Identification, Characterization, and Epitope Mapping of Tree Nut Allergens Jason M

Florida State University Libraries

Electronic Theses, Treatises and Dissertations The Graduate School

2006 Identification, Characterization, and Epitope Mapping of Tree Nut Allergens Jason M. Robotham

Follow this and additional works at the FSU Digital Library. For more information, please contact [email protected]

THE FLORIDA STATE UNIVERSITY

COLLEGE OF ARTS AND SCIENCES

IDENTIFICATION, CHARACTERIZATION, AND EPITOPE MAPPING OF

TREE NUT ALLERGENS

JASON M. ROBOTHAM

A Dissertation submitted to the Department of Biological Science in partial fulfillment of the requirements for the degree of Doctor of Philosophy

Degree Awarded: Spring Semester, 2006

The members of the Committee approve the Dissertation of Jason M. Robotham

defended on January 27, 2006.

______Kenneth H. Roux Professor Directing Dissertation

______Shridhar K. Sathe Outside Committee Member

______Peter G. Fajer Committee Member

______Thomas C.S. Keller III Committee Member

______Kenneth A. Taylor Committee Member

Approved:

______Timothy S. Moerland, Chair, Department of Biological Science

______Joseph Travis, Dean, College of Arts and Sciences

The Office of Graduate Studies has verified and approved the above named committee members

ACKNOWLEDGEMENTS

I would like to thank Kenneth Roux for being a patient advisor, mentor, and most importantly a good friend over the course of my dissertation work. I would like to thank my committee members Shridhar Sathe, Thomas Keller, Peter Fajer, and Kenneth Taylor for their assistance in my development as a scientist. Dr. Suzanne Teuber made much of my research possible through a fruitful collaboration. I am also grateful to Drs. Kirsten Beyer and Hugh Sampson for their generosity in supplying our lab with sera and general advice and Dr. Rudolf Jung for supplying me with the soybean glycinin gene when it seemed like nobody else would. I’d like to thank all the former and current lab technicians, graduate students, and postdoctoral fellows from the Roux lab. In particular, I’d like to thank Shirley Roux for helping me to adjust to graduate school early on, Pallavi Tawde for our many thought- filled conversations regarding our work together, Ping Zhu for demonstrating a work ethic that I still strive to achieve, Fang Wang for helping introduce me to Molecular Biology, and both Dillon Fritz and Henry Grise for their support as scientist and friends. I’d like to thank all the support personnel that made the completion of this project possible, including the Analytical lab, the Sequencing lab, the Cloning lab, and the KLB facilities. In particular, I’d like to thank Margaret Seavy for her patience and hard work throughout my many efforts at protein purification and Rani Dhanarajan for teaching me nearly everything I know about the technical aspects of Molecular biology and cell culturing. I’d also like to thank my fellow graduate students throughout the department for their friendship and for creating both a work and non-work community of which I am proud to say I was a part. Finally, a special thanks goes out to my family. I could never have succeeded in graduate school without the advice and support of my older brother and closest friend, Claude Robotham. Similarly, I could have never succeeded in life without the unconditional love and support of my parents, Maria and Kenrick Robotham. How they managed to put two sons through a combined 20 years of higher education is beyond me and if I could love them more I would, but it is not possible.

iii

TABLE OF CONTENTS

List of Tables ...... Page vi

List of Figures ...... Page vii

Abstract ...... Page viii

1. FOOD HYPERSENSITIVITY AND TREE NUT ALLERGY...... Page 1

2. IDENTIFICATION AND CHARACTERIZATION OF IMPORTANT ALLERGENS IN THE ENGLISH WALNUT AND CASHEW TREE NUT BELONGING TO THE 2S ALBUMIN AND 11S GLOBULIN FAMILIES

Introduction ...... Page 4

Methods ...... Page 5

Results ...... Page 12

Discussion ...... Page 18

3. LINEAR IGE-EPITOPE MAPPING OF ENGLISH WALNUT AND CASHEW NUT 2S ALBUMIN AND 11S GLOBULIN ALLERGENS

Introduction ...... Page 22

Methods ...... Page 23

Results ...... Page 26

Discussion ...... Page 40

4. IDENTIFICATION OF A CASHEW 11S GLOBULIN CONFORMATIONAL EPITOPE

Introduction ...... Page 42

Methods ...... Page 44

Results ...... Page 50

Discussion ...... Page 57

5. HOMOLOGY MODELING OF VARIOUS 11S GLOBULIN ALLERGENS

Introduction ...... Page 61

Methods ...... Page 62

Results ...... Page 62

Discussion ...... Page 69

CONCLUSION ...... Page 72

REFERENCES ...... Page 74

BIOGRAPHICAL SKETCH ...... Page 83

LIST OF TABLES

2.1 Primers used for subcloning of walnut and cashew 2S albumin and 11S globulin genes into the pMAL-c2-His expression vector...... Page 7

2.2 Clinical characteristics of cashew and walnut-allergic subjects...... Page 10

2.3 Proteins showing identity and similarity to Ana o2...... Page 13

2.4 Proteins showing identity and similarity to Ana o 3...... Page 13

3.1 Mutational analysis of the Jug r 1 IgE-binding epitope, E1 ...... Page 28

3.2 Ana o 3 IgE-binding peptides and relative intensity...... Page 31

3.3 Jug r 4 IgE-binding peptides and relative intensity ...... Page 33

3.4 Ana o 2 IgE-binding peptides and relative intensity...... Page 34

4.1 Primers used for the amplification and subcloning of cashew and soybean Page 45 11S globulin genes and constituent subunits ......

4.2 Primers used to amplify Ana o 2 and Gy1 glycinin large and small subunits Page 46 for use in blunt end ligation reactions......

5.1 Percent identity and similarity of five 11S globulin allergens...... Page 63

LIST OF FIGURES

2.1 Purification of the expressed recombinant Ana o 2-MBP fusion protein... Page 14

2.2 Western immunoblotting and rJug r 1 inhibition blotting of walnut extract in reducing conditions with patient #3 sera ...... Page 15

2.3 Western immunoblotting and rJug r 4 inhibition blotting of walnut extract in reducing conditions with patient #3 sera...... Page 16

2.4 Western immunoblotting and rAna o 3 inhibition blotting of cashew extract in reducing conditions with patient #31 sera...... Page 17

2.5 Western immunoblotting and rAna o 2 inhibition blotting of cashew extract in reducing conditions with patient #31 sera ...... Page 18

3.1 Schematic representation of solid phase overlapping peptide synthesis of Jug r 1 ...... Page 27

3.2 IgE binding analysis of Jug r 1 SPOTs ...... Page 27

3.3 Isolation and inhibition of E1-specific IgE...... Page 30

3.4 Amino acid sequence and linear epitope comparison of cashew (Ana o 3), walnut (Jug r 1), sesame, and mustard seed (Sin a 1, segment) 2S albumins Page 32

3.5 Mutational analysis of two immunodominant Ana o 2 IgE-binding peptides Page 35

3.6 Comparison of the IgE-binding regions of Jug r 4 and Ana o 2 with various allergenic 11S globulins...... Page 38

4.1 BLAST alignment of the deduced aa sequences of cashew (Ana o 2) and soybean (Gy1 glycinin) 11S globulins...... Page 51

4.2 A schematic representation of the A2 LG-Gy1 SM (left) and Gy1 LG-A2 SM (right) chimeric molecules ...... Page 51

4.3 Monoclonal antibody dot blot immunoblotting of rAna o 2 and rGy1 Page 53 full-length, subunit, and chimeric fusion proteins ......

4.4 Binding of mAbs 1F5 and 2B5 to rAna o 2 and rA2 LG-Gy1 SM chimera Page 54 under various denaturing conditions......

4.5 Binding of mAbs 1F5 and 2B5 to whole cashew extract and HPLC Page 55 fraction #30 under various denaturing conditions ......

vii

4.6 Inhibition of Ana o2-specific human IgE with mAbs 2B5 and 1F5...... Page 56

4.7 Inhibition of mAb 2B5 reactivity with sera from two cashew sensitive Page 57 patients…………………………..…………………………...……………

5.1 Homology modeling of four allergenic 11S globulins and surface mapping Page 64 of their respective IgE-binding linear epitopes…………………………...

5.2 Surface epitope mapping of the Ana o 2 trimer model…………………… Page 66

5.3 Sequence alignment of six identified 11S globulin allergenic “hot spots” Page 67

5.4 Surface mapping of six identified 11S globulin allergenic “hot spots” Page 67

viii

ABSTRACT

Recent years have seen a dramatic worldwide increase in allergies and asthma that is now described as reaching epidemic proportions.1, 2 Food allergies affect approximately 2% of the adult population and up to 6% of the pediatric population. Tree nut allergies, in particular, affect about 0.5% of the US population.3 Five commonly consumed tree nuts in the US are almonds, walnuts, cashews, pistachios, and pecans, all of which are allergenic to a subset of the consuming population. Among the proteins thus far associated with tree nut allergies include the seed storage proteins belonging to the 2S albumin and 11S globulin gene families. Complementary DNA expression libraries were created from English walnut and cashew nut embryos. To identify the 2S albumin and 11S globulin genes in both nuts, the libraries were used either directly as targets for degenerate primers in PCR ‘fishing’ experiments or transferred to nitrocellulose membranes and screened with nut allergenic patient sera. Once the genes were identified and amplified, they were modified with restriction enzymes, ligated into an expression vector, and expressed as fusion proteins. These purified fusion proteins were used to screen for walnut and cashew nut-allergic patient IgE-reactivity in direct immunoblots and to identify the native counterparts of these proteins in crude nut extracts via inhibition immunoblots. Synthetic overlapping peptide libraries were created using the SPOTs technology and the linear IgE- binding epitopes of each allergen subsequently determined by screening with pooled patient sera. For the walnut 2S albumin and cashew 11S globulin the most reactive epitopes were examined further and amino acids critical or influential for IgE binding identified. Some reactive peptides appeared to be conserved in sequence position and composition among the 2S albumins and 11S globulins of the walnut, cashew, and other previously characterized plant allergens when analyzed in comparative epitope maps. Structural motifs that might explain the conservation of these antibody-binding regions among the proteins were identified through homology modeling experiments. Largely unexplored in terms of allergen epitope characterization are those of a conformational nature, comprised of amino acids distant in the proteins primary structure but adjacent once the protein folds. A chimeric molecule-based strategy for identifying the existence and proposed location of a conformational IgE-binding epitope was successfully carried out for the cashew 11S globulin.

CHAPTER 1

FOOD HYPERSENSITIVITY AND TREE NUT ALLERGY

A body’s ability to fend off potentially harmful agents can allow it to thrive under otherwise difficult circumstances. As humans (Homo sapiens), our body’s most formidable line of defense is its immune system. While many common infections can be controlled through innate (non-adaptive) immunity, jawed vertebrates are also equipped with a more versatile means of defense through adaptive (acquired) immunity. One such adaptive immune response is the ability of lymphocytes to detect and help mediate the eventual elimination of harmful foreign organisms and molecules (antigens). In some cases, an adaptive immune response is initiated when an antigen presenting cell (APC) encounters a foreign antigen. The three major APCs are macrophages, dendritic cells, and B-lymphocytes (B-cells). In the case of a patrolling B-cell, antibody receptors on the surface of the cell can bind antigens circulating in the blood or the lymphoid tissues. This interaction can trigger both cellular and humoral immune responses involving the internalization of the antigen, activation of the B-cell, and the presentation of small peptide fragments in conjunction with major histocompatibility complex (MHC) to a T- lymphocyte (T-cell). The type of antigen and class of MHC with which it is presented to the T- cell will, in part, determine what type of effector function will be carried out. For a humoral immune response to occur, the B-cell will present antigen via MHC class II to a CD4+ helper T- cell (Th). These Th cells will proliferate and produce cytokines that will help activate those B- cells bearing the appropriate antigen specific receptors. Successful interaction between these two cells can lead to the eventual proliferation and differentiation of some B-cells into antibody secreting plasma cells. The isotype of the secreted immunoglobulin is dependent on the profile of the cytokines secreted by the Th cell, which, in turn, is predicated on the signal initially sent by the antigen presenting B-cell. Those B-cell presented antigens that interact with Th1 cells can cause proliferation and differentiation, as described earlier, and stimulate the production of the protective immunoglobulin-G class of antibodies that help neutralize the antigen without causing allergic symptoms. In contrast, those otherwise innocuous molecules (i.e., allergens that cause Type-I hypersensitive reactions) that interact with Th2 cells can stimulate plasma cell production of IgE antibodies.

There are two major classes of allergens, those that enter the body through the respiratory tract (aeroallergens such as pollens, animal danders, and dust mites) and those that are ingested (food allergens).4, 5 IgE-mediated hypersensitivity reactions to food occur rapidly following ingestion or inhalation of a multivalent dietary antigen to which an individual is sensitized. This class of antibody binds to specific cells in the blood (basophils) or other tissues of the body (mast cells) via high affinity FcεRI receptors on the cells for the Fc portion of the immunoglobulin.6 Subsequent exposure to even trace amounts of the same or a cross-reactive allergen causes the receptors on the basophils and mast cells to be cross-linked and induces the rapid release of histamines and other highly active molecules (e.g., histamines and leukotrines) to cause the allergic symptoms (hives, rash, mucus secretion, smooth muscle contraction, bronchial and tracheal constriction, and, in extreme cases, death).5 Food allergies affect approximately 2% of the adult population and up to 6% of the pediatric population and are estimated to cause more than 100 fatalities/year in the United States.7-9 According to a 1995 report by the Food and Agriculture Organization for the United Nations, eight foods have been identified as the most frequent human food allergens and account for ~90% of food related allergies.3 These foods are often referred to as “The Big 8” and include milk, eggs, fish, crustacea, wheat, peanuts, tree nuts, and soy. While allergies to milk, egg, wheat, and soy in infants are usually outgrown by age three, allergies to fish, shellfish, peanuts, and tree nuts often persist throughout life.5 Up to 1% of the US population is allergic to peanuts, tree nuts, or both.10 Walnuts and cashew nuts are associated with IgE-mediated anaphylaxis and are the two most commonly reported tree nut allergies in the US.11 The major allergens of cod, shrimp, soy, peanut, milk, and egg have been described in some detail and the genes encoding these proteins have been cloned.5 In contrast, until our laboratories began working in this area, relatively little had been reported about the proteins responsible for tree nut allergies even though 25% of those experiencing tree nut allergies have severe reactions.9, 12-18 The proteins most associated with tree nut allergies thus far include the seed storage proteins, lipid transfer proteins, and several other pathogenesis-related proteins (reviewed in16). The concept of classifying plant proteins, including those of nuts and legumes, according to their solubility was formalized by T.B. Osborne in 1924.22 Osborne classified proteins into four groups based on their solubility in water (albumins), dilute salt solutions (globulins), alcohol/water mixtures, and dilute alkali (glutelins).23 The sequential fractionation of plant

proteins following this solubility scheme is termed Osborne fractionation. In seeds of almost all mono- and dicotyledonous plants the salt-soluble globulin fractions are highly enriched in storage proteins. Storage proteins are deposited in seeds during development to provide a store of amino acids (aa) and carbon skeletons for germination and seedling growth.23 There are two groups of globulin storage proteins with sedimentation coefficients of about 7 to 8 and 11 to 12. These families of seed storage proteins are commonly referred to as 7S and 11S globulins or vicilins and legumins, respectively. Storage proteins found in the water-soluble albumin fraction of plant proteins typically have a sedimentation coefficient of about 2, are widespread in seeds of dicotyledonous plants, and are simply termed 2S albumins.23 The most fundamental quality of any allergen, from an immunological perspective, is the nature of its antibody binding sites (epitopes). The two broadest classes of B-cell epitopes are linear (also termed sequential or continuous) and conformational (discontinuous). Linear epitopes may be defined as short contiguous stretches of aa that are reactive with antibody (Ab) whereas conformational epitopes are frequently composed of non-contiguous peptide segments that are only proximally associated and correctly presented in the native folded protein.4, 5 Because so few tree nut allergens had been identified, no epitope mapping studies were reported at the time this project was undertaken. In this study, we identify, biochemically and clinically characterize, and map the linear IgE-binding epitopes of allergenic proteins belonging to the 2S albumin and 11S globulin families of the English Walnut (Juglans regia) and cashew nut (Anacardium occidentale). We also prove the existence of at least one IgE-binding conformational epitope on the cashew nut 11S globulin. Our aim is to ultimately create a composite epitope map comparing the cognate sequences of the homologous allergens among the studied tree nuts and discern patterns or protein motifs that are commonly recognized among each. In certain instances we set out to identify aa both necessary and influential for patient IgE binding to occur and demonstrated the possibility of genetically reengineering the genes encoding the offending proteins to eliminate the aa that induce allergic reactions and enhance the features that provide protective (non- or anti-allergenic) immunity.4, 24- 27 Our efforts may lead to advances in the diagnosis and treatment, and particularly in the prevention of the development of allergy, thus benefiting the agricultural and food communities by lessening the regulatory pressure on contamination and avoidance.

CHAPTER 2

IDENTIFICATION AND CHARACTERIZATION OF MAJOR ALLERGENS IN THE ENGLISH WALNUT AND CASHEW NUT BELONGING TO THE 2S ALBUMIN AND 11S GLOBULIN FAMILIES

Introduction

Recent data from the Food Allergy and Anaphylaxis Network’s Peanut and Tree Nut Registry (FAAN), reveals that 34% of nut-allergic patients self-reported reactions to walnut, 20% to cashew, 15% to almond, 9% to pecan, and 7% to pistachio.28 The fact that all of the food induced fatal allergic reactions for individuals over the age of 6, reported to a US national registry, have been caused by either peanuts or tree nuts demonstrates the seriousness of tree nut allergy.29 Most identified nut allergens have been characterized as seed storage proteins, the major components of which include 2S albumins and 11S and 7S globulins. The 2S albumins are typically methionine-rich heterodimeric proteins comprised of subunits of about 30-40 (MW ~3 kD) and 60-90 (MW ~9 kD) residues, respectively. They are synthesized as precursor proteins, which are co-translationally transported into the lumen of the endoplasmic reticulum (ER). After the formation of four intra-chain disulphide bonds, involving eight conserved cysteine residues, the folded protein is transported to the storage vacuole where proteolytic processing can occur. This processing yields the small and large subunits which remain associated by two disulphide bonds with the loss of short linker, flanking, and signal peptide sequences.30 The 2S albumins from a variety of seeds and nuts have been shown to bind IgE from allergic patients’ sera.31-43 Allergenic proteins from castor bean (Ric c 1),44 sesame seed (Ses i 2),38 mustard seed (Sin a 1, Bra j 1),32, 33 hazelnut (Cor a 1),45 Brazil nut (Ber e 1),46 and almond,42 for example, have all been classified as 2S albumins. Like 2S albumins, 11S globulins are synthesized as single preproteins that get co- translationally cleaved into two subunits joined by an intermolecular disulfide bond.47 The constituent 30-35 kD acidic and 20-25 kD basic subunits are a result of processes that do, however, differ from that of the 2S albumin modifications. While a short signal peptide targets the preprotein to the ER and is removed as in the 2S albumin, the resultant proprotein then

assembles into trimers. The trimeric complex then gets sorted to protein storage vacuoles where a specific post-translational cleavage occurs, resulting in a trimer of mature heterodimers consisting of the acidic and basic polypeptides. Lastly, two trimers come together in a face-to- face fashion to form a 300-380 kD hexamer of about 11S. Legumins represent the major storage protein in cashew and walnut, accounting for approximately 50% of the total seed protein.35, 48 Previous studies have identified legumins from a variety of plant sources as food allergens. These include Ara h 313 and Ara h 449 of the peanut, the soybean G1 and G2 glycinins,50-51 Cor a 9 of the hazelnut,53 and possibly a legumin from buckwheat.54, 55 We set out to identify new 2S albumin and 11S globulin allergens in cashew and compare them both biochemically and clinically to their previously identified counterparts in the English walnut.40, 56

Methods

Walnut and Cashew Nut cDNA Expression Library Construction Standard recombinant DNA methodologies were used by Teuber et al to construct a cDNA library from walnut somatic embryos as previously described.40 The cashew cDNA library was constructed from four cashew nuts in late maturation by Dr. Fang Wang in our lab.57 Briefly, nuts were chopped, frozen in liquid nitrogen, and ground with a mortar and pestle. Total RNA was extracted in TRIzol (GIBCO BRL Life Technologies Inc., Rockville, MD) as previously described57 and mRNA was isolated using a PolyATtract kit (Promega Co., Madison, WI) according to the manufacturer’s instructions. Synthesis and cloning of cDNA was performed using the Uni-ZAP XR Gigapack Cloning Kit (Stratagene Inc., Cedar Creek, TX) following the manufacturer’s instructions. The double-stranded cDNAs with EcoRI (using a 5’ end adapter) and XhoI (using a 3’ end PCR primer) cohesive ends were cloned into the lambda Uni-ZAP XR expression vector. The library was amplified on E. coli strain XL1-Blue.

Expression Library Screening and Immunopositive Clone Identification After amplification the authors screened the libraries as previously described.40, 57 Briefly, the transfected XL1-Blue cells were incubated with E. coli (strain Y1090) for 15 minutes (min) at 37° C and then plated in a layer of Luria-Bertani (LB) soft agar for approximately 5

hours (hr) at 42° C at a density of approximately 20,000 plaques per plate. A nitrocellulose filter soaked in 10 mmol/L isopropyl thiogalactose (IPTG) was dried, placed on the agar plate, and further incubated 10 to 14 hr at 37° C. The filter was placed in blocking solution (4% nonfat milk in phosphate-buffered saline, pH 7.4) for 1 hr, and primary sera from a patient or patients that displayed IgE reactivity to multiple bands on whole walnut or whole cashew immunoblots was diluted 1:5 in 4% nonfat milk and allowed to incubate at 4° C overnight with the filter. 125I labeled anti-human IgE (Hycor Biomedical, Garden Grove, CA) was added after washing and allowed to incubate overnight. The filter was exposed for 96 hr to Kodak XRT film and developed. Positive plaques were purified to homogeneity. In cases where cashew patient sera was scarce, screens were performed with rabbit anti-cashew sera diluted 1:5,000 in blocking buffer. The rabbit IgG was considered a sufficient substitute because it had previously shown an IgG binding profile to cashew proteins similar to that of the patient sensitive IgE. Bound rabbit IgG was detected using HRP-conjugated goat anti-rabbit IgG antibody (Sigma Co., St. Louis, MO) at a 1:50,000 dilution and the ECL Plus chemiluminescent kit (ECL+, Amersham Pharmacia Biotech Inc., Piscataway, NJ) followed by exposure to Kodak X-OMAT x-ray film. Clones were subsequently screened with antiserum from a cashew-allergic patient (at 1:20 dilution) to demonstrate IgE reactivity. The immunopositive clones were picked, plaque- purified, and stored in SM buffer supplemented with 2% chloroform at 4°C.

Sequencing, Amplifying, and Analysis of Selected Genes Inserts from the selected phage clones were amplified by the authors with M13 forward and reverse primers via PCR; both strands of the PCR products were then sequenced on an ABI 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA) using capillary electrophoresis and Version 2 Big Dye Terminators as described by the manufacturer.40, 57 Similarity searches and alignments of deduced aa sequences were performed on Genetics Computer Group (GCG) software (Accelrys Inc., San Diego, CA) using the basic local alignment search tool (BLAST) 2.0 program. In the case of the cashew 2S albumin, which could not be identified by screening the cDNA expression library, the gene was amplified directly from the library by PCR with the aid of a degenerate forward primer 5’- CGTCAAGAGTCTYTTCGTGARTGYTGYCAR - 3’ (Y =C/T , R = A/G ), deduced from the previously published cashew 2S albumin N-terminal aa

sequence data,43 and a lock dock reverse primer 5’-TTTTTTTTTTTTTTTTTTNN –3’ (N= A, C, G, and T, Qiagen, Inc. Valencia, CA).58 Both strands of the PCR product were sequenced as described above. Like the cashew 2S albumin, the walnut legumin could not be isolated from the cDNA expression library. Instead, the authors56 used the synthesized cDNA and preformed both 3' and 5' RACE experiments with primers based on a partial coding sequence for walnut legumin-like protein from a walnut embryo cDNA library. PCR products were cloned into pCR4-TOPO according to the manufacturers instructions (Invitrogen Life Technologies, Carlsbad, CA) and the corresponding inserts were sequenced. Based on this deduced sequence, new 5' and 3' primers (Table 2.1) were synthesized and used in an RT-PCR reaction to clone the full-length legumin cDNA.

Subcloning of cDNA cDNA coding sequences were modified by the addition of restriction sites at the 5’ and 3’ ends by PCR using PfuTurbo DNA polymerase (Stratagene). The restriction sites were chosen after creating a RE map of each gene using the Sequencher Software package (Gene Codes Co., Ann Arbor, MI) and identifying non-cutting REs. The primers used for subcloning each gene are summarized in Table 2.1

Table 2.1. Primers used for subcloning of walnut and cashew 2S albumin and 11S globulin genes into the pMAL-c2-His expression vector.

Source Gene Primer (5’- 3’) Walnut 2S albumin Forward: GTCGACGGATCCGCAGCTCTCCTTGTAGCCC Reverse: AAGCTTCTGCAGCTAGAACCAGCTTCTGCG Walnut 11S globulin Forward: GTCTGGATCCATGGCCAAGCCCATCTTGC Reverse: TCTAGAGTCGACTTAAACTTCAGCCCTCC Cashew 2S albumin Forward: GTCTGGATCCATGGCAAAGTTCTTACTCC Reverse: TCTAGAGTCGACTTACTAATAAGATGACTGAACTG Cashew 11S globulin Forward: TCTAGAGTCGACCGCCAGGAATGGCAAC Reverse: GCTTGCCTGCAGTTAAGCATCATCCCTCATG

1Restriction sites underlined, GGATCC= BamHI, CTGCAG= PstI, GTCGAC = SalI.

After each gene was modified and amplified as described above, the PCR products were purified using the QIAquick PCR Purification Kit (Qiagen Inc., Valencia, CA) and subsequently digested using the appropriate REs (see Table 2.1) and reaction buffers (Invitrogen) according to the manufacturers instructions. Cut products were then purified using a QIAquick Gel Extraction Kit (Qiagen) and ligated with T4 DNA Ligase H.C. (Invitrogen) to their respective sites of the RE cut maltose binding protein (MBP) fusion expression vector pMAL-c2 (New England BioLabs Inc., Beverly, MA), according to the manufacturers instructions. The pMAL- c2 expression vector (pMAL-c2-His) used for all experiments was previously modified to contain a His-tag, thrombin cleavage site, and modified restriction sites.

Expression and Purification of MBP-Fusion Proteins For expression of the cDNA/pMAL-c2-His plasmids, competent E. coli DH5-α cells were transformed with complete ligation mixtures. Positive clones were selected via PCR using gene specific primers (Table 2.1) and used to innoculate 5 ml LB overnight cultures containing 100 µg/ml ampicillin. Plasmid preps were performed using 3 ml of the overnight culture and the QIAprep Spin Miniprep Kit (Qiagen) following the manufacturers instructions. A portion of the purified plasmids were cut with the appropriate REs to verify the presence of the inserted gene and insert positive plasmids were subsequently used to transform competent BL21 (DE3) cells (Novagen Inc, Madison, WI). Transformed cells were plated overnight and positive clones were identified via PCR as described above. Single colonies were grown at 37°C to an OD600 of 0.5- 0.8, at which point protein synthesis was induced with the addition of isopropyl-D- thiogalactopyranoside (IPTG) to a final concentration of 0.3 mM for 3-4 hr. The cells were harvested by centrifugation at 4000g for 20 min, resuspended in amylose resin column buffer (20 mM Tris-HCl, 200 mM NaCl, pH 7.4) supplemented with 10 mM β-mercaptoethanol (β−ME) and 1 mM EDTA and lysed using a microfluidizer. The whole lysate was centrifuged as described above and the fusion protein containing supernatant diluted 1:4 in column buffer, passed over an amylose affinity column, (New England Biolabs) and eluted with column buffer containing 10 mM maltose. Pure recombinant fusion proteins were dialyzed overnight vs. borate saline buffer (BSB; 100 mM boric acid, 25 mM sodium borate, 75 mM NaCl, pH 8.2), while changing the dialysis buffer every 4 hr, concentrated using Centricon Centrifugal Filter Devices

(YM-30, Millipore Corporations, Bedford, MA) and either stored (briefly) at 4°C until use or frozen at -80°C.

Human Sera Blood samples were drawn after informed consent from patients with life-threatening systemic reactions to walnut or cashew nut and the sera frozen at -70°C until use. The study was approved by the Institutional Review Board of the University of California at Davis. The presence of walnut and cashew-reactive IgE was confirmed by Pharmacia ImmunoCAP assays (Pharmacia Diagnostics, Uppsala, Sweden) or by Western immunoblotting (described below). Control sera were obtained from patients with a history of anaphylaxis to tree nuts but who reported tolerance of those assayed. Characteristics of the subjects are shown in Table 2.2.

Walnut and Cashew Nut Protein Extracts Soluble protein extract was prepared from defatted cashew or walnut nut flour as previously described.61 Protein concentrations were measured by the Bradford protein assay (Bio-Rad Laboratories Inc., Hercules, CA) using bovine serum albumin (BSA; Sigma, St. Louis, MO) as the standard protein.

Gel Electrophoresis (SDS-PAGE) Recombinant proteins (1 µg/4 mm well width) or whole extract samples (200-300 µg/4 mm well width) were boiled for 5 min in sample buffer (60 mM Tris-HCl, 2% SDS, 10% glycerol, 0.01% bromophenol blue, pH 6.8) with (reducing) or without (non-reducing) 100 mM dithiothreitol (DTT) and loaded onto Tris-HCl mini-gels (7X10 mm), 12% monomer acrylamide concentration resolving gel and 4% stacking gel) made according to Laemmli62 and electrophoresis carried out at 125 V using a Mini Protean 3 electrophoresis unit (Bio-Rad). To allow for optimal separation and enhanced resolution of the low molecular weight 2S albumin bands, total walnut and cashew extract samples were electrophoresed on large (16x18 mm) 18% Laemmli gels in a Hofer SE 600 Series standard dual cooled gel electrophoresis unit (Hoefer Scientific Instruments, San Francisco, CA). All gels were either stained with colloidal Coomassie stain (Sigma), or transferred to Protran Pure nitrocellulose transfer membrane (Schleicher & Schuell Inc., Keene, NH) for blotting assays as described below.

Table 2.2. Clinical characteristics of cashew and walnut-allergic subjects. ______No. Sex Age Age of onset Other food allergy Cashew ImmunoCAP, of cashew RAST (kU/l allergy or Class), or positive IgE immunoblot (+blot)

1 M 25 3 pecan, hazel 5.66 3 F 26 2 PN, pist 9.51 4 M 27 5 brazil, coconut hazel 6.95 5 F 54 10 pecan, hazel 1.62 7 F 30 10 PN, hazel 4.04 8 F 43 1 PN, TNs 41.9 9 F 35 2 pecans, almond 35.1 10 F 31 2 sunflower 4.42 11 M 50 1 TNs 5.19 12 F 26 3 PN, TNs 2.41 13 F 39 1 PN, TNs 9.53 14 F 39 5 pist, TNs 94.7 15 M 50 2 TNs 1.87 16 F 44 1 PN +blot 17 F 39 6 pine nut +blot 18 F 62 1 PN, almond +blot 20 F 48 1 PN, hazelnut, chestnut +blot 22 M 27 7 pecans, brazil nut +blot 23 F 27 child macadamia 0.43 27 M 43 2 PN, pecans Class 0 29 F 49 3 PN, TNs, sesame 0.52 30 F 53 15 TNs except almond +blot 31 F 44 1 almond, sesame +blot 32 M 38 1 pecan, hazel, pist +blot 33 F 42 1 PN <0.35 34 M 33 1 PN, almond, hazel 1.42 35 F 54 2 pecan, hazel Class 3 36 M 43 2 none Class 0 37 M 57 1 PN, almond not done 38 F 50 2 PN, pecan, pine nut 12.6. 39 F 52 1 brazil nut 3.56 40 F 2 2 none 5.97 ______

1Patient numbers 1-40 correspond to those in previous publications.57-59 2Anaphylaxis grades (class) assigned by Teuber and Beyer based on the scoring system suggested by Sampson.60 Pist = pistachio, PN= peanut, TNs = all other tree nuts.

Gel Electrophoresis (SDS-PAGE) Recombinant proteins (1 µg/4 mm well width) or whole extract samples (200-300 µg/4 mm well width) were boiled for 5 min in sample buffer (60 mM Tris-HCl, 2% SDS, 10%

glycerol, 0.01% bromophenol blue, pH 6.8) with (reducing) or without (non-reducing) 100 mM dithiothreitol (DTT) and loaded onto Tris-HCl mini-gels (7X10 mm), 12% monomer acrylamide concentration resolving gel and 4% stacking gel) made according to Laemmli62 and electrophoresis carried out at 125 V using a Mini Protean 3 electrophoresis unit (Bio-Rad). To allow for optimal separation and enhanced resolution of the low molecular weight 2S albumin bands, total walnut and cashew extract samples were electrophoresed on large (16x18 mm) 18% Laemmli gels in a Hofer SE 600 Series standard dual cooled gel electrophoresis unit (Hoefer Scientific Instruments, San Francisco, CA). All gels were either stained with colloidal Coomassie stain (Sigma), or transferred to Protran Pure nitrocellulose transfer membrane (Schleicher & Schuell Inc., Keene, NH) for blotting assays as described below.

Direct and Inhibition Immunoblotting

Electrophoresed proteins were transferred to 0.22 µm (2S albumins) or 0.45 µm (11S globulins) nitrocellulose (Schleicher & Schuell) overnight at 30 mA using a Mini Protean III transfer cell (Bio-Rad). Nitrocellulose membranes containing blotted recombinant fusion proteins were cut into 3-4 mm wide strips and blocked for 1 hr at room temperature in 20 mM phosphate buffered saline (PBS; pH 7.3) containing 3% nonfat dry milk and 0.2% Triton X-100 (TX-100). Diluted sera (1:5 in the blocking buffer or 1:20 for highly reactive sera) were added to the strips and incubated overnight at 4°C. Additional strips were incubated with non-cashew or non-walnut-allergic patient sera in blocking buffer to check for non-specific IgE or secondary antibody binding. The strips were then washed for 20 min three times in PBS containing 0.01% TX-100 and incubated overnight at 4°C with equine polyclonal 125I-anti-human IgE (Hycor Biomedical) diluted 1:10 in blocking buffer. Washing was repeated as above and the strips exposed to x-ray film (Kodak X-OMAT). To identify the molecular weight of the native allergenic proteins in whole nut extracts, inhibition experiments were carried out in which 10-50 µg of recombinant fusion proteins or MBP alone (negative control) were pre-incubated with allergen specific human antiserum for 3 hr at room temperature prior to probing whole walnut or cashew extract blotted strips, as described above.

Results

Identification of Genes Encoding for the English Walnut and Cashew Nut 2S Albumin and 11S Globulin Seed Storage Proteins Walnut-allergic patient sera showed IgE reactivity to several of the walnut cDNA expression library clones. The first such reactive clone was identified by Teuber et al40 and designated Jug r 1 by following the guidelines of the International Union of Immunological Societies’ Allergen Nomenclature Subcommittee. Analysis of the nearly 660 base pair (bp) gene (GenBank ID U66866) revealed an open reading frame encoding a 139 aa 2S albumin seed storage protein precursor possessing important homologies with other 2S albumin proteins from Brazil nut, cottonseed, castor bean, and mustard.40 Unable to identify a clone encoding the full-length 11S globulin gene using the above method, the authors used fresh walnut mRNA in 3’ and 5’ RACE experiments. Products from the resulting RT-PCR reaction were sequenced and a 1539 bp full-length clone, named Jug r 4, was obtained. The gene translated a precursor protein with an open reading frame 513 aa in length and showing good homology to another known allergenic tree nut 11S globulins, Cor a 9 (hazelnut). Cashew-allergic patient sera showed IgE reactivity to 50 of the cashew cDNA expression library clones. Sequencing of one of these 1670 bp genes (GenBank ID AF453947) and comparison to GenBank proved it to be homologous with genes encoding the legumin (11S globulin) family of seed storage proteins and the deduced 457 aa protein was designated, Ana o 2. The protein was found to have high sequence homology to several other legumins as summarized in Table 2.3. None of the clones identified by directly screening the cashew cDNA library contained a gene encoding for the cashew 2S albumin. Therefore, degenerate primers were made as described above and used to PCR amplify the clone using the cDNA library as template. The resulting 585 bp PCR product (GenBank ID AY081853) encodes a 138 aa protein designated Ana o 3. When compared to the GenBank database, the sequence proved to be homologous to other members of the 2S albumin family of seed storage proteins as summarized in Table 2.4.

Table 2.3. Proteins showing identity and similarity to Ana o 2.

Protein Description Organism Accession # AA Overlapa Identity Similarity Legumin-like protein Ricinus communis (castor bean) AAF73007 1-446 58% 74% 11S globulin-like Corylus avellana AF449424 2-444 55% 71% protein (hazelnut) Legumin precursor Quercus robur (English oak) CAA67879 2-445 55% 72% 11S globulin Sesamum indicum (sesame) AAK15087 20-451 51% 69% Legumin precursor Magnolia salicifolia S54206 1-454 50% 68% (grain amaranth) S54208 1-441 50% 68% 11S globulin Amaranthus hypochondriacus S49422 1-454 48% 68% (grain amaranth) Glycinin G2 subunit Glycine max (soybean) AAB23210 1-450 47% 66% Glycinin G1 subunit Glycine max (soybean) AAB23209 2-439 47% 65% 11S storage globulin Coffea arabica (coffee) AAC61881 1-457 46% 64% Legumin A precursor Vicia Sativa (spring vetch) S44294 1-439 45% 65% 11S globulin β-subunit Cucurbita pepo AAA33110 3-455 45% 64% precursor (winter squash) Glucinin Ara h 3 Arachis hypogaea (peanut) AF093541 21-439 43% 59% Glucinin Ara h 4 Arachis hypogaea (peanut) AF086821 4-439 42% 58%

Table 2.4. Proteins showing identity and similarity to Ana o 3.

Protein Description Organism Accession # AA Overlap Identity Similarity 2S albumin precursor Ricinis communis (Castor bean) P01089 36-134 42 65 2S albumin seed Juglans nigra (Black walnut) AAM54365 7-133 41 64 storage protein 2S albumin seed Juglans regia (English walnut) AAB41308 7-133 42 61 storage protein precursor Putative allergen Carya illinoinensis (Pecan) AAO32314 7-133 42 60 Allergen Ses i 2 Sesamum indicum (Sesame seed) AAK15088 1-151 37 58 2S albumin precursor Cucurbita sp. (Squash) T10257 11-134 41 57 2S albumin precursor Vitis cinifera L. (Grape) AAP31047 1-104 41 56 Allergen Ber e 1 Bertholletia excelsa (Brazil nut) P04403 1-128 30 55 Allergen Bra j 1 large Brassica juncea (Oriental P80207 68-137 34 53 chain mustard) Allergen Ses i 1 Sesamum indicum (Sesame seed) AAD42943 1-146 30 53 Allergen Sin a 1 large Sinapsis alba (Yellow mustard) S65447 35-137 28 51 chain

Fusion Protein Expression and Purification Fusion proteins were expressed and purified with yields ranging from 20-30 mg/L of cultured cells. After dialysis in BSB the purified proteins reached nearly 100% homogeneity. An example of the purification scheme and purified cashew 11S globulin protein is represented in Figure 2.1

1234

250 kD 148 kD

60 kD

42 kD

30 kD

Figure 2.1. Purification of the expressed recombinant Ana o 2-MBP fusion protein. Lane 1, marker; Lane 2, cell lysate; Lane 3, cell lysate centrifugation supernatant; Lane 4, amylose column eluate after dialysis.

Characterization of Major Allergens Belonging to the 2S Albumin and 11S Globulin Families via Direct and Inhibition Immunoblotting Teuber et al40 demonstrated that immunoblotting of their Jug r 1 fusion protein with patient sera revealed 75% (12/16) had IgE reactivity. The authors also carried out an inhibition immunoblot to show that they had cloned a gene encoding an allergen present in the mature walnut kernel and help deduce the approximate molecular weight of this presumed 2S albumin. Preabsorption of a patient’s serum with rJug r 1 abolished IgE binding on immunoblotting (reducing conditions) to proteins from mature walnut kernel at approximately 14, 10-12, and 5

kD (Figure 2.2). Such findings along with observed sequence similarities to Brazil nut and other 2S albumin seed storage proteins led the authors to infer that the native Jug r 1 allergen consists of a large subunit of about 10 kD and a small subunit of about 4 to 5 kD. They also made note of a14 kD band that was inhibited on the blot and describe it as the in tact heterodimeric protein which was presumably incompletely reduced because of partial inaccessibility of the disulfide bonds or re-association after reduction.

123

66 kD

45 kD 39 kD

31 kD

21.5 kD

14.5 kD

7.5 kD 6.5 kD

Figure 2.2. Western immunoblotting and rJug r 1 inhibition blotting of walnut extract in reducing conditions with patient #3 sera. Lane 1, sera diluted 1:5; Lane 2, sera diluted 1:5 and pre-incubated with rJug r 1 fusion protein; and Lane 3, sera diluted 1:5 and pre-incubated with recombinant pyruvate dehydrogenase (control).

In 2003 Teuber et al56 proved through Western immunoblotting that their expressed recombinant 11S globulin (Jug r 4) was the fourth important allergen to be identified in the English walnut. The assay showed that 73% of walnut sensitive patient sera tested (16/22) had IgE reactivity to the recombinant fusion protein. Preabsorption of a patient’s serum (#9) with rJug r 4 nearly abolished IgE binding on immunoblotting (reducing conditions) to proteins from mature walnut kernel at approximately 35 and 22 kD (Figure 2.3). Inhibition of these mature

processed peptides is consistent with the fact that legumin group proteins are processed into a ~33 kD acidic subunit and ~22 kD basic subunit connected by disulphide bonds.

66 kD

45 kD

31 kD

21 kD

14 kD

7 kD

Figure 2.3. Western immunoblotting and rJug r 4 inhibition blotting of walnut extract in reducing conditions with patient #9 sera. Lane 1, sera diluted 1:5 in blocking buffer; Lane 2, sera diluted 1:5 and pre-incubated with rJug r 4 fusion protein.

The 2S albumin and 11S globulin proteins proved to be important allergens in the cashew nut as well as the walnut. Western immunoblotting of the expressed recombinant 2S albumin from the cashew (Ana o 3) showed that IgE from 21 of 26 sera (81%) from patients with self- reported reactions to cashews bound to the recombinant protein. Three apparent isoforms of the large subunit of cashew 2S albumin have been reported33 and sera from one walnut-allergic patient (#31, Table 2.2) showed IgE reactivity to each (Figure 2.4, Lane 1). Inhibition assays with this patient’s serum pre-incubated with rAna o 3 were performed to confirm the molecular weight of these large subunit isoforms and that of the Ana o 3 small subunit in whole cashew extract. The results (Figure 2.4, Lane 2) reveal the loss of IgE binding to peptides in the 8 and 10 kDa range in a reducing gel.

123

42 kD

30 kD

22 kD

17 kD

6 kD

4 kD

Figure 2.4. Western immunoblotting and rAna o 3 inhibition blotting of cashew extract in reducing conditions with patient #31 sera. Lane 1, sera diluted 1:5 in blocking buffer; Lane 2, sera diluted 1:5 and pre-incubated with rAna o 3 fusion protein; Lane 3, sera diluted 1:5 and pre- incubated with and MBP (control).

Western immunoblotting of the expressed recombinant Ana o 2 revealed that IgE from 13 of 21 (62%) sera from patients with a history of life-threatening reactions to cashews showed reactivity. Having proved that Ana o 2 is an important cashew allergen, we sought to identify the band or bands in a typical total cashew immunoblot that correspond(s) to the cloned polypeptide storage protein precursor. Figure 2.5 shows the results of our inhibition immunoblot with sera from patient #9. The first strip (Lane 1) shows this serum reacting with three major bands at ~53, ~36 and ~33 kD, presumably corresponding to the unsuccessfully reduced portion of the in tact precursor protein, the large subunit, and the small subunit respectively. The second through fourth strips (Lanes 2-4) were incubated with the same human sera pre-incubated with rAna o 2, rAna o 1 + rAna o 2, and MBP (negative control), respectively. Note that Ana o 2 inhibits the strong band at ~33 kD and weakens the band at ~53 kD, suggesting that there were two polypeptides at this higher molecular weight – both the vicilin and a small quantity of legumin-group protein (most likely the intact precursor protein). The addition of Ana o 1

inhibits the residual signal at ~53 kD. MBP has no inhibitory effect as expected. The remaining band at ~36 kD is as yet unidentified.

1234

213 kD 128 kD

85 kD

43 kD

31 kD

18 kD

Figure 2.5. Western immunoblotting and rAna o 2 inhibition blotting of cashew extract in reducing conditions with patient #9 sera. Lane 1, sera diluted 1:5 in blocking buffer; Lane 2 sera diluted 1:5 and pre-incubated with rAna o 2 fusion protein; Lane 3, sera diluted 1:5 and pre- incubated with both rAna o 1 and rAna o 2 fusion proteins; Lane 4, sera diluted 1:5 and pre- incubated with MBP (control).

Discussion

This study, coupled with the work of Teuber et al,40, 56 has identified those proteins belonging to the 2S albumin and 11S globulin families of seed storage proteins responsible for cashew and walnut-allergic patient sera IgE binding. It is now believed that approximately 1% of the US population is allergic to tree nuts and up to 57% of those reporting allergy to tree nuts list sensitivity to walnut and 41% to cashews.10 Together, we have cloned and characterized a total of four important allergenic proteins in these nuts. For many years, 2S albumin seed storage proteins have been associated with IgE responses in the literature.40 It is not surprising that allergens from walnut and cashew nut are also 2S albumin proteins. We have described the cloning and sequencing of a 2S albumin seed storage protein precursor, Jug r 1, from a walnut cDNA library and a product of a PCR-amplified cashew cDNA sequence also encoding a 2S albumin-family seed storage protein, Ana o 3.

These genes show sequence homology to other 2S albumins, in particular those found to be allergenic in other plant sources such as Brazil nut (Ber e 1),46 Castor bean (Ric c 1),44 mustard seed (Bra j 1, Sin a 1),32, 35 sesame seed (Ses e 1)63 and almond.42 Both proteins demonstrate characteristics expected of those belonging to the 2S albumin family. They are apparently synthesized as precursor proteins before undergoing post-translational modifications that processes them into heterodimers consisting of large subunits ranging from 6-10 kD and small subunits of ~ 4 kD. It is becoming apparent that the 2S albumin storage proteins can be viewed as pan-allergens among seeds in which they occur and that they are associated with IgE responses through airborne and oral exposure.40 The ability of both recombinant proteins to inhibit IgE binding to the peptides in their respect ive whole nut extracts by patient sera in our inhibition assays confirmed the identity of the native subunits. Further, these findings demonstrate the relevance of using the recombinant precursor proteins in our patient screening assays despite the fact that our constructs lacked the post-translational modifications that alter the final structure and size of the native protein. In such assays it was found that rJug r 1 was bound by IgE from 75% of the walnut-allergic patient sera tested40 while rAna o 3 showed reactivity with IgE from 83% of the cashew-allergic patient sera.58 Binding of IgE from more than 50% of nut allergic patient sera has, in the past, been evidence enough to classify these proteins as major allergens. There has, however, been some movement away from using immunoblot data as a standard for allergen importance because instances have arisen where these data do not correlate to the intensity of the reaction in cellular assays and clinical symptons.64 We have also shown that the walnut and cashew nut 11S globulins can be considered major a llergens when the above standard is applied. We have described the cloning and sequencing of these seed storage protein precursors; Jug r 4, from walnut cDNA via degenerate primer PCR amplification and Ana o 2 via IgE screening of a cashew cDNA expression library. These genes show sequence homology to other allergenic 11S globulins including Ara h 313 and Ara h 449 of the peanut, the soybean G1 and G2 glycinins,50-52 Cor a 9 of the hazelnut,53 and possibly a legumin from buckwheat.54, 55 Both Jug r 4 and Ana o 2 proteins demonstrate characteristics of the 11S globulin family. Like the 2S albumins, they are apparently synthesized as precursor proteins before undergoing post-translational modifications that processes them into heterodimers consisting of large and small subunits. The intact precursor protein is generally 50-

60 kD, consisting of a large (acidic) subunit ranging from 30-40 kD and a small (basic) subunit of ~20 kD that are generally covalently associated in seeds.65 Unlike the 2S albumins, the 11S globulins often polymerize in their native state, eventually forming hexamers ranging in molecular weight from 300-450 kD. 11S globulins have proven to be reactive with IgE from a majority of tested nut-allergic patients and have been the most thoroughly studied tree nut allergens to date.11 As we observed with the 2S albumins in our inhibition assays, the recombinant walnut and cashew legumins displayed the ability to inhibit reactive-patient IgE binding to peptides corresponding to the protein’s large and small subunits in whole walnut and cashew nut extracts, respectively. These findings helped confirm the identity and molecular weight of the acidic and basic chains and further demonstrate the relevance of using the recombinant precursor proteins in our patient screening assays. These assays, which, as previously stated, allowed us to classify Jug r 4 and Ana o 2 as major allergens, found that rJug r 4 was bound by IgE from 73% of the walnut- allergic patient sera tested56 while rAna o 2 showed reactivity with IgE from 62% of the cashew- allergic patient sera.59 By characterizing and recombinantly expressing these new major allergens in cashew nut and walnut we have contributed to a new approach for identifying and clinically diagnosing reactivity to tree nuts; determining antibody reactivity to recombinant allergenic proteins. Most commercial kits available for food allergen detection or, from a clinical standpoint, the initial diagnosis of patient allergen sensitivity, are comprised of crude extracts that very often differ in protein concentrations from batch to batch. Similarly, the use of crude extracts tell the investigator little with regards to the identify of the proteins most threatening to the patient, thereby making it difficult to design a rational plan for therapy. Recombinant allergen-based diagnostic tests enable the dissection and monitoring of the IgE-reactivity profiles of allergic patients, resulting in more specific detection, diagnosis, disease monitoring, and therapy.66-68 With the reported incidence of tree nut allergies clearly on the rise,10 our efforts to identify new cashew and walnut allergenic proteins and, further, to express them as recombinant proteins, represent a step forward in the diagnosis and possible treatment of this allergic disease. Beyond specific protein reactivity, there is now a belief that the most useful information will come from screening for reactivity to just those IgE-binding peptides (epitopes) within the protein.67

Given the broad allergenicity of these protein classes, IgE-epitope mapping and T-cell epitope research may provide further insight into common motifs involved in IgE recognition and prove to be potential targets for treatment.

CHAPTER 3

LINEAR IGE-EPITOPE MAPPING OF ENGLISH WALNUT AND CASHEW NUT 2S ALBUMIN AND 11S GLOBULIN ALLERGENS

Introduction

Despite the rapidly increasing number of described recombinant allergens, relatively few have had their IgE-reactive B-cell epitopes defined.69 Given the critical role allergen-specific IgE plays in the allergic reaction, determination of allergen-specific IgE binding epitopes, be they linear or conformational, appears to be of great importance for gaining a better understanding of the allergenic nature of foods and for possible therapeutic intervention. While conformational IgE-binding epitopes are prevalent and important to the etiology of aeroallergen- mediated allergic reactions, linear epitopes are important for food allergens mainly because the immune system will eventually encounter such allergens after they, in theory, have been partially denatured and digested by the human GI tract. Therefore, the linear IgE-binding epitopes of food allergens have attracted more attention than the less prevalent conformational epitopes.70 Linear epitopes may be defined as short contiguous stretches of aa that are reactive with antibody. We are hopeful that as the database for linear epitopes of allergens grows, patterns of IgE reactivity will emerge. With this information, studies involving the analysis of new allergens and the generation of engineered allergens for plant breeding and therapeutic agents can be better focused. Similarly, diagnostic assays involving the screening of patient sera with IgE specific epitopes rather than full-length proteins in crude extracts may prove to be a more sensitive means of identifying patient allergenicity.67 Herein I report the linear IgE epitope mapping of the important walnut (Jug r 1, Jug r 4) and cashew nut (Ana o 3, Ana o 2) 2S albumin and 11S globulin allergens through screening of solid-phase overlapping peptide libraries. This is the most detailed study of IgE-binding linear epitopes of tree nut allergens to date. By combining our findings with those related epitope mapping studies previously undertaken, we have been able to create composite epitope maps and propose a basis for cross-reactivity among patient IgE-binding to 2S albumin and 11S globulin proteins. We also identify aa found to be critical and influential for IgE binding to

immunodominant epitopes of Jug r 1 and Ana o 2 and demonstrate the feasibility of reducing or even eliminating this reactivity through epitope specific IgE removal and/or protein mutagenesis experiments.

Methods

Solid-phase Peptide (SPOTs) Synthesis Peptides were synthesized on derivatized cellulose sheets using 9-fluorenlymethoxy carbonyl-derived (Fmoc) aa as described by the manufacturer (Genosys Biotechnologies, Inc., The Woodlands, TX). Briefly, cellulose membranes containing free hydroxy groups were esterfied with an Fmoc-aa dissolved in 1-methyl-2 pyrrolidione. The coupling reaction was followed by washing in N,N-dimethylformamide (DMF) and the aa were acetylated with acetic anhydride. The membranes were deprotected by washing in DMF and incubation with a solution of 20% piperidine in DMF followed by another wash in DMF. Coupling, acetylation, and deprotection steps were repeated for each cycle. During the final cycle, a mixture of dichloromethane, trifluoroacetic acid, and triisobutylsilane (1:1:0.05) was used to deprotect the acid-labile aa side chains. Based on the published aa sequence of Jug r 140 and our unpublished data determining the subunit cleavage sites, twenty-five 13-aa peptides, offset by three aa, were synthesized which corresponded to the entire 64-aa length of the large subunit (Figure 3.1.1) and 26-aa length of the small subunit (Figure 3.1.2). Additional peptides bearing the target epitope (QGLRGEEMEEMV), mutated forms of this peptide, and a non-IgE binding peptide (LSQRGLQSSSV) were also synthesized. Thirteen versions of the target peptide were created via a single site alanine substitution at each position along the aa sequence, and six mutated peptides were synthesized using multiple alanine substitutions (Table 3.1). Based on the cDNA-derived aa sequence of the 138 aa Ana o 3 protein (excluding the presumptive 20 aa leader sequence which was predicted using the SignalP V.1 World Wide Web Prediction Server, www.cbs.dtu.dk/services/SignalP/), 36 overlapping 12-aa peptides, each offset by three aa, were synthesized. Discard peptides resulting from post-translational cleavage events other then the leader peptide were not omitted as they were with Jug r 1 nor have any alanine scanning experiments been conducted on the proteins epitopes to this date.

Due to the relatively large size of the 11S globulins, as compared to the 2S albumins, larger peptides needed to be synthesized and were offset by more aa. SPOTs membranes displaying either the 508 aa sequence of Jug r 4 or the 457 aa sequence of Ana o 2 in the form of 62 or 57, 15-aa peptides, each offset by 8 residues, were synthesized. We were able to identify target epitopes on each preprotein (leader and discard peptides included) and chose three immunodominant epitopes on Ana o 2 for further analysis via alanine scanning mutagenesis. Mutagenesis was performed as it was with the Jug r 1 epitope, whereby membranes containing the Ana o 2 target epitopes, alanine mutants, and a non-IgE binding peptides were also synthesized (Figure 3.5).

Human Sera Blood samples were drawn after informed consent from patients with life-threatening systemic reactions to walnut or cashew nut and characterized as described above (Table 2.2). Limitations on the amount of available patient sera with known IgE reactivity to the specific 2S albumin and/or 11S globulin proteins in immunoblotting experiments precluded analysis of individual patients’ sera with some of the epitope arrays described above. However, to gain some sense of patient reaction variability, each reactive serum sample was assigned to one of several pools. In all cases the pools were limited to a maximum of six patients so not to over dilute each sera and always included at least one serum that strongly reacted to the recombinant protein being represented in the epitope array.

IgE binding to SPOTs The peptide-containing membranes were washed in TBS and incubated o/n at room temperature in a proprietary blocking solution as directed by the manufacturer (Genosys Biotechnologies Inc., The Woodlands, TX). Membranes were then washed in TBS-T for 10 min and incubated o/n at 4°C with individual patient’s serum or pooled patients’ sera diluted 1:5 in blocking buffer (total sera:blocking buffer). This incubation was followed by three 5-min washes in TBS-T and an o/n (4°C) incubation with 125I labeled anti-human IgE (Hycor Biomedical) diluted 1:10 in a mixture of PBS, 5% nonfat dry milk, and 0.05% Tween-20. Three final 10 min washes in PBS containing 0.05% Tween-20 (PBS-T) were performed and IgE-

peptide reactivity identified after a 48 hr to 1 week exposure at -70°C to Kodak Biomax x-ray film.

Soluble Jug r 1-epitope Peptide Synthesis A soluble form of the identified Jug r 1-reactive peptide (QGLRGEEMEEMV) was synthesized by Fmoc protocols on an automated peptide synthesizer (Model 433A, Applied Biosystems).

Isolation and Inhibition of Epitope-specific IgE The epitope-reactive IgE was isolated from patients’ sera by affinity chromatography. A 500 µl pool of patients’ sera known to be reactive to the synthetic solid-phase epitope peptide, was slowly added to a 5 ml disposable polypropylene column (Pierce Chemical Co., Rockford, IL) containing cyanogen-bromide-activated beads (Sigma) (to which 5 mg of the soluble reactive peptide had been covalently coupled, as described by the manufacturer) and the effluent collected. Bound IgE was eluted with the addition of 200 mM glycine sulfate, pH 2.3, collected in a beaker containing 1 ml of 1% BSA in BSB, and subsequently neutralized with 1.0 M Tris. The column was then rinsed with BSB and the initial effluent re-passed over the column. The process was repeated for a total of three times, yielding three eluates and the column effluent. To test the specificity and reactivity of the epitope-specific IgE, the control, pooled and fractionated patients’ sera were pre-incubated overnight at 4°C with different amounts (70, 7.0, and 0.7 mg) of the soluble epitope peptide, or a soluble non-specific peptide. The pre-incubated sera were used to probe either solid phase IgE-reactive and non-IgE-reactive peptides or rJug r 1. Detection of IgE binding to the solid-phase synthetic peptides and dot-blotted rJug r 1 was detected using 125I labeled anti-human IgE (described above) and ECL+, respectively, as described below.

Dot Blotting of rJug r 1 and Detection of IgE Reactivity For dot-blot assays, a graphite pencil was used to circumscribe 4 mm X 8 mm elipses on a dry 0.4 µm nitrocellulose membrane (Shleicher & Schuell). The nitrocellulose was incubated in distilled-deionized water (ddH20) on a rocking table for 5 min and allowed to air dry before each dot was loaded with 0.11 µg of rJug r 1 in 2 µl of ddH20. The antigen-loaded nitrocellulose

was then rinsed in ddH20 as described above, placed protein side up on 3MM paper (Whatman Intl. Ltd., Maidstone, England), and dried under a 60-watt light until all moisture was removed. Once dry, dots were rinsed for 2 min in Tris-buffered saline (TBS: 20 mM Tris, 137 mM NaCl, pH 7.6) containing 0.2 % Tween-20, blocked for 1 hr at room temperature in TBS-T containing 2% BSA and then incubated o/n at 4 °C with sera diluted 1:40 in TBS-T. Membranes were then washed once for 15 min, and 3X for 5 min in TBS-T prior to being incubated for 1 hr at room temperature with horseradish peroxidase-labeled goat-anti-human IgE Biosource Intl., Camarillo, CA) diluted 1:2000 in TBS-T. Washing in TBS-T was repeated as above and the reactive dots were identified after a 5-min incubation in ECL+ (Amersham Pharmacia), as previously described, and subsequent exposure to Kodak X-OMAT x-ray film.

Results

Linear IgE-epitope Mapping of Jug r 1 (SPOTs analysis) The entire lengths of both the large and small subunits of Jug r 1, shown schematically in Figure 3.1, were studied by probing the overlapping solid phase synthetic peptides with sera from 20 patients randomly assigned to four pools (as described above). Each pool recognized three adjacent peptides from the large subunit (Figure 3.2.1); two peptides were recognized very strongly (#’s 11 and 12) and one less so (#10); no peptides were identified from the small subunit. A common sequence, GLRGEEM, was observed in all three large subunit peptides (Figure 3.2.2). A fourth partially overlapping sequence (#9) showed a slight positive reaction. Some other peptides showed slight positive signals in some assays (#7) but were not reproducible, leading us to examine only the identified dominant linear epitope-bearing peptides. SPOTs containing the most immunoreactive 12-aa peptide (#12, QGLRGEEMEEMV), designated E1, and SPOTs containing a negative control (-) peptide (LSQRSQQQCRQ), selected from the large subunit of Jug r 1, were used to test the degree of individual reactivity and specificity of allergic and control sera. Of the 20 patients tested, five exhibited strong recognition of E1, six moderate, four weak, and only five (25%) showed no recognition; none reacted with the (-) peptide (data not shown).

(1)

(2)

Figure 3.1. Schematic representation of solid phase overlapping peptide synthesis of Jug r 1. 3.1.1, Jug r 1 large subunit, aa 1-64 shown (leader peptide excluded) and a GGG linker are shown. 3.1.2, Jug r 1 small subunit, aa 1-26 and a GG linker are shown.

(1)

(2)

Figure 3.2. IgE binding analysis of Jug r 1 SPOTs. 3.2.1, SPOTs array depicting all 25 - overlapping 13-aa Jug r 1 peptides probed with pooled sera from six walnut-sensitive patients. 3.2.2, Sequence comparison of three adjacent overlapping peptides, 10-12, recognized by patient sera. The aa sequence common to each has been depicted with shading.

Additional peptides were tested in which alanine was substituted at each of the 12 amino residues of the IgE-reactive peptide, E1. In addition, peptides were tested with varying numbers of alanines substituted at the N- and C-termini. Together, these data demonstrated that the core aa RGEE, at positions 36 to 39, and an additional glutamic acid residue at position 42, were necessary for maximum IgE binding to occur (Table 3.1).

Table 3.1. Mutational analysis of the Jug r 1 IgE-binding epitope, E1.

Amino Acid Sequence1 Substitution IgE binding2 AGLRGEEMEEMV Q1A ++ QALRGEEMEEMV G2A ++ QGARGEEMEEMV L3A + QGLAGEEMEEMV R4A − QGLRAEEMEEMV G5A − QGLRGAEMEEMV E6A − QGLRGEAMEEMV E7A − QGLRGEEAEEMV M8A ++ QGLRGEEMAEMV E9A ++ QGLRGEEMEAMV E10A + QGLRGEEMEEAV M11A ++ QGLRGEEMEEMA V12A ++ QAARGEEMEEMV G2A,L3A + QAAAGEEMEEMV G2A,L3A,R4A − QAAAAEEMEEMV G2A,L3A,R4A,G5A − QGLRGEEMEAAV E10A,M11A + QGLRGEEMAAAV E9A,E10A,M11A − QGLRGEEAAAAV M8A,E9A,E10A,M11A −

1The critical core aa residues, located at positions 36-39 of the large subunit, and an influential glutamic acid residue at position 42, are indicated in bold type. Mutated residues are underlined. 2Peptides were probed with one patient pool comprised of sera from six walnut-sensitive patients.

Isolation and Inhibition of the Jug r 1 Linear Epitope (E1)-specific IgE To determine if the identified IgE-binding peptide (E1) is the major epitope recognized by patients’ sera, we separated E1-specific IgE from the total walnut sensitive patient antiserum and tested both fractions for reactivity with rJugr 1. Epitope-specific IgE was isolated from

patient serum by passage over an E1-affinity column as described above. To assure removal of anti-E1 Abs, the three sequential eluate and the serum effluent fractions were first assayed against our solid-phase IgE binding (+) and non-IgE binding (-) peptides. IgE binding to the positive (E1) peptide in unfractionated patient serum, progressively less binding in the first two eluate fractions, and no binding in the third was observed, indicating complete removal of the E1-reactive Ab by the peptide affinity column (Figure 3.3.1). Significantly, the E1-adsorbed serum (effluent) also showed no reactivity to the E1 peptide, confirming complete removal of the E1-specific IgE from the serum (Figure 3.3.1). The eluate and effluent fractions were subsequently used in dot-blot assays to determine if IgE Abs to the E1 epitope represented the bulk of the rJug r 1-reactive IgE. As expected, binding of IgE to rJug r 1, from total sera and the first two eluate (anti-E1) fractions, but not the third (anti-E1-depleted), was demonstrated (Figure 3.3.2). However, there was considerable rJug r 1-specific IgE remaining in the effluent (anti-E1 depleted fraction), demonstrating the presence of additional IgE Ab specific for one or more presumably conformational epitopes. To further examine the specificity of both our peptide- reactive fractions and unfractionated Jug r 1-reactive IgE, a soluble form of the epitope peptide was pre-incubated with whole patient serum as well as the E1-specific IgE Abs in an attempt to inhibit their reaction with both our solid-phase peptides and rJugr 1. Varying amounts (70, 7.0, 0.7 mg) of the peptide were incubated with whole patients’ sera prior to probing the E1 epitope (positive) and negative solid-phase peptides. Partial to complete inhibition was achieved (Figure 3.3.3), demonstrating that the solid and fluid phase versions of the peptide are similarly recognized. The amount of soluble peptide (70 mg) needed to completely inhibit binding of E1- specific IgE to the solid phase version of this epitope was similarly used for inhibition studies involving unfractionated patient serum IgE and affinity column purified IgE fractions against the rJug r 1. Inhibition was again observed in the eluate (E1 epitope-specific) IgE fractions. Together, these data demonstrate that the epitope is similarly recognized in both the peptide and the recombinant walnut protein. However, IgE reactivity in both whole patient serum and the E1-adsorbed effluent fraction reacted with rJug r 1 (Figure 3.3.4), again potentially demonstrating the presence of additional (presumably conformational) IgE-reactive epitopes on the recombinant protein.

(1) (2)

(3) (4)

Figure 3.3. Isolation and inhibition of E1-specific IgE. 3.3.1, Strips containing the E1 epitope peptide (top SPOT, +) and a negative control peptide (bottom SPOT, – ) probed with whole patient serum (#3), each of three E1 eluates (1st, 2nd, 3rd) and the effluent (EF). 3.3.2, Strips dotted with rJug r 1 and probed as in A. C3: sera from an atopic subject not allergic to walnut. 3.3.3, Strips containing SPOTs as in A and probed with whole patient serum (#3) containing varied amounts (in mg) of soluble E1 peptide. 3.3.4, Strips dotted with rJug r 1 and probed as in 3.3.2, but with added soluble E1 inhibitor.

Linear IgE-epitope Mapping of Ana o 3 (SPOTs analysis) In probing the entire length of the cashew 2S albumin precursor protein, Ana o 3, several IgE-binding peptides were identified across the primary sequence. This was contrary to the finding of only one immunodominant IgE-binding peptide on Jug r 1. The 36-overlapping peptides were probed with four pools of patient sera comprised of patients known to have varying degrees of IgE reactivity to rAna o 3. As presented in Table 3.2, the four pools collectively reacted weakly with nine, moderately with three, and strongly with four peptides, many of which overlapped with one another. Only two of the identified peptides (#’s 6 and 7) were bound by patients’ sera from all four pools. Peptide #6 was bound moderately by pools 1 and 2, and weakly by pools 3 and 4, whereas #7 was bound strongly by pools 1 and 2 and moderately by pools 3 and 4.

Table 3.2. Ana o 3 IgE-binding peptides and relative intensity. 1Bold print denotes strongly reacting epitopes. 2Underlining denotes those epitopes bound by sera from all four patient pools.

IgE Binding Intensity Epitope AA AA sequence position Pool 13 Pool 2 Pool 3 Pool 4 11 33-44 SGREQSCQRQFE +++ + 2 39-50 CQRQFEEQQRFR ++ + + 3 42-53 QFEEQQRFRNCQ + + 4 45-56 EQQRFRNCQRYV + + 5 48-59 RFRNCQRYVKQE + + 62 54-65 RYVKQEVQRGGR ++ ++ + + 7 57-68 KQEVQRGGRYNQ +++ +++ ++ ++ 8 66-77 YNQRQESLRECC + 9 69-80 RQESLRECCQEL + 10 72-83 SLRECCQELQEV + +++ +++ 11 75-86 ECCQELQEVDRR + + + 12 78-89 QELQEVDRRCRC + + + 13 99-110 LQQQEQIKGEEV ++ 14 102-113 QEQIKGEEVREL +++ + 15 105-116 IKGEEVRELYET + 16 123-134 ICSISPSQGCQF +

Comparison of Jug r 1 and Ana o 3 IgE-binding Peptides With Those of Other Known 2S Albumin Allergens To compare the IgE-binding peptides of the walnut 2S albumin (Jug r 1) and cashew 2S albumin (Ana o 3) to those of sesame,71 and yellow mustard seed (Sin a 1)32 2S albumins, we aligned the sequences using the BLAST program and highlighted the corresponding IgE-binding regions (Figure 3.4). Because only a small portion of Sin a 1 was screened and found to bind patient IgE,35 only this portion was included in our analysis. The lone linear IgE-binding epitope of the large subunit of Jug r 1 showed significant sequential overlap with one of the strongly IgE- binding Ana o 3 peptides (#14, Table 3.2) and, though differing in sequence, shared positional homology with epitopes in the yellow mustard 2S albumin and oriental mustard 2S albumin (Bra j 1, data not shown).35 Comparison of the Jug r 1 and Ana o 3 shared epitope showed that both contained the four aa found to be critical (shown in red, Figure 3.4) and the one residue found to be influential (blue, Figure 3.4) for IgE binding to Jug r 1 and shared 100% similarity and 80% identity overall. In contrast, the sequence similarity between the single mustard seed epitope and

the corresponding epitope in Ana o 3 was much less (46%). Consistent with this observation is the finding that only two of the cashew-allergic patients reported concomitant mustard allergy, but all patients that listed specific tree nut allergies, specified walnut as an allergen. In comparing the Ana o 3 epitope map to that of the sesame seed 2S albumin, we found that an extended segment (residues 24 to 94, blue shading in Figure 3.4) showed considerable positional overlap with three of the four strongly reactive Ana o 3 epitopes (#’s 1, 7, and 10). Two of the cashew epitopes, #’s 1 and 7, displayed minimal homology with the corresponding sesame sequence, but the third, epitope #8, showed 75% similarity (58% identity).

Figure 3.4. Amino acid sequence and linear epitope comparison of cashew (Ana o 3), walnut (Jug r 1), sesame, and mustard seed (Sin a 1, segment) 2S albumins. Epitope-containing regions are highlighted or shaded. Strongly reactive Ana o 3 epitopes are boxed. Critical and influential Jug r 1 epitope residues are in red and blue underlined type, respectively. Symbols: “|” = identical residues, “·” = similar residues, “–“space added for alignment.

Linear IgE-epitope Mapping of Jug r 4 (SPOTs analysis) The aa sequence of rJug r 4 was screened for IgE-binding linear epitopes by probing 62- overlapping solid phase synthetic peptides with sera from tree nut allergic patients which show clinical reactions to walnut and have previously shown IgE reactivity to rJug r 4 on Western blots.56 Each reactive serum sample was assigned to one of four pools based on their varying IgE reactivity to rJug r 4 in the blotting assay. As presented in Table 3.3, the four pools collectively reacted weakly with twelve, moderately with seven, and strongly with three

peptides, many of which overlapped with one another and could represent partial epitopes. Only three of the identified peptides (#’s 1, 13 and 14) were bound by patients’ sera from all four pools. Peptide #1 was bound moderately by pools 1 and 4, and weakly by pools 2 and 3; peptide #13 was bound weakly by all pools; and peptide #14 was bound strongly by pools 1 and 3, weakly by pool 2 and moderately by pool 4. Peptide #’s 10, 11, and 19 did not react with sera from all four pools but did react strongly with sera from at least one pool.

Table 3.3. Jug r 4 IgE-binding peptides and relative intensity.

IgE Binding Intensity Epitope AA AA sequence position Pool 1 Pool 2 Pool 3 Pool 4 11 1-15 MAKPILLSIYLFLIV ++ + + ++ 2 57-61 IESWDPNNQQFQCAG + 3 89-103 YSNAPQLVYIARGRG + + + 4 105-119 TGVLFPGCPETFEES ++ 5 121-135 RQSQQGQSREFQQDR + 6 129-143 REFQQDRHQKIRHFR + + + 7 145-159 GDIIAFPAGVAHWSY + + 8 201-215 QGQQEYEQHRRQQQR + + + 9 209-223 HRRQQQRQQRPGEHG + + 102 233-247 VFSGFDADFLADAFN +++ 11 257-271 QSENDHRRSIVRVEG ++ +++ + 12 265-279 SIVRVEGRQLQVIRP + ++ + 13 273-287 QLQVIRPRWSREEQE + + + + 14 281-295 WSREEQEREERKERE +++ + +++ ++ 15 313-327 DDNGLEETICTLRLR + ++ + 16 377-391 PHWNLNAHSVVYALR + + 17 385-399 SVVYALRGRAEVQVV ++ 18 417-431 LTIPQNFAVVKRARN ++ + 19 425-439 VVKRARNEGFEWVSF ++ +++ ++ 20 465-479 LATAFQIPREDARRL ++ + 21 481-495 FNRQESTLVRSRPSR + 22 489-503 VRSRPSRSRSSRSER +

1Underlining denotes those epitopes bound by sera from all four patient pools. 2Bold print denotes strongly reacting epitopes.

Linear IgE-epitope Mapping of Ana o 2 (SPOTs analysis)

The entire aa length of Ana o 2 was studied by probing 58-overlapping solid phase synthetic peptides with sera from cashew-allergic patients. Patients were assigned to one of three pools based on their varying degree of IgE reactivity to rAna o 2 in immunoblotting assays. Collectively, the three pools reacted weakly with 12, moderately with three, and strongly with seven linear IgE-binding epitopes. The 22 reactive epitopes were distributed throughout the length of the protein (Table 3.4) with 68% (15/22) of all epitopes residing on the acidic chain region of the protein, including 86% (6/7) of those epitopes that were found to react strongly with pooled patient IgE. Only two of the identified epitopes were bound by patient sera from all three pools. Peptide #6 was bound strongly by pools 1 and 3, and moderately by pool 2, whereas epitope #3 was bound strongly by pool 1, but moderately by pools 2 and 3.

Table 3.4. Ana o 2 IgE-binding peptides and relative intensity.

Epitope AA AA sequence Phosphoimaging dataa position Pool 1 Pool 2 Pool 3 1 1-15 LSVCFLILFHGCLAS + 2 9-23 FHGCLASRQEWQQQD 2.4 31 15-29 SRQEWQQQDECQIDR +++ + + 4 17-31 QEWQQQDECQIDRLD + 5 89-103 QGEGMTGISYPGCPE + 62 105-119 YQAPQQGRQQGQSGR +++ ++ +++ 7 113-127 QQGQSGRFQDRHQKI + + 8 121-135 QDRHQKIRRFRRGDI + 9 137-151 AIPAGVAHWCYNEGN ++ 10 169-183 LDRTPRKFHLAGNPK ++ 11 185-199 VFQQQQQHQSRGRNL +++ 12 217-231 RLIKQLKSEDNRGGI ++ + 13 233-247 KVKDDELRVIRPSRS + +++ 14 241-255 VIRPSRSQSERGSES +++ 15 257-271 EESEDEKRRWGQRDN +++ +++ 16 313-327 LKWLQLSVEKGVLYK + 17 329-343 ALVLPHWNLNSHSII + 18 337-351 LNSHSIIYGCKGKGQ + 19 377-391 QNFAVVKRAREERFE + + 20 385-399 AREERFEWISFKTND + 21 417-431 PEEVLANAFQISRED + + 22 425-439 FQISREDARKIKFNN +++

1Bold print denotes strongly reacting epitopes. 2Underlining denotes those epitopes bound by sera from all three sera pools.

(1) (2)

SPOT AA Sequence Mutation SPOT AA Sequence Mutation 1 YQAPQQGRQQGQSGR (+) 1 SRQEWQQQDECQIDR (+) 2 AQAPQQGRQQGQSGR Y105A 2 ARQEWQQQDECQIDR S15A 3 YAAPQQGRQQGQSGR Q106A 3 SAQEWQQQDECQIDR R16A 4 YQAPQQGRQQGQSGR A107A 4 SRAEWQQQDECQIDR Q17A 5 YQAAQQGRQQGQSGR P108A 5 SRQAWQQQDECQIDR E18A 6 YQAPAQGRQQGQSGR Q109A 6 SRQEAQQQDECQIDR W19A 7 YQAPQAGRQQGQSGR Q110A 7 SRQEWAQQDECQIDR Q20A 8 YQAPQQARQQGQSGR A111A 8 SRQEWQAQDECQIDR Q21A 9 YQAPQQGAQQGQSGR P112A 9 SRQEWQQADECQIDR Q22A 10 YQAPQQGRAQGQSGR Q113A 10 SRQEWQQQAECQIDR D23A 11 YQAPQQGRQAGQSGR Q114A 11 SRQEWQQQDACQIDR E24A

12 SRQEWQQQDEAQIDR C25A 12 YQAPQQGRQQAQSGR G115A

13 SRQEWQQQDECAIDR Q26A 13 YQAPQQGRQQGASGR Q116A

14 SRQEWQQQDECQADR I27A 14 YQAPQQGRQQGQAGR S117A

15 YQAPQQGRQQGQSAR G118A 15 SRQEWQQQDECQIAR D28A 16 SRQEWQQQDECQIDA R29A 16 YQAPQQGRQQGQSGA R119A

17 RFRRGDIIAIPAGVA (-) 17 RFRRGDIIAIPAGVA (-)

Spot # Spot #

Figure 3.5. Mutational analysis of two immunodominant Ana o 2 IgE-binding peptides. SPOTs layout and IgE reactivity (represented graphically) to peptides #6 (3.5.1, SRQEWQQQDECQIDR) and #7 (3.5.2, YQAPQQGRQQGQSGR). Peptides in which an alanine substitution decreased IgE reactivity by 50% or more are highlighted in yellow.

These peptides were considered immunodominant and studied further in alanine scanning mutagenesis experiments (Figure 3.5). Probing of the mutated peptides with pooled patient sera demonstrated that, for peptide #6, five aa from residues 20-24 would reduce patient-IgE

reactivity by more than 50% if mutated to alanine (Figure 3.5.1). Similarly, IgE binding to peptide #7 was reduced by greater than 50% when aa Q109 was mutated to alanine and completely eliminated when aa Q114 or G115 were mutated (Figure 3.5.2).

Comparison of Jug r 4 and Ana o 2 IgE-binding Peptides With Those of Other Known 11S Globulin Allergens To compare the linear epitopes of walnut (Jug r 4) and cashew (Ana o 2) legumins to those of the peanut (Ara h 3),13 hazelnut (Cor a 9, unpublished data) and soybean legumins (G2 and G1 glycinin),72, 73 we aligned the sequences using the BLAST program and highlighted the corresponding linear epitopes (Figure 3.6). Those Jug r 4 and Ana o 2 peptides that bound at least one pool of walnut or cashew-sensitive patient IgE strongly are outlined in blue boxes. One significant finding is that when the aa sequences of these proteins are aligned, two of the strong IgE-binding peptides from Jug r 4 overlap in position and show considerable sequence homology with epitope-containing regions previously identified on several of the other allergens (Figure 3.6, dotted black boxes). It was also interesting to find that the two other strongly reactive Jug r 4 peptides, also outlined in blue, overlapped in position and again showed sequence homology to only Ana o 2 IgE-reactive peptides. In fact, all but two of the 12 IgE-binding regions mapped on Jug r 4 overlapped at least slightly in position and in most cases showed sequence homology to Ana o 2 epitopes. This finding was consistent with the shared epitope motif observed between the 2S albumin allergens from these nuts, Jug r 1 and Ana o 3. Beyond the comparison of the walnut and cashew 11S globulins, there are several allergenic ‘hot spots’ highlighted in the composite map. These six regions (Figure 3.6, dotted red and black boxes) identify shared areas of IgE binding in four or more of the epitope mapped legume and tree nut allergens and suggest the possibility of a shared structural motif favoring immunogenicity. Looking more closely at just the tree nut allergenic legumins, Jug r 4, Ana o2, and Cor a 9, a great deal of sequence overlap and homology is obvious among the identified epitopes. In fact, eight of the nine Cor a 9 IgE-binding regions showed sequence homology to epitope containing areas of Ana o 2. The same sort of homology was observed among the legume allergens, Ara h 3 and the G1 glycinin. This was not the case, however, when comparing Ana o 2 epitopes with those of the peanut 11S globulin. Interestingly, only two of the 22 Ana o 2 linear epitope-bearing peptides (#’s 13 and 15, Figure 3.4) showed significant (>7 aa)

positional overlap with any of the four previously identified peanut Ara h 3 epitope-bearing peptides.13 Even within these two overlapping regions, there was little sequence similarity between the Ana o 2 and Ara h 3 epitopes. This same pattern was observed when comparing the Jug r 4 and Cor a 9 sequences to the Ara h 3 epitopes; while some positional overlap occurred, very little sequence homology was observed. A much greater degree of correspondence was observed when the sequences of Ana o 2 and G2 glycinin were aligned; nine of the 11 (89%) previously identified G2 glycinin epitopes showed significant positional overlap with the Ana o 2 epitopes.

Figure 3.6. Comparison of the IgE-binding regions of Jug r 4 and Ana o 2 with various allergenic 11S globulins. Amino acid sequence and linear epitope comparison of cashew (Ana o 2), walnut (Jug r 4), hazelnut (Cor a 9), soybean (G2 and G1 glycinin), and peanut (Ara h 3) legumins. Segments indicating regions of allergens expressing one or more IgE-binding peptides are highlighted in different colors; Red- Ana o 2, Olive- Jug r 4, Purple- Cor a 9, Teal- G2 glycinin, Yellow- Ara h 3, Green- G1 glycinin. Strongly reactive epitopes for Jug r 4 and Ana o 2 are boxed in blue. Identical aa are indicated with a “·” while similar residues are depicted with “|”. A “–“ represents positions where a space was added and a “/” represents positions where part of the sequence was deleted to maximize alignment. Dotted black and red boxes represent allergenic “hot spots” on the 11S globulins as described in the manuscript.

A o 2: 1 LSVCFLI--LFHGCLA--SRQEWQQQD--ECQIDRLDALEPDNRVEYEAGTVEAWDPNHEQFRCAGVALVRHTIQP 70 J r 4: 12 ----···VA······--QSGGR···QFG·········T···A···V··············R··· 81 C a 9: 10 -··L·--·····G-/R··QYFG--··N·······T···A··CQ···H·D········R··· 83 G2gly: 6 ·····---··S··F·--L·-A··--····Q········S·G·F·T···P·······S·C·NR 73 A h 3: 2 ------·FQP--A··FQ···QR·····S·G·Y·T····E······S·LVR 57 G1gly: 7 -·····--FSGC·F·F-·S·QP··--····Q········S·G·L·T···P·······S·C·NR 76

A o 2: 71 NGLLLPQYSNAPQLIYVVQGEGMTGISYPGCPETYQAPQ-QGRQ-QGQ-SGRFQ-DRHQKIRRFRRGDII 136 J r 4: 82 ···············A·R···LF······ES·-QS·-···-·RE··Q·······H··E···· 149 C a 9: 84 ···············E·R···L······D··-·QS·-···/QRSE·-·······H··E···· 149 G2gly: 74 ·A·RR·S··G··E··Q··N·F·I····S···E··-S·-·-·Q·P·-·····H···E··· 139 A h 3: 58 ·A·RR·F······E·Q··R·YF·I····RH·E·HT···/D·-·QQ-·S···H··DE··· 136 G1gly: 77 ·A·RR·S··G··E··Q···F·I·····S·E··-·P·-··-·S·P·-······YN··E··· 142

A o 2: 137 AIPAGVAHWCYNEGNSPVVTVTLLDVSNSQNQLDRTPRKFHLAGNPKDVF-Q-QQ-QQHQSR-GRN 198 J r 4: 150 ·F·······S······A···T·A····N··N······D·E·-/·-··R·Q·/·N· 232 C a 9: 150 ····················HT·YA····EN··H······D·EH-·-·/···GE-·N· 239 G2gly: 140 ··T···W·M··NE···A··TL····M·······QQE·/·-··-··/E-·S· 215 A h 3: 137 ··T···F·L··HD··A··T·T·D····F·······TQE·/Y-··-·SR···/·G· 241 G1gly: 143 ··T···W·M··NE···A··TL····M·······QQE·/-H·-·EE-·G 219

A o 2: 199 LFSGFDTELLAEAFQVDER-LIKQLKS-E-DNRGGIVKVKDDELRVIRP--S-RS-QSE-RGS-ESE 256 J r 4: 233 ·····AF····N··TE-TA··-·N··RS···GR·····RW·-·E-Q·-·EE/·R· 297 C a 9: 240 ·····A·F····N··VD-TA··N-·K·RN····GR-····--·E··-Q·/ER-··· 303 G2gly: 216 L···AP·F·K···G·M-N·/-·-S·A··T··GG-···TA·--/P-·Q·-EDD-D· 276 A h 3: 242 ····TP·F·E······QN·G-·/EE·A··T·GG-··S·--D-·K/E·-EYD-·D· 306 G1gly: 220 L···TL·F·EH··S··-A·N·G-·/·A··T··GGLSVKP·---DE-·Q-·PQ-·E· 280

Å Acidic Subunit ↓ Basic Subunit Æ

A o 2: 257 EES-EDEK-R-RWGQ-RD-NGIEETICTMRLKENINDPARADIYTPEVGRLTTLNS 307 J r 4: 298 R··-·S·-·/·-G·-D··············G········E·A····· 351 C a 9: 304 ·R-·R-·-G·G··V··F··········CTR······E···N··· 356 G2gly: 277 ··Q/·T·/·-SK-SR-·············GQNSP···N·A·S··AT· 336 A h 3: 307 Y·Y-·D-·-·R·/·G-·········ASA···GRNRSP···N·A·LSK·A·D 360 G1gly: 281 ··E-····-P-CKG-·/···········H··GQTSP···N·A·S··AT· 346

A o 2: 308 LNLPILKWLQLSVEKGVLYKNALVLPHWNLNSHSIIYGCKGKGQVQVVDNFGNRVFDGEVREGQM 372 J r 4: 352 HT········A··A··S··Y·········AL·A········QT···D····· 416 C a 9: 357 NT········A··D·QEG·Y·········AI·A·····N··T···D···· 421 G2gly: 337 ·F·A·WL···AY·S·R···F··T······ALN·AL····CN·E········G 401 A h 3: 361 ···L····G··A·Y·N·····FA··T······RL·AH·····N·····E···H 426 G1gly: 347 ·F·A·S····A·F·S·R···F·········ALN·AL···CN·E········ 406

A o 2: 373 LVVPQNFAVVKRAREERFEWISFKTNDRAMTSPLAGRTSVLGGMPEEVLANAFQISREDARKIKFNNQQT 442 J r 4: 417 ·T···········N·G········N··V········ARA······T····P········R· 486 C a 9: 422 ·T······A···ES·G········N·QI········ARA················R·E· 491 G2gly: 402 ········AASN·········PSIGN···AN··NA····QT·NKSQ···N··PF 470 A h 3: 427 ·········AGS·N·····S·PSIN···EN··DN······GQ··Q···N··--- 493 G1gly: 407 ······V·AA·SN········TP·IGT···AN··NA····QT·NKSQ····N··--- 478

Discussion

Although the 2S albumins are common seed and tree nut allergens, few have been tested for cross-reactivity. Cross-reactivity has only been demonstrated between rapeseed and mustard seed35 and between almond, walnut, and hazelnut 2S albumins.42 The basis for the observed cross-reactivity, however, has not been determined. One strongly reactive Ana o 3 epitope overlapped in aa position with the lone linear epitope of the large subunit of the English walnut allergenic 2S albumin, Jug r 1, with which it shares considerable homology (81% similarity). Significantly, all five aa found to be critical or influential for IgE binding to the Jug r 1 epitope69 were conserved in Ana o 3 (100% similarity, 80% identity).58 This finding may partly explain why the majority of cashew-allergic patients in our test group are also allergic to walnut and was supported by the results of our 11S globulin epitope mapping studies. A recent study by Barre et al74 involving homology modeling of the cashew, walnut, peanut, and castor bean 2S albumin allergens actually showed a shared structural motif between the identified Jug r 1 epitope and those corresponding regions on the molecular surface of the other allergens. They describe their findings as a molecular basis for IgE-binding cross-reactivity among such dietary nuts as cashew and walnut. Our linear epitope mapping studies of 11S globulin allergens have shown that allergenic epitope ‘hot spots’ are present among the tree nut allergens Jug r 4 (walnut), Ana o 2 (cashew), and Cor a 9 (hazelnut). All but two of the 12 IgE-binding regions mapped on Jug r 4 overlapped in at least position and, in most cases, showed sequence homology, with the Ana o 2 epitopes. This overlap included three of the strongly reactive IgE-binding peptides of Jug r 4. This finding was consistent with the shared epitope motif observed among the 2S albumin allergens, Jug r 1 and Ana o 3, in these nuts. Similarly, eight of the nine Cor a 9 IgE-binding regions showed sequence homology to epitope containing areas of Ana o 2. Recent studies support our epitope mapping studies and analysis in that Wallowitz et al75 have shown that cross-reactive patient IgE directed at 11S globulins may exist among patients sensitive to tree nuts. Inhibition blots revealed pre-incubation of sera with walnut, cashew, or hazelnut extracts resulted in a significant decrease in IgE binding to the each of the corresponding 11S globulin recombinant proteins, indicating allergenic cross-reactivity.

Our finding that the soybean G1 glycinin acidic chain may share cross-reactive epitopes with the peanut Ara h 373 was also not surprising when one considers that both are members of the legume family and patients allergic to one sometimes have serum IgE antibodies that immunologically cross-react with other legumes.76 It has even been shown that the IgE-binding epitopes of the soybean G2 glycinin also bind IgE from peanut sensitive patients.72 Clinical information regarding cross-reactivity between tree nuts and legumes has not been well documented, but based on our findings we can predict that while there may be a basis for cross-reactivity between cashew and soybean, it is unlikely that such cross-reactivity occurs between tree nut and peanut legumins, at least with regard to linear epitopes. Similarly, we previously reported that cross-reactivity between the vicilin linear epitopes of cashew (Ana o 1) and peanut (Ara h 1) would also be unlikely,59 observations which, together, further explains the lack of reported cross-reactivity between tree nut- and peanut-reactive patients’ sera.28 As the database for linear epitopes of allergens grows, patterns of additional IgE-reactive hot spots are likely to emerge. With this information, studies involving the analysis of new allergens and the engineering of proteins for plant breeding and therapeutic agents can be better focused. We have also demonstrated that IgE reactivity to epitopes identified on the walnut 2S albumin as well as the cashew 11S globulin could be significantly reduced and in some cases completely eliminated by simply altering one aa, a finding that provides the most useful information for the efforts described above. To date, no common structural characteristics of 11S globulin linear-IgE epitopes have been identified, but our findings that shared linear IgE- binding epitope motifs are present within the primary aa sequences of this family of seed storage proteins suggests that epitope mapping can provide important clues to the nature of allergenicity.77 Three-dimensional modeling of such allergenic proteins displaying their linear IgE-binding epitopes may provide additional insights towards identifying shared structural characteristics and will be discussed in chapter 5.

CHAPTER 4

IDENTIFICATION OF A CASHEW NUT 11S GLOBULIN (ANA O 2) CONFORMATIONAL EPITOPE

Introduction

In chapter 3, I pointed out the importance of and identified the linear IgE-binding epitopes of the 2S albumin and 11S globulin allergens of the English walnut and cashew nut. Largely left unexplored in the allergy field is the identification of conformational epitopes. It is now becoming clear that studying this class of epitopes can lead to a deeper understanding of the structural features that are immunogenic in these molecules and to a determination of whether such features can be correlated to clinical symptoms and histories. The identification of conformational binding sites, however, presents a special challenge since most are composed of non-contiguous (discontinuous) peptide segments that are only presented in the correct array in the native folded protein. Some estimates actually suggest that the majority of binding sites on protein antigens are conformational.20, 78 It has also been suggested that a differential responsiveness to linear versus conformational epitopes in allergic patients is of considerable clinical and diagnostic significance.20, 79-81 For example, children who outgrow allergies to ovomucoid, a major allergen in egg82 and casein in cow’s milk83 may develop IgE to mostly conformational epitopes whereas those that remain allergic to these proteins typically target linear epitopes. At a more fundamental level, the process by which the immune system targets specific epitopes is still largely unknown and, with regard to allergy, the role that epitope configuration and presentation to the IgE-charged mast cells and basophils plays in triggering needs elucidation. Certainly, issues relating to epitope stability during food processing, digestion and absorption following consumption, and the ability of IgE to cross-link are key factors that must also be considered. However, such studies are predicated on proof that the binding of IgE to conformational epitopes does indeed occur in the native protein. In this study we provide evidence that a conformational epitope on the cashew nut legumin, Ana o 2, binds IgE from cashew-sensitive patient sera. The 11S globulins of the tree nuts represented a logical choice in beginning to search for conformational epitopes as the probability of such binding sites being present would seem greatest within folded regions of the

individual large and small subunits or as a result of these subunits joining to form a heterodimer. Moreover, there is also the possibility that conformational epitopes could potentially form when the monomeric form of the heterodimeric protein formed trimers or eventually hexamerized. The relative importance of these proteins in the seeds of tree nuts is also noteworthy as they are found in abundance in all tested tree nuts, are reactive with IgE from a majority of tested nut allergic-patients, and have been the most thoroughly studied tree nut allergens to date.11 Of particular significance to our study was the fact that we had already characterized and mapped the linear epitopes of Ana o 259 (Figure 3.6), thus allowing us to focus on the more intractable, yet potentially more important, conformational epitopes of this molecule. Unfortunately, because the technology for identifying linear epitopes is considerably more straightforward than the characterization of conformational epitopes, rather little attention has been paid to the latter.69, 77 To date, most of the efforts aimed at allergen conformational epitope identification rely on expressing and probing PCR-generated gene fragments of various length and overlap.84, 85 In such experiments, researchers usually identify a large segment (~80 residues) of the allergen to which IgE binds and provide proof that smaller comprising segments do not, individually or in summation, mimic this binding.84 While this procedure is used to address conformational epitopes that form through relatively proximal aa coming together or influencing one another in a somewhat non-contiguous manner, it does little for identifying those epitopes that form strictly as a result of protein folding or the protein’s oligomerization state. We propose a strategy that will help solve this problem and address the concern that isolated fragments might not possess all IgE-binding conformational epitopes simply because the segments, when not constrained by the remainder of the allergen structure, may not retain or acquire the correct native configuration needed for epitope expression. The technique is based largely on the approach used by Karisola et al86 in their identification of the major conformational IgE-binding an epitope on the latex allergen, hevein (Hev b 6.02). In this study, the authors utilized a novel PCR-based approach to graft suspected conformational epitope bearing N-and C-terminal segments of the latex allergen to corresponding sides of the core of a homologous antimicrobial protein.86 The authors observed a lack of IgE binding to the core protein as well a reduced binding to the hevein N- and C-terminal fragments when expressed alone. They then found that the chimeric molecule had the ability to reconstitute latex-allergic patient IgE binding in all tested patients, leading them to conclude that such binding is

essentially determined by reactivity to the N- and C-terminal portions of the protein in a conformational manner. Based on this approach we grafted a gene encoding the large subunit of Ana o 2 to a gene encoding the small subunit of the homologous soybean 11S globulin, Gy1 glycinin. Similarly, we created another chimeric molecule in which the large subunit of the Gy1 glycinin was expressed in association with the small subunit of Ana o 2. Additionally, as controls, we expressed each protein as a full-length precursor-MBP fusion protein and did the same for the individual large and small subunits. Ana o 2-specific mAbs were then screened for reactivity to each of the constructs and those of interest were used along with patient sera in inhibition ELISAs and dot blot assays to determine if there was a correlation between the mouse- IgG and human-IgE epitope binding. We have found, through this chimeric-based allergen screening approach, that human IgE binds to a conformational epitope on Ana o 2. Our studies may help lay the foundation for future work in determining the physical state (conformation) of the allergen during the priming and reactive phases of allergy and might reveal subtle differences in the manner by which allergens are digested and passaged into the body in allergic and non-allergic subjects.

Methods

PCR-based Amplification of Ana o 2 and Gy1 Glycinin Full-length and Subunit-Expressing Genes A recombinant pMAL-c2-His vector containing the full-length Ana o 2 gene (described chapter 2) was used as a template for PCR-based cloning of the gene’s large and small subunits. The target coding sequences were modified by the addition of restriction sites at the 5’ and 3’ ends during the PCR reactions using PfuTurbo DNA polymerase (Stratagene) and the appropriate primers (Table 4.1). For the soybean Gy1 glycinin gene (Genbank accession # X15121), which was graciously provided to us in a pET28 bacterial expression vector (Invitrogen) by Dr. Rudolf Jung (Pioneer Hi-Bred Intl. Inc., Johnston, IA), this same strategy was employed and the full-length protein as well as it’s large and small subunits were amplified. The restriction sites for all constructs were chosen after creating a RE map of each gene as previously described.

Table 4.1. Primers used for the amplification and subcloning of cashew and soybean 11S globulin genes and constituent subunits.

Gene Target Primer (5’- 3’) Ana o 2 Full length Forward: AAACTGCAGTCTCGCCAGGAATGGCAAC Reverse: CCCCAAGCTTTTAAGCATCATCCCTCATGTGG Ana o 2 Large subunit Forward: AAACTGCAGTCTCGCCAGGAATGGCAAC Reverse: CCCCAAGCTTTTAATTGTCACGCTGTCCCCATC Ana o 2 Small subunit Forward: AAACTGCAGGGGATTGAAGAGACCATTTGC Reverse: CCCCAAGCTTTTAAGCATCATCCCTCATGTGG Gy1 Full length Forward: AAACTGCAGTCCAGAGAGCAGCCTCAGC Reverse: CCCCAAGCTTCTAAGCCACAGCTCTCTTCTG Gy1 Large subunit Forward: AAACTGCAGTCCAGAGAGCAGCCTCAGC Reverse: CCCCAAGCTTCTAATTTCTTCTGCTTTTGCTTTGGC Gy1 Small subunit Forward: AAACTGCAGGGCATTGACGAGACCATATGC Reverse: CCCCAAGCTTCTAAGCCACAGCTCTCTTCTG

1 Restriction sites underlined; CTGCAG= PstI, AAGCTT = HindIII.

PCR-based Construction of Cashew and Soybean Chimeric 11S Globulin Genes In addition to cloning the full-length and corresponding large and small subunit- expressing genes for both Ana o 2 and Gy1 glycinin, we created two chimeric molecules that bore fragments from each gene. The first chimeric construct was comprised of the large subunit of Ana o 2 and the small subunit of Gy1 glycinin (A2 LG-Gy1 SM) whereas the second chimera was constructed from the large subunit of Gy1 glycinin and the small subunit of Ana o 2 (Gy1 LG-A2 SM). To do this, PCR reactions separate from those previously mentioned, had to be performed so that the constituent subunit-bearing portions of the full length genes did not have thymidine (T) hangovers, a processing residual of the PfuTurbo DNA polymerase that would hinder any subsequent blunt end ligations. To overcome this modification, Platinum Pfx DNA polymerase (Invitrogen) was used in all PCR reactions as described by the manufacturer. We also did not include restriction sites or stop codons on the 5’ ends of the large subunit reverse primers or small subunit forward primers because these ends of the genes needed to be blunt end ligated and then serve as full-length templates for Pfu Turbo DNA polymerase in subsequent

PCR reactions. Only details of those primers used in the Platinum Pfx DNA polymerase PCR reactions that differed from those summarized in Table 4.1, are shown in Table 4.2.

Table 4.2. Primers used to amplify Ana o 2 and Gy1 glycinin large and small subunits for use in blunt end ligation reactions.

Gene Target Primer (5’- 3’) Ana o 2 Large subunit Reverse: TTAATTGTCACGCTGTCCCCATC Ana o 2 Small subunit Forward: GGGATTGAAGAGACCATTTGC Gy1 Large subunit Reverse: CTAATTTCTTCTGCTTTTGCTTTGGC Gy1 Small subunit Forward: GGCATTGACGAGACCATATGC

1All primers were phosphorylated on the 5’ ends.

After amplification of the modified large and small subunits of both the Ana o 2 and Gy1 genes via PCR, products were purified using the QIAquick PCR Purification Kit (Qiagen) and blunt end ligations were carried out using these purified templates and T4 DNA Ligase H.C. (Invitrogen) as described by the manufacturer. A working dilution of the ligation mixtures, ranging from 0.01-1.0 µl, was used as template DNA and the full-length chimeric molecules were amplified with the appropriate primers; Ana o 2 large subunit forward and Gy1 small subunit reverse primer (as shown on Table 4.1) for the A2 LG-Gy1 SM ligation mixture and Gy1 large subunit forward and Ana o 2 small subunit reverse primer (also shown on Table 4.1) for the Gy1 LG-A2 SM ligation mixture.

Subcloning of Ana o 2, Gy1 Glycinin, A2 LG-Gy1 SM, Gy1 LG-A2 SM, and Subunit Genes After each full-length gene, chimeric construct, or individual subunit was amplified as described above, the PCR products were purified using the QIAquick PCR Purification Kit (Qiagen Sciences) and digested using PstI and HindIII and the appropriate reaction buffers (Invitrogen) according to the manufacturers instructions. Cut products were purified using a QIAquick Gel Extraction Kit (Qiagen Sciences) and ligated with T4 DNA Ligase H.C.

(Invitrogen) to their respective sites of the RE cut pMAL-c2-His expression vector. Sequences were checked for identity to their respective templates using the BLAST program after both strands of the transfected vectors were sequenced as previously described.

Expression and Purification of MBP-Fusion Proteins Expression and purification of all cDNA/pMAL-c2-His plasmids was carried out as described in chapter 2.

Monoclonal Antibody (mAb) Production and Ana o 2 Reactivity Ana o 2-reactive monoclonal antibodies (mAbs) were generated in the Hybridoma Core Facility at Florida State University according to standard procedures as described below. Briefly, BALB/c mice were immunized with 25 µg of whole cashew extract in RIBI adjuvant (Corixa Inc., Hamilton, MT) according to the manufacturers instructions. The initial immunization was given ½ subcutaneously and ½ interperitoneal (i.p.). Three weeks after the immunization, healthy mice were given a primary boost with 15-20 µg of antigen, again using the RIBI adjuvant system. Two weeks following the boost, mice were test bled for and their Ab titer against whole cashew extract determined via ELISA using plates coated with 10 µg/ml of whole cashew extract and probed as previously described.87 When the titer approached 1:10,000, the mice were given a final boost with 15 µg antigen in saline, ½ intravenously (i.v.) in the tail vein and ½ i.p. Three days after the final boost, the mice were sacrificed and their spleens were removed. Spleen cells were dispersed, washed, and red blood cells removed with ammonium chloride prior to fusion with NS-2 myeloma cells (ATCC, Manassas, VA) which were mixed at a ratio of 1:5 (myeloma cells:spleen cells). Fused cells, achieved using polyethylene glycol (PEG), were plated in HAT media and allowed to grow for two weeks. Surviving hybridoma cells were screened for reactivity to whole cashew extract via ELISA. Those mAbs found to be reactive with whole cashew in this initial screening were then checked for reactivity to rAna o 2 in a subsequent ELISAs using plates coated with 5 µg/ml of the recombinant fusion protein. Reactive clones of interest were subcloned until a total of 20 monospecific Ana 2-binding mAbs were obtained.

Purification of Native Ana o 2 From Whole Cashew Extract To compare the fusion protein reactivity of the mAbs with that of native Ana o 2, the 11S globulin fraction was purified from whole cashew extract, prepared as described earlier, by using a Superdex 200 column (Amersham Pharmacia) and high performance liquid chromatography (HPLC, Pump 126, UV Detector 168, Beckman Coulter, Fullerton, CA).

Dot Blot Screening of Proteins With Ana o 2-specific mAbs For each fusion protein 0.4 µM in 100 µl BSB was dotted onto 0.2 µm nitrocellulose (Schleicher & Schuell) using the Bio-Rad Dot Blot apparatus (Bio-Rad) according to the manufacturers instructions. As a negative control 0.4 µM of MBP (New England Biolabs) was also dotted onto each membrane to make certain that any mAb reactivity to the fusion protein was specific for the expressed portion of the Ana o 2 or Gy1 gene. Because samples were not completely homogenous, 0.2 µg whole cashew extract and purified native Ana o 2 were dotted onto membranes for mAb blotting. Dotted membranes were blocked overnight at 4ºC in 5% non-fat dry milk in PBS-T and cut into strips containing the desired antigen array. Monoclonal antibodies were diluted 1:50-1:2500 depending on signal intensity observed in previous assays and incubated for 1 hr at room temperature with the membrane. Three 15-min washes were performed with TBS-T followed by three 5-min washes. Membranes were then probed with an HRP-secondary Ab, goat-anti-mouse-IgG (Jackson Immunoresearch, West Grove, PA) for 1 hr at room temperature and washed as described above. Reactive dots were identified after a 5-min incubation in ECL+ (Amersham Pharmacia), prepared following manufacturer’s instructions, and subsequent exposure to Kodak X-OMAT x-ray film.

Conformational Sensitivity of mAb Binding to Reactive Fusion and Native Proteins Based on screening of the fusion proteins with the mAbs and previous unpublished data of mAb binding to Ana o 2 SPOTs, linear and conformational epitope binding mAb candidates were chosen. Each mAb was used to probe versions of their target proteins that had been subjected to various denaturing conditions prior to dot blotting. Proteins were either boiled or not boiled in 100 µl BSB alone or BSB containing either 0.1-10% SDS, 250 mM-1 M β-ME, or 100 mM-6 M urea and dotted onto nitrocellulose for use in blotting assays as described above.

For native proteins, 0.5 µg aliquots of whole cashew extract or HPLC fractionated extract samples in 100 µl BSB were used per dot as described above.

Mouse mAb (IgG)/Human IgE Inhibition ELISA The ability of mouse IgG to inhibit Ana o 2-specific IgE reactivity was tested via ELISA.

Plates were coated with 20 µg/ml of Ana o 2 in coating buffer (0.1 M NaHCO3, pH 9.6) for 2 hr at 37ºC, blocked in 5% non-fat dry milk in PBS-T for 1 hr at 37ºC and washed three times with PBS-T. Each undiluted mAb (50 µl) was then added to each appropriate well and incubated overnight at 4ºC. A non-Ana o 2 binding mAb (walnut specific) and PBS-T were used as inhibitors for positive controls. After three washes with PBS-T the plates were incubated with Ana o 2-reactive human sera, diluted 1:20 in blocking buffer, overnight at 4ºC. Washing was again performed with PBS-T and an HRP-labeled secondary Ab, mouse-anti-human-IgE (Zymed Labs, San Francisco, CA) was diluted 1:1000 in blocking buffer and incubated on the plate for 1 hr at 37ºC. Detection was achieved after washing by incubation with a O-Phenylenediamine tablet (Zymed) diluted in 12 ml substrate buffer (0.1 M Citrate phosphate buffer, pH 5.0) for 10 min. Plates were read at a wavelength at 495 nm using a µQuant ELISA reader (Bio-Tek Instruments, Highland Park, VT) and accompanying software according to the manufacturers instructions. All assays were at least repeated in triplicate.

Dot Blot Inhibition With Human Sera of mAb Binding To show complete inhibition between the mouse monoclonal IgG and human polyclonal IgE classes, we performed dot blot inhibition assays whereby dotted proteins were blocked as described earlier and pre-incubated with human sera diluted 1:10 in blocking buffer overnight at 4°C. Membranes were then washed with TBS-T and probed again with the appropriate mAb diluted 1:10,000 in TBS-T. For positive controls, a non-Ana o 2 binding mAb (walnut specific) was used as an inhibitor. Direct human-IgE binding to the dotted proteins was also performed after blocking by blotting with patient sera, washing with TBS-T, and probing with the same HRP-labeled mouse-anti-human-IgE secondary Ab used in the ELISA assay (Zymed) diluted 1:5,000. Reactive dots were identified after a 5-min incubation in ECL+ (Amersham Pharmacia) as described earlier.

Results

PCR-based Amplification of Ana o 2 and Gy1 glycinin Full-length and Subunit- expressing Genes The 1332 bp open reading frame of the Ana o 2 gene, excluding the leader peptide, encoding for a 443 aa protein (A2, Figure 4.1), was amplified as described above. Similarly, with this gene serving as a template, a 771 bp fragment corresponding to the 257 aa large subunit (Figure 4.1, blue) and a 558 bp fragment corresponding to the 186 aa small subunit (Figure 4.1, red) of Ana o 2 were also amplified. The leader sequence-deleted open reading frame of the Gy1 glycinin was cloned and the 1425 bp gene encoded for a 475 aa protein (G1, Figure 4.1). With this gene serving as a template, a 867 bp fragment corresponding to the 289 aa large subunit (Figure 4.1, green) and 558 bp fragment corresponding to the 186 aa small subunit (Figure 4.1, purple) of Gy1 were also amplified. Deduced aa sequence alignment of the two legumins (Figure 4.1) using the BLAST program revealed a 47% identity and 65% similarity match. Alignments of the proteins’ subunits led to similar results; 46% identity/65% similarity between the two large subunits and 52% similarity/71% similarity between the two small subunits. These results led us to hypothesize that the Gy1 glycinin subunits would serve as suitable surrogates for the Ana o 2 subunits when attempting to create chimeric molecules with structural properties similar to that of the original protein.

PCR-based Construction of Cashew and Soybean Chimeric 11S globulin Genes The first 11S globulin chimeric construct we created included the large subunit of Ana o 2 (Figure 4.1, blue) and the small subunit of Gy1 glycinin (Figure 4.1, purple). The second chimera was constructed from the large subunit of the Gy1 glycinin (Figure 4.1, green) and the small subunit of Ana o 2 (Figure 4.1, red). The A2 LG-Gy1 SM and Gy1 LG-A2 SM chimeras are shown schematically in Figure 4.2.

A2 3 QEWQQQDECQIDRLDALEPDNRVEYEAGTVEAWDPNHEQFRCAGVALVRHTIQPNGLLLP 62 |· ··|···· |·|··|····|· · · |· ·|··|| ·|······ · ·| · · · G1 2 REQPQQNECQIQKLNALKPDNRIESEGGLIETWNPNNKPFQCAGVALSRCTLNRNALRRP 61

A2 63 QYSNAPQLIYVVQGEGMTGISYPGCPETYQAPQQGRQQGQSGRFQDRHQKIRRFRRGDII 122 ·|· ·· ··| ··|·| ·| ····· ·|| ··· |·|··· · ······· ·· ··|· G1 62 SYTNGPQEIYIQQGKGIFGMIYPGCPSTFEEPQQPQQRGQSSRPQDRHQKIYNFREGDLI 121

A2 123 AIPAGVAHWCYNEGNSPVVTVTLLDVSNSQNQLDRTPRKFHLAGNPKDVF------172 ·|· ··· · ·· ||··· ·|||· || |····| ··|·|···· | · G1 122 AVPTGVAWWMYNNEDTPVVAVSIIDTNSLENQLDQMPRRFYLAGNQEQEFLKYQQEQGGH 181

A2 173 -----QQQQQHQSRGRNLFSGFDTELLAEAFQVDERLIKQLKSE---DNRGGIVKVKDDE 224 | ··| || · || ··· · · ·· ··||| · ·| · |||· ·· ·· G1 182 QSQKGKHQQEEENEGGSILSGFTLEFLEHAFSVDKQIAKNLQGENEGEDKGAIVTVKGGL 241

A2 225 LRVIRPSRSQSERGSESEEESEDEKRRWGQRD------NGIEETICTMRLK 269 | ·| · |· · ··· ···· | |· ···|········| G1 242 SVIKPPTDEQQQRPQEEEEEEEDEKPQCKGKDKHCQRPRGSQSKSRRNGIDETICTMRLR 301

A2 270 ENINDPARADIYTPEVGRLTTLNSLNLPILKWLQLSVEKGVLYKNALVLPHWNLNSHSII 329 ·· | ··· ·| · |·· ··| · · ··|·· · · · ···| |··|···||··· G1 302 HNIGQTSSPDIYNPQAGSVTTATSLDFPALSWLRLSAEFGSLRKNAMFVPHYNLNANSII 361

A2 330 YGCKGKGQVQVVDNFGNRVFDGEVREGQMLVVPQNFAVVKRAREERFEWISFKTNDRAMT 389 · ·| |···| · ······||··||·|····· · ·|| | ··||······ · G1 362 YALNGRALIQVVNCNGERVFDGELQEGRVLIVPQNFVVAARSQSDNFEYVSFKTNDTPMI 421

A2 390 SPLAGRTSVLGGMPEEVLANAFQISREDARKIKFNN 425 ··· ·|· |····| | · | | ··|·· ·· G1 422 GTLAGANSLLNALPEEVIQHTFNLKSQQARQIKNNN 457

Figure 4.1. BLAST alignment of the deduced aa sequences of cashew (Ana o 2) and soybean (Gy1 glycinin) 11S globulins. Identical aa are indicated with a “·” while similar residues are depicted with “|”. A “–“ represents positions where a space was added to maximize alignment. Amino acids corresponding to the Ana o 2 (A2) large (blue) and small (red) subunits as well as the Gy1 glycinin (G1) large (green) and small (purple) subunits are shown.

A2 LARGE Gy1 SMALL Gy1 LARGE A2 SMALL

Figure 4.2. A schematic representation of the A2 LG-Gy1 SM (left) and Gy1 LG-A2 SM (right) chimeric molecules. Colors for each subunit correspond to those of their sequences in Figure 4.1.

Expression and Purification of MBP-Fusion Proteins After each insert was successfully subcloned into the pMAL-c2-His expression vector and the correct sequence confirmed, fusion proteins were expressed and purified as previously described. Yields ranged from 5-50 mg/L of cultured bacteria with the lowest yields usually belonging to the low molecular weight small subunit fusion proteins.

Monoclonal Antibody Production and Ana o 2 Reactivity Twenty mAbs proved to be whole cashew-reactive when screened via ELISA. Of these 20, only eight demonstrated IgG binding to rAna o 2 in subsequent ELISAs. We chose to use five of the strongly reactive mAbs for future experiments; 4C3, 5G10, 2B5, 4H9, and 1F5.

Purification of Native Ana o 2 From Whole Cashew Extract Our HPLC size exclusion separation of various proteins in the whole cashew extract showed that peak #30 correlated to the fraction most enriched for the cashew 11S globulin. This finding was confirmed by SDS-PAGE and immunoblotting with both Ana o 1 and Ana o 2- specific mAbs (data not shown).

Dot Blot Screening of Ana o 2 and Gy1 glycinin Full-length, Chimeric, and Subunit Expressing Genes The five selected Ana o 2-specific mAbs were tested for reactivity to a total of eight fusion proteins; Ana o 2 and Gy1 glycinin full-length, Ana o 2 large and small subunits, Gy1 glycinin large and small subunits, and the A2 LG-Gy1 SM and Gy1 LG-A2 SM chimeric genes. The dot blot set-up and results are shown in Figure 4.3. It is clear that all but one of the mAbs tested (2B5) showed specific recognition of the Ana o 2 large subunit in each of the proteins bearing this portion of the cashew 11S globulin (rAna o 2 full, rAna o 2 large subunit, and rA2 LG-Gy1 SM chimera). 2B5, however, recognizes only the full length rAna o 2 fusion protein and the rA2 LG-Gy1 SM chimeric molecule, suggesting binding to a conformational epitope not fully displayed on either Ana o 2 subunits alone or on the Gy1 LG-A2 SM chimeric molecule. The negative control mAb, 3G1, which was specific for walnut 2S albumin, did not show any reactivity to the dotted fusion proteins. None of the mAbs bound to the full-length rGy1 glycinin or it’s constituent subunits, again confirming it’s usefulness as a surrogate in the chimera

experiments. Similarly, none of the mAbs bound to the MBP protein alone, proving that binding to the fusion proteins in the immunoblots was specific for the target protein.

3G1 4C3 4H9 1F5 5G10 2B5 rAna o 2 full length-MBP

rAna o 2 LG subunit-MBP

rAna o 2 SM subunit-MBP

rA2 LG- Gy1 SM-MBP

rGy1 full length-MBP

rGy1 LG subunit-MBP

rGy1 SM subunit-MBP

rGy1 LG- A2 SM-MBP

MBP

Figure 4.3. Monoclonal antibody dot blot immunoblotting of rAna o 2 and rGy1 full-length, subunit, and chimeric fusion proteins. Monoclonal antibody clones are identified atop each strip and the dot blot set-up for each nitrocellulose membrane is depicted to the far right.

Conformational Sensitivity of the Binding of mAb 2B5 to Reactive Fusion and Native Proteins Monoclonal antibodies 2B5 and 1F5 were tested for binding to denatured variants of the reactive Ana o 2 and A2 LG-Gy1 SM recombinant fusion proteins. The mAb 2B5 was chosen because it displayed what we viewed as conformational epitope specificity in our initial dot blot with rAna o 2 and the rA2 LG-Gy1 SM chimera. MAb 1F5 was chosen for the opposite reason;

it bound to the large subunit of Ana o 2 when it was expressed either alone or as a part of the full-length or chimeric fusion protein. This antibody had previously been shown to bind a specific peptide within the large subunit in SPOTs assays and thus 1F5 was deemed a linear epitope binding mAb (data not shown). The dot blot assay, depicted in Figure 4.4, showed that the linear peptide binding of mAb 1F5 was not affected by any of the denaturing conditions. Boiling of the proteins under these same denaturing conditions and subsequent immunoblotting similarly had no effect (data not shown). Binding of 2B5, however, was lessened and in some cases completely eliminated when unboiled samples were treated with varying amounts of different denaturing agents prior to blotting. In fact, boiling alone prior to immunoblotting led to a complete loss of 2B5 binding to even the otherwise untreated rAna o 2 and rA2 LG-Gy1 SM fusion proteins (data not shown).

1F5 2B5 1F5 2B5 1F5 2B5 rAna o 2-MBP rAna o 2-MBP rAna o 2-MBP

rAna o 2-MBP rAna o 2-MBP rAna o 2-MBP (1M β-ME) (10% SDS) (6M Urea)

rAna o 2-MBP rAna o 2-MBP rAna o 2-MBP (500 mM) (1% SDS) (1M Urea)

rAna o 2-MBP rAna o 2-MBP rAna o 2-MBP (250 mM) (0.1% SDS) (100 mM)

rA2 LG-Gy1 rA2 LG-G1 rA2 LG-Gy1 SM-MBP SM-MBP SM-MBP

rA2 LG-Gy1 SM rA2 LG-Gy1 SM rA2 LG-Gy1 SM MBP (1M β-ME) MBP (10% SDS) MBP (6M Urea)

rA2 LG-Gy1 SM rA2 LG-Gy1 SM rA2 LG-Gy1 SM MBP (500 mM) MBP (1% SDS) MBP (1M Urea)

rA2 LG-Gy1 SM rA2 LG-Gy1 SM rA2 LG-Gy1 SM MBP (250 mM) MBP (0.1% SDS) MBP (100 mM) Figure 4.4. Binding of mAbs 1F5 and 2B5 to rAna o 2 and rA2 LG-Gy1 SM chimera under various denaturing conditions. Denaturing conditions are listed to the left of each set of two strips and the mAb used for probing shown directly atop.

In addition to probing the denatured recombinant fusion proteins with our mAbs we also performed dot blots with denatured variants of whole cashew extract and the 11S globulin

enriched HPLC fraction #30. The results, shown in Figure 4.5, correlate well with those of the fusion protein in that binding is lessened when the proteins are probed with mAb 2B5 and no such effect is seen for 1F5 binding (Figure 4.5). Boiling the samples further decreased and, in most cases, eliminated binding of 2B5 to the native cashew proteins but again had no effect on 1F5 binding (data not shown).

1F5 2B5

Whole Cashew

Whole Cashew (500 mm β-ME)

Whole Cashew (10% SDS)

Whole Cashew (6M Urea)

WC HPLC #30

WC HPLC #30 (500 mm β-ME)

WC HPLC #30 (10% SDS)

WC HPLC #30 (6M Urea)

Figure 4.5. Binding of mAbs 1F5 and 2B5 to whole cashew extract and HPLC fraction #30 under various denaturing conditions. Denaturing conditions are listed to the left of each set of two strips and the mAb used for probing shown directly atop.

Mouse mAb (IgG)/Human IgE Inhibition ELISA The ability of our mouse monoclonal IgG Abs to inhibit binding of Ana o 2-reactive IgE from cashew sensitive human subjects was first demonstrated through an inhibition ELISA. We demonstrated that both our linear epitope binding (1F5) and conformational epitope binding (2B5) mAbs could block human-IgE binding sites when used to pre-incubate rAna o 2 ELISA- coated plates prior to probing with the human sera. The averaged results are summarized for two different patients, #’s 9 (4.6.1) and 8 (4.6.2) in Figure 4.6. A slight degree of inhibition (~16%)

is observed in both patients when the 2B5 mAb was used as an inhibitor and to an even larger extent with 1F5. Much less inhibition (~1-3.5%) is observed when a control mAb (3G1, anti- walnut 2S albumin) is used as negative control.

(1)

IgE Binding

(Ads at 495 nm)

Inhibition of Monoclonal Antibodies

(2)

IgE Binding

(Ads at 495 nm)

Inhibition of Monoclonal Antibodies

Figure 4.6. Inhibition of Ana o 2-specific human IgE with mAbs 2B5 and 1F5. 4.6.1, shows the inhibition of IgE from patient #9 with both mAbs. Similar inhibition is observed for patient #8, 4.6.2. Standard error bars are shown for all measurements.

Dot Blot Inhibition of mAb 2B Binding With Human Sera To further prove that mAb 2B5 was binding to an epitope recognized by human IgE, we performed a dot blot inhibition assay. The assay showed that pre-incubation of any of the mAb- reactive proteins (2B5 reactivity to each protein is shown in Figure 4.7.1) with human IgE from patient #’s 9 (4.7.3) or 8 (4.7.4) prior to probing with 2B5 would completely eliminate binding. When sera from a non-cashew sensitive individual was used as a negative control inhibitor, no such inhibition was observed (Figure 4.7.2).

1 2 3 4

WC HPLC #30

rAna o 2- MBP

rA2 LG- G1 SM- MBP

Figure 4.7. Inhibition of mAb 2B5 reactivity with sera from two cashew sensitive patients. 4.7.1, binding of 2B5 to whole cashew HPLC fraction #30, rAna o 2, and rA2 LG-Gy1 SM chimera. The same strip pre-incubated with sera from a non-cashew sensitive patient, 4.7.2; patient # 9, 4.7.3; or patient #8, 4.7.4.

Discussion To help understand the molecular basis of Type I allergic immune responses, identification of conformational IgE-binding epitopes on the surface of allergen molecules is required. Unfortunately, because the technology for identifying linear epitopes is considerably

more straightforward than the characterization of conformational epitopes, rather little attention has been paid to the latter.69, 77 During the last decade, for example, the IgE-binding epitopes of various allergens have been described, as a rule, by overlapping peptide mapping as we described in chapter 3. This method, however, only enables immunological properties to be studied in terms of short peptides (usually 8-15 aa), making it useful only for the identification of linear IgE-binding epitopes. Despite the movement in the field of allergy to focus on the identification of these peptide bearing epitopes, currently available data from crystallographic studies suggest that most B-cell epitopes on proteins are conformational.86 From a practical standpoint, a detailed knowledge of both the linear and conformational epitope profiles of tree nut allergens would best facilitate the engineering of their recombinant counterparts to possess those desirable traits present in the native forms. Such variants might be used in patient testing and therapeutics (i.e., for desensitization and vaccination) because while allergen-specific immunotherapy using authentic IgE-binding allergens is effective in controlling some allergies, the high risk of systemic reactions during therapy with food allergies remains a serious problem.86 Using a common approach for proving the existence of conformational epitopes, Schramm et al84 used Asp f 3 reactive patient sera to probe PCR-generated fragments of various lengths and overlap corresponding to the sequence of this fungus gene. The authors found that sera from certain patients did not recognize any of the smaller overlapping fragments which, together, covered a larger IgE-binding region of the protein, indicating that there is another, possibly discontinuous, epitope on this larger fragment which was destroyed by expressing it as smaller peptides. In our study we set out to use a "gain in function" approach. We wanted to transfer a conformationally sensitive IgE-binding epitope of a known allergen, Ana o 2, to a homologous adaptor protein, Gy1 glycinin, to verify the structural dependence of this binding. To do this, we designed our experiments around the novel chimera-based allergen epitope mapping strategy employed by Karisola et al86 in their identification of the major conformational IgE-binding epitope of hevein (Hev b 6.02). We proved that, with respect to our cashew-reactive mAbs and our Ana o 2 reactive patient sera, the Gy1 glycinin did not bind antibody. Further verification that this protein would serve as a viable surrogate in our chimeric molecule assays was seen when we expressed the large and small subunits of the Gy1 glycinin as individual fusion proteins

and again saw no antibody binding. With a non-reactive homologous adaptor protein in hand, we searched for mAbs with unique and potentially informative binding patterns to Ana o 2. One mAb, 2B5, was of particular interest because unlike the other four mAbs assayed, it alone showed binding to only the full-length rAna o 2 fusion protein and not to either of it’s constituent subunits. This binding pattern suggested that 2B5 was reacting to an epitope that formed as a result of structural changes brought upon by the Ana o 2 large and small subunits being expressed as a full-length protein or when this protein interacted with other heterodimeric monomers to form trimers. Because our recombinant protein was expressed in E. coli, it could not have undergone the post-translational cleavage event necessary for hexamer formation, so any potential binding sites that may have been created from this process were ruled out. Further, we were able to reconstitute this epitope and re-establish 2B5 binding to the previously non- reactive Ana o 2 large and Gy1 small subunits by creating a chimeric molecule of the two that more readily mimicked the native conformation of the rAna o 2. We also proved that in addition to the recombinant precursor protein, 2B5 could bind to the native cashew 11S globulin and that binding in both cases, and the case of the chimeric molecule, was conformationally sensitive. Binding of 2B5 to it’s epitope on the native or recombinant proteins was severely diminished and, in most cases, eliminated by boiling or treating the targets with various denaturants prior to use in dot blot assays. Further experiments need to be done to determine the precise location of the 2B5 epitope. It is possible that the binding site is located within the monomer or at the monomer-to-monomer interface within the trimer. The solved crystal structure of the homologous soybean 11S globulin trimer,88 for example, shows the interaction of helices located near the N-termini of one and C-termini of an adjoining monomer, creating a molecular surface that could be the 2B5 binding site. The importance of this conformational epitope in terms of human-IgE binding was confirmed by two inhibition assays, both of which suggested that 2B5 was binding to an epitope also recognized by IgE from Ana o 2-reactive patient sera. Our inhibition ELISA incorporating the mAb showed a reduced binding of patient IgE but complete inhibition was not observed. Such a finding is not surprising when one considers the polyclonal nature of the human IgE response. To circumvent this problem we reversed the assay and showed that the human IgE could indeed completely inhibit the mAb binding to the target proteins in dot blot experiments. We also recognize the possibility that the inhibitory effect mAb 2B5 has on patient IgE-binding

to Ana o 2 may be a result of steric hindrance and not a direct result of binding to the same epitope or same amino acids within the epitope. Further studies need to be carried out in order to determine if 2B5 and the patient IgE do, as we suggest, bind the same epitope. Together, the above approaches could provide new possibilities for defining IgE-binding conformational allergenic epitopes. This information, coupled with our linear epitope mapping studies, could prove useful in the effort to create specific reagents for tree-nut allergen identification, disease diagnosis, immunotherapy, and vaccine development.

CHAPTER 5

HOMOLOGY MODELING OF VARIOUS 11S GLOBULIN ALLERGENS

Introduction

The preceding experiments demonstrated the presence of both linear (chapter 3) and conformational epitopes (chapter 4) on the cashew 11S globulin allergen, Ana o 2. A recent study by Wallowitz et al75 demonstrated that a great deal of cross-reactivity is observed among legumins of various tree nuts. Cross-reactivity among allergens is of considerable scientific as well as clinical interest.89 Shared structural features of epitopes, which are the most important determinants of cross-reactivity, have yet to be elucidated among the 11S globulin allergens. An analysis of the three-dimensional (3D) structure of an allergen is important in understanding the molecular basis of allergenicity. In the past few years, the primary structures of a large number of different allergens have been reported.72 Recently, homology based modeling of allergens and surface mapping of their IgE-binding epitopes has been used as tools to help define a structural basis for epitope sharing and thus cross-reactivity.74, 89, 90 In this study we model and compare the 3D structures of allergenic legumins from cashew (Ana o 2), walnut (Jug r 4), hazelnut (Cor a 9), and peanut (Ara h 3) and map the antigenic surfaces of their IgE- binding epitopes. In the case of Ana o 2, we have also shown the accessibility of the epitopes on the higher order trimeric structure. The goal of these studies was to compare the structure of the 11S globulin allergen epitopes and define any shared structural features that might provide a molecular basis for cross- reactivity among these proteins. Defining the structural basis for shared and unique epitopes will help identify critical features of IgE binding that may be used to develop mimotopes (a small peptide that mimics the structure of the original epitope) or identify allergen homologues for vaccine development and help screen for potential allergenicity among genetically modified food crops.90 Our findings demonstrate both a molecular and structural basis for allergen cross- reactivity based on epitope sequence homology and conformation conservation, respectively. Such findings suggest that both factors regarding epitope relatedness need be considered in

attempting to predict the potential allergenicity of a new, or cross-reactivity of an existing, protein and identify the candidates usefulness in allergen therapy, vaccine development, and food engineering.

Methods

Sequence Alignment of 11S Globulin Allergens To determine the amount of sequence homology observed between each of the studied 11S globulins, sequence alignments of the Ana o 2, Jug r 4, Cor a 9, Ara h 3, and the A1aB1b glycinin were carried out using the BLAST program as described earlier.

Homology Modeling and Surface Mapping of Ana o 2, Jug r 4, Cor a 9, and Ara h 3 Homology based 3D modeling was performed for Ana o 2, Jug r 4, Cor a 9, and Ara h 3 based on the solved crystal structure of soybean the proglycinin A1aB1b homotrimer (PDB 1FXZ) to determine the spatial arrangement of the legumin IgE epitopes. Models were constructed using SWISS MODEL.91 Briefly, appropriate quaternary structure templates for each protein were identified using a BLAST search of the ExNRL-3D database (PDB). Template and target sequences were aligned using CLUSTAL-X.92 The target sequence and the quaternary structure template were loaded into SwissPDB Viewer (SPDBV)93 and the target sequence was threaded onto the template after manually aligning the sequences in SPDBV based on the Clustal-X data. This PDB file was submitted to the Swiss Model automated protein homology modeling server (http://swissmodel.expasy.org/SWISS-MODEL.html). The resulting structure file was subjected to successive rounds of steepest descent energy minimization using GROMOS 96 (http://www.igc.ethz.ch/gromos/) until delta E between steps was below 0.050 kJ/mol.

Results

Sequence Alignment of 11S Globulin Allergens

Amino acid sequence comparison of the glycinin, cashew, walnut, hazelnut, and peanut legumins revealed a high incidence of identity and similarity among all of the proteins as summarized in Table 5.1.

Table 5.1. Percent identity and similarity of five 11S globulin allergens. Percent identity values are listed first followed by the percent similarity. The number of aa included in each alignment are shown in parentheses.

Glycinin Ana o 2 Jug r 4 Cor a 9 Ara h 3 Glycinin 100% 47/65 (457) 47/66 (470) 45/64 (466) 56/69 (474) Ana o 2 100% 59/74 (442) 54/70 (452) 43/59 (424) Jug r 4 100% 73/83 (506) 46/60 (454) Cor a 9 100% 46/61 (470) Ara h 3 100%

Homology Modeling and Surface Mapping of Ana o 2, Jug r 4, Cor a 9, and Ara h 3 Homology-based models of the tertiary structures of four legumins were generated to predict the 3D location of epitopes on the molecules. While the degree of aa sequence similarity is high among the 11S globulins studied (Table 5.1), the structural similarities of their epitopes required elucidation. In this study, the previously identified IgE-binding epitopes of Ana o 2 (Table 3.2), Jug r 4 (Table 3.3), Cor a 9 (unpublished data), and Ara h 3,13 were localized onto their respective models (Figure 5.1). Interestingly, the epitopes of the allergens were located throughout the exterior of the molecules when modeled as monomers (Figure 5.1) or, in the case of Ana o 2, as a trimer (Figure 5.2). It seemed clear that nearly all of those epitopes recognized strongly by one (red), all (green) or at least weakly by all patient sera pools (orange), were located on what could be considered protruding regions of the molecules.

Figure 5.1. Homology modeling of four allergenic 11S globulins and surface mapping of their respective IgE-binding linear epitopes. Epitopes are highlighted based on the strength and frequency of pooled sera binding. Those bound weakly to moderately and unrecognized by at least one pool of patient sera are highlighted in yellow, those bound with similar strength but by all pools of patient sera, orange, and those bound strongly by at least one pool of patient sera, red. In the cases where epitopes were bound strongly and by all pools of patient sera highlighting was done in green. 5.1.1 and 5.1.2, Ana o 2, N-termini oriented left and right (rotated about Y-axis 180 °) respectively; same orientation applies to Jug r 4 (5.1.3 and 5.1.4), Cor a 9 (5.1.5 and 5.1.6), and Ara h 3 (5.1.7 and 5.1.8).

Ana o 2

(1) (2)

Jug r 4

N N (3) (4)

Cor a 9

(5) (6)

Ara h 3

(7) (8)

IA Face IE Face

Figure 5.2. Surface epitope mapping of the Ana o 2 trimer model. Epitopes are colored as previously described and labeled on the molecules IA (left) and IE (right) faces as described by Adachi et al.88

We also note that no Ana o 2 epitopes are found within a stretch of residues homologous to that of a C-terminal region on the soybean A1aBb1 protein believed to be involved in trimer formation.94 Similarly, the position of most of our epitopes correlated with information regarding potential contact residues involved in hexamer formation.95 Adachi et al95 identified certain aa that contributed to the soybean 11S globulin’s decrease in accessible surface area upon hexamerization, making them potential binding sites between the two trimers and thus unlikely to be available for antibody binding assuming the protein is predominantly expressed in this state. When aligned with the aa sequence of this protein, it appeared as though very few suspected contact residues involved in hexamer formation overlap with those involved in IgE binding to Ana o 2 (data not shown). To determine whether or not our mapped linear epitopes were part of a structural motif common to the 11S globulin allergens, those epitopes which were previously identified as allergenic “hot spots” (Figure 3.6 and summarized in Figure 5.3), based on their positional and sequence overlap in several of the allergens, were chosen for further modeling experiments. Five

of the six hot spots were located on the surface of the models and only one (Figure 5.3, #1) could not be mapped. This epitope (Figure 5.3, #1) resides at the N-terminal end of the molecule that was missing from the electron density map of the template 11S globulin from soybean. Failure of the N-terminal region of this protein to be present in our models was not surprising as Helm et al72 described the same problem in their homology modeling of the soybean G2 11S globulin and explained that this region is one of high mobility on the surface of the molecule.

Hot spot #1 Hot spot #2

Ana o 2: 1 LSVCFLI--LFHGCL 13 Ana o 2: 84 LIYVVQGEGMTGISY 98 Jug r 4: 12 ----···VA····· 22 Jug r 4: 95 ··A·R···LF 109 Cor a 9: 10 -··L·--····· 21 Cor a 9: 97 ···E·R···L 111 Ara h 3: ------Ara h 3: 71 E·Q··R·YF·I 85

Hot spot #3 Hot spot #4

Ana o 2: 184 NPKDVF-Q-QQ-QQHQSR- 195 Ana o 2: 199 LFSGFDTELLAEAFQV 214 Jug r 4: 194 ··D·E·-/·-··R·Q·/ 229 Jug r 4: 233 ·····AF····N· 248 Cor a 9: 194 ··D·EH-·-·/···GE- 236 Cor a 9: 240 ·····A·F····N· 255 Ara h 3: 181 ·TQE·/Y-··-·SR···/ 237 Ara h 3: 242 ····TP·F·E···· 257

Hot spot #5 Hot spot #6

Ana o 2: 240 RVIRP--S-RS-QSE-RGS-ESE 256 Ana o 2: 421 LANAFQISREDARKIKFNN 438 Jug r 4: 277 ····RW·-·E-Q·-·EE/·R· 297 Jug r 4: 465 ··T····P········R 482 Cor a 9: 284 ···--E-··-Q·/ER-··· 303 Cor a 9: 470 ··············R 487 Ara h 3: 286 ·S·--D-·K/E·-EYD-·D· 306 Ara h 3: 472 ··GQ··Q···N·· 489

Figure 5.3. Sequence alignment of six identified 11S globulin allergenic “hot spots”. Primary sequence alignment of the “hot spots” from Ana o 2, Jug r 4, Cor a 9, and Ara h 3, as described in chapter 3 and Figure 3.6. Segments indicating regions of allergens expressing one or more IgE-binding peptides are boxed.

Figure 5.4. Surface mapping of five identified 11S globulin allergenic “hot spots”. Surface rendering homology models of Ana o 2, Jug r 4, Cor a 9, and Ara h 3. The five allergenic “hot spots” present on the models are mapped onto the surface in light blue and numbered according to Figure 5.3. 5.4.1 and 5.4.2, Ana o 2, N-termini oriented left and right respectively (rotated about Y-axis 180 °); this same orientation applies to Jug r 4 (5.4.3 and 5.4.4), Cor a 9 (5.4.5 and 5.4.6), and Ara h 3 (5.4.7 and 5.4.8). Arrows on models indicate general surface area covered by the epitope.

5 5

(2) Ana o 2 6 6 2

N N 3 3 (2) (1)

6 5 6

2 2

N (4) N Jug r 4 4 3 (3) 3

6 5 6 2 2 4 N 3 N 3 4 Cor a 9 5 5 (6) (5)

5 5

4 4 N N Ara h 3 (8) (7) 4 4 N N Ara h 3 (7) (8)

With regard to those hot spots that could be mapped on the surface of the proteins, it appears as though several structural similarities and differences can be observed (Figure 5.4). Hot spot #6 (Figure 5.4), for example, is comprised of a helix-loop-helix component common to each of the four allergenic legumins and HS#5 appears to be quite distinct, structurally, in most of the allergens. HS#2 also showed structural similarity among Jug r 4 and Cor a 9 and HS#3 among Ana o 2 and Jug r 4. It was also interesting to note that despite the structural similarity of HS#6 on the peanut legumin, Ara h 3, with HS#6 of the other tree nuts, no IgE binding to this region of the protein was identified in epitope mapping experiments.13 HS#4 of the peanut, however, did show structural similarity to the same region of Jug r 4 and Cor a 9.

Discussion

IgE cross-reactivity among tree nut extracts has been previously reported but the offending proteins and underlying basis for this cross-reactivity has not been fully explained.96, 97 Recently we identified an IgE-binding epitope on the cashew 2S albumin58 which shared sequence homology and positional overlap with the lone linear epitope of the walnut 2S albumin, Jug r 1,69 a finding which could explain the basis for clinical cross-reactivity to these two tree nuts. It is important to note that even though past experiments have shown a basis for allergen cross-reactivity involving aa and epitope sequence homology98 it has recently been shown that sequence homology alone was a poor predictive value for cross-reactivity among some allergens.99 Human-IgE reactivity to melon profilin, for example, strongly depends on the highly conserved conformational structure, rather than a high degree of aa sequence identity.99 That is not to say that structure alone is to be viewed as the sole factor involved in determining the potential cross-reactivity of epitopes in closely related allergenic proteins from different species. For example, the soybean glycinin and that of peanut (Ara h 3) share IgE-binding regions, however alanine scanning of the peptides indicates that the structures recognized by IgE antibodies are probably different in these two proteins.90 Based on our findings and those of Barre et al74 it appears as though epitope cross-reactivity is a function both of a proteins sequence and structure. Through homology modeling Barre et al74 mapped previously identified linear epitopes on the surfaces of Ana o 3, Jug r 1, and other related 2S albumin proteins. In

doing so, they identified a structural basis for the observed cross-reactivity common to this family of proteins.74 These findings coincide with the identified shared sequence homology and positional overlap we previously observed. Recent studies show that in addition to 2S albumins, the11S globulins of these nuts are also likely to be involved in tree nut cross- reactivity.75 To date, however, no shared linear epitope or structural motif has been identified as a candidate for the basis of this discovery. To help solve this problem, we have identified the linear IgE-binding epitopes of several tree nut legumins and through homology modeling mapped the epitopes of these allergens to their respective protein surface. As explained in chapter 3, epitope mapping studies of Ana o 2, Jug r 4, and Cor a 9 identified several reactive epitopes with varying degrees of IgE reactivity. It was not surprising that homology modeling and surface mapping of the allergen epitopes revealed that most of those demonstrating strong IgE-binding were found on highly accessible and somewhat protruding regions of the molecules. All of the weakly reactive epitopes appeared to be dispersed somewhat randomly across the surface of the monomeric molecules (Figure 5.1) and regardless of the intensity to which they reacted with IgE, epitopes were localized to both ‘faces’ of Ana o 2 when it was viewed as a trimer (Figure 5.2). This finding was not consistent with data accumulated regarding the dispersion of epitopes around the G2 glycinin.72 Epitope mapping of this molecule followed by homology modeling experiments demonstrated that, as with our proteins, most epitopes were located on the molecular surface, but were also localized to one face of the trimer.72 The authors suggest this would be a reasonable conclusion since when the glycinin trimers stack and form hexamers they do so in a face-to-face pattern, leaving one face of each trimer readily exposed to the surrounding solvent and the other largely occluded by the binding face of the other trimer. We observe that while this may be true, there is actually a shift in the mobile disordered region of the contacting face of the trimer which can expose some of the ~70 presumed contact residues to the side of the protein before hexamerization occurs.95 Therefore, residues shown to be on one face of the trimer may not be in that position when the hexamer is formed and could therefore be available for antibody binding. Also, it is known that while most of the 11S globulins do get processed into trimers and eventually hexamers, these are only the typical events and through misfolding and failed cleavage events there remains a small percentage of the unprocessed preprotein in monomeric and trimeric

configurations. The presence of the molecules in either such state would readily allow antibody binding to areas of the protein free of the constraint that would otherwise be present following the formation of higher ordered structures. Sequence alignment of the allergenic 11S globulins from cashew, walnut, hazelnut, and peanut revealed several frequently reactive homologous stretches of amino acids. We termed these IgE-binding epitopes, ‘hot spots’, and mapped five of the six conserved antigenic surfaces onto our models (Figure 5.4). It appears that while not all the hot spots share a similar structure and thus could not be considered suitable for explaining cross-reactivity among these proteins, HS #6 does appear to be structurally homologous, in terms of 3D conformation, in all of the studied 11S globulins. Additionally, when comparing only the Jug r 4 and Cor a 9 models, HS #’s 4 and 5 also appear to be structurally similar. Analysis of the model of the Ara h 3 hot spots revealed that, in most cases, these regions differed by their 3D conformation from the corresponding regions of the tree nut legumins, perhaps explaining the low incidence of reported tree-nut cross-reactivity, especially between cashew and peanut. A potential basis for cross- reactivity among walnuts, hazelnuts, and peanuts could be HS#4 which was the lone peanut hot spot to show structural similarity to any of the nut allergens. It has long been known that Ab-mediated cross-reactivity is, at least in part, dependent on shared structural features and that conserved aa may be involved in the binding of IgE.89 Moreover, these Ab-binding sites should be exposed on the surface of the protein. Here we have identified a structural basis for the observed cross-reactivity common to the 11S globulin family of proteins in tree nuts. This data supports and in some regards could serve as an explanation for the findings of Wallowitz et al75 regarding the observed IgE-cross-reactivity among tree nut legumins. Additionally, our work coincides with previous observations that a shared sequence homology is present among tree nut 2S albumin epitopes and, according to Barre et al,74 there is also a structural basis for such a finding. Defining the structural basis for shared and unique epitopes will help to identify critical features of IgE epitopes that can be used to develop better assays for initial allergen identification, potential cross-reactivity among related proteins, and even allergen homologues for vaccine development.90

CONCLUSIONS

This study set out to help address a growing health concern; that of the rising incidence of tree nut allergy in the US. It is now believed that approximately 1% of our population is allergic to peanuts or tree nuts,10 a statistic that has doubled since 1997.3 We planned to identify, biochemically and clinically characterize, and map IgE-binding epitopes of allergenic proteins belonging to the 2S albumin and 11S globulin families of the English Walnut (Juglans regia) and cashew nut (Anacardium occidentale). Our efforts were aimed at contributing to a new area of commercial tree nut allergen identification and, more broadly, food allergy diagnosis and treatment, involving recombinant proteins. In the first set of experiments we helped to identify and/or characterize major 2S albumin and 11S globulin allergens in walnut, Jug r 1 and Jug r 4, respectively, and cashew, Ana o 2 and Ana o 3. By demonstrating that our recombinant proteins mimicked the IgE-binding reactivity of their native counterparts in crude extracts, we provided evidence to support the idea that recombinant allergens can serve as suitable surrogates in diagnostic and perhaps even therapeutic advances. Further, through our linear-IgE epitope mapping studies (chapter 3) we were able to create composite epitope maps of those 2S albumin and 11S globulin proteins studied. These maps allowed us to compare the cognate sequences of the homologous allergens and identify IgE-binding regions commonly recognized in each. The identification of epitope homology among these broadly allergenic proteins, in part, could serve as a molecular basis for some incidences of tree nut cross-reactivity. In the case of Jug r 1 and Ana o 2, we even identified aa both influential or necessary for patient-IgE binding to occur and demonstrated the possibility of genetically re-engineering the genes encoding the offending proteins to eliminate the aa that helped induce allergic reactions and enhance the features that provide protective (non- or anti- allergenic) immunity.4, 24-27 In this regard, our efforts may lead to advances in the diagnosis, treatment, and, particularly, the prevention of the development of allergy, benefiting not only those involved in the field from a clinical standpoint but the agricultural and food communities which could see a lessening in the regulatory pressure on contamination and avoidance. With recent experiments providing proof that aa sequence homology alone cannot predict cross-reactivity among homologous allergens99 we also made an effort to elucidate any shared structural motifs that might be present when the IgE-binding epitopes were mapped on the

surface of 11S globulins. This strategy was based on the success of Barre et al74 who found that the lone linear-IgE binding epitope of Jug r 169 was structurally related to epitopes of other phylogenetically related 2S albumin proteins. In doing so, they identified a structural basis for the observed cross-reactivity common to this family of proteins74 thereby supporting the shared sequence homology and positional overlap we previously observed.58 In line with these results we identified a structural basis for the observed cross-reactivity common to the 11S globulin family of proteins in tree nuts. The approaches we used in chapters 3 through 5 demonstrate the value and limitations of using aa sequence identity and 3D homology modeling to define new epitopes and to characterize the structural basis of epitope sharing. We also demonstrated that even though technology for identifying conformational allergenic epitopes is less direct than the techniques used for linear-IgE epitope identification, this is not an area that can be ignored. Conformational epitopes are an important part of tree nut IgE-binding profiles, as we demonstrated for the cashew legumin, Ana o 2. Taken together, our experiments represent the most comprehensive study regarding tree nut allergen epitope mapping to date. The identification of shared and unique epitopes, be they linear or conformational, may have both fundamental and clinical value. An understanding of the structural requirements for the IgE responses and reactions should provide a basis for determining the allergenic potential of previously unstudied proteins, helping to prevent their introduction into genetically modified foods or reducing the extent of contact by humans. In addition, development of new generations of immunological reagents for allergy testing and specific immunotherapy could be expedited if well-characterized reagents could be used in groups of patients who are allergic to homologous allergens.90

REFERENCES

1. Cookson W. "The Alliance of Genes and Environment in Asthma and Allergy." Nature 402 (1999): B5.

2. Holgate ST. "The Epidemic of Allergy and Asthma." Nature 402 (1999): B2-B4.

3. Sicherer, S. H., A. Munoz-Furlong, A. W. Burks, and H. A. Sampson. "Prevalence of Peanut and Tree Nut Allergy in the US Determined by a Random Digit Dial Telephone Survey." J Allergy Clin Immunol 103 (1999): 559-62.

4. Valenta, R., S. Vrtala, M. Focke-Tejkl, A. Bugajska-Schretter, T. Ball, A. Twardosz, S. Spitzauer, H. Gronlund, and D. Kraft. "Genetically Engineered and Synthetic Allergen Derivatives: Candidates for Vaccination Against Type I Allergy." Biol Chem 380 (1999): 815-24.

5. Yunginger J.W. "Food Antigens. In: Food Allergy: Adverse Reactions to Foods and Food Additives.” DD Metcalfe, HA Sampson, and RA Simon. Cambridge, Massachusetts: Blackwell Science (1997): 49-63

6. Barrett, K. E. "Mast Cells, Basophils, and Immunoglobulin E. In: Food Allergy: Adverse Reactions to Foods and Food Additives.” DD Metcalfe, HA Sampson, and RA Simon. Cambridge, Massachusetts: Blackwell Science (1997): 27-48.

7. Sampson, HA. "Immediate Reactions to Food in Infants and Children. In: Food Allergy: Adverse Reactions to Foods and Food Additives.” DD Metcalfe, HA Sampson, and RA Simon. Cambridge, Massachusetts: Blackwell Science (1997): 169-82.

8. Sampson, H. A., M. D. Dean, and D. D. Metcalfe. "Food Allergies." JAMA 268 (1992): 2840-2844.

9. Dutau G, Rittie KJL, Rance F, Juchet A, and Bremont F. "Les Nouveaux Allergenes Alimentaires." La Presse Medicale 28 (1999): 1553.

10. Sicherer, S. H., A. Munoz-Furlong, and H. A. Sampson. "Prevalence of Peanut and Tree Nut Allergy in the United States Determined by Means of Random Digit Dial Telephone Survey: a 5-Year Follow-Up Study." J Allergy Clin Immunol 112 (2003): 1203-7.

11. Roux, K. H., S. S. Teuber, and S. K. Sathe. "Tree Nut Allergens." Int Arch Allergy Immunol 131 (2003): 234-44.

12. Teuber, S. S., and K. Beyer. "Peanut, Tree Nut and Seed Allergies." Curr Opin Allergy Clin Immunol 4 (2004): 201-3.

13. Rabjohn, P., E. M. Helm, J. S. Stanley, C. M. West, H. A. Sampson, A. W. Burks, and G. A. Bannon. "Molecular Cloning and Epitope Analysis of the Peanut Allergen Ara h 3." J Clin Invest 103 (1999): 535-42.

14. Nordlee, J. A., S. L. Taylor, J. A. Townsend, L. A. Thomas, and R. K. Bush. "Identification of a Brazil-Nut Allergen in Transgenic Soybeans." N Engl J Med. 334 (1996): 688-92.

15. Sathe, S. K., K. H. Roux, W. R. Peteson, and S. S. Teuber. "Characterization of Allergens From Cashew." J Allergy Clin Immunol 103 (1999): S101.

16. Teuber, S. S., S. S. Comstock, S. K. Sathe, and K. H. Roux. "Tree Nut Allergy." Curr Allergy Asthma Rep 3 (2003): 54-61.

17. Etesamifar, M., and B. Wuthrich. "IgE-Vermittelte Nahrungs-Mittelallergien Bei 383 Patienten Unter Beruchsichtigung Des Oralen Allergie-Syndroms. " Allergologie 21 (1998): 451-57.

18. Teuber, S. S., K. C. Jarvis, A. M. Dandekar, W. R. Peterson, and A. A. Ansari. "Identification and Cloning of a Complementary DNA Encoding a Vicilin-Like Proprotein, Jug r 2, From English Walnut Kernel (Juglans Regia), a Major Food Allergen." J Allergy Clin Immunol 104 (1999): 1311-20.

19. Gendel SM. "Assessing the Potential Allergenicity of New Food Proteins." Food Biotechnology 12 (1998): 175-85.

20. Aalberse, RC. "Structural Biology of Allergens." J Allergy Clin Immunol 106 (2000): 228-38.

21. Gendel SM. "Sequence Databases for Assessing the Potential Allergenicity of Proteins Used in Transgenic Foods." Advances in Food and Nutrition Research 42 (1998): 63-92.

22. Osborne, T. B. "The Vegetable Proteins.” 2nd Ed. London: Longsman-Green. (1924).

23. Shewry, P. R., J. A. Jenkins, F. Beaudoin, and E. N. Mills. "The Classification, Functions and Evolutionary Relationships of Plant Proteins in Relation to Food Allergies. In: Plant food allergens.” E. N. Mills, and P. R. Shewry. Malden, MA: Blackwell Publisning (2003): 24-41.

24. Schmid-Grendelmeier, P., and R. Crameri. "Recombinant Allergensd for Skin Testing." Int Arch Allergy Immunol 125 (2001): 96-111.

25. Bhalla, P. L., I. Swoboda, and M. B. Singh. "Reduction in Allergenicity of Grass Pollen by Genetic Engineering." Int Arch Allergy Immunol 124 (2001): 51-54.

26. Vrtala S, Fischer S, Grote M, Vangelista L, Pastore A, Sperr WR, Valent P, Reichelt R, Kraft D, and Valenta R. "Molecular, Immunological, and Structural Characterization of Ph1 p 6, a Major Allergen and P-Particle-Associated Protein From Timothy Grass (Phleum Pratense) Pollen." J Immunol 163 (1999): 5489-96.

27. Ferreira, F, M Wallner, H Breiteneder, A Hartl, J Thalhamer, and C Ebner. "Genetic

Engineering of Allergens: Future Therapeutic Products." Int Arch Allergy Immunol 128 (2002): 171-78.

28. Sicherer, S. H., T. J. Furlong, A. Munoz-Furlong, A. W. Burks, and H. A. Sampson. "A Voluntary Registry for Peanut and Tree Nut Allergy: Characteristics of the First 5149 Registrants." J Allergy Clin Immunol 108 (2001): 128-32.

29. Bock, S. A., A. Munoz-Furlong, and H. A. Sampson. "Fatalities Due to Anaphylactic Reactions to Foods." J Allergy Clin Immunol 107 (2001): 191-93.

30. Shewry, P. R., and M. J. Pandya. "The 2S Albumin Storage Proteins. In: Seed Proteins." (1999): 563-86.

31. Thorpe SC, Kemeny DM Panzani RC McGurl B Lord M. "Allergy to Castor Bean. II. Identification of the Major Allergens in Castor Bean Seeds." J Allergy Clin Immunol 82 (1988): 67-72.

32. Menendez-Arias L, Dominguez J, Moneo I, and Rodriquez R. "Epitope Mapping of the Major Allergen From Yellow Mustard Seeds." Molecular Immunology 27 (1990): 143-50.

33. Gonzalez, De La Pena M. A., R. I. Monsalve, E. Batanero, M. Villalba, and R. Rodriguez. "Expression in Escherichia Coli of Sin a 1, the Major Allergen From Mustard." Eur J Biochem 237 (1996): 827-32.

34. Youle RJ, Huang AH. "Albumin Storage Protein and Allergens in Cottonseeds." J Agric Food Chem 27 (1979): 500-3.

35. Monsalve, R. I., M. A. Gonzalez de la Pena, L. Menendez-Arias, C. Lopez-Otin, M. Villalba, and R. Rodriguez. "Characterization of a New Oriental-Mustard (Brassica Juncea) Allergen, Bra j IE: Detection of an Allergenic Epitope." Biochem J 293 (1993): 625-32.

36. Kelly, J. D., and S. L. Hefle. "2S Methionine-Rich Protein (SSA) From Sunflower Seed Is an IgE-Binding Protein." Allergy 55 (2000): 556-60.

37. Pastorello, E. A., C. Pompei, V. Pravettoni, O. Brenna, L. Farioli, C. Trambaioli, and A. Conti. "Lipid Transfer Proteins and 2S Albumins As Allergens." Allergy 56 (2001): 45-47.

38. Beyer, K., L. Bardina, G. Grishina, and H. A. Sampson. "Identification of Sesame Seed Allergens by 2-Dimensional Proteomics and Edman Sequencing: Seed Storage Proteins As Common Food Allergens." J Allergy Clin Immunol 110 (2002): 154- 59.

39. Gander ES, Holmstroem KO De Paiva GR De Castro LA Carneiro M Grossi de Sa MF. "Isolation, Characterization and Expression of a Gene Coding for a 2S Albumin From Bertholletia Excelsa (Brazil Nut)." Plant Mol Biol 16 (1991): 437-48.

40. Teuber, S. S., A. M. Dandekar, W. R. Peterson, and C. L. Sellers. "Cloning and Sequencing of a Gene Encoding a 2S Albumin Seed Storage Protein Precursor From English Walnut (Juglans Regia) - a Food Allergen." J Allergy Clin Immunol 101 (1998): 807-14.

41. Bannon, GA, M Ling, G Cockrell, and HA Sampson. "Cloning, Expression and Characterization of Two Major Allergens, Jug n 1 and Jug n 2, From the Black Walnut, Juglans Niger." J Allergy Clin Immunol 107 (2001): S140.

42. Poltronieri, P., M. S. Cappello, N. Dohmae, A. Conti, D. Fortunato, E. A. Pastorello, C. Ortolani, and G. Zacheo. "Identification and Characterisation of the IgE-Binding Proteins 2S Albumin and Conglutin Gamma in Almond (Prunus Dulcis) Seeds." Int. Arch Allergy Immunol 128 (2002): 97-104.

43. Teuber, S. S., S. K. Sathe, W. R. Peterson, and K. H. Roux. "Characterization of the Soluble Allergenic Proteins of Cashew Nut (Anacardium Occidentale L.)." Agric Food Chem 50 (2002): 6543-9.

44. Bashir ME, Hubatsch I Leinenbach HP Zeppezauer M Panzani RC Hussein IH. "Ric c 1 and Ric c 3, the Allergenic 2S Albumin Storage Proteins of Ricinus Communis: Complete Primary Structures and Phylogenetic Relationships." Int Arch Allergy Immunol (1998): 73-82.

45. Breiteneder H, Ferreira F Hoffmann-Sommergruber K Ebner C Breitenbach M Rumpold H Kraft D Scheiner O. "Four Recombinant Isoforms of Cor a I, the Major Allergen of Hazel Pollen, Show Different IgE-Binding Properties." Eur J Biochem 212 (1993): 355-62. .

46. de Castro LB, Lacerda Z Aramayo RA Sampaio MM Gander ES. "Evidence for a Precursor Molecule of Brazil Nut 2S Seed Proteins From Biosynthesis and cDNA Analysis." Mol Gen Genet 206 (1987): 338-43.

47. Platts-Mills, T. A., E. A. Erwin, A. B. Allison, K. Blumenthal, M. Barr, D. Sredl, H. Burge, and D. Gold. "The Relevance of Maternal Immune Responses to Inhalant Allergens to Maternal Symptoms, Passive Transfer to the Infant, and Development of Antibodies in the First 2 Years of Life." J Allergy Clin Immunol 111 (2003): 123-30.

48. Clark AT, Ewan PW. "Interpretation of Tests for Nut Allergy in One Thousand Patients, in Relation to Allergy or Tolerance." Clin Exp Allergy 33 (2003): 1041-5.

49. Kleber-Janke, T., R. Crameri, U. Appenzeller, M. Schlaak, and W. Becker. "Selective Cloning of Peanut Allergens, Including Profilin and 2S a l Bumins, by Phage Display Technology." Int Arch Allergy Immunol 119 (1999): 265-74.

50. Zeece, M. G., T. A. Beardslee, J. J. P. Markwell, and G. Sarath. "Identification of an IgE- Binding Region in Soybean Acidic Glycinin G1." Food Agric Immunol. 11

(1999): 83-90.

51. Helm, R. M., G. Cockrell, C. Connaughton, H. A. Sampson, G. A. Bannon, V. Beilinson, D. Livingstone, N. C. Nielsen, and A. W. Burks. "A Soybean G2 Glycinin Allergen. 1. Identification and Characterization." Int Arch Allergy Immunol. 123 (2000): 205-12.

52. Beardslee, T. A., M. G. Zeece, G. Sarath, and J. P. Markwell. "Soybean Glycinin G1 Acidic Chain Shares IgE Epitopes With Peanut Allergen Ara h 3." Int Arch Allergy Immunol 123 (2000): 299-307.

53. Beyer, K., G. Grishina, L. Bardina, A. Grishin, and H. A. Sampson. "Identification of an 11S Globulin As a Major Hazelnut Food Allergen in Hazelnut-Induced Systemic Reactions." J Allergy Clin Immunol 110 (2002): 517-23.

54. Fujino, K., H. Funatsuki, M. Inada, Y. Shimono, and Y. Kikuta. "Expression, Cloning, and Immunological Analysis of Buckwheat (Fagopyrum Esculentum Moench) Seed Storage Proteins." J Agric Food Chem 49 (2001): 1825-29.

55. Yamada, K., A. Urisu, Y. Morita, Y. Kondo, E. Wada, H. Komada, M. Yamada, Y. Inagaki, and S. Torii. "Immediate Hypersensitive Reactions to Buckwheat Ingestion and Cross Allergenicity Between Buckwheat and Rice Antigens in Subjects With High Levels of IgE Antibodies to Buckwheat." Ann Allergy Asthma Immunol. 75 (1995): 56-61.

56. Teuber, SS, WR Peterson, S Uratsu, A Dandekar, KH Roux, and SK Sathe. "Identification and Cloning of Jug r 4, a Major Food Allergen From English Walnut Belonging to the Legumin Group." J Allergy Clin Immunol 111 (2003): S248.

57. Wang, F., J. M. Robotham, S. S. Teuber, P. Tawde, S. K. Sathe, and K. H. Roux. "Ana o 1, a Cashew (Anacardium Occidental) Allergen of the Vicilin Seed Storage Protein Family." J Allergy Clin Immunol. 110 (2002): 160-166.

58. Robotham, J. M., F. Wang, V. Seamon, S. S. Teuber, S. K. Sathe, H. A. Sampson, K. Beyer, M. Seavy, and K. H. Roux. "Ana o 3, an Important Chasew Nut (Anacardium Occidentale L.) Allergen of the 2S Albumin Family." J Allergy Clin Immunol. 115 (2005): 1284-90.

59. Wang, F., J. M. Robotham, S. S. Teuber, S. K. Sathe, and K. H. Roux. "Ana o 2, a Major Cashew Nut Allergen of the Legumin Family." J Allergy Clin IMmunol 111 (2003): S249.

60. Sampson HA. "Anaphylaxis and Emergency Treatment." Pediatrics 111 (2003): 1601-8.

61. Sathe, S. K., K. W. C. Sze-Tao, W. J. Wolf, and B. R. Hamaker. "Biochemical Characterization and in Vitro Digestibility of the Major Globulin in Cashew Nut (Anacardium Occidentale)." J Agric Food Chem 45 (1997): 2854-60.

62. Laemmli UK. "Cleavage of Structural Proteins During the Assembly of the Head of Bacteriophage T4." Nature 227 (1970): 680-685.

63. Tai, S. S., L. S. Wu, E. C. Chen, and J. T. Tzen. "Molecular Cloning of 11S Globulin and 2S Albumin, the Two Major Seed Storage Proteins in Sesame." J Agric Food Chem 47 (1999): 4932-38.

64. Palmer, G. W., D. A. Dibbern Jr., A. W. Burks, G. A. Bannon, S. A. Bock, H. S. Porterfield, R. A. McDermott, and S. C. Dreskin. "Comparative Potency of Ara h 1 and Ara h 2 in Immunochemical and Functional Assays of Allergenicity." Clin Immunol 115 (2005): 302-12.

65. Shewry, P. R. "Seed Proteins. In: Distribution and Some Properties of Seed Globulins.” R Casey. Boston: Kluwer Academic (1999): 159-69.

66. Ghunaim, N., H. Gronlund, M. Kronqvist, R. Gronneberg, L. Soderstrom, S. Ahlstedt, and M. van Hage-Hamsten. "Antibody Profiles and Self-Reported Symptoms to Pollen-Related Food Allergens in Grass Pollen-Allergic Patients From Northern Europe." Allergy 60 (2005): 185-91.

67. Linhart, B., A. Hartl, B. Jahn-Schmid, P. Verdino, W. Keller, M. Krauth, P. Valent, F. Horak, U. Wiedermann, J. Thalhamer, C. Ebner, D. Kraft, and R. Valenta. "A Hybrid Molecule Resembling the Epitope Spectrum of Grass Pollen for Allergy Vaccination." J Allergy Clin Immunol 115 (2005): 1010-1016.

68. Kurup VP, Sussman GL Yeang HY Elms N Breiteneder H Arif SA Kelly KJ Bansal NK Fink JN. "Specific IgE Response to Purified and Recombinant Allergens in Latex Allergy." Clin Mol Allergy 10 (2005): 9-11.

69. Robotham J.M., Teuber S.S., Sathe S.K., and Roux K.H. "Linear IgE Epitope Mapping of the English Walnut (Juglans Regia) Major Food Allergen, Jug r 1." J Allergy Clin Immunol 109 (2002): 143-49.

70. Mills EN, Breiteneder H. "Food Allergy and Its Relevance to Industrial Food Proteins." Biotechnol Adv 23 (2005): 409-14.

71. Wolff, N., U. Cogan, A. Admon, I. Dalal, Y. Katz, N. Hodos, N. Karin, and S. Yannai. "Allergy to Sesame in Humans Is Associated Primarily With IgE Antibody to a 14 KDa 2S Albumin Precursor." Food Chem Toxicol 41 (2003): 1165-74.

72. Helm, R. M., G. Cockrell, C. Connaughton, H. A. Sampson, G. A. Bannon, V. Beilinson, N. C. Nielsen, and A. W. Burks. "A Soybean G2 Glycinin Allergen. 2. Epitope Mapping and Three-Dimensional Modeling." Int Arch Allergy Immunol 123 (2000): 213-19.

73. Beardslee, T. A., M. G. Zeece, G. Sarath, and J. P. Markwell. "Soybean Glycinin G1 Acidic Chain Shares IgE Epitopes With Peanut Allergen Ara h 3." Int Arc Allergy Immunol 123 (2000): 299-307.

74. Barre, A., J. P. Borges, R. Culerrier, and P. Rouge. "Homology Modelling of the Major Peanut Allergen Ara h 2 and Surface Mapping of IgE-Binding Epitopes." Immunol Lett 100 (2005): 153-58.

75. Wallowitz, M. L., S. S. Teuber, K. Beyer, H. A. Sampson, K. H. Roux, S. K. Sathe, F. Wang, and J. M. Robotham. "Cross-Reactivity of Walnut, Cashew, and Hazelnut Legumin Proteins in Tree Nut Allergic Patients." J Allergy Clin Immunol 113 (2003): S156.

76. Eigenmann, P. A., A. W. Burks, G. A. Bannon, and H. A. Sampson. "Identification of Unique Peanut and Soy Allergens in Sera Adsorbed With Cross-Reacting Antibodies." J Allergy Clin Immunol 98 (1996): 969-78.

77. Bredehorst, R., and K. David. "What Establishes a Protein As an Allergen?" Chromatog 756 (2001): 33-40.

78. Berzofsky, J. A. "Intrinsic and Extrinsic Factors in Protein Antigenic Structure." Science 229 (1985): 932-40.

79. Shreffler, W. G., K. Beyer, T. H. Chu, A. W. Burks, and H. A. Sampson. "Microarray Immunoassay: Association of Clinical History, in Vitro IgE Function, and Heterogencity of Allergenic Peanut Epitopes." J Allergy Clin Immunol 113 (2004): 776-82.

80. Nowak-Wegrzyn, A. "Future Approaches to Food Allergy." Pediatrics 111 (2003): 1672- 80.

81. Van Regenmortel, M. H. V. "Mapping Epitope Structure and Activity: From One- Dimensional Prediction to Four-Dimensional Description of Antigenic Specificity." Methods 9 (1996): 465-72.

82. Cooke, S. K., and H. A. Sampson. "Allergenic Properties of Ovomucoid in Man." J Immunol 159 (1997): 2026-32.

83. Vila, L., K. Beyer, K. M. Jarvinen, P. Chatchatee, L. Bardina, and H. A. Sampson. "Role of Conformational and Linear Epitopes in the Achievement of Tolerance in Cow's Milk Allergy." Clin Exp Allergy 31 (2001): 1599-606.

84. Schramm G, Bufe A Petersen A Haas H Merget R Schlaak M Becker WM. "Discontinuous IgE-Binding Epitopes Contain Multiple Continuous Epitope Regions: Results of an Epitope Mapping on Recombinant Hol l 5, a Major Allergen From Velvet Grass Pollen." Clin Exp Allergy 31 (2001): 331-41.

85. Ramachandran H, Jayaraman V Banerjee B Greenberger PA Kelly KJ Fink JN Kurup VP. "IgE Binding Conformational Epitopes of Asp f 3, a Major Allergen of Aspergillus Fumigatus." Clin Immunol 103 (2002): 324-33.

86. Karisola, P., H. Alenius, J. Mikkola, N. Kalkkinen, J. Helin, O. T. Pentikainen, S. Repo,

T. Reunala, K. Turianmaa, M. S. Johnson, T. Palosuo, and M. S. Kulomaa. "The Major Conformational -IgE-Binding Epitopes of Hevein (Hev B6.02) Are Identified by a Novel Chimera-Based Allergen Epitope Mapping Strategy." J Biol Chem 277 (2002): 22656-61.

87. Roux, K. H., S. S. Teuber, J. M. Robotham, and S. K. Sathe. "Detection and Stability of the Major Almond Allergen in Foods." J Agric Food Chem 49 (2001): 2131-36.

88. Akari, H., T. Fukumori, and A. Adachi. "Cell-Dependent Requirement of Human Immunodeficiency Virus Type 1 Gp41 Cytoplasmic Tail for Env Incorporation into Virions." J Virol 74 (2000): 4891-93.

89. Roy D, Ghosh D Gupta-Bhattacharya S. "Homology Modeling of Allergenic Cyclophilins: IgE-Binding Site and Structural Basis of Cross-Reactivity." Biochem Biophys Res Commun 307 (2003): 422-29.

90. Midoro-Horiuti T, Schein CH Mathura V Braun W Czerwinski EW Togawa A Kondo Y Oka T Watanabe M Goldblum RM. "Structural Basis for Epitope Sharing Between Group 1 Allergens of Cedar Pollen." Mol Immunol 43 (2005): 509-18.

91. Schwede, T., J. Kopp, N. Guex, and M. C. Peitsch. "Swiss-Model: an Automated Protein Homology-Modeling Server." Nucleic Acids Res 31 (2003): 3381-85.

92. Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin, and D. G. Higgins. "The Clustal X Windows Interface: Flexible Strategies for Multiple Sequence Alignment Aided by Quality Analysis Tools." Nucleic Acids Res 25 (1997): 4876- 82.

93. Guex, N, and MC Peitsch. "Swiss-Model and the Swiss-PdbViewer: an Environment for Comparative Protein Modeling." Electrophoresis 18 (1997): 2714-23.

94. Adachi, M., Y. Takenaka, A. B. Gidamis, B. Mikami, and S. Utsumi. "Crystal Structure of Soybean Proglycinin A1aB1b Homotrimer." J Mol Biol 305 (2001): 291-305.

95. Adachi, M., J. Kanamori, T. Masuda, K. Yagasaki, K. Kitamura, B. Mikami, and S. Utsumi. "Crystal Structure of Soybean 11S Globulin: Glycinin A3B4 Homohexamer." Proc Natl Acad Sci 100 (2003): 7395-400.

96. Asero, R., G. Mistrello, D. Roncarolo, and S. Amato. "Walnut-Induced Anaphylaxis With Cross-Reactivity to Hazelnut and Brazil Nut." J Allergy Clin Immunol (2004): 358-60.

97. de Leon, M. P., I. N. Glaspole, A. C. Drew, J. M. Rolland, R. E. O'Hehir, and C. Suphioglu. "Immunological Analysis of Allergenic Cross-Reactivity Between Peanut and Tree Nuts." Clin Exp Allergy 33 (2003): 1273-80.

98. Ayuso, R., G. Reese, S. Leong-Kee, M. Plante, and S. B. Lehrer. "Molecular Basis of Arthropod Cross-Reactivity: IgE-Binding Cross-Reactive Epitopes of Shrimp,

House Dust Mite and Cockroach Tropomyosins." Int. Arch Allergy Immunol 129 (2002): 38-48.

99. Sankian, M., A. Varasteh, N. Pazouki, and M. Mahmoudi. "Sequence Homology: a Poor Predictive Value for Profilins Cross-Reactivity." Clin Mol Allergy 3 (2005).

BIOGRAPHICAL SKETCH

Personal

Jason Michael Robotham Born July 13, 1976 in Flushing, N.Y.

Education

April, 2006 Ph.D. in Biological Science Florida State University, Tallahassee, FL

May, 1998 B.S. Biology, Cum Laude, University of Scranton, Scranton, PA

June, 1994 William Floyd High School, Mastic Beach, NY

Professional Experience

1999-2005 Graduate Research Assistant Department of Biological Science, Florida State University

Publications

Robotham JM, Tawde P, Hoffman GG, Teuber SS, Sathe SK, Roux KH. Identification of a human-IgE binding conformational epitope on the cashew nut (Anacardium occidentale L.) important allergen, Ana o 2 (in preparation).

Robotham JM, Hoffman GG, Teuber SS, Beyer K, Sampson H, Sathe SK, Roux KH. Linear IgE-epitope mapping and homology modeling of the 11S globulins from English walnut (Juglans regia), Jug r 4, and Hazelnut, Cor a 9 (in preparation).

Robotham JM, Wang F, Seamon V, Teuber SS, Sathe SK, Sampson H, Beyer K, Seavy M, Roux KH. Ana o 3, an important cashew nut (Anacardium occidentale L.) allergen of the 2S albumin family. J Allergy Clin Immunol. 115:1284-90, 2005

*Wang F, *Robotham JM, Teuber SS, Sathe SK, Roux KH. Ana o 2, a major cashew nut (Anacardium occidentale L.) allergen of the legumin family. Int Arch Allergy Immunol. 132:27-39, 2003 (*both authors contributed equally to this work)

#Wang F, Robotham JM, Teuber SS, Tawde P, Sathe SK, Roux KH. Ana o 1, a cashew (Anacardium occidentale) allergen of the vicilin seed storage protein family. J Allergy Clin Immunol. 110:160-6, 2002 (# This work singled out for positive editorial comment)

Robotham JM, Teuber SS, Sathe SK, Roux KH. Linear IgE epitope mapping of the English walnut (Juglans regia) major food allergen, Jug r 1. J Allergy Clin Immunol. 109:143-9, 2002

Roux KH, Teuber SS, Robotham JM, Sathe SK. Detection and stability of the major almond allergen in foods. J Agric Food Chem. 49:2131-6, 2001.

Selected Presentations

"Linear IgE Epitope Mapping of Various Tree Nut and Legume 11S globulins, " 12th International Congress of Immunology and 4th Annual Conference of FOCIS, Montreal, Canada. 2004

"Ana o 2, a Major Cashew Nut Allergen of the Legumin Family," American Academy of Allergy, Asthma and Immunology, 59th Annual Meeting, Denver, CO. 2003

"Linear IgE epitope mapping of the English walnut (Juglans regia) major food allergen, Jug r 1," American Academy of Allergy, Asthma and Immunology, 57th Annual Meeting, New Orleans, LA. 2001

Abstracts

"Linear IgE Epitope Mapping of Various Tree Nut and Legume 11S globulins, " 12th International Congress of Immunology and 4th Annual Conference of FOCIS, Montreal, Canada. 2004.

"Hypoallergenic Variants of the Legumin, Ana o 2, a Major Cashew Allergen, " 12th International Congress of Immunology and 4th Annual Conference of FOCIS, Montreal, Canada. 2004.

"Ana o 3, a Major Cashew Nut Allergen of the 2S Albumin Family," American Academy of Allergy, Asthma and Immunology, 60th Annual Meeting, San Francisco, CA. 2004.

"Cross-reactivity of Walnut, Cashew, and Hazelnut Legumin Proteins in Tree Nut Allergic Patients," American Academy of Allergy, Asthma and Immunology, 60th Annual Meeting, San Francisco, CA. 2004.

"Mutational Analysis of the Cashew Vicilin Allergen, Ana o 2," American Academy of Allergy, Asthma and Immunology, 60th Annual Meeting, San Francisco, CA. 2004.

"Ana o 2, a Major Cashew Nut Allergen of the Legumin Family," American Academy of Allergy, Asthma and Immunology, 59th Annual Meeting, Denver, CO. 2003.

"Linear IgE epitope mapping of the English walnut (Juglans regia) major food allergen, Jug r 1," American Academy of Allergy, Asthma and Immunology, 57th Annual Meeting, New Orleans, LA. 2001.

Awards and Honors

2004-2005 Dissertation Research Grant Award

2003-2004 Margaret Menzel Endowed Award for Outstanding Performance as a Graduate Student

2002-2003 Florida State University Graduate Student Publication Award

2000-2002 NSF Fellowship in Macromolecular Assemblies

1999-2000 NSF Pilot Fellowship in Macromolecular Assemblies