Article pubs.acs.org/JAFC
Elusive Structural, Functional, and Immunological Features of Act d 5, the Green Kiwifruit Kiwellin Lesa R. Offermann,† Ivana Giangrieco,§ Makenzie L. Perdue,† Sara Zuzzi,#,Δ Mario Santoro,#,Δ Maurizio Tamburrini,§ Daniel J. Cosgrove,⊥ Adriano Mari,#,Δ Maria Antonietta Ciardiello,*,§ and Maksymilian Chruszcz*,† †
Department of Chemistry and Biochemistry, University of South Carolina, Columbia, South Carolina 29208, United States Institute of Biosciences and Bioresources, CNR, Via Pietro Castellino 111, I-80131 Napoli, Italy # Center for Molecular Allergology, IDI-IRCCS, Rome, Italy Δ Associated Centers for Molecular Allergology, Rome and Latium, Italy ⊥ Department of Biology, Pennsylvania State University, University Park, Pennsylvania 16802, United States §
S Supporting Information *
ABSTRACT: Kiwellin (Act d 5) is an allergenic protein contained in kiwifruit pulp in high amounts. The aim of this study was to investigate the three-dimensional structure of the natural molecule from green kiwifruit and its possible function. Kiwellin was crystallized, and its structure, including post-translational modifications, was elucidated. The molecular weight and structural features, in solution, were analyzed by gel filtration and circular dichroism, respectively. Although structurally similar to expansin, kiwellin lacks expansin activity and carbohydrate binding. A specific algorithm was applied to investigate any possible IgE reactivity correlation between kiwellin and a panel of 102 allergens, including expansins and other carbohydrate-binding allergens. The available data suggest a strong dependence of the kiwellin structure on the environmental/experimental conditions. This dependence therefore poses challenges in detecting the correlations between structural, functional, and immunological features of this protein. KEYWORDS: kiwellin, kiwifruit, allergen, protein crystallization
■
INTRODUCTION Kiwellin is the most abundant protein in gold kiwifruit (Actinidia chinensis) and one of the three most abundant proteins in green kiwifruit (Actinidia deliciosa).1 Although gene sequences and mRNA transcripts coding for kiwellin homologous proteins have been found in several plant organisms, including mulberry, grape, rice, potato, citrus fruits, and others, the expressed protein, until now, has been purified and characterized in kiwifruit only. Kiwellin is a 189-residue protein with a molecular mass of approximately 20 kDa.2 In unripe kiwifruit, kiwellin is mostly extracted from the cell wall by 0.5 M NaCl, whereas in the ripe fruit the amount of soluble protein exceeds the binding capacity of the cell wall. In the ripe green species, kiwellin may undergo in vivo proteolytic processing catalyzed by the cysteine protease actinidin, thus producing kissper and KiTH, which correspond to the 4 kDa N-terminal domain and 16 kDa C-terminal domain, respectively.2 In contrast, the two separated domains have never been detected in gold kiwifruit, although kiwellin purified from this species was cleaved in vitro by actinidin from green kiwifruit. Therefore, the apparent absence of processing in gold kiwifruit can be correlated to the low level, or absence, of the protease actinidin, rather than to different structural features between the molecules contained in the two species. The two kiwellin domains, kissper and KiTH, were purified from green kiwifruit and investigated. In experiments on model synthetic planar lipid membranes, kissper revealed pH-depend© XXXX American Chemical Society
ent and voltage-gated pore-forming activity together with anion selectivity and channeling.3 In addition, kissper showed antiinflammatory and antioxidant effects in experiments performed on cultured intestinal cells and ex vivo colonic tissues from subjects suffering from Crohn’s disease.4 The molecular mechanism underlying the biological effects of kissper in human model systems, which is still unknown, and a possible correlation of its function with the ion transport activity are both intriguing topics that are worth investigating soon. Kiwellin is an allergenic protein that is officially registered by the WHO-IUIS (www.allergen.org) with the allergen name Act d 5 from green kiwifruit and Act c 5 from the golden species.5 In a test group of 29 subjects known to be allergic to kiwifruit, 38% of patients, who were analyzed by the skin prick test, were positive to kiwellin. The subjects that tested positive to kiwellin reported one or more of the following symptoms: edema, oral symptoms, vomiting, asthma, atopic dermatitis, skin itching, and urticaria.5 The prevalence of subjects allergic to kiwifruit and producing IgE antibodies specifically recognizing kiwellin, was >80% by IgE immunoblot analysis.6 Therefore, it appears that the prevalence of subjects producing kiwellin-specific IgE antibodies is much higher than the percentage of those reacting Received: May 2, 2015 Revised: June 30, 2015 Accepted: July 6, 2015
A
DOI: 10.1021/acs.jafc.5b02159 J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Article
Journal of Agricultural and Food Chemistry
mg/mL and 50 mM, respectively. The Act d 5−glucose solution was mixed with mother liquor [0.1 M HEPES, pH 7.5, 25% (w/v) polyethylene glycol (PEG) 3350] in a 1:1 ratio and was hung over a reservoir of 26% PEG 400. The crystals were cryo-protected with PEG 400 and cooled in liquid N2. Data were collected from a single crystal at 100 K at SER-CAT 22ID beamline at the Advanced Photon Source (APS), Argonne National Laboratory (ANL). The crystal belonged to the P21 spacegroup with two subunits per asymmetric part of the unit cell. The data were processed with the HKL-2000 software package.14 The data collection statistics are reported in Table 1.
in in vivo tests. In addition to IgE antibodies, a significant percentage of subjects, both kiwifruit allergic and tolerant, were also reported to produce kiwellin-specific IgG4.7 The subjects positive to kiwellin in IgE-immunoblotting were also positive to KiTH, purified from the natural source, thus suggesting that at least the linear sequence epitopes recognized by IgE antibodies could be located on the kiwellin C-terminal domain.2 The immunological properties, especially those correlated with the conformational epitopes, were reported to be modulated by the physicochemical features of the environment/experimental conditions, such as solvent pH and polarity.5 The effects of kiwellin on human health have so far received more attention than the physiological role of this protein in the natural source. The few papers describing putative homologues from other species provide only vague information on the possible implications of kiwellin in complex plant processes without any attribution of a specific function. For instance, in grape berries, a kiwellin homologue was classified as a “ripening-related protein” because of a marked increase in the mRNA levels associated with the fruit ripening.8 A DIGE-based quantitative proteomic analysis of the grape berry suggested kiwellin’s involvement in fruit development and stress response.9 In addition, a pathogenesis-related function was hypothesized following experiments in potato leaves showing a strong up-regulation of kiwellin mRNA transcription after infection with Phytophthora infestans, an oomycete that causes late blight in Solanum spp.10 It is worth noting that a clear orthologue of kiwellin is not present in Arabidopsis thaliana,11 the model plant for which the entire genome is described. This observation suggests that kiwellin might have a specialized function associated with physiological processes particular to some plant species/families, rather than a basic function found ubiquitously throughout the plant kingdom. In summary, the biochemical function of kiwellin, as well as its physiological role in kiwifruit, is still unknown. However, the available literature indicates that kiwellin, and its disjoined domains kissper and KiTH, may have some effects on human health, including intriguing pharmacological activities. Very recently, the partial (about 83%) crystal structure of kiwellin from the golden kiwifruit species, registered as allergen Act c 5, was reported.12 Here we describe an investigation of the structure of kiwellin from green kiwifruit, allergen Act d 5, determined using X-ray crystallography and refined to 2.1 Å. Additionally, attempts to correlate the structural features with possible biological and immunological activities were performed.
■
Table 1. Crystallographic Data and Refinement Statistics for Act d 5a PDB code
4X9U
data collection wavelength (Å) unit cell (a, b, c; β) (Å; °) space group solvent content (%) protein chains in AU resolution range (Å) highest resolution shell (Å) unique reflections redundancy completeness (%) Rmerge (%) average I/σ(I) CC1/2
1.00 52.4, 60.42, 55.60; 95.9 P21 43.0 2 50.0−2.10 2.14−2.10 20290 5.9 (4.3) 99.8 (96.6) 13.0 (75.0) 15.3 (2.1) 0.94(0.83)
refinement R (%) Rfree (%) mean B value (Å2) B from Wilson plot (Å2) RMS deviation bond lengths (Å) RMS deviation bond angles (deg) no. of amino acid residues
no. of water molecules
16.8 (18.4) 22.3 (26.1) 39.0 25.2 0.02 1.80 326 residues total subunit A: 183 residues subunit B: 143 residues 245
Ramachandran plot most favored regions (%) additional allowed regions (%)
95.0 5.0
a
Values in parentheses refer to the highest resolution shell. AU, asymmetric unit.
MATERIALS AND METHODS
Protein Purification. Kiwellin was purified from the pulp of green kiwifruit (A. deliciosa cv. Hayward) following the already described procedure.13 As a final purification step, the protein was subjected to size exclusion chromatography on a Superdex 75 HR10/30 column (Amersham Biosciences, Uppsala, Sweden), equilibrated and eluted with 10 mM Tris-HCl, pH 7.5, containing 0.25 M NaCl. The protein fractions were pooled and used for crystallization experiments. The protein concentration was estimated on the basis of the molar extinction coefficient at 280 nm (23335 M−1 cm−1). The purity of the protein preparation was checked by SDS-PAGE, RP-HPLC, and Nterminal amino acid sequencing, as previously reported.13 Crystallization, Data Collection, and Processing. Crystallization experiments were performed at room temperature using the hanging-drop vapor diffusion method and NeXtal plates (Qiagen, Chatsworth, CA, USA). A solution of purified Act d 5 was mixed with glucose such that the final protein and glucose concentrations were 3.5
Structure Determination and Refinement. The Act d 5 structure was solved by molecular replacement using a truncated version (residues 176−263) of Achromobacter lyticus protease I (PDB ID 1ARB) as a starting model. The starting model was identified after a BLAST15 search with a sequence of Act d 5 against sequences of proteins with their structures determined experimentally, as reported in the Protein Data Bank.16 Molecular replacement was performed with HKL-300017 integrated with MOLREP18 and selected programs from the CCP4 package.19 Models were rebuilt using COOT,20 Buccaneer,21 and ARP/wARP22 and refined with REFMAC.23 MOLPROBITY24 and ADIT25 were used for structure validation. The refinement statistics are summarized in Table 1. Analysis of Protein Molecular Weight by Size Exclusion Chromatography. The entire molecule, kiwellin, and the C-terminal domain, KiTH, were loaded on a gel filtration column Superdex 75 HR10/30 (Amersham Biosciences), eluted with 10 mM Tris-HCl, pH B
DOI: 10.1021/acs.jafc.5b02159 J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Article
Journal of Agricultural and Food Chemistry 7.5, 0.25 M NaCl. Fractions of 0.5 mL were collected, and the absorbance at 280 and 230 nm was recorded. Circular Dichroism (CD) Experiments. CD spectra were recorded on a JASCO J-810 spectropolarimeter (Easton, MD, USA) as already reported.5 A quartz cell of 0.1 cm path length was used to record the spectra over the wavelength range of 260−190 nm with a bandwidth of 1.0 nm. The purified protein was first extensively dialyzed against 0.2 M NaCl to eliminate Tris-HCl contained in the solution and then against 2 mM NaCl. After that, the protein solution was concentrated to 1 mg/mL by ultrafiltration. For CD experiments, the protein solution was diluted 10-fold in the appropriate buffer and left for 2 h at 25 °C before acquisition of the spectra. Measurements were performed in 5 mM sodium phosphate, pH 7.4, and 5 mM sodium acetate, pH 4.5, at SDS concentrations from 0 to 100 mM, at 25 °C. The protein concentration was 0.10 mg/mL. Each spectrum was baseline corrected for the contribution of the solvent. Endoglucanase Activity Assay. Endo-1,4-β-D-glucanase activity was assayed using dyed soluble azo-carboxymethylcellulose (AZOCMC) as the substrate, according to the manufacturer’s protocol (Megazyme, Intl., Bray, Ireland). Five, 20 and 50 μg of kiwellin were incubated overnight, at 37 °C, in the presence of 2% AZO-CMC, at pH 4.0, 5.0, 6.0, and 7.4, in 100 mM sodium acetate or sodium phosphate buffer, in a final volume of 1 mL. The reaction was stopped by adding 96% ethanol. Then, the mixture was centrifuged, and the supernatant absorbance was measured at 590 nm. 3,5-Dinitrosalicylic Acid (DNS) Assay for Reducing Sugars. Kiwellin’s capacity to hydrolyze different polysaccharides made of glucose or mannose was analyzed by using the DNS colorimetric method, allowing for a quantitative estimation of released reducing sugars.26 The substrates used were cellobiose, cellotetraose, mannan, laminarin, curdlan, and lichenan. Each one was analyzed in the following buffers: sodium phosphate, pH 4.0 and 7.4; sodium acetate, pH 5.0; MES, pH 6.0. Amounts of kiwellin corresponding to 12.5, 25, and 50 μg were added to the incubation mixture containing 100 μg of a substrate in 0.1 M buffer, in a final volume of 125 μL. Incubations were carried out at 37 °C for 1 and 3 h and overnight. Then, after 10 min at room temperature, 0.1 mL of DNS reagent was added to the samples, and the mixtures were incubated for 5 min in a boiling water bath. After cooling in an ice bath, the absorbance at 546 nm was measured. Standard curves with glucose as reducing sugar were used as a reference. All assays were carried out in duplicate. Analysis of Released Glucose by HPLC. To assess the possible hydrolytic activity of kiwellin on cellobiose, the release of glucose was analyzed by chromatographic separation. Twenty-five micrograms of kiwellin was incubated at 37 °C in 50 mM sodium acetate, pH 5.0, in the presence of 100 μg of cellobiose, in a final volume of 0.3 mL. Fiftymicroliter aliquots were removed at incubation times of 0, 3, 24, and 72 h and analyzed by an HPLC (Dionex, Sunnyvale, CA, USA), equipped with an anionic exchange column (Carbopac PA-100). Monosaccharides were separated by isocratic elution with 16 mM NaOH at a flow rate of 0.25 mL/min. Glucose was used as an internal standard. Glycan Array Analysis. Biotinylated Act d 5 was diluted to 0.5 mg/mL in 100 mM Na2HPO4, 20 mM NaCl, pH 7.6, and sent to the Consortium for Functional Genomics (CFG) at Emory University to determine if protein−glycan interactions occur. The CFG facility ran the mammalian glycan array assay (version 5.1), testing 610 glycans in replicates of 6 with bacterial lectin as the positive control. The protein was directly labeled with a fluorescent tag for detection. This CFG mammalian glycan array contains glycans of potential interest for the biological effects of kiwellin in human being. It does not include plant cell wall glycans such as cellulose, xyloglucan, and pectins, but recent results indicate kiwellin has negligible affinity for these plant components.12 Cell Wall Creep Analysis. Cucumber seedlings were grown in the dark, and hypocotyls were harvested, frozen, abraded, and prepared for wall extension assays as described previously.27 Seeds of wheat (Triticum aestivum L. cv. Pennmore) were germinated on Kimpak Seed Germination Paper K-22 (Seedburo Equipment Co., Chicago, IL, USA) soaked with distilled water, in flats, 50 × 25 × 6 cm, with lids of
the same dimensions. Seedlings were grown in the dark for 4 days at 28 °C. Coleoptiles from etiolated grass seedlings were cut, gently abraded by rubbing them between two fingers coated with a slurry of well washed carborundum (320 grit, Fisher Scientific Inc., Fair Lawn, NJ, USA), separated from primary leaves, and then stored at −80 °C before use. Etiolated wheat coleoptiles and cucumber hypocotyls were inactivated by a 15 s dip in boiling water and clamped in a constant-force extensometer28 with 5 mm between clamps, a 20-g force, in 200 μL of 20 mM sodium acetate buffer, pH 5.5. To increase the wall sensitivity to wall-loosening agents, the buffer was exchanged at 15 min for 50 mM NaOH for ∼35 min (for coleoptile walls) or 50 mM CDTA (for cucumber walls) and then rinsed in 20 mM sodium acetate buffer. Finally, the buffer was replaced with fresh buffer containing 500 μg/mL kiwellin. All experiments were performed in replicates of four. Protein Sequencing. Protein direct sequencing was carried out by automated Edman degradation on a Procise 492 automatic sequencer, connected with an online phenyl thiohydantoin (PTH) amino acid analyzer (Applied Biosystems, Foster City, CA, USA). Amino acids were separated by RP-HPLC and identified, as their PTH derivatives, on the basis of coelution with appropriate standards (20 PTH-amino acid standard solution, PerkinElmer Life Sciences). Immunological Data Collection, Storing, Processing, and Statistics. Act d 5 IgE reactivity, as well as IgE reactivity toward other food and inhaled allergens, was collected by routinely testing 30836 allergics using a molecule-based microarray (ImunoCAP ISAC 103, Phadia Multiplexing Diagnostics, PMD, Vienna, Austria) bearing 103 allergenic proteins. All diagnostic data were stored during testing by online connectivity to the InterAll e-record (Allergy Data Laboratories, Italy). The study was approved by the Institutional Review board (35/ IDI-CE/2006). To understand the immunological behavior of Act d 5, a new statistical procedure was applied to this IgE data set. Act d 5 IgE reactivity was compared to the other 102 allergens using an aggregative hierarchical clustering algorithm with Ward’s linkage method. The distance between two allergens was built as a dissimilarity for the categorical value (0 = patient IgE negative to the allergen, 1 = patient IgE positive to the allergen). The chosen dissimilarity was d=
n01 + n10 n11 + n01 + n10
where n11 is the number of patients positive to both allergens, n10 the number of patients positive to the first allergen, and n01 the number of patients positive to the second allergen, respectively. To remove the effect of the frequency of IgE positive patients, the allergens were divided into four frequency classes, broadly dispersed in a range from 33 for Bla g 5.0101 to 8774 patients for Phl p 1.0101 (Table S1). The results of the analysis are shown in cluster dendrogram graphics. The approximately unbiased (AU) p values for the cluster dendrograms were calculated via multiscale bootstrap resampling (pvclust package in R). To understand whether the findings obtained with the newly applied data analysis were reliable, the same analytical method was applied to a second green kiwifruit molecule, namely Act d 1, actinidin, which is known to have cysteine protease activity.29 The IgE binding frequency of this allergen was evaluated in comparison with that of other three cysteine proteases, Der p 1, Der f 1, and Ana c 2, immobilized on the same ISAC system, usually grouped together with Act d 1 on the basis of their similar biochemical features.30 The statistical analysis and graphical visualization of the data were performed using the R software (www.r-project.org). The statistically significant level was set at a p value
