Effect of food matrix on amandin, almond (Prunus dulcis L.) major protein, immunorecognition and recovery

LWT - Food Science and Technology 43 (2010) 675–683

Contents lists available at ScienceDirect

LWT - Food Science and Technology journal homepage: www.elsevier.com/locate/lwt

Effect of food matrix on amandin, almond (Prunus dulcis L.) major protein, immunorecognition and recovery Rashmi S. Tiwari a,1, 3, Mahesh Venkatachalam a, 2, 3, Girdhari M. Sharma a, 3, Mengna Su a, Kenneth H. Roux b, Shridhar K. Sathe a, * a b

Department of Nutrition, Food and Exercise Sciences, Florida State University, Tallahassee, FL 32306, USA Department of Biological Science, Florida State University, Tallahassee, FL 32306, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 29 August 2009 Received in revised form 26 November 2009 Accepted 28 November 2009

Amandin, the primary storage protein in almonds, contains key polypeptides recognized by almondallergic patients. A variety of food matrices representing diverse categories of foods were analyzed to assess the effect of food matrix on amandin recognition and recovery using rabbit polyclonal antibody based immunoassays. Food matrices from dairy, nuts, and vegetables typically resulted in over-estimation of amandin. Some foods representing legumes and cereals resulted in over-estimation while others in under-estimation of amandin. The amandin recovery range was 116–198 mg/100mg (dairy) 110–292 mg/100mg (tree nuts), 43–304 mg/100mg (legumes), 106–183 mg/100mg (most cereals- with the exception of barley, whole-wheat flour, wild rice and raisin bran whole mix). Amandin recovery from spices was typically low (2–85 mg/100mg) with a few exceptions where higher recoveries were observed (121–334 mg/100mg). Salt (black and white), tea, confectionery (sugar, cocoa, dark chocolate), and fruits (1–83 mg/100mg) generally resulted in lower recoveries. Tested food matrices did not adversely affect amandin immunorecognition in Western blots. The pH and the extraction buffer type affected amandin recovery. The results suggest that food matrix effects as well as extraction conditions need to be carefully evaluated when developing immunoassays for amandin detection and quantification. Ó 2009 Elsevier Ltd. All rights reserved.

Keywords: Almond Amandin Food matrix Polyclonal antibody ELISA

1. Introduction Globally, the US is the largest producer (50.35% of 2007 global production) and exporter (80.41% of the 2006 global export value) of shelled almonds (FAO Stats, 2007). Valued for their nutritional quality and sensory properties, almonds are eaten as snacks and are extensively used in confectionery and baked goods. Tree nuts are one of the eight food groups that account for the majority of food induced allergies and sensitive individuals are therefore susceptible ˜ ozto experiencing adverse reactions (Angus, 1998; Sicherer, Mun Furlong, & Sampson, 2003). During 2001–2006, of the 31 confirmed fatalities due to food allergies, 17 were attributed to peanuts and 8 ˜ oz-Furlong & Sampson, 2007). At least one to tree nuts (Bock, Mun * Corresponding author. 402 Sandels Building, Department of Nutrition, Food and Exercise Sciences, College of Human Sciences, Florida State University, Tallahassee, FL 32306-1493, USA. Tel.: þ1 850 644 5837; fax: þ1 850 645 5000. E-mail address: [email protected] (S.K. Sathe). 1 Present Address: PepsiCo R and D, 100 E Stevens Avenue, Valhalla, NY 10595, USA. 2 Present Address: The Hershey Company, Technical Center, 1025 Reese Ave, Hershey, PA 17033, USA. 3 RST, MV, and GMS equally contributed in all phases of the work reported in this manuscript. 0023-6438/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.lwt.2009.11.012

out of these 8 deaths, that of a 29 year old male, was directly attributed to almond exposure upon candy consumption in an office, despite the ready accessibility of epinephrine. The Food Allergen Labeling and Consumer Protection Act (FALCPA) became effective in the US in 2006 (http://www.cfsan.fda.gov/wdms/ alergact.html) and was designed to protect consumer from unintended exposure. FALCPA specifically requires declaration of the source of ingredients derived from common allergenic foods (milk, eggs, fish, crustacean shellfish, peanuts, soybeans, tree nuts, and wheat). However, the finding that ‘‘consumers with food allergy are increasingly ignoring advisory labeling’’ (Hefle, Furlong, Niemann, Lemon-Mule, Sicherer, & Taylor, 2007) is troubling. In the absence of a cure for treatment of food allergies, avoidance of the offending food, even if present at mg/g levels, therefore remains the prudent choice for sensitive individuals. Bargman, Rupnow and Taylor (1992) using patient sera IgE and Western blotting identified several almond polypeptides recognized by the patient IgE. Sathe (1993) reported almond proteins to be highly soluble in aqueous buffers and that a single globulin, AMP or amandin, dominated the almond protein composition of all the major marketed varieties of almonds in the USA. Sathe, Wolf, Roux, Teuber, Venkatachalam, and Sze-Tao (2002) developed a column

676

R.S. Tiwari et al. / LWT - Food Science and Technology 43 (2010) 675–683

1:10 w/v). The defatted samples were dried for 24 h under a fume hood at room temperature (RT, 25  C) and further homogenized in an Osterizer blender to obtain a uniform powder (w400 mm). The powdered samples were stored in airtight plastic containers at 20  C until further use.

chromatographic procedure to isolate and purify the major storage protein in almonds. Investigations by Wolf and Sathe (1998) using ultracentrifuge demonstrated that amandin accounted for w65 g/ 100 g of the soluble seed proteins. Roux, Teuber, Robotham, and Sathe (2001) demonstrated that amandin contained the key reactive polypeptides recognized by the patient sera IgE. For these and several other reasons, amandin was targeted as a useful marker protein to develop rabbit polyclonal antibodies (pAbs) with the intent of developing a sensitive immunoassay for detection of trace amounts of almonds. Development of a rabbit pAb-based inhibition Enzyme Linked Immunosorbent Assay (ELISA) with a sensitivity of 20 ng amandin/ml ensued (Acosta, Roux, Teuber, & Sathe, 1999). Further studies on molecular properties of amandin (Sathe et al., 2002), global applicability of amandin as a marker protein, regardless of almond cultivar/hybrid (Sathe, Teuber, Gradziel, & Roux, 2001), and the stability of amandin towards various food processing methods (Roux et al., 2001; Venkatachalam, Teuber, Roux, & Sathe, 2002; Su, Venkatachalam, Teuber, Roux & Sathe, 2004) suggested that amandin is a globally applicable marker protein that may be used as a target for the purpose of detecting trace quantities of almonds. Though useful, these studies did not evaluate possible interference from a variety of foods and food ingredients in the pAb-based ELISA targeting detection of trace amounts of amandin. Scheibe, Weiss, Rue¨ff, Przybilla, and Go¨rg (2001) investigated the effect of chocolate on almond protein detection and found that trace amounts (0.5 mg/g) of almonds could be detected using rabbit polyclonal sera and Western blotting. The rabbit polyclonal sera used in these experiments recognized both almond and hazelnut proteins. The authors stated that despite the observed cross reactivity, both protein sources could be easily distinguished based on polypeptide patterns. More recently, a competitive indirect ELISA capable of detecting peanut, hazelnut, almond, cashew and Brazil nuts in chocolate in a single run at <1 mg/g protein with good specificity has also been reported (Rejeb, Abbott, Davies, Cleroux, & Delahaut, 2005). Almonds are often used in a variety of foods, other than chocolate, and may also be present in a myriad of foods and food ingredients as contaminants in trace amounts. Consequently, sensitive individuals may be inadvertently exposed to trace amounts of undeclared almonds. Evaluating the potential influence of several food matrices and select environmental conditions on amandin immunorecognition by rabbit pAbs was therefore of interest.

Competitive inhibition ELISA assays were performed as described previously (Acosta et al., 1999; Roux et al., 2001). Amandin Recovery (mg/100 mg) ¼ [(amandin in spiked sample  amandin in the corresponding unspiked control)  100]/amandin used for spiking

2. Materials and methods

2.6. Electrophoresis and immunoblotting

The foods/food ingredients used in the study were purchased from the local markets. Certain foods/ingredients were processed in the laboratory. Micro titer 96 well ELISA plates were from Costar (Cambridge, MA). BSA (bovine serum albumin), alkaline phosphatase labeled goat anti-rabbit IgG, phosphatase substrate [p-nitro phenyl phosphate, disodium (PNPP)], were from Sigma Chemical Co., St. Louis, MO. Electrophoresis chemicals and supplies were from sources described earlier (Sathe, 1993). Nitrocellulose and blotting papers were from Schleicher and Schuell, Inc. (Keene, NH). All other chemicals and supplies were of reagent or better grade and were purchased from Fisher Scientific, Pittsburgh, PA. Production of anti-amandin rabbit pAbs (6235) and the rabbit perimmune serum has been described earlier (Acosta et al., 1999).

SDS-PAGE in the presence of b-mercaptoethanol (b-ME, 20 ml/L) was carried out according to the method of Fling and Gregerson (1986) and immunoblotting was done as described earlier (Acosta et al.,1999). The antibody dilutions, all diluted in Tris-buffered saline with Tween 20 (TBS-T; 10 mmol/L Tris, 0.9 g NaCl/100 ml, 0.05 ml Tween 20/100 ml, pH 7.6), used were: anti-amandin rabbit pAb ¼ 1:10,000 (v/v), and rabbit preimmune serum ¼ 1:10,000 (v/ v). The blots were incubated at RT for 1 h with horseradish peroxidase (HRP)-labeled goat anti-rabbit IgG (1:40,000v/v) diluted in TBS-T (the secondary antibody). The reactive bands were visualized using chemiluminescent substrates (ECL Plus, Amersham Pharmacia Biotech, Piscataway, NJ) as described by the manufacturer.

2.2. Amandin preparation Amandin was prepared using isoelectric precipitation (pH 5.0) method as previously described (Sathe et al., 2002). Lyophilized amandin was stored in an airtight plastic bottle at 20  C until further use. 2.3. Food spiking Defatted foods and food ingredients (100 mg) were spiked with 100 ml antigen solution containing desired quantity of amandin or BSB soluble protein from defatted almond flour. Select food samples (1 g) were also spiked with known quantities of defatted almond flour to determine possible effects of food matrix on amandin immunorecognition when food was spiked with defatted almond flour instead of amandin. 2.4. Protein extraction and quantitation The antigen spiked samples and the corresponding unspiked controls were extracted with 0.9 ml of borate saline buffer (BSB; 0.1 mol/L H3BO3, 0.025 mol/L Na2B4O7, 0.075 mol/L NaCl, pH 8.45) with continuous vortexing for 1 h at RT, centrifuged (13,600 g,10 min, RT), and the aliquots of the supernatants analyzed for soluble protein by the method of Lowry, Rosebrough, Farr and Randall (1951). The remaining supernatants were stored at 4  C until further use. Appropriate blanks were used in all assays. Standard curve for BSA (0–200 mg/ml) was prepared in appropriate buffer for each assay. 2.5. ELISA

2.7. Effects of antigen extraction conditions 2.1. Food matrix preparation The food matrices and appropriate food ingredients (listed in Table 1) were ground in an Osterizer blender (at the ‘‘Grind’’ setting) and defatted for 8 h in a Soxhlet apparatus using petroleum ether (boiling point range 38.3–53.1  C; sample to solvent ratio

2.7.1. Buffer type The antigen-spiked samples and the corresponding controls were extracted with the desired buffer type, buffer volume, and desired time. The buffers used were: 1) BSB, 2) antigen coating citrate buffer used in ELISA (48.5 mmol/L citric acid, 0.103 mmol/L

R.S. Tiwari et al. / LWT - Food Science and Technology 43 (2010) 675–683

677

Table 1 Recovery of amandin in 100 mg of food samples spiked with different levels of amandin. No.

Food category/Matrix

Spike level 100 mg/100 mg

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61

Dairy Mozzarella cheesea,f,g,j Pasteurized American cheesea,f,h,i Swiss cheesea,e,h,i Parmesan cheesea,e,g,j Vanilla ice-creama,f,h,i Macaroni cheesea,f,h,i Romano cheesea,f,g,i Spices (powder) Allspicesb,e,g,j Bay leafb,f,h,j Garlicd,f,h,j Dried parsleyb,e,g,j Cardamomf,h,j Gingerf,h,j Cuminc,e,g,j Dried oreganob,f,h,j Nutmegb,f,h,i Cayenne pepperb,f,h,j Anise seeda,f,h,i Mustardd,f,h,i Black pepperb,f,h Clovesb,f,h,j Corianderb,f,g,i Cinnamond,f,h,j Onionb,f,h,j Dill seeda,f,h,j Indian spice mix 1c,f,h,j Turmericb,f,h,i Indian spice mix 2b,f,h,i Fruits Pears dried with skinb,f,g,i,j, k Pears dried without skinb,f,h,k Tropical mixed fruitb,e,g,i Apple chipsb,f,h,j Mango dried w/o skinb,h,k Pears skinb,g,j,k Mango skinb,e,h,i,k Raw mango powderb,f,g,i Mango pulpb,f,h,i,k Dried bananab,f,h,i Coconut sweetenedb,e,h,i Dried mangob,e,h,i Peachesb,f,h,i Dried cranberriesb,e,h,j Dried cherriesb,g,i Blue cherriesb,f,g,j Mixed fruitb,e,h,i Plumsb,e,g,j Figsb,f,h,i Pineappleb,f,h,j Raisinsb,f,h,j Papayab,f,h,i Apricotb,f,h,j Nuts Brazil nuta,f,h,i Pistachioa,f,g,i Hazelnuta,e,h,i Walnutc,f,h,j Macademiaa,e,h,i Pinenuta,f,h,j Peanuta,f,h,j Pecana,f,h,j Cashew nutc,e,g,j Dried almond skina,f,h,j

10 mg/100 mg

pHx

1 mg/100 mg

LSD

Recovery

SEM

Recovery

SEM

Recovery

137.3 130.2 150.3 141.7 165.7 132.8 198.8

3.99 7.73 7.06 4.14 6.1 10.69 8.01

132.2 116.4 166.4 151.6 140.9 178.5 139.8

6.48 3.4 6.87 3.65 2.19 7.19 21.51

192.1 172.94 137.49 153.7 117.37 172.88 159.43

6.73 3.92 6.84 15.7 9.69 4.43 11.48

7.5 6.8 7.8 7.5 7.0 7.3 7.5

19.84 21.5 24.7 37.8 19.27 24.1 61.58

15.8 21.31 42.73 61.85 18.2 43.87 121.8 17.12 29.55 5.79 157.4 155 38.15 59.79 85.55 70.79 33.49 142.7 131.2 27.26 38.63

2.07 1.34 7.4 4.47 1.25 6.19 10.64 2.22 1.42 1.28 12.14 16.22 2.74 4.75 8.14 7.25 3.43 8.14 13.95 2.55 4.96

10.6 12.72 66.39 64.63 16.55 78.71 128 17.68 14.83 3.87 131.9 156.01 34.85 53.63 50.8 11.72 38.21 148.3 176.9 19.58 49.07

1.68 1.52 1.5 3.2 2.16 3.84 8.94 2.14 2.2 0.57 9.59 2.26 3.68 14.83 4.2 2.75 3.87 7.54 17.57 1.11 4.51

15.61 61.29 140.11 65.16 72.27 26.86 160.29 36.38 8.14 15.65 82.13 124.19 47.83 269.98 54.28 334.25 3.63 43.03 330.5 2.56 84.35

5.12 11.12 26.72 3.87 8.38 3.3 20.21 7.27 1.66 2.51 14.89 22.72 4.77 36.6 12.86 47.68 0.99 6.51 49.23 0.33 10.57

6.6 6.7 7.0 7.0 7.9 7.3 7.5 7.1 7.3 5.2 7.4 6.5 8.3 5.0 7.6 7.1 7.0 8.0 7.5 8.3 7.5

10.19 26.7 60.7 15.04 20.34 17.71 48.06 15.8 7.28 6.06 39.33 52.23 14.38 62.88 27.8 98.95 8.39 18.21 125 6.27 17.37

12.72 38.89 69.78 18.69 49.87 14.38 29.12 10.08 46.41 37.82 45.81 43.02 12.33 16.39 7.16 21.7 19.4 31.14 70.96 61.55 52.45 83.87 74.52

1.92 7.71 4.08 1.6 4.08 2.66 2.3 2.28 3.08 3.43 6.42 6.12 2.24 1.62 1.87 3.77 2.18 2.43 1.65 2.98 7.74 10.26 4.61

19.28 34.27 54.83 19.62 62.72 12.54 35.93 15.31 38.54 49.95 34.13 28.91 7.93 12.55 12.63 26.35 28.03 33.52 74.83 56.72 43.53 102.9 83.22

3.08 7.38 2.16 1.76 5.45 1.81 1.88 1.28 2.74 2.87 2.93 3.77 0.75 2.93 1.11 4.27 4.09 2.29 2.27 4.56 4.55 2.28 5.58

26.75 11.81 64.92 3.63 41.7 12.01 23.15 16.51 1.69 14.59 58.38 43.63 3.31 26.25 15.21 38.19 16.5 31.88 16.98 83.31 305.13 58.69 31.09

4 1.79 4.39 0.53 4.33 0.71 3.02 3.1 0.48 1.83 3.54 5.91 1.07 4.37 0.91 3.25 1.41 3.13 2.71 3.76 27.82 6.73 7.03

8.0 4.9 6.4 5.5 5.0 5.2 5.6 4.0 5.0 5.5 6.0 4.5 5.5 5.5 5.4 6.2 5.0 4.9 5.5 6.1 4.5 5.3 5.0

39.33 22.33 15.01 5.96 15.75 7.07 9.57 9.08 9.38 7.94 16.6 19.51 1.84 10.38 3.94 14.57 11.04 9.64 5.39 14.7 24.55 29.63 22.38

22.65 22.66 14.83 14.01 13.18 17.35 8.16 5.03 2.96 31.08

150.1 163.6 122.9 118.4 123.6 199.4 152.2 132 108.4 104

5.16 4.04 1.5 10.49 7.91 26.45 20.88 11.93 4.16 28.24

96.5 178.01 206.24 566.63 190 82.31 99.39 110.89 114.13 4469.9

7.74 21.78 21.8 62.96 13.83 9.57 14.07 10.57 10.14 252.4

8.3 8.0 7.6 7.5 8.0 7.7 8.0 7.6 8.4 7.8

43.05 62.25 61.99 120.8 46.91 78.53 58.78 31 25.9 415.4

240.6 237.2 184 167.7 192.1 247.9 142.9 129.6 112.4 292.6

SEM

(continued on next page)

678

R.S. Tiwari et al. / LWT - Food Science and Technology 43 (2010) 675–683

Table 1 (continued ) No.

Food category/Matrix

Spike level 100 mg/100 mg

62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117

Vegetables Green pepperd,f,g,i,k Spinachb,f,h,i,k Collard greenc,f,h,j,k Green salada,f,h,j,k Turnip greensa,f,h,i,k White potato skina,f,h,k Brussels sproutd,f,h,j,k Mushrooma,f,g,i,k Red potatoesc,f,h,j,k Broccolib,e,g,j,k White potatoa,f,h,i,j,k Cauliflowerd,f,h,i,k Red potato skina,h,k Cut carrotsa,f,h,i,k Asparagusa,f,h,j,k Legumes Fava beansa,f,h,j Flax seedc,e,h,j Soybeana,e,h,i Tepary bean floura,f,h,i Black gramd,e,h,i Chick pea flourb,f,h,j Cereals Milletsb,f,g,i,j Oat quickd,f,h,i Oat brana,f,g,i Rice brana,f,h Rice floura,f,g Barleyb,e,h,i Whole wheat flourb,f,h Rice wildb,f,h,i Amarantha,f,g,i Corn wholea,e,g Wheat bulgura,f,g,i All purpose Wheat floura,f,h,j Quinoab,f,h,j Buckwheatb,f,h,j Ryea,e,h,i Breakfast cereals Puffed ricec,f,h,i Puffed wheata,e,h,i All bran wholed,e,h,j Oat cereal whole graina,e,g Post raisin brana,f,h,j Raisin bran w/o raisinc,e,g,j Multigrain flakesa,e,h,i Raisin bran whole mixb,f,h,j Corn flaxesd,e,g,i Confectionery Baker’s sweet Chocolateb,f,g,i Baker’s unsweetened Chocolateb,f,h,j Jaggerryb,f,h,j Cocoab,f,h,i Dark chocolateb,f,h,j Brown sugarb,f,h,j Salt Black saltb,f,h,i Saltb,f,h,i Tea and color Annattod,f,h,j Black tead,f,h,j Green teab,f,h,j

10 mg/100 mg

1 mg/100 mg

pHx

LSD

Recovery

SEM

Recovery

SEM

Recovery

221.8 68.77 246.1 198.4 171.9 213.6 116.9 329.3 134.5 58.49 134.3 118 114.8 226.6 171.8

23.94 6.92 17.99 18.75 15.91 16.73 5.65 37.72 3.77 3.5 6.8 13.21 12.4 32.87 18.39

176.03 41.99 133.1 203.6 208.1 176.2 90.2 151.1 134 61.34 131.5 159.46 104.3 140.7 127.4

6.82 5.91 10.1 24.37 12.38 12.79 4.81 10.94 5.27 4.67 6.97 4.14 11.13 9.4 4.63

245.94 239.5 787.3 509.9 31.44 129.1 343.3 128.9 109.7 79.29 100.2 211.3 82.64 421.1 491.3

7.83 10.71 146.4 16.74 2.28 10.69 43.48 34.69 5.97 23.2 11.49 11.44 13.8 37.94 26.46

6.9 8.4 7.7 6.9 7.5 7.6 7.0 7.5 7.9 7.3 8.0 6.7 8.1 6.5 7.3

55.6 27.49 357.2 81.23 38.04 53.85 105.9 115.7 16.17 40.86 35.42 37.12 42.63 102.8 77.18

138.8 108.2 304.7 137.1 122.1 49.26

9.81 6.39 32.14 8.02 6.13 3.07

150.7 132.1 211.3 126.1 161.37 49.3

19.47 21.21 8.32 7.71 8.55 2.66

304.2 90.04 285.6 180.4 119.1 83.45

26.71 20.73 33.94 14.56 9.98 16.23

7.9 8.1 7.5 7.7 7.5 8.5

66.84 67.19 88.63 22.78 29.16 22.9

72.24 116.4 155.4 131.4 151.2 57.57 49.84 46.87 153.2 169.2 140.8 134.1

3.98 3.44 7.09 8.53 10.75 3.93 1.98 1.95 5.57 18.15 7.65 4.48

54.94 106.6 106.5 145.6 110.4 45.79 47.56 40.69 128.5 140.3 128.7 127.9

4.18 5.52 2.4 13.76 1.4 3.58 1.97 1.91 9.75 11.55 5.19 6.81

43.76 83.89 123.3 107.3 106.3 62.73 35.5 25.38 125.8 157.4 119.9 91.49

10.89 2.38 7.24 11.93 8.91 2.04 12.58 3.7 8.22 9 11.45 6.1

7.5 8.4 8.5 8.3 8.5 8.3 8.4 8.2 8.1 8.2 8.5 8.5

29.04 16.12 25.38 41.17 28.11 11.95 21.97 9.87 25.61 56.57 33.17 24.32

55.29 77.82 151.4

1.97 1.39 10.11

56.68 74.69 115.2

2.75 3.68 5.45

74.88 96.16 183

7.88 5.96 19.98

8.2 7.4 8.4

8.71 16.3 46.07

130 147.4 73.44 121.9 151.3 112.4 129.2 21.38 183.4

3.79 6.4 7.67 6.49 3.59 5.61 4.75 2.39 20.08

132.9 120.8 103.9 127.4 131.4 119.8 92.78 24.56 103.3

12.57 5.91 2.04 8.78 10.95 14.17 9.41 3.81 5.83

176.1 139.5 59.96 22.65 110.9 146.1 136.4 42.67 35.31

18.3 4.24 3.14 14.01 3.06 50.08 8.12 4.52 12.25

8.3 8.4 8.1 8.3 8.1 7.9 8.4 6.4 8.2

54.42 21.19 17.39 25.05 27.2 121.8 30.88 10.84 49.01

45.84

2.21

32.01

4.97

24.8

1.9

7.8

13.23

19.76

1.31

26.75

1.8

79.78

6.9

6.9

17.53

49.57 52.37 58.06 46.95

17.13 2.32 6.85 4.23

49.16 76.05 63.34 51.06

4.93 1.36 7.69 4.28

148.8 258.7 108.4 31.75

10.75 8.63 7.34 5.68

7.2 7.0 8.0 7.9

47.1 16.19 28.55 14.61

38.24 32.7

2.1 2.18

46.42 45.93

2.78 3.69

17.49 27.22

2.45 4.61

8.2 7.7

9.8 11.1

74.23 24.14 29.01

5.03 3.59 3.02

85.42 28.91 16.73

2.17 1.29 1.98

9.3 93.95 26.55

7.2 6.2 6.4

23.12 230.6 65.55

123.3 420.3 197.9

SEM

R.S. Tiwari et al. / LWT - Food Science and Technology 43 (2010) 675–683

679

Table 1 (continued ) No.

Food category/Matrix

Spike level 100 mg/100 mg

118 119 120 121 122

Seeds Poppy seeda,f,h,i Elucine coracanab,f,h,i Tapiocab,e,g,i Sesame seeda,f,h,j Sunflower seedsa,f,h,i LSD

10 mg/100 mg

Recovery

SEM

Recovery

125.4 23.71 67.31 224.4 174.2 15.15

6.21 1.5 8.31 7.96 3.95 11.88

104 29.51 45.32 228.3 147.4 41.2

pHx

1 mg/100 mg SEM 8.48 3.34 9.94 6.7 7.75

Recovery 170.6 18.87 60.94 229.81 125.8

LSD

SEM 16.8 3.27 3.97 7.17 7.59

7.9 8.3 7.5 7.8 5.6

24.72 7.09 28.73 26.01 25.32

SEM ¼ Standard error of mean. pHx ¼ pH of the supernatant obtained after extracting the spiked amandin with food matrix in buffer. n ¼ 8. a Food matrices showing higher detection of amandin than theoretical yield. b Food matrices showing lower detection of amandin than theoretical yield. c Food matrices showing no difference in detection of amandin than the theoretical yield. d Food matrices showing no specific pattern for detection of amandin. e Food matrices showing no significant difference in recovery at spike levels 100 and 1 mg/100 mg of food. f Food matrices showing significant difference in recovery at spike levels 100 and 1 mg/100 mg of food. g Food matrices showing no significant difference in recovery at spike levels 10 and 1 mg/100 mg of food. h Food matrices showing significant difference in recovery at spike levels 10 and 1 mg/100 mg of food. i Food matrices showing no significant difference in recovery at spike levels 100 and 10 mg/100 mg of food. j Food matrices showing significant difference in recovery at spike levels 100 and 10 mg/100 mg of food. k These samples were dried in the laboratory while all the rest were purchased as dry samples. The Theoretical yield ¼ 100 mg/100 mg. Data are expressed as ratio of amandin detected over the theoretical yield. LSD ¼ Least significant difference. The differences between the two means, within the same row/column, exceeding the corresponding LSD value are significant (p ¼ 0. 05).

Na2HPO4, pH 5.0), 3) sodium phosphate buffer (0.02 mol/L, pH 7.5), 4) carbonate (0.1 mol/L NaHCO3, pH 9.6), 5) Tris-HCl (0.1 mol/L, pH 8.1), and 6) Phosphate Buffered Saline (PBS; 0.1 mol/L, pH 7.2, 0.9 g NaCl/100 ml). At the end of the extraction period, samples were centrifuged (13,600 g, 10 min, RT) and aliquots of the supernatants were analyzed for total soluble protein and amandin recovery. 2.7.2. pH The antigen-spiked samples were prepared by adding 0.1 ml of antigen (100 mg amandin in BSB buffer) to 100 mg of the defatted ground food prior to extracting the food with 0.9 ml of BSB. Extraction was carried out by vortexing the spiked sample and the corresponding unspiked control continuously for 1 h at RT. Samples and the controls were centrifuged (13,600 g, 10 min, RT). The pH of each supernatant was measured using a pH meter (accuracy of the pH meter was 0.05 pH units). When needed, the pH of the aliquots of food extracts was adjusted to the desired value with 1 mol/L HCl or 1 mol/L NaOH, samples centrifuged (13,600 g, 10 min, RT) to remove any insolubles, soluble protein content of the supernatants determined by the method of Lowry et al. (1951), and aliquots of the supernatants were used for further analyzes. For assessing the influence of pH adjustment on amandin recovery, 100 mg amandin was dissolved in 0.1 mol/L BSB and the pH of the solution was adjusted to the desired value with 1 mol/L HCl or 1 mol/L NaOH and the final volume was adjusted to 1 ml with distilled deionized water. The resulting solutions were then centrifuged (13,600 g, 10 min, RT) and the supernatants used for ELISAs. 2.7.3. Extraction step Since buffer-insoluble components of food may potentially influence amandin interaction with the food matrix, each of the select foods (100 mg) was analyzed for amandin recovery as follows. 1) Food matrix spiked with 100 mg amandin/100 ml BSB (pH 8.45) and the spiked food matrix extracted with 0.9 ml BSB (pH 8.45); 2) Food matrix was first extracted with 0.9 ml BSB (pH 8.45), centrifuged as described above, and the supernatant spiked with 100 mg amandin/100 ml extraction buffer; and 3) Food matrix alone was extracted with 0.9 ml extraction buffer (the control), centrifuged as described above, and to the supernatant was added 100 ml

extraction buffer. The supernatants were analyzed for Lowry protein content followed by amandin determination using the rabbit pAb-based ELISA. 2.8. Data analysis and statistical procedures All experiments were done at least in duplicate and data expressed as Mean  SEM, unless indicated otherwise. Data were analyzed using SPSS (version 11.0.1) for significant differences (p ¼ 0.05) using ANOVA and Fisher’s Least Significance Difference test as described by Ott (1977). 3. Results and discussion The ELISA standard curve with an amandin detection range of 0.03–0.3 mg/ml (30–300 ng/ml) and an IC50 value of w0.075 mg/ml (75 ng/ml) was obtained (Fig. 1). 3.1. Amandin recovery ELISA results of amandin recovery from various foods and food ingredients spiked with three different levels of amandin (100, 10 and 1 mg per 100 mg food matrix) exhibited a wide range (Table 1).

Fig. 1. Mean ELISA standard curve for assays (n ¼ 100) performed on different days.

680

R.S. Tiwari et al. / LWT - Food Science and Technology 43 (2010) 675–683

At 20 g/100 g protein content in whole almond seeds (as is basis) and w65 g/100 g of the total proteins accounted by amandin (Wolf and Sathe, 1998), 1 mg amandin would represent 7.692 mg whole almonds. Typically amandin recoveries from tree nuts and dairy products were 110–292 mg/100 mg and 116–198 mg/100 mg of the theoretical yield. With the exception of chickpea, spiking legumes typically resulted in amandin recovery in the range 108–304 mg/100 mg. Spiked chickpea flour, especially at 10 and 100 mg levels, had only about 50 mg/100 mg recovery. Mixed recoveries, low range 2–85 mg/ 100 mg and high range 121–334 mg/100 mg, were observed when various spices were spiked with amandin. Spiking cereal grain flours and cereal-based products typically resulted in amandin overestimation (range 106–183 mg/100 mg). Barley, buckwheat, finger millets, quinoa, whole-wheat flour, whole bran cereal, wild rice, and raisin bran whole mix were exceptions with lower (range 21–96 mg/ 100 mg) recoveries compared to the theoretical yield. Most vegetables registered increased recovery (range 110–400 mg/100 mg) while salt (black and white), tea, and confectionery (sugar, cocoa, jaggery, dark chocolate) decreased amandin recovery. Fruit matrices decreased amandin recovery (range 1.0–83 mg/100 mg). Certain edible seeds caused over-estimation of amandin recovery (range 104–224 mg/100 mg). Poppy, sesame, and sunflower seeds increased (104–230 mg/100 mg of the theoretical yield) while tapioca, a starchy food, decreased (range 45–67 mg/100 mg) amandin recovery. The pattern of interference (over- or under- estimation) for a specific food matrix typically followed similar trend at all spike levels (100, 10 and 1 mg) tested. However, the detected amandin quantity varied considerably across different food matrices for a specified spike level (notably at 1 mg spike level). These data suggest cross-reactivity or non-immunological interference in the assay by various food matrices and food components. Whether amandin recovery from a variety of spiked food matrices was statistically significant or not, the rabbit pAb based ELISA was able to detect amandin at levels used for spiking in the current study. 3.2. Matrix effects on amandin recognition Possible causes of interference in ELISA by food matrix may include a) amandin reacting with food matrix components thereby possibly resulting in loss (or masking) of epitopes recognized by the rabbit pAbs directed against those epitopes, b) food matrix components reacting with assay reagents (excluding amandin), c) food matrix components decreasing amandin solubility thereby decreasing the estimated amount by ELISA, or d) a combination of two or more factors a–c. To better estimate the possible influence of food matrices on ELISA robustness, ratio (R) was determined: R ¼ IC50 of amandin alone/IC50 of inhibitor. The IC50 inhibitor refers to IC50 for food matrix spiked with the fixed quantity (e.g. 100 mg) of amandin. The summary of ranges of R values is shown in Table 2. The rationale for using IC50 as an indicator of amandin reactivity under assay conditions was based on the expectation that alteration in amandin reactivity would be captured in the most sensitive manner when reactivity was assessed using the most sensitive portion of the ELISA curve (values at or near IC50). The R value was used to estimate possible matrix interference as follows: when R ¼ 1, IC50 of the inhibitor ¼ IC50 of amandin, suggesting no change in amandin reactivity. If R < 1, then IC50 of the inhibitor >IC50 of amandin, suggesting loss in amandin activity due to interference by matrix components in some manner. If R > 1, then IC50 of the inhibitor
Table 2 The ratio of IC50 values of spiked amandin in food (spike level 100 mg/100 mg of food). Food Category

IC50 (mg/ml)

Ratio

Control Dairy Nuts Vegetables Seeds Legumes Cereals

0.05–0.8 0.0068–0.014 0.01–0.06 0.0009–0.06 0.0007 0.0009–0.06 0.015–0.04

R>1 R>1 R>1 R>1 R>1 R>1

Breakfast cereals Fruits Confectionery Salt Tea & Color Spices

0.013–0.036 0.2–0.5 0.2–0.33 0.1–0.2 0.1–0.5 0.1–1.2

R>1 R<1 R<1 R<1 R<1 R<1

Matrices with IC50 different from the food category trend

Dried almond skin Red potato skin Ragi seed, Tapioca Barley, Rice wild, All purpose wheat flour, Millets Multigrain flakes, All bran whole Jaggery

Anise, Garlic, Coriander, Dill seed, Cumin, Mustard

R is typically 1. Most legumes, cereals, and cereal products also showed similar pattern with the exception of millets, wild rice, barley and all-purpose wheat flour where R < 1. For fruits, R was significantly <1. Similarly R was <1 for confectionery, salt, tea and color with the exception of jaggery where R > 1. Garlic powder, anise seeds, mustard, coriander, cumin, and dill seeds registered R > 1 indicating possible cross-reactivity with rabbit pAbs. Amandin recovery and matrix effect data, taken together, suggest that while the rabbit pAbs targeting amandin as the marker protein for almond detection are useful, the assay may be susceptible to interferences by certain food matrices. The level of interference appears to be dependent on the food matrix and the assay conditions. For example, the R range for fruits was 0.2–0.5 which suggests that IC50 of the control amandin (0.05 mg/ml) changed significantly (range 0.25–0.1 mg/ml) in presence of fruit matrix. In practice, under the assay conditions used in the current investigation this change equates to a shift from ability to detect 50 ng/ml to 100–250 ng amandin/ml (i.e. decrease in assay sensitivity by a factor of 2–5). Depending on the desired level of assay sensitivity such shift may or may not be of concern.

3.3. Effect of extraction buffer on amandin recognition and recovery In order to determine a suitable buffer for extracting amandin from various food matrices, several buffers commonly used for protein extraction and ELISAs were evaluated for their effect on amandin determination as assessed by IC50 values (Fig. 2A). The results indicated that buffers used in this study did not significantly influence the amandin IC50 thereby suggesting that buffer per se did not alter amandin in a manner that would affect amandin recognition by the rabbit pAbs. These experiments however did not permit assessment of amandin extraction efficiency from the targeted food matrix by a test buffer. We therefore evaluated these same buffers for their ability to extract amandin from a few selected food matrices (Fig. 2B). The choice of food matrix was partly based on the matrix interference observed in the preliminary studies (e.g. spice, fruit, nut, chocolate, and vegetable) shown in Table 1. The results indicated a wide variation in amandin (mg/100 mg) recovery. Among the buffers tested, BSB was judged to be effective for amandin extraction. Although the current study used a small number of select food matrices for antigen extraction, the data suggest buffer-food matrix interaction (e.g. note the higher amandin extraction from turnip green versus dark chocolate using the same buffer) is important and should be taken into account when selecting the extraction buffer for optimal antigen solubilization and determination.

R.S. Tiwari et al. / LWT - Food Science and Technology 43 (2010) 675–683

681

Fig. 2. Effect of extraction buffer (A, B) and pH (C, D) on amandin recognition and recovery. All data are represented as Mean  Standard Deviation (SD), where n ¼ 8 for A, B, and D and n ¼ 6 for C. LSD (p ¼ 0.05) ¼ 56.45 and 34.15 for A and C, respectively. LSD values for each food category in B are given above the data set for the food matrix. The % recovery is calculated based on expected theoretical yield. Compared to citrate buffer, only carbonate buffer registered significantly higher amandin recovery (A). Acid pH (pH 2) significantly decreased amandin recovery (C). The original (,) pH of extracts (pH range 4.0–5.0) was adjusted to a final value (-) (pH range 8.0–9.0) using 1 mol/L NaOH. This pH adjustment of the extracts significantly improved amandin recovery for all but raisin matrix (D). Different buffers used in (B) include ,BSB, Citrate, Phosphate, - Carbonate, Tris-HCl, and PBS.

3.4. Effect of pH on amandin recovery Based on our previous experience with amandin protein extraction (Sathe et al., 2002; Wolf and Sathe, 1998), the pH and ionic strength were anticipated to influence amandin solubility which in turn may affect amandin recovery. Therefore, the influence of pH on amandin recovery was evaluated. Only the extreme acid pH (pH ¼ 2) had a statistically significant effect on amandin recovery (Fig. 2C), a result consistent with our earlier observations on reduced amandin pAb reactivity at extreme acid and alkali pH (Acosta et al., 1999). Possible explanation for such a loss in reactivity at pH 2 may include: 1) epitope denaturation and/or destruction, 2) epitope modification, 3) epitope masking as a result of partial

protein denaturation, or 4) a combination of two or more of these factors. Further studies will be needed to clarify the relative contribution of these factors. Immunoreactivity of amandin at acid pH is of interest as the pH range of 1–4 is physiologically relevant with regards to almond allergies in humans. Loss in amandin solubility as a result of possible acid induced protein aggregation at pH 2 may not account for all of the loss in antigenic reactivity as amandin prepared by isoelectric precipitation method was demonstrated to retain the majority of the antigenicity compared to the activity of its native counterpart (Sathe et al., 2002). The current study also suggests the IC50 values for amandin over a pH range 4–10 are not statistically significantly different indicating remarkable stability of amandin immunoreactivity over a wide pH range.

Fig. 3. Amandin recovery as affected by (A) matrix spiked with defatted almond flour (,) and amandin (-); (B) spiked with amandin before (,) and after (-) extraction; (C) matrix:buffer ratio of 1:10 w/v (,) and 1:20 w/v (-); and (D) chocolate matrix spiked with indicated amandin amounts.

682

R.S. Tiwari et al. / LWT - Food Science and Technology 43 (2010) 675–683

0.9 ml buffer and the resultant protein extracts used for IC50 determinations. The pH of the extracts ranged from 3.5 to 5.7 before adjustment. The final pH range of spiked food matrices upon pH adjustments, was 8.0–9.9. Depending on the matrix, pH adjustment of the extract improved amandin recovery by 5–48 mg/100 mg (Fig. 2D). 3.5. Amandin recovery from select foods spiked with defatted almond flour or amandin

Fig. 4. Amandin recovery from select food matrices spiked with 100 mg/g (-), 50 mg/g (,), and 10 mg/g ( ) of amandin (top) and defatted almond flour protein extract (bottom). Values are mean  SD (n ¼ 4). Spiking was done prior to protein extract preparation.

Since acidic pH seemed to affect amandin recovery and as a majority of fruits and vegetables are acidic in nature, select fruit/vegetable based food matrices were tested for amandin recovery by extracting two sets of paired samples of the same food matrix (unspiked control with only 100 ml BSB added and spiked sample with 100 mg amandin in 100 ml of BSB) with

ELISA results from select food matrices spiked with defatted almond flour or amandin did register differences in amandin recovery (Fig. 3A). The food matrix and the non-amandin protein components of the almond flour used for spiking may be responsible for the observed differences. It was of interest to learn whether the observed variable recovery in ELISA was due to matrix proteins interacting with amandin in the presence of insoluble components of the food matrix as compared to buffer-soluble food matrix components. For these reasons, amandin recoveries were compared from select food matrices spiked with amandin before and after matrix protein extraction. The results suggested the effect to be matrix dependent (Fig. 3B). For example, cloves and dark chocolate exhibited significantly lower recovery when spiked with amandin before extraction, which was not the case for amaranth and coriander. Together these results suggest possible interactions of the buffer-insoluble components of the food matrix with the targeted antigen. To further explore matrix interference, we selected eleven matrices that are likely to be encountered when almonds are used in food preparations. Of these, salt and nutmeg are not expected to be used at the levels we tested in the current investigation but were included for comparative purposes. Testing with these additional different matrices, using different levels of amandin or almond protein extract spiking (10, 50, and 100 mg/g) revealed the amandin recovery was not only matrix dependent but was also affected by spiking level (Fig. 4). Although we tested amandin recovery from a limited number of select matrices, the data suggest spiking food matrices prior to protein extraction may provide better amandin recovery.

Fig. 5. Ponceau S stain (top) and Western blot (bottom) of amandin spiked chocolate matrices using rabbit anti-amandin pAbs. S-standards; A-10 mg amandin. Spiking level (mg/g): 1- none; 2- 1000; 3- 500; 4- 100; 5- 50; 6- 20; 7- 10; 8- 5. 100 ml of extract was loaded in the corresponding lane.

R.S. Tiwari et al. / LWT - Food Science and Technology 43 (2010) 675–683

Almonds are often used in chocolate and confectionery products. Dark chocolate, a food that contains large quantities of phenolic compounds, is reportedly a good source of antioxidants (Lee, Kim. Lee, & Lee, 2003; Hurst, Glinski, Miller, Apgar, Davey, & Stuart, 2008; Niemenak, Rohsius, Elwers, Ndoumou, & Lieberei, 2008). Since almonds are also a rich source of vitamin E, an antioxidant, several manufacturers have dark chocolate products on the market that incorporate almonds and berries (e.g. cranberries) that contain phenolics compounds. Investigating amandin recovery from chocolate matrix in general and dark chocolate in particular was therefore of interest. The results suggest dark chocolate significantly decreased amandin recovery when dark chocolate was spiked with amandin prior to protein extraction (Figs. 3B, and 4A). The mechanism by which dark chocolate may affect antigen recovery remains unknown. Since chocolate matrix negatively affected ELISA recovery, it was of interest to visually assess whether the low recovery was due to a lack of amandin recovery or chocolate matrix proteins being cross-reactive with rabbit anti-amandin pAb. Therefore, 100 mg defatted dark, milk, and white chocolate matrices were spiked with amandin (5–1000 mg/g) and extracted proteins were probed with rabbit anti-amandin pAb. Amandin was detected at spike levels of 50 mg/g in white, and 100 mg/g in dark and milk chocolates (Fig. 5). Similar detection sensitivity has been reported for hazelnut detection from hazelnut spiked chocolate matrix using Western blot assay (Scheibe et al., 2001). 4. Conclusion In summary, the ELISA successfully detected amandin residues in all tested food matrices. Cross-reactivity of tested matrices in anti-amandin pAb based assay was not significant in that the assay could reliably detect the presence of amandin. The developed ELISA will be useful as a screening tool for qualitative and quantitative detection of trace quantities, w30–150 mg/L, of almonds in a variety of tested foods and ingredients. Acknowledgments This work was supported by grants from the College of Human Sciences (Research Initiative Award program), Council for Faculty Research (COFRs), Florida State University; and the USDA NRICGP (#9901530). References Acosta, M. R., Roux, K. H., Teuber, S. S., & Sathe, S. K. (1999). Production and characterization of rabbit polyclonal antibodies to almond (Prunus dulcis L.) major storage protein. Journal of Agricultural and Food Chemistry, 47, 4053–4059. Angus, F. (1998). Nut allergens. In D. Watson (Ed.), Natural toxicants in foods (pp. 84–104). Boca Raton: CRC Press.

683

Bargman, T. J., Rupnow, J. H., & Taylor, S. J. (1992). IgE-binding protein in almonds (Prunus amygdalus): identification by immunoblotting with sera from almond allergic adults. Journal of Food Science, 57, 717–720. ˜ oz-Furlong, A., & Sampson, H. A. (2007). Further fatalities caused by Bock, A. S., Mun anaphylactic reactions to food, 2001–2006. The Journal of Allergy and Clinical Immunology, 119, 1016–1018. FAO Stats 2007. http://faostat.fao.org Accessed 06.08.09. Fling, S. P., & Gregerson, D. S. (1986). Peptide and protein molecular weight determination by electrophoresis using a high-molarity tris buffer system without urea. Analytical Biochemistry, 155, 83–88. Hefle, S. L., Furlong, T. J., Niemann, L., Lemon-Mule, H., Sicherer, S., & Taylor, S. L. (2007). Consumer attitudes and risks associated with packaged foods having advisory labeling regarding the presence of peanuts. The Journal of Allergy and Clinical Immunology, 120, 171–176. Hurst, W. J., Glinski, J. A., Miller, K. B., Apgar, J., Davey, M. H., & Stuart, D. A. (2008). Survey of the trans-resveratrol and trans-piceid content of cocoa-containing and chocolate products. Journal of Agricultural and Food Chemistry, 56, 8374–8378. Lee, K. W., Kim, Y. J., Lee, H. J., & Lee, C. Y. (2003). Cocoa has more phenolic phytochemicals and a higher antioxidant capacity than teas and red wine. Journal of Agricultural and Food Chemistry, 51, 7292–7295. Lowry, O. H., Rosebrough, N. J., Farr, A. L., & Randall, R. J. (1951). Protein measurement with Folin phenol reagent. The Journal of Biological Chemistry, 193, 265–275. Niemenak, N., Rohsius, C., Elwers, S., Ndoumou, D. O., & Lieberei, R. (2008). Comparative study of different cocoa (Theobroma cacao L.) clones in terms of their phenolics and anthocyanins contents. Journal of Food Composition and Analysis, 19, 612–619. Ott, L. (1977). An introduction to statistical methods and data analysis. Belmont, CA: Duxbury Press, A Division of Wadsworth Publishing Company. Rejeb, S. B., Abbott, M., Davies, D., Cleroux, C., & Delahaut, P. (2005). Multiallergen screening immunoassay for the detection of protein markers of peanut and four tree nuts in chocolate. Food Additives and Contaminants, 22, 709–715. Roux, K. H., Teuber, S. S., Robotham, J. S., & Sathe, S. K. (2001). Detection and stability of the major almond allergen in foods. Journal of Agricultural and Food Chemisty, 49, 2131–2136. Sathe, S. K. (1993). Solubilization, electrophoretic characterization and in vitro digestibility of almond proteins. Journal of Food Biochemistry, 16, 249–264. Sathe, S. K., Wolf, W. J., Roux, K. H., Teuber, S. S., Venkatachalam, M., & SzeTao, K. W. C. (2002). Biochemical characterization of amandin, the major storage protein in almond (Prunus dulcis l.). Journal of Agricultural and Food Chemistry, 50, 4333–4341. Sathe, S. K., Teuber, S. S., Gradziel, T. M., & Roux, K. H. (2001). Electrophoretic and immunological analyses of almond (Prunus dulcis L.) genotypes and hybrids. Journal of Agricultural and Food Chemistry, 49, 2043–2052. Scheibe, B., Weiss, W., Rue¨ff, F., Przybilla, B., & Go¨rg, A. (2001). Detection of trace amounts of hidden allergens: hazelnut and almond proteins in chocolate. Journal of Chromatography B, 756, 229–237. ˜ oz-Furlong, A., & Sampson, H. A. (2003). Prevalence of peanut Sicherer, S. H., Mun and tree nut allergy in the United States determined by means of a random digit dial telephone survey: a 5-year follow-up study. The Journal of Allergy and Clinical Immunology, 112, 1203–1207. Su, M., Venkatachalam, M., Teuber, S. S., Roux, K. H., & Sathe, S. K. (2004). Impact of g irradiation and thermal processing on the antigenicity of almond, cashew nut and walnut proteins. Journal of the Science of Food and Agriculture, 84, 1119–1125. Venkatachalam, M., Teuber, S. S., Roux, K. H., & Sathe, S. K. (2002). Effects of roasting, blanching, autoclaving, and microwave heating on antigenicity of almond (Prunus dulcis L.) proteins. Journal of Agricultural and Food Chemistry, 50, 3544–3548. Wolf, W. J., & Sathe, S. K. (1998). Ultracentrifugal and polyacrylamide gel Electrophoretic studies of extractability and stability of almond meal proteins. Journal of Agricultural and Food Chemistry, 78, 511–521.