Contributed by Clarence A. Ryan, Jr.
The isolation to homogeneity of the 160-kDa systemin cell-surface receptor (SR160) from plasma membranes of suspension cultured cells of Lycopersicon peruvianum is reported. The purification procedure resulted in recovery of 13 μg of pure receptor protein, representing an 8,200-fold purification. Gel blot analyses using SR160-specific antibodies confirmed that a cross-reacting protein in the membranes of suspension-cultured cells comigrates with both the native and a deglycosylated form of the radiolabeled receptor. Internal amino acid sequences of the purified protein facilitated the isolation of a cDNA clone encoding the 160-kDa receptor. The identity of the encoded protein as SR160 was further confirmed by a comparison of its sequence with a mass spectral fingerprint of the SR160 protein. The deduced amino acid sequence of SR160 revealed that it is a member of the leucine-rich repeat (LRR) receptor kinase family, closely related to the brassinolide receptor kinase, BRI1.
Polypeptide hormones have recently been identified in plants as key regulatory signals for defense, reproduction, growth, and development (1–8). As with animal polypeptides hormones, plant polypeptide signals act as extracellular ligands that are functionally perceived by high-affinity plasma-membrane-bound receptors (9–14). Systemin, an 18-aa polypeptide, was identified in 1991 as the primary signal for the systemic activation of defense genes in leaves of wounded tomato plants (1). Systemin causes a cascade of intracellular signaling events leading to the release of linolenic acid from membranes, and its conversion to oxylipin molecules that signal defense gene expression (15–20). These events are initiated by the release of systemin at wound sites and its interaction with a membrane-bound receptor (9, 10, 21). A photoaffinity analog of systemin was used to identify a 160-kDa cell-surface receptor protein in membranes of Lycopersicon peruvianum suspension cultured cells that possessed characteristics of a systemin receptor (10). We report here that the radioactive photoaffinity label provided a marker for purification of the receptor protein to homogeneity. Amino acid sequence analysis allowed the cloning of the receptor cDNA from a L. peruvianum cDNA library and the identification of the cDNA-encoded protein as a leucine-rich repeat (LRR) receptor kinase with high amino acid identity and domain similarities to the BRI1 receptor kinase from Arabidopsis.
Suspension cultures of L. peruvianum were grown in media under conditions previously described (10). One-liter cultures were grown in 4-liter Fernbach flasks initially inoculated with 50 ml of 1-week-old cells grown in 125-ml Erlenmeyer flasks. The cells were grown for 1 week under the same growth conditions and media reported (10).
N-[4-(p-azidosalicylamido) butyl]-3′[2′-Cys-3,Ala-15-systemindithiol] propionamide, or azido-Cys-3,Ala-15-systemin, was prepared and used for iodinations as described (10), except for the following modifications. Ten nanomoles of azido-Cys-3,Ala-15-systemin were iodinated in 500 μl of 0.1 M sodium monophosphate, pH 7.6, by adding 50 mCi Na125I (2200 Ci/mmol/0.1 mCi/μL; DuPont; 1 Ci = 37 GBq) and three iodobeads (Pierce), prerinsed in the same buffer. The reaction was allowed to proceed for 5 min, agitated, and allowed to incubate for an additional 3 min. After allowing the iodobeads to settle, the reaction solution was separated from the beads with a syringe. The iodobeads were rinsed with 1 ml of 0.1% trifluoroacetic acid (TFA) and the solution was recovered as above. The radiolabeled peptide was purified by C18 RP-HPLC on a 0–50% acetonitrile/90 min gradient in 0.1% TFA. The fractions (1 ml) were analyzed by gamma counting (Isodata 2020, Isodata, Palatine, IL) and the three peak tubes were pooled and stored in the dark at 4°C. The procedure was performed twice, yielding a final recovery of ≈2 mCi of 125I-azido-Cys-3,Ala-15-systemin (2,200 Ci/mmol specific activity).
A bulk in vivo procedure was used to react the receptor on the surface of the cells with a photoaffinity label. For labeling, 1 liter of 7 day suspension cells treated for 15 h with 50 μM methyl jasmonate was introduced into a 3.5-cm-deep, 33.0-cm diameter plastic dish cut from the base of a 32-gallon waste bucket. One mCi of 125I-azido-Cys-3,Ala-15-systemin was added to the cells under a red light (to prevent activation of the azide moiety) while on an orbital shaker rotating at 90 rpm. After 3 min, a UV-B lamp (F15T8.UV-B, 15 W; Ultraviolet Products) positioned 8.0 cm above the suspension cultured cell medium was switched on to allow irradiation of the rotating cells. After a 10-min irradiation period, the cells were filtered on Whatman paper in a Buchner funnel aided by a vacuum and washed with 2 liters of 4°C H2O. After washing, the cells were transferred into 500 ml of extraction buffer (21) containing 3.0% β-mercaptoethanol (β-Me) and 3 ml of protease inhibitor mixture for plant extracts (Sigma) on ice. The cells were homogenized in a Parr Cell Disruption Bomb (Parr Instrument, Moline, IL) under 1,500 psi (nitrogen), and the cellular debris was pelleted by centrifugation at 10,500 × g for 15 min at 4°C. The labeling and homogenization was repeated using a second liter of cells. The combined supernatants from the 10,500 × g centrifugations were termed the “crude extract” and the membranes were subsequently sedimented from the extract by ultracentrifugation at 100,000 × g for 3 h at 4°C, employing a Beckman Ti 45 preparative ultracentrifuge rotor (600 ml total volume). The cell debris obtained from the 10,500 × g centrifugation was resuspended in a total of 500 ml of extraction buffer containing β-Me and inhibitors, rehomogenized in the Parr Cell Bomb, and the supernatant was recovered and ultracentrifuged as above. Sixteen membrane pellets containing a total of 1,064 mg of protein were obtained from 1,500 ml of crude extract containing 2,730 mg of protein. The pellets were stored at −20°C for at least 1 month.
Preparative slab electrophoresis was performed with a 0.6 cm × 14.0 cm × 8.5 cm 7.5% acrylamide gel that was allowed to polymerize for 1 h. A 4% stacking gel 2.0 cm deep was added and polymerized for 1 h. Each membrane pellet (≈65 mg protein) was solubilized by adding 2 ml of 10% SDS/3 ml Laemmli sample buffer containing 5.0% β-Me. Solubilization was aided by placing the mixture in a 50-ml conical tube with stirring in a boiling water bath. The solution was cooled on ice for 15 min and loaded on the preparative gel. Electrophoresis was performed overnight (15 h) at room temperature at a constant 38 V and terminated 30 min after the dye front exited the gel. The gel was sliced in half vertically and each half cut horizontally into 0.5 cm × 0.25 cm segments by using a long-blade gel slicer. The gel slices were placed in glass vials and analyzed for radioactivity by gamma counting. The fractions containing radioactivity corresponding to the 160-kDa protein were pooled.
The gel slices were macerated to allow for protein extraction. Maceration was as described using a 20-ml syringe (22), and the material was extracted in a buffer containing 0.1 M Mes, pH 4.5, 1 mM MnCl2, 1 m CaCl2, and 0.5 M NaCl by end-over-end mixing at room temperature for 30 min. The gel material was centrifuged at 2,000 × g for 5 min at 4°C and the supernatant containing the extracted proteins was recovered. The gel was re-extracted with a second 50 ml of extraction buffer. The extracted proteins from the gels were pooled (400 ml total) for further purification.
Concanavalin A (Con A) Sepharose affinity resin (Pharmacia) was equilibrated in the same buffer used for gel extraction. To 50-ml aliquots of the pooled proteins, 75 μl of Con A Sepharose was added and incubated at 4°C with constant rotation on an orbital shaker (150 rpm) for 48 h. The Con A resin quantitatively bound the radioactive proteins. The resin was pelleted by centrifugation at 2,000 × g for 10 min and washed three times with 50 ml of 0.5 M α-d-mannopyranoside/0.1 M borate, pH 6.5, followed by three washes with distilled H2O. The Con A resin was centrifuged at 12,000 × g for 2 min to remove excess water, and the protein was solubilized by boiling in 300 μl of 0.1% SDS for 5 min. The resin was pelleted by centrifugation for 5 min and the supernatant was recovered. The resin was dispersed in 1 ml of water and boiled for an additional 5 min, microcentrifuged for 5 min, and all of the supernatant were pooled. The pooled solution was frozen, lyophilized, and resuspended in 500 μl of Laemmli sample buffer.
The final purification step was performed by gel electrophoresis on a 11.0 × 14.0 × 0.15 cm resolving gel with a 7.5% acrylamide concentration and a 4.0%, 4.0-cm stacking gel and polymerized overnight. The sample obtained from the Con A affinity step was boiled, loaded to the gel and subjected to SDS–gel electrophoresis for 6 h at 4°C with 175 V (constant voltage). After electrophoresis, the gel was cut vertically into 3-cm strips and each strip was sliced horizontally on a gel slicer to a width of 0.1 cm. The slices were analyzed for radioactivity by gamma counting, and the active fractions were pooled, macerated with a 3-ml syringe as described above, and extracted in 10 ml of H2O. The resultant purified receptor was frozen, lyophilized, resuspended in 100 μl of H2O, and stored indefinitely at −20°C.
Purified receptor protein was separated on a 7.5% SDS–polyacrylamide minigel and digested in gel with lysyl endopeptidase (Wako, Richland, VA). The resultant peptides were analyzed by matrix-assisted laser desorption ionization–time-of-flight (MALDI-TOF) mass spectrometry. Tryptic digestion, HPLC purification of peptides, and amino acid sequencing was performed by the Rockefeller University Protein/DNA Technology Center. Analytical SDS/PAGE was performed using a MiniProtean gel apparatus (Bio-Rad) and stained with Coomassie brilliant blue. Gels were subsequently dried between cellophane sheets and analyzed for radioactivity by phosphorimaging (Bio-Rad). Protein determinations were made with both Bradford (Bio-Rad) and BCA reagents (Pierce).
An L. peruvianum suspension cell cDNA library was made according to manufacturer's protocol (CLONTECH) and the library was double screened with two probes. One probe was made from a PCR product generated by using primers TGCTAACAACACCAATTGGAAGCT and ACTTTGCAATAACCAAGAAGAG (Arabidopsis BRI1) and an Arabidopsis cDNA library, and the second probe was made from a PCR product by using primers GGCACGTCTAGCTGGAAGCTT and TCTTGCAATATCCGAGCAGA (Tomato Est BE357576) and a Lycopersicon esculentum cDNA library. The plaques hybridizing to both probes were isolated and sequenced (GenBank accession no. AY112661).
The purified protein and the protein digested with PNGase F (Glyco, Novato, CA) were subjected to electrophoresis on 7.5% mini gels and blotted to nylon. The blot was probed with preimmune and immune serum generated to the synthetic peptide (C)LSINAAFEKPLR by Pocono Rabbit Farm and Laboratory (Canadensis, PA). Immunodetection was performed according to manufacturer's protocol for the SuperSignal West Femto kit (Pierce).
Genomic DNA was isolated from L. peruvianum suspension cultured cells and 5 μg of DNA were digested for 8 h with either SpeI or NcoI and XbaI together. Digested DNA was separated on 0.8% agarose/TBE gel and blotted to nylon. A PCR product of ≈450 bp was produced from the SR160 cDNA by using primers TGGTGAATGTTCAAGTTTGGA and AGCCTGCAAGTACATCAACTC. A 32P-labeled probe was generated from the PCR product by using DECAprime II labeling kit (Ambion) and hybridized at 65°C overnight with the membrane. Washes were: 2× SSC for 30 min at 22°C (1× SSC = 0.15 M sodium chloride/0.015 M sodium citrate, pH 7); 2× SSC/1%SDS for 45 min at 65°C (twice) and then exposed for 72 h to x-ray film.
The isolation of the systemin receptor was facilitated by photoaffinity radiolabeling of the suspension cultured cells in bulk quantity. Only the 160-kDa protein was labeled, which provided a method of specifically tagging the protein for identification throughout the purification process. Furthermore, by using the radioactivity as an assay, the protein could be isolated under denaturing conditions, allowing for identification by analytical SDS/PAGE/phosphorimaging and quantification by gamma radiation detection.
Membranes were isolated from 2 liters of suspension-cultured cells labeled with 125I-Cys-3,Ala-15,systemin. The membrane proteins were solubilized and separated electrophoretically by preparative SDS/PAGE, in which up to 65 mg of protein could be loaded per gel, and the radiolabeled protein was quantitatively recovered using a syringe maceration-extraction technique (22). A Con A binding buffer was used to extract the proteins from the gel and the extract was directly applied to Con A resin, which specifically and quantitatively bound the protein. Elution of the protein, however, could only be achieved by boiling the resin in at least 0.1% SDS. The conditions ordinarily used for elution of proteins from Con A resin, α-d-mannopyranoside, was unable to displace the receptor, suggesting that interactions other than with the carbohydrate moiety may be involved in binding. Other detergents, including Triton X-100, CHAPS {3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate}, and octyl-glucoside, at concentrations at up to 5.0%, were also unable to elute the protein, even if boiled. The purification steps were monitored by SDS/PAGE and are shown in Fig. 1.
Electrophoretic analysis of the systemin receptor during purification. (A) Proteins were monitored by staining with Coomassie brilliant blue. (B) The radiolabeled receptor was detected in gels by phosphorimaging.
As a final step for the isolation of the 160-kDa protein, eluted proteins from the Con A resin were separated on polyacrylamide gels, and the pure protein was recovered as described in Materials and Methods. The yield was 13 μg of protein, with an 8,200-fold purification representing 3.9% of the initially radiolabeled receptor (Table 1). The purified receptor was initially digested with lysyl endopeptidase to yield a peptide mass fingerprint (Fig. 2A); however, the mass information did not identify any corresponding proteins in available databases. Subsequently, the protein was subjected to tryptic digestion and the fragments were separated by HPLC and sequenced by Edman degradation. The amino acid sequences of four major peaks are shown in Fig. 2B. Although no sequence matches were found in tomato EST databanks, homologous sequences were identified within the LRR receptor kinase BRI1 from Arabidopsis. However, the sequences of only peptides I and II (Fig. 2B) matched identically with sequences in BRI1. Two 200-bp oligonucleotide probes were prepared corresponding to the region containing peptide I (see Materials and Methods) and used to screen an L. peruvianum library. A cDNA was identified in the L peruvianum library from which a protein was deduced that contained all four of the peptides (cf. Fig. 2B) with 100% identity (Fig. 2C). Furthermore, the mass spectral fingerprint of the purified protein (Fig. 2A) coincided with masses deduced from the cDNA sequence (Fig. 2C), confirming that the cDNA coded for the isolated protein. Twelve of fourteen predicted sequence masses were identified, whereas a mass search of the Arabidopsis BRI1 gave no positive matches.
Protocol for the purification of the systemin receptor from L. peruvianum
Structural analysis of SR160. (A) A representative mass spectrum obtained from proteolytic digestion by lysyl endopeptidase of the purified, radiolabeled SR160. (B) Amino acid sequences of four peptides recovered from tryptic digests of the purified 160-kDa protein. (C) The deduced amino acid sequence of SR160 cDNA. The four tryptic peptides from (B) are underlined, and 12 of the 14 peptide masses obtained by mass spectral fingerprinting are indicated in bold with their mass number.
Subjecting SR160 to enzymatic hydrolysis with PNGase F, an N-linked carbohydrase, shifted the molecular weight by about 15 kDa (Fig. 3A), indicating that the isolated protein was a glycoprotein. Mass spectral analysis of the smaller species gave an identical fragmentation profile as the parent SR160, indicating that the backbone of the deglycosylated protein coincided with the parent glycosylated protein. Antibodies were prepared in rabbits against synthetic peptide I (Fig. 2B) and used in gel blots to correlate immunogenicity with radioactively labeled SR160 before and after digestion with PNGase F. The immunological reactivity corresponded with the shift of radioactivity (Fig. 3B), confirming that the deglycosylated form of the receptor retained immunological identity.
Deglycosylation of SR160. (A) The systemin receptor, SR160, was radiolabeled with 125I-azido-Cys-3,Ala-15-systemin, and deglycosylated with PNGase F. The native and deglycosylated forms of the radiolabeled protein were separated by SDS/PAGE and detected by phosphorimaging. (B) Samples of radiolabeled SR160 untreated or deglycosylated with PNGase F were separated by SDS/PAGE and probed with preimmune or immune serum raised to a synthetic peptide derived from the receptor sequence. The gel blots were visualized by chemiluminescence and the position of radioactively labeled SR160 was determined by phosphorimaging of the gel blot.
The amino acid sequence of the systemin receptor (Fig. 4) contains a putative amino acid signal sequence, a 3 heptad leucine zipper, 25 LRR repeats interrupted by a 68-aa island between the 21st and 22nd LRR, a transmembrane domain, and a Ser/Thr protein kinase domain. The sequence exhibits a high percentage of amino acid identity with the brassinolide receptor, BRI1, from Arabidopsis, and a closely related receptor, OsBRI1, from rice (23, 24). A comparison of the deduced amino acid sequences of SR160 with those of BRI1 are shown in Fig. 4. Comparing amino acid sequences, both receptors have all of their domains in common. The transmembrane domain and the kinase domain are the most highly conserved domains between the two receptors, exhibiting 83% and 90% identity, respectively. There are 18 cysteines in the putative extracellular domain of BRI1 and 16 in SR160, indicating that there is one less disulfide pair in SR160. A pair of half cysteines occur N-terminal to the first LRR of the 21 LRRs preceding the 68-aa island (Fig. 4). These N-flank cysteine motifs are shorter than the N-flank cysteine motifs found in animal LRR regions (25, 26). A similar cysteine pair (C-flank) follows the final LRR of the protein.
Comparison of the amino acid sequences and domain structures of SR160 with the BRI1 sequence from Arabidopsis. The amino acids identical between the two proteins are highlighted with a black background. Putative domain structures are indicated above the SR160 sequence. Thin overline, signal sequence; thick overline, 3 heptad leucine zipper; asterisk, N-flank and C-flank cysteines; chevron, the first amino acid of each LRR; dotted overline, 68-aa island; dashed overline, transmembrane domain; boxed, the Ser/Thr kinase.
Mutations in BRI1 that inactivate the receptor have been found in the 70-aa island within the LRR region (23, 27). Loss of function has been attributed to the loss of brassinosteroid (BR) binding or possibly conformational changes (28, 29). The three mutated glycines in BRI1 that account for loss of function are conserved in the SR160 sequence. Eventual site-directed mutagenesis of SR160 at these positions may reveal whether these residues are also required for a functional systemin receptor.
The high percentage of conservation of amino acids in the kinase domain between SR160 and BRI1 indicates that the two receptors may have some downstream intracellular signaling components in common. An early event in BR signaling is autophosphorylation of BRI1, and a similar event is predicted for SR160, based on the similarity of its kinase domain to that of BRI1 (29, 30).
The largest differences between the two receptor sequences are in the extracellular LRR regions and may reflect the structural conformations that are required to interact in a highly specific way with different ligands. To investigate possible functional relationships between BRI1 and SR160, we performed competition experiments between brassinolide (BL) and systemin, using the alkalinization assay, and between BL and the photoaffinity label (10). At BL concentrations up to 1.0 mM, no competition was observed in the systemin-induced medium alkalinization of tomato suspension cultured cells or with 125I-azido-Cys-3,Ala-15-systemin labeling of SR160 (data not shown). This does not, however, preclude the existence of independent binding sites within the same protein. The percentage of identity between SR160 and BRI1 is higher than between BRI1 and any other LRR receptor kinase from Arabidopsis. This indicates that SR160 may have recently evolved from a common ancestral BRI1 gene, because BL signaling is ubiquitous in plants and systemin signaling in known only in the Solonaceae family. The direct interaction of the small organic ligand BR with BRI1 has been hypothesized (29), but the possibility remains open that BL may bind to a polypeptide to facilitate receptor binding, or that a polypeptide may bind to BRI1 to accommodate BL binding. Such scenarios may allow a variety of small molecules to regulate physiological processes through LRR receptors, activating signal-transduction cascades.
Results from DNA gel blot analyses, using a 450-bp oligonucleotide probe from the LRR region, indicated that SR160 is a single-copy gene (Fig. 5). The differences in nucleotides in the probe region between SR160 and other LRR receptor kinases, including BRI1, likely would have precluded the identification of other LRR receptor genes under the stringency used for the SR160 gene.
DNA blot analysis of SR160 gene. Genomic DNA was isolated from L. peruvianum suspension-cultured cells and 5 μg of DNA were digested with SpeI or NcoI and XbaI together. Digested and undigested DNA were separated on an agarose gel, blotted to nylon membranes, and probed with an ≈450-bp oligonucleotide probe derived from the LRR region of the SR160 cDNA.
The identification of the systemin receptor as a member of the LRR receptor kinase family further supports a hypothesis that members of this family of receptors play major roles in plants in recognition of a variety of polypeptide signals. The LRR receptor kinases have previously been identified in plants, as receptors for a polypeptide originating from pathogens (FLS2; ref. 31), as receptors for polypeptide hormones involved in developmental processes (CLV1; ref. 12), and now, as a receptor for the plant polypeptide wound signal systemin (SR160). The possibility has already been raised that LRR signaling in plants and animals may have evolved from ancestral LRR receptors that interacted with polypeptide ligands (32), with the LRR regions providing variable scaffolds to potentially accommodate the evolution of perception for a vast variety of polypeptide signals (26). Research on polypeptide signaling is in its infancy. A further understanding of the structural and functional relationships of SR160 with its ligand systemin, and the intracellular signaling that results from this interaction, should provide unique insights into the mechanisms of polypeptide perception and signal transduction in plants.
We thank Dr. David Kramer and Dr. Frank Loewus for generous use of the Parr Cell Disruption Bomb, Dr. Jerry Reeves for iodination facilities, and the DNA and Protein Technology Center at Rockefeller University for internal amino acid sequencing. This research was supported by Project 1791 of the Washington State University College of Agriculture and Home Economics and National Science Foundation Grant IBN 0090766.
