Seed germination of Striga spp. (witchweeds), one of the world's most destructive parasitic weeds, cannot be induced by light but is specifically induced by strigolactones. It is not known whether Striga uses the same components for strigolactone signaling as host plants, whether it has endogenous strigolactone biosynthesis and whether there is post-germination strigolactone signaling in Striga.
Strigolactones could not be detected in in vitro grown Striga, while for host-grown Striga, the strigolactone profile is dominated by a subset of the strigolactones present in the host.
Branching of in vitro grown Striga is affected by strigolactone biosynthesis inhibitors. ShMAX2, the Striga ortholog of Arabidopsis MORE AXILLARY BRANCHING 2 (AtMAX2) – which mediates strigolactone signaling – complements several of the Arabidopsis max2-1 phenotypes, including the root and shoot phenotype, the High Irradiance Response and the response to strigolactones. Seed germination of max2-1 complemented with ShMAX2 showed no complementation of the Very Low Fluence Response phenotype of max2-1.
Results provide indirect evidence for ShMAX2 functions in Striga. A putative role of ShMAX2 in strigolactone-dependent seed germination of Striga is discussed.
Striga species (witchweeds) are among the world's most destructive weeds as they parasitize important crop species and reduce yields by 30–90% (Scholes & Press, 2008; and references there in). More than 50 million hectares of the arable land used for cereal and legume production in sub-Saharan Africa are infested with at least one Striga species, resulting in annual losses of more than US$10 billion (Westwood et al., 2010, and refenences there in).
Striga species have evolved a life strategy in which germination is delayed until a potential host is sensed through the presence of strigolactones in the host root exudate (Bouwmeester et al., 2003; Matusova et al., 2004). After germination, the Striga radicle will grow towards and then invade the host root through a parasitic structure called a haustorium (Kuijt, 1969).
Strigolactones also have an endogenous signaling role in plants acting as a plant hormone controlling shoot branching/tillering by inhibiting the outgrowth of axillary lateral buds (Stirnberg et al., 2002; Booker et al., 2004; Gomez-Roldan et al., 2008; Umehara et al., 2008). Strigolactones also affect different developmental processes like primary root, lateral root and root hair growth (Kapulnik et al., 2011a,b; Ruyter-Spira et al., 2011; Mayzlish-Gati et al., 2012).
The effect of strigolactones on branching requires the F-box protein MAX2 (Stirnberg et al., 2002; Gomez-Roldan et al., 2008; Umehara et al., 2008). MAX2 also has a role in seed germination and seedling de-etiolation (Shen et al., 2012; Toh et al., 2012). However, the role of strigolactones in these MAX2-related phenotypes is disputed, mainly because strigolactone biosynthesis mutants often lack these phenotypes (Shen et al., 2012). At least some of the strigolactone-induced phenotypes in nonparasitic plants require the presence of an LRR F-box protein (MAX2 in Arabidopsis) and a member of the α/β-hydrolase superfamily (e.g. OsD14 in rice, AtD14 in Arabidopsis, DAD2 in petunia) which has been identified as a second strigolactone signaling component (Ishikawa et al., 2005; Stirnberg et al., 2007; Arite et al., 2009; Gao et al., 2009; Liu et al., 2009; Gaiji et al., 2012; Hamiaux et al., 2012; Waters et al., 2012). Some α/β-hydrolase family members are also involved in signaling by karrikins, which are smoke-derived compounds with a structure that resembles the active part of strigolactones and are potent seed germination stimulants (Nelson et al., 2009, 2012; Waters et al., 2012). Recently, DAD2, the D14 homolog from petunia was shown to bind to and cleave the synthetic strigolactone GR24 and the presence of GR24 increases DAD2 binding to PhMAX2A (Hamiaux et al., 2012).
Strigolactones are derived from all-trans-β-carotene through isomerization, cleavage and several (putative) modification reactions (Matusova et al., 2005; Alder et al., 2012; Fig. 1). The presence of strigolactones is conserved in the plant kingdom (Delaux et al., 2012; Waters et al., 2012; Brewer et al., 2013), but endogenous strigolactones, strigolactone biosynthesis genes or strigolactone signaling components have not been reported in parasitic plants. Only a putative CCD7 was reported from Phelipanche ramosa recently (Péron et al., 2012). The dependence of parasitic plant seed germination on exogenous strigolactones suggests the absence of endogenous strigolactones or a malfunctioning of strigolactone-related signaling in their seeds. Here we present evidence for endogenous strigolactone biosynthesis and signaling during Striga plant development and we focus on characterization of a homolog of MAX2 in Striga (ShMAX2). Because of the lack of mutants and problems with transformation of Striga, functional characterization of the ShMAX2 was achieved by overexpression in an Arabidopsis thaliana max2-1 mutant background and characterization of the complementation of several different max2-1 phenotypes. Results suggest that ShMAX2 can function in Arabidopsis as most max2-1 related phenotypes are fully complemented by ShMAX2, except for the VLFR seed germination phenotype.
Materials and Methods
Plants and growing conditions
Arabidopsis thaliana (L.) Heynh seeds of max2-1 (Stirnberg et al., 2002) and the parental Columbia-0 wild-type were kindly provided by Prof. Ottoline Leyser (University of Cambridge, UK). Striga (Striga hermonthica Del. Benth) seeds were collected from a maize field in Kibos, Kenya, as described by Matusova et al. (2005).
Arabidopsis plants were grown on sterile plates and then on soil in 16-h photoperiods as described by Bennett et al. (2006). Other conditions used are outlined in the paragraphs below.
Inhibitor treatment of in vitro grown Striga hermonthica
Three in vitro grown S. hermonthica explants were grown in a plastic jar with half strength MS medium (containing 20 g l−1 sucrose) as a pool. For carotenoid biosynthesis inhibitor treatments, fluridone and D2 (dissolved in acetone) were added to the medium to reach the desired concentration. An equal amount of acetone was added to all treatments and the controls. For each treatment, four replicates were used. After 3 wk in a climate chamber (25°C, 16 h day length), the branch numbers were scored.
Strigolactone extraction from Striga hermonthica
Strigolactones were extracted from in vitro grown Striga (see Methods S1) and from in vivo grown Striga, grown on rice (Oryza sativa), maize (Zea mays) and sorghum (Sorghum bicolor). Striga underground shoot and root were carefully removed from the host with all host material excluded. For strigolactone analysis, 500 mg of Striga material was extracted with ethyl acetate and purified with Silica gel Grase Pure SPE (200 mg 3 m−1) columns as described by Kohlen et al. (2011) or extracts were fractionated by HPLC for Phelipanche ramosa seed germination bioassays (see Supporting Information Methods S1 and S2).
Gene expression analysis by real time PCR
Gene expression analysis for ShCCD7, ShCCD8 and ShMAX2 was done by real time PCR with total RNA isolated from Striga tissues (see Methods S3).
Isolation of ShMAX2
Genomic DNA was isolated from Striga shoot material using the Plant Genomic DNA miniprep Kit (Sigma-Aldrich) according to manufacturers' protocol. Degenerate primers were designed based on the protein sequences of Arabidopsis MAX2, pea RMS4 and Vitis vinifera and used to amplify fragments of the ShMAX2 gene from DNA. Then Genome-walking (Kilstrup & Kristiansen, 2000) was used to obtain additional sequence information. DNA sequence analysis showed that the ShMAX2 does not contain any introns and the full-length ShMAX2 coding sequence was amplified for cloning into an expression vector. For primer sequences and details of cloning into expression vector and plant transformation see Methods S4. Alignment of predicted amino acid sequences of ShMAX2 compared with other reported MAX2 was performed by CLC workbench (Qiagen, USA).
The Arabidopsis decapitation assay was performed according to the description of Stirnberg et al. (2007) with modifications. Following imbibition in water for 3 d at 4°C, seeds were sown onto soil in 49 cm2 pots (four seeds per pot later thinned to one). They were first grown at 20°C : 18°C in an 8 h : 16 h, light : dark photoperiod at a light intensity of 120 μmol cm−2 s−1. After 30 d, the plants were moved to a 12 h light photoperiod (same light intensity) for 1 wk and then to a 16 h light photoperiods to induce flowering. To encourage branching from the rosette, the primary inflorescence of each plant was removed when it was 2–4 cm long. Rosette branches with a length of at least 2 cm were counted 10 d later. Sample size varied between 7 and 30 plants per treatment between the different experiments.
Seed germination assays
Plates with germinating seeds were imaged by a digital camera (Nikon D80 with Nikkor AF-S 60 mm f/2.8 G Micro ED; Nikon, http://www.nikon.com). The experiment was carried out in an air-conditioned room (20°C). Image processing and data analysis were performed with the GERMINATOR package (Joosen et al., 2010). Germination was defined by emergence of the radicle from seeds. All germination assays were tested with two independent transgenic lines.
Seed germination rate under continuous white light
Six-well (12.7 × 8.5 cm) plates (Greiner bio-one, Frickenhausen, Germany) were used for germination experiments according to the description of Joosen et al. (2010) with modifications. One layer of white filter paper and one layer of blue filter paper (Anchor Paper Company, http://www.seedpaper.com) were placed in each well and 1 ml of water added. Approximately 50–80 mature Arabidopsis seeds were dispersed on the filter paper using a mask to ensure an accurate and reproducible spacing. Clustering of seeds was prevented as much as possible. Then the plates were sealed by parafilm (Pechiney Plastic Packaging Company, Chicago, IL, USA). A maximum of 20 plates were stacked together. On top of the stack, three empty plates were wrapped in aluminum foil to ensure equal distribution of light to other plates. The whole stack was wrapped in a closed transparent plastic bag and placed in a 20°C incubator (type 5042; Seed processing Holland, http://www.seedprocessing.nl).
Seed germination percentage after FR or R light pulse treatment
After being stratified at 4°C for 4 d, seeds were surface-sterilized and plated on ½MS growth medium (0.5 × MS slats supplemented with ×1 Gamborg's B5 vitamin mix, 0.8% (w/v) agar (Daishin; Duchefa Biochemie BV, Haarlem, the Netherlands), without sucrose at pH 5.8) (50–100 seeds per genotype). Then they were placed under dark conditions, or exposed to 5 min of far-red (FR) light or red (R) light before returning to the dark for 2–3 d at room temperature.
Seed dark germination percentage in response to GR24
For GR24 treatment, seeds were treated as described above and then plated on ½ MS growth medium with minimal or various concentrations of GR24 as indicated.
Hypocotyl elongation assay
Before sowing on MS plates, seeds were surface-sterilized in a container containing 50 ml bleach and 1.5 ml concentrated HCl for 4 h and then transferred to a flow cabinet for 1 h. Seeds were imbibed on wet filter paper at 4°C for 3 d and plated on ½MS agar plates. Plants were grown on near vertical plates in a climate chamber under a 22°C : 18°C 16 h-light : 8 h-dark regime (80 μmol m−2 s−1). Image capturing was done by a digital camera (Nikon D80 with Nikkor AF-S 60 mm f/2.8 G Micro ED; Nikon, http://www.nikon.com). Hypocotyl length was measured with ImageJ from digitally computed images of 5-d-old seedlings.
Sequence data from this article can be found in the GenBank data libraries under the following accession numbers: JX565467 for ShMAX2, NM_129823 for AtMAX2, DQ403159 for PsRMS4, Q5VMP0 for OsD3, XP_003607592 for putative MtORE9, and XP_002320412 for putative PtMAX2.
Evidence for endogenous strigolactone production in Striga
Strigolactones are not detected but germination stimulants are present in in vitro grown Striga
In order to determine if Striga produces strigolactones, seeds were stimulated to germinate with GR24 in vitro and the resulting shoots were sub-cultured in vitro at least three times on medium without GR24. Subsequently, shoot tissue was extracted for analysis of strigolactones. No signal was detected for any of the known strigolactones using LC-MRM-MS/MS. The extracts of in vitro grown Striga were assessed for germination stimulant bioactivity in a Phelipanche ramosa seed germination bioassay (Kohlen et al., 2011). Extracts were fractionated by HPLC into 30 fractions and several fractions were able to stimulate P. ramosa seed germination (Fig. 2). Many of these fractions coincide with the elution position of known strigolactones or active fractions of nonparasitic plant root exudates.
Host-derived strigolactones are detectable in extracts of host attached Striga plants
In order to determine if strigolactones are present in Striga, Striga plants attached to a host, rice (Oryza sativa), maize (Zea mays) or sorghum (Sorghum bicolor), were extracted for strigolactone analysis using LC-MRM-MS/MS. For comparison, the roots of the host plants were also extracted and analyzed for the presence of strigolactones. In the extract of Striga attached to rice, only the strigolactone 2′-epi-5-deoxystrigol was identified, while in rice root extracts the strigolactones 2′-epi-5-deoxystrigol, orobanchol and two putative methoxy-5-deoxystrigol isomers were identified (Table 1, Fig. S1a,d). In Striga grown on sorghum, 5-deoxystrigol and sorgomol were identified, while 5-deoxystrigol, sorgomol, and orobanchol were detected in the corresponding sorghum root extracts (Table 1, Fig. S1b,d). A number of unknown strigolactones, which previously have been reported from maize (Jamil et al., 2012), were detected in Striga grown on maize. Again, these unknown strigolactones were a subset of the strigolactones present in the maize root extract (Table 1, Fig. S1c,d). Overall the results show that different types of strigolactones are detected in host-attached Striga and that, in all cases, these strigolactones represent a subset of the strigolactones present in the host root, suggesting that strigolactones are transported from the host to Striga.
Table 1. Strigolactones identified from in vivo grown Striga plants and its host by LC- MRM-MS/MS
Nevertheless, the bioassays suggested that Striga can also produce strigolactones itself. In the Striga EST sequences (http://ppgp.huck.psu.edu/; M. Das et al., unpublished) homologs of the strigolactone biosynthesis genes AtMAX3 (CCD7) and AtMAX4 (CCD8) were identified that were named ShCCD7 and ShCCD8, respectively. The expression of both ShCCD7 and ShCCD8 genes in different tissues and stages of Striga was quantified by real time RT-PCR (Fig. 3). In general, ShCCD8 expression was higher than that of ShCCD7 but their expression patterns were otherwise similar in that most stages were statistically indistinguishable (Fig. 3a,b). Preconditioning of Striga seeds is required to make them responsive to strigolactones (Matusova et al., 2004). Surprisingly, expression of ShCCD7was transiently upregulated during seed conditioning, suggesting endogenous strigolactone biosynthesis in Striga seeds at this stage, while at the same time seeds become strictly responsive to exogenous strigolactones for induction of germination.
Evidence for strigolactone signaling in Striga
Strigolactones have been shown to suppress branching in plants. As strigolactones are derived from carotenoids, branching can be induced by inhibition of strigolactone biosynthesis with the carotenoid biosynthesis inhibitor fluridone or a more specific inhibitor D2 (Sergeant et al., 2009). In order to get further confirmation of the presence of strigolactones in parasitic plants and to test whether they have a similar function in parasitic plant development, in vitro grown Striga plants were treated with different concentrations of fluridone and D2. Treatment with both fluridone (Fig. 4a,b) and D2 (Fig. 4c) resulted in enhanced branching in Striga, showing that also in Striga endogenous strigolactone biosynthesis and strigolactone signaling are involved in the regulation of branching.
MAX2 homolog expression in Striga
One of the proteins required for strigolactone signaling is MAX2 (Stirnberg et al., 2002; Gomez-Roldan et al., 2008; Umehara et al., 2008; Hamiaux et al., 2012). AtMAX2 encodes an F-box leucine-rich repeat protein, which probably functions as the substrate-recognizing subunit of SCF-type ubiquitin E3 ligase in protein ubiquitination (Stirnberg et al., 2007). We isolated the Striga homolog of AtMAX2, ShMAX2. Comparison of the ShMAX2 amino acid sequence with MAX2 orthologs from Arabidopsis, pea and rice shows an N-terminal extension of 25 amino acids in ShMAX2 compared with the longest MAX2 from nonparasitic plants (Fig. S2). The expression of ShMAX2 was analyzed by real time RT-PCR in Striga seeds. Because Striga seeds need to perceive exogenous strigolactones to trigger germination (Matusova et al., 2004), the ShMAX2 could be involved in this specific signaling event. Indeed, ShMAX2 expression increased upon preconditioning at 30°C (Fig. 3c), which correlates with an increase in strigolactone sensitivity during preconditioning (Matusova et al., 2004). This pattern of expression is different from AtMAX2 in Arabidopsis seeds, which declines in seeds upon imbibition (Fig. S3). ShMAX2 expression was detected in all tissues analyzed, including young plants, underground shoots and roots, and floral buds (Fig. 3c).
Functional analysis of ShMAX2 in Arabidopsis
In order to test the functionality of ShMAX2 in Striga, it would be best to obtain mutants or to interfere with ShMAX2 gene expression by use of specific RNAi or amiRNA constructs. However, currently there is no transformation protocol for this species. Therefore, the function of ShMAX2 was tested by complementation analysis in the Arabidopsis max2-1 mutant. The Arabidopsis max2-1 mutant displays increased branching (Stirnberg et al., 2002), delayed seed germination (Shen et al., 2007) and has an increased hypocotyl length in light-grown seedlings (Tsuchiya et al., 2010). For expression in Arabidopsis, the ShMAX2 full-length coding sequence was cloned into a binary expression vector (pBIN:35S:ShMAX2) under the control of the Cauliflower Mosaic Virus 35S promoter (see Methods S5). To test whether the N-terminal extension has an effect on ShMAX2 functionality, we also cloned a truncated version of ShMAX2 without the first ATG and with the transcription start site at a similar position as AtMAX2 (35S:ShMAX2S; Fig. S2). As a positive control, 35S:AtMAX2, which was provided by Professor Otoline Leyser, was used for complementation of the max2-1 mutant. 35S:ShMAX2, 35S:ShMAX2S and 35S:AtMAX2 were introduced into the Arabidopsis max2-1 mutant background by floral dip transformation.
For each construct, 9–23 independent transgenic lines (T0 generation) were selected on kanamycin. Preliminary analysis of T0 plants indicated that all three constructs were able to suppress the branching phenotype of max2-1 (Fig. S4). For each construct two lines displaying the highest reduction in branching were selected for further analysis. The branching phenotype was quantified in kanamycin-resistant T1 progeny plants using the decapitation assay developed by Leyser and coworkers (Stirnberg et al., 2007). Figure 5(a) shows the branching phenotype for each of the constructs. Under the conditions used, wild-type (WT) plants had on average 5.1 branches, while the max2-1 mutant had 28.1. In max2-1 mutant plants with ectopic expression of AtMAX2 the number of side branches was strongly reduced and not significantly different from the WT, indicating that ectopic expression of AtMAX2 can fully complement the branching phenotype (Fig. 5b). The four selected max2-1 lines with ectopic expression of ShMAX2 or ShMAX2S also showed significant reduction in the number of branches (between 16.9 and 8.1) compared with untransformed max2-1, but still had significantly higher numbers of branches than WT.
ShMAX2 complements the Arabidopsis max2-1 root growth phenotype
Strigolactones affect root phenotypes in a MAX2-dependent way (Kapulnik et al., 2011a; Ruyter-Spira et al., 2011). To determine whether ShMAX2 can complement the root growth phenotype of the max2-1 mutant, the primary root length of 5-d-old seedlings of 35S:ShMAX2, 35S:ShMAX2S and 35S:AtMAX2 transgenic plants was determined. Both ShMAX2 and AtMAX2 showed suppression of root elongation of the max2-1 mutant, demonstrating that they can rescue MAX2 signaling in roots (Fig. 5c). Constitutive expression of ShMAX2S reduced the mean length of the root from 1.104 cm in max2 to 0.631–0.517 cm, resulting in plants indistinguishable from WT, with a root length of 0.594 cm. Constitutive expression of ShMAX2 was slightly less effective, reducing the primary root length to 0.694–0.755 cm. Combining these results suggested that ShMAX2 gene functions similar to the AtMAX2 gene in complementing the root growth phenotype of max2-1.
ShMAX2 complements the Arabidopsis max2-1 high irradiance response phenotypes
MAX2 is involved in light-regulated changes in gene expression and the resulting photomorphogenesis such as light-repressed hypocotyl elongation and seed germination under continuous white light (Tsuchiya et al., 2010). Both of these processes represent a response to radiation with relatively high energy for a relatively long period of time and therefore are representative of the High Irradiance Response (HIR) of plants (Briggs et al., 1985). To determine whether ShMAX2 can complement the HIR hypocotyl phenotype of the max2-1 mutant, the hypocotyl length of representative transformants with 35S:ShMAX2, 35S:ShMAX2S or 35S:AtMAX2 was quantified. Arabidopsis max2-1 has a similar hypocotyl length as WT when grown in the dark (Shen et al., 2012), but when grown in continuous light or grown in the dark after germination was triggered with a red (R) light pulse, it has a longer hypocotyl than WT (Tsuchiya et al., 2010). Both ShMAX2 and AtMAX2 complemented the mutant hypocotyl phenotype resulting in a hypocotyl length that was not significantly different from WT (Fig. 6a). Also, there was no difference in complementation between the full length ShMAX2 and its truncated version ShMAX2S, suggesting that the N-terminal extension does not influence the function of ShMAX2 in this process.
In order to determine whether ShMAX2 can complement the HIR seed germination phenotype of the max2-1 mutant, the seed germination rate under continuous white light and seed germination percentage after a R light pulse were quantified for max2-1 transformants with 35S:ShMAX2, 35S:ShMAX2S or 35S:AtMAX2. For all our seed germination assays we made sure that seeds were harvested at the same time from plants grown under the same conditions and seeds were stored under similar conditions to ensure equal after-ripening for the different genotypes. Under continuous white light max2-1 seeds have a lower germination rate than WT seeds (Shen et al., 2007; Tsuchiya et al., 2010). Also in our experiments the germination rate of max2-1 was significantly lower than that of WT seeds (Fig. 6b). Although overexpression of AtMAX2 in this mutant background resulted in a significant increase in the seed germination rate, it was still significantly lower than that of the WT. ShMAX2 generally complemented the reduced seed germination rate of max2 to a similar extent as AtMAX2, with the strongest complementation (the same rate as WT) achieved in line ShMAX2S7. Previously it was shown that the branching phenotype of the Arabidopsis max2 mutant can also almost fully be complemented by one of the MAX2 homologs from petunia (PhMAX2A; Drummond et al., 2011). However, in these studies other Arabidopsis max2 related phenotypes, such as germination, HIR, VLFR, were not characterized, so it is not known whether PhMAX2A homologs can complement these function in the Arabidopsis max2-background as well.
Seed germination was also quantified in response to a R light pulse. Seeds were stratified at 4°C for 4 d after which they received a 5 min pulse of red light in the absence or presence of GR24 to test if strigolactones affect the response at this stage. Two days after the R light pulse, seeds of all genotypes, except max2-1, showed almost 100% germination (Fig. 6c), regardless of the presence of GR24.
ShMAX2 complements the max2-1 seed germination response to GR24
Max2-1 mutant seeds show reduced germination in the dark after a R or far-red (FR) light pulse (Shen et al., 2007), which cannot be complemented by addition of the synthetic strigolactone GR24 (Tsuchiya et al., 2010). To determine whether this GR24 response can be restored by ShMAX2, the seed germination phenotype of WT, max2-1 and max2-1 complemented with ShMAX2 or AtMAX2 was assessed at four different concentrations of GR24 (Fig. 7). After 96 h, seed germination of WT and max2-1 complemented with AtMAX2 was close to 100% with or without GR24, while no germination was observed in max2-1 seeds. In ShMAX2 complemented max2-1 the germination percentage increased to 81.9% with increasing GR24 concentrations, but at the highest GR24 concentration declined again to 45.5%. The whole experiment was repeated with a different line from each genotype, and showed similar results. These results show that ShMAX2 can only partially restore the seed germination response to GR24 in max2-1 and that endogenous strigolactone levels are already sufficient for the full germination response in WT and max2-1 complemented with AtMAX2.
ShMAX2 does not complement the VLFR of max2-1 seed germination
MAX2 is required for seed germination under very low light intensities or after a FR light pulse (Botto et al., 1996; Shen et al., 2007), a response that can be classified as a Very Low Fluence Response (VLFR; Botto et al., 1996). To determine whether the hyposensitivity to FR of max2-1 seed germination can be restored by ShMAX2, the FR light-induced seed germination phenotype of WT, max2 and max2 complemented with ShMAX2 or AtMAX2 was tested. Seeds were stratified at 4°C for 4 d in dark after which they received a 5 min pulse of FR light in the absence or presence of GR24. Three days after the far-red light pulse, max2-1 showed no seed germination, while germination of WT and max2-1 complemented with AtMAX2 were as high as 50% and 80%, respectively (Fig. 8). Under these conditions, there was only 2% germination of max2-1 complemented with ShMAX2. In these assays the addition of GR24 had either no effect or resulted in a slight decrease in germination percentage. The results suggest that, contrary to AtMAX2, ShMAX2 cannot complement for the Very Low Fluence Response (VLFR) of seed germination.
Strigolactones are required for germination of the parasitic plant Striga, but before this study it was unknown whether Striga itself produces strigolactones, whether strigolactone signaling plays a role during Striga development after germination, or whether Striga has similar signaling components as nonparasitic plants, for example such as AtMAX2 in Arabidopsis (Stirnberg et al., 2007). The results presented in this paper provide evidence for strigolactones and strigolactone signaling in Striga beyond germination and we demonstrate that ShMAX2, the AtMAX2 homolog in Striga, can replace most AtMAX2 functions except for its role in the Very Low Fluence Response (VLFR) of seed germination.
Evidence for strigolactones and strigolactone signaling in Striga
The germination of most parasitic plants seeds, for example Striga, Orobanche and Phelipanche spp., does not respond to light, but requires the presence of exogenous stimulants, for instance strigolactones (Matusova et al., 2004). The strict dependence of germination of the parasitic plant Striga on host-derived strigolactones raised the question whether these parasitic plants have lost the capacity for endogenous strigolactone biosynthesis, thus making them strictly dependent on exogenous strigolactones. Although direct measurement of strigolactones in different Striga tissues was not successful (no known strigolactones could be detected by LC-MRM-MS/MS), the HPLC fractioned extracts of sub-cultured in vitro grown Striga did contain fractions that were able to trigger germination of P. ramosa seeds (Fig. 2). Some active fractions coincided with the elution time of known strigolactones, suggestive of strigolactones presence. It is worthy to note that a similar experiment conducted independently by Das et al. (unpublished) reached the same conclusion. Even though the concentration of these putatively present strigolactones was extremely low (below the detection limit of our LC-MRM-MS machine), they were still functional in suppression of branching as inhibition of endogenous strigolactone biosynthesis by fluridone resulted in an increase in branching (Fig. 4). For Striga plants attached to host plants, the strigolactone profile differed depending on the host to which the Striga plant was attached, and the detected strigolactones always matched a subset of the strigolactones detected in the host. In nonparasitic plants, strigolactone biosynthesis is most active in the roots, but Striga does not have an extensive root system of its own. Strigolactones are transported through the xylem to the shoot (Kohlen et al., 2011) and presumably the vascular connection between Striga and its host allows for the transfer of host-produced strigolactones to the parasite. This explains why the strigolactone profile in Striga is dominated by host-derived strigolactones. Putative strigolactone biosynthesis genes are present and expressed in Striga (ShCCD7, ShCCD8; Fig. 3). The overall patterns of expression of these genes in different plant tissues are only slightly different from those of AtCCD7 (MAX3) and AtCCD8 (MAX4) in Arabidopsis (compare Fig. 3 with Fig. S1).
The increased branching in in vitro grown Striga as a result of the application of strigolactone biosynthesis inhibitors strongly suggests that strigolactones play a similar role in the control of shoot branching of Striga as in nonparasitic plants. Although none of the experiments on their own are conclusive, combined these results give sufficient evidence that Striga can produce its own strigolactones, and that endogenous strigolactone signaling in relation to shoot branching is conserved in Striga.
ShMAX2 complements Arabidopsis max2-1 shoot, root and HIR phenotypes
Germination of Striga seeds cannot be induced by light. This suggests that in Striga a component required for light-induced germination was lost or altered. In Arabidopsis, MAX2 is such a component as it is involved in both the high irradiance response (HIR) and very low fluence response (VLFR) of seed germination (Shen et al., 2007).
Because no fast transformation procedure exists for Striga plants and no relevant Striga mutants are available at present, the functionality of the ShMAX2 was tested by complementing the Arabidopsis max2-1 mutant. This mutant is characterized by its increased shoot branching (Stirnberg et al., 2002), a HIR-related seedling hypocotyl (Nelson et al., 2011), a HIR and VLFR seed germination phenotype (Shen et al., 2007), and a root elongation phenotype (Ruyter-Spira et al., 2011). Our results show that, when transformed into max2, ShMAX2 complements the branching, HIR and root phenotypes (Figs 5-7). This indicates that ShMAX2 has not lost any of these known strigolactone-related signaling functions, suggesting that in Striga ShMAX2 may have similar functions as that of AtMAX2.
ShMAX2 does not complement the Arabidopsis max2-1 VLFR seed germination phenotype
The max2-1 mutant also has reduced sensitivity to Far-Red (FR) induced germination (Shen et al., 2007), which is a typical VLFR phenotype. Our germination assay in response to a short FR light pulse indicates full complementation by the AtMAX2 but no complementation by the ShMAX2 (Fig. 8). Results also show that strigolactones affect the VLFR for max2-1/ShMAX2 seeds (Fig. 8). Because VLFR is supposedly mediated by PHYA, this suggests that ShMAX2 has little interaction with the PHYA signaling components of Arabidopsis. The higher germination percentage of the max2-1/AtMAX2 seeds than WT suggests that this MAX2 mediated VLFR may normally be limited by endogenous MAX2 expression levels: overexpression of AtMAX2 in the max2-1 background results in a higher sensitivity to FR than in WT Arabidopsis (Fig. 8).
It is not clear whether the max2 seed germination phenotypes relate to strigolactone signaling as there are conflicting reports about the seed germination phenotype of the different strigolactone biosynthesis mutants (Shen et al., 2007, 2012; Nelson et al., 2009; Tsuchiya et al., 2010; Toh et al., 2012). Also, MAX2 promotes seedling de-etiolation in response to light (Shen et al., 2007), but again the strigolactone biosynthetic mutants (max1-1, max3-100, and max4-100) do not display a seedling de-etiolation phenotype in response to different fluence rates of red or FR light (Shen et al., 2012). Moreover, high levels of the synthetic strigolactone analog GR24 promote photomorphogenesis by inducing nuclear exclusion of COP1, which targets the photomorphogenesis related transcription factor HY5 for degradation and this effect seems to be at least partly independent of MAX2 (Tsuchiya et al., 2010). It might be that seed germination and de-etiolation are extremely sensitive to strigolactones so that even very low levels in leaky strigolactone biosynthesis mutants are sufficient to promote photomorphogenic responses (for recent reviews see Foo & Reid (2012) and Brewer et al. (2013)).
The lack of a germination response after a FR light pulse (seeds were in darkness before and after FR light pulse) of max2-1 or max2-1 complemented with ShMAX2 was not rescued by the addition of GR24 (Fig. 8). In strong contrast to this was the germination response in total darkness (Fig. 7). Here, germination was dependent on a functional MAX2 gene, while the ShMAX2 was less effective in complementing the max2-1 phenotype. Moreover, in complete darkness the max2-1/ShMAX2 seeds responded to added strigolactones (GR24) in a dosage dependent manner (Fig. 7). Combined, the results show that there is an interaction between strigolactones and MAX2, but that effects of strigolactones are very much condition dependent. In full darkness exogenous strigolactones stimulate germination in max2-1/ShMAX2, while for WT and max2-1/AtMAX2 presumably the endogenous strigolactone levels are already sufficient. However, after a FR light pulse this stimulatory role of endogenous strigolactones is abolished in max2-1/ShMAX2 seeds and results in reduced germination in WT and max2-1/AtMAX2 seeds (Fig. 8). It has been shown that AtMAX2 functions in the D14 and KAI2 signaling in combination with the signaling molecules strigolactones and karrikins (Waters et al., 2012). While the interaction of KAI2 with MAX2 is mostly triggered by binding of karrikins, the interaction of D14 with MAX2 is triggered by binding of strigolactones (Waters et al., 2012). Therefore, the difference in the complementation action of ShMAX2 in the max2-1 mutant background (compared to AtMAX2) could be due to a difference in interaction of ShMAX2 with AtD14 and the KAI2 protein of Arabidopsis. Such a difference has been observed for the germination VLFR complementation of max2-1 (Fig. 8). It has been shown that in Arabidopsis KAI2 signaling is most active during seed germination, while during control of shoot branching the signaling of D14 is more prominent (Waters et al., 2012). Therefore, the action of ShMAX2 during germination is most likely due to interaction with KAI2, while the complementation of the branching phenotype is most likely due to interaction of ShMAX2 with AtD14. The lack of response to FR light in max2-1/ShMAX2 indicates a reduced VLFR, which is mediated by PHYA. This result therefore suggests that the putative signaling complex ShMAX2/KAI2 has reduced interaction with one of the PHYA signaling components. The lack of complementation of the VLFR seed germination response by ShMAX2 in Arabidopsis max2-1 could be that the F-box from Striga cannot recognize the protein(s) from Arabidopsis, hence resulting in less efficient interaction of ShMAX2 with KAI2 than of AtMAX2 with KAI2.
Why is Striga seed germination dependent on exogenous strigolactones?
Strigolactone signaling through AtMAX2 has been shown to be involved in the control of Arabidopsis shoot branching. Here we have shown that ShMAX2 can function in this strigolactone signaling pathway in Arabidopsis by demonstrating complementation of the max2-1 branching phenotype by ShMAX2 (Figs 5-7). As in vitro grown Striga also shows a branching phenotype in response to strigolactone biosynthesis inhibitors (Fig. 4) we assume that ShMAX2 also functions in strigolactone-controlled shoot branching in Striga itself. The inhibitor experiment suggests that Striga itself can produce strigolactones. Indeed, putative strigolactone biosynthesis genes, ShCCD7 and ShCCD8, were identified and are expressed in different tissues (Fig. 3). The expression profile of ShCCD7 and ShMAX2 in pre-conditioned Striga seeds is different from the expression profile of the Arabidopsis CCD7 (MAX3) and AtMAX2 as retrieved from publicly available microarray expression data at TAIR (expression data for AtCCD8/MAX4/at4g32810 are not available). Both AtMAX3 and AtMAX2 have relatively high expression in developing and dry seeds, especially in the chalazal seed coat, in Arabidopsis (Winter et al., 2007; Bassel et al., 2008). However, in Arabidopsis, expression of both genes rapidly declines upon imbibition. By contrast, in Striga expression of ShCCD7 and ShCCD8 show a transient maximum c. 7 d of seed conditioning, after which expression of these genes declines to low levels in germinating seeds (Figs 3, S1). Most interestingly, the expression of ShMAX2 increases during Striga seed conditioning, and only declines in germinating seeds. This expression profile of ShMAX2 during seed conditioning correlates well with the increased sensitivity of Striga for strigolactones in response to the preconditioning treatment (Fig. 3; Matusova et al., 2004). Germination of Striga seeds depends on exogenous strigolactones (also after conditioning). Therefore, it is not clear what the significance is of transient ShCCD7 and ShCCD8 expression during preconditioning. Possibly, Striga endogenous strigolactones do play a role in seed germination but are not sufficient to induce germination. Possibly, the lack of a VLFR in Striga seed germination is due to ShMAX2. However, the lack of a HIR in Striga seed germination does not seem to be caused by ShMAX2, as it can complement this phenotype in the Arabidopsis max2-1 mutant. Identifying the target proteins of the SCFShMAX2 protein complex will be of interest to elucidate this enigma.
We thank Prof. O. Leyser (University of Cambridge, UK) and Dr P. Stirnberg (University of York, UK) for providing us with the binary vector containing 35S:AtMAX2. We thank Ronny V.L. Joosen and Hanzi He for their help in seed germination analysis with GERMINATOR. We thank the Netherlands Organization for Scientific Research (to H.J.B. VICI grant, 865.06.002 and Equipment grant, 834.08.001) for providing funds. This project is co-financed by the Centre for BioSystems Genomics (CBSG) which is part of the Netherlands Genomics Initiative/Netherlands Organization for Scientific Research. M.F-A., J.H.W., M.P.T and K.H. were supported by National Science Foundation grant DBI-0701748, with additional support from the Marie Curie postdoctoral fellowship PIOF-GA-2009-252538 (M.F.-A.), USDA Hatch Project no. 135798 (J.H.W.), and NSF award IBN-0322420 and Kirkhouse Trust (M.P.T. and K.H.).