RESULTS

Photoperiod stress increases the CK content in wild-type plants

Plants impaired in CK biosynthesis or signaling are sensitive to photoperiod stress (Nitschke et al., 2016). To investigate whether photoperiod stress influences the CK concentration, we have measured CK in leaves of SD-grown wild-type plants exposed to a PLP of 32 h, which is the standard stress treatment used in this study (Fig. 1A). The altered light regime caused an elevated total CK concentration at the end of the PLP and in the middle of the following night (Fig 1B; time points 2 and 3). The concentration of CK free bases was elevated up to three-fold in PLP plants compared to control plants at the end of the PLP and in the middle and at the end of the following night (Fig 1C; time points 2, 3 and 4). A similar pattern was observed for the concentration of CK ribosides (Fig. 1D). CK nucleotides were increased two-fold in PLP plants compared to control plants after 16 h of additional light (Fig.1E; time point 1) and stayed elevated in PLP plants in comparison to control plants until the end of the night following the PLP (Fig. 1E.; time points 2 and 3). Concentrations of CK O -glucosides were elevated in PLP plants during and at the end of the night following the PLP while concentrations of N -glucosides did not differ between stressed and control plants (Fig. 1F, G). The increase in the sum of free bases, nucleosides and nucleotides was reflected by the increased concentrations of the respective individual iP-, t Z- and DHZ-type CK metabolites already during the PLP (Table S1; time points 1 and 2). In contrast, the concentrations of c ZR and c ZRMP levels were decreased in PLP plants at early time points but strongly increased at the end of the night following the PLP and the day after (Table S1; time points 1, 2, 4, 5). Taken together, photoperiod stress treatment led to an increase of all types of isoprenoid class CKs including the bioactive CKs iP and t Z as well as their transport forms and precursors.

Root-derived tZ-type CKs protect plants from photoperiod stress

Since stressed wild-type plants increased the concentration of the functionally most relevant CKs - iP and t Z - we wondered which of these two CKs might be protective against photoperiod stress. Therefore, we investigated the involvement of t Z-type CKs by exposing mutants impaired in either the biosynthesis of t Z-type CKs (cypDM ; Kiba et al., 2013) or their transport from the root to the shoot (abcg14 ; Ko et al., 2014; Zhang et al., 2014) to photoperiod stress. Only results of PLP-treated plants are shown since control plants do not show any differences in Fv/Fm, lesion formation nor an altered gene expression during the course of the experimental treatment (Nitschke et al., 2016).
Over 80 % of the leaves ofcypDM and abcg14 mutants showed lesion formation after photoperiod stress treatment, which was a four-fold increase compared to wild-type plants (Fig. 2B, Suppl. Fig. 1A). Furthermore, photoperiod stress caused a drop in Fv/Fm to 0.35 in these mutants while wild-type leaves had an Fv/Fm value of 0.8 (Fig. 2C). The transcript abundance of the stress marker genes BAP1 and ZAT12 was increased in the response to stress two- to three-fold higher in the mutants as compared to wild type (Fig. 2D, E). The abundance of CAB2transcript was strongly decreased in all genotypes but much stronger in both mutants compared to wild type 15 h after the PLP (Fig. 2F). Summing up, these results support a protective function of root-derivedt Z-type CKs against photoperiod stress.

Watering of cypDM plants with tZ ort ZR reduces the response to photoperiod stress

A recent study by Osugi et al. (2017) demonstrated that under long-day conditions root-derived t Z has distinct functions in the shoot as compared to root-derived t ZR, for example in regulating the size of leaves and of the SAM. In order to dissect the role of root-derivedt Z and t ZR in photoperiod stress, we watered cypDMplants with either 10 µM t Z or 10 µM t ZR daily during the whole cultivation period and exposed them subsequently to photoperiod stress. The effectiveness of the treatment was tested by determining the expression of CK response genes ARR5 and ARR6(Supplemental Fig. S2). Expression of both genes was lower in controlcypDM plants compared to wild type but could be rescued by application of t ZR and t Z.
Moreover, t ZR application reduced lesion formation incypDM plants in response to photoperiod stress by about 15 % compared to untreated cypDM plants (Fig. 3A, Suppl. Fig. 1B). Also, the decrease in photosynthetic capacity of t ZR-treated plants was lower compared to untreated cypDM controls and almost like wild type (Fig. 3B). These results indicate that t ZR applied through roots has a protective effect against photoperiod stress. Watering plants with t Z suppressed the photoperiod stress syndrome in cypDM plants almost completely suggesting that also root-derived t Z protects plants during photoperiod stress (Fig. 3A, B). At the molecular level, DMSO treatment itself lowered the expression of stress marker genes ZAT12 and BAP1 (Fig. 3C, D). t ZR and t Z supplementation reduced the induction of these genes as well. The rescue of gene regulation as a response to photoperiod stress by t Z was particularly evident in the case ofCAB2 (Fig. 3E). In summary, supplementation experiments indicated that lesion formation, the decrease in photosynthetic capacity and the transcriptional response can be rescued to a different extent byt Z and t ZR.

AHP2, AHP3 and AHP5 act redundantly in photoperiod stress signaling

In Arabidopsis , AHK receptors transduce the CK signal to AHPs and phosphorylated AHP1 to AHP5 activate type-B ARRs (Hutchison et al., 2006). Although AHPs are involved in several developmental processes and responses to stress (Hutchison et al., 2006), their role in photoperiod stress has not been investigated so far. Thus, the ahp2,3,5 triple mutant as well as the corresponding double mutants were exposed to photoperiod stress.
Compared to wild-type plants, about twice more leaves showed lesion formation in ahp2,3 and ahp2,3,5 plants (Fig. 4A, Supplemental Fig. 1C). In correspondence, the photosynthetic capacity ofahp2,3,5 plants was decreased compared to all other genotypes (Fig. 4B). Functional redundancy of AHPs in the response to photoperiod stress was also reflected by the response of marker genes. While the stronger induction of BAP1 and ZAT12 expression during the night following the PLP was apparent in all ahp double and triple mutants compared to wild type, the amplitude was the highest inahp2,3 and ahp2,3,5 (Fig. 4C, D). Similarly, a decrease ofCAB2 transcript levels (two- to three-fold) was more apparent inahp2,3 and ahp2,3,5 plants than in ahp2,5 andahp3,5 15 hours after the PLP (Fig. 4E).
Summing up, AHPs were shown to act redundantly in photoperiod stress signaling with AHP2 and AHP3 having a more prominent role in comparison to AHP5.

Loss of ARR10 and ARR12 rescues the photoperiod stress sensitivity of arr2mutants

After phosphorylation by AHPs, type-B ARRs regulate the CK signaling output. Three members of the type-B ARR family - namely ARR2, ARR10 and ARR12 - act in photoperiod stress signaling (Nitschke et al., 2016). However, the analysis was limited to changes in Fv/Fm and the combination of all three mutant alleles was not tested. Hence, we created arr2,10,12 triple mutant plants and exposed them to a PLP treatment along with the corresponding double and single mutants.
Consistent with the findings of Nitschke et al. (2016), the percentage of lesion forming leaves in arr2 plants was increased 2.5-fold compared to wild-type plants after photoperiod stress treatment. In contrast, arr10 , arr12 and arr10,12 mutants did not differ from wild type with respect to lesion formation (Fig. 5A, Suppl. Fig. 1D). Surprisingly, also arr2,10 and arr2,12 plants were indistinguishable from wild type while arr2,10,12 plants were much more sensitive to photoperiod stress. This indicated that ARR2, ARR10 and ARR12 may interact in a complex manner to regulate the response to photoperiod stress. Measurement of the photosynthetic capacity after photoperiod stress treatment confirmed that arr2leaves were more affected after the PLP compared to all other genotypes except for arr2,10,12 , which were even stronger affected (Fig. 5B). At the molecular level, the response of the different arrmutants varied (Fig. 5C-E). The abundance of BAP1 andZAT12 did not give clear indications whether the mutants tested differed in their photoperiod stress response as the majority of differences were not statistically significant (Fig. 5C, D). In contrast, 15 h after the exposure to photoperiod stress CAB2 was less abundant in arr2 and arr2,10,12 in comparison to all other genotypes (Fig. 5E). Consistent with the similar phenotypic response in terms of lesion formation and Fv/Fm, CAB2 expression was lowered to a similar level in all other genotypes and wild type 7.5 h and 15 h after the PLP.
In summary, the results confirmed the results of Nitschke et al. (2016) who reported a positive regulatory function of ARR2 in photoperiod stress. In addition, the results suggested that ARR2, ARR10 and ARR12 interact in a complex manner to regulate the response to photoperiod stress.