https://orcid.org/0000-0002-9425-6538
Abstract
The kinases SNF1-RELATED KINASE 1 (SnRK1) and TARGET OF RAPAMYCIN (TOR) are central sensors of the energy status, linking this information via diverse regulatory mechanisms to plant development and stress responses. Despite the well-studied functions of SnRK1 and TOR under conditions of limited or ample energy availability, respectively, little is known about the extent to which the 2 sensor systems function and how they are integrated in the same molecular process or physiological context. Here, we demonstrate that both SnRK1 and TOR are required for proper skotomorphogenesis in etiolated Arabidopsis (Arabidopsis thaliana) seedlings, light-induced cotyledon opening, and regular development in light. Furthermore, we identify SnRK1 and TOR as signaling components acting upstream of light- and sugar-regulated alternative splicing events, expanding the known action spectra for these 2 key players in energy signaling. Our findings imply that concurring SnRK1 and TOR activities are required throughout various phases of plant development. Based on the current knowledge and our findings, we hypothesize that turning points in the activities of these sensor kinases, as expected to occur upon illumination of etiolated seedlings, instead of signaling thresholds reflecting the nutritional status may modulate developmental programs in response to altered energy availability.
Background: Plants adjust their development to light and metabolic signals to make best use of their available resources. Dark-grown seedlings undergo etiolation with increased hypocotyl elongation, while illumination causes greening of leaves to initiate photosynthesis. These alternative developmental programs require reprograming of many genes via various mechanisms. Accordingly, exposing etiolated seedlings to light or sugar causes changes in splicing, a critical step in the generation of mature mRNAs. Key sensors for metabolic signals are SNF1-RELATED KINASE 1 (SnRK1) and TARGET OF RAPAMYCIN (TOR), which were shown to be activated under conditions of limited and ample energy availability, respectively.
Question: Given the importance of SnRK1 and TOR in sensing the metabolic state, we examined their role in light-dependent seedling development and splicing. We were particularly interested in establishing if they have opposing functions and whether this differs for seedlings grown in the presence or absence of light.
Findings: We have established mutants in the model plant Arabidopsis thaliana for inducible repression of SnRK1 and TOR. Mutant analysis revealed that both components are needed for regular seedling development in darkness and light. Furthermore, SnRK1 and TOR are involved in controlling light-regulated splicing events. Knockdown of either SnRK1 or TOR caused similar changes in splicing as those observed upon exposing etiolated seedlings to light or sucrose. Our findings demonstrate that concurring activities of these 2 energy sensors are indispensable for proper regulation of gene expression and seedling development.
Next steps: SnRK1 and TOR are assumed to have antagonistic functions in energy sensing. Resolving their crosstalk in various developmental conditions will be a key aspect of future research. Moreover, it will be of interest to examine the mechanism by which SnRK1 and TOR can alter the splicing outcome.
Introduction
The photoautotrophic lifestyle of plants requires precise monitoring of the ambient light conditions and its integration with the metabolic status to adjust the plant's physiology on multiple levels, ranging from biochemical to developmental adaptations. The underlying components of signal perception and transduction, including multiple classes of photoreceptors for a spectrum of light signals (Galvão and Fankhauser 2015) and the central energy sensor kinases SNF1-RELATED KINASE 1 (SnRK1) and TARGET OF RAPAMYCIN (TOR), have been intensively studied (Li and Sheen 2016; Sakr et al. 2018; Margalha et al. 2019; Li et al. 2021; Jamsheer et al. 2021; Henriques et al. 2022). Previous studies revealed the antagonistic functions of these 2 energy sensing kinases. While SnRK1 is activated under low energy conditions to inhibit growth and energy-consuming processes, TOR is activated when ample levels of nutrients are available to promote growth and cell proliferation. Despite the structural and functional conservation of both complexes in yeast, mammals, and plants, they also have evolved several plant-specific features and functions (Emanuelle et al. 2015; Ramon et al. 2019; Broeckx et al. 2016; Liu and Xiong 2022). This may represent an adaptation of SnRK1 and TOR to the peculiarities and challenges of a photoautotrophic lifestyle.
Plant SnRK1 kinases can form heterotrimeric complexes that contain 1 catalytic α subunit, similar to their orthologs in yeast and mammals (Broeckx et al. 2016; Sakr et al. 2018). In Arabidopsis thaliana, the α subunit is mainly encoded by the homologs SnRK1.1 (KIN10) and SnRK1.2 (KIN11), while the third homolog SnRK1.3 (KIN12) is expressed only at low levels and in specific tissues (Margalha et al. 2019). The conserved TOR kinase forms 2 types of complexes, TOR complex 1 (TORC1) and TOR complex 2 (TORC2) in yeast and mammals, whereas in plants TOR function is restricted to TORC1 according to our current knowledge (van Dam et al. 2011, Liu and Xiong 2022). TOR signaling can be suppressed by SnRK1, as shown for human cells (Gwinn et al. 2008) and also for plants (Nukarinen et al. 2016). Recently, the sequestration of SnRK1 in SnRK2-containing complexes was described as a novel mechanism that protects TOR from inhibition under growth-promoting conditions in A. thaliana (Belda-Palazón et al. 2020). Conversely, SnRK1 gets released under stress conditions that involve the phytohormone abscisic acid, leading to TOR inhibition and growth retardation (Belda-Palazón et al. 2020). Moreover, a negative feedback loop restricting TOR signaling under favorable conditions was recently described by Jamsheer et al. (2022). Accordingly, TOR promotes the levels of the FCS-LIKE ZINC FINGER protein 8 (FLZ8), which activates SnRK1 that in turn inhibits TORC1. This intricate crosstalk within the SnRK1-TOR network allows balancing the antagonistic functions of growth and stress responses both under regular conditions and when plants are challenged, thereby improving plant fitness.
The plant's energy status is typically defined by the availability of assimilation products and adequate illumination; however, little is known about the steps and mechanisms integrating the underlying signaling pathways that sense light and sugar signals. Interestingly, 2 studies provided evidence for an involvement of TOR in this context. First, Pfeiffer et al. observed that TOR is required for the integration of light and metabolic signals during stem cell activation at the shoot apical meristem in A. thaliana (Pfeiffer et al. 2016). Second, TOR has been demonstrated to contribute to photomorphogenesis of etiolated A. thaliana seedlings via translational enhancement, acting downstream of photoreceptors and CONSTITUTIVE PHOTOMORPHOGENESIS 1 (COP1), a negative regulator of light-dependent development (Chen et al. 2018). Given the importance of SnRK1 and TOR in sensing the metabolic state, both signaling pathways could be critical for the integration of light and sugar signals during photomorphogenesis.
The photomorphogenic response has been demonstrated to be accompanied and driven by massive transcriptome re-programming, including changes in alternative precursor mRNA splicing (AS) upon illumination of etiolated A. thaliana seedlings (Shikata et al. 2014; Hartmann et al. 2016). Accordingly, the splicing regulators REDUCED RED-LIGHT RESPONSES IN cry1cry2 BACKGROUND 1 (RRC1; Shikata et al. 2012; Xin et al. 2019), SPLICING FACTOR FOR PHYTOCHROME SIGNALING (SFPS; Xin et al. 2019; Xin et al. 2017), and SUPPRESSOR-OF-WHITE-APRICOT/SURP RNA-BINDING DOMAIN-CONTAINING PROTEIN 1 (SWAP1; Kathare et al. 2022) promote photomorphogenesis and interact with the red light receptor PHYTOCHROME B (PHYB). However, light-responsive AS events can also be regulated in a photoreceptor-independent manner (Hartmann et al. 2016; Petrillo et al. 2014; Mancini et al. 2016). In this regard, external sugar supply was found to elicit similar AS changes as illumination in etiolated seedlings (Hartmann et al. 2016), and evidence for an involvement of retrograde signaling from the chloroplast to the nucleus in AS control was provided (Petrillo et al. 2014). In a recent report, inhibition of TOR signaling in seedlings grown in constant light before exposure to an extended dark period and subsequent treatment with light and/or sugars revealed the requirement of TOR for proper AS responses (Riegler et al. 2021). However, the role of SnRK1 in this context has remained unclear. Illumination of etiolated seedlings is expected to affect both SnRK1 and TOR signaling, but their function in the regulation of AS and the skoto-/photomorphogenesis-related developmental processes are unresolved. In this study, we demonstrate that repression of SnRK1 and TOR signaling can similarly alter light-dependent AS events and developmental attributes in etiolated seedlings, providing insight into the activity windows, target processes, and physiological implications of these 2 central energy sensor kinases in plants.
Results and discussion
To examine the functions of SnRK1 and TOR in light-dependent gene regulation and development, we developed silencing constructs based on artificial microRNAs (amiRs) to specifically knock down these critical signaling components (Fig. 1A). Previous studies showed that downregulation of both SnRK1.1 and SnRK1.2 via virus-induced gene silencing (VIGS) causes early senescence (Baena-González et al. 2007), whereas single snrk1 mutants do not display an obvious aberrant growth phenotype (Baena-González et al. 2007; Mair et al. 2015), probably due a functional redundancy of the 2 SnRK1 homologs. We assumed that simultaneous targeting of SnRK1.1 and SnRK1.2 via constitutively expressed amiRs (Fig. 1A; Supplemental Fig. S1, A and B) in A. thaliana may result in a remaining level of SnRK1 activity that prevents lethality on the one hand and causes discernible phenotypes on the other hand. Following these mutants' development revealed the frequent occurrence of phenotypical alterations, including dwarfism, early bolting, increased mortality rates, and reduced hypocotyl elongation of etiolated seedlings (Supplemental Fig. S1, C to G). Moreover, ∼30% of the plants that were resistant to the selection agent and therefore must carry the transgene died after the selection step before seed generation (Supplemental Table S1). Analyzing 3 subsequent generations (up to 2 generations after the primary transformants) for several independent lines did not result in a nonsegregating mutant, suggesting homozygosity causes lethality due to a dose effect. In line with our observations, simultaneous downregulation of SnRK1.1 and SnRK1.2 via VIGS or a combination of VIGS and RNA interference in A. thaliana also strongly reduced growth and in addition caused anthocyanin accumulation (Baena-González et al. 2007; Mair et al. 2015).
Establishing inducible knockdown lines for SnRK1 and TOR. A) Full SnRK1.1/1.2 and partial TOR transcript structures based on representative gene models, with target sites of amiRs (rhomb symbols; dotted line indicates positioning over an exon–exon border). T-DNA insertion site in snrk1.1-3 is indicated by a white triangle. Lines correspond to introns; black and gray shapes depict UTRs and coding exons, respectively. First, last, and amiR-targeted exons are numbered. B) Relative transcript levels of SnRK1.1 and SnRK1.2 in 6-d-old etiolated WT and mutant seedlings allowing inducible amiR expression (i-amiR), treated with either Mock (white bars) or estradiol (Est, gray bars) for 3 d. Seedlings were grown in liquid culture. Data are mean values (n = 3; individual data points as dots) ± Sd, normalized to WT Mock samples. A 1-sample t-test was performed in comparison to WT Mock. C) Immunoblot detection of phosphorylated SnRK1.1 (black triangle) and SnRK1.2 proteins (white triangle, upper panel) in WT and different snrk1 mutant seedlings. Ponceau S staining is shown as loading control (lower panel). Other details of plant growth and treatments as described in B). D) Relative transcript level of the SnRK1 target DIN1 in WT and i-amiR-SnRK1 seedlings. Sample information, data normalization, and statistics as described in B). E) Relative transcript level of TOR in liquid-grown 6-d-old etiolated WT and i-amiR-TOR mutant seedlings, treated with either Mock (white bar) or Est (gray bar) for 3 d. Other display details and statistical analyses here as defined in B). F) (Left) Immunoblot detection of pS6K1 from samples as described in E). Tubulin detected on a separate membrane served as loading control. (Right) Quantification of relative pS6K1 protein level. Data are mean values (n = 3) ± Sd, normalized to WT Mock samples. G) Relative transcript level of DIN1 and DIN6 in WT and i-amiR-TOR mutant seedlings as defined in E). Asterisks indicate significant difference of i-amiR-TOR Est to corresponding Mock control based on 2-tailed Student's t-test. A 1-sample t-test was performed for the comparison of WT Est to WT Mock. P values: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
We concluded from these data that our amiRs are likely functional but need to be expressed in an inducible manner to avoid more general growth defects and increased mortality rates. In comparison to the approach from Nukarinen et al. (2016) using an inducible amiR against SnRK1.2 in a snrk1.1 T-DNA mutant background, we expected regular activities of both SnRK1 homologs under noninduced conditions. Plant transformants carrying a transgene for estradiol-inducible amiR expression (i-amiR-SnRK1) showed normal development when grown without application of the inducer, indicating sufficiently tight control of amiR production (Supplemental Fig. S2, A and B). Comparing hypocotyl lengths in mock- and estradiol-treated etiolated seedlings revealed reduced hypocotyl lengths upon amiR induction for both i-amiR constructs and several independent lines (Supplemental Fig. S2C). This observation was in agreement with the phenotype seen for our constitutive amiR lines upon seedling growth in darkness (Supplemental Fig. S1G). One line from each i-amiR construct was propagated in the following generations to obtain homozygous mutants for further characterisation. Exposing etiolated seedlings from such homozygous mutants for 3 d to estradiol strongly diminished transcript levels of SnRK1.1 and SnRK1.2 for both i-amiR-SnRK1 constructs (Fig. 1B). SnRK1α and related kinases require activation via the phosphorylation of a conserved threonine (Hawley et al. 1996; Baena-González et al. 2007; Broeckx et al. 2016). Using an antibody directed against phospho-AMPKα (AMP-activated protein kinase alpha phosphorylated at threonine 172) confirmed diminished protein levels of the active fractions of SnRK1.1 (upper signal) and SnRK1.2 (lower signal) upon amiR induction (Fig. 1C; Supplemental Fig. S3A). Moreover, reduced transcript levels of DARK INDUCED1 (DIN1; Fig. 1D), a transcriptional regulation target of SnRK1 (Baena-González et al. 2007), provided evidence that the inducible SnRK1 knockdown also reduced SnRK1 activity and signaling. Upon 3 d estradiol treatment in liquid culture, hypocotyl length was not yet significantly affected (Supplemental Fig. S3B). Thus, our analyses indicated that 3 d amiR induction is well suited to analyze the effects of impaired kinase signaling on a molecular level, also before major changes in growth became apparent.
Having successfully established an inducible knockdown system for SnRK1 and based on previous reports of conditional TOR suppression (Deprost et al. 2007; Caldana et al. 2013; Chen et al. 2018), we decided to generate an i-amiR-TOR mutant. This would not only overcome the issue that also knocking out TOR is lethal, as previously observed for mutants carrying T-DNA insertions in the TOR gene (Menand et al. 2002; Ren et al. 2011) but also enable us to compare the effects of inducible impairment of SnRK1 and TOR signaling under identical experimental settings and using the same knockdown system. Thus, we generated an i-amiR-TOR mutant, which displayed regular development in the absence of estradiol (Supplemental Fig. S4A). Induction of amiR expression at the seedling stage caused strongly reduced TOR transcript level (Fig. 1E). Using an antibody directed against TOR protein did not reveal significant changes in the immunosignal (Supplemental Fig. S4B); however, this may not represent the fraction of active TOR protein as the following observations, besides the reduced TOR transcript levels, also supported diminished TOR activity. First, determining the level of the phosphorylated form of the known TOR target S6K1 (pS6K1, Xiong and Sheen 2012) showed a strong decline upon amiR induction (Fig. 1F) compared to a constant pattern of total S6K1/2 (Supplemental Fig. S4C). Second, upon 3 d of estradiol treatment, the i-amiR-TOR seedlings displayed shortened hypocotyls (Supplemental Fig. S4D), being in line with the phenotype previously reported for a tor-RNAi line (Chen et al. 2018). Interestingly, this impaired hypocotyl elongation was also reminiscent of the phenotype observed upon extended downregulation of SnRK1 (Supplemental Fig. S2C). We also measured DIN1 transcript levels as a potential indicator of the SnRK1 signaling status in the i-amiR-TOR line. Transcript levels of DIN1 and also the other tested marker gene DIN6 were reduced upon estradiol treatment (Fig. 1G). Further studies are needed to validate whether this observation results from suppression of SnRK1, which has been recently shown to be activated by TOR under nonstress conditions (Jamsheer et al. 2022). On the one hand, the levels of these DIN transcripts might also be affected by factors other than SnRK1 activity. On the other hand, measuring the phosphorylated forms of SnRK1.1 and SnRK1.2 upon TOR repression did not reveal a significant difference between mock- and estradiol-treated seedlings (Supplemental Fig. S4E). It also needs to be taken into account that SnRK1 activity may be altered via other means such as a change in its subcellular distribution between the nucleus and other compartments, as previously reported (Blanco et al. 2019; Ramon et al. 2019).
Energy signaling is expected to have a critical function in skoto- and photomorphogenesis, respectively, due to the depletion of resources during etiolated growth and the generation of new assimilates upon the light-induced onset of photosynthesis. Having established our inducible knockdown lines allowed us to study the roles of SnRK1 and TOR in direct comparison in both of these developmental processes. The initial analysis of the i-amiR lines already provided evidence for reduced hypocotyl elongation in darkness, indicating an impaired skotomorphogenic response. For a more detailed characterisation, we grew the mutants for 6 d on estradiol-containing plates under dark conditions. We observed that knocking down SnRK1.1 and SnRK1.2 strongly reduced hypocotyl elongation, whereas the single T-DNA insertion mutant snrk1.1-3 showed wild-type (WT)-like skotomorphogenic development (Fig. 2, A and B). The unaltered phenotype of the snrk1.1-3 is in line with previous studies indicating functional redundancy of the homologs SnRK1.1 and SnRK1.2 (Baena-González et al. 2007; Mair et al. 2015). Interestingly, the inducible TOR knockdown resulted in an even more pronounced impairment of hypocotyl growth. Our observation that SnRK1 and TOR can regulate hypocotyl elongation is in line with previous studies from other groups. In this regard, TOR has been shown to play a role in glucose-promoted hypocotyl elongation of dark-grown seedlings via phosphorylation of ETHYLENE-INSENSITIVE PROTEIN 2 (EIN2; Fu et al. 2021). TOR can also support hypocotyl growth via the accumulation of BRASSINAZOLE-RESISTANT 1 (BZR1), a transcription factor in the brassinosteroid signaling pathway (Zhang et al. 2016). Furthermore, for plants grown in light/dark cycles, an inhibitory role of SnRK1.1 in sucrose-induced hypocotyl elongation was previously reported (Simon et al. 2018).
Both SnRK1 and TOR knockdown impedes hypocotyl elongation in darkness. A, B) Representative photographs A) and quantitation B) of hypocotyl lengths from 6-d-old WT (2 sets grown together with the respective mutants), snrk1.1-3, and i-amiR seedlings. The seedlings were grown in darkness on Mock or estradiol (Est)-containing plates. White scale bar in A) indicates 0.5 cm. The plot in B) depicts interquartile range, maximum as well as minimum of the data set as box and whiskers, respectively. The middle line and the cross represent the median and mean value, respectively; dots show outliers. Asterisks indicate significant difference compared to corresponding Mock control based on 1-way ANOVA with post hoc Tukey test. n is indicated above each condition. C) Relative transcript levels of 3 hypocotyl marker genes determined via RT-qPCR from 6-d-old etiolated seedlings treated with either Mock (white bars) or Est (gray bars) for 3 d. Seedlings were grown in liquid culture. Data are mean values (n = 2 to 6; individual data points as dots) ± Sd, normalized to WT Mock samples. A 1-sample t-test was performed for comparison of Mock and Est samples to WT Mock. P value: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
We hypothesized that the reduced hypocotyl elongation upon kinase knockdown may be the result of a general growth impairment, e.g. due to limited energy supply as a consequence of disturbed sensing or metabolization of sugars. This phenotype could also manifest an impaired skotomorphogenic program due to altered developmental regulation. Obviously, these 2 scenarios are not mutually exclusive as metabolic and developmental regulations are expected to be tightly intertwined. Interestingly, further developmental alterations in the i-amiR-TOR mutant provided more evidence that skotomorphogenesis is directly linked to the activity of at least this sensor kinase. Accordingly, dark-grown i-amiR-TOR seedlings displayed increased root length, hypocotyl bending as an indicator of a reduced negative gravitropism, and partial cotyledon opening (Supplemental Fig. S5, A to C). These additional phenotypical changes of the i-amiR-TOR mutant are also in line with an impairment of skotomorphogenesis that typically involves reduced root growth, negative gravitropism of the hypocotyl, and closed cotyledons (Hangarter 1997; Gommers and Monte 2018). A more pronounced reduction in hypocotyl elongation and the other phenotypical alterations suggested that skotomorphogenesis was even more suppressed upon knockdown of TOR compared to SnRK1, which may be explained by different degrees of knockdown efficiencies or that the 2 pathways can contribute to this process to a different extent or in a different manner. Furthermore, the reduced hypocotyl elongation in the i-amiR mutants could not be rescued by supplementing the medium with sucrose or glucose at various concentrations (Supplemental Fig. S5, D and E), further supporting that this phenotype does not result just from limited energy availability.
Our findings substantiated previous studies (Chen et al. 2018; Simon et al. 2018) linking SnRK1 and TOR function to hypocotyl elongation. On the one hand, we were surprised to find similar phenotypes upon knocking down either of the 2 kinases as they represent signaling hubs with presumably overall antagonistic functions (Li and Sheen 2016; Margalha et al. 2019). On the other hand, more evidence has recently been provided that both components are required for regular development, e.g. SnRK1 was reported to be required for the establishment of seedlings grown in dark/light cycles (Henninger et al. 2022) and to regulate sucrose homeostasis and gene expression in plants grown under favorable conditions (Peixoto et al. 2021), besides its functions under energy deprivation. To determine whether impaired hypocotyl elongation upon inducible kinase knockdown involves specific signatures of gene expression, we determined the transcript level of organ-specific photomorphogenic marker genes that were introduced by Martín and Duque (2021). From this study, we selected 4 genes positively corresponding in their expression with hypocotyl elongation (Martín and Duque 2021). Transcript levels of those genes as well as 3 members from the AUXIN/INDOLE-3-ACETIC ACID (IAA) gene family, which are also known to promote cell and hypocotyl elongation (Luo et al. 2018), were analyzed in etiolated WT and i-amiR lines upon mock and estradiol treatment. In line with the phenotype of strongly reduced hypocotyl elongation, inducible repression of SnRK1 and TOR caused a pronounced and consistent decrease for all 7 hypocotyl marker genes in darkness (Fig. 2C; Supplemental Fig. S6). This specific re-programming of growth regulators could also explain why providing external sugars as additional energy resource is not sufficient to restore hypocotyl elongation in the i-amiR mutants. However, at this point it remains open whether downregulation of the hypocotyl marker genes occurs early upon kinase knockdown and thus rather represents a cause or consequence of reduced hypocotyl elongation. In summary, both SnRK1 and TOR are required for regular hypocotyl elongation during skotomorphogenic seedling growth. Kinase repression causes decreased transcript levels of hypocotyl marker genes, as observed in regular development during the light-induced switch from skoto- to photomorphogenesis (Martín and Duque 2021).
Light-mediated changes in gene expression of etiolated seedlings also involve complex alterations on the level of AS (Shikata et al. 2012, 2014; Hartmann et al. 2016; Xin et al. 2017, 2019). Given that not only light but also sugar availability can trigger these AS changes (Hartmann et al. 2016), we were particularly interested in establishing if the corresponding AS patterns are also responsive to the signaling status of SnRK1 and TOR. We selected for our analysis 4 genes previously shown to exhibit AS changes in etiolated seedlings upon exposure to light or external sucrose, namely SERINE/ARGININE-RICH PROTEIN 30 (SR30), ARGININE/SERINE-RICH SPLICING FACTOR 31 (RS31), RRC1, and PEAPOD 2 (PPD2;Hartmann et al. 2016; Riegler et al. 2021). In our experiment, 6-d-old etiolated seedlings, treated from day 4 onwards for 3 d with mock or estradiol solution, were incubated for an additional 6 h under one of the following conditions before sampling for the AS analysis (experimental scheme provided in Supplemental Fig. S7A): (i) in darkness and a mannitol solution serving as osmotic control, (ii) in darkness and a sucrose solution, and (iii) in light and a mannitol solution. Consistent with previous observations (Hartmann et al. 2016), a 6-h treatment with light or sucrose caused for all genes pronounced AS changes in WT seedlings, compared to control samples kept in darkness and being exposed to mannitol (Fig. 3). The absence or presence of estradiol had no effect on the AS output in the WT seedlings under any condition. In contrast, repressing SnRK1 via estradiol treatment in darkness caused pronounced AS shifts for all 4 investigated genes, similar to the AS changes seen for the WT in response to light and sucrose. In the case of SR30, RS31, and RRC1, the combined treatment of the i-amiR-SnRK1 mutants with estradiol and sucrose caused an even stronger change in the AS ratio compared to mock/sucrose samples. This difference was statistically significant in several cases (Supplemental Data Set 1) and suggested that the external sucrose supply does not always elicit a saturated response. When comparing the AS patterns of PPD2 in the 2 i-amiR-SnRK1 lines (Fig. 3D), no consistent difference between the mock and estradiol treatment in the sucrose-exposed samples was observed. For the light treatment, an additional effect of combined illumination and estradiol exposure could only be seen in the case of RRC1 (Fig. 3C). This observation and different degrees of AS ratio changes in response to sucrose and light exposure can be explained by the existence of alternative signaling thresholds and/or routes.
Knockdown of SnRK1 and TOR triggers similar AS changes as light and sugar. A to D) Models of splicing variants (top) analyzed from SR30A), RS31B), RRC1C), and PPD2D). Lines correspond to introns; dark and light gray boxes depict UTRs and coding exons, respectively. Asterisk marks position of premature termination codon, and arrowheads depict binding sites of primers used for RT-qPCR (dotted line indicates primer binding over an exon–exon border). Scale bar: 500 bp. AS ratios of different light-regulated AS events were quantified via RT-qPCR from liquid-grown 6-d-old seedlings treated with either Mock or estradiol (Est, striped bar) for 3 d (bottom). Six-day-old etiolated seedlings were further treated either with 1.06% mannitol (Man) or 2% sucrose (Suc) and either retained in darkness or exposed to white light for 6 h. Displayed are mean values ± Sd (n = 2 to 4; from 2 independent experiments; individual data points as dots), and data were normalized to corresponding Dark Man Mock control. Statistical significance was determined by 2-tailed Student's t-test against corresponding control and a 1-sample t-test for comparison of Est Dark Man to Mock Dark Man samples (P values: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). An additional statistical analysis was performed to compare the Est response to Suc and Light, respectively. For extended statistical data, please see Supplemental Data Set 1.
Inducible repression of TOR signaling triggered for SR30, RS31, and RRC1 equivalent AS changes as seen upon light and sucrose treatment as well as SnRK1 repression (Fig. 3). Accordingly, inhibition of either of the 2 kinases in etiolated seedlings suppresses etiolation-specific signatures, including the characteristic AS patterns, expression of hypocotyl elongation genes, and hypocotyl elongation. Strikingly, in the case of PPD2, the treatment with estradiol and mannitol in darkness did not trigger an AS response as observed upon sucrose and light exposure, but actually even caused a slight but significant AS shift into the opposite direction (Fig. 3D). Moreover, for RRC1 a considerably weaker change in the AS ratio was detected upon TOR knockdown compared to the effects triggered upon illumination, sucrose exposure, or SnRK1 suppression (Fig. 3C). The distinct response patterns suggest that in these cases an additional input signal may also contribute to the AS outcome in a TOR-dependent manner. We hypothesized that the medium exchange for the 6-h treatment might be responsible for the distinct AS responses of PPD2 and RRC1 upon TOR knockdown, e.g. due to alterations in the availability of other nutrients. To exclude such effects, we also analyzed the AS patterns in 6-d-old etiolated seedlings treated with mock or estradiol solution for 3 d, but then without an additional medium change (Supplemental Fig. S7). Under this condition, both TOR and SnRK1 inhibition resulted in a pronounced decrease of the AS ratio for all 3 analyzed AS events, namely SR30, RRC1, and PPD2. This observation further supports the critical link between the 2 kinases and light-/sugar-responsive AS regulation, but also points to the existence of additional, event-specific input signals. Signal integration at the level of TOR had also been described in the context of glucose-dependent regulation of the circadian clock, which can be blocked by elevated levels of the metabolite nicotinamide in A. thaliana (Zhang et al. 2019). Moreover, the differences in the AS ratios for RRC1 and PPD2 upon TOR knockdown in comparison of the 0-h and 6-h time points also highlights that the AS responses after 3 d of amiR induction are still highly dynamic and responsive to external signals.
We conclude from our observations that plant SnRK1 possesses an expanded action spectrum, which in addition to the previously reported major functions in the regulation of metabolism and transcription via the phosphorylation of enzymes and transcription factors (Sakr et al. 2018; Li and Sheen 2016; Broeckx et al. 2016; Jamsheer et al 2021), respectively, also includes the control of AS decisions. The recent report of SnRK1-mediated phosphorylation of a histone demethylase allowing derepression of starvation-responsive genes in rice (Wang et al. 2021) further highlights that SnRK1 can control gene expression via a diverse range of mechanisms. Our findings similarly support a role of TOR in the regulation of light-responsive AS events during photomorphogenesis. Interestingly, TOR kinase was also reported to be required for proper AS responses in roots of light-grown seedlings after an extended dark period (Riegler et al. 2021). The authors observed that inhibition of TOR signaling using either chemical treatment or an RNAi approach can impair sugar- and light-regulated AS pattern changes in roots and to some extent also in leaves. Considering these and our data, TOR signaling contributes to AS regulation in different tissues and under various physiological conditions.
We next examined whether impairment of either SnRK1 or TOR signaling also similarly affects the developmental response in the transition phase from darkness to light, based on the measurement of cotyledon opening kinetics for etiolated seedlings upon exposure to white light. Knocking down SnRK1 and TOR both strongly impaired cotyledon opening, as evidenced on the level of opening percentages and average cotyledon opening angles measured after 6, 12, 24, 48, and 72 h of white light exposure (Fig. 4, A to C; Supplemental Fig. S8, A to F). In general, cotyledon opening was even more strongly delayed in the i-amiR-TOR line compared to the 2 i-amiR-SnRK1 lines, consistent with our observation that skotomorphogenesis was also more strongly impaired in the i-amiR-TOR mutant. A similar extent of delayed cotyledon opening in response to white light was also seen in a previous study knocking down TOR via an inducible RNAi approach (Chen et al. 2018). As an additional parameter of light-dependent development, we compared the formation of the apical hook, based on measuring the angle between an axis from the tip to the base of the cotyledons and the hypocotyl, before and after 72 h of light exposure (Supplemental Fig. S8, A and G). Knocking down SnRK1 did not affect regular formation of the apical hook in darkness, whereas it was almost completely suppressed in the i-amiR-TOR mutant. The absence of an apical hook is in line with the conclusion that skotomorphogenesis is more strongly impaired in the i-amiR-TOR line than in the i-amiR-SnRK1 mutants, as also supported by the observation of agravitropic growth, increased root length, cotyledon opening (Supplemental Fig. S5), and the more pronounced inhibition of hypocotyl elongation (Fig. 2) upon TOR knockdown in etiolated seedlings. Upon 72 h light exposure, the apical hook was completely resolved in the WT and mock-treated seedlings. As for cotyledon opening, the i-amiR-SnRK1 mutants showed also here a delayed light response. To exclude the possibility that the diminished photomorphogenic responses upon kinase knockdown is the consequence of an irreversible developmental arrest and to confirm that the seedlings are still viable, we transferred seedlings to glucose-containing plates and scored their further development in light (Supplemental Fig. S9). Both the SnRK1 and TOR knockdown seedlings continued growth under these conditions, with the latter seedlings being most impaired, and displayed cotyledon opening and greening as well as further development of the root system, verifying their viability. We conclude from these data that both SnRK1 and TOR are required to elicit a full photomorphogenic response upon illumination of etiolated seedlings, whereas in darkness both kinases support skotomorphogenic development.
SnRK1 and TOR are also required in the photomorphogenic response. A) Representative photographs of the upper region from WT, i-amiR-SnRK1-II, and i-amiR-TOR seedlings that were grown for 4 d on estradiol (Est)-containing plates and then illuminated for the indicated durations with continuous white light (15 μmol m−2 s−1). Scale bar = 0.2 cm. B) Cotyledon opening angles of indicated genotypes grown on Mock or Est-containing plates as described in A). Data show mean ± Se and are derived from 3 replicates with 14 to 33 samples per replicate. For statistical analysis, please see Supplemental Fig. S8F and Supplemental Data Set 1. C) Cotyledon opening angles after 72 h cWL illumination. Details of plant growth and treatments as described in A). Data were analyzed from 3 independent experiments. In all box plots, median is represented by the central line, mean by the cross; box limits show the 25th and 75th percentiles, and whiskers extend to 1.5 × the interquartile range. Statistical significance was determined by 1-way ANOVA with post hoc Tukey test, and asterisks indicate significant difference compared to corresponding Mock control (P value: ****P < 0.0001). n is indicated above each condition. D) Turning point model integrating the current knowledge of SnRK1/TOR signaling and crosstalk with our findings. Hypothetical activity profiles of SnRK1 (red dashed line) and TOR (golden dashed line) during skoto- and photomorphogenesis. Darkness and light intervals indicated by dark and light boxes at bottom. Representative AS switch and light-induced cotyledon opening are schematically depicted.
Finally, we compared seedling growth upon cultivation for 14 d in continuous light and in the absence or presence of sucrose in the medium, providing the plants continuously and at high level with energy sources, other than in the previously tested conditions. In the absence of sucrose, growth of both kinase mutants was strongly diminished compared to the WT (Supplemental Fig. 10A, lower left). This phenotype was most pronounced for the i-amiR-TOR line and could in this case not be rescued by supplementing the medium with sucrose (Supplemental Fig. 10A, lower right). In contrast, in the presence of light and sucrose in the medium, the i-amiR-SnRK1 mutants showed almost WT-like development, indicating that SnRK1 signaling is largely dispensable under these artificially energy-rich conditions. This experiment also showed that the phenotype of the i-amiR-SnRK1 mutants can be rescued by sugar supplementation; however, this is dependent on the growth conditions as hypocotyl elongation in darkness could not be recovered by exogenous sugars (Supplemental Fig. S5, D and E). The effect of energy availability on the phenotype of the i-amiR-SnRK1 mutants was further corroborated by growing seedlings under long day conditions and at different light intensities. The most strongly impaired growth was observed at the lowest light intensity (Supplemental Fig. S10, B and C). Improved growth of the i-amiR-SnRK1 lines was observed at the 2 higher light intensities. However, their development was still more affected, including seedlings with a dwarfed and pale phenotype, than in the previous experiment, when the medium was supplemented with sucrose and continuous light was used, instead of dark/light cycles. In agreement with our findings that SnRK1 is indispensable for normal development of plants cultivated in day/night rhythms, Henninger et al. (2022) recently showed the requirement of SnRK1 for the mobilization of seed storage compounds and proper seedling establishment upon growing A. thaliana in light/dark cycles.
In conclusion, we have demonstrated that both SnRK1 and TOR are required for proper skoto- and photomorphogenic seedling development. Accordingly, both kinases support the developmental program as defined by the environmental condition, i.e. darkness or light in our experimental setup. This feature clearly distinguishes their function in light-dependent seedling development from classical photomorphogenic repressors, such as COP1 and DE-ETIOLATED 1 (DET1), which mainly act by degrading photomorphogenesis-promoting factors (Lau and Deng 2012). Moreover, we have shown that both SnRK1 and TOR function in the upstream signaling of light- and sugar-regulated AS events, identifying an additional mode of action used by these 2 central energy sensor kinases to control gene expression and possibly physiological responses. The observation that knocking down SnRK1 in etiolated seedlings triggers similar AS shifts as observed upon light and sugar exposure (Hartmann et al. 2016, 2018) is in line with the assumption that this signaling pathway is mainly active under energy-deprived conditions. However, studying the role of TOR signaling in parallel revealed that its suppression comparably alters the AS outcome in dark-grown seedlings as observed upon SnRK1 knockdown. This finding was unexpected as TOR is assumed to be mainly active under conditions of sufficient energy supply and to exert opposing functions compared to SnRK1. Intriguingly, this identical response pattern upon SnRK1 and TOR knockdown is not only observed on the level of AS, but could also be shown for expression of DIN and hypocotyl marker genes.
Previous transcriptome studies provided evidence for at least partially antagonistic SnRK1- and TOR-specific patterns of gene expression (Margalha et al. 2019). However, little is known about what the individual contributions and the direct crosstalk of SnRK1 and TOR signaling in specific processes are, as most previous studies were restricted to only 1 of these 2 pathways. In the case of autophagy, the analysis of overexpression and partial knockdown mutants for SnRK1 and TOR revealed their role as activator and repressor, respectively (Soto-Burgos et al. 2018). SnRK1 homologs in animals and yeast have been demonstrated to act upstream of TOR as repressors (Margalha et al. 2019) and evidence for the existence of a similar regulatory relationship in plants was provided (Nukarinen et al. 2016). Accordingly, Jamsheer et al. (2022) identified a negative feedback loop of TOR signaling via SnRK1 activation. The inducible knockdown lines established here provide suitable tools to study under identical experimental conditions specific and overlapping functions of SnRK1 and TOR. It seems likely that not only seedling development but also other physiological processes require concurrent activities of SnRK1 and TOR, which are fine-balanced in response to altered energy availability as opposed to a complete switch between the 2 sensor systems. Such a simple “yin-yang” model with opposing activities of SnRK1 and TOR was also questioned in the context of plant stress responses (Rodriguez et al. 2019). Moreover, Peixoto et al. (2021) proposed oscillation in SnRK1 activity reaching its maximum and minimum at the end of the night and day, respectively. In agreement with our observation and the study from Henninger et al. (2022) that SnRK1 is also needed for regular seedling development upon growth in light/dark cycles, SnRK1's requirement in the regulation of metabolism and gene expression in rosettes of A. thaliana plants grown under favorable conditions was recently reported (Peixoto et al. 2021). Interestingly, we found SnRK1 to be largely dispensable at the seedling stage when plants were grown in continuous light and in the presence of sucrose, i.e. a condition of artificially high availability of energy resources.
Based on our current knowledge of SnRK1 and TOR signaling and the observations from this study, we propose a working model that simultaneous activities of SnRK1 and TOR occur and are required throughout regular plant development (Fig. 4D). Changes in the nutritional status may affect SnRK1 and TOR signaling in an antagonistic manner, e.g. in response to the consumption of storage compounds during skotomorphogenesis. Moreover, day/night cycles could trigger oscillating activity patterns for both kinases, as proposed in the case of SnRK1 (Peixoto et al. 2021) and in line with the finding that TOR signaling is linked to the circadian clock (Zhang et al. 2019). We propose that key for the activation of SnRK1- and TOR-dependent downstream processes, including an altered AS output and altered DIN levels as described in this study, may be the turning points in their activity profiles at major transitions, i.e. a reversal in the signaling status. This can occur during photomorphogenesis when illumination of etiolated seedlings triggers the onset of photosynthesis, causing the transition from growth based on the progressive depletion of storage compounds to phototrophy. As another example, the response of a light-grown plant to severe stress can result in an altered energy balance. Based on our current knowledge, these conditions are expected to alter the activities of SnRK1 and TOR in an opposite manner. The inducible knockdown of SnRK1 and TOR used in this study could mimic such a state transition. Accordingly, the resulting changes on the level of AS and DIN expression could be part of an adaptation process that is usually activated in response to an altered nutritional status to maintain metabolic homeostasis despite the fluctuating and often highly variable environmental conditions. However, further studies are needed to gain deeper insight into the activity profiles of SnRK1 and TOR during plant development and also to reconcile the classical “yin-yang” model with more recent reports including the present study revealing concurrent activities of these 2 energy sensor systems. Of particular interest here is the crosstalk of the 2 systems, which can improve plant fitness by optimizing the tradeoff between growth and stress resilience. Interestingly, SnRK1 was reported to be able to inhibit TOR signaling via interaction with and phosphorylation of RAPTOR1B (Nukarinen et al. 2016). The zinc finger protein FLZ8 can bridge this interaction and is positively regulated by TOR signaling as part of a negative feedback loop under favorable conditions (Jamsheer et al. 2022). Further defining this crosstalk and the underlying mechanisms, in particular in the course of developmental processes linked to altered energy availability as examined in this study, will be critical to obtain a more thorough understanding of how the activities of these 2 sensor systems dynamically change in order to balance plant growth and stress responses. Moreover, we also need to obtain a better understanding of the mechanistic links of SnRK1 and TOR signaling to AS regulation, e.g. by testing whether SnRK1 and TOR can directly phosphorylate splicing regulators to induce AS shifts. Interestingly, a phosphoproteome analysis of light-grown A. thaliana seedlings treated with Torin2 for the inhibition of TOR signaling revealed the differential phosphorylation of multiple RNA-binding proteins including RNA splicing regulators (Scarpin et al. 2020). Follow-up studies making use of our inducible mutants to suppress SnRK1 and TOR signaling during the response of etiolated seedlings to light and sugar signals will help to address the questions of whether splicing regulators are directly phosphorylated by these energy sensor kinases and how this can contribute to the observed AS pattern changes. Recent profiling studies capturing the protein interaction networks of TOR (van Leene et al. 2019) and SnRK1 (van Leene et al. 2022) also provide excellent starting points to further dissect the downstream signaling routes of these 2 central players in energy signaling.
Materials and methods
Plant material, growth conditions, and phenotyping
All mutant and WT seeds used in this study were in A. thaliana Col-0 background. The i-amiR-TOR line additionally contained the WUSCHEL/CLAVATA3 reporter (Pfeiffer et al. 2016). The T-DNA insertion line snrk1.1-3 (GABI_579E09) was previously described (Mair et al. 2015). Seeds were surface sterilized in 3.75% (v/v) NaOCl and 0.01% (v/v) Triton X-100 and plated on ½ strength Murashige and Skoog (MS) medium (M0222.0050, Duchefa, Haarlem, Netherlands), pH 5.7 to 5.8, containing 0.8% (w/v) plant agar (Duchefa). Depending on the experiment, MS media was lacking or containing 1% or 2% (w/v) sucrose, as well as 5 µM β-estradiol (E2758-1G; Sigma-Aldrich, St Louis, MO, USA) or an equivalent concentration of dimethyl sulfoxide (DMSO; mock).
For segregation analyses of constitutive amiR-SnRK1 lines, seeds were plated singly on ½ MS plates containing 1% (w/v) sucrose, 5 mg/L Basta (Bayer, Leverkusen, Germany), and 0.8% (w/v) plant agar. After stratification (2 d at 4 °C), plates were transferred to regular white light (SciBrite LEDs, ∼100 µmol m−2 s−1), and seedlings were grown for 2 wk at 22 °C and 60% relative humidity under long day (16-h light/8-h dark) conditions. Plates without sucrose and Basta for WT growth served as controls. After 2 wk, resistant seedlings were transferred to soil. Transferred plants were grown under a long day regime (16-h light/8-h dark) with a regular white light intensity (∼100 µmol m−2 s−1) at 22 °C. For phenotypic analysis, plants were rated daily regarding their developmental stage and photographs were taken weekly with a Nikon D3200 camera.
Hypocotyl assay
To measure hypocotyl length, seedlings were grown on either 5 µM β-estradiol-containing or mock (with DMSO) solid ½ MS plates without sucrose. Alternatively, seedlings were grown in liquid ½ MS media, and β-estradiol or DMSO (mock) solution was added after 3 d of growth, respectively. After stratification (2 d at 4 °C), germination was induced by a 2 h light treatment (∼100 µmol m−2 s−1) and plates were placed in darkness. Six-day-old etiolated seedlings were transferred to ½ MS plates containing 1.5% agar. Plates were scanned, and hypocotyl length was measured using ImageJ.
Root assay
Seeds were surface sterilized using 70% (v/v) ethanol and 0.1% (v/v) Triton X-100 for 10 min, washed twice with sterile water and kept at 4 °C for 3 d. Seeds were plated singly onto 100 µm nylon meshes containing ½ MS (Duchefa, Haarlem, Netherlands) plates with 0.8% (w/v) phytoagar, and germination was induced by a 12 h white light treatment (∼150 µmol m−2 s−1). Seedlings were then transferred with the mesh to ½ MS plates supplemented with 10 µM β-estradiol or DMSO and further grown vertically for 4 d in darkness. Plates were scanned and root length was measured using ImageJ.
Rescue assay
To analyze whether mutant seedlings can be rescued upon transfer to estradiol-free medium, seedlings were first grown vertically on ½ MS plates containing 1.5% (w/v) agar and either 5 µM β-estradiol or DMSO for 4 d in darkness. Four-day-old etiolated seedlings were then transferred to new ½ MS plates (+ 1.5% (w/v) agar) supplemented with 3% (w/v) glucose. Plates were scanned (4 d Dk timepoint) and placed in continuous white light (30 µmol m−2 s−1) for further 6 d. Plates with 10-d-old seedlings were scanned (+6 d WL timepoint) and phenotypes were analyzed.
To analyze whether mutant seedlings can be rescued by sugar and/or light, seedlings were grown on ½ MS plates (+ 0.8% [w/v] agar) supplemented with either 5 µM β-estradiol or DMSO. Additionally, plates contained either no sugar (without), 2% sucrose or 0.5%, 1%, or 3% (w/v) glucose, respectively. Equimolar concentrations of mannitol were included as osmotic control for the glucose treatments. Seedlings were grown for 6 d in darkness and 14 d under continuous white light conditions (140 µmol m−2 s−1), respectively.
Cotyledon opening assay
Four-day-old etiolated seedlings were grown horizontally on ½ MS plates supplemented with 5 µM β-estradiol or DMSO. The percentage of seedlings displaying cotyledon opening was determined in darkness, by analyzing 3 independent experiments. Seedlings with the seed coat still attached to the cotyledons were counted separately.
Cotyledon opening kinetic assay
Surface sterilized seeds were placed singly on ½ MS plates containing 1.5% (w/v) plant agar, in the presence or absence of β-estradiol. After vernalization for 2 d at 4 °C, germination was induced by white light (∼100 µmol m−2 s−1) for 2 h. Seedlings were grown vertically in darkness for 4 d and then shifted to white light (15 µmol m−2 s−1). If seed coat was still present on cotyledons, it was gently removed at the 0 h time point. Cotyledon opening percentages and angles between the cotyledons were analyzed at 6, 12, 24, 48, and 72 h; apical hook angles were measured at 0 and 72 h using ImageJ. Cotyledons open more than 8° were counted as opened cotyledons. The degree of cotyledon opening was measured with ImageJ.
Plasmid constructions and generation of transgenic plants
Two independent amiR sequences for targeting both SnRK1.1 and SnRK1.2 (Supplemental Fig. S1A) were identified, and corresponding cloning primers (Supplemental Table S2) were designed using the Web MicroRNA Designer (WMD3, http://wmd3.weigelworld.org; Schwab et al. 2006; Ossowski et al. 2008). The 35S promoter-driven constructs for constitutive amiR expression (amiR-SnRK1-I/II) were generated via Gateway cloning (Invitrogen, Carlsbad, CA, USA). Site-directed mutagenesis on pRS300 was performed in 3 single PCR reactions (i to iii) using Herculase II Fusion DNA Polymerase (Agilent Technologies, Santa Clara, CA, USA), as specified in the protocol “Cloning_of_artificial_microRNAs” (http://wmd3.weigelworld.org/downloads/). The following primer combinations were used to amplify fragment i-iii: for amiR-I, (i) SL11/TW026, (ii) TW024/TW025, and (iii) SL12/TW023; for amiR-II, (i) SL11/TW030, (ii) TW028/TW029, and (iii) SL12/TW027. The purified PCR products were mixed with the primer pair SL11/SL12 to perform an overlap PCR (iv) for each construct generating the corresponding amiR precursor sequence flanked by attB sites. Through performing the BP reaction, the DNA insert was introduced into the entry vector pDONR201 (Karimi et al. 2002). Subsequently, the amiR-containing sequences were transferred into the expression vector pB7WG2 via the LR reaction (Karimi et al. 2002).
The constructs for inducible amiR expression were cloned using the GreenGate system (Lampropoulos et al. 2013). The DNA fragments corresponding to the amiR sequences were generated by overlap PCR, as described before for the constitutive constructs, but using TW080/TW081 as outer primers. The amiR precursor sequences were integrated into the entry vector pGGC000 by restriction with BsaI HF (NEB; Ipswich, MA, USA) and subsequent ligation using T4 DNA ligase (Thermo Fisher Scientific; Waltham, MA, USA), resulting in pGGCTW01 and pGGCTW02 with the amiR-SnRK1 sequences, and in pAP039 with the amiR-TOR sequence. The expression cassettes for XVE and the amiR-containing sequences were assembled and ligated to intermediate vectors following the procedure described in Lampropoulos et al. (2013). The module pGGC124, containing the cds of chimeric TF XVE, was generated by amplification of the cds from the plasmid pLB12 (PMID: 16896232; Brand et al. 2006) in 2 PCRs with primers P-878/P-879 and P-880/P-881 to remove the internal BsaI site, followed by cloning the 2 fragments into pGGC. The combination of modules for generating intermediate vectors pGGMTW01, pGGNTW01, pGGNTW02, pAP039, and pAP043 are listed in Supplemental Table S3. Finally, the expression cassettes were combined using the FH and HA adapter sequences. In case of the i-amiR-SnRK1 constructs, the XVE-encoding vector pGGMTW01 was mixed with the destination vector pGGZ003 and either pGGNTW01 or pGGNTW02, resulting in the final constructs pGGZTW01 and pGGZTW02, respectively. For the i-amiR-TOR construct, the amiR sequence-containing vector pAP039 was mixed with the destination vector pGGZ003 and pAP043, resulting in the final construct pAP044.
For generating the respective A. thaliana (Col-0) mutants, the final Gateway and GreenGate constructs were transformed into Agrobacterium tumefaciens strain C58C1 or ASE, respectively, followed by the floral dip method (Clough and Bent 1998). The i-amiR-TOR construct was transformed into the A. thaliana (Col-0) background of the WUSCHEL/CLAVATA3 reporter (Pfeiffer et al. 2016). This line was in parallel described in Stitz et al. 2023.
Chlorophyll content measurements
For chlorophyll content analyses, seeds were plated on ½ MS medium supplemented with either 5 µM β-estradiol or DMSO and stratified for 2 d at 4 °C. Plates were transferred to low (10 µmol m−2 s−1), regular (140 µmol m−2 s−1), or high (300 µmol m−2 s−1) intensities of white light, followed by plant growth under long day conditions (16-h light/8-h dark). After 14 d, representative seedlings were transferred to ½ MS plates containing 1.5% agar and used for scanning. Furthermore, seedlings (22 to 250 mg) were harvested for chlorophyll content measurements, resuspended in 200 µL phosphate buffer (25 mM KH2PO4, 25 mM K2HPO4 pH 7.0, and 2 mM EDTA [pH 8.0]) and chlorophyll was extracted with 800 µL 100% acetone. Mixtures were incubated for 1 h at room temperature under constant shaking. Subsequently, samples were centrifuged for 2 min at 10,000 g and 4 °C and supernatants were used for spectrophotometric analysis at 646, 663 and 750 nm, respectively. Total chlorophyll was calculated using the previously described formula 17.76 × OD646 + 7.34 × OD663/1,000 ×V/FW, where V indicates the volume (mL) and FW the fresh weight (g) (Porra et al. 1989).
Light and sucrose treatments
For light and sucrose treatments, seedlings were grown in liquid ½ MS medium in darkness. Expression of amiRs was induced in 3-d-old etiolated seedlings, by adding 5 µM β-estradiol to the medium. After 6 d, medium was exchanged with ½ MS medium supplemented with 1.06% (w/v) mannitol or 2% (w/v) sucrose. Subsequently, seedlings were either kept in darkness or transferred to white light (∼100 µmol m−2 s−1) and incubated for 6 h (see also experimental scheme provided in Supplemental Fig. S7A).
RNA extraction, reverse transcription quantitative PCR, and PCR product analyses
RNA isolation was performed using the Universal RNA purification kit (Roboklon; EURx, Berlin, Germany). Possible DNA contaminants were eliminated with an on-column DNaseI digest. cDNAs were generated with Superscript II Reverse Transcriptase (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. Reverse transcription quantitative PCR (RT-qPCR) was performed using the MESA GREEN qPCR Mastermix and a CFX384 real-time PCR cycler (Bio-Rad, Hercules, CA, USA). Quantitative PCR (qPCR) program conditions were as followed: 95 °C for 5 min (initial denaturation), 95 °C for 15 s (denaturation) and 60 °C for 45 s (annealing and elongation). PP2AA3 served as reference transcript for normalisation. A detailed protocol for the RT-qPCR analysis was previously described (Stauffer et al. 2010). For some experiments, SR30 splice variants were co-amplified via RT-PCR using the following PCR program: 95 °C for 3 min (initial denaturation), 95 °C for 30 s (denaturation), 59 °C for 30 s (annealing), 72 °C for 30 s (elongation), and 72 °C for 10 min (final elongation). Subsequently, isoform concentrations were determined using the 2100 Bioanalyzer with the DNA1000 kit (Agilent Technologies, Santa Clara, CA, USA). All transcript and splice variant data were normalized to control samples separately for each replicate set. The oligonucleotides used for RT-qPCR and RT-PCR are listed in Supplemental Table S4.
Protein extraction and immunoblot analyses
If not stated otherwise, immunoblot analyses were carried out as previously described (Hartmann et al. 2016). In brief, 0.2 g of 6-d-old etiolated seedlings were mortared to powder in liquid nitrogen and homogenized in 0.2 mL extraction buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.1% [v/v] Tween 20, 0.1% [v/v] β-mercaptoethanol, 1 × protease inhibitor cocktail [Roche], and PhosSTOP [Roche]). Lysates were clarified by centrifugation at 15,000 g and 4 °C for 15 min and proteins were denatured by boiling (5 min at 95 °C) in SDS sample buffer. Samples were separated by SDS-PAGE and transferred to a nitrocellulose membrane using semi-dry transfer (transfer buffer: 25 mM Tris, 142 mM glycine, 20% [v/v] ethanol). Membranes were probed with commercial antibodies: rabbit α-AKIN10 (dilution 1:500; AS10919; Agrisera, Vännäs, Sweden), rabbit α-pAMPK (T172) (dilution: 1:1000; 2531; lot no. 16; Cell Signaling Technology, Danvers, MA, USA), rabbit α-S6K1/2 (dilution 1:750; AS121855; Agrisera, Vännäs, Sweden), rabbit α-S6K1 (phospho T449) (dilution: 1:500; ab207399; abcam, Cambridge, UK), and rabbit α-tubulin (dilution: 1:2000; AS10680; Agrisera, Vännäs, Sweden). Anti-rabbit secondary antibody conjugated with horseradish peroxidase (A6154; Sigma-Aldrich, St Louis, MO, USA) was used at 1:10,000 in 5% skim milk in TBST for 1 h.
For immunoblotting TOR and pAMPK in WT and i-amiR-TOR seedlings, a modified protocol was used. In brief, etiolated WT and i-amiR-TOR seedlings (50 to 200 mg) were grown in liquid ½ MS for 3 d and subjected to 5 µM β-estradiol- or mock treatment for further 3 d. Seedlings were collected in darkness, flash frozen and ground in liquid nitrogen. Proteins were extracted with 1:1 ratio (mg/µL) 95 °C preheated denaturing buffer (100 mM MOPS [pH 7.6], 100 mM NaCl, 5% [w/v] SDS, 10% [v/v] glycerol, 4 mM EDTA, 40 mM β-mercaptoethanol, 2 mM phenylmethanesulfonyl fluoride (PMSF), 1 × protease inhibitor cocktail [Roche]) and boiled at 95 °C for 5 min. Cell debris was removed by 2 centrifugation steps (10 min, 16,000 g, room temperature). Proteins were separated on a 6% (w/v) (for TOR) and 10% (w/v) (for pAMPK) SDS gel and transferred to PVDF using wet transfer (standard settings: 1 h, 110 V at 4 °C). For immunoblotting TOR, proteins were transferred in 1 h, 400 V at 4 °C using a modified transfer buffer (25 mM Tris, 142 mM glycine, 5% methanol). Membranes were blocked for 1 h and then incubated with rabbit α-pAMPK (T172) (dilution: 1:1000; 40H9; 2535; lot no. 21; Cell Signaling Technology, Danvers, MA, USA) or rabbit α-TOR (dilution: 1:2000; AS12 2608, Agrisera, Vännäs, Sweden) at least overnight at 4 °C.
Chemiluminescence was imaged using the Fusion Fx system (Vilber, Collégien, France). The relative band intensities were quantified using the quantification tool of the Evolution-Capt Edge program. Tubulin was detected as a loading control and results were quantified by calculating the volume ratio of pAMPK, pS6K1 or TOR to tubulin.
Statistical analysis
All data measurements were taken from distinct samples. Statistical analyses were performed with GraphPad Prism 8.0.2 (GraphPad Software, Inc., CA, USA). Statistical details of each experiment including biological replicates (n), types of error bars, and used test including whether they were 1- or 2-sided are defined in the results and figure legend sections and in addition in Supplemental Data Set 1. The significance level was set to 0.05 in all cases. The following tests and parameter settings were used according to the description in GraphPad Prism. Standard unpaired t-test was based on the assumption of a Gaussian distribution and the same standard deviation of both groups. Unequal variance t-test (Welch t-test) for 2 groups of data sampled from Gaussian populations, but assuming varying standard deviation. One-sample t-test comparing the mean of a data set derived from an assumably Gaussian population to a provided hypothetical mean of 1, by calculating the t ratio from dividing the difference between the actual and hypothetical means by the standard error of the mean. One-way ANOVA was performed with post hoc Tukey test, assuming Gaussian distribution and equal standard deviation, and using multiple comparisons of every mean with every other mean. Further details and results of all statistical analysis are provided in Supplemental Data Set 1.
Created box plots range from the 25th to 75th percentiles, and whiskers extend to 1.5 × the interquartile range (IQR) or are delimited by the smallest or largest value in case it lies within the 1.5 IQR. Outliers are considered if the value is greater than the sum of the 75th percentile plus 1.5 IQR or smaller than the 25th percentile minus 1.5 IQR.
Accession numbers
Sequence data from this article can be found in the EMBL/GenBank data libraries under accession numbers: AT3G01090 (SnRK1.1), AT3G29160 (SnRK1.2), AT1G50030 (TOR), At4G35770 (DIN1), AT3G47340 (DIN6), AT3G61860 (RS31), AT1G09140 (SR30), AT5G25060 (RRC1), AT4G14720 (PPD2), and AT1G13320 (PP2AA3).
Acknowledgments
We thank Claudia König and Moritz Denecke for technical assistance.
Author contributions
J.S., T.W.-K., and A.W. designed experiments. J.S. and T.W.-K. performed experiments and analyzed data. T.W.-K. initiated the study by generating and characterizing the constitutive amiR-SnRK1 lines (with the help of D.M.O.) as well as the inducible amiR-SnRK1 lines. The i-amiR-TOR line was generated and initially characterized by A.P. and D.J. under supervision of J.L. J.S. was mainly performing experiments involving the i-amiR-TOR line except the hypocotyl assay (6 d dark) that was done by T.W.-K. J.S. supported T.W.-K. in splice pattern analyses (0 h time point), performed all splice pattern studies in response to light and sugar, and analyzed hypocotyl marker genes. Furthermore, J.S. performed cotyledon opening experiments, rescue and chlorophyll assays. K.E. performed phenotyping experiments of constitutive amiR-SnRK1 lines. The manuscript was mainly written by J.S., T.W.-K., and A.W. All authors contributed to data interpretation and approved the manuscript. A.W. conceived the project and supervised the research.
Supplemental data
The following materials are available in the online version of this article.
Supplemental Figure S1. Construct design and phenotypes of constitutive snrk1 knockdown lines.
Supplemental Figure S2. Construct design and phenotypes of i-amiR-SnRK1 lines.
Supplemental Figure S3. Analysis of i-amiR-SnRK1 lines in liquid culture assay.
Supplemental Figure S4. Characterization of i-amiR-TOR line.
Supplemental Figure S5. Suppressed skotomorphogenesis does not correlate with a general growth arrest due to energy depletion.
Supplemental Figure S6. Additional hypocotyl marker genes.
Supplemental Figure S7. Experimental cartoon and AS patterns in amiR lines at 0 h time point.
Supplemental Figure S8. Light-induced cotyledon opening in snrk1 and tor mutants.
Supplemental Figure S9. Inducible knockdown seedlings are viable and growth can be rescued upon transfer to glucose-containing medium.
Supplemental Figure S10. The amiR-SnRK1 phenotype can be partially rescued in the presence of light and sucrose.
Supplemental Table S1. Segregation and mortality rates at post-seedling stage of amiR-SnRK1 mutants.
Supplemental Table S2. Primers for cloning (i-)amiR-SnRK1 and i-amiR-TOR constructs.
Supplemental Table S3. GreenGate cloning modules and destination constructs.
Supplemental Table S4. Sequences of primers for qPCR and co-amplification PCR.
Supplemental Data Set 1. Statistical analyses.
Funding
This research project was supported by the German Research Foundation (Deutsche Forschungsgemeinschaft—DFG project no. 232631280: SFB1101/C03 to A.W. and SFB1101/B07 to J.L.), as well as the European Research Council (ERC grant 282139 “StemCellAdapt” to J.L.).
References
Author notes
The author(s) responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://academic.oup.com/plcell/) is: Andreas Wachter ([email protected])
Conflict of interest statement. None declared.
