Sugar signaling modulates SHOOT MERISTEMLESS expression and meristem function in Arabidopsis

Significance The shoot apical meristem (SAM) generates all the aboveground plant organs and is hence crucial for plant adaptation to the environment. However, little is known of how the SAM perceives environmental information and how this impacts meristem activity and plant growth. Here, we show that sugars promote the accumulation of SHOOT MERISTEMLESS (STM), a transcription factor necessary for stem cell identity and proliferation. This is counteracted by SUCROSE NON-FERMENTING1-RELATED KINASE 1 (SnRK1), which is activated when sugar levels decline, and interacts with STM. On the other hand, silencing SnRK1 in the SAM showed that it is needed for meristem integrity. Overall, our data support a dual function for SnRK1 in plant growth and a need to finely balance its activity.

Because of their sessile lifestyle, plants continuously adjust their development to changes in the environment, and this is reflected in the dynamic nature of the SAM.In addition to its maintenance by a network of TFs and hormonal signals, the SAM also responds to environmental cues that influence the relative size of its subdo mains and the type and number of organs it produces.One of the external factors that affect meristem activity is light, which can exert a direct effect through photoreceptor-mediated signaling and an indirect effect by driving photosynthesis and sugar production ( 20 -22 ).Both light and metabolic signals activate the TARGET OF RAPAMYCIN (TOR) protein kinase, which in turn promotes cell proliferation in the SAM via an increase in the expression of S-phase genes ( 23 , 24 ).In addition, TOR induces WUS expres sion, partially through an effect on CK degradation ( 22 , 25 ).
TOR activity is often antagonized by SUCROSE NON-FERMENTING1-RELATED KINASE 1 (SnRK1) which, like TOR, translates environmental information into metabolic and developmental adaptations ( 26 -28 ).SnRK1 is a heterotrimeric pro tein kinase complex, composed of an α-catalytic subunit and two regulatory βand γ-subunits.In Arabidopsis , the α-subunit is present in two major isoforms, SnRK1α1 and SnRK1α2 (also known as KIN10 and KIN11) ( 28 ).The SnRK1 complex is activated under low carbon conditions to promote energy-saving and nutrient remo bilization strategies, while TOR is activated in response to nutrient abundance to promote cell proliferation and growth ( 26 -28 ).
Despite the importance of STM for establishing and maintaining SAM function, little is known of its potential regulation by environ mental signals.Here, we make use of plants expressing transcriptional and translational STM reporters to investigate this question.We show that STM protein accumulation does not respond to CK but that it is clearly induced by photosynthesis-derived sugars.We also show that suboptimal light conditions activate the SnRK1 kinase in the SAM and that SnRK1 interacts with STM and represses its function.Finally, we show that, despite being generally considered a growth repressor, SnRK1 is necessary under favorable conditions to maintain meristem organization and integrity.

Results
Light Promotes STM Protein Accumulation.Light is essential for proper plant development and physiology.To investigate a potential regulatory role of light on STM levels, we made use of an Arabidopsis (Col-0) reporter line in which a fluorescently tagged form of the STM protein (STM-VENUS) is expressed under the control of the STM promoter [pSTM::STM-VENUS (29,30)].We measured STM-VENUS levels in inflorescence meristems from 5-wk-old plants grown under, or transiently treated with different light conditions.In one set of experiments, we compared STM-VENUS levels between plants grown under two different irradiances [60 vs. 170 μmol m −2 s −1 , referred to as low light (LL) and high light (HL), respectively].Irradiance had a strong impact on STM accumulation, with the mean STM-VENUS levels of plants grown under LL being 76% of those grown under HL (Fig. 1 A and B).In a second set of experiments, we compared STM-VENUS levels between HL-grown plants transferred to darkness for up to 72 h and their corresponding controls maintained under HL conditions.Incubation under darkness had a very severe impact on STM accumulation, with STM-VENUS levels decreasing to 39% over the course of the 72 h dark treatment (Fig. 1 A and C).
To assess whether the impact of light on STM levels was a general effect on protein abundance in meristems, we extracted total proteins from SAMs of plants constantly grown under HL, LL, or treated with 48 h of darkness and compared STM levels to those of the reference protein TUBULIN (TUB) by immunob lotting ( Fig. 1D ).These analyses confirmed the microscopy results regarding STM-VENUS accumulation, showing that, in LL and dark-treated plants, STM levels were 71% and 23%, respectively, of the STM levels in HL.The immunoblots revealed no impact of the light conditions on TUB accumulation, indicating that the lower STM levels were not caused by a general decrease in protein accumulation.Finally, to assess whether low STM accumulation could be due to reduced STM transcript abundance, we dissected SAMs of plants kept under HL conditions or subjected to 48 h darkness and analyzed STM transcript levels by qPCR.STM levels were not significantly affected by the dark treatment ( Fig. 1E ), showing that the differences in protein accumulation are not due to changes in STM transcription or transcript stability.On the other hand, the levels of AINTEGUMENTA-LIKE 7 (AIL7 ) and HOMEOBOX PROTEIN 25 (HB25 ), two known gene targets of STM ( 10 ), were reduced upon dark treatment ( Fig. 1E ).This is also consistent with the lower STM-VENUS abundance and indi cates decreased STM activity in the SAM in these conditions.

The Response of STM to Light is CK-Independent and Involves
Sugars.Several lines of evidence suggest that STM expression could be, like WUS, directly regulated by CK (12,31).To investigate whether CK could also regulate STM at the protein level and hence be involved in the response of STM to light, we first tested whether light influenced CK signaling in inflorescence meristems.To this end, we used plants expressing the synthetic CK reporter pTCSn::GFP (32) in similar experiments as described for the STM-VENUS reporter line.In plants grown in LL or subjected to a 48 h dark treatment, pTCSn::GFP levels were 74% (SI Appendix, Fig. S1 A and B) and 46% (SI Appendix, Fig. S1 A and C) of those in HL plants, respectively.These observations show that CK signaling in inflorescence meristems is, like in vegetative meristems (22), affected by light.We next examined whether CK could impact STM levels in inflorescence meristems.For this, we excised SAMs of HL-grown STM-VENUS plants and maintained them under HL in vitro (14) for different periods of time in the absence or presence of 500 nM 6-benzylaminopurine (BAP), a synthetic CK.Dissection of the SAMs led to a strong reduction of the STM-VENUS (SI Appendix, Fig. S1D) and the pTCSn::GFP (SI Appendix, Fig. S1E) reporter signals, as previously described for the pTCSn::GFP and pWUS::GFP reporters (14).However, in contrast to pTCSn::GFP [SI Appendix, Fig. S1E; (14)], CK could not sustain STM-VENUS levels (SI Appendix, Fig. S1D), indicating that the effect of light on STM-VENUS is likely CK-independent.
Light plays direct signaling functions through various photore ceptors but also signals indirectly through sugars produced by pho tosynthesis.We therefore wondered whether the effect of light on STM was direct or mediated by sugars.To investigate this, we first measured the levels of sucrose, glucose, and fructose in the rosettes (SI Appendix, Fig. S2A ) and SAMs ( Fig. 2A ) of HL-and dark-treated plants.We also measured the levels of Tre6P, a regulatory sugarphosphate that reflects the sucrose status, and that is crucial for sucrose homeostasis, growth promotion, and developmental pro gression ( 33 ).In the light, the levels of sucrose, glucose, and Tre6P were, respectively, 2.1-, 2.2-, and 7.9-fold higher in the SAM than in the rosette.Fructose accumulated to comparable levels in the two organs ( Fig. 2A and SI Appendix, Fig. S2A ).Incubation in the dark led to a marked depletion of sucrose and fructose both in rosettes (15% and 11% of the levels in the light, respectively) and SAMs (8% and 4% of the levels in the light, respectively), with a much milder reduction being observed for glucose, the most abundant sugar in the SAM (44% and 35% of the levels in the light in rosettes and SAMs, respectively).Tre6P levels were also much lower in dark-treated plants (11% and 3% of the levels in the light in rosettes and SAMs, respectively), reflecting the drop in sucrose levels.To further distinguish between a light and a sugar effect, we excised inflorescences at around 3 cm from the apex and placed them for 48 h in liquid medium.Similarly to what was observed in dissected SAMs (SI Appendix, Fig. S1D ), STM-VENUS signal decreased markedly in cut inflorescences as compared to the uncut controls ( Fig. 2B ).Furthermore, light alone was not sufficient to sustain STM-VENUS expression, as the signal was comparable in cut inflorescences incubated in the light and in the dark (47% and 43% of the levels in the uncut control, respectively).These results, obtained with a double reporter line (pSTM::STM-VENUS/pSTM:: TFP-N7 ; Fig. 2B ), were similar to those obtained for plants express ing STM-VENUS alone (SI Appendix, Fig. S2B ).
To test whether the decrease in STM levels was due to sugar deprivation, we first incubated (48 h in darkness) excised inflores cences of the double marker line in medium supplemented with increasing concentrations of sucrose.Sorbitol, which is not a readily metabolized carbon source, was used as an osmotic control.Sucrose was able to sustain STM-VENUS accumulation, and its effect was largely dose-dependent, leading to STM-VENUS levels close to those of uncut inflorescences when supplied at a 5% concentration (72% of the uncut control values as compared to 18% in the cor responding 2.5% sorbitol control; Fig. 2C ).STM-VENUS levels did not increase in response to sorbitol, indicating that the effects of sucrose were not osmotic.Similar results were obtained for the single STM-VENUS reporter line (SI Appendix, Fig. S2C ).To test whether the observed effects were transcriptional, we monitored the activity of the pSTM::TFP-N7 transcriptional reporter.Quantifi cation of the pSTM::TFP-N7 signal revealed no significant repres sion of STM promoter activity upon inflorescence excision and incubation in darkness (SI Appendix, Fig. S2D ), consistent with the results obtained by qRT-PCR in intact plants ( Fig. 1E ).In addition, incubation in sucrose or sorbitol-containing media had minor effects on TFP levels (SI Appendix, Fig. S2E ) as compared to STM-VENUS ( Fig. 2C ), with the most severe condition (2.5% sorbitol) leading to 65% of the signal in the uncut control as com pared to the 18% of the equivalent STM-VENUS samples.This indicates that the effect of sugar deprivation on STM levels does not rely on transcriptional regulation of STM .
The SnRK1 Sugar Sensor Is Expressed in the SAM and Influences STM Levels.One major component of sugar signaling is the SnRK1 protein kinase, that is activated under conditions of low carbon availability and is conversely repressed by sugars (34).Given its well-established role as a sugar sensor and the increasing number of studies implicating SnRK1 in developmental processes (26,28), we next investigated whether SnRK1 could be involved in the regulation of SAM function through STM.To this end, we used a line expressing SnRK1α1-GFP under the control of the SnRK1α1 promoter and other gene regulatory regions (35).A clear SnRK1α1-GFP signal was detected in the SAM, showing a stronger intensity in the peripheral regions, and developing organs (Fig. 3A).To further confirm this expression and to assess whether SnRK1α1 might be enriched in the SAM relative to other organs of the plant, we extracted total proteins from rosettes and shoot apices of 6-to 7-wk-old plants and compared the relative levels of SnRK1α1 by immunoblotting (Fig. 3B).For the same amount of total protein, shoot apices contained higher amounts of SnRK1α1 suggesting that SnRK1 is relatively more abundant in the SAM than in rosette leaves.
To test whether SnRK1 is functional in the meristem, we used SAMs dissected from HL-or dark-treated plants (48 h) to measure the activity of the SnRK1 signaling pathway using the expression of downstream target genes ( 34 ) as readout of in vivo SnRK1 activ ity ( Fig. 3C ).As expected, SnRK1-regulated starvation genes were barely expressed under control conditions.However, after 48 h of darkness, a marked upregulation of these genes was observed ( Fig. 3C ), indicating an activation of SnRK1 signaling in the SAM.The induction of SnRK1-regulated genes in darkness was accom panied by a reduction in total SnRK1α1 levels ( Fig. 3D ), consistent with a tight coupling between SnRK1 activity and degradation ( 35 , 36 ), and by an increase in the relative phosphorylation of the SnRK1α1 (T-loop) that is essential for SnRK1 activity ( 34 ).
To investigate whether SnRK1 is involved in STM regulation, we used a line overexpressing SnRK1α1 that displays no obvious growth or developmental defects [35S::SnRK1α1 , hereafter ref erred to as SnRK1α1-OE ; ( 37 )].We introgressed the pSTM::STM-VENUS reporter construct into this line and monitored STM-VENUS levels in different light conditions.When plants were grown under HL, the levels of STM-VENUS in SnRK1α1-OE were 60% of those in control plants ( Fig. 3E ), a decrease that could not be explained by differences in STM transcript levels (SI Appendix, Fig. S3 ).However, the differences between the two genotypes became smaller when plants were grown in LL (STM-VENUS levels in SnRK1α1-OE were 77% of those in HL plants; Fig. 3E ) and negligible when subjected to a 48 h dark treatment (STM-VENUS levels in SnRK1α1-OE were 97% of those in con trol plants; Fig. 3F ).STM-VENUS levels thus appeared to be constitutively low and largely insensitive to the light conditions in SnRK1α1-OE plants.This contrasted with control plants which, in response to restrictive light conditions, reduced STM-VENUS accumulation to levels equivalent to those of SnRK1α1-OE.Lower STM-VENUS levels in SnRK1α1-OE in HL could not be explained by lower sugar accumulation, as these plants had a higher content of sucrose, glucose, and fructose both in the SAM ( Fig. 3G ) and rosettes (SI Appendix, Fig. S4 ), although the differences were not always statistically significant due to large variation in the SnRK1α1-OE samples.The levels of Tre6P, known to inhibit SnRK1 activity ( 38 -40 ), were also markedly higher in SnRK1α1-OE SAMs (5.6-fold) and rosettes (fivefold), consistent with previous observations in SnRK1α1-OE rosettes ( 41 ).During the dark treat ment, however, all sugars were largely depleted, reaching similarly low levels in control and mutant samples.Altogether these results suggest that SnRK1 is active in the SAM and that it contributes to the adjustment of STM protein levels, inhibiting STM accumulation when sugar levels decline.To fur ther investigate the involvement of SnRK1 on STM regulation, we first used yeast-two-hybrid (Y2H) assays to test whether SnRK1α1 can interact directly with STM ( Fig. 3H ).We observed that yeast coexpressing SnRK1α1 with STM were able to grow in selective medium but this was not the case when SnRK1α1 or STM were expressed individually with the corresponding empty vector controls, suggesting that these two proteins can interact.To determine whether the SnRK1α1-STM interaction can also occur in planta, we next performed coimmunoprecipitation (co-IP) experiments using Col-0 mesophyll cell protoplasts expressing SnRK1α1-HA with STM-GFP or with GFP as a neg ative control.Immunoprecipitation with an anti-GFP antibody and subsequent immunoblot analyses revealed that SnRK1α1 interacts with STM-GFP ( Fig. 3I ), but not with GFP, indicating that the interactions revealed by Y2H may also occur in planta .To explore the functional implications of this interaction, we developed a reporter of STM activity by fusing LUCIFERASE (LUC ) to the promoter of the STM target gene CUC1 ( 42 ) and used the reporter in transient protoplast-based assays ( 43 ).STM expression triggered a 17-fold induction of the CUC1::LUC reporter and this induction was suppressed by 60% when STM was coexpressed with SnRK1α1, which reduced STM accumula tion ( Fig. 3J ).This repressive effect was no longer visible when we used a catalytically inactive SnRK1α1 variant [SnRK1α1 K48M , ( 34 )], indicating that the impact of SnRK1 on STM depends on the protein-phosphorylating activity of SnRK1.Taken together, our results support that SnRK1α1 regulates STM protein accu mulation and activity.
Silencing SnRK1α in the SAM Disrupts Meristem Function.To investigate further the possibility that SnRK1 acts locally in the meristem, we designed artificial microRNAs (amiRNAs) targeting both SnRK1α1 and SnRK1α2 in two different regions of the transcripts (amiRα-1 and amiRα-2) and expressed these amiRNAs under the 5.7 kb promoter of STM in STM-VENUS plants (Fig. 4A).Immunoblot analyses confirmed a decrease in the activated form (phosphorylated in the T-loop) of SnRK1α in all lines, but this was accompanied by a decrease in total SnRK1α1 levels only in lines expressing amiRα-2 (Fig. 4B).
To our surprise, SnRK1α depletion resulted in decreased STM-VENUS accumulation in optimal growth conditions ( Fig. 4 C -E ) with the signal reaching 14% (amiRα-2#1 and amiRα-2#2 lines) and 37% (amiRα-1#1 ) of the control plant levels.This was accompanied by a strong reduction in STM transcript levels (SI Appendix, Fig. S5 ), contrasting with the effect of SnRK1α1 over expression ( Fig. 3 E and F and SI Appendix, Fig. S3 ).The extent of SnRK1α depletion correlated with defects in SAM development, including altered phyllotaxy, reduced bulging, and the appearance of bract-like structures in some floral meristems, as well as fusions between adjacent floral meristems ( Fig. 4F ).Organ fusions were also visible later in development and affected cauline and rosette leaves, petals, siliques, and stems ( Fig. 5 D and F ).All amiRα lines displayed defects in internode elongation, with an increased frequency of aber rantly long and aberrantly short internodes ( Fig. 5 A , B , G , and H ). Defects in internode elongation resulted in clusters of siliques ( Fig. 5  A , B , and E ) and what appeared to be aerial rosettes on the main inflorescence ( Fig. 5 C , D , and F , Table 1 , and SI Appendix, Fig. S6A ).These phenotypes were more severe in the amiRα-2 plants ( Fig. 4 C -E ), which also exhibited reduced apical dominance with one or two axillary meristems often becoming activated well before flower ing (39% and 28% of the amiRα-2#1, and amiRα-2#2 plants, respectively; n = 18; SI Appendix, Fig. S6 B and C ).A less frequent termination of the main meristem was also observed, after which growth resumed from an axillary meristem (17% of amiRα-2#1 plants; n = 18; SI Appendix, Fig. S6C ).Plants expressing pST-M::amiRα were also compared to the double reporter line as control (pSTM::STM-VENUS/pSTM:: TFP-N7 ) supporting that the observed phenotypes were not caused by the introgression of an additional STM promoter in the genome of the STM-VENUS line ( Fig. 4 C , D , and F and SI Appendix, Fig. S6D ).
To further examine the consequences of depleting SnRK1 activ ity in the SAM, we expressed amiRα-1 and amiRα-2 under the control of the RIBOSOMAL PROTEIN 5A (RPS5A ) or FD pro moters, both highly active in inflorescence meristems ( 44 ).Three independent lines of each construct were analyzed at the T2 plant stage, revealing phenotypes largely similar to those of the pST-M::amiRα lines.All the twelve analyzed lines showed a higher fre quency of defective internode elongation than the control, with ten showing also a higher frequency of pedicel fusion (SI Appendix, Fig. S7 A -E and I ).This was accompanied by a higher incidence of more than one silique per node (SI Appendix, Fig. S7J ).For some of the pFD::amiRα-1 plants, we also observed severe stem and flower organ fasciation and even meristem abortion (SI Appendix, Fig. S7 F -H ).
Collectively, these results indicate that SnRK1 plays critical functions in meristem organization and function.

Discussion
The capacity to generate organs throughout development is crucial for plant adaptation to the environment.However, how the SAM perceives environmental information and how this is translated into changes in meristem activity are poorly understood.
Here, we show that light promotes STM accumulation through sugars.First, a clear correlation between STM-VENUS and SAM sugar levels was observed across different light conditions.STM-VENUS levels were lower in LL-grown or dark-treated plants than in plants grown and maintained under HL ( Fig. 1 A -C ).A similar pattern was observed for sugar accumulation in the inflorescence mer istems ( Fig. 2A ), consistent with a previous report on the impact of limiting photosynthetic rates (and thereby sugar supply to the sinks) on the growth and development of reproductive organs and meristem function ( 45 ).Second, STM-VENUS levels declined rapidly when inflorescences were excised from rosettes and this decline was similar in inflorescences maintained in the light or transferred to dark, show ing that light is not sufficient to sustain STM levels in this system ( Fig. 2B ).The reason for this could be that light is sensed in leaves, generating a systemic light-related signal that is disrupted upon excision of the inflorescence.An alternative explanation is that the signal regulating STM levels is not light itself, but rather photosynthesisderived sugars.The fact that the decline in STM-VENUS levels trig gered by inflorescence excision could be largely suppressed by supple menting sucrose in darkness argues that sucrose is sufficient to sustain STM-VENUS levels and that the effect of light observed in intact plants is indirect via photosynthesis and sugar production.The impact of sucrose on STM is in line with the reported effects of nutrients on WUS expression and on meristem function.Sugars contribute to meristem activation by inducing WUS in young seedlings ( 22 ) and nitrogen promotes WUS expression and meristem growth in the inflo rescence via systemic CK signaling ( 14 ).However, in contrast to WUS , for which transcriptional regulation plays a major role ( 14 , 22 ), we did not detect significant changes in STM transcript levels under our different growth conditions or treatments ( Fig. 1E and SI Appendix, Fig. S3 ), indicating that STM is regulated at the protein level.Despite reports linking CK signaling to STM expression ( 12 , 31 ), STM-VENUS levels did not increase in excised meristems treated with CK (SI Appendix, Fig. S1D ).Even though we did not measure STM tran script accumulation under these conditions, this could mean that CK signals may influence STM more indirectly, e.g., by affecting WUS expression and stem cell number ( 3 ).The rescue of STM-VENUS levels by sucrose in excised inflores cences suggests that sucrose is sensed locally in the meristem.This is in accordance with the enrichment and activity of the SnRK1 sugar sensor in the SAM ( Fig. 3 A -D ).Ubiquitous SnRK1α1 overexpres sion caused a reduction in STM-VENUS levels under HL conditions ( Fig. 3E ) despite the high accumulation of soluble sugars and Tre6P in the rosettes and SAMs of the SnRK1α-OE plants ( Fig. 3G ).In addition, SnRK1α1 interacted physically with STM in yeast cells and mesophyll cell protoplasts ( Fig. 3 H and I ) and repressed STM accu mulation and activity in the latter ( Fig. 3J ), suggesting that SnRK1 modulates STM function locally in the meristem.
Depletion of SnRK1 activity via amiRs further demonstrated that SnRK1 acts locally in the SAM and that its functions extend beyond STM protein regulation ( Figs. 4 and 5 and SI Appendix, Figs.S5-S7 ).Reduced SnRK1 activity caused a wide range of developmental defects which contrasts with the absence of morphological anomalies in the SnRK1α-OE line in our and previous studies ( 26 , 28 ).
The finding that sucrose promotes STM-VENUS protein accu mulation together with the fact that sugars repress SnRK1 activity, may at first sight appear to conflict with abnormal meristem function and the decline in STM transcript and protein accumulation observed upon SnRK1α silencing.However, despite being generally considered a growth repressor, SnRK1 is also required for cell cycle progression ( 46 ) and for normal growth and development ( 47 ).Indeed, transient SnRK1α1/α2 down-regulation via virus-induced gene silencing leads to full growth arrest of plants ( 34 ) and double snrk1α1 snrk1α2 null mutants could thus far not be recovered, suggesting that complete loss of SnRK1α is embryo lethal ( 48 ).A similar duality is observed for the AMP-activated protein kinase (AMPK), the homologue of SnRK1 in animals.Despite serving as a brake for cell proliferation through downregulation of TOR activity ( 49 ), AMPK is also essential for normal growth and development.For example, complete loss of the AMPKß1 subunit leads to cell cycle defects in neural stem and progenitor cells, causing profound abnormalities in brain develop ment in mice ( 50 ).Along the same lines, hematopoietic stem cell function in mammals is disrupted both upon inactivation and over activation of TOR signaling, indicating that a fine balance of this central regulator is required for coordinating cell proliferation, differ entiation, and regeneration ( 51 ).
The effects of light, sucrose, and SnRK1α1 overexpression on STM indicate that the underlying mechanisms do not rely on changes in STM transcript abundance ( Fig. 1 -3 and SI Appendix, Fig. S3 ).Furthermore, SnRK1α1 and STM proteins interact phys ically and functionally, and SnRK1α1 kinase activity is needed for the repression of STM function ( Fig. 3 H -J ).This suggests that SnRK1 impacts STM protein function, either through phosphoryl ation of STM or of an STM interactor required for its activity.The consequences of SnRK1α silencing, on the other hand, reveal a more complex scenario, causing severe developmental abnormal ities that are accompanied by a reduction in the STM transcript and a more pronounced reduction in STM protein accumulation ( Figs. 4 and 5 and SI Appendix, Figs.S5-S7 ).In this case, the impact of SnRK1α on STM is likely to be largely indirect, involving processes (e.g., hormone signaling, the cell cycle, others) that remain to be determined.
Altogether, our work demonstrates that sucrose promotes STM accumulation and that this is counteracted by the SnRK1 sugar sensor, likely to adjust SAM activity to the environment ( Fig. 6 ).Nevertheless, SnRK1 is also essential for the maintenance of  meristem functions in optimal growth conditions, adding to the evidence that SnRK1 performs a dual function in the regulation of growth and that its activity needs to be finely balanced.

Materials and Methods
A list of all primers, plant lines, and antibodies used in this study is provided in SI Appendix.Details of plant growth and treatment conditions, protein extraction and quantification, immunoblotting, RNA extraction, cDNA synthesis, qRT-PCR, sugar measurements, yeast-two-hybrid assays, and protoplast assays are described in SI Appendix.For the time-lapse experiments examining the effect of exogenous CK, meristems were dissected from HL-grown plants and placed in a box of ACM with 1% (w/v) sucrose and 500 nM BAP or the equivalent volume of the BAP solvent (DMSO) as control.Meristems were thereafter returned to the constant HL cabinet for the indicated times and covered with water for imaging.
For the experiments examining the effect of exogenous sugar, inflorescences were dissected at about 3 cm from the apex from HL-grown plants at the beginning of the flowering stage and placed in a 2 mL Eppendorf tube containing 2 mL of liquid ACM (no sucrose) covered with parafilm pierced with a needle so that the inflorescence could be held in air while the base of the stem was submerged in the solution, supplemented with the indicated concentrations of sucrose or sorbitol as osmotic control.Sorbitol and sucrose were used at near isosmotic concentrations, with 1% (w/v) sorbitol and 2% (w/v) sucrose corresponding to 54 mM and 58 mM, respectively.Inflorescences were thereafter kept in darkness inside the growth cabinet for the indicated times.Meristems were then dissected from the excised inflorescences, transferred to a box containing the same sucrose-or sorbitol-supplemented solid [1.6% (w/v) agarose)] medium and covered with water for imaging.
Data, Materials, and Software Availability.All study data are included in the article and/or SI Appendix.

Fig. 1 .
Fig. 1.Effect of light on STM expression.(A-C), STM-VENUS expression in SAMs of pSTM::STM-VENUS plants grown under HL (170 μmol m −2 s −1 ) or LL (60 μmol m −2 s −1 ) conditions or transferred from HL to darkness (D) or kept under HL for the indicated times.(A) Representative STM-VENUS images of SAMs from HL and LL-grown plants and of plants transferred to D for 48 h.(Scale bar, 50 µm.)(B and C) Quantification of STM-VENUS signal.(B) Plots show SAM measurements of plants grown as three independent batches normalized by the mean of the HL condition of each batch (HL, n = 44; LL, n = 45).Student's t test (P-value shown).(C) Plots show SAM measurements of plants grown as two to three independent batches normalized by the mean of the HL condition of each batch (0 h, n = 18; 24 h L, n = 19; 24 h D, n = 18; 48 h L, n = 19; 48 h D, n = 18; 72 h L, n = 9; 72 h D, n = 12).The 0 h sample serves as control for both L and D treatments.Different letters indicate statistically significant differences (Kruskal-Wallis with Dunn's test; P < 0.05).(D) Immunoblot analyses of STM and TUBULIN (TUB) protein levels in SAMs of pSTM::STM-VENUS plants grown under HL or LL conditions or grown in HL and transferred to D for 48 h.Ponceau staining serves as loading control.Numbers refer to mean STM-VENUS amounts in LL and D as compared to HL (n = 2; each a pool of five SAMs; in parentheses, SEM).(E) RT-qPCR analyses of STM and STM target genes AIL7 and HB25 in SAMs of pSTM::STM-VENUS plants grown in HL and transferred to D or kept in HL for 48 h.Graphs show the average of three independent samples, each consisting of a pool of five SAMs.Paired ratio t test (P-values shown).

Fig. 2 .Fig. 3 .
Fig. 2. Effect of sugars on STM levels.(A) Effect of light on the levels of soluble sugars in SAMs of pSTM::STM-VENUS plants grown in HL and transferred to darkness (D) or kept in HL for 48 h.Suc, sucrose; Tre6P, trehalose 6-phosphate; Glc, glucose; Fru, fructose.Plots show measurements of five to six samples, each consisting of a pool of five SAMs from plants grown as two independent batches.Welch's t test (P-value shown).(B) Effect of light on STM-VENUS levels in cut inflorescences.Inflorescences of pSTM::STM-VENUS/pSTM::TFP-N7 plants grown under HL were cut and placed in medium without sugar for 48 h under HL (L) or dark (D) conditions, after which the SAMs were dissected and imaged (VENUS).Upper panel, representative STM-VENUS images of SAMs.(Scale bar, 50 µm.)Lower panel, plots showing SAM measurements of plants grown as one to two independent batches normalized by the mean of the uncut condition of each batch (uncut, n = 31, two batches; 48 h L, n = 14, one batch; 48 h D, n = 21, two batches).Different letters indicate statistically significant differences (Kruskal-Wallis with Dunn's test; P < 0.05).(C) Effect of sugar on STM-VENUS levels in cut inflorescences.Inflorescences of pSTM::STM-VENUS/pSTM::TFP-N7 plants grown under HL condition were cut and placed under darkness for 48 h in medium with sucrose (Suc; 2% and 5%) or sorbitol (Sor; 1% and 2.5%) as osmotic control.SAMs were thereafter dissected and imaged (VENUS).Upper panel, representative STM-VENUS images of SAMs.(Scale bar, 50 µm.)Lower panel, plots showing SAM measurements of plants grown as one to three independent batches normalized by the mean of the uncut condition of each batch (uncut, n = 31, three batches; 1% Sor, n = 21, two batches; 2% Suc, n = 28, three batches; 2.5% Sor, n = 7, one batch; 5% Suc, n = 6, one batch).Different letters indicate statistically significant differences (Kruskal-Wallis with Dunn's test; P < 0.05).

Fig. 5 .
Fig. 5. Silencing of SnRK1α in the SAM affects meristem function and plant architecture.(A-F) Representative images of control (STM-VENUS, A and C) and amiRα (B and D-F) plants showing irregular internode length (A and B), clusters of leaves (C, D, and F) and siliques (A, B, and E), and termination of the main inflorescence (F) in the amiRα lines.Insets show organ fusion between leaves of an aerial rosette (D) and between pedicels and the stem (F).(G and H) Quantification of the internode length defects in control and two independent lines of amiRα-1 and amiRα-2 mutants.Internode length was determined by measuring the length of the internodes between paraclades (G) and between the first 12 siliques (H), all counted acropetally.Internode length was scored in the indicated size ranges from the main inflorescence of 18 plants of each genotype.Graphs show the relative frequencies of each size class in the total number of internodes scored.All phenotypes were scored from plants grown under equinoctial conditions until the completion of flowering.

Fig. 6 .
Fig. 6.Model for the role of sugars and SnRK1 signaling in the SAM.Left, under favorable conditions, basal SnRK1 activity is required for meristem organization, with local SnRK1α silencing causing severe phenotypes related to SAM dysfunction and, likely as a consequence, reduced STM expression.The mechanisms underlying these SnRK1 effects remain unknown (indicated by a question mark).Right, under limiting light conditions or other unfavorable situations, sugar levels decrease, leading to a strong activation of SnRK1 signaling.This results in decreased STM protein accumulation, potentially through direct action of SnRK1α1 on STM or an STM partner to reduce SAM activity and growth.

Table 1 . Number of organs per stem node in amiRα plants
Measurements were taken from the main inflorescence of 18 plants of the amiRα lines and the STM-VENUS control line after flowering was completed.Numbers are averages and SD.P-values refer to differences between each mutant and the control (Kruskal-Wallis with Dunn's test).