Comprehensive Analysis of DWARF14-LIKE2 (DLK2) Reveals Its Functional Divergence from Strigolactone-Related Paralogs

Abstract

Strigolactones (SLs) and related butenolides, originally identified as active seed germination stimulants of parasitic weeds, play important roles in many aspects of plant development. Two members of the D14 α/β hydrolase protein family, DWARF14 (D14) and KARRIKIN INSENSITIVE2 (KAI2) are essential for SL/butenolide signaling. The third member of the family in Arabidopsis, DWARF 14-LIKE2 (DLK2) is structurally very similar to D14 and KAI2, but its function is unknown. We demonstrated that DLK2 does not bind nor hydrolyze natural (+)5-deoxystrigol [(+)5DS], and weakly hydrolyzes non-natural strigolactone (-)5DS. A detailed genetic analysis revealed that DLK2 does not affect SL responses and can regulate seedling photomorphogenesis. DLK2 is upregulated in the dark dependent upon KAI2 and PHYTOCHROME INTERACTING FACTORS (PIFs), indicating that DLK2 might function in light signaling pathways. In addition, unlike its paralog proteins, DLK2 is not subject to rac-GR24-induced degradation, suggesting that DLK2 acts independently of MORE AXILLARY GROWTH2 (MAX2); however, regulation of DLK2 transcription is mostly accomplished through MAX2. In conclusion, these data suggest that DLK2 represents a divergent member of the DWARF14 family.

Keywords: AtD14; DLK2; KAI2; MAX2; butenolide; light; strigolactone.

FIGURE 1

DLK2 shares high structural similarity with KAI2 and AtD14. (A) Alignment of amino acid sequences of AtD14, KAI2 and DLK2 showing that DLK2 shares 40 and 42% identity at the protein level with KAI2 and AtD14, respectively. Dark blue color shows identity in all three proteins; light blue coloring shows identity in two of the three proteins. The amino acids of the catalytic triad are marked with red rectangles and arrows. The residues required for the physical interaction of AtD14 with MAX2 are marked with green rectangles and arrows. (B–D) Crystallized tertiary structures of AtD14 (B, green; PDB code 4IH4), KAI2 (C, blue; PDB code 4IH1) and predicted structure of DLK2 (D, yellow; (I-TASSER server; http://zhanglab.ccmb.med.umich.edu/I-TASSER/). (B,C) The predicted structure of DLK2 (yellow) is overlaid on those of AtD14 and KAI2, respectively, using Swiss-PdbViewer (Guex and Peitsch, 1997). An expanded view of the catalytic triad residues of DLK2 (Ser-102, Asp-223 and His-253) and the predicted cavity are shown in (D).

FIGURE 2

DLK2 stereospecifically binds and hydrolyses (–)5DS. (A) Differential Scanning Fluorimetry (DSF) assay curves of AtD14 and DLK2 proteins (purified from E. coli Rosetta cells) in the presence of (+) and (–)5DS. Protein-ligand mixtures with SYPRO fluorescent dye were gradually heated in a real-time PCR instrument and the change in fluorescence emission was monitored and plotted against temperature. Curve fitting was accomplished with SimpleDSFviewer (Sun et al., 2015). ΔT_m is calculated as a difference from mock (DMSO) T_m; calculated T_m values of three protein samples and four technical replicates from different protein batches are shown; asterisks represent a significant difference from mock (buffer) under the same conditions (mean ± SD; ANOVA, P < 0.01). (B) (+)5DS and (–)5DS (m/Z = 330) show emission peaks at 235 nm. Peak of the putative degradation product is also shown (m/Z = 298). Amounts of deoxystrigols and degradation products were estimated as the areas of peaks. (C) Hydrolysis of (+) and (–)5DS in the presence of AtD14 and DLK2. Hydrolysis of the compounds was assessed by HPLC after 2 h of incubation at 20°C. Data represent the percentages of decrease in the peak areas of substrates. The spontaneous hydrolysis was around 10%. Measurements were repeated at least three times using different protein batches. Asterisks represent significant differences from mock (buffer) under the same conditions (mean ± SD; ANOVA, P < 0.01).

FIGURE 3

Seedling growth responses to low light conditions of the dlk2-3 mutant and its combinations with htl-3 and d14-1 mutants. (A) Seedling phenotypes of 5-day-old dlk2-3, d14-1 and htl-3 mutants and their combinations either untreated or exposed to 10 μM of (+)5DS and (–)5DS. Seeds were sown on 0.5XMS plates with 1% sucrose and supplemented with 10 μM of each compound as indicated. To initiate germination, seeds were dark stratified and were kept in red light (10 μmol m^-2 s^-1; LED) for 10 min. Plates were incubated for 5 d under low light conditions (7 μmol m^-2 s^-1; 21°C). (B,C) Hypocotyl elongation (B) and cotyledon expansion (C) responses of dlk2-3, d14-1 and htl-3 mutants and their combinations to 10 μM of (+)5DS and (–)5DS applications as compared to wild type Col-0 seedlings grown in low light for 5 days. Data are means of 3 independent experiments, >30 seedlings in each. Bars with the same letter are not significantly different from each other (mean ± SD; ANOVA, P < 0.01, Tukey’s HSD test).

FIGURE 4

DLK2 overexpression (OE) lines exhibit elongated hypocotyl response to low light conditions. (A) Hypocotyl elongation responses of low light grown DLK2 OE (Ler) and DLK2 OE (dlk2-2) lines (10 days old). Seeds were germinated and grown on 0.5 × MS plates with 1% sucrose and supplemented with 1, 5, and 10 μM of rac-GR24 applications [DLK2 OE (Ler)] as compared to wild type seedlings. Data are means of 5 independent experiments, >30 seedlings in each. Asterisks represent a significant difference from Ler or Col-0 under the same treatment (mean ± SD; ANOVA, P < 0.025). Real-time PCR was corroborated in all lines with three biological replicates (n = 3, 15 10-day-old seedlings in each; Col-0 was set as calibrator); reactions were performed in quadruplicates. (B) Hypocotyl elongation responses of DLK2 OE (dlk2-3 d14-1 kai2-2) lines. Seeds were germinated and grown on 0.5 × MS plates with 1% sucrose and supplemented with 10 μM of (+)5DS and (–)5DS. Data are means of 3 independent experiments, >30 seedlings in each. Asterisks represent a significant difference from dlk2-3 d14-1 kai2-2 under the same treatment (mean ± SD; ANOVA, P < 0.025). Real-time PCR was corroborated in all lines with three biological replicates (n = 3, 15 10-day-old seedlings in each; Col-0 was set as calibrator); reactions were performed in quadruplicates. (C) Cotyledon expansion responses of DLK2 OE (dlk2-3 d14-1 kai2-2) lines. Seeds were germinated and grown on 0.5 × MS plates with 1% sucrose. Data are means of 3 independent experiments, >30 seedlings in each. Asterisks represent significant differences from dlk2-3 d14-1 kai2-2 under the same treatment (mean ± SD; ANOVA, P < 0.025).

FIGURE 5

DLK2 transcripts are upregulated in the dark through KAI2 and PIFs. Dark adaptation upregulates DLK2 in 5-day-old dark-grown seedlings. Seeds were germinated and grown on 0.5 × MS plates with 1% sucrose. Real-time PCR was conducted on three biological replicates (n = 3, 15 seedlings in each). Asterisks represent significant differences from Col-0 (low light; mean ± SD; ANOVA, P < 0.025). Reactions were performed in quadruplicates.

FIGURE 6

Spatio-temporal regulation of DLK2 promoter activity in DLK2pro:GUS transgenic seedlings and adult plants (bars = 1 mm). (A) GUS histochemical activity in mock and rac-GR24 (10 μM) treated 10-day-old seedlings grown under low light conditions (7 μmol m^-2 s^-1). Insets show close-ups of the basal section of the hypocotyls. (B) GUS histochemical activity in 6 days old dark grown seedlings. (C) GUS histochemical activity in 4 weeks old whole plants. (D) GUS histochemical activity in the aerial parts of 4 weeks old plants. (i) Stem (ii) Hand section of a stem segment adjacent to the first cauline leaf. (iii) Close-up of the basal part of stem. (iv) Close-up of the basal part of the petiole with the stipules and bud. (E) GUS histochemical activity in the roots of 4 weeks old plants. (i) Whole root (ii) Root segment close to the hypocotyl root junction; root in the differentiation zone; root cap. (iii) Hand section of the root in the differentiation zone.

FIGURE 7

DLK2 does not show rac-GR24-specific degradation. (A) Representative captures show that DLK2:sGFP accumulates upon rac-GR24 treatment in the primary leaves of 14 days old DLK2pro:DLK2:sGFP plants. GFP is detected in the cytoplasm and nucleus. (B) DLK2-sGFP expression in guard cells of the primary leaves of 14 days old DLK2pro:DLK2:sGFP plants (bar = 5 μm). (C) Two weeks old seedlings of 35Spro:DLK2:sGFP were treated with 10 μM rac-GR24 for 2, 6, 24 h. Total protein was extracted, run on SDS-PAGE and blotted. DLK2-sGFP fusion protein was visualized using GFP specific antibodies. DLK2-sGFP protein levels show slight increase during rac-GR24 treatment compared to untreated samples.