Báo cáo khoa học: Gain of structure and IgE epitopes by eukaryotic expression of the major Timothy grass pollen allergen, Phl p 1

Gain of structure and IgE epitopes by eukaryotic expression of the major Timothy grass pollen allergen, Phl p 1 Tanja Ball1,2, William Edstrom2, Ludwig Mauch3, Jacky Schmitt3, Bernd Leistler3, Helmut Fiebig4, Wolfgang R. Sperr5, Alexander W. Hauswirth5, Peter Valent5, Dietrich Kraft1, Steven C. Almo2 and Rudolf Valenta1

1 Department of Pathophysiology, Center for Physiology and Pathophysiology, Medical University of Vienna, Austria 2 Albert Einstein College of Medicine, Department of Biochemistry, NY, USA 3 Pharmacia Diagnostics, Freiburg, Germany 4 Allergopharma KG, Reinbek, Germany 5 Division of Hematology, Department of Internal Medicine I, Medical University of Vienna, Austria

Keywords allergen; allergy; epitope; eukaryotic expression; Phl p 1

Correspondence R. Valenta, Division of Immunopathology, Department of Pathophysiology, Center for Physiology and Pathophysiology, Medical University of Vienna, Waehringer Guertel 18–20, A-1090 Vienna, Austria Fax: +43 1 40 400 5130 Tel: +43 1 40 400 5108 E-mail: rudolf.valenta@meduniwien.ac.at

(Received 6 August 2004, revised 21 September 2004, accepted 22 September 2004)

doi:10.1111/j.1432-1033.2004.04403.x

Approximately 400 million allergic patients are sensitized against group 1 grass pollen allergens, a family of highly cross-reactive allergens present in all grass species. We report the eukaryotic expression of the group 1 aller- gen from Timothy grass, Phl p 1, in baculovirus-infected insect cells. Domain elucidation by limited proteolysis and mass spectrometry of the purified recombinant glycoprotein indicates that the C-terminal 40% of Phl p 1, a major IgE-reactive segment, represents a stable domain. This domain also exhibits a significant sequence identity of 43% with the family of immunoglobulin domain-like group 2 ⁄ 3 grass pollen allergens. Circular dichroism analysis demonstrates that insect cell-expressed rPhl p 1 is a is well folded species with significant secondary structure. This material behaved and is adequate for the growth of crystals that diffract to 2.9 A˚ resolution. The importance of conformational epitopes for IgE recognition of Phl p 1 is demonstrated by the superior IgE recognition of insect-cell expressed Phl p 1 compared to Escherichia coli-expressed Phl p 1. More- insect cell-expressed Phl p 1 induces potent histamine release and over, leads to strong up-regulation of CD203c in basophils from grass pollen allergic patients. Deglycosylated Phl p 1 frequently exhibits higher IgE binding capacity than the recombinant glycoprotein suggesting that rather the intact protein structure than carbohydrate moieties themselves are important for IgE recognition of Phl p 1. This study emphasizes the important contribution of conformational epitopes for the IgE recognition of respiratory allergens and provides a paradigmatic tool for the structural analysis of the IgE allergen interaction.

family of

a

allergens and more than 95% of them display IgE reactivity to group I allergens [3–6]. Group 1 allergens represent glycoprotein allergens of approximately 30 kDa that occur as cross-reactive antigens in almost all grasses and corn species [6,7].

Type I allergy is an IgE-mediated hypersensitivity dis- ease affecting more than 25% of the population [1,2]. Grass pollen allergens belong to the group of most fre- quently recognized allergenic components [3]. At least 40% of allergic patients are sensitized to grass pollen

Abbreviations PrPhl p 1, prokaryotic recombinant Phl p 1; ErPhl p 1, eukaryotic recombinant Phl p 1; GST, glutathione S-transferase.

FEBS Journal 272 (2005) 217–227 ª 2004 FEBS

217

T. Ball et al.

Eukaryotic expression of Phl p 1

insect cell-expressed and E. coli-expressed Phl p 1 was compared in histamine release experiments and by CD203c expression in basophils from grass pollen allergic patients [26]. The finding that proper folding of insect-cell expressed Phl p 1 is related to increased IgE reactivity and allergenic activity is discussed as a general feature of respiratory allergens and has rele- vance for the development of allergy vaccines which are based on the reduction of allergen fold.

Results

They are exclusively expressed in mature pollen grains where they are localized mainly in the cytoplasm [8]. Using immuoelectronmicroscopy two mechanisms for the release of group 1 allergens have been demonstra- ted. First, contact of intact pollen grains with mucosal surfaces (e.g. nasal epithelium) leads to hydration and rapid diffusion of the allergens [9]. Second, it has been demonstrated that rain water induces the expulsion of respirable micron size allergen-containing particles from grass pollens [10,11]. The small size of these sub- cellular particles allows them to reach the deeper air- ways and may explain the frequent occurrence of heavy asthma attacks after rainfalls [12,13].

Comparison of natural and recombinant group 1 grass pollen allergens

Although the majority of rPhl p 1 was detected in the insoluble pellet fraction of infected insect cells, up to 0.75 mgÆL)1 of soluble rPhl p 1 could be purified from the culture supernatant by Ni2+-affinity chromato- graphy under nondenaturing conditions. Purified insect cell-expressed Phl p 1 migrated at slightly higher than the natural Phl p 1, E. coli- molecular mass expressed Phl p 1 and the Phl p 1-homologous allergen from rye grass (Lol p 1) (Fig. 1A). Insect cell-expressed Phl p 1 as well as natural group 1 allergens (nPhl p 1 and nLol p 1) reacted with a rabbit antiserum raised against purified E. coli-expressed Phl p 1 (Fig. 1B) but not with the corresponding preimmune serum (Fig. 1C). A band of approximately double the molecular mass as the purified allergens, possibly rep- resenting a dimer, was detected in the bacterial and insect cell-expressed Phl p 1 and, to a lower degree, in the nPhl p 1 preparations.

several major

cDNAs coding for group 1 allergens from several grasses have been isolated and showed high sequence similarity [14–20]. The recombinant group 1 allergen from Timothy grass, rPhl p 1, expressed in Escherichia coli contained many of the T cell epitopes of natural group 1 allergens and cross-reacted with the naturally occuring isoallergens from Timothy grass and other grass species [6,21,22]. However, several post-transla- tional modifications (e.g. glycosylation, occurrence of hydroxyprolines) and the formation of disulphide bonds have been described for group I allergens [23,24]. These modifications do not occur when pro- teins are expressed in prokaryotic expression systems and hence E. coli-expressed group 1 allergens exhibit impaired structural and immunological properties. The importance of conformational epitopes for IgE recog- nition of group I allergens is highlighted by IgE com- petition experiments using recombinant fragments of Phl p 1 representing continuous IgE epitopes as even a mixture of IgE-epitope-containing rPhl p 1 fragments does not completely inhibit IgE binding to intact Phl p 1 [25].

Biochemical and biophysical characterization of insect-cell expressed Phl p 1

The molecular mass of insect cell-expressed Phl p 1 was determined by mass spectrometry to be 28 122 Dalton (data not shown). The difference of 956 Da between the calculated (27 166 Da) and the deter- mined molecular mass is attributed to glycosylation. Limited proteolysis in combination with mass spectro- metry was performed to identify structural domains [27,28]. Fundamental to this approach is the notion that protection against proteolysis is conferred in regions of the protein that are within a rigid struc- ture, while proteolytic cleavage of a multiple-domain protein is biased towards solvent accessible regions (i.e. exposed loops, interdomain linker chains). We identified two proteolytically stable structural domains of rPhl p 1 by limited proteolysis, one comprising C78–K118 and a second domain spanning from

To obtain properly folded rPhl p 1, the cDNA cod- ing for the mature allergen was expressed in baculo- virus-infected insect cells. An expression strategy was chosen which led to the secretion of the recombinant allergen into the cell culture supernatants. rPhl p 1 was purified to homogeneity, characterized by mass spectro-metry and the presence of post-translational modification (i.e. glycosylation) was investigated. The secondary structure content of insect cell-expressed rPhl p 1 was examined by circular dichroism analysis and diffraction quality crystals of the recombinant allergen were grown. The IgE binding properties of insect cell-expressed Phl p 1 were compared with those of E. coli-expressed and natural Phl p 1 by competition studies performed under native conditions and the importance of glycosylation for IgE reactivity was examined by enzymatic deglycosylation of insect cell- expressed Phl p 1. Finally, the biological activity of

FEBS Journal 272 (2005) 217–227 ª 2004 FEBS

218

T. Ball et al.

Eukaryotic expression of Phl p 1

A

B

C

Fig. 1. Coomassie staining and immunoreactivity of purified recombinant and natural group 1 allergens. (A) Coomassie stained SDS ⁄ PAGE containing natural Lol p 1 (nLol p 1), natural Phl p 1 (nPhl p 1), eukaryotic recombinant Phl p 1 (ErPhl p 1) and bacterial recombinant Phl p 1 (PrPhl p 1). B and C represent immunoblots probed with rabbit anti-(Phl p 1 Ig) antiserum and the corresponding preimmune serum, respect- ively.

(data not shown). Only the properly folded insect the E. coli-expressed Phl p 1 cell-expressed but not afforded crystals. These crystals belonged to the ortho- rhombic space group P212121 and diffracted X-rays to a resolution of 2.9 A˚

(Fig. 4).

Phl p 1 and Phl p 2 belong to different clusters of proteins as determined by phylogenetic analysis

insect

K147–K241 (data not shown). The latter corresponds to the region homologous to group 2 allergens. To confirm that the difference between the calculated and determined molecular mass is due to glycosyla- tion of the insect cell-expressed Phl p 1, glycan detec- tion was performed (Fig. 2A). Nitrocellulose-blotted insect cell-expressed Phl p 1, but not E. coli-expressed Phl p 1 showed a positive staining for glycan moieties (blue color). Also, a nonglycosylated control protein, creatinase, and the marker proteins gave negative reaction in the glycan staining and appear brown (Fig. 2A). Finally, enzymatic deglycosylation with PNGase F resulted in a reduction of molecular mass of cell-expressed Phl p 1 as visualized by SDS ⁄ PAGE (Fig. 2B).

Insect cell-expressed Phl p 1 represents a folded protein with considerable b-sheet structure that crystallizes as thin plates

Amino acid sequences of seven group 1 pollen aller- gens and four group 2 ⁄ 3 allergens were subjected to phylogenetic analysis using the phylip 3.6a2 package (http://evolution.genetics.washington.edu/phylip.html) (Fig. 5). The allergens formed three main clusters, one comprising Zea m 1 and Cyn d 1, a second consisting of Lol p 3, Dag g 3, Lol p 1, Phl p 1, Ory s 1 and Tri a 3 and a third cluster including Lol p 2, Hol l 1, Phl p 2 and Pha a 1. Although Phl p 1 and Phl p 2 are derived from Phleum pratense and share high sequence identity, the phylogenetic analysis shows that they are less related to each other than group 1 and group 2 ⁄ 3 allergens from different species.

Insect cell-expressed Phl p 1 contains the IgE epitopes of natural Phl p 1

the CD spectrum of

A comparison of the IgE binding capacity of E. coli- and insect cell-expressed Phl p 1 under nondenaturing conditions in a dot-blot assay showed that insect cell- expressed Phl p 1 was more potent than the E. coli- derived allergen (Table 1). IgE competition studies performed under native conditions confirmed this result (Fig. 6A). Preincubation of sera from four grass pollen allergic patients with E. coli-expressed Phl p 1 completely inhibited IgE binding to the very same pro- tein, but not to the insect cell-expressed Phl p 1. An

Insect cell- and E. coli-expressed Phl p 1 were analyzed by circular dichroism (CD) spectroscopy to determine their secondary structural content (Fig. 3). The CD spectrum of insect cell-expressed, eukaryotic Phl p 1 (ErPhl p 1) suggested the presence of substantial anti- parallel b-sheet, while the CD spectrum for the Phl p 1 expressed in bacteria (prokaryotic: PrPhl p 1) indicated a considerable amount of unordered structure. The characteristics of insect cell- expressed Phl p 1 indicates structural similarity with Phl p 2, an almost exclusively b-sheet containing aller- gen with 43% sequence identity to the C-terminal third of Phl p 1 [29,30]. Thermal denaturation of insect cell- expressed Phl p 1 was monitored by far-UV CD in the range of 20 (cid:1)C to 90(cid:1) and showed an irreversible unfolding transition, with a melting point of (cid:1) 42 (cid:1)C

FEBS Journal 272 (2005) 217–227 ª 2004 FEBS

219

T. Ball et al.

Eukaryotic expression of Phl p 1

despite preincubation with an excess of E. coli- expressed Phl p 1.

almost complete reduction of IgE binding to insect cell-expressed Phl p 1 was only observed for serum of patient 2, whereas considerable IgE reactivity of sera 1, 3 and 4 to insect cell-expressed Phl p 1 was observed

A

Whether insect cell-expressed Phl p 1 contains the IgE epitopes of a natural Phl p 1 preparation was inves- tigated by IgE competition experiments (Fig. 6B). Pre- incubation of sera from grass pollen allergic patients with insect cell-expressed Phl p 1 led to a strong or complete inhibition of IgE reactivity to natural Phl p 1 (Fig. 6B).

insect

Next we studied the influence of glycosylation on the IgE binding capacity of cell-expressed Phl p 1 (Fig. 6C). Five out of 10 patients showed stronger IgE reactivity to deglycosylated insect cell- expressed Phl p 1 than to the untreated protein (Fig. 6C, 1, 2, 3, 5, 7). Three patients exhibited com- parable IgE reactivity to both protein forms (Fig. 6C, #4, 9, 10) and two sera reacted stronger with the gly- cosylated allergen version (Fig. 6C, 6, 8). Finally, we studied the possible presence of cross-reactive IgE epi- topes between Phl p 1 and Phl p 2. Preincubation of sera from pollen allergic patients with insect cell- expressed, E. coli-expressed Phl p 1 or an unrelated control allergen (birch pollen allergen, rBet v 1) had no effect on IgE binding to rPhl p 2 (data not shown).

Allergenic activity of insect cell-expressed Phl p 1

B

The allergenic activity of insect cell-expressed Phl p 1 was analyzed by basophil histamine release (Fig. 7) and CD203c expression (data not shown). Basophils from two grass pollen allergic patients were exposed to different concentrations of E. coli- or insect cell- expressed Phl p 1 (Fig. 7A,B). In both patients, insect cell-expressed Phl p 1 was more potent, inducing hista- mine release at lower concentrations (10)3 lgÆmL)1) than E. coli-expressed Phl p 1 (10)2 lgÆmL)1). Meas- urement of CD203c expression on blood basophils of three Phl p 1 allergic patients confirmed these results. Incubation with insect cell-expressed Phl p 1 always led to stronger upregulation of CD203c than incuba- tion with E. coli-expressed Phl p 1 (data not shown).

Fig. 2. Biochemical and biophysical characterization of insect cell- (A) Glycan detection. Nitrocellulose blotted expressed Phl p 1. insect cell-expressed rPhl p 1 (ErPhl p 1), rPhl p 1 expressed in bacteria (PrPhl p 1), Creatinase and marker proteins (M) were sim- ultaneously stained for sugar moieties (blue) and reactive amino groups (fluorescent). Molecular masses are indicated on the left (B) SDS ⁄ PAGE containing insect cell-expressed Phl p 1 margin. before (ErPhl p 1-) and after (ErPhl p 1 +) enzymatic deglycosyla- tion. Lane M: Molecular mass marker.

FEBS Journal 272 (2005) 217–227 ª 2004 FEBS

220

T. Ball et al.

Eukaryotic expression of Phl p 1

Fig. 3. Comparison of E. coli- and insect cell-expressed Phl p 1 by circular dichroism spectroscopy. Far-UV CD spectra of E.coli- (grey) and insect cell-expressed Phl p 1 (black), expressed as mean resi- due ellipticity (y-axis), were recorded at 20 (cid:1)C in the wave length range displayed on the x-axis.

the sequence and phylogenetic relationship Fig. 5. Analysis of among group 1 and group 2 ⁄ 3 allergens from various grass spe- cies. A phylogenetic tree was reconstructed on the basis of amino- acid sequences of group 1 (Zea m 1: Zea mays, Cyn d 1: Cynodon dactylon, Pha a 1: Phalaris aquatica, Hol l 1: Holcus lanatus, Ory s 1: Oryza sativa, Lol p 1: Lolium perenne, Phl p 1: Phleum pra- tense) and group 2 ⁄ 3 allergens (Lol p 3: Lolium perenne, Dag g 3: Dactylis glomerata, Lol p 2: Lolium perenne, Tri a 3: Triticus aesti- vum, Phl p 2: Phleum pratense) using the PROTDIST and KITSCH program of the PHYLIP package.

Fig. 4. Crystal growth of insect cell-expressed Phl p 1.Phl p 1 crys- tallizes as thin plates of 0.35 · 0.35 · 0.15 mm.

and thus will yield the three-dimensional structure of this allergen.

Discussion

Phl p 1 belongs to the group 1 family of highly cross- reactive grass pollen allergens. The C-terminal domains of these allergens display sequence similarity to group

Table 1. Comparison of the IgE binding capacity of E. coli- and insect cell-expressed Phl p 1. IgE reactivity to recombinant Phl p 1, expressed in E. coli (PrPhl p 1) and baculovirus-infected insect cells (ErPhl p 1).

PrPhl p 1

ErPhl p 1

IgE binding (c.p.m.)

IgE binding (c.p.m.)

Patient number

only

1 2 3 4 5 6

158 4544 724 1010 963 1193

865 6370 2309 2980 3918 3889

Phl p 1 represents one of the most important respiratory allergens known to date. As Phl p 1 is a glycoprotein containing seven cysteines, we expressed the allergen in eukaryotic insect cells to obtain a post-translationally modified and folded protein. As demonstrated by mass spectrometry, glycan detection and deglycosylation experiments, insect cell-expressed Phl p 1 was obtained as a glycoprotein. The seemingly correct folding of insect cell-expressed Phl p 1 is demonstrated by the fol- lowing experiments: insect cell-expressed Phl p 1 but not E. coli-expressed Phl p 1 exhibited a secondary structure consisting mainly of b-sheets when analyzed by CD spectroscopy. Furthermore, insect cell-expressed Phl p 1 grew diffraction quality crystals

FEBS Journal 272 (2005) 217–227 ª 2004 FEBS

221

T. Ball et al.

Eukaryotic expression of Phl p 1

A

B

C

Fig. 6. (A) Superior IgE binding capacity of insect cell- vs. E. coli-expressed rPhl p 1. Sera from four grass pollen allergic patients were prein- cubated with bacterially expressed rPhl p 1 and exposed to dot-blotted bacterial recombinant (PrPhl p 1+) or eukaryotic recombinant Phl p 1 (ErPhl p 1+). PrPhl p 1- and ErPhl p 1- show the IgE binding without preadsorption of sera. (B) Inhibition of IgE binding to natural Phl p 1 (nPhl p 1) by insect cell-expressed Phl p 1. Sera from three grass pollen allergic patients (1–3) were tested for IgE reactivity to nitrocellulose- dotted eukaryotic recombinant Phl p 1 (ErPhl p 1) and natural Phl p 1 (nPhl p 1). Sera were preadsorbed with BSA (A), natural Phl p 1 (B), or insect cell-expressed Phl p 1 (C). (C) IgE binding capacity of deglycosylated insect cell-expressed Phl p 1.IgE reactivity of 10 sera from grass pollen allergic patients (1–10) to untreated (–) and deglycosylated (+) Phl p 1 is shown.

suggesting that

conformational

IgE epitopes

2 ⁄ 3 grass pollen allergens, another family of major grass pollen allergens that exhibit an immunoglobulin-like fold composed almost exclusively of b-sheet structure [29,30]. As shown by circular dichroism spectroscopy, Phl p 1 showed also almost exclusively b-sheet secon- dary structure. The results from limited proteolysis com- bined with mass spectrometry indicated a two domain organization of the protein with a C-terminal portion homologous to group 2 allergens. Despite these findings, Phl p 1 and Phl p 2 appear to represent immunological- ly independent allergens because significant crossreactiv- ity of IgE antibodies was not observed and both proteins belonged to different phylogenetic clusters.

The analysis of Phl p 1 IgE epitopes using recom- binant allergen fragments had indicated the presence

of several continuous IgE epitopes, of which the most prominent could be allocated to the C-terminal por- tion of Phl p 1 [25]. We have identified this portion as an intact domain by the limited proteolysis experi- intact and folded Phl p 1 ment domains represent the primary targets for patients’ IgE antibodies. The latter assumption is also suppor- ted by the fact that only an incomplete inhibition of IgE reactivity to the Phl p 1 allergen could be obtained after preincubation of patients’ sera with small recombinant protein fragments suggesting the importance of [25]. Therefore, we further tested the importance of struc- tural integrity on IgE binding capacity and allergenic activity of Phl p 1 by comparing insect cell-expressed

FEBS Journal 272 (2005) 217–227 ª 2004 FEBS

222

T. Ball et al.

Eukaryotic expression of Phl p 1

The importance of native tertiary structure for the IgE recognition of Phl p 1 seems to be a general princi- ple applicable to the most common respiratory aller- gens. For example, it has been demonstrated that disruption of native structure by fragmentation has led to a strong reduction of the IgE binding capacity and allergenic activity of the major birch pollen allergen, Bet v 1 [31], the cross-reactive calcium-binding allergens Aln g 4 [32] and Phl p 7 [33], the major mite allergen Der p 2 [34], and the major bovine allergen, Bos d 2 [35]. We consider the possibility that respiratory aller- gens may predominantely contain conformational IgE epitopes as important for at least three reasons. First, it indicates that respiratory sensitization occurs preferen- tially against intact and folded protein antigens which elute from respirable particles (e.g. pollen, mite faeces, animal dander). Second, our study emphasizes that it is important to choose an optimal expression strategy for obtaining native properly folded recombinant allergens which closely mimic the immunological properties of the natural counterparts for diagnostic purposes.

Induction of basophil histamine release with recombinant Fig. 7. Phl p 1 preparations. Granulocytes from patients (A, B) allergic to grass pollen were incubated with various concentrations (x-axis) of bacterial rPhl p 1 (PrPhl p 1) and eukaryotic rPhl p 1. The percentage of histamine released into the supernatant is displayed on the y-axis.

Finally, and perhaps most importantly, IgE recogni- tion of mainly conformational epitopes has important implications for the design of safe allergy vaccines with reduced allergenic activity. Disruption of the native structure of respiratory allergens allows for the main- tenance of important T cell epitopes of a given allergen and simultaneously preserves sequences relevant for the induction of protective antibody responses [36]. Con- trolled reduction of the fold of respiratory allergens by recombinant DNA technology or synthetic peptide chemistry thus seems to be a generally applicable strat- egy for the generation of recombinant allergy vaccines with reduced allergenic activity [37].

Experimental procedures

Materials, patients’ sera and antibodies

The Sf9 cell line was purchased from the German Collec- tion of Microorganisms and Cell Cultures (Braunschweig, Germany). After informed consent was obtained, sera were following the collected from Phl p 1 allergic patients, Helsinki guidelines. Allergenic patients were characterized by case history, skin prick test, and the demonstration of allergen-specific serum IgE antibodies by RAST (Pharmacia Diagnostics, Uppsala, Sweden). Natural group 1 grass pollen allergens from Timothy grass (nPhl p 1) and rye grass (nLol p 1) were purified as described [38]. Purified E. coli-expressed rPhl p 1, rPhl p 2 and rBet v 1 were obtained from BIOMAY (Vienna, Austria). A rabbit anti- Phl p 1 antiserum was obtained by immunizing rabbits with purified rPhl p 1 using complete Freunds’ adjuvant (Charles

Phl p 1 with E. coli-expressed Phl p 1. This compar- ison revealed a higher IgE-binding capacity and more pronounced allergenic activity of insect cell-expressed rPhl p 1 compared to E. coli-expressed rPhl p 1, as determined by basophil activation assays. Deglycosy- lation experiments demonstrate that the higher IgE- binding capacity and increased allergenic activity of insect cell-expressed Phl p 1 is due to intact structural integrity rather than to IgE recognition of carbo- hydrate moieties. In fact, we found that deglycosyla- tion rather increased the IgE binding capacity of Phl p 1. This may be due to the exposure of protein epitopes by removing carbohydrates from a poten- tially hyperglycosylated insect cell-expressed Phl p 1. On the other hand, it is unlikely that the authentic, plant-derived carbohydrates represent per se import- ant targets for patients’ IgE antibodies because insect cell-expressed Phl p 1 showed almost identical IgE reactivity as natural Phl p 1.

FEBS Journal 272 (2005) 217–227 ª 2004 FEBS

223

T. Ball et al.

Eukaryotic expression of Phl p 1

tion coefficient of the protein was calculated from the tyro- sine and tryptophan content [40].

Mass spectrometry

River, Kissleg, Germany). Alkaline phosphatase-conjugated goat anti-(rabbit Ig) and rabbit anti-(mouse Ig) serum was pur- chased from JacksonImmunoResearch Laboratories (West Grove, PA, USA), a mouse monoclonal anti-Hexahistidine antibody was obtained from Dianova (Hamburg, Germany). The 125I-labeled anti-human IgE immunoglobulins were pur- chased from Pharmacia Diagnostics. (Liquid Chromatography-Mass

Construction of recombinant baculovirus

Purified baculovirus-expressed Phl p 1 was analyzed by LC-MS Spectrometry) using a VYDAC (Hesperia, CA, USA) C4 column on a Waters HPLC 2690 (Waters Corp., Milford, MA, USA) which fed into an electrospray Thermo Finnigan LCQ quadrupole ion-trap mass spectrometer (ThermoQuest Inc., San Jose, CA, USA).

Limited proteolysis followed by LC-MS

The Phl p 1-encoding cDNA [16] was PCR amplified and cloned into the BamHI and KpnI restriction sites of the pBacPAK8 vector (Clontech Inc., Palo Alto, CA, USA), containing the baculovirus-derived ecdysteroid UDPgluco- syltransferase signal peptide [39] for enhanced secretion of the recombinant protein into the culture supernatant and a C-terminal His6-tag. The pBacPAK8 construct was con- firmed by DNA sequencing and cotransfected with the line- arized pBacPAK6 viral DNA (Clontech Inc., Palo Alto, CA, USA) into Sf9 insect cells. The clones with the highest level of protein secretion were chosen by Western blotting for virus amplification.

Expression and purification of rPhl p 1 from baculovirus-infected insect cells

Purified baculovirus-expressed Phl p 1 was subjected to lim- ited proteolysis by trypsin, Arg-C, Lys-C, Asp-N and Glu-C. Ten microliter aliquots containing 18 lm rPhl p 1 were diges- ted with protease in the following ratios 1 : 5, 1 : 15, 1 : 50, 1 : 150 and 1 : 500 (protease:rPhl p 1; w ⁄ w) for 1 h at room temperature. Proteolysis was halted by freezing at )70 (cid:1)C. Aliquots were analyzed by SDS ⁄ PAGE. Samples showing multiple bands, indicative for successful partial digest were then selected for further investigation by LC-MS. A Vydac C18 column was used on a Waters HPLC 2690 (Waters Corp.) followed by electrospray into a Thermo Finnigan LCQ Ion Trap Mass Spectrometer (ThermoQuest Inc.). The spectra were deconvoluted using Thermo Finnigan’s xcali- bur software and the spectra were also verified by hand cal- culations of charge states. The proteolytic fragments were identified using the paws software program (version 8.1.1, for Macintosh; Genomic SolutionsTM, Ann Arbor, MI, USA; http://bioinformatics.genomicsolutions.com/paws.html).

The expression of rPhl p 1 in insect cells was optimized by infecting Sf9 cells with different amounts of virus and by expression for various periods. Aliquots of the culture sup- ernatants and cell pellets were analyzed by SDS ⁄ PAGE and immunoblotting with a rabbit anti-Phl p 1 antiserum and a monoclonal anti-hexahistidine antibody. Rabbit anti- rPhl p 1 Igs were detected with an alkaline phosphatase (AP)-labeled goat anti-(rabbit Ig) antiserum. Bound anti- hexahistidine Igs were detected with AP-labeled rabbit anti- mouse Igs.

Detection of glycoproteins and deglycosylation treatment

FEBS Journal 272 (2005) 217–227 ª 2004 FEBS

224

Optimal expression of Phl p 1 was achieved by infection of 2 · 106 Sf9 cells per mL with recombinant baculovirus at a multiplicity of infection (MOI) of 5 with culturing in 3 L spinner ⁄ flasks in Insect-Xpress medium (BioWhittaker Inc., Walkersville, MD, USA) containing 2% fetal bovine serum. At day two postinfection, supernatants were separ- ated by centrifugation (8000 g, 4 (cid:1)C, 30 min) and dialyzed against start buffer [50 mm sodium phosphate (pH 8.0), 300 mm NaCl] at 4 (cid:1)C overnight. Insect cell-expressed rPhl p 1 was purified using Ni-nitrilotriacetic acid superflow matrix (Qiagen, Hilden, Germany) under nondenaturing conditions by stepwise elution with increasing (20–250 mm) imidazole concentrations. The eluted samples were dialyzed against 10 mm Tris HCl (pH 8.0), 100 mm NaCl and con- centrated by Centricon ultrafiltration (Millipore, Bedford, MA, USA). Protein concentrations of purified samples were estimated using BCA reagent (Pierce Chemicals, Rockford, IL, USA) and UV absorption at 280 nm. The molar extinc- Purified E. coli- and insect cell-expressed Phl p 1 proteins were separated by SDS ⁄ PAGE and transferred to nitrocellu- lose followed by detection of sugars using a DIG glycan ⁄ pro- tein double labeling kit (Boehringer Mannheim GmbH, Mannheim, Germany). Briefly, glycans were oxidized to pro- duce aldehyde groups allowing the covalent attachment of the steroid hapten digoxigenin (DIG). The latter was then detected using horseradish peroxidase-conjugated anti-digo- xigenin Igs yielding a blue color reaction. Creatinase and bacterial rPhl p 1 were used as nonglycosylated controls which were stained by labeling of amino groups with fluo- rescein and detection with alkaline phosphatase-conjugated anti-fluorescein Igs (Boehringer) giving a brown color reac- tion. Enzymatic deglycosylation was performed with gluta- thione S-transferase (GST)–PNGase F (Hampton Research, CA, USA) by using reaction ratios of GST–PNGase F:glyco- protein of 1 : 2. Deglycosylation was carried out for 15 h at

T. Ball et al.

Eukaryotic expression of Phl p 1

room temperature in 10 mm Tris (pH ¼ 8.0), 100 mm NaCl. The GST–PNGase F was removed from the target protein using glutathione-Sepharose.

Circular dichroism (CD) measurements

excess [7]. Three micrograms of the purified recombinant proteins were dotted onto nitrocellulose strips and incuba- ted with sera from Phl p 1 allergic patients. Bound IgE antibodies were detected with 125I-labeled anti-(human IgE) Igs (Pharmacia) and quantified by c-counting (Wallac, LKB, Turku, Finland) [25].

IgE inhibition experiments under conditions of antigen excess were performed as described [7]. Patients’ sera were incubated with 5 lgÆmL)1 of each allergen (or the same amount of BSA for control purposes) overnight at 4 (cid:1)C. The next day, preincubated sera were exposed to 3 lg of nitrocellulose-dotted natural Phl p 1, rPhl p 2, E. coli and baculovirus expressed Phl p 1. Bound serum IgE was detec- ted as described for the IgE immunoblotting and quantified by c-counting [25].

Basophil activation experiments

All recombinant proteins were subjected to CD analysis to access stability and secondary structure composition. Far UV-CD spectra were collected on a Jasco-J720 spectropola- rimeter (Jasco, Tokyo, Japan) at room temperature, at final protein concentrations of 10–25 lm in either 0.5 or 0.01 mm path-length quartz cuvettes. The molar ellipticity was calculated according to [h] ¼ ⁄ 10cl, where h is the ellip- ticity, l is the cuvette path-length in cm and c is the protein concentration in molÆL)1. Three independent measurements were recorded and averaged for each spectral point in all experiments. Thermal denaturation was monitored in the range of 20 (cid:1)C to 90 (cid:1)C. The reversibility of the unfolding process was checked by measuring the CD signal upon cooling to the starting temperature.

Phylogenetic analysis of the relationships among group 1 and group 2/3 allergens from various grass species

Granulocytes were isolated from heparinized blood samples of individuals allergic to Phl p 1 by dextran sedimentation. The capacity of E. coli- and insect cell-expressed Phl p 1 to induce basophil degranulation was tested by incubation of granulocytes with various concentrations of the purified proteins and by measuring histamine released into the cell-free supernatant by radioimmunoassay (Immunotech, Marseille, France). Histamine release was measured in trip- licates and expressed as a percentage of total histamine determined after cell lysis, as described [48]. Up-regulation of CD203c expression on basophils after allergen exposure was measured as described [26].

Acknowledgements

Multiple alignment of sequences homologous to Phl p 1 as identified by a BLAST search [41] was generated by using clustalx [42]. A distance matrix among sequences was constructed using the protdist program of the phylip 3.6a2 [43,44] package. The distance matrix was used as input to the KITSCH program from the phylip package for the construction of a phylogenetic tree. This program implements the Fitch–Margoliash least-squares methods with the assumption of an evolutionary clock.

study was

SDS/PAGE analysis and immunoblotting

technical assistance of We acknowledge the skillful circular dichroism Miriam Gulotta regarding the experiments. This supported by grants Y078GEN, F01801, F01809, J1835 and J2122 of the Austrian Science Fund, by the CeMM Project of the Austrian Academy of Sciences, and by a research grant from BIOMAY, Vienna, Austria.

References

1 Wills-Karp M, Santeliz J & Karp CL (2001) The germ- less theory of allergic disease: revisiting the hygiene hypothesis. Nat Rev Immunol 1, 69–75. Samples were resolved on 12.5% polyacrylamide gels under reducing conditions. Proteins were stained with Coomassie blue or transferred to nitrocellulose membranes [45]. Blot- ted proteins were probed with sera from Phl p 1 allergic patients¢, anti-His Igs, rabbit anti-Phl p 1 antiserum and the corresponding preimmune serum. Patients’ bound IgE antibodies were detected with 125I-labeled anti-human IgE [46], anti-His Igs with AP-conjugated rabbit anti-(mouse Ig), and bound rabbit Igs with an AP-conjugated goat anti- (rabbit IgG) serum [47].

IgE-binding capacity and cross-reactivity of allergens as determined by nondenaturing dot-blot experiments

FEBS Journal 272 (2005) 217–227 ª 2004 FEBS

225

The IgE reactivity of the recombinant Phl p 1 molecules was determined by dot-blot under conditions of antigen 2 Wu¨ thrich B, Schindler C, Leuenberger P & Ackermann- Liebrich P (1995) Prevalence of atopy and pollinosis in the adult population of Switzerland (SAPALDIA study). Swiss Study on Air Pollution and Lung Diseases in Adults. Int Arch Allergy Immunol 106, 149–156. 3 Andersson K & Lidholm J (2003) Characteristics and immunobiology of grass pollen allergens. Int Arch Allergy Immunol 130, 87–107.

T. Ball et al.

Eukaryotic expression of Phl p 1

4 Johnson P & Marsh DG (1965) ‘Isoallergens’ from rye 16 Laffer S, Valenta R, Vrtala S, Susani M, vanRee R, grass pollen. Nature (London) 206, 935–937.

Kraft D, Scheiner O & Ducheˆ ne M (1994) Complemen- tary DNA cloning of the major allergen Phl p 1 from timothy grass (Phleum pratense); recombinant Phl p 1 inhibits IgE binding to group 1 allergens from eight dif- ferent grass species. J Allergy Clin Immunol 94, 689– 698. 5 Freidhoff LR, Ehrlich-Kautzky E, Grant JH, Meyers DA & Marsh DG (1986) A study of human immune response to Lolium perenne (rye) pollen and its compo- nents, Lol, p. 1 and Lol, p. 2 (rye 1 and rye 2). J Allergy Clin Immunol 78, 1190–1201.

17 Suphioglu C & Singh MB (1995) Cloning, sequencing and expression in Escherichia coli of Pha a 1 and four isoforms of Pha a 5, the major allergens from canary grass pollen. Clin Exp Allergy 25, 853–865. 18 Xu H, Theerakulpisut P, Goulding N, Suphioglu C, 6 Laffer S, Duchene M, Reimitzer I, Susani M, Mannhal- ter C, Kraft D & Valenta R (1996) Common IgE-epi- topes of recombinant Phl p 1, the major timothy grass pollen allergen and natural group I grass pollen isoaller- gens. Mol Immunol 33, 417–426. 7 Niederberger V, Laffer S, Froeschl R., Kraft D,

Singh MB & Bhalla PL (1995) Cloning, expression and immunological characterization of Ory s 1, the major allergen of rice pollen. Gene 164, 255–259.

Rumpold H, Kapiotis S, Valenta R. & Spitzauer S (1998) IgE antibodies to recombinant pollen allergens (Phl p 1, Phl p 2, Phl p 5, and Bet v 2) account for a high percentage of grass pollen-specific IgE. J Allergy Clin Immunol 101, 258–264.

19 Smith PM, Suphioglu C, Griffith IJ, Theriault K, Knox B & Singh MB (1996) Cloning and expression in yeast Pichia pastoris of a biologically active form of Cyn d 1, the major allergen from Bermuda grass pollen. J Allergy Clin Immunol 98, 331–343.

8 Grote M, Dolecek C, vanRee R., Kraft D, Scheiner O & Valenta R (1994) Immunogold electron microscopic localization of timothy grass (Phleum pratense) pollen major allergens Phl p 1 and Phl p 5 after anhydrous fixation in acrolein vapor. J Histochem Cytochem 42, 427–431. 20 Schramm G, Bufe A, Petersen A, Haas H, Schaak M & Becker WM (1997) Mapping of the IgE-binding epi- topes on the recombinant major group 1 allergen of vel- vet grass pollen, rHol 1 1. J Allergy Clin Immunol 99, 781–787.

9 Vrtala S, Grote M, Duchene M, vanRee R., Kraft D, Scheiner O & Valenta R (1993) Properties of tree and grass pollen allergens; reinvestigation of the linkage between solubility and allergenicity. Int Arch Allergy Immunol 102, 160–169.

10 Grote M, Vrtala S, Niederberger V, Wiermann R, Valenta R & Reichelt R (2001) Release of allergen- bearing cytoplasm from hydrated pollen: a mechanism common to a variety of grass (Poaceae) species revealed by electron microscopy. J Allergy Clin Immunol 108, 109–115. 11 Taylor PE, Flagan RC, Valenta R & Glovsky MM 21 Vrtala S, Susani M, Sperr WR, Valent P, Laffer S, Dolecek C, Kraft D & Valenta R (1996) Immuno- logical characterization of purified recombinant timothy grass pollen (Phleum pratense) allergens (Phl p 1, Phl p 2, Phl p 5). J Allergy Clin Immunol 97, 781–787. 22 Schenk S, Breiteneder H, Susani M, Najafian N, Laffer S, Duchene M, Valenta R, Fischer G, Scheiner O, Kraft D & Ebner C (1995) T-cell epitopes of Phl p 1, major pollen allergen of timothy grass (Phleum pratense): evi- dence for cross-reacting and non-crossreacting T-cell epitopes within grass group 1 allergens. J Allergy Clin Immunol 96, 986–996.

(2002) Release of allergens as respirable aerosols: a link between grass pollen and asthma. J Allergy Clin Immu- nol 109, 51–56.

23 Petersen A, Becker WM, Moll H, Blu¨ mke M & Schlaak M (1995) Studies on the carbohydrate moieties of the timothy grass pollen allergen Phl p 1. Electrophoresis 16, 869–875. 12 Suphioglu C, Singh MB, Taylor P, Bellomo R, Holmes P, Puy R & Knox RB (1992) Mechanism of grass-pol- len-induced asthma. Lancet 339, 569–572.

24 Petersen A, Schramm G, Schlaak M & Becker WM (1998) Post-translational modifications influence IgE reactivity to the major allergen Phl p 1 of timothy grass pollen. Clin Exp Allergy 28, 315–321.

13 Knox RB, Suphioglu C, Taylor P, Desai R., Watson HC, Peng JL & Bursill LA (1997) Major grass pollen allergen Lol p 1 binds to diesel exhaust particles: impli- cations for asthma and air pollution. Clin Exp Allergy 27, 246–251. 14 Perez M, Ishioka GY, Walker LE & Chesnut RW

25 Ball T, Fuchs T, Sperr WR, Valent P, Vangelista L, Kraft D & Valenta R (1999) B cell epitopes of the major timothy grass pollen allergen, Phl, p. 1, revealed by gene fragmentation as candidates for immunother- apy. FASEB J 13, 1277–1290. (1990) cDNA cloning and immunological characteriza- tion of the rye grass allergen Lol p 1. J Biol Chem 265, 16210–16215. 26 Hauswirth AW, Natter S, Ghannadan M, Majlesi Y, 15 Broadwater AH, Bedinger PA, Rubinstein AL, Chay

FEBS Journal 272 (2005) 217–227 ª 2004 FEBS

226

CH & Klapper DG (1993) Zea m 1, the maize homolog of the allergen-encoding Lol p 1 gene rye grass. Gene 131, 227–230. Schernthaner GH, Sperr WR, Buehring HJ, Valenta R & Valent P (2002) Recombinant allergens promote expression of CD203c on basophils in sensitized indivi- duals. J Allergy Clin Immunol 110, 102–109.

T. Ball et al.

Eukaryotic expression of Phl p 1

27 Cohen SL (1996) Domain elucidation by mass spec- trometry. Structure 4, 1013–1016.

28 Cohen SL, Ferre-D’Amare AR, Burley SK & Chait BT (1995) Probing the solution structure of the DNA-bind- ing protein Max by a combination of proteolysis and mass spectrometry. Protein Sci 4, 1088–1099. 37 Valenta R (2002) The future of antigen-specific immu- notherapy of allergy. Nat Rev Immunol 2, 446–453. 38 Suck R, Hagen S, Cromwell O & Fiebig H (1999) Rapid and efficient purification of Phleum pratense major allergens Phl p 1 and group 2 ⁄ 3 using a two-step procedure. J Immunol Methods 229, 73–80. 39 Murphy CI, McIntire JR, Davis D vR, Hodgdon H,

29 Fedorov AA, Fedorov EV, Dolecek C, Susani M, Val- enta R & Almo SC (1997) Crystal structure of Phl p 2, a major timothy grass (Phleum pratense) pollen aller- gen. Structure available in the PDB (1WHO). 30 De Marino S, Castiglione Morelli MA, Fraternali F, Seals JR & Young E (1993) Enhanced expression, secre- tion, and large-scale purification of recombinant HIV-1 gp120 in insect cells using the baculovirus egt and p67 signal peptides. Protein Expr Purif 4, 349–357. 40 Gill SC & Hippel PH (1989) Calculation of protein

extinction coefficients from animo acid sequence data. Anal Biochem 182, 319–326.

Tamborini E, Musco G, Vrtala S, Dolecek C, Arosio P, Valenta R & Pastore A (1999) An immunoglobulin-like fold in a major plant allergen: the solution structure of Phl p 2 from timothy grass pollen. Structure 7, 943–952. 31 Vrtala S, Hirtenlehner K, Vangelista L, Pastore A, 41 Altschul SF, Gish W, Miller W, Myers EW & Lipman DJ (1990) Basic local alignment search tool. J Mol Biol 215, 403–410.

42 Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F & Higgins DG (1997) The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 24, 4876– 4882. 43 Felsenstein J (1989) Phylip (Phylogeny Inference Pack- age), Version 3.2. Cladistics 5, 164–166.

44 Felsenstein J (1993) phylip (Phylogeny Inference Pack- age), Version 35c, Department of Genetics. University of Washington, Seattle.

Eichler HG, Sperr WR, Valent P, Ebner C, Kraft D & Valenta R (1997) Conversion of the major birch pollen allergen, Bet v, 1, into two nonanaphylactic T cell epi- tope-containing fragments: candidates for a novel form of specific immunotherapy. J Clin Invest 99, 1673–1681. 32 Hayek B, Vangelista L, Pastore A, Sperr WR, Valent P, Vrtala S, Niederberger V, Twardosz A, Kraft D & Valenta R (1998) Molecular and immunologic charac- terization of a highly cross-reactive two EF-hand cal- cium-binding alder pollen allergen, Aln g 4: structural basis for calcium-modulated IgE recognition. J Immunol 161, 7031–7039. 33 Verdino P, Westritschnig K, Valenta R & Keller W 45 Towbin H, Staehlin T & Gordon J (1979) Electrophore- tic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 76, 4350–4354.

(2002) The cross-reactive calcium-binding pollen aller- gen, Phl p 7, reveals a novel dimer assembly. EMBO J 21, 5007–5016. 34 Takai T, Mori A, Yuuki T, Okudaira H & Okumura Y

46 Valenta R, Duchene M, Ebner C, Valent P, Sillaber C, Deviller P, Ferreira F, Tejkl M, Edelmann H, Kraft D & Scheiner O (1992) Profilins constitute a novel family of functional plant pan-allergens. J Exp Med 175, 377– 385.

(1999) Non-anaphylactic combination of partially deleted fragments of the major house dust mite allergen Der f 2 for allergen-specific immunotherapy. Mol Immu- nol 36, 1055–1065. 35 Zeiler T, Taivainen A, Rytkonen M, Rautiainen J,

47 Flicker S, Laffer S, Steinberger P, Alhani B, Zhu Y, Laukkanen ML, Keinanen K, Kraft D & Valenta R (2000) Engineering, purification and applications of His- tagged recombinant antibody fragments with specificity for the major birch pollen allergen, Bet v, 1. Biol Chem 381, 39–47. Karjalainen H, Mantyjarvi R, Tuomisto L & Virtanen T (1997) Recombinant allergen fragments as candidate preparations for allergen immunotherapy. J Allergy Clin Immunol 100, 721–727. 36 Valenta R & Kraft D (2001) Recombinant allergen

FEBS Journal 272 (2005) 217–227 ª 2004 FEBS

227

molecules: tools to study effector cell activation. Immu- nol Rev 179, 119–127. 48 Valent P, Besemer J, Muhm M, Majdic O, Lechner K & Bettelheim P (1989) Interleukin 3 activates human blood basophils via high-affinity binding sites. Proc Natl Acad Sci USA 86, 5542–5547.