Phytohormone ethylene-responsive Arabidopsis organ growth under light is in the fine regulation of Photosystem II deficiency-inducible AKIN10 expression

For photoautotrophic plants, light-dependent photosynthesis plays an important role in organismal growth and development. Under light, Arabidopsis hypocotyl growth is promoted by the phytohormone ethylene. Despite well-characterized ethylene signaling pathways, the functions of light in the hormone-inducible growth response still remain elusive. Our cell-based functional and plant-system-based genetic analyses with biophysical and chemical tools showed that a chemical blockade of photosystem (PS) II activity affects ethylene-induced hypocotyl response under light. Interestingly, ethylene responsiveness modulates PSII activity in retrospect. The lack of ethylene responsiveness-inducible PSII inefficiency correlates with the induction of AKIN10 expression. Consistently, overexpression of AKIN10 in transgenic plants suppresses ethylene-inducible hypocotyl growth promotion under illumination as in other ethylene-insensitive mutants. Our findings provide information on how ethylene responsiveness-dependent photosynthetic activity controls evolutionarily conserved energy sensor AKIN10 that fine-tunes EIN3-mediated ethylene signaling responses in organ growth under light.


Results and Discussion
Ethylene promotes hypocotyl growth through PSII regulation in light. In our previous work an abnormal chloroplast structure was found in etr1-1 resulting from a second site mutation on arc3-3 21 . Since etr1-7 is an intragenic suppressor of etr1-1 12 , etr1-7 also enclosed abnormal chloroplasts in mesophyll cells similar to those of etr1-1 (Fig. 1a). To separate the etr1-1 and etr1-7 alleles from the second mutation arc3-3, etr1-1 and etr1-7 were crossed with the wild type (WT; Col-0). Individual allele-specific mutants were isolated without arc3-3 and named as etr1-1sg and etr1-7sg, respectively. Each genetic background was confirmed by using derived cleaved amplified polymorphic sequences (dCAPS) analysis and DNA sequencing (Fig. S1).
Unlike WT, several large chloroplasts with an abnormal structure were observed in protoplasts of ethylene insensitive etr1-1 and ethylene hypersensitive etr1-7 (Fig. 1a). Without the arc3-3 allele, many small chloroplasts were observed in etr1-1sg and etr1-7sg as in WT, confirming that arc3-3 caused the chloroplast abnormality. In WT, etr1-1sg, etr1-7sg, and arc3-3, the accumulation of two representative photosynthetic proteins, PSII D1 (PsbA) and Rubisco large subunit (RbcL) were detected at a similar level by protein blot analysis using specific antibodies (Fig. 1b). Thus, both the PSII protein complex and the downstream metabolic-output machinery in photosynthesis seem to be normal in etr1-1sg and etr1-7sg.
In the light, Arabidopsis WT hypocotyl growth promotion was pronounced on an MS-deficient growth medium in the presence of 50 µM ACC 3, 4 , but not those of ethylene insensitive etr1-1, etr1-1sg, and another well-characterized strong ethylene insensitive mutant ein2-1 (Fig. 1d,e). On the other hand, the hypocotyl growth promotion of two arc3 alleles, arc3-2 and arc3-3, was again pronounced in the presence of ACC and was similar to that of WT. To further investigate photosynthesis functions uncoupled from other effects of light on organ growth promotion, Arabidopsis hypocotyl growth was quantitatively analyzed in the presence and absence of DCMU, which blocks the PSII pathway 39 . Surprisingly, ethylene dependent hypocotyl growth promotion in light was largely suppressed in the presence of DCMU (Fig. 1d,e). Since DCMU is a sensitive blocker that interrupts the electron-transport chains in photosynthesis, its loss-of-photosynthetic activity may modulate cellular energy status and ethylene-dependent hypocotyl elongation under light.
To understand the molecular basis of light-grown Arabidopsis seedling responses to ethylene/ACC in terms of PSII regulation, the expression pattern of marker genes including skotomorphogenesis-related PIF3 5 , photosynthate/sucrose-responsive SUCROSE TRANSPORTER1 (SUC1) and SUC4 40 were examined using quantitative reverse transcriptase-dependent real-time PCR (RT-qPCR) with total RNA extracted from the shoots of 5-day-old WT and etr1-1sg seedlings grown in light. The expression of PIF3, SUC1 and SUC4 was induced by ACC in WT, but not in ethylene insensitive etr1-1sg (Fig. 1f). Moreover, the ACC-inducible marker gene expression was largely suppressed in WT by DCMU. Because EIN3-induced PIF3 plays main regulatory role in hypocotyl growth elongation, regulation of EIN3 level would result in the modulation of ACC/ethylene dependent hypocotyl growth promotion under light. All results suggest that the receptor-dependent ethylene signaling controls EIN3-dependent gene expression leading to hypocotyl growth promotion and perhaps PSII activity is involved in the ethylene-inducible organ growth regulation.
Sugars, the products of photosynthesis, function not only as energy resources, but also as signaling molecules that modulate the expression of genes such as SUC1 and SUC4 40 . To examine whether sugar signaling plays a role in ethylene-inducible hypocotyl growth promotion under light, hypocotyl growth of gin2-1, the plant glucose sensor HEXOKINASE1-null mutant, was monitored under light. The hypocotyls of gin2 grew more or less like those of WT (Ler) when grown in combinations of ACC and DCMU under light (Fig. S2); thus, glucose/sugar as a nutrient resource, but not as an intracellular signal that antagonizes ethylene signaling, apparently has a major role in connecting PSII activity in chloroplasts to ethylene-inducible hypocotyl growth promotion under light. In line with this notion, exogenous 1% (w/v) sucrose could pronounce ACC-dependent and -independent hypocotyl growth promotion in Col-0, but not in ein2-1 and ein3eil1 blocking EIN3-dependent ethylene signaling. This is consistent with our observation that EIN3-dependent PIF3 is necessary in Arabidopsis hypocotyl elongation in the presence of ACC/ethylene under light. Moreover PSII-dependent photosynthetic activity may modulate EIN3 and its downstream regulator PIF3 in ethylene-inducible hypocotyl elongation under light (Fig. S3).
FLIM-based CFL measures PSII activity at single cell resolution. Next we investigated whether ethylene responsiveness could affect PSII activity. First, we measured chlorophyll fluorescence using traditional pulse-amplitude modulation (PAM) fluorimetry and expressed these values as photochemical yields (F v /F m ). Because each leaf of a plant ages differently, only the third and fourth leaves from 3-to 5-week-old soil-grown plants were used for analyses. F v /F m was always higher in WT than in etr1-1 at 3 to 5 weeks after germination ( Fig. 2a,b).
We then measured PSII activity in the chloroplasts at single-cell resolution. To do so, an FLIM technique was applied to obtain CFLs in mesophyll protoplasts freshly isolated from mature leaves of ethylene sensitive WT and ethylene insensitive etr1-1. When color-coded CFLs were superimposed over leaf cell images, many small chloroplasts were largely shown to be bluish-green in WT protoplasts, but several large chloroplasts were yellowish-red in etr1-1 protoplasts (Fig. 2c,d). These results indicated that chlorophyll fluorescence decays rather quickly in the chloroplasts of WT than in those of etr1-1. The convoluted intensity-weighted average CFLs of WT and etr1-1 were estimated to be ca.797 ps and ca.1144 ps, respectively (Fig. 2e, Table 1). The CFL value obtained from WT protoplasts was much shorter than the values obtained from isolated antenna complexes, which were around 2 to 4 ns 41 . This may imply that photo energy transfer to the reaction center occurs rather quickly and efficiently in the cellular environment than in isolated light-harvesting complexes.
CFL reflects the efficiency of PSII activity in the experiment. PSII fluorescence is predominant at 700(+/−20) nm and nearly 100 times greater than that of PSI at room temperature. Furthermore, the plastoquinone pool is fully oxidized by dark incubation before CFL analysis 25 . Under such conditions, CFL values most likely echo the quenching efficiency of the PSII pathway by energy transfer to the reaction center, i.e., photochemical quenching 42,43 . Therefore, when PSII efficiency is low, photochemical quenching is delayed/slow and CFL becomes longer in etr1-1 than in WT ( Fig. 2c-e). The differences in CFLs between WT and etr1-1 were consistent throughout different developmental stages from 3 (mature) to 5 weeks (just before bolting), indicating that ethylene sensitivity/responsiveness modulates PSII activity throughout plant leaf development (Fig. 2g,h). In summary, both FLIM-and PAM fluorimetry-based chlorophyll fluorescence analyses consistently suggest that PSII photochemical activity in abnormal chloroplasts of ethylene-insensitive etr1-1 plants is less efficient than in normal chloroplasts in ethylene-sensitive WT plants.
To further compare FLIM-based CFL and PAM fluorescence analysis, we tried to obtain photochemical yields with FLIM-based fluorescence analysis. When the chlorophyll fluorescence of PSII was measured in mesophyll protoplasts, parts of reaction centers must be closed by laser excitation. In the results, fluorescence values can be attributed to two different states of the reaction centers, open and closed. In WT and etr1-1, these two peaks were found under optimized fitting conditions (Fig. 2i). By referring to previous studies on the CFL of leaves and isolated chloroplasts, shorter (~500 ps) and longer (~1000 ps) peaks in WT were identified for reaction centers at open (τ open ) and partially closed states, respectively (Fig. 2i) [44][45][46][47] . Interestingly, fluorescence values at both peaks were always relatively longer in etr1-1 than in WT. When DCMU that chemically closes PSII pathway was applied, a new fluorescence component of 2700 ps was identified for reaction centers at completely closed state (τ closed ) in WT (Fig. 2j). Then, photochemical yields was calculated as 0.81 as follow- 48 . This F v /F m value calculated from CFL values was close to the values obtained by PAM fluorimetry (Fig. 2a), supporting its fidelity in analyzing PSII activity. The advancement of cellular imaging microscopy now allows specific observations of cellular structures and their functions in vivo. Its application to the measurement of chlorophyll fluorescence from chloroplasts in a single cell enables the accurate and dynamic detection of PSII photochemical efficiency at high resolution in live single cells of a higher plant, Arabidopsis.
Ethylene responsiveness modulates the chlorophyll fluorescence lifetime in PSII. Next we examined whether the chloroplast abnormality observed in etr1-1 may affect the chlorophyll fluorescence of PSII. Unlike etr1-1, the CFLs of etr1-7 and arc3-2 were similar to those of WT (Fig. 3a,b). The arc3-2 allele produces defective chloroplast division, resulting in development of giant chloroplasts as in etr1-7 21,49 . We then examined whether ethylene responsiveness affects the photochemical efficiency of PSII. A FLIM device was fabricated in house and used to monitor the CFLs of other strong ethylene insensitive ein2-1 protoplasts, in addition to ethylene sensitive WT, ethylene insensitive etr1-1, and ethylene hypersensitive etr1-7. CFLs were again longer in etr1-1 and ein2-1 than in WT and etr1-7 (Fig. 3c,d) indicating that the CFL of PSII is under the control of ethylene responsiveness.  To further verify our original observation that ethylene responsiveness modulates the chlorophyll fluorescence of PSII (Fig. 2), FLIM-and PAM fluorimetry-based analyses were conducted with mesophyll protoplasts isolated from etr1-1sg and etr1-7sg, and arc3-3 (Fig. 4a). As shown in etr1-1 (Fig. 3a,b), CFL was also longer in etr1-1sg than in WT, etr1-7sg, and arc3-3 (Fig. 4a,b). Consistently, the quantum yields of PSII were relatively lower in etr1-1sg, but higher in etr1-7sg, than those in ethylene sensitive WT (Fig. 4c). The photochemical efficiency of arc3-3 that is segregated from etr1-1 was similar to that of the WT (Fig. 4a-c), confirming that abnormal chloroplast structure does not affect PSII activity. As in etr1-1 (Fig. 3e), the quantum yields of NPQ in etr1-1sg were relatively higher than those in WT (Fig. 4c). Both FLIM and PAM fluorescence data reliably demonstrated that ethylene responsiveness, but not chloroplast structure, modulates the photochemical efficiency of PSII in the chloroplasts.
To substantiate that the lack of ethylene responsiveness leads to photochemical inefficiency of PSII, changes in CFL were measured from WT protoplasts concomitantly with etr1-1 expression. Protoplasts were co-transfected with a construct encoding a nuclear marker-GFP that harbored DOF1a 50 , allowing for the selection of transfection-positive ones. CFL was relatively longer in WT protoplasts expressing etr1-1 than in those expressing the N-terminal end of split-YFP as a control (Fig. 4d), indicating that the reconstitution of lack of ethylene responsiveness by etr1-1 expression results in a longer CFL of PSII, i.e., low photochemical efficiency. (Figs 2-4) and has a regulatory role in photosynthesis outputs 15,16 , we reasoned that inefficient PSII activity caused by the lack of ethylene responsiveness might affect the cellular energy supply. To test this possibility, the cellular energy stress sensor AKIN10 expression was monitored in WT protoplasts transfected with an etr1-1 construct using a semi-quantitative RT-PCR. Notably, AKIN10 expression was clearly increased by the lack of ethylene responsiveness imposed by etr1-1 expression (Fig. 5a). More specifically mature AKIN10 mRNA expression increased (Fig. 5b), but its pre-mRNA that retains the first intron decreased (Fig. 5c) in our RT-qPCR analysis.

Photochemical activity affects ethylene-inducible hypocotyl elongation under light through the cellular energy sensor SnRK1/AKIN10. Since ethylene responsiveness controls PSII efficiency
To further examine the notion of ethylene responsiveness-dependent AKIN10 expression, we monitored AKIN10 expression in genetically predisposed ethylene insensitive etr1-1sg together with ethylene-sensitive WT. Mature AKIN10 mRNA transcripts accumulated to a much higher level in etr1-1sg than in WT (Fig. 5d). Pre-mRNA transcripts of AKIN10, however, accumulated to a relatively lower level in etr1-1sg than in WT (Fig. 5e). These consistent but conditional results indicated that post-transcriptional regulation seemingly plays a role in AKIN10 induction under the lack of ethylene responsiveness in a condition sensitized to their differential photosynthetic capacities. The mature AKIN10 mRNA accumulated again more in both ethylene-insensitive ein2-1 and ein2-5 than in WT (Fig. 5f), supporting that ethylene responsiveness modulates AKIN10 expression, perhaps through photosynthetic activity regulation. In line with this, mature mRNA of AKIN10 accumulated highly in the presence of DCMU (Fig. S4a). Moreover, such induction of AKIN10 expression seemingly leads to protein kinase activation, because the expression of AKIN10 target genes DIN1 and DIN6 was consistently induced in etr1-1sg (Fig. S4b) and also in WT only in the presence of DCMU (Fig. S4c). Taken together, the lack of ethylene responsiveness leads to PSII inefficiency, resulting in cellular energy deprivation to induce AKIN10 expression and its kinase activation.
To genetically examine whether AKIN10 modulates ethylene-inducible hypocotyl growth under light, hypocotyl growth of transgenic plants expressed with AKIN10 (ref. 19) was observed in the presence of ACC under light together with WT, ethylene-insensitive etr1-1sg, and ethylene-constitutive responsive ctr1-1. The WT hypocotyl growth was again promoted in the presence of ACC/ethylene under light, but not in AKIN10-expressing transgenic plants much as for etr1-1sg (Fig. 5g,h). Consistently, the organ growth was increased in ctr1-1 in the absence and presence ACC under light 3, 4 , thereby reflecting the constitutive ethylene responsive nature of the genetic background. Interestingly, both ACC-dependent and independent growth promotion of ctr1-1 was repressed by DCMU as in WT (Fig. 5g,h). Taken all together, these results strongly support that lack of ethylene responsiveness affects PSII activity and modulates hypocotyl growth promotion by regulating AKIN10 expression under light conditions.
In this study we took advantage of a time-correlated and space-resolved fluorescence microscopic technique to characterize CFLs inversely reflecting PSII activity in vivo utilizing uniform and abundant mesophyll protoplasts isolated from mature Arabidopsis leaves. The CFL-based PSII activity in live single cells correlates well with chlorophyll fluorescence analysis using conventional PAM fluorimetry analysis at the tissue level. Here, we have described that Arabidopsis hypocotyl growth is under the regulation of ethylene signaling pathway (Fig. 5i). Ethylene responsiveness modulates PSII-dependent photosynthesis activity, controlling cellular energy sensor AKIN10 expression. Cellular AKIN10 activity could involve in negative modulation of EIN3-mediated ethylene responsive gene expression. In turn ethylene responsiveness could secure its signaling responses by assisting full activation of PSII efficiency. Further studies on the functional mechanisms of ethylene signaling underlying the regulation of photosystem II efficiency is needed to yield novel strategies for more efficient solar-energy harvesting in higher plants.
The ethylene response assay of etiolated seedlings was performed in complete darkness for 3 days with or without ACC. After 4 days of stratification, seeds were germinated on MS agar plates supplemented with 1% sucrose to ensure uniform germination.
For hypocotyl elongation assays, seedlings were grown on water-agar plates (without added salts or sugars) containing either no ACC or 50 µM ACC and/or 5 µM DCMU for 5 days under photon irradiance of 80 µmol·m −2 s −1 . Experiments were repeated at least three times with 50 seedlings each time.

Transient expression of Arabidopsis mesophyll protoplasts. Protoplast isolation and transient
expression of genes of interest were carried out using previously described methods 22 . The etr1-1 construct was generated by inserting cDNA between the 35SC4PPDK promoter and the NOS terminator in a plant expression vector. The transfected protoplasts were incubated in W5 solution for 15 h and the CFL was measured. Photochemical efficiency measurement. MicroTime 200 (PicoQuant, Germany) was used for fluorescence lifetime imaging microscopy (FLIM) analysis. A diode-pulsed laser (470 nm, 40 MHz, 480 nW, LDH-P-C-470B/PDL 800-B; PicoQuant) was used to excite the chlorophyll, and emissions from 665 to 685 nm were collected in 40,000-pixel mode. Images with Chi-squared values of 0.70 to 1.30 were used in the analysis. Protoplasts were placed on an inverted microscope without any actinic light. The lifetime data of living protoplasts were collected according to the manufacturer's recommendations. The collection time was approximately 1 min for each protoplast image. FLIM data were analyzed and fitted using the Symphotime Program provided by PicoQuant. The photochemical efficiency of PSII was measured using a portable plant efficiency analyzer (Hansatech Instruments, King's Lynn, Norfolk, UK).
PAM fluorimetry analysis. The quantum yield of PSII photochemistry, nonphotochemical quenching, and intrinsic decay were estimated using equations developed previously 4 . Chlorophyll fluorescence from bulk leaves was measured by FMS2 fluorimetry (Hansatech instruments, Norfolk, England). The saturating pulse and actinic light intensity were fixed at 5000 µE and 500 µE, respectively. RNA isolation and transcript accumulation analysis. ACC and/or DCMU-mediated gene regulation was investigated using monitoring marker gene expression in WT and etr1-1sg mutants. For gene expression analysis, total RNA was isolated using RNAiso Plus (Takara Bio, Otsu, Japan) from the shoots of 5-day-old seedlings grown on water-agar plates in the presence and absence of ACC and/or DCMU, and 1 µg of total RNA was used for cDNA synthesis using Moloney murine leukemia virus reverse transcriptase (Promega, Madison, WI, USA). Gene expression was measured using real-time PCR (CFX Connect TM Real-Time System with C1000 Thermal Cycler, Bio-Rad), with seedlings grown on water-agar plates serving as controls. RT-qPCR was carried out using an iQ SYBR Green dye-supplemented PCR mix (Bio-Rad). Either Tubulin4 (TUB4, At1g04820) or Elongation initiation factor4a (EIF4a, At3g13920) transcripts were used as real-time PCR controls with gene-specific primers. Primer sequences are presented in Supplemental Table S1. The production of a single gene product by each primer set in the PCR reactions was validated before each experiment. Experiments were repeated three times (using more than 20 seedlings each time), with consistent results.
Protein extraction and protein blot analysis. Total proteins were extracted from the third and fourth leaves of each plant using an extraction buffer containing 150 mM NaCl, 50 mM Tris-HCl (pH7.5), 5 mM EDTA, 1 mM DTT, 1% (v/v) Triton X-100, and complete protease inhibitor. Protein blot analysis was performed using 40 μg of total protein and protein amounts were quantified using the Bradford method (Bio-Rad). Anti-RbcL-rabbit or anti-PsbA-rabbit antibody (Agrisera) was used for detection.