A major latex-like protein is a key factor in crop contamination by persistent organic pollutants.

This is the first report, to our knowledge, to reveal important factors by which members of the Cucurbitaceae family, such as cucumber (Cucumis sativus), watermelon (Citrullus lanatus), melon (Cucumis melo), pumpkin (Cucurbita pepo), squash (C. pepo), and zucchini (C. pepo), are selectively polluted with highly toxic hydrophobic contaminants, including organochlorine insecticides and dioxins. Xylem sap of C. pepo ssp. pepo, which is a high accumulator of hydrophobic compounds, solubilized the hydrophobic compound pyrene into the aqueous phase via some protein(s). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of xylem sap of two C. pepo subspecies revealed that the amount of 17-kD proteins in C. pepo ssp. pepo was larger than that in C. pepo ssp. ovifera, a low accumulator, suggesting that these proteins may be related to the translocation of hydrophobic compounds. The protein bands at 17 kD contained major latex-like proteins (MLPs), and the corresponding genes MLP-PG1, MLP-GR1, and MLP-GR3 were cloned from the C. pepo cultivars Patty Green and Gold Rush. Expression of the MLP-GR3 gene in C. pepo cultivars was positively correlated with the band intensity of 17-kD proteins and bioconcentration factors toward dioxins and dioxin-like compounds. Recombinant MLP-GR3 bound polychlorinated biphenyls immobilized on magnetic beads, whereas recombinant MLP-PG1 and MLP-GR1 did not. These results indicate that the high expression of MLP-GR3 in C. pepo ssp. pepo plants and the existence of MLP-GR3 in their xylem sap are related to the efficient translocation of hydrophobic contaminants. These findings should be useful for decreasing the contamination of fruit of the Cucurbitaceae family as well as the phytoremediation of hydrophobic contaminants.

Numerous agricultural fields and crops have been contaminated with persistent organic pollutants (POPs), including dioxins, such as polychlorinated dibenzop-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs); dioxin-like compounds, such as coplanar polychlorinated biphenyls (PCBs) and the insecticide dichlorodiphenyltrichloroethane; drins, such as aldrin, dieldrin, and endrin; and chlordane (Hashimoto, 2005;Uegaki et al., 2006;Hilber et al., 2008). POPs show carcinogenicity, teratogenicity, immunotoxicity, and estrogenicity toward humans and wildlife after accumulation through the food chain. Despite the fact that the use of PCBs and these insecticides was prohibited several decades ago, environmental and crop contamination remains a problem due to their high hydrophobicity and chemical stability.
Members of the Cucurbitaceae family, such as cucumber (Cucumis sativus), watermelon (Citrullas lanatus), melon (Cucumis melo), pumpkin (Cucurbita pepo), and zucchini (C. pepo), are some of the major crops in the world. Previous studies reported that members of the Cucurbitaceae family, particularly C. pepo, which includes pumpkin and zucchini, accumulated higher levels of PCDDs and PCDFs (Hülster et al., 1994;Inui et al., 2008), 2,2-bis(p-chlorophenyl) 1,1-dichloroethylene (p,p9-DDE; White et al., 2003), PCBs (Aslund et al., 2008;Inui et al., 2008), chlordane (Mattina et al., 2004), and drins (Otani et al., 2007) compared with the levels in other plant species. Thus, it appears that the Cucurbitaceae family has unique mechanisms of POP uptake and translocation. Lunney et al. (2004) reported that the shoots of pumpkin and zucchini plants showed much higher concentrations of dichlorodiphenyltrichloroethane than those of tall fescue (Festuca arundinacea), alfalfa (Medicago sativa), and ryegrass (Lolium multiflorum), whereas concentrations in roots were similar among these plants. Likewise, significant differences were found between C. pepo ssp. pepo and ssp. ovifera in concentrations of dioxins and dioxin-like compounds in the aerial parts, whereas the concentrations in their roots were similar . These results suggest that the mechanisms causing the high accumulation of POPs in C. pepo plants mainly occur during translocation from the roots to the aerial parts.
The transport of substances such as nutrients and signal molecules over long distances in higher land plants is mediated by the vascular bundles, which consist of phloem and xylem strands. In addition to inorganic salts, organic nutrients such as amino acids, sugars, and organic acids are translocated through the xylem from the roots to the aerial parts (Satoh, 2006). Furthermore, the fact that POPs such as chlordane, dieldrin, and PCBs were detected in xylem sap of C. pepo suggests that their accumulation in the aerial parts of plants occurs during the translocation from roots to aerial parts in xylem sap (Mattina et al., 2004;Murano et al., 2010b;Greenwood et al., 2011). A recent study revealed that there were protein-like materials with the ability to dissolve dieldrin in xylem sap (Murano et al., 2010a). However, these materials have yet to be identified, and the mechanisms underlying the high transport ability and high accumulation of POPs in C. pepo plants are not fully understood.
In this study, to clarify the molecular mechanisms of the efficient uptake and high accumulation of POPs by C. pepo plants, xylem sap proteins related to the transport of POPs in xylem sap were identified. The aim of this research is to provide a means of preventing cucumber, melon, watermelon, pumpkin, and zucchini fruits from being contaminated by POPs.

Solubilization Activities of Xylem Sap toward Pyrene
Pyrene-adsorbed Tenax was incubated with each xylem sap from nine C. pepo cultivars, and fluorescence from pyrene in the supernatant, which is desorbed from the Tenax and glass tube, was measured. The time for the incubation of pyrene-adsorbed Tenax with xylem sap was decided as 4 h because a solubilization activity of cv Patty Green (PG) and cv Magda (MG) at 4 h was almost saturated (Supplemental Fig. S1). There were differences in the pyrene solubilization activities of xylem sap among the nine tested C. pepo cultivars (Fig. 1A). The activity of water was lower than those of all xylem sap samples. The activity of PG xylem sap was the lowest among cultivars, whereas that of MG was three times higher than that of PG. To inactivate proteins in xylem sap, the xylem sap of MG was boiled. The activity of boiled MG sap was similar to that of PG sap. The pyrene solubilization activities of xylem sap were significantly positively correlated with bioconcentration factors (BCFs) of PCBs (Matsuo et al., 2011;r = 0.781 [Pearson's correlation coefficient], P , 0.05; Fig. 1B). The solubilization activities and BCFs of PCDDs and PCDFs have a trend to be positively correlated (Matsuo et al., 2011; r = 0.636, P = 0.0656 for PCDDs, r = 0.658, P = 0.0538 for PCDFs; data not shown). The BCFs used here were calculated by dividing the concentration of PCDDs, PCDFs, and PCBs in aerial parts including leaves and stems by the concentration in soil.

SDS-PAGE for Xylem Sap
Proteins in xylem sap were compared among the nine C. pepo cultivars by means of SDS-PAGE ( Fig. 2A).
Although multiple common bands were detected, a quantitative difference was observed in protein bands at approximately 17 kD. Band intensities at 17 kD were significantly positively correlated with BCFs of PCBs (Matsuo et al., 2011; r = 0.669, P , 0.05; Fig. 2B). The band intensities and the BCFs of PCDDs and PCDFs and phytoextraction of p,p9-DDE were also significantly positively correlated (White et al., 2003;Matsuo et al., 2011; r = 0.711 for PCDDs, r = 0.742 for PCDFs, and r = 0.780 for p,p9-DDE, P , 0.05 in all cases; Supplemental Fig. S2). The intensities of the other fluctuated bands at approximately 87, 37, 23, and 10 kD did not show significant correlation with BCFs of PCBs (data not shown).

Binding Activities of Xylem Sap Proteins to PCBs
The existence of PCB-binding proteins in xylem sap was examined by using PCB-immobilized or control beads. 4-Hydroxy-29,3,39,49,59-pentachlorobiphenyl (4OH-PeCB106) was used for the preparation of PCB-immobilized beads instead of pyrene, because 4OH-PeCB106 has a hydroxylated group suitable for binding to the beads, and the 29,3,39,49,59-pentachlorobiphenyl (PeCB106) immobilized on beads will be bound to proteins responsible for translocation. Xylem sap from MG was mixed with the magnetic beads binding 4OH-PeCB106, and proteins binding PeCB106 eluted by heating for 5 min at 98°C were subjected to SDS-PAGE. In the PCBimmobilized bead treatment, more 17-kD proteins were eluted than in the control bead treatment (Fig. 3), whereas there were no bands at the last washing step for both concentrations of 4OH-PeCB106 (Supplemental Fig. S3).

Identification of the Proteins in Xylem Sap and Cloning of MLP Genes
The partial amino acid sequences VYGFFR, YIIYEAVEGD, and WSVVFEK of 17-kD proteins from the xylem sap of cv Gold Rush (GR) were obtained (Fig. 4). Using the melon unigene library in the Cucurbit Genomics Database, these three sequences showed homology with MU50778, MU52905, and MU57801, which are annotated as major latex-like proteins (MLPs). By using reverse transcription (RT)-PCR and 59 and 39 RACE-PCR, three MLP genes (MLP-PG1 from PG and MLP-GR1 and MLP-GR3 from GR) were cloned. Each of these genes had 471 bp (Supplemental Fig. S4). The three MLPs were approximately 17 kD, and sequence identities in amino acids were 99.4%, 78.8%, and 79.5% for MLP-PG1/MLP-GR1, MLP-PG1/ MLP-GR3, and MLP-GR1/MLP-GR3, respectively. A conserved motif for Bet v 1 allergen was found from amino acids 122 to 156. The hydropathy plot for MLP-GR3 toward those for MLP-PG1 and MLP-GR1 was different in the C-terminal region (Supplemental Fig.  S5). MLP-GR3 from around amino acids 120 to 140 in part of the conserved motif showed more hydrophobicity than the corresponding regions of MLP-PG1 and MLP-GR1. No signal sequences related to secretion were found in these three MLPs.

Expression Analysis of MLP Genes in C. pepo
The expression levels of the three MLP genes in the roots of eight cultivars were determined relative to those of PG. MLP-PG1 and MLP-GR1 were not separated with the primers used due to their high sequence identity. The expression levels of MLP-PG1 tended to decrease with an increase in a cultivar's ability to take up dioxins and dioxin-like compounds ( Fig. 5A; Matsuo et al., 2011). In contrast, the expression levels of MLP-GR3 in C. pepo ssp. pepo, including cv Black Beauty (BB), GR, cv Raven, and MG, were much higher than those of PG (Fig. 5B). GR showed the highest relative expression levels. Significant correlations between the relative MLP-GR3 expression levels and band intensities of 17-kD proteins (r = 0.821, P , 0.01; Fig

Binding Activities of Recombinant MLPs to PCBs
The binding activities of recombinant MLPs to PeCB106 attached to magnetic beads were investigated (Fig. 6). Recombinant MLP-GR3 was detected on the beads prepared with 25 mM 4OH-PeCB106, whereas recombinant MLP-PG1 and MLP-GR1 did not show any bands. Furthermore, a larger amount of recombinant MLP-GR3 bound to beads prepared with 50 mM 4OH-PeCB106. The initial amount of recombinant MLPs applied to the beads was similar among the three recombinant MLPs, and there were no bands at the last washing step (Supplemental Fig. S7).

DISCUSSION
Previous studies have shown that translocation of POPs from roots to shoots is responsible for their high accumulation in the Cucurbitaceae family. Although no differences in the root concentrations of POPs were observed between Cucurbitaceae and Poaceae, significant differences were found in their shoots (Lunney et al., 2004). Another study found that the accumulation profiles of chlordane in scions of cucumber were reflected by the uptake ability of rootstocks of C. pepo (Mattina et al., 2007). Therefore, xylem sap is thought to be the primary pathway for POP translocation. However, because high transpiration rates were not related to high accumulation in shoots (Murano et al., 2010a), some material(s) with binding activity to POPs within xylem sap might be important. Highly hydrophobic compounds tend to accumulate in the lower parts of C. pepo stems (Aslund et al., 2008), indicating that xylem vessels may trap them on their inner surfaces. Xylem vessels are usually lignified, and lignins can form a complex with hydrophobic compounds. Therefore, xylem sap must be able to solubilize POPs before they can be translocated to the upper parts of plants. We found a significant positive correlation between BCFs for PCBs and solubilization activities of xylem sap toward the hydrophobic compound pyrene (Fig. 1). Based on this and previous results (Murano et al., 2010a), it seems that POPs adsorbed on xylem vessels may be desorbed by xylem sap containing some proteins, and then POPs are translocated into the upper parts of plants and accumulated there.
Satoh and colleagues reported that xylem sap of cucumber plants contains Gly-rich proteins (Sakuta and Satoh, 2000), xylem sap protein30 (Oda et al., 2003), and chitinase (Masuda et al., 2001). In this study, significant positive correlations were observed between the amounts of 17-kD proteins and BCFs of PCDDs, PCDFs, and PCBs (    Supplemental Fig. S2C). Thus, despite the structural differences among POPs, all the tested compounds could be translocated and accumulated by xylem sap proteins; hydrophobicity is a major factor. The 17-kD proteins in xylem sap bound PeCB106 (Fig. 3), suggesting that they related to the efficient translocation of POPs in C. pepo plants.
According to the melon unigene library, the 17-kD proteins showed similarity to MLPs in a subfamily with Bet v 1, the birch (Betula spp.) pollen allergen. Bet v 1 proteins are members of the ubiquitous Pathogenesis-Related10 (PR-10) family and are acidic proteins of 15 to 18 kD (Radauer et al., 2008). The most distinctive feature of the Bet v 1 conformation is a large solvent-accessible hydrophobic cavity, which may function as a ligandbinding site for hydrophobic compounds such as fatty acids, plant hormones, and flavonoids (Mogensen et al., 2002). Three-dimensional structures of birch Bet v 1 and yellow lupine (Lupinus luteus) PR-10 revealed that the former formed a complex with the hydrophobic compound deoxycholate and the latter with plant hormones in the brassinosteroid (Markovi c-Housley et al., 2003) and cytokinin (Fernandes et al., 2009) groups. The a3helix located in the C-terminal region at around amino acids 130 to 155 of the birch Bet v 1 contains amino acids related to ligand binding (Gajhede et al., 1996;Markovi c-Housley et al., 2003). Therefore, the hydrophobic property in this region of MLP-GR3 may give a higher affinity toward hydrophobic compounds than those of MLP-PG1 and MLP-GR1 (Kyte and Doolittle, 1982). It was reported that the melon MLP gene MEL7 and the cucumber MLP gene Csf2 were expressed in roots and stems and fruits during ripening (Aggelis et al., 1997;Suyama et al., 1999). The melon phloem-sap proteome revealed that MLPs were detected in phloem sap from young petiole and stem of plants infected with Cucumber mosaic virus (Malter and Wolf, 2011). Blanca et al. (2011) recently reported that there were more than 30 expressed sequence tags of MLP genes in the unigene library of C. pepo. The amino acid sequence WSVVFEK from 17-kD proteins of xylem sap of GR was not completely conserved in the deduced amino acid sequences from the cloned genes, whereas the others were detected. It may be due to the coexpression of MLP multigenes in C. pepo. The uptake of POPs is not likely a necessary function in plants; rather, some MLPs that serve other functions in the Cucurbitaceae family happen to bind and deliver them to shoots.
It is suggested that the 17-kD proteins mainly consisted of MLP-GR3, and the differential expression of MLP-GR3 caused cultivar-specific uptake and accumulation of dioxins and dioxin-like compounds ( Fig. 5;  Supplemental Fig. S6). In a three-dimensional model, MLP-PG1 and MLP-GR1 showed structures similar to that of MLP-GR3 (data not shown), because about 80% of the sequence is identical. However, only recombinant MLP-GR3 bound PeCB106 (Fig. 6). These data indicated that MLP-GR3 recognized and interacted with PeCB106 with greater affinity than the other proteins.
Based on the results of this study, the high transport and accumulation abilities of C. pepo ssp. pepo plants are due to the following mechanism. POPs taken up into roots are accumulated in endodermis and pericycle cells, which are located inside of endodermis cells (data not shown). MLP-GR3 binds POPs in those cells because the gene for PR-10, which shows structural similarity to MLPs, is reported to be expressed in pericycle cells (Dembinsky et al., 2007). The MLP-GR3-POP complex is then secreted and transferred through xylem sap into aerial parts. Free POPs are adsorbed onto root tissues and xylem vessels, resulting in high concentrations of POPs in roots and stems. The complex distributed into each organ is deposited as the complex or free POPs dissociated from them (White et al., 2003). However, there may be another possibility: that the MLP-GR3-POP complex translocates into aerial parts through phloem or a latex duct. This mechanism is likely applicable to many other highly hydrophobic pollutants, even those that cannot be taken up by other plant families.

CONCLUSION
This study identified the proteins in xylem sap that play a crucial role in the transport of POPs in the Cucurbitaceae family. Our findings clearly show a high affinity of MLPs toward hydrophobic compounds and that the high expression of MLP genes resulted in the high accumulation of POPs that cause crop contamination. Understanding this mechanism should help to control the risk of contamination with highly hydrophobic pollutants in three ways: (1) preventing important vegetable crops such as cucumber, watermelon, and melon from unexpected contamination with POPs; (2) selecting appropriate plants from the Cucurbitaceae family to serve as hosts with efficient phytoremediation capacities for POPs; and (3) expressing the genes responsible for the efficient uptake of POPs in other plant species to endow them with a high phytoremediation ability, with the hope of overcoming inappropriate environmental conditions for the Cucurbitaceae family.

Plant Materials
Cucurbita pepo ssp. ovifera cultivars PG, Starship, Sunburst, Sunray, and Zephyr and C. pepo ssp. pepo cultivars BB, GR, Raven, and MG were used in this study. With the exception of BB, seeds of C. pepo plants were purchased from Johnny's Selected Seeds. The seeds of BB were purchased from Tanenomori. In the previous study (Matsuo et al., 2011), these nine C. pepo cultivars were cultivated in pots supplemented with soil contaminated with dioxins and dioxin-like compounds collected from the periphery of an incinerator. There are differences in BCFs of dioxins and dioxin-like compounds, which are calculated by dividing the concentration of PCDDs, PCDFs, and PCBs in aerial parts including leaves and stems by its concentration in soil, among these cultivars. These cultivars also show different 1,1-dichloroethylene (DDE) phytoextraction efficiencies (%), which are the sum of total DDE in vegetative tissues expressed as a percentage of DDE present in soil when they were cultivated in a DDE-contaminated field (White et al., 2003).

Collection of Xylem Sap
The seeds of C. pepo plants were immersed in tap water for 1 d at 4°C. They were then sown and cultivated in vermiculite or soil for about 10 d. Seedlings were transferred to a field at Kobe University and cultivated for 2 to 3 months. The aerial part of each plant was cut about 5 to 15 cm above the soil level and removed, and a conical flask was fitted to the root side of the stem. The flask was wrapped in aluminum foil and placed on ice to protect xylem sap from light and heat, and xylem sap was collected for 6 to 14 h. Xylem sap prepared by this method has the possibility to be contaminated by phloem sap, even though phloem sap is collected from the shoot side of the cut stem (Buhtz et al., 2004). Furthermore, since phloem vessels are sealed via callose secretions during the few minutes after cutting of the stem (Greenwood et al., 2011), phloem sap is diluted by the large volume of xylem sap relative to that of phloem sap during a long-term collection. Xylem sap was collected from several plants of each cultivar and used as separate batches for the experiments. The collected volumes of xylem sap from one plant were about 50 to 500 mL during this period. The collected xylem sap was filtered through cellulose filters (no. 2; Toyo Roshi Kaisha). The xylem sap was stored at -30°C until use.

Solubilization Activities of Xylem Sap toward Pyrene
A total of 50 mg of Tenax (HR C206X; Toho Tenax), an adsorbent material for hydrophobic compounds, was added to 10 mL of 5 mM pyrene containing 5% dimethyl sulfoxide in a glass tube and incubated for 48 h at 25°C. The amount of pyrene adsorbed on Tenax was calculated by measuring the concentration of pyrene in the supernatant with a fluorescence spectrophotometer (F-2500; Hitachi High-Technologies), with the fluorescence excitation set at 340 nm and emission set at 382 nm. After discarding the supernatant completely, 10 mL of water or xylem sap was added to the glass tube containing pyrene-adsorbed Tenax, and the tube was incubated at 25°C. In some experiments, xylem sap was heated for 5 min at 98°C to denature proteins. To measure the concentration of pyrene in the supernatant, 1 mL of each solution was sampled after incubation for 4 h. The concentration of pyrene in the supernatant was measured with a fluorescence spectrophotometer as described above. Pyrene solubilization activities were calculated with the following equation: pyrene solubilization activity (%) = (amount of pyrene in the supernatant/amount of pyrene adsorbed on Tenax) 3 100.

SDS-PAGE for Xylem Sap
A 5-mL volume of xylem sap of C. pepo plants was concentrated to approximately 100 mL by ultrafiltration using a Centricut Ultramini ultrafiltration device (M r cutoff = 10,000; Kurabo Industries). Concentrations of the proteins in the concentrated xylem sap were measured by using the method of Bradford (1976). The samples containing the same amount of total proteins were prepared by mixing with a sample buffer solution with a reducing reagent (Nacalai Tesque) and subjected to SDS-PAGE on a 15% acrylamide gel. Protein bands were detected by incubation of gels in Coomassie Brilliant Blue Stain One (Nacalai Tesque). The intensity of bands at 17 kD was quantified by using ImageJ 1.440 (National Institutes of Health).
A 17-mL volume of xylem sap from MG plants was concentrated by using a Vivaspin 20 ultrafiltration membrane (M r cutoff = 10,000; GE Healthcare Bio-Sciences) to about 440 mL. This solution was centrifuged at 20,700g for 30 min to remove insoluble debris. Then, 100 mM KCl buffer was added to bring the volume to 1.44 mL (the final concentration of xylem sap proteins was 40 mg mL -1 ). This solution was added to 0.5 mg of 4OH-PeCB106-bound beads or control beads (no 4OH-PeCB106). After dispersion of the beads, the solutions were incubated for 8 h at 4°C with continuous rotation at 25 rpm. After the binding reaction, these solutions underwent magnetic separation, and the supernatants were removed. Next, 200 mL of 100 mM KCl buffer was added to the remaining beads, and the beads were dispersed for washing. This washing step was repeated eight times. After extensive washing, 35 mL of 100 mM KCl buffer and 7 mL of the sample buffer solution were added to these beads, and the beads were heated for 5 min at 98°C. After magnetic separation, the supernatants were subjected to SDS-PAGE on a 15% acrylamide gel, and the gel was stained using the Silver-Stain MS kit (Wako).

Identification of the Proteins in Xylem Sap
A 2-mL volume of xylem sap from GR plants was concentrated to about 100 mL by ultrafiltration using Centricut Ultramini. The concentrated xylem sap samples were subjected to Tricine-SDS-PAGE, and the gel was stained with Coomassie Brilliant Blue, as described above.
Bands of 17-kD proteins from GR were excised. After treatment of the bands with trypsin at 35°C for 20 h, partial amino acid sequences were determined at APRO Science by reverse-phase liquid chromatography with a Symmetry C18 3.5-mm column (1.0 3 150 mm; Nihon Waters) using a mobile phase of 0.1% (v/v) trifluoroacetic acid in 2% (v/v) acetonitrile and 0.09% (v/v) trifluoroacetic acid in 90% (v/v) acetonitrile with a flow rate of 50 mL min -1 at room temperature. Individual components were detected at 210 and 280 nm.
Homology searches were conducted for each determined partial amino acid sequence in the Cucurbit Genomics Database (http://www.icugi.org/). Hydropathy plots were constructed by the method proposed by Kyte and Doolittle (1982) based on a window size of seven.

Cloning of MLP Genes
Total RNAs extracted from the roots of PG and GR plants grown in soil for 40 d at 26°C under a 16-h-light/8-h-dark cycle were used for complementary DNA (cDNA) synthesis using the ThermoScript RT-PCR System (Life Technologies) according to the manufacturer's instructions. The primers S-1 and AS-1 (Supplemental Table S1) corresponding to the partial amino acid sequences were used for RT-PCR.
We performed 59 and 39 RACE-PCR by using the SMART RACE cDNA Amplification kit (Takara Bio) with primers AS-2 and S-1, respectively. The primers 59-UTR and 39-UTR, which included the 59 and 39 untranslated regions, respectively, were used for amplification of the whole sequences of MLP genes. MLP genes (accession nos. AB753855 [MLP-PG1], AB753856 [MLP-GR1], and AB753857 [MLP-GR3]) were subcloned into the pT7Blue vector (Merck). A motif search of the deduced amino acid sequences of MLPs was carried out in the BLOCKS database (http://www.genome.jp/tools/motif/).

Expression Analysis of MLP Genes in C. pepo
After treatment with DNase I, total RNAs extracted from the roots of nine cultivars of C. pepo grown in soil for 40 d at 26°C under a 16-h-light/8-h-dark cycle were used for cDNA synthesis with ReverTra Ace qPCR RT Master Mix (Toyobo) according to the manufacturer's instructions. Thunderbird SYBR qPCR Mix (Toyobo) was used for quantitative PCR (Light Cycler 480 II; Roche Applied Science) by PG1/GR1-S and PG1/GR1-AS, GR3-S and GR3-AS, and Actin-S and Actin-AS for the detection of MLP-PG1/MLP-GR1, MLP-GR3, and Actin, respectively. Reaction conditions were 1 min at 95°C followed by 40 cycles of 15 s at 95°C, 30 s at 60°C, and 5 s at 95°C, and then 1 min at 65°C. Subsequently, samples were heated to 97°C and then cooled to 40°C.

Expression of MLP Genes in Recombinant Escherichia coli
The MLP genes with NcoI and XhoI sites at the 59 and 39 ends were amplified with the primers 59-NcoT and 39-XhoCT for MLP-PG1 and 59-NcoT and 39-XhoGA for MLP-GR1 and MLP-GR3, with MLP genes in the pT7Blue vector as a template. After the insertion of amplified fragments into the pT7Blue vector, fragments digested with NcoI and XhoI were cloned into the pET-28b (+) vector (Merck), resulting in pET28MLPPG1, pET28MLPGR1, and pET28MLPGR3. E. coli Rosetta-gami 2 (Merck) was used as a host for expression of the three MLP genes by introduction of these plasmids, and MLPs were produced as hexa-His-tagged proteins.

Purification of MLPs Produced in Recombinant E. coli
Recombinant E. coli cells harboring each expression plasmid were inoculated in Luria-Bertani medium containing 100 mg mL -1 kanamycin and 50 mg mL -1 chloramphenicol at 37°C. When the optical density at 600 nm reached approximately 0.8, isopropyl-b-D-thiogalactopyranoside was added to the culture medium at a final concentration of 0.1 mM to cause transcription, and incubation was continued for an additional 20 to 24 h at 20°C. The cells containing the expressed protein were harvested and disrupted by sonication. The insoluble components were removed by centrifugation at 26,700g for 20 min at 4°C. The clear supernatant was loaded onto a HiTrap Chelating HP column (GE Healthcare Bio-Sciences) equilibrated with 20 mM Tris-HCl buffer (pH 7.0) containing 500 mM NaCl and 20 mM imidazole. After washing with the same buffer, the proteins were eluted by imidazole. Fractions containing MLPs were dialyzed in 20 mM Tris-HCl buffer and loaded onto an ion-exchange column chromatograph (Resource Q; GE Healthcare Bio-Sciences). Fractions containing MLPs were combined and dialyzed in 20 mM Tris-HCl buffer (pH 7.0).

Binding Activities of MLPs to PCBs
Magnetic beads binding 4OH-PeCB106 in 25 and 50 mM solutions were prepared as described above. We added 200 mL of MLP solution dissolved in 100 mM KCl buffer at 0.25 mg mL -1 to the beads and mixed for 4 h at 4°C with continuous rotation. The washing step described above was repeated eight times. Next, 35 mL of 100 mM KCl buffer and 7 mL of the sample buffer (Nacalai Tesque) were added and applied to a 15% acrylamide gel for SDS-PAGE after heating for 5 min at 98°C. Silver staining was conducted as described above.

Supplemental Data
The following materials are available in the online version of this article.
Supplemental Figure S1. Pyrene solubilization activities of xylem sap from C. pepo cultivars.
Supplemental Figure S2. Correlation between intensity of the bands at approximately 17 kD and BCFs of dioxins and dioxin-like compounds.
Supplemental Figure S3. SDS-PAGE of xylem sap proteins from MG plants eluted at the last washing step.
Supplemental Figure S4. Alignment of DNA sequences of MLP genes.
Supplemental Figure S6. Correlation between band intensities of 17-kD proteins and relative MLP-GR3 expression levels, and correlation between relative BCF values of dioxins and dioxin-like compounds and relative MLP-GR3 expression levels.
Supplemental Figure S7. Specific binding of MLPs to PeCB106.
Supplemental Table S1. The sequences of primers used in this study.