In Arabidopsis, low blue light enhances phototropism by releasing cryptochrome 1-mediated inhibition of PIF4 expression

Shade-avoiding plants including Arabidopsis thaliana display a number of growth responses elicited by shade cues including elongation of stem-like structures and repositioning of leaves. Shade also promotes phototropism of de-etiolated seedlings through repression of phytochrome B (phyB) presumably to enhance capture of unfiltered sunlight. Light cues indicative of shade include a reduction in the blue and red portions of the solar spectrum and a low red to far-red ratio. Here we show that in Arabidopsis seedlings both low blue and a low red to far-red ratio are required to rapidly enhance phototropism. However, prolonged low blue treatments through reduced cryptochrome 1 (cry1) activation are sufficient to promote phototropism. The enhanced phototropic response of cry1 mutants in the lab and in response to natural canopies depends on PHYTOCHROME INTERACTING FACTORs (PIFs). In favorable light conditions, cry1 limits the expression of PIF4 while in low blue light PIF4 expression increases, which contributes to phototropic enhancement. The analysis of a quantitative DII auxin reporter indicates that low blue light leads to enhanced auxin levels in the hypocotyl and, upon phototropic stimulation, a steeper auxin gradient across the hypocotyl. We conclude that phototropic enhancement by canopy shade results from the combined activities of phytochrome B and cry1 that converge on PIF regulation. ONE SENTENCE SUMMARY The persistent depletion of blue light in natural canopy shade relieves the inhibitory effect of cryptochrome 1 on PIF4, enhancing phototropism in de-etiolated Arabidopsis seedlings. Financial support This work was supported by the University of Lausanne and a grant from the Swiss National Science foundation (n° 310030B_179558 to C.F.); Human Frontier Science Program organization (HFSP) Grant RPG0054-2013, ANR-12-BSV6-0005 grant (AuxiFlo) to T.V.; The University of Buenos Aires (Grant 20020100100437 to J. J. C.), and Agencia Nacional de Promoción Científica y Tecnológica of Argentina (Grant PICT-2018-01695 to J. J. C.). Alessandra Boccaccini and Martina Legris are funded by Marie Curie fellowships H2020-MSCA-IF-2017 grants CRoSh 796283 and Flat-Leaf 796443 respectively.

show that in Arabidopsis seedlings both low blue and a low red to far-red ratio are required to 48 rapidly enhance phototropism. However, prolonged low blue treatments through reduced In natural environments, light conditions are highly dynamic and heterogeneous and given the 60 importance of light for their survival, plants evolved sophisticated photosensory systems to 61 integrate multiple light cues (Casal, 2000;Paik and Huq, 2019). The presence of dense vegetation is 62 not well tolerated by sun-loving plants, such as Arabidopsis thaliana. Plants detect neighbors by 63 sensing the low red (R) to far red (FR) ratio (abbreviated as LRFR), which is a consequence of FR 64 reflection by leaves. If the vegetation becomes denser a canopy filters sunlight, creating an 5 and 7 (PIF4, PIF5, PIF7), which promote expression of YUCCA genes (YUC2, YUC5, YUC8), 85 encoding enzymes for auxin biosynthesis (Hornitschek et al., 2012;Li et al., 2012;Kohnen et al., 86 2016). This increase of auxin synthesis in the cotyledons promotes hypocotyl re-orientation in 87 LRFR (Goyal et al., 2016). 88 Phenotypical experiments of seedlings defective for another class of blue light photoreceptors, 89 called cryptochromes (cry), revealed that they modulate phototropism with a positive role in 90 etiolated seedlings (Whippo and Hangarter, 2003;Ohgishi et al., 2004;Tsuchida-Mayama et al., 91 2010), and a potentially negative role in de-etiolated seedlings (Goyal et al., 2016). The Arabidopsis 92 genome encodes two crys, cry1 and cry2, which coordinate blue light-mediated gene expression by  Wang et al., 2018;Xu et al., 2018;He et al., 2019;Mao et al., 2020). 96 Light-induced activation of cry1 and cry2 is controlled by BIC1 (Blue light Inhibitor of 97 Cryptochrome 1) and BIC2 (Wang et al., 2016). Cry1 and cry2 are associated with chromatin where 98 they are proposed to control transcription factor activity through incompletely characterized 99 mechanisms (Ma et al., 2016;Pedmale et al., 2016). When expressed in a heterologous system cry2 100 has the ability to interact with DNA and promote gene expression in a blue light-induced manner 101 .
suggested that LBL typical of canopy shade also influences hypocotyl re-orientation (Goyal et al.,  Figure 1A). In all conditions analyzed, the seedlings were exposed to supplementary horizontal blue 134 light (8 μmol m −2 s −1 ) during phototropic stimulation ( Figure 1A). We measured deviation from 135 vertical growth after 6 hours of lateral blue light treatment. The overall bending of WT (Col-0) 136 seedlings in SUN/SUN, SUN/LBL and SUN/LRFR was modest, indicating that neither LBL nor 137 LRFR alone were sufficient to trigger a significant enhancement of hypocotyl curvature ( Figure   138 1B,C). However, in SUN/LBL seedlings showed the tendency to bend more and phototropism was 139 significantly enhanced when LBL was combined with LRFR (SUN/CS) ( Figure 1B,C). The LRFR 140 condition described in Goyal et al., 2016, did stimulate phototropism; however, here seedlings were 141 grown in long-days under stronger white light, to more closely mimic a natural environment.

142
Interestingly, LBL starting during the pretreatment period enhanced the phototropic response at 143 equal conditions during the exposure to the BL gradient (compare LBL/LBL vs SUN/LBL, Figure   144 1). In the presence of the same amount of blue light provided unilaterally, the yellow filter used to 145 create the LBL environment changes the blue light differential between the top and the illuminated side. However, this does not appear to be enough to affect phototropism and allowed us to  Figure S1B). Moreover, a 151 treatment with a neutral filter to reduce PAR intensity the day before phototropic stimulation did 152 not affect phototropism (Supplemental Figure S1C). To better define when the LBL pre-treatment 153 was most effective to promote phototropism the following day, LBL treatment was started or ended 154 at different times of the first day (Supplemental Figure S1D). This experiment showed that to be effective the LBL treatment had to occur during the last 4-7 hours of the day prior to phototropic 156 stimulation (Supplemental Figure S1D). We therefore conclude that specifically a prolonged 157 reduction of blue light in the environment promotes phototropism and this is not merely a 158 consequence of enhanced hypocotyl elongation.

159
Persistent LBL relieves the inhibitory effect of cry1 on phototropism. 160 Cryptochomes are the photoreceptors sensing blue light reduction in canopy shade (Keller et al., 161 2011;de Wit et al., 2016;Pedmale et al., 2016) and they also modulate hypocotyl re-orientation in 162 etiolated seedlings (Whippo and Hangarter, 2003;Ohgishi et al., 2004;Tsuchida-Mayama et al., 163 2010). To define cryptochrome function during shade-enhanced phototropism, we compared 164 hypocotyl growth re-orientation of the wild type and cry1 mutant in response to different SUN and 165 LBL (pre-)treatment combinations (Figure 2A Figure S2B). In addition, the increased 175 gradient in SUN never led to the phenotype observed in LBL/LBL (Supplemental Figure S2B).

176
Lastly, in our light conditions, only cry1 and cry1cry2, but not cry2, exhibited a de-repressed 177 phototropic response similar to LBL-pretreated WT seedlings ( Figure 2C). Taken together our 178 experiments indicate that cry1 suppresses the phototropic response in SUN conditions and reduced 179 cry1 activation in LBL releases this suppression. 180 phot1 is needed for phototropism in LBL phot1 is the major photoreceptor initiating phototropism towards relatively low blue light intensities 182 phototropism. We compared the response of the phot1cry1 with cry1 and phot1 single mutants 185 ( Figure 3A). phot1cry1, as well as phot1 hypocotyls re-oriented much less than WT in persistent 186 LBL (LBL/LBL), indicating that phot1 was needed for cry1-mediated phototropism enhancement.

187
Moreover, NPH3, which is essential for phototropism in etiolated and green seedlings (Motchoulski 188 and Liscum, 1999;Goyal et al., 2016) was also required for the response in our conditions ( Figure   189 3B). One of the first steps in phot1 signaling is NPH3 de-phosphorylation, which was recently 190 implicated in modulating the phototropic response as reduced NPH3 de-phosphorylation correlates 191 with accelerated phototropism in seedlings treated for a few hours with light prior to phototropic 192 stimulation (Sullivan et al., 2019). We therefore tested whether the LBL treatment that accelerates 193 phototropism led to changes in NPH3 phosphorylation. NPH3 immunoblots did not reveal any 194 differences among the tested light conditions suggesting that the differences in hypocotyl curvature 195 triggered by LBL were not a consequence of altered NPH3 phosphorylation status ( Figure 3B). We 196 therefore conclude that LBL-enhanced phototropism requires phot1 and NPH3 but we have no 197 evidence for a role of LBL-regulated NPH3 phosphorylation in this process.  Figure 4B). Seedlings were grown in the lab for 4 days before being placed on the south side of a 212 grass canopy (Southern hemisphere) ( Figure 4B). Both phyB and cry1 mutants re-oriented more 213 than WT seedlings, while pif4pif5pif7 showed a weaker phototropic response ( Figure 4B).

214
Interestingly, in this condition, pif4pif5pif7, but not pif4pif5, fully suppressed the cry1 phenotype as 215 observed in the lab in SUN-CS conditions ( Figure 4A, B). Moreover, while pif4pif5pif7 was fully 216 epistatic over cry1, this triple mutant did not fully suppress the phyB phenotype ( Figure 4B). Taken ZT0 to ZT3, but we did not observe an effect of the LBL pretreatment on PIF protein levels ( Figure   229 5B, C). Given that LBL-enhanced phototropism is most effective with a LBL pretreatment we also 230 determined whether this pretreatment altered PIF4 and 5 levels the day prior to the phototropic 231 assay ( Figure 5E, F). In SUN, we observed diel regulation of PIF4 ( Figure 5E) and PIF5 ( Figure   232 5F), with a peak in the middle of the day (ZT8) and a decrease during the last hours of the day 233 (ZT13-17). LBL had a strong effect on PIF4 protein levels ( Figure 5D, E). PIF4 levels remained 234 high for much longer during the day and only returned to the same levels as in SUN-treated samples at ZT19 ( Figure 5E). LBL had a more modest effect on PIF5 protein levels which declined slightly 236 slower in LBL than in SUN conditions ( Figure 5F). As a control, we probed the membrane with 237 CRY1 and CRY2 antibodies. As reported previously (Shalitin et al., 2002) LBL led to higher levels 238 of CRY2 protein but not CRY1 ( Figure 5E, F). We conclude that LBL has a strong effect on PIF4 239 protein levels particularly towards the end of the day. To determine how LBL regulates PIF protein abundance we determined the effect of this light 242 treatment on PIF transcript abundance using RT-qPCR. These experiments revealed that PIF4 243 ( Figure 6A), but not PIF5 (Supplemental Figure S3), transcript levels increased in LBL, as 244 described previously (Pedmale et al., 2016). However, the LBL treatment did not alter the diel 245 expression profile of PIF4 and PIF5 ( Figure 6A, Supplemental Figure S3). PIF4 levels were higher 246 in SUN-grown cry1 mutants, which expressed PIF4 levels similar to those observed in LBL-grown 247 WT seedlings ( Figure 6A). The negative effect of cry1 on PIF4 abundance was also observed by 248 immunoblotting using a PIF4 antibody ( Figure 6B). This effect on PIF4 protein abundance was  Our experiments indicate that high cry1 activity and low PIF4 levels limit phototropism in high 265 light (SUN) conditions. A prediction of this model is that a mutant with high cry activity is 266 expected to have reduced PIF4 levels and be less responsive to blue light gradients. We tested this using the bic1bic2 (b1b2) double mutant which has higher cry activity (Wang et al., 2016). b1b2 268 had the same phototropic response than the WT in SUN/LBL conditions. However, in persistent 269 LBL which strongly promotes phototropism in the WT, b1b2 showed a reduced phototropic 270 response and reduced levels of PIF4 ( Figure 6C, D). Taken together our data indicate that cry1 271 controls phototropism at least in part by controlling PIF4 levels.

272
Phototropism in LBL requires auxin transport, but also biosynthesis and signaling. 273 Asymmetrical hypocotyl growth is ensured by differential auxin distribution, which is mediated by

311
Our experiments revealed that LBL strongly contributes to phototropic enhancement. For LBL to be 312 effective on its own, it is required for several hours the day prior and during phototropic 313 stimulation ( Figure 1B, Supplemental Figure S1). This might be due to the slower effect of LBL,  Figure S1). The fact that LBL alone when applied from the day prior to phototropic 318 stimulation was sufficient to promote phototropism allowed us to specifically study the role of this 319 component of canopy shade in phototropism enhancement. Altering blue light from the top to 320 generate LBL also modifies the horizontal blue light gradient in our experimental setup ( Figure  1A). However, several experiments allowed us to demonstrate that phototropism enhancement in 322 LBL is not simply a consequence of a modified light gradient (Figure 1, Supplemental Figure S2). 323 We conclude that ambient LBL is an important feature of canopy shade enhancing phototropism. 324 cry1 has a negative effect on hypocotyl re-orientation of green seedlings 325 Our experiments show that in de-etiolated seedlings cry1 inhibits phototropism in favorable (SUN-326 mimicking) light conditions (Figures 2, Supplemental Figure S2). In the conditions we tested phot1 327 is the primary photoreceptor controlling hypocotyl reorientation and the enhanced response of cry1 328 mutants depends on phot1 (Figure 3). In seedlings treated with a few hours of light to initiate de-329 etiolation NPH3 phosphorylation has an effect on phototropism (Sullivan et al., 2019). In our 330 conditions, we found that NPH3 is essential for phototropism but we did not detect an effect of LBL  The importance of PIF4 and PIF5 in controlling LBL-induced phototropism downstream of cry1 342 prompted us to analyze PIF4/PIF5 regulation by light and cry1. LBL treatment leads to elevated 343 PIF4-HA and to a lesser extent PIF5-HA towards the end of the day ( Figure 5). This data was 344 confirmed for PIF4 using an antibody recognizing the endogenous protein ( Figure 6). PIF4 seems to 345 have a predominant role in regulating hypocotyl elongation in LBL, in fact, pif4 elongates similarly to pif4,5 double mutant, but less than pif5 and pif4 alone abolishes cry1 elongation phenotype 347 (Pedmale et al., 2016). Our data showed that cry1 regulates PIF4 levels, as shown by the analysis of 348 PIF4 levels in cry1 and bic1bic2 mutants. The cry1 mutant has higher PIF4 levels than the WT in 349 SUN simulating conditions and the bic1bic2 double mutant, with higher cry activity (Wang et al.,350 2016; Wang et al., 2017), has lower PIF4 levels in LBL correlating with a reduced phototropic 351 response ( Figure 6). The effect of LBL and cry1 on PIF4 levels could, at least in part, be due to 352 transcriptional regulation given that PIF4 transcript levels are higher in LBL than in SUN 353 conditions ( Figure 6A). Moreover, LBL-regulated PIF4 levels were essentially absent in cry1 354 mutants which always expressed higher PIF4 levels than SUN-treated WT ( Figure 6A). These data Several reports have demonstrated impaired hypocotyl elongation responses to LBL in mutants 374 defective in auxin transport and auxin biosynthesis (Pierik et al., 2009;Keuskamp et al., 2011;de 375 Wit et al., 2016). Deficient enhancement of the phototropic response by LBL in the sav3, 376 yuc2yuc5yuc8yuc9, pin3pin4pin7, tir1 and msg2 mutants (Figure 7) indicate that this process 377 requires normal auxin synthesis, transport, perception and signaling. A priori, the phenotype of 378 these mutants might simply indicate that normal auxin synthesis, transport, perception and signaling 379 are a condition for the LBL effects, or that the auxin system carries LBL information. In this regard,  (Goyal et al., 2016). Therefore, we used qDII-Venus to investigate 391 whether the auxin system actually carries the LBL information. Our data indicate that the latter is 392 actually the case because a LBL pretreatment leads to higher auxin levels and/or sensitivity in the 393 hypocotyl ( Figure 7C) and a steeper gradient of auxin levels and/or sensitivity upon phototropic 394 stimulation ( Figure 7B), which correlates with enhanced phototropism (Figure 1). 395 We conclude that phototropic enhancement by canopy shade involves changes in activity of at least 396 three photoreceptors: phot1, cry1 and phyB (Figures 2, 3)  (LBL) and it was measured by white diffuser filter combined with blue filter (400-500nm). The R 429 (640-700 nm)/FR (700-760 nm) ratio was measured using the Ocean Optics USB2000+ 430 spectrometer. In SUN and LBL the R/FR ratio was 1.25 and in LRFR and CS it was 0.3, obtained 431 by adding FR LED to white light lamps. The LEE filter number 298 0.15ND was used in Figure   432 Supplemental S1 to reduce PAR similarly to the yellow filter used for LBL treatment. Transmission 433 through ND filter is 69.3%, the transmission of PAR through the yellow filter is about 76.6%.

435
Phototropic stimulation by the application of lateral blue light was always started after the lights 436 turned on, between ZT0 and ZT0.5 by removing one side of the black boxes and supplying 437 unilateral blue light by LED. The LED source was placed 60 cm distant from the black boxes to 438 raise the horizontal blue light up to 8μmol m −2 s −1 . In addition, we calculated the blue light 439 differential between the illuminated side and the top (i.e., the light coming from above), because the 440 yellow filter in the LBL condition affected this differential and we had to evaluate its potential 441 physiological impact. The log10 of the side-to-top differential of blue light was 0.0 (log10 (side The genotype used was pPDF1::DII-n7-Venus-2A-mTurquoise-sv40 t35 in the Col-0 background.

478
Seedlings were grown as for phototropic experiment. On the fourth day, half of the plates were 479 shifted in LBL and half of them were kept in SUN. The fifth day, all plates were covered with a 480 yellow filter before the start of the day and 1h after the start of the day one side of the black box 481 was opened to perform the phototropism assay. Confocal images were taken between ZT23,5 and 482 ZT0,5 and between 1h and 2h after starting the phototropic assay. All pictures were taken in the 483 epidermis in the elongation zone using an LSM710 confocal microscope ( differences at p-value < 0.05 obtained by two way ANOVA followed by the post hoc Tukey's HSD.