Evidence for species-dependent biosynthetic pathways for converting carlactone to strigolactones in plants

Strigolactones, a novel class of phytohormones, regulate plant architecture and act as rhizosphere signals. Species-specific biosynthetic pathways convert a common precursor, carlactone, to strigolactones.


Expression of SbCCD7 and SbCCD8 and preparation of carlactone (1)
Escherichia coli cells of strain BL21 were transformed with pGEX-SbCCD7 and pGEX-SbCCD8 constructs. Codons unfavorable for E. coli were replaced by modifying the nucleotide sequences of cDNAs in order to enrich the A/T residues without changing the amino acid sequence. The transformed cells were grown overnight in Luria-Bertani (LB) medium supplemented with 100 μg ml −1 ampicillin and incubated in 250 ml modified LB medium containing the same concentration of ampicillin. The medium, incubated at 37 °C with continuous shaking at 200 rpm for 3 h, was subsequently supplemented with 0.1 mM isopropyl β-D-thiogalactoside. The culture was subsequently incubated at 18 °C with continuous shaking at 150 rpm for 24 h, and E. coli cells were harvested by centrifugation at 13 420 g for 3 min. The resulting spheroplasts, sonicated in 200 mM HEPES-NaOH (pH 7.8), were subjected to centrifugation at 13 420 g for 3 min. The supernatant was subjected to enzyme reactions as described by Alder et al. (2012). A reaction product exhibiting identical chromatographic behavior to an authentic sample of CL was purified using semi-preparative HPLC. The CD spectrum showed positive maxima at 218 and 266 nm and matched that reported for 11R-carlactone (Seto et al., 2014).

Feeding of SL precursors
Sorghum seedlings were transferred to test tubes and grown hydroponically in Long Ashton nutrient solution (two-fifths strength) in a growth chamber at 28 °C, with a 16 h light/8 h dark photoperiod. Cotton seedlings were transferred to test tubes and grown hydroponically in Hoagland nutrient solution (one-fourth strength) in a growth chamber under the same temperature and photoperiod conditions as those for the sorghum seedlings. Cowpea and sunflower seedlings were transferred to test tubes and grown hydroponically in Hoagland nutrient solution (one-fourth strength) in a growth chamber at 23 °C and a 16 h light/8 h dark photoperiod. After 2 weeks, the culture medium was replaced by tap water supplemented with 1 µM fluridone, an inhibitor of phytoene desaturase, a key enzyme involved in carotenoid biosynthesis. Three days later, plants were transferred to test tubes containing tap water (50 ml) with any one of the SL precursors CL (1) (final concentration, 0.18 µM), CLA (2) (0.1 µM), 4-DO (4) (0.012 µM), and 5-DS (9) (0.012 µM) and incubated for 24 h before filtration. Each aquaculture filtrate was supplemented with epi-strigol (50 pmol) as an internal standard and then extracted with EtOAc (15 ml×3). The combined organic layer, washed with 0.2 M K 2 HPO 4 (pH 8.3) and dried over Na 2 SO 4 , was concentrated in vacuo. Each crude EtOAc extract was subjected to column chromatography on silica gel (0.4 g). Strigolactones were eluted with 30% EtOAc in CHCl 3 . Each eluate, dried in vacuo, was re-dissolved in 100 μl acetonitrile. For sunflower, owing to low production of heliolactone, three plantlets were transferred to separate plastic cups, each containing 200 ml tap water. Fluoridone (1 µM) was added to each cup and the aquaculture, supplemented with any of the plausible biosynthetic precursors as described above, was incubated for 24 h and filtered. Each filtrate was extracted with EtOAc (50 ml×3), and SLs semi-purified as above were dissolved in 100 μl acetonitrile. An aliquot (5 μl) of each resulting solution was subjected to LC-MS/MS analysis.
Moonseed roots, transferred 50 d after sub-culturing to 25 ml of fresh B5 medium deficient in phosphorus and supplemented with fluridone at 1 μM, were incubated at 26 °C for 24 h before replacement of the medium with fresh phosphorus-deficient B5 medium supplemented with fluridone (1 μM) and CL (1), CLA (2), or 5-DS (9). The cultures were re-incubated for 24 h, and the SLs were extracted, semi-purified, and analysed as described above.

Expression of AtCYP711A1
Escherichia coli cells, strain DH5α, were transformed with a pCW-AtCYP711A1 construct. Codons unfavorable for E. coli were replaced by modifying the nucleotide sequence of the cDNA in order to enrich A/T residues without changing the amino acid sequence. The transformed cells, grown overnight in LB medium supplemented with 100 μg ml −1 ampicillin, were subsequently incubated in 50 ml modified Terrific Broth medium containing 100 μg ml −1 ampicillin, 0.2% w/v glucose, and 0.5 mM γ-aminolevulinic acid. The medium was incubated at 37 °C with continuous shaking at 225 rpm for 2.5 h and subsequently supplemented with isopropyl β-D-thiogalactoside (0.1 mM) and chloramphenicol (1 μg ml −1 ). The culture was incubated at 25 °C with continuous shaking at 150 rpm for 72 h before E. coli cells were harvested by centrifugation at 2330×g for 20 min. The resulting spheroplasts, sonicated in a buffer solution containing 50 mM potassium phosphate (pH 7.25), 20% (w/v) glycerol, 1 mM EDTA, and 0.1 mM dithiothreitol, were subjected to centrifugation at 100 000 g for 1 h. The pellets were sonicated in the same buffer, and the resulting microsomal fractions were subjected to enzyme reactions. The reaction mixture (250 µl), containing 50 mM potassium phosphate buffer (pH 7.25), 30 µM CL, 2.5 mM NADPH, NADPH-cytochrome P450 reductase, and 25 µl microsomal fractions, was incubated at 30 °C for 1 h. The reaction was terminated by adding 25 µl 2 M HCl. The mixture was extracted with EtOAc (250 µl), followed by evaporation of the solvent in vacuo. The residue, dissolved in 50 μl acetonitrile, was subjected to LC-MS/MS analysis.
SL analyses SL fractions obtained as described above from root exudates were analysed using LC-MS/MS to evaluate the SLs quantitatively and qualitatively. The ultra-performance liquid chromatograph parameters were as follows: column: 100 × 2.0 mm, 2.5 μm, COSMOSILpacked column 2.5C 18 -MS-II, 30 °C; solvent: 50-100% MeOH in H 2 O (0-20 min, linear gradient); flow rate: 0.2 ml min −1 . The mass spectrometer was operated in the positive electrospray ionization (ESI) mode for CL, 19-OH-CL, sorgomol, strigol, strigyl acetate, orobanchol, orobanchyl acetate (alectrol), 5-DS, 4-DO, heliolactone, and MeCLA. The capillary voltage was 3 kV, source temperature was 120 °C, and desolvation gas temperature was 350 °C. The nebulizer and desolvation N 2 gas flow rates were 50 and 600 l h −1 , respectively. Fragmentation was performed using collision-induced dissociation with argon at 0.1 ml min −1 . CLA was measured in the negative ESI mode. Multiple reaction monitoring (MRM) was used to detect the presence of precursors and SLs. The MRM transitions, set according to the mass spectra obtained for authentic samples, are summarized in Table 1. epi-Strigol, added as an internal standard, was detected under the same conditions as those for strigol. Metabolites derived from CL in moonseed were probed by performing a parent ion scan on the fragment ion at m/z 97, which corresponds to the D-ring, over the range of 50-400 m/z, with a cone voltage of 30 V and collision energy of 20 eV. Three products were detected, one of which exhibited major ions at m/z 319 and m/z 301.

SL biosynthesis in sorghum
The LC-MS/MS analysis results confirmed the SL profile previously reported for the aquaculture filtrate of sorghum cultivar Sudax (Motonami et al., 2013). In the culture filtrate, 5-DS (9) and sorgomol (10) were minor and major SLs, respectively ( Fig. 2A). Fluridone restricted the production of both SLs to negligible levels (Fig. 2B). Concurrent application of CL (1) to the aquaculture partially restored the accumulation of 5-DS (9) and sorgomol (10), and the recovery of the exogenously applied CL was below the detection limit (Fig. 2C). In the same experiment, CLA (2) was detected, although its signal was weak (Fig. 3A). Peaks of 5-DS (9) and sorgomol (10) were also detected when CLA (2) was applied to the culture (Fig. 2D). The efficient conversion of 5-DS (9) to sorgomol (10) (Fig. 2E) was consistent with the findings of our previous study (Motonami et al., 2013). Numbers in bold for the strigolactones refer to Fig. 1. IS, internal standard.

SL biosynthesis in cotton
Treatment of the cotton cultivars Wawata and Tonko with fluridone reduced SL accumulation to negligible levels ( Supplementary Fig. S1B at JXB online and Fig. 4B).
(2) (Fig. 3D). The culture converted both CL (1) and CLA (2) to strigol (7) (Fig. 5C, D). 5-DS (9) was not detected in either culture. However, a small signal of strigol was detected after the culture was fed with 5-DS (9) (Fig. 5E). In a separate experiment, feeding the culture with [6ʹ-D]5-DS confirmed that the labeled substrate with a deuterium atom was not converted to strigol (6) (data not shown). Accordingly, the signal of strigol detected in Fig. 5E was not owing to the conversion of exogenously applied 5-DS, but an endogenous product that escaped from the biosynthesis inhibition by fluridone (Fig. 5B).
Feeding moonseed root culture with CL (1) at 10-fold concentration, in addition to strigol (retention time (Rt) 7.0 min), resulted in major peaks at Rt 8.3 min, 9.6 min, and 10.1 min (Fig. 6A). The ESI-MS analyses indicated that the unidentified peaks are associated with molecular masses of 316 Da (Fig. 6C), 318 Da (Fig. 6D), and 316 Da (Fig. 6E), respectively. The peak at Rt 9.6 min showed a fragmentation pattern similar to that of strigol (Fig. 6B) and a fragment ion of m/z 301 (Fig. 6D), corresponding to [M + H − H 2 O], suggesting a hydroxylated product of CL (1). Ions corresponding to dehydrated fragments were not observed in compounds eluted at Rt 8.3 min (Fig. 6C) and Rt 10.1 min (Fig. 6E). Abe et al. (2014) reported that 19-hydroxy CL was detected along with CLA (2) in the enzyme reaction mixture of recombinant AtCYP711A1 protein and CL (1) as a substrate. Our experiments confirmed their findings (see Supplementary Fig. S2). The chromatographic behavior of the unidentified product eluted at Rt 9.6 min (Fig. 6A) differed from that of the 19-hydroxy CL eluted at Rt 13.0 min ( Supplementary Fig. S2C). The possibility that the product was 18-hydroxy CL cannot be ruled out because its chromatographic behavior has not been reported elsewhere, and no authentic sample is available for comparison. However, considering that strigol (7) is the only SL identified in the moonseed culture and has a hydroxy group at C-5 (corresponding to C-4 in the carlactone skeleton), the unidentified metabolite eluted at Rt 9.6 min would probably be 4-hydroxy CL. The fact that 5-DS (9) was below the detection limit (Fig.  5A), that exogenously applied 5-DS (9) was not converted to strigol (7), but metabolized or decomposed (Fig. 5E), and that CL (1) might be hydroxylated at C-4 ( Fig. 6) are evidence that 5-DS (9) is not a biosynthetic precursor of strigol (7) in moonseed.

The common and species-specific pathways of SL biosynthesis
In all plant cultures tested in this study, SL production was effectively restricted by fluridone, a phytoene desaturase inhibitor, confirming that both canonical and non-canonical SLs are derived from carotenoids. The conversion of CL (1) to CLA (2) was found to be a common pathway (Fig. 3). CLA (2) is converted to canonical or non-canonical SLs in all plants. However, the exact type of SL produced varies across plant species.
Only limited literature is available on BC-ring formation associated with canonical SLs. Zhang et al. (2014) reported that 4-DO (4) is formed from CLA (2) through multistep oxidation reactions catalysed by CLA oxidase. This enzyme is encoded by one of five rice MAX1 homologs, namely Os01g0700900. The authors also reported that 4-DO (4) was converted to orobanchol (5) by orobanchol synthase, which is encoded by another rice MAX1 homolog, Os01g0701400 (Zhang et al., 2014). According to the proposed scheme for the bioconversion of CL into SLs (Zwanenburg et al., 2016), C-18 and C-19 of CL are oxidized to alcohol and carboxylic acid, respectively. Proton transfer from the carboxylic acid to the alcohol provides a species that is prone to undergoing a concerted cyclization to form the BC-ring. The present study revealed that deoxy SL is not always an intermediate in oxygenated SL biosynthesis, and the conversion routes of CLA (2) to SLs are different even among plants that produce the same oxygenated SL as revealed in moonseed and the cotton cultivar Tonko, both of which produce strigol (7).

Co-production of deoxy SL and its corresponding oxygenated SL
5-DS has been detected in sorghum ( Fig. 2A) and cotton (Fig. 4A), and 4-DO in rice and tobacco (Xie et al., 2013). Nevertheless, deoxy SL and its corresponding oxygenated SL is not always co-produced. Moonseed produces strigol, but only undetectable amounts of 5-DS (Fig. 5A), whereas cowpea produces large amounts of orobanchol and alectrol, but negligible quantities of 4-DO (Fig. 7A). Conversion of CLA (2) to 5-DS (9) or 5-DS to strigol (7) was not detected in the moonseed culture. In cowpea, the conversion of CLA (2) to 4-DO (4) and 4-DO to orobanchol (5) was not found. In plant species in which deoxy SLs did not participate in oxygenated SL biosynthesis, the recovery of exogenously applied deoxy SLs was very poor. No exogenously applied 5-DS (9) was detected in moonseed (Fig. 5E), and only a limited amount of 4-DO Fig. 9. Plant species-dependent biosynthetic pathways for the conversion of carlactone to strigolactones. Solid arrows indicate the conversion steps confirmed in this study. Dashed arrows indicate the established pathways in rice reported by Zhang et al. (2014). *Intermediacy of the compound is not confirmed.
(4) was found in cowpea (Fig. 7E). Their conversion to strigol (7) or orobanchol (5) was not detected. Conversely, significant amounts of unconverted 5-DS (9) accumulated in sorghum and in both cotton cultivars (Figs 2E and 4E, Supplementary  Fig. S1E). These plants converted the substrate to strigol (7) or sorgomol (10). The results suggest that SLs irrelevant to endogenous SL biosynthesis are recognized and treated by plants as xenobiotics, which might be decomposed, converted to extraction-recalcitrant forms, or retained in the plants. Thus, co-production is an indication that deoxy SLs are converted to their corresponding oxygenated SLs in plants.

Strigol biosynthesis in moonseed and orobanchol biosynthesis in cowpea
The conversions of CL (1) and CLA (2) to strigol (7) (Fig. 5C, D) as well as that of CL to CLA (Fig. 3D) and putative 4-hydroxy CL (Fig. 6) suggest that a metabolic grid exists in moonseed that uses 4-hydroxy CLA as an immediate precursor of strigol (7). Excess amounts of CL and/or its hydroxylated product may be recognized as xenobiotics and subsequently catabolized to yield the products eluting at Rt 8.3 min and Rt 10.1 min (Fig. 6A). The irrelevance of 5-DS in strigol biosynthesis in moonseed is rational if hydroxylation at C-4 precedes BC-ring formation. However, no possible intermediate was detected between CLA (2) and orobanchol (5) in cowpea root exudates. Notably, gold-catalysed cyclization of cis-2,4-dien-1-als generates the corresponding cyclic compounds (Lin et al., 2007). If the C-18 of CLA is oxidized in two steps to form an aldehyde, and then the aldehyde is subjected to a similar cyclization procedure via a catalyst, orobanchol isomers could be generated. This reaction mechanism rationalizes how CLA (2) is converted to orobanchol (5) without passing through 4-DO (4), although whether elaborate machinery exists, such as a dirigent protein enabling stereo-controlled BC-ring closure, needs to be determined.

Heliolactone biosynthesis in sunflower
The bioconversion of CL (1) and CLA (2) to heliolactone (6) was unambiguously demonstrated (Fig. 8). Heliolactone (6) is a non-canonical SL having a hydroxylated ε-ring, whereas CL (1) and CLA (2) have the same β-ring as β-carotene. Accordingly, the conversion of CL (1) to heliolactone (6) requires double bond migration and hydroxylation at C-3, followed by oxidation to ketone in the A-ring. However, the participation of precursors such as α-carotene, lutein, and zeaxanthin derived from other plant carotenoids, cannot be ruled out. These carotenoids have a β-ring and an ε-ring, a hydroxylated ε-ring and a hydroxylated β-ring, and hydroxylated β-rings, respectively. The isomerization of α-carotene to its 9-cis isomer by rice D27 (Bruno and Al-Babili, 2016) and the oxidative cleavage of 9-cis-lutein and 9-cis-zeaxanthin by Arabidopsis and pea CCD7 (Bruno et al., 2014) support the involvement of these carotenoids in heliolactone biosynthesis. A previous study (Ueno et al., 2014) could not provide the absolute configuration at C-11 of heliolactone (6) because the Cotton effect around 250 nm, which indicates the configuration at C-2ʹ of SLs (Welzel et al. 1999), was hidden by the superimposition of the curve derived from the cyclohexenone/dienoate chromophoric pair. CL (1) used in this experiment was prepared by incubating 9-cis-β-carotene with an enzyme mixture consisting of SbCCD7 and SbCCD8. The CD spectrum of the enzymatically prepared CL indicated its C11R configuration, the same as that prepared by OsCCD8 (Zhang et al., 2014). Accordingly, the conversion of CL (1) to heliolactone (6) (Fig. 8C) provides evidence for the C11R configuration of the latter molecule.
The results obtained in this study are summarized in Fig. 9. In conclusion, the conversion of CL (1) to CLA (2) was a reaction common to all plants investigated irrespective of whether they produced canonical or non-canonical SLs. 5-DS (9) was not a precursor of strigol (7) in moonseed, and 4-DO (4) was not a precursor of orobanchol (5) in cowpea. These results indicate that deoxy SLs are not necessarily the precursors of hydroxy SLs, although the conversion of 5-DS (9) to sorgomol (10) and 4-DO (4) to orobanchol (5) was established in sorghum (Motonami et al., 2013) and rice (Zhang et al., 2014), respectively. The involvement of deoxy SLs in the biosynthesis of oxygenated SLs is consistent with the detection of deoxy SLs in root exudates. Elucidation of BC-ring formation mechanisms in plants other than rice might reveal the entire canonical SL biosynthesis pathway. The conversion of CLA (2) to the non-canonical SLs MeCLA (3) and heliolactone (6) was detected in sunflower. Further studies are warranted to comprehensively analyse the biosynthetic pathways of canonical and non-canonical SLs.

Supplementary data
Supplementary data are available at JXB online. Fig. S1. Conversion of CL to CLA and 19-hydroxy CL by AtCYP711A1. Fig. S2. Conversion of CL, CLA, and 5-DS to sorgomol in Wawata cotton. Fig. S3. Conversion of CLA to MeCLA in sunflower.