A Constitutively Active Allele of Phytochrome B Maintains Circadian Robustness in the Absence of Light

: The sensitivity of the circadian system to light allows entrainment of the clock, permitting coordination of 56 plant metabolic function and flowering time across seasons. Light affects the circadian system both via 57 photoreceptors, such as phytochromes and cryptochromes, and sugar production by photosynthesis. In the 58 present studies, we introduce a constitutively active version of phytochrome B (phyB-Y276H, YHB) into 59 both wild-type and phytochrome null backgrounds of Arabidopsis thaliana to distinguish the effects of 60 photoreceptor signalling on clock function from those of photosynthesis. We find that the YHB mutation is 61 sufficient to phenocopy red light input into the circadian mechanism and to sustain robust rhythms in 62 steady-state mRNA levels even in plants grown without light or exogenous sugars. The pace of the clock is 63 insensitive to light intensity in YHB plants, indicating that light input to the clock is constitutively activated 64 by this allele. Mutation of YHB so that it is retained in the cytoplasm abrogates its effects on clock function, 65 indicating that nuclear localization of phytochrome is necessary for its clock regulatory activity. We also demonstrate a role for phytochrome C as part of the red light sensing network that modulates phytochrome B signalling input into the circadian system. Our findings indicate that phytochrome signaling in the nucleus plays a critical role in sustaining robust clock function under red light, even in the absence of photosynthesis or exogenous sources of energy. We next examined the effects of activated phytochrome containing the G767R mutation on circadian pace. The YHB or YHB-G767 transgenes were crossed into phyABCDE mutants the reporter and rhythms in bioluminescence activity were assessed in dark-adapting plants grown in the presence of exogenous sucrose. While YHB ( phyABCDE) plants had a shorter period of 23.68 ± 0.08 h compared to the parental phyABCDE (26.57 ± 0.48 h), the period length of YHB- G767R ( phyABCDE ) plants was indistinguishable from the control (27.04 ± 0.37 h). These data indicate that shortening circadian YHB its enhancement of in transcript abundance nuclear localization.


INTRODUCTION
7 Tryon et al., 2007) and so we generated Columbia plants containing both the YHB allele and the 146 CCR2::LUC reporter in addition to introducing a CCA1::LUC2 reporter into existing YHB (Landsberg 147 erecta; Ler) lines (Hu et al., 2009). Introduction of the YHB allele confers shortened hypocotyls and 148 expanded cotyledons in Ler seedlings grown in darkness or in constant light (Su and Lagarias, 2007;Hu et 149 al., 2009). Similar short hypocotyls were observed in our newly generated YHB (Col) lines when compared 150 with wild-type Col controls grown under white light (p<0.05, Fig. 1A-B). YHB was moderately 151 overexpressed compared to endogenous phyB in these lines (Fig. 1C).

152
We initially tested whether constitutively 'active' YHB protein would be sufficient to maintain 153 robust luciferase rhythms in light/dark (L/D)-entrained plants transferred to constant darkness ( Fig. 2A-D).

154
Assessments of circadian rhythms have historically used sucrose as a media supplement in order to enhance 155 bioluminescence in transgenic plants (Millar et al., 1992), although recent work has demonstrated that this 156 exogenous sucrose can itself act as an entrainment signal (Dalchau et al., 2011;Haydon et al., 2013). We 157 consequently compared luciferase activity in our YHB lines in either the presence or absence of exogenous 158 sucrose to facilitate comparison with historical and more recent datasets. We observed that rhythms of 159 CCR2-driven luciferase activity in YHB seedlings grown with sucrose had increased amplitude throughout 160 the experiment compared to wild-type controls (p<0.001, Fig. 2A), although there was no significant 161 difference in period length between the two populations (τ=25.4±0.21 and 24.93±0.08 for wild-type and 162 YHB respectively, p=0.1). CCA1-driven luciferase activity was similarly increased in YHB (Ler) lines in the 163 presence of sucrose (Fig. 2B). In the absence of sucrose, we observed that rhythmic bioluminescence in

172
In order to better understand the role of exogenous sucrose in the maintenance of clock-regulated 173 gene expression in darkness, we examined the steady-state levels of CCR2 and CCA1 transcripts in both 174 YHB (Col) and wild-type control lines ( Fig. 2E-H). In the absence of sucrose, CCR2 transcript 175 accumulation in wild type mirrored the luciferase activity data, with one rhythmic peak of transcript 176 accumulation before dampening towards arrhythmia (Fig. 2E). Rhythms were more robust in YHB plants 8 assessed transcript levels of several core circadian clock genes (Hsu and Harmer, 2014). The presence of 183 YHB was sufficient to maintain rhythms in transcript levels of the morning-phased genes CCA1 and PRR9 184 as well as the evening-phased genes GIGANTEA (GI), TOC1, and ELF4 . In all 185 cases, transcript accumulation rhythms dampened more significantly in wild-type seedlings than in YHB 186 seedlings. These results indicate that, independent of the presence of sucrose, YHB acts to sustain rhythmic 187 expression of core clock transcripts in constant darkness, a phenomenon not seen in wild-type seedlings.

188
Circadian period of YHB plants is insensitive to fluence rate 189 We next examined YHB influence on circadian rhythms under a range of intensities of constant red light 190 (Rc). Following an entrainment period, wild-type, YHB and phyB-9 seedlings grown on sucrose-free media 191 were released into Rc, with circadian period and amplitude measured via activity of the CCR2::LUC 192 reporter (Fig. 4). Similar to previous reports (Somers et al., 1998;Palagyi et al., 2010), the circadian period 193 of wild-type plants shortened from 25.1 ± 0.33 h to 22.9 ± 0.13 h as the fluence rate increased from 12 to 194 184 μmol m -2 s -1 (Fig. 4A). PhyB-9 seedlings exhibited a similar, albeit more exaggerated response over the 195 fluence rates tested, as was reported previously (Fig. 4A, Somers et al., 1998;Palagyi et al., 2010). By 196 contrast, YHB seedlings were essentially unresponsive to increasing fluence rates of Rc (up to 184 μmol m -2 197 s -1 ), with period length remaining approximately 23.5 h at all fluence rates tested (Fig. 4A). This 198 unresponsiveness resulted in the greatest period difference between YHB and wild type at the lowest 199 fluence rate of Rc (12 μmol m -2 s -1 Rc) tested, where a ~1.5 h shorter period was observed in the transgenic 200 plant (Fig. 4A). By contrast, under high intensity Rc the period lengths of all genotypes were nearly 201 identical. Taken together, these results indicate that period length in YHB plants is nearly insensitive to 202 increasing fluence rates of red light.

203
Differences in the amplitude of bioluminescence rhythms were also detected for YHB and wild-204 type plants after transfer to high fluence rates of Rc. Although similar for all genotypes under 12 μmol m -2 205 s -1 Rc, the amplitude of bioluminescence was greatly enhanced in the wild type with increasing fluence rate 206 of Rc, whereas the responsiveness of the YHB plants was reduced. Indeed, the discrepancy between the 207 amplitude of these genotypes was most pronounced at the highest fluence rates examined, where YHB 208 seedlings were half as bright as wild-type controls . Similar to the period phenotype, these 209 results indicate that clock amplitude in YHB plants is less responsive to increasing fluence rates of Rc than 210 in wild type.

212
Type II phys (phyB-E) form homo-and heterodimers that complicate interpretation of phenotypes of loss-213 of-function phy mutants and of gain-of-function YHB transgenics (Sharrock and Clack, 2004;Clack et al., 214 2009;Hu et al., 2009). The additive circadian defect of phyABD mutants compared to phyAB has been 215 assumed to indicate an ability of phyD to provide light input into the circadian system in the absence of 216 phyB (Devlin and Kay, 2000). However, the loss of phyD potentially also alters the amount of phyC-phyE 217 heterodimers in the two genotypes, providing an alternative explanation for their distinct phenotypes.

218
Similarly, introduction of the YHB allele would alter the amounts of the homo-and hetero-dimeric species 9 of endogenous phyB-E proteins. To better understand how YHB and by extension phyB influence the 220 circadian system, we introduced the YHB allele into the recently isolated phyABCDE quintuple mutant (Hu 221 et al., 2013). In contrast to the photomorphogenesis-challenged phenotypes of the phyABCDE parental line 222 under 12:12 L/D cycles, YHB(phyABCDE) plants looked similar to wild-type seedlings with short 223 hypocotyls and expanded cotyledons ( Fig. 5A-B).

224
To further explore the effect of YHB in the absence of other phytochromes, we assessed the 225 accumulation of core clock gene transcripts in the phyABCDE mutant background using qRT-PCR. The 226 phyABCDE mutant was generated in the Ler background; hence we used a previously reported YHB line 227 (YHB(Ler); Su and Lagarias, 2007;Hu et al., 2009) as a control. As observed for YHB(Ler) plants,

228
YHB(phyABCDE) seedlings transferred to constant darkness displayed robustly rhythmic CCA1 229 accumulation -a response that is strongly damped in Ler wild type (Fig. 6A). YHB(phyABCDE) also 230 sustained rhythmic expression of PRR9, GI and ELF4 transcripts in prolonged darkness ( Fig. 6B-D). All of 231 these clock genes displayed dampened oscillations in dark-adapting Ler wild type, similar to our results in 232 the Col accession ( Fig. 2 and Fig. 3). These data indicate that endogenous phytochromes are not required 233 for YHB-mediated maintenance of robust circadian rhythms in darkness.

234
Cytosolic YHB has little effect on clock gene expression or circadian pace

235
In contrast to endogenous phytochrome, YHB does not require light activation to migrate from the 236 cytoplasm to nucleus and is instead constitutively targeted to the nucleus (Su and Lagarias, 2007). In the 237 nucleus, YHB acts similarly to the P fr form of phyB by binding PIF bHLH transcription factors and 238 targeting them for degradation (reviewed by Bae and Choi, 2008). More recently, signaling roles for P fr in 239 the cytoplasm have been reported (Paik et al., 2012;Hughes, 2013). In order to evaluate the contribution of 240 cytoplasmic P fr signaling into the circadian system, we introduced the G767R mutation into the YHB allele 241 (YHB-G767R). Phytochromes containing the G767R mutation are retained in the cytoplasm (Wagner and 242 Quail, 1995;Ni et al., 1999;Matsushita et al., 2003). This has been attributed to the inability of the G767R 243 mutant to interact with PIF3 and then be imported into the nucleus (Pfeiffer et al., 2012). Surprisingly, the 244 double phytochrome mutant partially complemented the phyABCDE null mutant: light-grown YHB-245 G767R(phyABCDE) lines exhibited expanded cotyledons and shorter hypocotyls compared with the 246 parental phyABCDE seedlings (Fig. 5A, B). However, these results contrast with the strong hyperactivity of 247 YHB in both null and wild type backgrounds (Fig. 5A, B). Thus it is clear that the G757R mutation largely 248 suppresses the gain-of-function activity of YHB, presumably by retaining it in the cytosol.

249
In order to assess the role of cytoplasmic YHB-G767R within the circadian system, we used qRT-250 PCR to assess its effects on transcript levels of genes with poor cycling in wild-type plants maintained in

277
PhyC mutants have hypocotyl growth defects (Franklin et al., 2003;Monte et al., 2003) implying 278 an important role of phyC in modulating the activity of other phytochromes (Franklin et al., 2003;Monte et 279 al., 2003;Hu et al., 2013). Our data show that phyC influences the activity of YHB under Rc (Fig. 7C), 280 supporting the hypothesis that phyC acts as a light input into the circadian system. To more directly test this  We have assessed circadian clock function in YHB-expressing seedlings, allowing us to evaluate the effects 293 of a single active phytochrome species on the circadian system independently from light effects on 294 photosynthesis. The YHB mimic of light-activated phyB was sufficient to sustain high amplitude, rhythmic 295 accumulation of CCA1, PRR9, TOC1 and GI transcripts in constant darkness in the absence of exogenous 296 sugar in both Col and Ler accessions (Fig. 2E, Fig. 2G, Fig. 3, and Fig. 6). This may be due to the 297 increased expression of ELF4 in YHB plants, which shows a robust peak of expression on the first 298 subjective day of free run in this genotype but not in wild type ( Fig. 3D and Fig. 6D). This difference 299 precedes the first-observed difference in cyclic amplitude in CCA1 and PRR9 transcripts in YHB and 300 control plants, which is not seen until the morning of the second subjective day in free run ( Fig. 2G, Fig.   301 3A, Fig. 6A, 6B). ELF4 forms part of the Evening Complex (Nusinow et al., 2011) which directly represses 302 expression of clock genes such as PRR7, PRR9, GI, and LUX (Herrero et al., 2012;Mizuno et al., 2014; 303 Box et al. 2015). Intriguingly, ELF4 also is necessary for red light mediated induction of CCA1 and LHY 304 (Kikis et al., 2005). We therefore suggest that sustained high amplitude expression of Evening Complex 305 components contributes to the maintenance of transcriptional rhythms in YHB plants in the dark.

306
Although luciferase and transcript oscillations were observed in YHB lines in constant darkness, 307 the activity of YHB was not sufficient to prevent lengthening of the circadian period under these conditions 308 when compared to even dim red light. We observed periods of 26 hours in YHB CCA1::LUC2 reporter 309 lines (Fig. 2D), and the later phase of peak CCA1 transcript accumulation also suggests a longer-than-24-

320
Dissecting the role of phytochromes as light inputs to the circadian system

321
The circadian period lengths of many diurnal species, including plants, are shortened in response to higher 322 fluence rates of constant light, a phenomenon known as Aschoff's rule (Aschoff, 1960;Somers et al., 1998; 323 Devlin and Kay, 2000). This pattern is apparent in the red light fluence rate response curves presented here Our analysis clearly indicates that YHB activity sustains phy-signaling input into the circadian 329 system in darkness regardless of the presence of exogenous sucrose. However, in continuous light it is also 330 clear that the clock receives additional red light-derived signaling cues, either from other phys, from the 331 effects of light driven chlorophyll synthesis, and/or from metabolic changes induced by photosynthesis 332 itself (Hu et al., 2013). PhyA, phyB and phyD have each been shown to contribute towards light perception 333 by the circadian system (Somers et al., 1998;Devlin and Kay, 2000). Recent studies reveal phyABDE and 334 phyABCDE mutants to have indistinguishable circadian phenotypes (Hu et al., 2013), consistent with the 335 evidence that phyC protein is unstable in the absence of other phytochromes (Clack et al., 2009). The 336 current study defines a role for phyC within the circadian system, by demonstrating both a circadian 337 phenotype in phyC mutants and modulation of YHB activity by phyC (Fig. 7). The long period phenotype  to LUX increasing while interactions with CCA1 and TOC1 diminish (Yeom et al., 2014). Since all of the 357 clock components function in the nucleus, yet require synthesis and transit through the cytosol, it is 358 possible that interactions with phytochromes could occur both in the nucleus and the cytosol. However, our 359 analyses indicate that cytosolic YHB-G767R is unable to sustain circadian rhythms seen in YHB lines in 360 constant darkness, nor does it shorten the clock period as measured by CCA1::LUC-dependent 361 luminescence (Fig. 6). These activities thus appear dependent on the nuclear localization of YHB. However, 362 YHB-G767R seems to evoke an advance in the phase of PRR9 expression during the early stages of free 363 run, suggesting a modest cytoplasmic role for YHB at least within this sub-loop of the circadian system 364 (Fig. 6B). We speculate that this response could be due to cytosolic retention of Pfr-interacting factors such 13 as TOC1 that inhibit expression of PRR9 in the nucleus (Huang et al., 2012), an intriguing possibility that 366 we will explore in future studies. 14 370

395
Bioluminescence from groups of 10 seedlings were pooled for each data point where seedlings were 396 transferred to constant darkness. Imaging was completed over 5 days and data was processed using 397 Metamorph software (Molecular Devices). Patterns of luciferase activity were fitted to cosine waves using 398 Fourier Fast Transform-Non-Linear Least Squares (Plautz et al., 1997) to estimate circadian period length.

401
RNA was isolated and qRT-PCR performed as previously described (Jones et al., 2010

409
Dark-grown, 4-d-old seedlings were harvested for protein extraction as previously described (Su and 410 Lagarias, 2007). After quantifying the total protein concentrations with the Pierce® BCA protein assay kit 411 (Thermo Scientific), equal amounts of proteins were separated on 4-20% ExpressPlus™ PAGE gels 412 (GenScript) and then semi-dry transferred onto Immobilon ® -FL PVDF membrane (EMD Millipore). PhyB 413 and actin were immunodetected by anti-phyB B1 (generous gift from Dr. Peter Quail, 1:300 dilution) and                 Previously described Ler seedlings similarly transformed with YHB are presented for comparison (Su et al. 2007). SEM is presented, n>20. * indicates significant difference for the indicated comparison (p<0.        , GIGANTEA (C) and ELF4 (D) mRNA were assessed. Plants were entrained for 10 d in 12:12 L/D cycles on sucrose-free MS media with 60 μmol m -2 s -1 white light before transfer to constant darkness. mRNA levels for each gene were normalized to PP2a; SEM is shown. (E) Circadian periodicity of phyABCDE, YHB(phyABCDE) and YHB-G767R(phyABCDE) seedlings expressing a CCA1::LUC2 reporter when grown on MS + sucrose plates. Plants were entrained for 6 days under 60 μmol m -2 s -1 white light in 12:12 L/D cycles before being transferred to constant darkness at ZT12. Bioluminescence from groups of 5 seedlings were pooled for each datapoint, n>8. SEM is shown, * indicates significant difference (Student's t test).  and YHB(phyABCDE) seedlings transformed with CCA1::LUC2 after transfer to constant darkness. Plants were grown under 60 μmol m -2 s -1 white light in 12:12 L/D cycles for 6 d with supplementary sucrose before being transferred to constant darkness at ZT12. Bioluminescence from groups of 5 seedlings were pooled for each datapoint, n>9. (B) Abundance of CCA1 transcripts under constant darkness in YHB(ABDE) and YHB(ABCDE) seedlings using qRT-PCR. Plants were entrained to 12:12 L/D cycles on sucrose-free MS media under 60 μmol m -2 s -1 white light for 10 d before transfer to constant darkness at ZT12. mRNA levels for each gene were normalized to PP2a; SEM is shown. (C) Circadian periodicity of Ler, YHB(phyABDE) and YHB(phyABCDE) seedlings transferred to dim red light (1 μmol m -2 s -1 ). Seedlings were grown on 0.5x sucrose-free MS media and entrained for 6 days in 12:12 L/D cycles under 60 μmol m -2 s -1 white light before being transferred to constant red light. Bioluminescence from groups of 5 seedlings were pooled for each datapoint, n>7. (D) Period estimates of Col-0, phyC-2 and phyC-4 seedlings under constant red light. Plants were entrained in 12:12 L/D cycles for 6 days before transfer to 20 μmol m -2 s -1 constant red light. The insert shows a schematic cartoon of the PHYC locus indicating T-DNA insertion locations for phyC-2 and phyC-4. 5' and 3' UTRs are shown in white boxes, exons in grey. T-DNA insertion points are indicated with white triangles. (E) Period estimates of seedlings transformed with a CCA1::LUC2 reporter. Wild type (Col-0, solid line), phyC-2 (green) and phyC-4 (orange) seedlings were entrained as described in (D) before being transferred to 20 μ mol m -2 s -1 constant red light. (F) Fluence rate response curve to evaluate the effect of phyC on the free running period of the circadian system. Wild type (Col-0, solid line), phyC-2 (green) and phyC-4 (orange) were entrained as described in (D) before being transferred to constant red light at the indicated fluence rate. SEM is shown, * indicates significant difference from wild-type (Bonferroni adjusted Student's t test).