Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=cfai20

Download by: [Heriot-Watt University] Date: 07 May 2016, At: 21:22

Food and Agricultural Immunology

ISSN: 0954-0105 (Print) 1465-3443 (Online) Journal homepage: http://www.tandfonline.com/loi/cfai20

Enzyme-linked immunoassay for the detection ofglyphosate in food samples using avian antibodies

A. Arul Selvi , M.A. Sreenivasa & H.K. Manonmani

To cite this article: A. Arul Selvi , M.A. Sreenivasa & H.K. Manonmani (2011) Enzyme-linkedimmunoassay for the detection of glyphosate in food samples using avian antibodies, Foodand Agricultural Immunology, 22:3, 217-228, DOI: 10.1080/09540105.2011.553799

To link to this article: http://dx.doi.org/10.1080/09540105.2011.553799

Published online: 16 Mar 2011.

Submit your article to this journal

Article views: 141

View related articles

Citing articles: 5 View citing articles

Enzyme-linked immunoassay for the detection of glyphosate in foodsamples using avian antibodies

A. Arul Selvia, M.A. Sreenivasab and H.K. Manonmania*

aFermentation Technology and Bioengineering Department, Central Food TechnologicalResearch Institute, Mysore 570 020, Karnataka, India; bFood Safety and Analytical QualityControl Laboratory, Central Food Technological Research Institute, Mysore 570 020,Karnataka, India

(Received 1 July 2010; final version received 8 January 2011)

A simple competitive immunoassay was developed for the measurement ofglyphosate in food samples. It employed the avian antibodies (IgY) thatrecognised glyphosate as a capture reagent and glyphosate�alkaline phosphataseconjugate as an enzyme label. The assay depended on the competitive bindingbetween the antibody and glyphosate derived from food samples for binding siteswith immobilised glyphosate�OVA conjugate. The concentration of glyphosate inthe food samples was quantified by the ability of the pesticide present in foodsamples to inhibit the binding of the enzyme conjugate to the antibody andsubsequently the colour formation in the assay. The assay was specific toglyphosate with a limit of detection of 2 ppb. Mean analytical recovery ofglyphosate in different food samples was 9.00�134.00%. The precision of theassay was satisfactory. The assay compared favourably with high-pressure liquidchromatography (HPLC) in its ability to accurately measure glyphosate in thefood samples.

Keywords: glyphosate; avian antibodies (IgY); enzyme-linked immunosorbentassays; herbicide; food samples; high-pressure liquid chromatography; immunol-ogy; hapten

Introduction

Glyphosate [N-(phosphonomethyl) glycine] is a broad-spectrum, non-selective,

systemic, post-emergence herbicide used both agriculturally and domestically

(Mensink & Janssen, 1994). It is an active ingredient in a number of commercial

herbicides produced by Monsanto, Cheminova and Zeneca Corp. Various formula-

tions are scheduled for use that include products with the trade names Clear It,

Expedite, Ezject, Glyfos, Laredo, Renegade, Roundup, Touchdown 480, Vision and

Wrangler. Glyphosate is a member of the amino acid herbicide family, and its mode

of action is through inhibition of 5-enolpyruvylshikimate-3-phosphate synthase

(EPSP), an enzyme of the shikimic acid pathway. This enzyme is important in the

biosynthesis of the aromatic amino acids � phenylalanine, tyrosine and tryptophan.

A blockage of the shikimic acid pathway leads to a depletion of the free pool of

aromatic amino acids in higher plants (Duke, 1988). Glyphosate translocates readily

in the plant, making it effective for controlling perennial weeds and overwintering

*Corresponding author. Email: manonmani_99@yahoo.com

Food and Agricultural Immunology

Vol. 22, No. 3, September 2011, 217�228

ISSN 0954-0105 print/ISSN 1465-3443 online

# 2011 Taylor & Francis

DOI: 10.1080/09540105.2011.553799

http://www.informaworld.com

Dow

nloa

ded

by [

Her

iot-

Wat

t Uni

vers

ity]

at 2

1:22

07

May

201

6

rhizomes and tubers. It is registered for preplant or postharvest treatment in crops

and on non-crop land (Omafra staff, 2002). The continued use of glyphosate

indiscriminately by agriculturalists raises the potential for residue accumulation in

food and water commodities. The maximum residue limit (MRL) concentration for

food crops has been set at 1.0 mg/ml by the Indian Prevention of Food Adulteration

act (PFA), the Canadian Food and Drug Act as well as the US Food and Drug

Regulations (USFDA). Glyphosate is persistent in soil with a half-life of 47 days(Wauchope, Buttler, Hornsby, Augustijn Beckers, & Burt, 1992). In many countries,

toxicology studies have been conducted to enable evaluation of the potential health

risk (Williams, Kroes, & Munro, 2000). Glyphosate’s laboratory testing has

discovered some negative short-term and long-term health effects. Additionally,

the glyphosate-containing product Roundup has been known to be used in suicide

cases in Japan and consumption of it results in symptoms such as intestinal pain,

vomiting, excess fluid in the lungs, pneumonia, clouding of consciousness and

destruction of red blood cells. Short-term exposure to glyphosate can cause breathing

difficulties, loss of muscle control and convulsions. Some glyphosate-containing

products belong to Toxicity Category I and II (acutely toxic) and are toxic to animals

and humans, and some symptoms include eye and skin irritation, headache, nausea,

numbness, elevated blood pressure and heart palpitations (Cox, 2000). Although

glyphosate is an acid, it is commonly used as the isopropylamine or trimethylsulfo-

nium salts and is usually distributed as water-soluble concentrates and powders.

Glyphosate analysis in environmental matrixes is problematic because it is a smallmolecule and has structural similarity to many naturally occurring plant materials

such as amino acids and secondary plant compounds. It is highly soluble in water

thereby making its extraction with solvents difficult. Therefore, glyphosate isolation

and quantitation poses a challenge to the analytical chemist due to the necessity of

removing matrix effects before analysis. In the last few years, many techniques have

been tested for analysis of glyphosate, including ion chromatography with UV

detection by Bedry and Parrot (1998), fluorescence detection (Mallat & Barcelo,

1998), conductivity detection (Zhu, Zhang, Tong, & Liu, 1999) or mass spectro-

metric detection (Bauer, Knepper, Maes, Schatz, & Voihsel, 1999). These methods

are not, however, sufficiently sensitive to reach the quality threshold for drinking

water. Gas chromatography continues to be of interest, although it suffers from

problems such as tedious sample preparation to clean-up and convert the molecules

into volatile derivatives (Borjesson & Torstensson, 2000). Another technique

investigated is high-pressure liquid chromatography (HPLC) with post-column

derivatisation with o-phthaldealdehyde-mercaptoethanol and fluorescence detection

(Reupert & Schlett, 1997; Winfield, 1990). The most difficult part of the analysis is,

in fact, the extraction of the compounds from food materials, so derivatisation ofglyphosate is usually conducted in this medium. The detection achieved using these

methods is generally at higher concentrations than those typically obtained using

enzyme-linked immunosorbent assays (ELISA) for most herbicides (Rubio et al.,

2003).

ELISA is therefore said to be a valuable tool in residue analysis and complements

conventional analytical methods (Johnson & Hall, 1996; Parnell & Hall, 1998).

ELISA provides rapid sample testing and accurate results and is more cost-effective

than conventional chromatographic analysis (Hall, Deschamps, & McDermott,

1990). ELISA has been used successfully for the quantitative analysis of numerous

218 A. Arul Selvi et al.

Dow

nloa

ded

by [

Her

iot-

Wat

t Uni

vers

ity]

at 2

1:22

07

May

201

6

pesticides in food matrices with little or no matrix interference (Lawruk et al., 1994;

Rubio, Itak, Scutellaro, Selisker, & Herzog, 1991).Previous publication on the indirect immunosorbent assay for the detection and

quantitation of glyphosate described by Clegg, Stephenson, and Hall (1999) using

mammalian antibodies (IgG) showed the limit of detection of 0.076 mg mL�1 which

indicates less sensitivity in comparison with the present studies which showed the

detection limit of 2 ng mL�1 using avian antibodies (IgY). This implies that the hen’s

antibodies (IgY) in particular have clear advantages over mammalian immunoglo-

bulins (Schade, Behn, Erhard, Hlinak, & Staak, 2001).

In this paper, we report the quantitative performance of ELISA for glyphosate

detection and quantitation in environmental food samples using avian antibodies.

Material and methods

Chemicals and reagents

Bovine serum albumin (BSA), Freund’s complete adjuvant, Freund’s incomplete

adjuvant, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), Ovalbumin (OVA),

2,4,6-Trinitrobenzenesulfonic acid (TNBS) and other pesticides standards were

purchased from Sigma-Aldrich Chemical Company (Saint Louis, MO, USA). The

analytical standard of glyphosate (99%) was obtained from Monsanto Co (St. Louis,

MO, USA). Rabbit anti-chicken IgY conjugated to alkaline phosphatase (ALP), was

purchased from Bangalore Genei (Bangalore, India). Microtiter plates were

purchased from NUNC (Roskilde, Denmark). ELISA plates were analysed using a

VERSAmax tunable microplate reader (Silicon valley, CA, USA). Dialysis cassettes

were purchased from Thermo Scientific (Rockford, IL, USA). All the other

chemicals used were of analytical grade and purchased from standard chemical

companies. Food samples such as turmeric, chili, tea samples from different regions,

coffee, coriander, ginger and rice were purchased from local market.

The hapten, OVA-pesticide conjugate and anti-glyphosate antibody for glypho-

sate assay were produced at CFTRI, Mysore, using 22-week-old single comb white

leg horn poultry that were purchased from Kateel Poultry Farm (Mysore, India).

Synthesis of pesticide�protein�hapten conjugates

The synthetic route for the preparation of conjugate was devised. The glyphosate�BSA immunogen was prepared by dissolving the acid form of glyphosate (0.3 mM)

and 100 mg of BSA in 10.0 mL of 0.1 M phosphate-buffered saline (PBS). This

solution was mixed with 130 mg of 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide

and allowed to react for 2 h at room temperature on a magnetic stirrer. The solution

containing the glyphosate�protein (gly-BSA) conjugates was dialysed extensively

using the dialysing cassettes (10K MW cutoff) against distilled water for 24 h at 48C.

The extent of conjugation was determined by using TNBS method (Plapp, Moore, &

Stein, 1971). Coating conjugate of glyphosate (gly-OVA) was synthesised by

conjugating glyphosate to ovalbumin (OVA). The conjugates thus obtained were

lyophilised and stored at �208C till further use.

Food and Agricultural Immunology 219

Dow

nloa

ded

by [

Her

iot-

Wat

t Uni

vers

ity]

at 2

1:22

07

May

201

6

Estimation of the level of conjugation

The TNBS reaction was carried out as follows. 200 mL of 0.6 M sodium borate

buffer, pH 9.5, was mixed with 200 mL of a solution of conjugate in 0.05 M sodium

acetate buffer, pH 4.7 and 50 mL of 0.2 M NaOH. The reaction was initiated by the

addition of 50 ml of a freshly prepared solution of TNBS (7.2 mg mL�1 of water). The

reaction mixtures were kept at 258C in dark for 60 min and then the absorbance at

367 nm was measured.

Immunisation of poultry

Twenty-two-week old poultry were injected intramuscularly, with 1 mg of immuno-

gen (Gly-BSA conjugate) in 0.5 mL of a 50 mM PBS, pH 7.5 and 0.5 mL Freund’s

complete adjuvant. The secondary immunisations were repeated once in 15 days for 6

months with 1 mg of immunogen in 0.5 mL of a 50 mM PBS, pH 7.5 and 0.5 mLFreund’s incomplete adjuvant. The eggs were collected everyday for the isolation of

antibodies.

Anti-glyphosate antibody isolation

Yolk was separated from egg white and suspended in PBS. The egg yolk was brokenand made into uniform suspension by mixing well in a magnetic stirrer for 10 min.

Mixing was continued for 30 more minutes with the addition of chloroform at room

temperature. The mixture was centrifuged at 10,000 g, 48C for 10 min. The

supernatant was decanted and Polyethylene Glycol 6000 was added at 14% (W/V)

level. Mixing was continued for 30 min at room temperature. Antibody was

separated by centrifuging at 10,000 g, 48C for 10 min. The precipitate (IgY) obtained

was dissolved in required quantity of PBS and stored at �208C (Schade et al., 2001).

Immunoassay procedure

Glyphosate-specific antibody titres were monitored as described by Campbell,

Burdon, and Knippenberg (1984) and Gee, Miyamoto, Goodrow, Buster, and

Hammock (1988). The protein concentration of IgY was determined by Bradford

method (1976). The optimum antibody concentration for coating onto the microwellplates and the best working concentration of the enzyme conjugate were determined

by checkerboard titration. The coating conjugate (glyphosate�OVA conjugate) was

diluted in carbonate buffer (50 mM, pH 9.6) at concentrations of 3.125 mg mL�1

through 625 pg mL�1/ml and added to the wells of the microwell plates (100 mL).

The plates were incubated at 48C overnight and were washed with 0.1% Tween 20 in

PBS (137 mM NaCl, 3 mM KCl and 10 mM sodium phosphate, pH 7.4) (PBS-T),

and the wells were blocked with 0.2% gelatine for 1 h. After a wash with PBS-T, anti-

glyphosate antibody (IgY) diluted in carbonate buffer at concentrations of 34.9 mgmL�1 through 69.5 ng mL�1 was coated onto microwell plates (100 mL). The plates

were incubated for 1h at 378C. After a thorough wash in PBS-T, anti-chicken rabbit

antibody (1: 10,000) conjugated with ALP was added (100 mL). After 1 h at 378C, the

plates were washed well in PBS-T. Colour was developed using p-nitrophenyl

phosphate (pNPP) (1 mg mL�1) in 1% diethanolamine buffer, pH 9.8 (150 mL). The

220 A. Arul Selvi et al.

Dow

nloa

ded

by [

Her

iot-

Wat

t Uni

vers

ity]

at 2

1:22

07

May

201

6

reaction was stopped by the addition of 50 mL of 3 M NaOH. The absorbance was

measured in an ELISA reader at 405 nm. Concentrations of antibody and enzyme

conjugate that yielded a signal between 0.8 and 1.2 absorbance units were used for

further testing of the competitive immunoassay procedures.

Optimisation of glyphosate concentration by ELISA

Coating conjugate was loaded to microwell titre plates (100 mL/well) and allowed to

incubate overnight at 48C. The plates were washed well with PBS-T and patted dry on

paper towels. Sites not containing coating conjugate were blocked with 100 mL/wellof 0.2% gelatin. After 60 min incubation, the plates were washed and dried as

previously described. Glyphosate standards or samples containing glyphosate were

prepared in carbonate buffer. Anti-glyphosate antibody and sample or standard

solutions were mixed 1:1 (v/v) and allowed to incubate in test tubes for 60 min. The

pre-incubated mixtures were transferred to the plates (100 mL/well). The plates were

incubated for another 60 min at room temperature before being washed with PBS-T.

The secondary antibody tagged with ALP was added to each well (100 mL) and

incubated at 378C for 60 min. The wells were then washed as described earlier. About150 mL of pNPP in diethanolamine buffer was added and incubated for 30 min.

Optical density (OD) at 405 nm was recorded after stopping the reaction with 50 mL

of 3 M NaOH.

Standard curve

The absorbance at 405 nm is inversely proportional to the concentration of

glyphosate in the standards and samples. To normalise the absorbance, the

background absorbance was subtracted from each value followed by division of

each value by the positive control (0 ng ml�1 glyphosate). The standard curves were

constructed by plotting the normalised absorbance values (A/A0) against the log

values of the glyphosate concentration. Glyphosate concentrations in the food

samples were interpolated from the standard curve.

Cross-reactivity

Few pesticides were tested for cross-reactivity to the anti-glyphosate antibody.

Glyphosine [N, N-bis (phosphonomethyl) glycine], a related herbicide, and few

structurally related smaller molecules were tested for their cross-reactivity. A 1000

mg mL�1 standard of each test chemicals was prepared using the carbonate bufferand tested against the anti-glyphosate antibodies. The IC50 value, per cent cross-

reactivity and least detectable dose for each of the compounds were determined.

High-pressure liquid chromatography (HPLC)analysis

Sample preparation was carried out by weighing one gram of food sample into a50 mL polypropylene centrifuge tube with stopper. About 10 mL of ultrapure

water and 5 ml dichloromethane were added. The tube was sonicated for 10 min

and shaken thoroughly for 1 min that was followed again by 10 min of sonication.

After that the tube was centrifuged for 10 min at 8000 g. The upper water phase,

Food and Agricultural Immunology 221

Dow

nloa

ded

by [

Her

iot-

Wat

t Uni

vers

ity]

at 2

1:22

07

May

201

6

approximately 9 mL, was transferred to another polypropylene tube. Evaporation

to dryness with stream of N2 at 558C on water bath was followed, and the sample

was redissolved in 1 mL of carbonate buffer prepared using ultrapure distilled

water.

Glyphosate was detected by HPLC using a cation exchange analytical column.

The compound was hydrolysed at 368C (Clegg & Ripley, 1996) with sodium

hypochlorite to form glycine. The glycine was then reacted with o-phthalaldehyde

(OPA) in the presence of mercaptoethanol at 558C to produce a highly fluorescent

isoindole, which was detected fluorometrically (excitation 330 nm, emission 465

nm). All glyphosate standards were prepared in distilled water and serially diluted

in potassium dihydrogen phosphate buffer to the concentration range required for

analysis. A volume of 10 mL of standard/sample was injected into the liquid

chromatograph, which had a mobile phase of potassium dihydrogen phosphate

buffer (0.005 M) with a flow rate set at 1.0 ml/min. The total chromatographic run

was 30 min. The mobile phase of potassium dihydrogen phosphate buffer was

isocratic for 17 min, column regenerant (potassium hydroxide) for 2 min and

finishing the chromatographic run with an 11 min continuation of potassium

dihydrogen phosphate buffer mobile phase. Peak areas of the standards were

plotted against the concentration of glyphosate, and the resulting standard curve

was used to interpolate glyphosate concentrations in the food samples.

Results and discussion

This study describes a sensitive enzyme immunoassay that quantifies glyphosate in

food samples using avian antibodies. The food sample containing glyphosate

competes with anti-glyphosate IgY which will be recognised by immobilised

glyphosate�OVA antigen (coating conjugate) in microwell plates. After removal of

unbound reagents, the amount of antibody bound to the immobilised glyphosate�OVA conjugate was determined by using a chromogenic substrate. The concentra-

tion of glyphosate in a food sample was quantified by the ability of the glyphosate

in food material to inhibit binding of anti-glyphosate antibody to the conjugate,

and the colour development was inversely proportional to the concentration of

glyphosate in the original sample.

It was observed that with the glyphosate�BSA conjugate, out of total 28 available

lysine molecules in BSA, 19 lysine molecules were found to be bound to glyphosate

molecule, i.e. the extent of conjugation was 68%.

Optimum assay conditions

The optimum concentrations of anti-glyphosate antibody required and the best

working concentration of the conjugate required for coating on to the microwell

plates were determined by checkerboard analysis. The analysis indicated that

antigen (glyphosate�OVA conjugate) dilution of 1:2.5 lakhs with carbonate buffer

(125 pg protein) and antibody dilution of 1:25 K (139 ng protein) gave the

optimum readings (data not shown). These concentrations were chosen for further

work.

222 A. Arul Selvi et al.

Dow

nloa

ded

by [

Her

iot-

Wat

t Uni

vers

ity]

at 2

1:22

07

May

201

6

Stability of anti-glyphosate antibody

Antibody stored at 48C and �208C was checked at different intervals of time for theamount of antibody activity. As shown in Figure 1, the antibody could be stored for

nearly six months at 48C. The loss in activity was 37% by the end of six months of

storage period at 48C. Thus the antibody (IgY) could be stored at 48C with little

activity loss up to six months. This reduces the cost of antibody production if stored

properly. The loss in activity for antibody stored at �208C was �2% by the end of six

months. Chicken egg yolk antibodies have been shown to have good activity up to 12

years (Schade et al., 2001) at �408C. It is clear that avian antibodies have longer half-

life than mammalian antibodies (IgG). These antibodies would be better choice tomammalian counterparts (Schade et al., 2001).

Calibration curves and sensitivity

Calibration curve was generated using glyphosate (99% pure) at concentrations from 2

to 1000 ng mL�1, prepared in carbonate buffer (Figure 2). The sensitivity of the assaywas determined by identifying the limit of detection defined as the lowest measurable

concentration of glyphosate that could be distinguishable from zero concentration. On

the basis of eight replicate measurements, the limit of detection was 2 ng mL�1. The

IC50 value of this avian antibody was 36.30 ng mL�1. In the similar studies by Clegg et

al. (1999) using IgG, the limit of detection was 0.076 mg mL�1 with an IC50 value of

1.54 mg mL�1 with anti-glyphosate rabbit antibody showing slightly less sensitivity.

Analytical performance of the assay

The glyphosate at different concentrations (2�250 ng mL�1) spiked to different food

samples was extracted using ultrapure water, and immunoassay was carried as

00

0.5

1

1.5

2

2.5

3

3.5

1 2 3 4 5 6Storage period (months)

O.D

. 405

4°C

–20°C

Figure 1. Stability of anti-glyphosate antibody.

Notes: Antibody stored at 48C and �208C was checked at different intervals of time for the

amount of residual antibody activity. Data points represent the mean of triplicate

determinations. The standard deviation in the mean is depicted by the error bars.

Food and Agricultural Immunology 223

Dow

nloa

ded

by [

Her

iot-

Wat

t Uni

vers

ity]

at 2

1:22

07

May

201

6

described in the experimental section. Each sample was analysed in triplicate for its

glyphosate content. The limit of detection of glyphosate in different food matrices

varied (Figure 3). The background effect of the glyphosate extract from turmeric, tea

and rice was very less. With chili and coffee as substrates for spiking, interference

could be observed. Ginger had tremendous interference with immunoassay. Linear

regression analysis of the results yielded a linear equation: y�0.9681x�0.3409,

R2 �0.9998 for turmeric, y�0.6618x�3.2148, R2 �0.9954 for chili, y�0.9792x�0.6683, R2 �0.9997 for tea (Assam), y�0.9825x�0.1669, R2 �0.9996 for coriander,

y�0.9638x�0.6991, R2 �0.9996 for coffee, y�0.9816x�0.1925, R2 �0.9998 for tea

(Ooty), y�0.9725x�0.6264, R2 �1 for rice and y�0.6733x�1.734, R2 �0.9895 for

ginger.

20

00 500 1000

40

60

80

100

120

y = 10.059x – 14.094R2 = 0.09649

Glyphosate concentration (ng mL–1)

Inh

ibit

ion

(%

)

Figure 2. Immunoassay with different concentrations of glyphosate.

Notes: The concentrations of the pesticide standard quoted on the X-axis (ng mL�1) were

diluted using carbonate buffer. Data were mean of eight independent experiments (n�8).

350

300

250

200

150

100

50

0

Turmeric Chili

Tea (Assam)

Tea (ooty)

RiceGinger

CorianderCoffee

Food materials

Co

nce

ntr

atio

n o

f g

lyp

ho

sate

(n

g)

2ng10ng100ng250ng

Figure 3. Immunoassay of glyphosate spiked to different food samples.

Notes: Data points represent the mean of three independent experiments (n�3). The standard

deviation in the mean is depicted by the error bars.

224 A. Arul Selvi et al.

Dow

nloa

ded

by [

Her

iot-

Wat

t Uni

vers

ity]

at 2

1:22

07

May

201

6

Specificity of the assay

The specificity of the assay for glyphosate was studied. The presence of other

pesticides at concentrations of 1000 ng mL�1 level was tested. Figure 4 shows the

cross-reactivity of different individual pesticides added in the assay. These data were

based upon triplicate determinations. The presence of any of the herbicide did not

interfere in the assay, i.e. the cross-reactivity was very low and it ranged between 1

and 5%. Glyphosine, a structurally related herbicide cross-reacted by 22% with anti-glyphosate antibody.

Comparison of immunoassay with high-pressure liquid chromatography (HPLC)

The mean analytical recovery was calculated as the ratio between the glyphosate

concentration found and the concentration added, and expressed as percentage.

Glyphosate-spiked food samples at 2�250 ng mL�1 levels were checked by both

immunoassay and HPLC as described in the experimental section. The comparisons

of these results are shown in Table 1. The recovery of glyphosate by immunoassay

(Table 1a) with turmeric as matrix was 85 and 95%, respectively, at 2 and 250 ngmL�1. Chili showed 88 and 60% recovery at 2 and 250 ng mL�1 level. This may be

due to the matrix effect of the sample. Glyphosate in tea could be recovered at 74 and

82%, respectively, at 2 and 250 ng mL�1. Glyphosate spiked to coriander could be

recovered by 53 and 80%, respectively, at 2 and 250 ng mL�1. Coffee-spiked

glyphosate had very low recovery of 9 and 56%, respectively, at 2 and 250 ng mL�1.

About 71 and 96% of the added glyphosate could be recovered from rice-spiked

samples. Ginger had the lowest recovery of 11 and 5%, respectively, at 2 and 250 ng

mL�1. This may be due to the interference of the matrix effect. The recovery ofglyphosate by HPLC (Table 1b) in turmeric was 91 and 96% at 2 and 250 ng mL�1

levels, respectively. In chili, it was 61 and 66%, respectively, for 2 and 250 ng mL�1.

In tea, coriander, coffee and rice, the recovery was almost complete. In ginger, only

52 and 66% of the added glyphosate could be recovered. The immunoassay method

1009080706050

Inh

ibit

ion

(%

)403020100

Glypho

sate

Atrazin

e de

seth

yl

Mon

ocro

toph

os2,

4-D

Parat

hion

Met

hyl P

arat

hion

Aceph

ate

Glypho

sine

Dichlor

ovos

Atrazin

e

Simaz

ine

Terbu

tryn

Diffrent pesticides used

Figure 4. Cross-reactivity/specificity of glyphosate immunoassay.

Notes: Few pesticides were tested for cross-reactivity to the anti-glyphosate antibody. Data

were based upon the mean of triplicate determinations. Competitive immunoassays were

performed as described in the experimental section using analytical grade standard pesticides.

Food and Agricultural Immunology 225

Dow

nloa

ded

by [

Her

iot-

Wat

t Uni

vers

ity]

at 2

1:22

07

May

201

6

compared well with HPLC, with few food materials indicating the ability of

immunoassay to accurately determine glyphosate in certain food samples.

Conclusion

A sensitive immunoassay for the determination of glyphosate in food samples has

been developed. The assay was made possible by using avian antibodies with high-

binding affinity for glyphosate. The matrix effect was found with few food samples.

Microwells immobilised with antigen and blocked could be stored at 48C. These

microwells could be used for routine assay. The assay involved two incubation steps

after which the plate was ready to generate the signal. The assay is very easy to

perform in a microwell plate, the number of wells to be chosen being dependent on

the number of samples. It is not time-consuming and an operator can analyse and

Table 1a. Comparison of immunoassay with HPLC � immunoassay data.

Immunoassay

with food

matrix

(ng/ml)

Concentration of glyphosate (ng/ml)

2 5 10 50 100 250

Turmeric 1.7090.20 4.5090.21 12.3290.02 48.4490.02 82.6090.13 238.3690.15

Chili 1.7691.74 3.8690.16 8.3192.24 46.3291.17 53.4893.24 149.8093.32

Tea (Assam) 1.4992.14 4.4693.20 7.9591.36 43.2190.21 84.3493.14 206.8191.13

Coriander 1.0790.52 4.2290.91 7.9690.94 46.8290.22 80.8293.20 200.3190.92

Coffee 0.1890.22 1.0290.31 1.3291.70 14.6493.20 30.3191.40 140.4093.60

Tea (Ooty) 1.0190.22 3.2891.50 6.3094.10 35.2093.60 70.8091.17 202.4093.95

Rice 1.4393.20 4.3592.10 8.6092.40 44.3190.22 83.4493.24 240.5891.50

Ginger 0.2291.50 0.3690.15 0.7690.22 4.1290.13 6.0291.74 13.8492.90

Notes: Samples were collected as described in experimental sections. Values were a mean of triplicatedeterminations 9 SD. Competitive immunoassays were performed as described in the experimentalsection.

Table 1b. Comparison of immunoassay with HPLC � HPLC data.

Concentration of glyphosate (ng/ml)

HPLC with

food matrix

(ng/ml) 2 5 10 50 100 250

Turmeric 1.8390.16 4.8590.21 9.5490.92 48.1591.36 100.0090.22 241.3891.74

Chili 1.2293.20 3.8290.91 8.3190.15 40.7492.24 75.8391.40 165.3691.50

Tea (Assam) 2.0090.22 4.7190.31 8.6293.20 45.6890.21 99.5891.17 243.7392.60

Coriander 1.8391.50 5.0090.22 9.0891.17 46.6190.22 101.6791.74 244.5190.52

Coffee 2.0992.10 4.8591.74 10.0091.32 43.5293.14 96.6791.91 240.6092.10

Tea (Ooty) 1.9192.20 4.7191.17 9.0892.20 47.5391.13 100.4292.20 244.5191.50

Rice 1.8391.70 4.7191.50 9.2391.80 47.2293.60 95.8392.14 242.9592.14

Ginger 1.0490.31 3.2491.13 6.3190.91 32.7291.40 82.5091.24 165.3691.16

Notes: Samples were collected as described in experimental sections. Values were a mean of triplicatedeterminations 9 SD. Competitive immunoassays were performed as described in the experimentalsection.

226 A. Arul Selvi et al.

Dow

nloa

ded

by [

Her

iot-

Wat

t Uni

vers

ity]

at 2

1:22

07

May

201

6

obtain results in less than two hours. It is able to detect glyphosate in food samples at

a concentration as low as 2 ng mL�1. The assay was comparable with HPLC and

future studies will include the exploitation of this approach to other food samples in

the presence of other contaminating pesticides. The stability of IgY being very good

at �208C, single batch preparation of antibodies could be used for a longer time.

Acknowledgements

The authors thank the Director, CFTRI, Mysore, for his encouragement and providingfacilities in the present work. Ms A. Arul Selvi is grateful to the Council of Scientific andIndustrial Research (CSIR) for Senior Research Fellowships. The Government of India isgratefully acknowledged for providing financial support under the Networking project(NWP). The authors are grateful to Dr Muthukumar for his help during immunisation ofpoultry and Mr Akmal Pasha for his help during preparation of conjugates.

References

Bauer, K.H., Knepper, T.P., Maes, A., Schatz, V., & Voihsel, M. (1999). Analysis of polarorganic micropollutants in water with ion chromatography-electrospray mass spectrometry.Journal of Chromatography A, 837, 117�128.

Bedry, R., & Parrot, F. (1998). Chromatography of amino-acids: A new method of dosage ofglyphosate (Round-up†). Archives des maladies professionnelles et de medicine, 59, 129�130.

Borjesson, E., & Torstensson, L. (2000). New methods for determination of glyphosate and(aminomethyl)phosphonic acid in water and soil. Journal of Chromatography A, 886, 207�216.

Bradford, M.M. (1976). A rapid and sensitive method for the quantitation of microgramquantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry,72, 248�254.

Campbell, A.M., Burdon, R.H., & Knippenberg, P.H. (1984). Assay technique. In R.H.Burdon & P.H. Knippenberg (Eds.) The production and characterization of rodent and humanhybridomas: In monoclonal antibody technology (Vol. 1). New York: Elsevier.

Clegg, B.S., & Ripley, B. (1996). The evaluation and implementation of rapid extraction andisolation techniques for analysis of glyphosate and its (AMPA) metabolite from complexmatrices (Document AF 0721). Ontario, Canada: Laboratory Services Branch, OntarioMinistry of Agriculture Food and Rural Affairs.

Clegg, B.S., Stephenson, G.R., & Hall, J.C. (1999). Development of an enzyme-linkedimmunosorbent assay for the detection of glyphosate. Journal of Agricultural and FoodChemistry, 47, 5031�5037.

Cox, C. (2000). Glyphosate factsheet. Journal of Pesticide Reform, 108(3 Fall 98).Duke, S.O. (1988). Glyphosate. In P.C. Kearney & D.D. Kaufman (Eds.) Herbicides chemistry,

degradation and mode of action (Vol. 3, pp 1�70). New York: Dekker.Gee, S.G., Miyamoto, T., Goodrow, M.H., Buster, D., & Hammock, B.D. (1988). Develop-

ment of an enzyme-linked immunosorbent assay for the analysis of the thiocarbamateherbicide molinate. Journal of Agricultural and Food Chemistry, 36, 863�870.

Hall, J.C., Deschamps, R.J.A., & McDermott, M.R. (1990). Immunoassays to detect andquantitate herbicides in the environment. Weed Technology, 4, 226�234.

Johnson, B.D., & Hall, J.C. (1996). Fluroxypyr and triclopyr-specific immunosorbent assays:Development and quantitation in soil and water. Journal of Agricultural and FoodChemistry, 44, 488�496.

Lawruk, T.S., Hottenstein, C.S., Fleeker, J.R., Hall, J.C., Herzog, D.P., & Rubio, F.M. (1994).Quantification of 2,4-D and related chlorophenoxy herbicides by a magnetic particle-basedELISA. Bulletin of Environmental Contamination Toxicology, 52, 538�545.

Mallat, E., & Barcelo, D. (1998). Analysis and degradation study of glyphosate and ofaminomethylphosphonic acid in natural waters by means of polymeric and ion-exchange

Food and Agricultural Immunology 227

Dow

nloa

ded

by [

Her

iot-

Wat

t Uni

vers

ity]

at 2

1:22

07

May

201

6

solid-phase extraction columns followed by ion chromatography-post-column derivatiza-tion with fluorescence detection. Journal of Chromatography A, 823(12), 129�136.

Mensink, H., & Janssen, P. (1994). Environmental Health Criteria 159. Geneva: Publishedunder the joint Sponsorship of the United Nations Environment Programme, theInternational Labour Organisation, and the World Health Organization.

Omafra staff. (2002). Guide to weed control (Publication 75). Ontario, Canada: OntarioMinistry of Agriculture and Rural Affairs.

Parnell, J.S., & Hall, J.C. (1998). Development of an enzyme-linked immunosorbent assay forthe detection of metolsulam. Journal of Agricultural and Food Chemistry, 46, 152�156.

Plapp, B.V., Moore, S., & Stein, W.H. (1971). Activity of bovine pancreatic desoxyribonucleaseA with modified amino groups. Journal of Biological Chemistry, 246, 939�945.

Reupert, R., & Schlett, C. (1997). Glyphosate into the Ruhr. Wasser Abwasser, 138, 559�563.Rubio, F., Linda, J., Veldhuis, B., Stephen Clegg, James R. Fleeker, & Christopher Hall, J.

2003. Comparison of a direct ELISA and an HPLC method for glyphosate determinationsin water. Journal of Agricultural and Food Chemistry, 51, 691�696.

Rubio, F.M., Itak, J.A., Scutellaro, A.M., Selisker, M.Y., & Herzog, D.P. (1991). Performance,characteristics of a novel magnetic particle-based enzyme-linked immunosorbent assay forthe quantitative analysis of Atrazine and related Triazines in water samples. Food andAgricultural Immunology, 3, 113�125.

Schade, R., Behn, I., Erhard, M., Hlinak, A., & Staak, C. (2001). Isolation of IgY from yolk.Chicken egg yolk antibodies production and application: IgY-Technology: Alternatives tolaboratory animals (ATLA) (Record number: 6893). Berlin, Heidelberg, New York:Springer lab manual. 29(3), 373�374.

Winfield, T.W. (1990). Method number 547. Determination of glyphosate in drinking water bydirect aqueous injection hplc, post-column derivatization and fluorescence detection.Cincinnati, Ohio: Technology Applications Inc., Environmental monitoring systemslaboratory office of research and development, US Environmental Protection Agency

Wauchope, R.D., Buttler, T.M., Hornsby, A.G., Augustijn Beckers, P.W.M., & Burt, J.P.(1992). SCS/ARS/CES pesticide properties database for environmental decision making.Environmental Contamination Toxicology, 123, 1�164.

Williams, G.M., Kroes, R., & Munro, I.C. (2000). Safety evaluation and risk assessment of theherbicide roundup and its active ingredient, glyphosate, for humans. Regulatory Toxicologyand Pharmacology, 31, 117�165.

Zhu, Y., Zhang, F., Tong, C., & Liu, W. (1999). Determination of glyphosate by ionchromatography. Journal of Chromatography A, 850, 297�301.

228 A. Arul Selvi et al.

Dow

nloa

ded

by [

Her

iot-

Wat

t Uni

vers

ity]

at 2

1:22

07

May

201

6