission in all adenosylmethionine to ACC is the rate-limiting step dur- three genotypes. WT seedlings treated with 1 mM ACC ing ET production and is catalyzed by ACC-synthase produced less, but still significant levels of ET compared to (ACS), which acts as a dimer of various combinations of control without ACC. Nevertheless, lincomycin did not af- ACS isoforms (Tsuchisaka et al., 2009). ACC is further oxi- fect ET emissions in any of the genotypes. dized to C2H4 (ET) by ACC oxidase (ACO). Tsuchisaka Coherent with the lack of increased ET synthesis, lincomy- et al. described that the amiR(acs) octuple mutant, defi- cin did not stabilize EIN3. We used the GFP-tagged, overex- cient in nine ACS isoforms, is unable to form functional pressed EIN3 protein (p35S::EIN3-GFP/ein3eil1) to measure dimers and as a consequence produces extremely low protein stability in our experimental set-ups by immuno- Plant Physiology, 2020 PLANT PHYSIOLOGY 2020: Page 5 of 10 | 5 Downloaded from https://academic.oup.com/plphys/advance-article/doi/10.1093/plphys/kiaa015/5991408 by Universitat Autònoma de Barcelona user on 27 January 2021 Figure 3 Lincomycin and ACC signaling pathways converge downstream of ETHYLENE INSENSITIVE3 (EIN3). A, Overlap of differentially expressed genes (p 5 0.05, FC 4 j1j) in 6-d-old low-light grown 0.5 mM linc-treated seedlings, 7-d-old long-day grown 10 mM ACC-treated seedlings, and 3- d-old dark-grown pif quadruple (pifq) mutants (Goda et al., 2008; Leivar et al., 2009; Ruckle and Larkin, 2009). Letters (a, b, c) refer to B. Asterisks represent significant overlap (hypergeometric distribution, p 5 0.05). B, Violin plots representing the Log2(fold change) of the genes differentially expressed by linc and at least one other condition (a, b, c denote the gene sets defined in A). C, Heatmap representing the Log2(fold change) of the genes in (B). D, Relative expression of EBP and RAP2.6 in WT Columbia-0 (Col-0), gun1 and ein2, grown in low light on regular growth medium, or medium supplemented with 10 mM ACC (left) or 0.5 mM linc (right), measured by RT-qPCR. Light control data are similar for left and right graphs. Dots represent biological replicates (n = 3), bars are averages. Significant differences are indicated by asterisks (Student’s t test, p 5 0.05). E, Overlap of differentially expressed genes (p 5 0.05, FC 4 j1j). Lincomycin, ACC and pifq datasets (as in A), are compared to 5-d-old light-grown and lincomycin-treated (200 lg/mL) gun1 mutants as compared to WT (A and B; GSE5770). blotting. Figure 2E shows that EIN3-GFP is stabilized by both pathways might co-target a group of photomorpho- 10 mM ACC but not by lincomycin. A smaller protein genesis repressing genes downstream of EIN3. We detected in lincomycin-treated seedlings corresponds to a compared previously published microarray data of low nonspecific product as it was also present in the ein3eil1 light-grown lincomycin-treated seedlings and ACC-treated mutant background samples (Supplemental Figure S3). seedlings and additionally included a dataset of pif quadru- Finally, we tested if the inhibition of cotyledon separation ple (pifq; pif1pif3pif4pif5) seedlings grown in darkness caused by ET is regulated via GUN1-mediated chloroplast (Goda et al., 2008; Leivar et al., 2009; Ruckle et al., 2012). signals. We used gun1 knock-out and GLK1-overexpressing Lincomycin-regulated genes showed strong overlap with seedlings, both with severely reduce sensitivity to lincomycin either ACC-regulated, or PIF-regulated genes, but only five (Martı́n et al., 2016). In contrast to lincomycin, ACC could genes were differentially regulated in all three datasets inhibit cotyledon separation in gun1 and GLK1-ox, similar to (Figure 3A). The violin plot in Figure 3B represents the WT (Supplemental Figure S4). fold change of the genes in the three Venn-intersections Together, these results indicate that: (1) lincomycin- in Figure 3A. It shows that lincomycin-repressed genes are induced RS and ET inhibit cotyledon separation in light- PIF-repressed in darkness (positive Log2(FC) in pifq; 317- grown seedlings; (2) chloroplast disruption by lincomycin gene subset), like we previously concluded (Martı́n et al., does not cause changes in ET emissions; (3) GUN1 and 2016). Interestingly, and in contrast to the PIF overlap, it is GLK1 are not required for the ACC-induced inhibition of predominantly the lincomycin-induced gene set that cotyledon separation; and (4) ET signaling downstream of strongly overlaps with ACC-induced genes (95-gene sub- EIN3 is required for the lincomycin-mediated inhibition of set). Within this 95-gene subset, 66 genes are co- cotyledon separation in light. upregulated by lincomycin and ACC, which is a substantial fraction of the complete lincomycin upregulated gene set. Lincomycin and ET co-target photomorphogenesis- These results are visualized together in the heatmap in repressing genes Figure 3C, which presents the same pattern showing that Because lincomycin and ET regulate similar phenotypes, the lincomycin-repressed (negative Log2(FC) depicted in but RS did not affect ET emissions, we hypothesized that yellow)/PIF-repressed (depicted in turquoise) gene set is 6 | PLANT PHYSIOLOGY 2020: Page 6 of 10 Gommers et al. indeed not regulated by ACC, and that many of the lincomycin-induced genes (positive Log2(FC) depicted in turquoise) overlap with the ACC-upregulated set. To ex- Downloaded from https://academic.oup.com/plphys/advance-article/doi/10.1093/plphys/kiaa015/5991408 by Universitat Autònoma de Barcelona user on 27 January 2021 plore the role of the EIN3-signaling pathway in the lincomycin-induced RS, we performed additional data comparisons that show no significant enrichment of EIN3 targets among the lincomycin-regulated genes, but does show a significant co-regulation with ein3eil1 differentially expressed genes (Supplemental Figure S5A; Chang et al., 2013). The lincomycin/ACC co-induced genes do not sig- nificantly overlap with PIF-regulated genes (Figure 3C), as also shown in Figure 3A. Further comparison to light regu- lated genes (continuous red light versus continuous dark- ness (Leivar et al., 2009)) shows that, as expected from the lack of PIF-overlap, 94 of the 100 lincomycin/ACC co- regulated genes are not regulated by red light (Supplemental Figure S5B, Supplemental Table S1). We confirmed that the pifq mutant has a WT-like phenotype when grown in lincomycin that can be partly rescued by Figure 4 Graphic summary. ET represses cotyledon opening via AgNO3 (Supplemental Figure S6), which supports that RS ETHYLENE INSENSITIVE3 (EIN3) and EIN3-LIKE1 (EIL1). Retrograde via the ET signaling cascade acts independently of PIF- signals, induced by lincomycin, repress cotyledon separation via two mediated signaling. pathways: a GENOMES UNCOUPLED1 (GUN1)-dependent repression Visual analysis of the lincomycin/ACC upregulated gene of photomorphogenesis-inducing factors (e.g. GOLDEN2-LIKE1, GLK1), set revealed several transcription factors members of the and a GUN1-independent induction of EIN3/EIL1 co-targets. “X” rep- MYB, bZIP and NAC protein families (Table S1). resents a repressor of cotyledon separation downstream of EIN3/EIL1 that is induced by ET and by chloroplast signals independently of GUN1. X is likely a gene (or set of genes) regulated redundantly by A GUN1-independent retrograde signal targets EIN3 and EIL1. ET signaling components To investigate whether these lincomycin and ET co- induced genes are possibly regulated by a GUN1-mediated pathway is functionally involved in the closed-cotyledon retrograde signal, we analyzed the expression of two genes phenotype, we treated gun1 mutants with AgNO3 to simul- encoding ET response factors: ETHYLENE-RESPONSIVE taneously repress the ET signaling pathway. Even though ELEMENT BINDING PROTEIN (EBP) and RELATED TO AP2 6 gun1 seedlings show almost completely separated cotyledons (RAP2.6). Both genes were significantly induced by 10 mM in the presence of lincomycin, the inhibition of ET percep- ACC and by 0.5 mM lincomycin (Figure 3D). The induc- tion significantly increased the angle between the cotyledons tion of EBP and RAP2.6 expression by both treatments was towards control-level (Supplemental Figure S6). This sup- lost in the ein2 mutant, which confirms that the induction ports the presence of a GUN1-independent, ET-dependent, of ET response factors by lincomycin-induced retrograde RS cascade, which inhibits photomorphogenesis in light- signals requires the functional ET-response pathway. grown seedlings. Interestingly, induced expression by lincomycin was main- tained in the gun1 mutant, which hints towards a GUN1- independent retrograde pathway affecting ET-regulated Discussion seedling development. Light strongly drives plant development and promotes pho- To further explore this possible GUN1-independence, we tomorphogenesis during seedling establishment. Activated extended our analysis and compared the genes co-targeted light receptors promote greening and expansion of cotyle- by lincomycin and ACC (Figure 3A) to a previously defined dons, opening of the apical hook, and inhibition of stem list of GUN1-dependent genes. This gene list was derived growth (Gommers and Monte, 2018). Nevertheless, this pro- from light grown, lincomycin-treated seedlings of WT vs. cess is inhibited when the development of chloroplasts is gun1 mutants (Koussevitzky et al., 2007). Only 1 out of the blocked by drugs such as lincomycin (Ruckle and Larkin, 100 lincomycin and ACC co-regulated genes is regulated by 2009). The retrograde signals released by disrupted chloro- GUN1-dependent retrograde signals according to the data plasts target the PIF-repressed phy-induced transcriptome of Koussevitzky et al. (Figure 3E). Contrastingly, out of the including genes like GLK1, to inhibit photomorphogenesis in 322 pifq and lincomycin co-targeted genes, 115 (35.7%) are the light (Martı́n et al., 2016). Our findings here support a regulated via GUN1 (Figure 3E). These data strongly suggest model whereby ET prevents excessive cotyledon separation that lincomycin induces ET-responsive genes independently under continuous light by regulating a set of ET-induced of GUN1 (Figure 4). To explore if this GUN1-independent genes downstream of the transcription factor ETHYLENE Plant Physiology, 2020 PLANT PHYSIOLOGY 2020: Page 7 of 10 | 7 INSENSITIVE3 (EIN3). In turn, when retrograde signaling (RS) levels. We consider the first two scenarios most likely, since is induced during early photomorphogenesis, inhibition of overexpression of EIN3 could significantly enhance the unse- cotyledon separation by chloroplast signals takes place at parated cotyledon phenotype compared to the WT control Downloaded from https://academic.oup.com/plphys/advance-article/doi/10.1093/plphys/kiaa015/5991408 by Universitat Autònoma de Barcelona user on 27 January 2021 least in part by further inducing this same ET signaling path- (Supplemental Figure S2A). way. Both signals converge to co-regulate a set of RS/ET- The origin of the GUN1-independent retrograde signal induced genes. The chloroplast signal is yet undefined, but it that targets the ET signaling pathway remains speculative. does not involve GUN1 (Figure 4). However, it is interesting to note that the ET pathway Interestingly, this RS/ET-targeted gene set does not has been associated with different stress-induced chloroplast overlap with the RS/PIF-repressed network previously de- signals before. The chloroplast stress-metabolite MEcPP was fined in our laboratory (Figure 3; Martı́n et al., 2016). This recently shown to repress ET synthesis and hypocotyl elon- finding suggests that RS optimizes seedling development gation in red light in an auxin-dependent fashion (Jiang in response to chloroplast disruption (caused for example et al., 2020). Additionally, reactive oxygen species (ROS) re- by high light) by targeting at least two distinct gene net- leased by mature (green) chloroplasts upon high light stress works that independently inhibit cotyledon separation: induce expression of ET response factors in the nucleus (1) a light-regulated network downstream of the phyto- (Vogel et al., 2014). Lastly, drought and high light stress in- chrome/PIF system, and (2) a light-independent ET duce 30 -phosphoadenisone 50 -phosphate (PAP) accumula- hormone-regulated pathway downstream of EIN3. tion in mature chloroplasts that act as RS and inhibit Existence of these two separate regulatory pathways is exoribonucleases (XRNs) such as XRN4, involved in the ET supported by the strong photomorphogenic phenotype of signaling pathway as it targets EBF1 and EBF2 mRNA for pifqein3eil1 in darkness compared to pifq and ein3eil1 (Shi degradation (Estavillo et al., 2011). Future work will be nec- et al., 2018). Remarkably, whereas RS antagonizes the essary to address whether these signals are also relevant in light-regulated network by repressing light-induced genes young seedlings grown in continuous white light, such (Martı́n et al., 2016), our results here suggest that RS fur- as the ones used in our study. We speculate that young ther induces an ET-induced network. This dichotomy seedlings with developing chloroplasts might release bio- resulting in the repression or promotion of nuclear gene genic signals via GUN1 as well as ROS and/or PAP to trigger expression might explain why the chloroplast utilizes two ET responses. distinct signaling pathways: a GUN1-dependent to repress To conclude, our results here together with previous work the light-induced gene network and a GUN1-independent demonstrate that lincomycin represses photomorphogenesis pathway to induce ET-regulated genes. Although retro- via at least two separate pathways. One requires GUN1- grade signals other than GUN1 have been described in re- mediated signals and represses PIF-repressed target genes. sponse to different stresses (Chan et al., 2016), GUN1 is The other requires GUN1-independent retrograde signals considered as a hub factor of RS during chloroplast bio- and induces the ET response pathway. Future work will ex- genesis (Mochizuki et al., 1996; León et al., 2013; plore if this GUN1-independent pathway involves any of the Hernandez-Verdeja and Strand, 2018; Wu et al., 2018). other known retrograde signaling mechanisms, or if it repre- Our findings here suggest the existence of at least one ad- sents a novel and yet undescribed pathway. ditional GUN1-independent pathway to optimally adjust seedling development to the light environment. Materials and methods Cotyledon closing by EIN3 in darkness requires activation Plant material and growth conditions of the ET signaling pathway by ACC (Shi et al., 2018). We Arabidopsis thaliana seeds used here were all described be- showed genetically and chemically that the inhibition of cot- fore, including ein2-5, ein3-1/eil1-1, p35S::EIN2-GFP/ein2-5, yledon separation by lincomycin also requires functioning p35S::EIN3-GFP/ein3-1/eil1-1, ctr1-1, amiR(acs), gun1-201, ET perception and EIN2 and EIN3/EIL1 signaling, but does p35S::GLK1/glk1glk2, and pifq, all in the Columbia-0 WT not enhance ET emission. We did not see a stabilization of background (Kieber et al., 1993; Leivar et al., 2008; Waters the EIN3 protein by lincomycin, which is coherent with the et al., 2008; Tsuchisaka et al., 2009; He et al., 2011; Ju et al., lack of induced ET synthesis. 2012; Wen et al., 2012; Xie et al., 2015; Martı́n et al., 2016). Despite the undetectable stabilization of EIN3-GFP, tran- For all experiments, seeds were surface-sterilized and sowed scriptional targets of the EIN2/EIN3 signaling module were on half-strength Murashige and Skoog (0.5 MS) medium upregulated by lincomycin treatment and ET signaling (0.8% w/v plant agar), followed by a 4-d stratification treat- mutants showed reduced lincomycin sensitivity. Different ment before moving to continuous low white light (approx. scenarios could explain our results: (1) lincomycin might 1.5 mmol m–2 s–1, T8 LED tube 4000K, Systion Electronics) minimally increase ET synthesis, undetectable in our set-up at 20 C. Phenotypes, as well as gene, protein, and ET quanti- but sufficient to allow the activation of the signaling cascade fication were analyzed after 3 d. downstream of EIN3, (2) lincomycin might stabilize EIN3 transiently during seedling establishment and we might have Pharmacological treatments missed it at the selected time point, or (3) lincomycin might 1-Aminocyclopropane-1-carboxylic acid (ACC; VWR P10007) promote an increase in EIN3 activity without affecting EIN3 was dissolved in sterile water (10 mM stock) and added to 8 | PLANT PHYSIOLOGY 2020: Page 8 of 10 Gommers et al. the growth medium to obtain different final concentrations GCC-30 ), RAP2.6 (AT1G43160; 50 -TGATTACCGGTTCAGCT (0.1, 1, 5, 10, 20, 50 mM). Lincomycin (Sigma-Aldrich L6004) GTG-30 and 50 -CTTGTGTGGGTCTCGAATCT-30 ). Expression was added directly to the growth medium to obtain 0.5 of PP2A was analyzed as a reference (AT1G13320; 50 -TAT Downloaded from https://academic.oup.com/plphys/advance-article/doi/10.1093/plphys/kiaa015/5991408 by Universitat Autònoma de Barcelona user on 27 January 2021 mM as described before (Martı́n et al., 2016). Silver nitrate CGGATGACGATTCTTCGT-30 and 50 -GCTTGGTCGACT (AgNO3; Sigma-Aldrich 209139) was dissolved in sterile wa- ATCGGAATG-30 ). Relative expression was calculated as ter (1 mM stock) and added to the growth medium in a fi- 2 ^(delta-delta CT). nal concentration of 5 mM. Norflurazon (Sigma-Aldrich 34364) was added directly to the growth medium to obtain Protein analysis a 5 mM solution. Protein extracts were prepared of 3-d-old seedlings, pooled per plate for each sample, flash-frozen, and homogenized by Phenotype analysis hand. Nuclear protein was extracted with an extraction After 3 d of growth, seedlings were photographed (Nikon buffer as described before (Soy et al., 2014; buffer: 100 mM D7000 camera) and pictures were analyzed with Fiji MOPS (pH 7.6), 2% SDS, 10% glycerol, 4 mM EDTA, 50 mM (Schindelin et al., 2012). For phenotyping experiments, Na2S2O5, 2 mgL–1 aprotinin, 3 mgL–1 leupeptine, 1 mgL–1 pep- n = 36–40 biological replicates (seedlings) per treatment and statin, and 2 mM PMSF). Total protein of the samples was genotype, divided over two petri dishes (18–20 seedlings of quantified using a Protein DC kit (Bio-Rad). b- each genotype per petri dish). Mercaptoethanol and loading dye was added to 125 mg of the samples, which were boiled at 95 C for 5 min before be- ET measurements ing loaded on a 7.5% SDS-PAGE gel. Proteins were then To measure ET emissions, seeds were sown in sterilized glass transferred to Immobilon-P membrane (Millipore) and EIN3- vials (10 mL “Headspace Vials” with screw heads, Restek) GFP was detected using an anti-GFP antibody (diluted which contained 7 mL of growth medium (with or without 1:10,000; Invitrogen A11122). Anti-rabbit secondary antibody pharmacological treatment). Each vial contained 120 seeds. (Sigma NA934) and SuperSignal West Femto chemilumines- Seeds were cold-stratified and germinated as described cence kit (Pierce) were used for protein detection in a Bio- above. Forty-eight hours after the transfer to light, the air Rad imaging system. The ImageLab program (Bio-rad) was headspace of the vials was flushed with clean air to set a used to quantify band intensity and sizes from blot images, starting point for ET accumulation. Twenty-four hours later, which was compared to a reference from the Coomassie- measurements were taken. Growth protocol was adapted stained blot. from (Jeong et al., 2016). Three mL air was taken from the vial’s headspace and transferred to clean (flushed) Restek vials. ET concentrations were measured by GC–MS with an Statistical analysis Agilent 7890A gas chromatograph coupled to a 5975C mass Multivariate comparisons were done in R and using the on- selective detector, as described before (Pereira et al., 2017). line MVApp (Julkowska et al., 2019). First, data were checked for equal variance by Levene’s test and when needed, data were LN-transformed to make variance equal. Multivariate Transcriptome re-analysis analyses were done by one- or two-way ANOVA with a Published and publicly available micro-array datasets (acces- post hoc Tukey test for pairwise comparisons. For experi- sion numbers: GSE24517, GSE17159, GSE39384, GSE5770, ments with nonequal variances, the nonparametric Kruskal– and GSE21762) were re-analyzed using the GEO2R online Wallis test was applied, with a post hoc Dunn test or tool (NCBI). Differentially expressed genes were selected by Mann–Whitney for pairwise comparison. Student’s t test 2-fold (Log2 fold 4 j1j) regulation and adjusted p5 0.05. pairwise comparisons were done in Microsoft Excel preceded For comparison with chromatin immunoprecipitation fol- by an F test to test for equal variances and LN-transformed lowed by sequencing (ChIP-seq) data (Supplemental Figure when needed. Hypergeometric distribution tests for signifi- S5A), we have removed the EIN3 target genes that are not cant overlap in Venn-diagrams were done in R. represented on the Affimetrix chip, which was used for the micro-array experiment. Accession numbers Gene expression analysis Sequence data from this article can be found in the To analyze gene expression, 20 seedlings were harvested and NCBI gene expression omnibus (GEO) data repository under pooled per sample, flash-frozen, homogenized and RNA was accession numbers: GSE24517, GSE17159, GSE39384, extracted using the Maxwell total RNA purification kit with GSE5770, and GSE21762. Arabidopsis gene identifiers for DNAse treatment (Promega). cDNA was synthesized using the major genes and proteins mentioned are: AT3G20770 the NZYtech first-strand cDNA synthesis kit with random (ETHYLENE INSENSITIVE3; EIN3), AT2G27050 (EIN3- primers and afterwards treated with RNAse. RT-qPCR was LIKE1; EIL1), AT5G03280 (ETHYLENE INSENSITIVE2; EIN2), performed using a Roche Lightcycler 480, SYBR Green mix AT2G31400 (GENOMES UNCOUPLED1; GUN1), AT2G20180 (Roche) and primers for EBP (AT3G16770; 50 -CCCA (PHYTOCHROME INTERACTING FACTOR1; PIF1), CCAACCAAGTTAACGT-30 and 50 -GTGGATCTCGAATCTCA AT1G09530 (PIF3), AT2G43010 (PIF4), AT3G59060 (PIF5), Plant Physiology, 2020 PLANT PHYSIOLOGY 2020: Page 9 of 10 | 9 AT2G20570 (GOLDEN2-LIKE1; GLK1), AT5G03730 References (CONSTITUTIVE TRIPLE RESPONSE1; CTR1). Binder BM (2020) Ethylene signaling in plants. J Biol Chem 295: 7710–7725 Downloaded from https://academic.oup.com/plphys/advance-article/doi/10.1093/plphys/kiaa015/5991408 by Universitat Autònoma de Barcelona user on 27 January 2021 Supplemental data Chan KX, Phua SY, Crisp P, McQuinn R, Pogson BJ (2016) Learning the languages of the chloroplast: Retrograde signaling and The following materials are available in the online version of beyond. Annu Rev Plant Biol 67: 25–53 this article. Chang KN, Zhong S, Weirauch MT, Hon G, Pelizzola M, Li H, Supplemental Figure S1. ACC inhibits photomorphogen- Carol Huang SS, Schmitz RJ, Urich MA, Kuo D, et al. (2013) Temporal transcriptional response to ethylene gas drives growth esis via ETHYLENE INSENSITIVE2 (EIN2) and ETHYLENE hormone cross-regulation in Arabidopsis. eLife 2: e00675 INSENSITIVE3 (EIN3). Crawford T, Lehotai N, Strand Å (2018) The role of retrograde sig- Supplemental Figure S2. ETHYLENE INSENSITIVE3 (EIN3) nals during plant stress responses. J Exp Bot 69: 2783–2795 and EIN3-LIKE1 (EIL1) act redundantly to repress cotyledon Das D, St Onge KR, Voesenek LACJ, Pierik R, Sasidharan R (2016) separation during lincomycin-induced stress, and ET percep- Ethylene- and shade-induced hypocotyl elongation share transcrip- tome patterns and functional regulators. Plant Physiol 172: tion is required for norflurazon-mediated inhibition of coty- 718–733 ledon separation. Dubois M, Van den Broeck L, Inzé D (2018) The pivotal role of eth- Supplemental Figure S3. Immunoblots for the EIN3-GFP ylene in plant growth. Trends Plant Sci 23: 311–323 protein. Estavillo GM, Crisp PA, Pornsiriwong W, Wirtz M, Collinge D, Supplemental Figure S4. ACC inhibits photomorphogen- Carrie C, Giraud E, Whelan J, David P, Marin E, et al. (2011) Evidence for a SAL1-PAP chloroplast retrograde pathway that func- esis independent of GENOMES UNCOUPLED1 (GUN1) and tions in drought and high light signaling in Arabidopsis. Plant Cell GOLDEN2-LIKE1 (GLK1). 23: 3992–4012 Supplemental Figure S5. Transcriptional comparison of Goda H, Sasaki E, Akiyama K, Maruyama-Nakashita A, lincomycin-regulated genes with ETHYLENE INSENSITIVE3 Nakabayashi K, Li W, Ogawa M, Yamauchi Y, Preston J, Aoki K, (EIN3) targets, EIN3- and EIN3-LIKE1 (EIL1)-regulated genes, et al. (2008) The AtGenExpress hormone and chemical treatment data set: experimental design, data evaluation, model data analysis and red light regulated genes. and data access. Plant J 55: 526–542 Supplemental Figure S6. AgNO3 inhibits lincomycin- Gommers CMM, Monte E (2018) Seedling establishment: a dimmer induced inhibition of cotyledon separation independent of switch-regulated process between dark and light signaling. Plant PIFs and GENOMES UNCOUPLED1 (GUN1). Physiol 176: 1061–1074 Supplemental Table S1. List of genes co-regulated by lin- He W, Brumos J, Li H, Ji Y, Ke M, Gong X, Zeng Q, Li W, Zhang X, An F, et al. (2011) A small-molecule screen identifies L-Kynurenine comycin and 1h ACC or pifq in darkness, which are marked as a competitive inhibitor of TAA1/TAR activity in in Figure 3A and plotted in Figure 3B, with fold-change and ethylene-directed auxin biosynthesis and root growth in adjusted p-value (Goda et al., 2008; Leivar et al., 2009; Ruckle Arabidopsis. Plant Cell 23: 3944–3960 and Larkin, 2009). Hernandez-Verdeja T, Strand A (2018) Retrograde signals navigate the path to chloroplast development. Plant Physiol 176: 967–976 Acknowledgments Jeong J, Kim K, Kim ME, Kim HG, Heo GS, Park OK, Park Y-i, Choi G, Oh E (2016) Phytochrome and ethylene signaling integra- We thank Jason Argyris (CRAG, IRTA) for his help to prepare tion in Arabidopsis occurs via the transcriptional regulation of the ET measurements, Zeguang Liu, Sjon Hartman, and genes co-targeted by PIFs and EIN3. Front Plant Sci 7: 1–14 Ronald Pierik (Utrecht University) for sharing seeds of the ET Jia Y, Tian H, Zhang S, Ding Z, Ma C (2019) GUN1-interacting pro- signaling mutants, and lab members Arnau Rovira, Nil teins open the door for retrograde signaling. Trends Plant Sci 24: Veciana, Liu Duan, and Ana Couso for their help to harvest 884–887 Jiang J, Xiao Y, Chen H, Hu W, Zeng L, Ke H, Ditengou FA, materials and support. Devisetty U, Palme K, Maloof J, et al. (2020) Retrograde induc- tion of phyB orchestrates ethylene-auxin hierarchy to regulate Funding growth. Plant Physiol 183: 1268–1280 Ju C, Yoon GM, Shemansky JM, Lin DY, Ying ZI, Chang J, Garrett This work was supported by grants from FEDER/Ministerio WM, Kessenbrock M, Groth G, Tucker ML, et al. (2012) CTR1 de Ciencia, Innovación y Universidades – Agencia Estatal de phosphorylates the central regulator EIN2 to control ethylene hor- Investigación (Project References BIO2015-68460-P and mone signaling from the ER membrane to the nucleus in Arabidopsis. Proc Natl Acad Sci USA 109: 19486–19491 PGC2018-099987-B-I00), from the Spanish Ministry of Julkowska MM, Saade S, Agarwal G, Gao G, Pailles Y, Morton M, Economy and Competitiveness (FJCI-2016-30876 to C.G.) Awlia M, Tester M (2019) MVApp—multivariate analysis applica- and from the CERCA Programme/Generalitat de Catalunya tion for streamlined data analysis and curation. Plant Physiol 180: (Project Reference 2017SGR-718) to E.M. We acknowledge 1261–1276 the financial support from the Spanish Ministry of Economy Kieber JJ, Rothenberg M, Roman G, Feldmann KA, Ecker JR (1993) CTR1, a negative regulator of the ethylene response pathway in and Competitiveness, through the “Severo Ochoa Arabidopsis, encodes a member of the Raf family of protein kin- Programme for Centres of Excellence in R&D” 2016-2019 ases. Cell 72: 427–441 (SEV-2015-0533). Koussevitzky S, Nott A, Mockler TC, Hong F, Sachetto-Martins G, Surpin M, Lim J, Mittler R, Chory J (2007) Signals from chloro- Conflict of interest statement. We declare no conflicts of plasts converge to regulate nuclear gene expression. Science 316: interest. 715–719 10 | PLANT PHYSIOLOGY 2020: Page 10 of 10 Gommers et al. Lacey RF, Binder BM (2014) How plants sense ethylene gas - the (2012) Fiji: an open-source platform for biological-image analysis. ethylene receptors. J Inorganic Biochem 133: 58–62 Nature Methods 9: 676. 10.1038/nmeth.2019 22743772 Leivar P, Monte E, Oka Y, Liu T, Carle C, Castillon A, Huq E, Shi H, Lyu M, Luo Y, Liu S, Li Y, He H, Wei N, Wang X Quail PH (2008) Multiple phytochrome-interacting bHLH tran- (2018) Genome-wide regulation of light-controlled seedling mor- Downloaded from https://academic.oup.com/plphys/advance-article/doi/10.1093/plphys/kiaa015/5991408 by Universitat Autònoma de Barcelona user on 27 January 2021 scription factors repress premature seedling photomorphogenesis phogenesis by three families of transcription factors. Proc Natl in darkness. Curr Biol 18: 1815–1823 Acad Sci 115: 2–7 Leivar P, Tepperman JM, Monte E, Calderon RH, Liu TL, Quail PH Shi H, Shen X, Liu R, Xue C, Wei N, Deng XW, Zhong S (2016) The (2009) Definition of early transcriptional circuitry involved in red light receptor phytochrome B directly enhances substrate-E3 li- light-induced reversal of PIF-imposed repression of photomorpho- gase interactions to attenuate ethylene responses. Dev Cell 39: genesis in young Arabidopsis seedlings. Plant Cell 21: 3535–3553 1–14 León P, Gregorio J, Cordoba E (2013) ABI4 and its role in chloro- Soy J, Leivar P, Monte E (2014) PIF1 promotes plast retrograde communication. Front Plant Sci 3 phytochrome-regulated growth under photoperiodic conditions in Martı́n G, Leivar P, Ludevid D, Tepperman JM, Quail PH, Monte E Arabidopsis together with PIF3, PIF4, and PIF5. J Exp Bot 65: (2016) Phytochrome and retrograde signalling pathways converge 2925–2936 to antagonistically regulate a light-induced transcriptional network. Tsuchisaka A, Yu G, Jin H, Alonso JM, Ecker JR, Zhang X, Gao S, Nat Commun 7: 11431 Theologis A (2009) A combinatorial interplay among the 1-amino- Mochizuki N, Susek R, Chory J (1996) An intracellular signal trans- cyclopropane-1-carboxylate isoforms regulates ethylene biosynthe- duction pathway between the chloroplast and nucleus is involved sis in Arabidopsis thaliana. Genetics 183: 979–1003 in de-etiolation. Plant Physiol 112: 1465–1469 Vogel MO, Moore M, Konig K, Pecher P, Alsharafa K, Lee J, Dietz Pereira L, Pujol M, Garcia-Mas J, Phillips MA (2017) Non-invasive K-J (2014) Fast retrograde signaling in response to high light quantification of ethylene in attached fruit headspace at 1 p.p.b. involves metabolite export, MITOGEN-ACTIVATED PROTEIN by gas chromatography–mass spectrometry. Plant J 91: 172–183 KINASE6, and AP2/ERF transcription factors in Arabidopsis. Plant Pesaresi P, Kim C (2019) Current understanding of GUN1: A key Cell 26: 1151–1165 mediator involved in biogenic retrograde signaling. Plant Cell Rep Waters MT, Moylan EC, Langdale JA (2008) GLK transcription fac- 38: 819–823 tors regulate chloroplast development in a cell-autonomous man- Pierik R, Tholen D, Poorter H, Visser EJW, Voesenek LACJ (2006) ner. Plant J 56: 432–444 The Janus face of ethylene: Growth inhibition and stimulation. Wen X, Zhang C, Ji Y, Zhao Q, He W, An F, Jiang L, Guo H (2012) Trends Plant Sci 11: 176–183 Activation of ethylene signaling is mediated by nuclear transloca- Ruckle ME, Burgoon LD, Lawrence LA, Sinkler CA, Larkin RM tion of the cleaved EIN2 carboxyl terminus. Cell Res 22: 1613 (2012) Plastids are major regulators of light signaling in Wu G-Z, Chalvin C, Hoelscher M, Meyer EH, Wu XN, Bock R Arabidopsis. Plant Physiol 159: 366–390 (2018) Control of retrograde signaling by rapid turnover of Ruckle ME, Larkin RM (2009) Plastid signals that affect photomor- GENOMES UNCOUPLED1. Plant Physiol 176: 2472–2495 phogenesis in Arabidopsis thaliana are dependent on GENOMES Xie LJ, Chen QF, Chen MX, Yu LJ, Huang L, Chen L, Wang FZ, Xia UNCOUPLED 1 and cryptochrome 1. New Phytol 182: 367–379 FN, Zhu TR, Wu JX, et al. (2015) Unsaturation of very-long-chain Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, ceramides protects plant from hypoxia-induced damages by modu- Pietzsch T, Preibisch S, Rueden C, Saalfeld S, Schmid B, et al. lating ethylene signaling in Arabidopsis. PLoS Genet 11: 1–33