|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
J. Biol. Chem., Vol. 279, Issue 37, 39035-39041, September 10, 2004
NMR Solution Structure of Ole e 6, a Major Allergen from Olive Tree Pollen*![]() ![]() ![]() ![]() ![]() ![]() **
From the
Received for publication, June 1, 2004 , and in revised form, June 30, 2004.
Ole e 6 is a pollen protein from the olive tree (Olea europaea) that exhibits allergenic activity with a high prevalence among olive-allergic individuals. The three-dimensional structure of Ole e 6 has been determined in solution by NMR methods. This is the first experimentally determined structure of an olive tree pollen allergen. The structure of this 50-residue protein is based on 486 upper limit distance constraints derived from nuclear Overhauser effects and 24 torsion angle restraints. The global fold of Ole e 6 consists of two nearly antiparallel -helices, spanning residues 3-19 and 23-33, that are connected by a short loop and followed by a long, unstructured C-terminal tail. Viewed edge-on, the structured N terminus has a dumbbell-like shape with the two helices on the outside and with the hydrophobic core, mainly composed of 3 aromatic and 6 cysteine residues, on the inside. All the aromatic rings lie on top of and pack against the three disulfide bonds. The lack of thermal unfolding, even at 85 °C, indicates a high conformational stability. Based on the analysis of the molecular surface, we propose five plausible epitopes for IgE recognition. The results presented here provide the structural foundation for future experiments to verify the antigenicity of the proposed epitopes, as well as to design novel hypoallergenic forms of the protein suitable for diagnosis and treatment of type-I allergies. In addition, three-dimensional structure features of Ole e 6 are discussed to provide a basis for future functional studies.
Type-I allergies are one of the most widespread pathologies in developed countries where they affect more than 20% of the population, provoking respiratory, skin, and intestinal disorders. The key factor responsible for the disease is the release of inflammatory mediators, such as histamine (1-3), after an IgE-mediated immunological response (4). The diagnosis and treatment of allergies are based on the use of extracts and vaccines, which require large amounts of biological materials obtained from natural sources for their production. These extracts can provoke lateral effects such as anaphylactic shock, are expensive to obtain, and do not assure a constant composition of allergens in extracts. Therefore, the production of the allergens via peptidic synthesis or molecular biology techniques could be a more effective and cheaper alternative source of allergens that could permit the standardization of treatments (5). The identification of the epitopes, regions of the antigen recognized by the immunoglobulin, is essential in order to design these synthetic vaccines. Until now, epitope-mapping techniques have been based on the search for linear or sequential epitopes (6-9), but they have failed in the identification of discontinuous or conformational epitopes. The shape and accessibility of each region of the allergen are essential for recognition by the IgEs (10), so to determine the structure of the allergens is a necessary step in order to define accurately the putative epitopes. This way we can limit the putative residues preferentially recognized by the IgEs, minimize the number of epitopes to be tested for mutating them, and verify their implication in the antigenic response to prepare hypoallergenic forms of the protein to be used as vaccines. Pollinoses, type-I allergies caused by anemophilous pollens, are one of the most widespread allergies (11), and olive tree (Olea europaea) pollen is one of the main causes of seasonal respiratory allergy in Mediterranean countries (12). Important cross-reactions with other Oleaceae pollens, like ash (Fraxinus excelsior), lilac (Syringa vulgaris), or privet (Ligustrum vulgare), have been described previously (13-19). At the moment, 10 allergens from olive pollen, Ole e 1-10, have been described (20-24), but their three-dimensional structures have not been determined. Ole e 6 is a 50-amino acid protein composed of a unique polypeptide chain with unknown function. Its allergenic prevalence varies for different geographical areas and reaches 55% in areas where the olive tree is extensively cultivated (21, 25). The amino acid sequence contains two sets of the sequential motif Cys-X3-Cys-X3-Cys (25) and shows similarity with a polypeptide sequence deduced from tobacco (Nicotiana tabacum) cDNA. This Cys-enriched motif has also been found in a putative protein deduced from the Caenorhabditis elegans genomic structure. Ole e 6 has not been shown to share IgE epitopes with other allergenic pollens, which makes Ole e 6 a good candidate to be included in synthetic vaccines for olive pollinosis because there are not expected significant homologues or similar sequences in other plants. Recently, we have produced in Pichia pastoris and characterized the recombinant form of Ole e 6 (22). In this work we have used its 15N-labeled form to determine its high resolution three-dimensional structure in solution. The analysis of this structure is essential to infer information about the accessibility, shape, and mobility of the different regions of the molecule, which can lead to the design of new vaccines specific for olive tree pollen. In addition, the structure determination can also provide a basis to understand or hypothesize on the biological function of this novel plant protein.
Protein SampleThe cDNA from Ole e 6 was obtained as described previously (25), and it was cloned into the vector pPIC9 to allow its expression in P. pastoris. The plasmid containing the cDNA was integrated in the P. pastoris GS115 his4 strain genome, and the production of the recombinant Ole e 6 (rOle e 6)1 was carried out as described previously (26). Briefly, a colony of plasmid-containing P. pastoris was grown in 10 ml of BMG (100 mM K2HPO4, pH 6, 0.34% yeast nitrogen base, 1% (NH4)2SO4, 4 x 10-5% biotin, and 1% glycerol) for 2 days at 30 °C. This culture was used as the preinoculum of a 1-liter culture in the same medium and conditions. After 2 days, the cells were grown in 200 ml of BMM, which was the same as BMG but with 0.5% methanol instead of glycerol, to induce the production of the recombinant protein. After a 4-day culture, the supernatant was collected and dialyzed in the presence of 20 mM NH4HCO3. The dialyzed protein was separated by size through a Sephadex G-50 chromatography column and later by reverse phase HPLC through a nucleosyl-C18 column.
Collected peaks from HPLC were resolved in 17% SDS-PAGE, and after Western blotting the peak containing the pure protein was determined. The protein from this peak was selected and used in the subsequent experiments. 15N-rOle e 6 was produced and isolated in the same way but substituting (NH4)2SO4 with (15NH4)2SO4 in the BMG and BMM media. All the samples were analyzed by amino acid analysis, N-terminal sequencing by Edman degradation, and mass spectroscopy.
Circular Dichroism AnalysesCD spectra of rOle e 6 were obtained on a Jasco J-715 spectropolarimeter (Japan Spectroscopic Co., Tokyo, Japan), fitted with a 150-W xenon lamp, and connected to a Neslab RTE-111 thermostabilizer bath. Far-UV spectra were registered in the range of 190-250 nm using an optical cell path of 0.1 cm. The spectra were recorded at 50 nm/min scanning speed, and the accumulation of four spectra was achieved. Samples were analyzed in 20 mM sodium phosphate, pH 7.2, at 20 and 85 °C, and the protein concentration was 0.2 mg/ml. Mean residue mass ellipticities were calculated based on 116 as the average molecular mass/residue, obtained from the amino acid composition of Ole e 6, and expressed in terms of Temperature-induced denaturation of rOle e 6 was monitored by CD measurements at 220 nm. The temperature was increased from 20 to 85 °C and then cooled to 20 °C at 0.5 °C/min.
NMR SpectroscopyrOle e 6 samples were prepared for NMR experiments at 0.5-1 mM in 95% H2O, 5% D2O or in D2O containing sodium-4,4-dimethyl-4-silapentane-1-sulfonate at pH 6.0. NMR spectra were carried out at 278, 298, or 308 K on a Bruker AV600 NMR spectrometer equipped with a triple-resonance cryoprobe and an active shielded z-gradient coil. Chemical shifts were referenced to the internal sodium-4,4-dimethyl-4-silapentane-1-sulfonate as described previously (27). On the unlabeled sample two-dimensional COSY (28), total correlation spectroscopy (65-ms mixing time) (29), and NOE spectroscopy (150-ms mixing time) (30) spectra were acquired in H2O and D2O. Two-dimensional 15N HSQC (31) and three-dimensional nuclear Overhauser effect spectroscopy-HSQC (80-ms mixing time) (32) spectra were acquired in H2Oonthe 15N sample. Additionally, 15N HSQC spectrum was recorded without decoupling during acquisition, using an
Structure Calculation1428 unambiguous NOEs were converted into 900 upper limits by the CALIBA program (36). The disulfide bond pattern of Ole e 6 was unambiguously determined on the basis of NOE evidences. For the three S-S pairs (8-34, 12-30, and 16-26), at least one intercysteine NOE could be found, and the corresponding restraints were introduced in the calculation. After DYANA filtering, 486 conformationally relevant distances were obtained. These constraints were supplemented with 24 The three-dimensional structure of rOle e 6 was determined with the program DYANA 1.5 (37). A total of 50 structures were calculated, and the 25 structures with the lowest target function were refined by restrained energy minimization in vacuum (500 steps) using the program AMBER7 (38). The program MOLMOL (39) was used to visualize and globally characterize the structures. The quality of the refined structures was evaluated with the program PROCHECK-NMR (40).
Thermal StabilityTo test the thermal stability of rOle e 6, CD spectra were obtained at 20 °C, after heating at 85 °C, and again at 20 °C after cooling (Fig. 1A). Temperature-induced changes in the secondary structure of rOle e 6 were registered by measure of CD values at 220 nm between 20 and 85 °C (Fig. 1B). The experiment showed a linear increase in the mean residue mass ellipticity without a transition between two states, making impossible the determination of a Tm value. The CD spectra obtained at 85 °C showed that the protein maintained elements of secondary structure. The recovery of the initial temperature induced almost the complete recuperation of the original spectrum, indicating the restoration of the structure of the allergen.
Description of the Ole e 6 StructureThe assignments of the 1H and 15N resonances were obtained following the standard strategy (41) and have been deposited in the BioMagResBank data base (42) under the accession number BMRB-6139 [BMRB] . The 1H and 15N assignments for the backbone and side chains are complete, except for Asp-1. As an example, the assigned 15NHSQC spectrum is shown in Fig. 2.
The coordinates of the 25 minimized conformers obtained with DYANA, on the basis of the NMR restraints, have been deposited in the Protein Data Bank under the accession number 1SS3 [PDB] . The resulting structures satisfied the experimental constraints with small deviations from the idealized covalent geometry, and most of the backbone torsion angles for nonglycine residues lay within the allowed regions in the Ramachandran plot. The statistics and details of quality and precision for the accepted structures are summarized in Table I, and a superposition of the selected structures is shown in Fig. 3. The root-mean-square deviation value for backbone heavy atoms of the most structured regions (excluding residues 1 and 2 from the N terminus and residues 34-50 from the unstructured C terminus) is 1.14 Å. The maximum violation is 0.29 Å, and the average sum of violations for the whole set of distance constraints is 4.0 Å.
The global fold of Ole e 6 consists of two nearly antiparallel -helices, spanning residues 3-19 and 23-33, that are bound by a short loop and maintained together with the help of the three disulfide bonds. The average angle formed between these two helices is 140° in the 25 final minimized structures. Both -helices have a very regular internal hydrogen bonding pattern. As expected, stretches of CO(i)-HN(i+4) hydrogen bonds are found in regions 3-20 and 22-33 in the resulting structures. The C-terminal end (residues 34-50) behaves as an unstructured tail. In addition to the mentioned hydrogen bonds participating in the secondary structure, we found an important one appearing in all conformers that involves the side chain of the histidine residue, His-13 H , and the carbonyl backbone atom of Phe-23. This hydrogen bond could be important for the stability of the protein as it keeps together residues in the two -helices.
Many side chain conformations of residues in rOle e 6 are well defined, with 22 residues (5, 8, 9, 10, 11, 12, 13, 15, 16, 17, 18, 19, 21, 24, 25, 26, 27, 29, 30, 31, 32, and 33) having The hydrophobic core of this small protein has an unusual composition. Residues with accessible surface below 20% are Phe-5, Cys-8, Tyr-9, Cys-12, His-13, Cys-16, Ser-17, Cys-26, Cys-30, and Cys-34. Thus, with the exception of Ser-17, the core of the protein is composed only of aromatic residues and cysteines (Fig. 4). Their relative disposition is also very specific; all of the aromatic rings are in one face of the interhelix surface, whereas the three S-S bonds are just opposite in the other face. Additionally, each disulfide bond is located on the top of the inner face of one ring: disulfide 16-26 on the top of His-13, disulfide 12-30 on Tyr-9, and disulfide 8-34 on Phe-5 (Fig. 4).
Ole e 6 has a great number of exposed residues, most of them bearing charged groups (11 negative and 9 positive charges). The surface of the protein shows a well defined charge distribution. The unstructured tail is mainly positive, whereas the structured part has a positive and a negative face as can be seen in Fig. 3B. The main cluster of positive charges is composed of residues Lys-19 and Lys-29, and the negative clusters are composed of Glu-15, Asp-18, and Asp-33 in one side and Glu-27 and Asp-31 in the other side.
Ole e 6 is the first olive pollen allergen whose three-dimensional structure has been determined. No homologue protein was found for Ole e 6 in the data base banks, and only polypeptide sequences derived of two genomic structures were detected to have a degree of similarity. The alignment of the amino acid sequence of Ole e 6 with those deduced from genes of N. tabacum and C. elegans rendered an identity degree of 52 and 28%, respectively, and the Cys-X3-Cys-X3-Cys motifs were conserved. Unfortunately, no data about these putative proteins or about their function or structure are available so far.
The solution structure of the Ole e 6 allergen determined here consists of a well defined globular N-terminal body composed of two antiparallel
On the other hand, the presence of a large amount of covalent, hydrophobic, and hydrogen bond interactions between the two helices suggests a high stability for this part of the protein. In fact, the absence of a two-state denaturation curve in the thermal transition from 20 to 85 °C and the extent of secondary elements that remained at high temperature indicate a great stability for the allergen, to which the disulphide bridges should also contribute. The presence of two grooves in the protein surface coincident with the two interhelix region faces is notable. Interestingly, the base of one of these grooves is composed of the outer faces of the aromatic rings of Phe-5, Tyr-9, His-13, and Phe-23. Their nearly planar disposition with respect to the surface suggests that they could interact with other groups as, for example, aromatic-aromatic or aromatic-
Moreover, the unstructured tail can also play a biological and/or immunological role. The conformational chemical shifts of the residues in the C-terminal tail suggest that there is a tendency to form an At the present, vaccination is the most effective treatment to prevent type-I allergies. Ole e 6 seems to be a good candidate for the design of vaccines against olive pollen allergy as it is a relevant allergen with high prevalence among patients allergic to olive pollen. Thus, an exquisite definition of the epitopes is mandatory to develop these vaccines. In this sense, the correct definition of the epitopes in the molecule should be facilitated by the determination of the three-dimensional structure as the solvent accessibility of each region precisely indicates which areas are more susceptible to interact with the immunoglobulin (49, 50). Crystallographic studies of Fab fragments of immunoglobulins complexed with their antigens have suggested that a contact surface covering approximately 600 Å2 is necessary for the binding (51). Thus, from the results of this work, it can be deduced that the most exposed regions of Ole e 6 are (i) the highly flexible tail at the C-terminal end of the molecule (from residues 34 to 50, more than 2000 Å2), (ii) the loop connecting both helices and its vicinity (from Asp-18 to Phe-25, 715 Å2), and (iii) the first residues of the protein (from Asp-1 to Lys-6, 742 Å2). Other possible epitopes are those formed by residues located in the external faces of the helices with accessible surface areas of 789 Å2 in helix I and 788 Å2 in helix II. All of these regions could be important for choosing key residues in the recognition and, as a consequence, how to design hypoallergenic forms of the protein.
Structural comparison of proteins that induce an allergenic response can be used to predict allergenic cross-responses (52), eventually to determine possible characteristics of IgE recognition (50), and to predict the regions recognized by the immunoglobulins (10, 53, 54). Comparison of the whole sequence of Ole e 6 with the Swiss-Prot data base had shown only homology with a cysteine-rich putative protein from N. tabacum (55) of unknown three-dimensional structure. However, using the previously defined regions of Ole e 6 with high accessibility, a detailed search of homologues in the specific Structural Database of Allergenic Protein (SDAP) (56) has shown similarities with fragments of some allergens (Table II). Among them, the only one with a known three-dimensional structure is Bet v 1 (1BTV
[PDB]
) (57). The peptide homologue of the C-terminal tail of Ole e 6 in Bet v 1 is part of a
The biological function of Ole e 6 is still unknown. To try to shed some light on this point, we have compared Ole e 6 with other published protein structures using the DALI server (58). The comparison reveals homology with protein fragments holding two -helices with similar relative orientation to those in Ole e 6. It is interesting to note that similarities have been found with protein fragments that are part of channels, such as ATP synthase (59) and the mechanosensitive channel MscS (60) from E. coli, and with a protein that interacts with lipids, saposin B (61) from humans. As mentioned above, the specific location of aromatic residues in the -helix surfaces of Ole e 6, and the existence of two well defined basic and acid clusters, and their conservation on the putative counterpart from tobacco pollen, allow us to suggest a possible involvement of these structural motifs in the biochemical activity of the protein. In conclusion, we have obtained the first well defined structure of an allergen from olive pollen, one of the most important causes of pollinosis in Mediterranean countries. The information derived from this structure can be used to simplify the mapping of the epitopes of the protein in order to design new diagnostic and therapeutic tools and to investigate the cellular activity of this pollen protein.
The atomic coordinates and structure factors (code 1SS3 [PDB] ) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).
* This work was supported by Grant CAM2002-07B/0054 from the Comunidad Autónoma de Madrid and Grant SAF2002-02711 from the Ministerio de Ciencia y Tecnología (Spain). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|| Recipient of a fellowship from the Ministerio de Educación, Cultura y Deporte (Spain). ** To whom correspondence should be addressed. Fax: 34-91-561-94-00; E-mail: mbruix{at}iqfr.csic.es.
1 The abbreviations used are: rOle e 6, recombinant Ole e 6; HPLC, high pressure liquid chromatography; W, watt; HSQC, heteronuclear single quantum correlation spectroscopy; NOE, nuclear Overhauser effect.
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||