YUCCA auxin biosynthetic genes are required for Arabidopsis shade avoidance

Plasmids

eDR5::LUC+ is described in Covington & Harmer (2007). The pZP-eDR5::LUC2 plasmid was constructed in two steps. First, the luciferase+ gene in the eDR5::LUC plasmid (Covington & Harmer, 2007) was replaced with the luciferase2 (luc2) gene (from pGL4.10, Promega, Madison, WI) using the HindIII and XbaI sites in the two plasmids. Second, the eDR5::LUC2 cassette was removed from the resulting plasmid using the BamHI and PstI sites and cloned into the BamHI and PstI sites of pPZPXomegaLUC+ (a derivative of pPZP221 (Hajdukiewicz, Svab & Maliga, 1994) that contains the RbcS E9 polyadenylation region). The resulting plasmid confers resistance to spectinomycin in bacteria and gentamycin in plants.

Plant materials and growth conditions

Plant transformations were performed by floral dip as previously described (Clough & Bent, 1998). eDR5::LUC2 transformants were selected on gentamycin-containing growth media. The T-DNA and transposon insertion lines were obtained from the Arabidopsis Biological Resource Center (ABRC), the Cold Spring Harbor Lab (CSHL) or GABI-Kat. Mutant yucca lines and plants carrying YUCCA promoter-GUS constructs were obtained from Yunde Zhao and have been previously described (Chen et al., 2014; Cheng, Dai & Zhao, 2006). Multiple mutant combinations were obtained by repeated crossing and PCR genotyping using described primers (Chen et al., 2014; Cheng, Dai & Zhao, 2006). Homozygous athb-2 mutants were obtained from SALK line_106790 (Alonso et al., 2003; O’Malley & Ecker, 2010). Homozygotes were identified by PCR genotyping using standard techniques and the primers listed in Table 1. A reverse-transcription PCR assay was used to confirm that no wild-type message was made.

For seedling stage EOD-FR analysis, seeds were surface sterilized with 70% ethanol, 0.1% TritonX-100 for 5 min, stratified for four days at 4 °C, then sown on medium containing 1/2X MS with minimal organics (Sigma M6899) and 0.7% agar (Sigma A1296). Seeds were grown in custom chambers outfitted with Quantum Devices Snaplite LEDs under short-day (8 hour day/16 hour night) conditions with 35 μmol m⁻² s⁻¹ “red” (peak wavelength 670 nm, half power spectral bandwidth 655–685 nm) and 5 μmol m⁻² s⁻¹ “blue” (peak wavelength 470 nm, half power spectral bandwidth 455–485 nm). EOD-FR treatment consisted of a 30 min, 14 μmol m⁻² s⁻¹ “far-red” (peak wavelength 730 nm, half power spectral bandwidth 715–745 nm) pulse given nightly for 1 or 4 nights before measurement. LED chamber temperature was 21 °C.

For seedling stage high and low R:FR analysis, seedlings were grown in the same custom chambers as described above for seedling EOD-FR analysis. Light conditions were continuous illumination with 35 μmol m⁻² s⁻¹ “red” (peak wavelength 670 nm, half power spectral bandwidth 655–685 nm) and 5 μmol m⁻² s^{−1 “} blue” (peak wavelength 470 nm, half power spectral bandwidth 455–485 nm). After 24 hours, “far-red” (peak wavelength 730 nm, half power spectral bandwidth 715–745 nm) illumination was added to bring the red-to-far-red ratio (R:FR) to 2. After an additional 48 hours the R:FR ratio in one chamber was lowered to 0.5 and plants were grown for an additional 4 days. The chambers assigned to high and low R:FR were swapped for each trial.

For analysis of juvenile plants under EOD-FR seeds were sown as above but plants were grown under 12/12 or short day (8 hr light:16 hr dark) conditions at 22 °C in a Conviron E7 chamber for approximately 18 days with cool white and incandescent lights (75 μmol m⁻² s⁻¹ PAR, R:FR 1.4). Two days prior to the EOD-FR pulse, plants were transferred to the LED chambers using the same light and temperature conditions as for seedlings (short day 35 μmol m⁻² s ⁻¹ red, 5 μmol m⁻² s ⁻¹ blue light; 21 °C.) and then pulsed as above.

For analysis of juvenile plants under high and low R:FR, stratified seeds were sown on soil and grown under long days in a Conviron walk-in chamber with cool white bulbs and far-red LEDs (Orbitec) (16 h light/8 h night; 100 μmol m ⁻² s⁻¹ PAR, R:FR 1.8, 22 °C). Two week old plants were transferred to shelves in the same chamber with increased FR (100 μmol m⁻² s⁻¹ PAR, R:FR 0.5) to stimulate the shade avoidance response or kept under high R:FR for ten days. Leaves were scanned and petiole length measured as described (Maloof et al., 2013). Plants for were grown under these same high R:FR conditions but were not transferred to low R:FR.

For N −1-naphthylphthalamic acid (NPA; Chem Service, PS-343, http://www.chemservice.com) treatment of eDR5::LUC juvenile plants, seeds were sown and grown as above. 24 hours and 1 hour prior to EOD-FR treatment each plate of plants was sprayed with 1.5ml of DMSO containing 100 μM NPA or an equivalent volume of DMSO alone. Powdered NPA was dissolved in DMSO and stored at −20 °C.

Quantitative RT-PCR

Columbia and athb-2 seedlings were grown as described above for seedling EOD-FR except that they had 30 min EOD-FR pulses on days 3 through 7 and were harvested on day 7, one hour after the end of the final EOD-FR pulse. RNA was prepared with Plant RNeasy (Qiagen) and cDNA prepared with Superscript II (Invitrogen). Real-time qRT-PCR was performed using an iCycler IQ™ 5 (Bio-Rad) in self-made buffer (final concentration: 40 mM Tris-HCL, pH 8.4, 100 mM KCl, 6 mM MgCl2, 8% glycerol, 20 nM fluorescein, 0.4× SYBR Green I (Molecular Probes), 1× bovine serum albumin (New England Biolabs), and 1.6 mM dNTPs) using primers described in Table 1, 10 ng of RNA-equivalent cDNA and Taq polymerase. Each of five to six independent cDNA preparations was assayed two times for each transcript analyzed. Data presented are normalized to the expression level of the control gene PP2a (At1g13320; Czechowski et al., 2005). Transcript abundance was calculated using the relative expression software tool (REST-MCS; (Pfaffl, Horgan & Dempfle, 2002)).

GUS staining

Columbia, YUCCA5::GUS, YUCCA8::GUS and YUCCA9::GUS seeds were grown as described for juvenile plants above. On day 2 in the LED chamber half of the plants were treated with an EOD-FR pulse. Two hours after the pulse plants were taken for GUS analysis. Plants were harvested in 80% acetone on ice and kept in acetone for 30 min. They were then washed twice with pre-staining solution (100 mM NaPO₄, pH 7.0, 0.1% (v/v) Triton X-100, 2 mM potassium ferrocyanide, 2 mM potassium ferricyanide, 1 mM EDTA), after which they were vacuum-infiltrated for 10 min with GUS-infiltration buffer (pre-staining solution + 1 mM X-gluc). Images were taken with a Zeiss Discovery-V12 stereo microscope and AxioCam MRC (Zeiss).

Imaging and analysis

For hypocotyl length measurements, whole seedlings were placed on transparency film and scanned with a flatbed scanner (Microtek ScanMaker 8700, http://www.microtek.com). For luminescence measurements, 24 hours prior to luciferase imaging each plant plate was sprayed with 1.5 ml of 3 mM D-luciferin (Biosynth AG) in 0.1% Triton X-100. Bioluminescence was captured with an XR/Mega-10Z ICCD camera (Stanford Photonics) and Piper Imaging software (Stanford Photonics) or an iKon M-934 CCD camera (Andor) controlled by LabView software (National Instruments). Photo analysis software ImageJ (Rasband, 1997) was used to measure both hypocotyl lengths and bioluminescence. Subsequent data analysis was performed in R (R Core Development Team, 2016) using base packages and the add-on packages ggplot2 (Wickham, 2009), reshape2 (Wickham, 2007), lme4 (Bates et al., 2015), lmerTest (Kuznetsova, Bruun Brockhoff & Haubo bojesen Christensen, 2014), and arm (Gelman & Su, 2014).

Data and scripts

The raw data and scripts to recreate plots are available on GitHub: https://github.com/MaloofLab/Mueller-Moule-PeerJ-2016.

End-of-day far-red treatment rapidly increases auxin responses

It is clear that changes in auxin biosynthesis and sensitivity are critical to shade avoidance responses (Bou-Torrent et al., 2014; Hornitschek et al., 2012; Li et al., 2012; De Wit, Lorrain & Fankhauser, 2014). To examine phytochrome/auxin pathway interactions in real-time we used an enhanced version of the synthetic auxin responsive promoter DR5 (Ulmasov et al., 1997) to drive the expression of firefly luciferase (LUC; Welsh & Kay, 2005), eDR5::LUC (Covington & Harmer, 2007). We initially used an end-of-day far-red (EOD-FR) pulse that, like low R:FR, will reduce the amount of active type II phytochromes, increases expression of auxin responsive genes (Kozuka et al., 2010), and increases stem elongation (Gorton & Briggs, 1980). Plants treated with EOD-FR displayed a strong increase in eDR5::LUC bioluminescence peaking two to three hours after the treatment, consistent with prior reports on eDR5::GUS (Carabelli et al., 2007). This response is found in both seedling stage (Fig. 1A) and juvenile (Fig. 1B) plants and occurred in cotyledons, hypocotyls, petioles, the shoot apex, and developing leaves (Figs. 1D and 1E).

To investigate the importance of auxin transport in eDR5::LUC activation we examined the effect of the auxin transport inhibitor N-1-naphthylphthalamic acid (NPA) on eDR5::LUC expression. Plants grown on NPA still responded with a peak of luminescence following an EOD-FR treatment (Fig. 1C), but in this case the increased bioluminescence was limited to the apex and young leaves (Fig. 1F). The magnitude of induction was somewhat lower on NPA because of higher basal luminescence, however the peak strongly resembles the response of the control plants without NPA (Fig. 1G) and occurs within a similar time-frame. These results suggest that auxin transport is not required to generate the peak of auxin reporter expression following EOD-FR treatment but that transport is required for increased auxin signaling in the petiole. Alternatively, it is possible that the lack of signal in the EOD-FR, NPA treated petioles is due to increased IAA conjugation that can occur in the presence of NPA.

Shade treatment induces expression of four YUCCA auxin biosynthetic genes

Shade treatment is known to lead to increased expression of some YUCCA auxin biosynthetic genes (Hornitschek et al., 2012; Li et al., 2012; Tao et al., 2008), so it seemed possible that the induction of eDR5 could be due to increased YUCCA expression. However, most studies of yucca mutants have not found strong shade avoidance phenotypes. One explanation for the observed weak shade phenotypes might be redundancy within the YUCCA gene family. To determine if this could be the case we asked which YUCCA genes were induced by EOD-FR or shade treatments. We first analyzed a published microarray data set (Sessa et al., 2005) and found that three members of this family, YUCCA5, 8, and 9, were all significantly and rapidly induced by low R:FR (P < 0.002; Fig. 2A), suggesting that they would be interesting targets for further analyses. A fourth member, YUCCA2, was more modestly induced (P < 0.03). All YUCCA genes returned to pre-induction levels after four days, indicating that they are involved in early response to shade conditions. We used quantitative real-time reverse transcription PCR (qRT-PCR) to confirm that YUCCA2, 5, 8, and 9 are induced after a series of EOD-FR treatments. One hour after the last EOD-FR treatment all four genes were significantly induced with mRNA levels up to 10 times higher than in control plants (Fig. 2B), consistent with previous microarray studies (Li et al., 2012; Tao et al., 2008).

YUCCA genes 2, 5, 8, and 9 are expressed in organs responsive to shade-treatment

To determine whether these genes were expressed in tissues relevant to shade avoidance, we examined staining in YUCCA2, 5, 8, or 9 promoter::GUS fusions (Fig. 3). All four genes were expressed in the hypocotyls and leaf veins (Figs. 3E–3L). YUCCA2 was also expressed strongly in the primary root, whereas the other three expressed more weakly in primary roots (Figs. 3M–3P). The YUCCA2 and 5 genes were expressed in the shoot apical meristem (Figs. 3E and 3F) and in very defined locations in the leaf. In the leaf they were highly expressed in the veins, petioles, and hydathodes (Fig. 3G). In the roots YUCCA5 was highly expressed at the branching points between primary and secondary roots (Fig. 3N), similar to reported patterns of eDR5::LUC (Moreno-Risueno et al., 2010) suggesting that it may play a role in defining these patterns. The YUCCA8 and 9 genes were expressed in a more diffuse pattern in the leaves starting from the leaf margins (Figs. 3K and 3L), similar to previously reported patterns of eDR5::GUS and Ptaa1::TAA1::GUS (Tao et al., 2008). They were also expressed in secondary roots (Figs. 3O and 3P) but not in the petioles or the shoot apical meristem. In summary, these genes are expressed in the main organs where shade induction of eDR5::LUC expression is observed: all four are expressed in leaves and YUCCA2 and 5 also in the shoot apex.

AtHB-2 is not required for YUCCA induction

The HD-zip transcription factor AtHB-2 is strongly induced by shade and affects both shade-avoidance traits and auxin-responsive processes (Carabelli et al., 1993; Carabelli et al., 1996; Morelli & Ruberti, 2002; Steindler et al., 1999). We were therefore curious if athb-2 mutations would affect YUCCA induction. However, we found full induction of YUCCA2, 5, 8, and 9 in athb-2 mutants (Fig. 2C). Although not statistically significant the induction appears higher in athb-2 than in wild type, perhaps hinting at a compensatory feedback loop. AtHB-2 may primarily affect auxin transport, as previously proposed (Morelli & Ruberti, 2002) but is not required for YUCCA expression.

YUCCA genes 2, 5, 8, and 9 are required for EOD-FR and low R:FR stimulation of auxin signaling and cell elongation

To determine the relative importance of YUCCA genes for EOD-FR or shade-mediated increases in auxin signaling and subsequent hypocotyl and petiole elongation, we constructed a quadruple mutant with insertions disrupting YUCCA2, 5, 8, and 9 (yucQd) and compared this to yucca5, 8, 9 (yucT) and yucca3, 5, 7 8, 9 (yucQt) mutant strains. The yucT and yucQt strains behaved similarly, partially reducing hypocotyl and petiole EOD-FR responses (Figs. 4A and 4B), similar to previous studies of yucca1, 4 or yucQt lines (Li et al., 2012; Tao et al., 2008; Won et al., 2011). In contrast, the quadruple mutant line completely disrupted EOD-FR in hypocotyls (Fig. 4C) and low R:FR growth responses in petioles (Fig. 4D). In separate experiments we also compared hypocotyl low R:FR response in the yucQd strain to yucca2, yucca5, yucca8, and yucca9 single mutants, a yucca1, 4 double mutant strain, and a yucca2, 5, 9 triple mutant strain (Fig. 5). In this assay all strains were shade responsive except for yucQd (Fig. 5). Across these different experiments the only consistent non-responder to low R:FR and EOD-FR is the yucQd. The difference between the yucQd mutant and the yucT and yucQt combinations is that the yucQd mutant is the only line missing the function of all four of the EOD-FR / low R:FR inducible YUCCA genes. Therefore, this result shows that YUCCA2, 5, 8, and 9 act additively and together are required for the shade avoidance response. In growing the mutant lines for these studies we did not observe any severe morphological defects, although yucQd had reduced fertility (Fig. 6).

The failure of the yucQd mutant to show a morphological shade avoidance response suggested that induction of eDR5::LUC2 by EOD-FR was likely also diminished. To investigate this possibility, the eDR5::LUC2 construct was transformed into the yucQd strain and wild-type plants. We found that EOD-FR induction of eDR5::LUC2 expression was essentially abolished in the yucQd mutant (juvenile plants; Fig. 4E). Thus, YUCCA2, 5, 8, and 9 are required for increased auxin signaling in response to EOD-FR and shade for the subsequent induction of hypocotyl and petiole elongation.