A high accumulation of silicon (Si) is required for overcoming abiotic and biotic stresses, but the molecular mechanisms of Si uptake, especially in dicotyledonous species, is poorly understood. Herein, we report the identification of an influx transporter of Si in two Cucurbita moschata (pumpkin) cultivars greatly differing in Si accumulation, which are used for the rootstocks of bloom and bloomless Cucumis sativus (cucumber), respectively. Heterogeneous expression in both Xenopus oocytes and rice mutant defective in Si uptake showed that the influx transporter from the bloom pumpkin rootstock can transport Si, whereas that from the bloomless rootstock cannot. Analysis with site-directed mutagenesis showed that, among the two amino acid residues differing between the two types of rootstocks, only changing a proline to a leucine at position 242 results in the loss of Si transport activity. Furthermore, all pumpkin cultivars for bloomless rootstocks tested have this mutation. The transporter is localized in all cells of the roots, and investigation of the subcellular localization with different approaches consistently showed that the influx Si transporter from the bloom pumpkin rootstock was localized at the plasma membrane, whereas the one from the bloomless rootstock was localized at the endoplasmic reticulum. Taken together, our results indicate that the difference in Si uptake between two pumpkin cultivars is probably the result of allelic variation in one amino acid residue of the Si influx transporter, which affects the subcellular localization and subsequent transport of Si from the external solution to the root cells.
Silicon (Si) is the most abundant mineral in the earth’s crust, and helps plants to overcome both abiotic and biotic stresses (Epstein, 1999; Ma and Yamaji, 2006). Silicon is effective in alleviating abiotic stresses, including drought, lodging, frost, salinity, heavy metal toxicity and nutrient imbalance (Ma and Yamaji, 2008). Silicon is also effective in the enhancement of pest and pathogen resistance in a wide range of plant species (Fauteux et al., 2005). Recently, the beneficial effects of Si in increasing resistance to pathogen stress have been investigated at the transcriptomic level in several species, including Arabidopsis (Fauteux et al., 2006), Oryza sativa (rice; Brunings et al., 2009) and Triticum (wheat; Chain et al., 2009). These studies found that Si effectively reduces the gene expression induced by pathogens.
Accumulation of Si in the shoot is required to benefit from the positive effects of Si (Ma and Yamaji, 2008); however, the molecular mechanism for Si accumulation is poorly understood, especially in dicotyledonous species. Although all plants contain Si, the level of Si accumulation in the shoots differs greatly among plant species (Ma and Takahashi, 2002; Hodson et al., 2005). Plants of the families Poaceae, Equisetaceae and Cyperaceae show a high accumulation of Si (>4% Si), the Cucurbitales, Urticales and Commelinaceae show an intermediate accumulation of Si (2–4% Si), whereas most other species show a low accumulation of Si (<2%; Ma and Takahashi, 2002; Hodson et al., 2005). These differences have been attributed to the capacity of the roots to absorb Si (Takahashi et al., 1990). Recently, transporters for Si have been identified in some gramineous plants, including rice, Zea mays (maize) and Hordeum vulgare (barley). In rice, which is a typical Si-accumulating species, Lsi1 and Lsi2 are the Si influx and efflux transporters, respectively (Ma et al., 2006, 2007). Both Lsi1 and Lsi2 are localized at the plasma membranes of exodermal and endodermal cells of the roots, but Lsi1 is localized at the distal side of the cells, whereas Lsi2 is localized at the proximal side. These transporters are responsible for the efficient uptake of Si by rice roots. On the other hand, another Si transporter, Lsi6, is localized at the xylem parenchyma cells of rice leaves and nodes (Yamaji et al., 2008; Yamaji and Ma, 2009). It is responsible for unloading Si from the xylem, and for the intervascular transfer of Si, which is important for preferential Si distribution to the panicle. Similar transporters have also been identified in barley and maize (Chiba et al., 2009; Mitani et al., 2009a,b), although the cell-specificity and the localization of these transporters differ between rice and barley/maize (Mitani et al., 2009a). However, no Si transporters have been identified from dicotyledonous plants so far. In the present study, we identified an influx Si transporter from two Cucurbita moschata Duch. (pumpkin) cultivars used for rootstocks of bloomless and bloom Cucumis sativus (cucumber).
The bloom is a white and fine powder on the surface of the cucumber fruits, and is primarily composed of silica (SiO2; Yamamoto et al., 1989). In the late 1980s, cucumber without any bloom (bloomless cucumber) became popular in Japan because of its more attractive and distinctly shiny appearance. Bloomless cucumber is produced by grafting cucumber on specific pumpkin cultivars. Previous studies have shown that these cultivars have lower Si accumulation compared with the rootstocks used for producing bloom cucumber (Yamamoto et al., 1989). For example, cucumber grafted onto a bloomless rootstock (cv. Yuuyuuikki) contained <10 g Si kg−1 in the leaves and 0.0–1.5 g Si kg−1 in the fruits, whereas the same cucumber grafted onto a bloom rootstock (cv. Kurodane) contained 15–50 g Si kg−1 in the leaves and 5 g Si kg−1 in the fruits (Seki and Hotta, 1997). However, the molecular mechanisms involved in Si uptake for bloomless cucumber are still unknown. In the present study, we found that one amino acid mutation in the Si influx transporter of the bloomless pumpkin rootstock affected its localization to the plasma membrane, resulting in a loss of transport of Si from an external solution to the root cells.
Silicon uptake in two pumpkin cultivars, and effect of HgCl2 on Si uptake
We first compared the Si accumulation in the shoots between two pumpkin cultivars: Shintosa (B+) and Super-unryu (B−), which are used for the rootstocks of bloom and bloomless cucumber, respectively (Sakata et al., 2006). When grown on soil, B+ accumulated 15 times more Si in the shoots than B− (Figure 1a). We then compared the uptake capacity per individual root by using a multicompartment transport box. A time-course experiment showed that the Si uptake was much higher in B+ than in B− at all time points (Figure 1b). We also used the whole root system for Si uptake. Both time- and concentration-dependent experiments showed that the Si uptake by B+ was much higher than that by B− (Figure S1a,b). The Si concentration in the xylem sap was also higher in B+ than in B− (Figure S1c). These results indicate that the difference in Si accumulation between B+ and B− is attributed to the difference in the Si uptake capacity of the roots.
We investigated the effect of HgCl2 on Si uptake in the two pumpkin cultivars. HgCl2 is an inhibitor of Si influx transporter (Mitani et al., 2008). The presence of HgCl2 significantly decreased Si uptake in Shintosa (B+) to a similar level as found in Super-unryu (B−) (Figure 2).
Isolation of pumpkin Si influx transporter
To understand the molecular mechanisms involved in different Si uptake in the two pumpkin rootstocks, we cloned Si influx transporter genes based on the similarity to the Si influx transporter OsLsi1 identified in rice (Ma et al., 2006). We obtained a full-length cDNA of Si influx transporter gene from Shintosa [designed as CmLsi1(B+)/CmNIP2-1(B+)] and Super-unryu [CmLsi1(B−)/CmNIP2-1(B−)], respectively. The predicted amino acid sequences of CmLsi1(B+) and CmLsi1(B−) share 58% identity with rice OsLsi1 (Figure S2, Appendix S1). The transporters encoded by these two genes contain typical two asparagine–proline–alanine (NPA) motifs conserved in most aquaporins, and belongs to nodulin 26-like intrinsic membrane protein III (NIP III) subgroup of plant aquaporins, which are characterized by a distinct ar/R (aromatic/arginine) selectivity filter composed from Gly (G), Ser (S), Gly (G) and Arg (R) (Figure S3) (Mitani et al., 2008). Comparison of the two sequences reveals that CmLsi1(B+) and CmLsi1(B−) only differ in two amino acids, at positions 75 and 242 (Figure S3).
Transport activity of the pumpkin Si influx transporter
To determine the transport activity of the pumpkin Si influx transporter for silicic acid, we expressed the genes isolated from different pumpkin cultivars in Xenopus oocytes. CmLsi1(B+) showed some transport activity, whereas CmLsi1(B−) did not (Figure 3a). Neither CmLsi1(B+) nor CmLsi1(B−) showed any transport activity for water and urea, in contrast to rice OsLsi1, which was included as a positive control (Figure 3b,c). Further analysis using site-directed mutation showed that a mutation at position 75 (V → A) did not affect the transport activity for silicic acid, whereas a mutation at position 242 (P → L) resulted in a loss of transport activity (Figure 3a). These results indicate that the amino acid at position 242 is critical for Si transport. To examine whether loss of transport activity caused by this amino acid substitution is to the result of mislocalization, we expressed CmLsi1(B+) and GFP-CmLsi1(B+)P242L fused with GFP in oocytes. The result showed that CmLsi1(B+) was localized at the plasma membrane (Figure S4), whereas GFP-CmLsi1(B+)P242L was not.
To further demonstrate that the amino acid mutation at position 242 in the Si influx transporter of pumpkin rootstocks was responsible for the loss of function in bloom production, we compared the sequences of CmLsi1 among seven other pumpkin cultivars, including three bloom rootstocks and four bloomless rootstocks. All bloomless rootstocks have the same mutation at position 242 (L instead of P) (Figure 4). Although a mutation is also found at position 75, this mutation is not related to the production of bloom (Figure 4). This result is consistent with the finding from site-directed mutagenesis that a change from V to A at position 75 did not result in a loss of function (Figure 3a).
Using the rice OsLsi1 promoter, we also transferred CmLsi1(B+) and CmLsi1(B−) genes into a rice mutant lsi1-3 that is defective in the influx transport of Si. Three independent lines were generated and subjected to Si uptake. In transgenic lines carrying CmLsi1(B+), the Si uptake was restored to the level of the wild-type rice (Figure 5a), whereas the lines carrying CmLsi1(B−) did not complement Si uptake in the mutant. Both CmLsi1(B+) and CmLsi1(B−) were highly expressed in the transgenic rice lines (Figure 5b). This result further indicates that CmLsi1(B+) is functional as an Si transporter, but that CmLsi1(B−) is not.
Expression pattern and localization of the pumpkin Si influx transporter
Both CmLsi1(B+) and CmLsi1(B−) were expressed in both the roots and shoots of pumpkin. The expression level of Lsi1 was slightly higher in B− than in B+ (Figure 6a). Immunostaining with an antibody of CmLsi1 shows that both CmLsi1(B+) and CmLsi1(B−) were localized at all root cells (Figure 6b,c). However, in contrast to CmLsi1(B+), which appeared to be localized to the plasma membrane (Figure 6b), CmLsi1(B−) was not found at the plasma membrane (Figure 6c).
To confirm this result, we performed Western blot analysis with the microsome fractions isolated from both B+ and B−. With the help of a sucrose gradient, we found that CmLsi1(B+) was present in the same fraction as H+-ATPase, a plasma membrane marker (Figure 6d), whereas CmLsi1(B−) was found in fractions similar to the luminal binding protein (BiP), an ER marker. As there was some overlap between ER and tonoplast (V-ATPase) fractions, we performed Western blot analysis in the presence of Mg2+ or EDTA (Ishikawa et al., 2005). In the presence of Mg2+, CmLsi1(B−) was found in the fractions similar to BiP (Figure 6e). When the membrane-bound ribosomes were removed from the rough ER by treatment with EDTA, the CmLsi1(B−) peak moved to lower-density fractions, as did the BiP peak (Figure 6e). This result further indicates that CmLsi1(B−) was localized at the ER.
Furthermore, we investigated the localization of pumpkin Lsi1 transporters in transgenic rice expressing either CmLsi1(B+) or CmLsi1(B−). Immunostaining clearly showed that CmLsi1(B+) was localized on the distal side of plasma membrane of both the exodermis and the endodermis (Figure 7a), whereas CmLsi1(B−) was localized to internal structures, probably the ER (Figure 7b,c). The localization of CmLsi1(B+) is the same as that of wild-type rice OsLsi1 (Ma et al., 2006), supporting the complementation test result (Figure 5).
Site-directed mutagenesis analysis showed that the amino acid at position 242 is critical for Si transport (Figure 3a). To demonstrate that this amino acid affects the localization, we constructed GFP fusion genes of wild-type CmLsi1(B+) and site-direct mutagenized CmLsi1(B+)P242L, and then co-expressed these with an ER-localized fluorescence protein marker, SP-DsRed-HDEL, in Allium cepa (onion) epidermal cells. SP-DsRed-HDEL is composed of the signal peptide (SP) of pumpkin 2S albumin followed by red fluorescence protein DsRed and a 12-amino-acid sequence including an ER-retention signal, HDEL (Mitsuhashi et al., 2000). GFP-CmLsi1(B+) was localized to the plasma membrane (Figure 8a,c,e), whereas SP-DsRed-HDEL was localized to the ER (Figure 8a,d,e). By contrast, GFP-CmLsi1(B+)P242L was co-localized with SP-DsRed-HDEL (Figure 8b,f–h). This result further confirms that the amino acid mutation at position 242 affects the localization of the Si influx transporter.
Our results show that the difference in the Si accumulation between two pumpkin cultivars used for rootstocks of bloom and bloomless cucumber, results from the capacity of the roots to take up Si (Figure 1 and Figure S1). The uptake of Si by the roots is mediated by influx (Lsi1) and efflux (Lsi2) transporters found in grass, including rice, maize and barley (Ma et al., 2006, 2007; Chiba et al., 2009; Mitani et al., 2009a,b). Lsi1 is responsible for Si uptake from the external solution to the root cells, and plays a crucial role in whole Si uptake. In rice, the knock-out of Lsi1 significantly causes decreased Si uptake (Ma et al., 2006). We therefore cloned an influx transporter of Si from these pumpkin cultivars. Sequence analysis showed that there were two amino acid differences in the influx Si transporter between the two pumpkin cultivars (Figure S3). Heterogeneous expression in both Xenopus oocytes and rice mutant defective in Si uptake showed that the influx transporter from the bloom pumpkin rootstock is functional in Si transport, whereas that from the bloomless rootstock is nonfunctional (Figures 3 and 5). These results indicate that the genotypic difference in Si uptake is probably caused by allelic variation in the Si influx transporter. This conclusion is supported by the finding that the Si uptake became similar between two cultivars when influx Si transport was inhibited by HgCl2, which is a specific inhibitor of aquaporin (Figure 2).
Silicon influx transporter Lsi1 belongs to the NIPs subfamily of aquaporins. In the NIPs subfamily, three groups have been divided based on the ar/R (aromatic/arginine) selectivity filter (Mitani et al., 2008), which is located in the narrowest region on the extramembrane mouth of the pore (Wallace and Roberts, 2004; Forrest and Bhave, 2007). Lsi1 belonging to the NIP III subgroup is characterized by a unique selectivity filter, which consists of smaller size residues: Gly (G), Ser (S), Gly (G) and Arg (R) (Mitani et al., 2008). Lsi1 isolated from two pumpkin cultivars have the same selectivity filter (Figure S3). Our results show that the loss of function of Lsi1 from the bloomless rootstock pumpkin cultivar results from a failure of localization to the plasma membrane. Lsi1 functions as an influx transporter; therefore, its localization at the plasma membrane is required. Lsi1 from the bloom rootstock pumpkin cultivar is localized at the plasma membrane; however, Lsi1 from the bloomless rootstock pumpkin cultivar is retained at the ER in pumpkin, rice and onion (Figures 6–8). The ER is a place for protein quality control (Vitale and Boston, 2008). In the ER, newly synthesized membrane proteins, secretory polypeptides and oligomeric proteins assemble before being sorted at their final destination. Retention of Lsi1 from the bloomless rootstock pumpkin cultivar in the ER suggests that it fails to fold or assemble properly.
We further showed that the different localization of Lsi1 in two pumpkin cultivars is caused by one amino acid mutation. There are two amino acid differences, at positions 75 and 242, between the sequences of Lsi1 taken from two cultivars (Figure S3). Site-directed mutagenesis analysis revealed that the change from a proline to a leucine at position 242, rather than the residue at position 75, results in the loss of Si transport activity arising from mislocalization (Figure 3 and Figure S4). This is also supported by allelic mutations in all cultivars tested: all rootstock cultivars for bloomless cucumber have the same mutation at position 242, but not at position 75 (Figure 4). Furthermore, heterogenous expression in onion epidermal cells also indicates that mutation of this amino acid results in the localization of Lsi1 at the ER (Figure 8). According to the modeling of rice Lsi1 (Ma et al., 2006), the amino acid at position 75 is located in the loop, whereas that at 242 is located in the membrane domain. The amino acid mutation at 242 might affect protein folding, therefore affecting the localization to the plasma membrane.
Although, like rice Lsi1, pumpkin Lsi1 also functions as an influx transporter of Si, there are differences in the transport substrates and cell-type specificity of localization between rice and pumpkin. Rice Lsi1 is also permeable to water and urea (Mitani et al., 2008), although the affinity for Si is higher. By contrast, functional pumpkin Lsi1 is not permeable to water and urea (Figure 3b,c). As rice and functional pumpkin Lsi1 share the same selectivity filter (Figure S3), the difference in the transport substrate specificity may be caused by other uncharacterized protein regions.
Pumpkin Lsi1 is localized at all root cells (Figure 6b,c). This is different from rice Lsi1 (Ma et al., 2006), but is similar to barley and maize (Chiba et al., 2009; Mitani et al., 2009b). This difference has been attributed to the difference in the root structure (Chiba et al., 2009; Mitani et al., 2009b). Rice roots form aerenchyma, but this is not developed under normal conditions in other plants, including barley, maize and pumpkin (Figure 6). Interestingly, unlike Lsi1 from rice, barley and maize (Ma et al., 2006; Chiba et al., 2009; Mitani et al., 2009b), pumpkin Lsi1 did not show polar localization. However, when pumpkin Lsi1 was expressed in rice under the control of the rice Lsi1 promoter, it showed polar localization at the distal side of both the exodermis and the endodermis (Figure 7a). These results suggest that polar localization depends on cell-specific factors rather than protein-specific properties, although the exact mechanisms remain to be examined in future.
An efflux transporter Lsi2 is also important for Si uptake, in addition to Lsi1 (Ma et al., 2007). Our preliminary results show that there was no difference in the sequence of Lsi2-like protein between two cultivars (data not shown). However, we have yet to examine whether Lsi2 is also involved in the difference in Si uptake between these two pumpkin cultivars.
In conclusion, we identified an influx transporter of Si in a dicotyledonous species, pumpkin, and suggested that one amino acid mutation at the position of 242 in this transporter could be responsible for the difference of Si accumulation between two cultivars, by affecting the subcellular localization.
Plant materials and growth conditions
Two cultivars of pumpkin (C. moschata Duch.), Shintosa (B+) and Super-unryu (B−), were used. These cultivars are used for the rootstock of bloom (B+) and bloomless (B−) cucumber, respectively. Seeds were soaked in deionized water for 4 h and then placed on moist filter paper in a Petri dish incubated overnight at 4°C. Germinated seeds were transferred to a floating nylon mesh on 0.5 mm CaCl2 solution in a growth cabinet at 25°C in the dark. Seedlings were transplanted into 3-l plastic pots containing continuously aerated one-fifth strength Hoagland’s solution [1 mm KNO3, 1 mm Ca(NO3)2, 0.4 mm MgSO4 and 0.2 mm KH2PO4, 20 μm Fe-EDTA, 3 μm H3BO3, 1 μm (NH4)6Mo7O24, 0.5 μm MnCl2, 0.4 μm ZnSO4 and 0.2 μm CuSO4; pH 6.0). The nutrient solution was replaced once every 2 days. The plants were grown in a growth chamber (25°C 14-h day/20°C 10-h night, light intensity of 40 W m−2 and 70% relative humidity).
Silicon accumulation and uptake
Seedlings prepared as described above were transferred to a pot containing 3 kg of soil. After 2 weeks, the shoots were harvested and analyzed for Si as described previously (Tamai and Ma, 2008). The Si uptake by an individual root was determined by using a multicompartment transport box, as described previously (Yamaji and Ma, 2007). Root segments (5.5 cm long) excised from 7-day-old seedlings were placed in a box with three compartments. The root tips (about 0–30 mm in length) were placed in the first compartment containing 0.5 mm Si as silicic acid, and the other parts were exposed to the solution without Si. The Si exuded from the cutting end in the third compartment was determined according to procedures described previously (Yamaji and Ma, 2007). The Si uptake by whole roots (14-days old) was also determined by exposing the plants to a solution containing different Si concentrations (0.2–1.6 mm) at different times. The xylem sap was collected according to the method described by Yamaji et al. (2008).
Cloning of influx Si transporter genes
RNA was extracted from pumpkin roots of the two cultivars by using an RNeasy Plant Mini kit (Qiagen, http://www.qiagen.com). Total RNA (0.5 μg) was reverse-transcripted into cDNA by the SuperScript™ First-Strand Synthesis System (Invitrogen, http://www.invitrogen.com). Primers were designed for the cloning of Si influx transporter genes based on the conserved sequence of rice Si influx transporter OsLsi1 and zucchini CpNIP1 (accession no. AJ544830). After obtaining a partial fragment by PCR, the full-length CmLsi1 cDNA was amplified by a rapid amplification of cDNA ends (RACE) method using the SMART™ RACE cDNA Amplification kit (Clontech, http://www.clontech.com) from total RNA prepared from pumpkin roots. The sequences of amplified fragments were confirmed by the 3130 Genetic Analyzer using the BigDye Terminators v3.1 Cycle Sequencing kit (Applied Biosystems, http://www.appliedbiosystems.com). The gene cloned from Shintosa was named CmLsi1(B+), and that from Super-unryu was named CmLsi1(B−).
Comparison of Si transporter sequences among different pumpkin cultivars
The sequences of CmLsi1 were compared between different cultivars, including four cultivars used for bloom rootstocks and five cultivars used for bloomless rootstocks. The plants were cultivated as described above and RNA was extracted from the roots of each cultivar. The cDNA including the full length of the CmLsi1 open reading frame (ORF) was amplified by PCR using the following primers: 5′-GAAACAAAGCGAGAAGACAAGG-3′ and 5′-GCTTTCATATACGCTGCTAGATG-3′. The sequence of each clone was determined as described above.
Oocytes were isolated from Xenopus laevis. Procedures for defolliculation, culture conditions and selection were the same as described previously (Mitani et al., 2008). For the synthesis of capped RNA, the ORF was amplified by RT-PCR with the following primers: 5′-GGATGATCAATGAGTTCTTCTCAGGATCCTCAGC-3′ and 5′-GGATGATCATCATCTTTCACCTTCACCAACGTCA-3′ for CmLsi1. The fragment containing the ORF was inserted into the BglII site of a Xenopus oocytes expression vector, pXβG-ev1. For site-directed mutagenesis, single nucleotide substitutions for V75A and P242L, respectively, were introduced by PCR using following primer pairs: 5′-GGATGATCAATGAGTTCTTCTCAGGATCCTCAGC-3′ and 5′-CAGCAAAAGCCGTGGTGACAGC-3′, 5′-GGCTGTCACCACGGCTTTTGCTG-3′ and 5′-GGATGATCATCATCTTTCACCTTCACCAACGTCA-3′.
Capped RNA was then synthesized from linearized pXβG-ev1 plasmids by in vitro transcription with the mMASSAGE mMACHINE High Yield Capped RNA Transcription kit (Ambion, http://www.ambion.com), according to the manufacturer’s instructions. A volume of 50 nl (1 ng nl−1) in vitro cRNA transcripts was injected into the selected oocytes using a Nanoject II automatic injector (Drummond Scientific Co., http://www.drummondsci.com). As a negative control, 50 nl of RNase-free water was also injected. After incubation at 18°C for 1 day, the oocytes were used for a transport activity assay of Si influx, as described previously (Ma et al., 2006).
The transport activity for urea was determined by exposing oocytes to a solution containing 14C-labeled 1 mm urea (200 mBq mmol−1) for 30 min (Mitani et al., 2008). For the determination of water permeability, pre-incubated oocytes in control MBS (osmolarity 200 mOsm kg−1) were transferred to one-fifth diluted MBS (40 mOsm kg−1). Changes in the oocyte volume were monitored every 20 s with a CCD camera operated by metaview (Universal Imaging Co., now Molecular Devices, http://www.moleculardevices.com). win roof (MITANI Corporation, http://www.mitani-corp.co.jp) was used to determine oocyte diameter, and volumes of oocytes were calculated assuming spherical geometry (Mitani et al., 2008).
Generation of transgenic rice
The promoter region of OsLsi1 (1.6 kb) was amplified from rice (cv. Nipponbare) genomic DNA using the following primers: 5′-ATAGATCTTCAGATGATATGTCGATTTCTG-3′ and 5′-TTTCTAGAGCTACTTGTGCAGCACGCG-3′. The ORFs of CmLsi1(B+) and CmLsi1(B−) were amplified and connected to the rice OsLsi1 promoter. These two fragments were inserted into a binary vector (pPZP2H-lac), and the final construct was named pPZP-CmLsi1(B+) or pPZP-CmLsi1(B−). The plasmid was transformed into Agrobacterium tumefaciens strain EHA101, and then transferred into the rice mutant lsi1-3 using an Agrobacterium-mediated transformation system.
We generated three independent transgenic lines, and used these for the Si uptake experiment as well as wild-type rice and the mutant with three replicates. The uptake experiment was conducted in a nutrient solution containing 0.5 mm Si as silicic acid for 24 h according to procedures described previously (Yamaji and Ma, 2007). The expression level of CmLsi1(B+) and CmLsi1(B−) in the transgenic lines were examined by semiquantitative RT-PCR. Total RNA was extracted from each line as described above. A PCR reaction was performed in a 20-μl reaction volume containing 1 μl of 1 : 5 diluted cDNA, 0.2 μm of each primer and 10 μl of 2× Quick TaqTM HS DyeMix (TOYOBO, http://www.toyobo.co.jp). PCR conditions were 2 min at 94°C, 26 cycles for CmLsi1 or 22 cycles for OsHistoneH3 of 30 s at 94°C, 30 s at 60°C and 30 s at 68°C. HistoneH3 was used as an internal control. Primer sequences used are as follows: CmLsi1 5′-GTTCTTCTCAGGATCCTCAGC-3′ (forward) and CmLsi1-semi 5′-CCTGATGGCGTGGTAGTGC-3′ (reverse), OsHistoneH3 5′-GGTCAACTTGTTGATTCCCCTCT-3′ (forward), and 5′-AACCGCAAAATCCAAAGAACG-3′ (reverse).
Roots and shoots of two pumpkin cultivars were sampled and subjected to RNA extraction, as described above. Absolute quantitative RT-PCR was performed in a 20-μl reaction volume containing 1 μl of 1 : 5 diluted cDNA, 0.3 μm each of gene-specific primers, 0.4 μl of ROX reference dye and 10 μl of THUNDERBIRD SYBR qPCR Mix (TOYOBO) using Mastercycler realprex4 (Eppendorf, http://www.eppendorf.com). PCR conditions were 30 s at 95°C, 40 cycles of 15 s at 98°C, 30 s at 55°C and 35 s at 72°C, followed by melting/dissociation curve analysis. The following primer pairs were used: CmLsi1, 5′-GTTCTTCTCAGGATCCTCAGC-3′ (forward) and 5′-GAACAACGATCCAAACTGGG-3′ (reverse). For each PCR run, a standard curve was generated by using diluted plasmid DNA as a template, and estimates of absolute copy number for each sample were obtained based on this standard curve.
The synthetic peptide CSLKLRRMSRSDVGEGER (positions 2–18 of CmLsi1) was used to immunize rabbits to obtain antibodies against both CmLsi1(B+) and CmLsi1(B−). The antiserum obtained was purified through the peptide affinity column before use. The roots of two pumpkin cultivars and transgenic rice were used for the immunostaining of CmLsi1, as described previously (Yamaji and Ma, 2007). Fluorescence of secondary antibody (Alexa Fluor 555 goat anti-rabbit IgG; Molecular Probes, available from Invitrogen, http://www.invitrogen.com) was observed with a fluorescence microscope (Axio Imager with Apotome; Carl Zeiss, http://www.zeiss.com).
Construction of fluorescent gene fusion
For construction of a translational GFP-CmLsi1 fusion, the ORF of CmLsi1 was amplified by PCR from root cDNA. The primer pair 5′-ATGTACAAGATGAGTTCTTCTCAGGATCCTC-3′ and 5′-TTGCGGCCGCTCATCTTTCACCTTCACCAA-3′ was used for amplification and the introduction of restriction sites. The amplified ORF was cloned between 35S promoter-GFP and NOS terminator in pBluescript vector. For construction of an ER-localized fluorescence marker protein gene SP-DsRed-HDEL, we amplified the ORF of DsRed-monomer from pDsRed-Monomer vector (Takara Bio Inc., http://www.takara-bio.com), and introduced sequences encoding the N-terminal signal peptide of pumpkin 2S albumin and C-terminal ER-retention signal HDEL, respectively, by PCR, following the methodology of Mitsuhashi et al. (2000). A primer pair used for amplification and introduction of the sequences was 5′-AGTCGACCATGGCCAGACTCACAAGCATCATTGCCCTCTTCGCAGTGGCTCTGCTGGTTGCAGATGCGTACGCCTACCGCACCATGGACAACACCGAG-3′ and 5′-TGCGGCCGCTCAAAGCTCATCGTGGTGGTGGTGGTGGTGCCCCCCCCCCTGGGAGCCGGAGTGG-3′. The amplified SP-DsRed-HDEL ORF was cloned between the 35S promoter and the NOS terminator in pBluescript vector.
Co-expression in onion epidermal cells
We observed the subcellular localization of GFP-CmLsi1(B+) and GFP-CmLsi1(B+)P242L, together with ER-marker protein SP-DsRed-HDEL, by transient expression in onion epidermal cells. Onion epidermal cells were bombarded with 1-μm gold particles coated with plasmid DNA carrying GFP-CmLsi1(B+) or GFP-CmLsi1(B+)P242L and SP-DsRed-HDEL, and then incubated in the dark at 25°C for 18 h. We observed the fluorescence by laser-scanning confocal microscopy (LSM700; Carl Zeiss).
Western blot analysis
Roots from 28-day-old plants were homogenized in 50 ml of ice-cold buffer solution comprising 100 mm Tris–HCl (pH 8.0), 150 mm KCl, 0.5% (w/v) polyvinylpolypyrrolidone, 5 mm EDTA, 3.3 mm DTT, 1 mm phenylmethylsulfonyl fluoride (PMSF) and 10% (v/v) glycerol (Sugiyama et al., 2007). After filtration, the homogenates were centrifuged at 8000 g for 10 min to yield the supernatant, and then centrifuged again under the same conditions. The supernatants were then ultracentrifuged at 100 000 g for 40 min. The pellets were resuspended in 1 ml of buffer solution containing 10 mm Tris–HCl (pH 7.6), 10% (v/v) glycerol, 1 mm EDTA, 1 mm DTT and 1/100 volume of Protease Inhibitor Cocktail for plant cell and tissue extracts (Sigma-Aldrich, http://www.sigmaaldrich.com). The suspensions (microsomal membranes) were layered on sucrose density gradients (10 ml, 0–50%) with 10 mm Tris–HCl (pH 7.6), 1 mm EDTA and 1 mm DTT centrifuged at 100 000 g for 19 h, and then fractionated into 1 ml each. The fractionated membranes were recovered by ultracentrifugation at 100 000 g for 40 min. Each pellet was resuspended in 50 μl of buffer solution supplemented with a 1/100 volume of Protease Inhibitor Cocktail for plant cell and tissue extracts (Sigma-Aldrich) and 1 mm DTT. Protein concentration was determined by BioRad protein assay (BioRad, http://www.bio-rad.com).
The same quantity of protein samples were subjected to SDS-PAGE and immunoblotting. The blots were treated with 1 : 100 dilutions of CmLsi1 antibody, 1 : 1000 H+-ATPase (plasma membrane marker; Agrisera, http://www.agrisera.com), 1 : 5000 of a V-ATPase (tonoplast marker; Agrisera) and the 1 : 2000 of ER luminal binding protein (BiP, ER marker; COSMO BIO, http://www.cosmobio.co.jp), respectively. Anti-Rabbit IgG (H+L) HRP Conjugate (1 : 10 000 dilution; Promega, http://www.promega.com) was used as a secondary antibody, and the ECL Plus Western Blotting Detection System (GE Healthcare, http://www.gehealthcare.com) was used for detection via chemiluminescence.
To further confirm the subcellular localization of CmLsi1 (B−), the microsome was extracted in the presence of 2 mm MgCl2 or 2 mm EDTA, according to the method of Ishikawa et al. (2005), but with some modifications. The crude membrane fraction was layered on sucrose density gradients (10 ml, 20–45%) with 20 mm Tris-acetate (pH 7.5), 1 mm EGTA, 2 mm DTT, and 20 μm PMSF with 2 mm MgCl2 or 2 mm EDTA. Centrifugation, fractionation and recovery steps were the same as described above. Each pellet was resuspended in 20 mm Tris-acetate (pH 7.5), with 2 mm MgCl2 or 2 mm EDTA.
This research was supported by a Grant-in-Aid for Scientific Research on Innovative Areas from the Ministry of Education, Culture, Sports, Science and Technology of Japan (22119002 to JFM), a grant from the Ministry of Agriculture, Forestry and Fisheries of Japan, Genomics for Agricultural Innovation IPG-0006 (to JFM) and Ohara Foundation for Agricultural Research.
Sequence data from this article can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under the following accession numbers: AB551949 for (CmLsi1(B+) and AB551950 for CmLsi1(B−), respectively.