PAPER www.rsc.org/analyst | Analyst

Publ

ishe

d on

04

Aug

ust 2

010.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

22/1

0/20

14 1

1:00

:44.

View Article Online / Journal Homepage / Table of Contents for this issue

Rapid quantification of melamine in milk using competitive1,10-oxalyldiimidazole chemiluminescent enzyme immunoassay

JooHee Choi,ab Young-Teck Kimc and Ji Hoon Lee*b

Received 11th June 2010, Accepted 3rd July 2010

DOI: 10.1039/c0an00396d

A novel competitive 1,10-oxalyldiimidazole (ODI) chemiluminescent enzyme immunoassay (CLEIA)

was developed as a method for rapid and simple screening of melamine in milk. Fat existing in milk acts

as an inhibitor in the competitive binding interaction of melamine and anti-melamine in the presence of

melamine-conjugated horseradish peroxidase. Thus, the calibration curve and sensitivity of competitive

ODI CLEIA for the quantification of melamine in fat free milk were wider and better than those in milk

containing fat. However, a centrifuge is not a good method for removing the inhibitor because

a portion of the melamine is also removed with the fat. The incubation time (20 min) for the competitive

binding interaction of anti-melamine and melamine in 20% milk diluted with PBS buffer of pH 7.4 was

longer than that (10 min) in 100% milk even though the sensitivity of the former was better than latter.

The limit of detection (1.12 ppb) determined in rapid ODI CLEIA (dynamic range: 3.8–125 ppb) for the

quantification of melamine in 20% milk not containing fat was lower than those (6.3 and 9.0 ppb)

calculated in relatively time-consuming luminol CLEIA and enzyme-linked immunosorbent assay

(ELISA). Also, we expect that ODI-CLEIA (dynamic range: 62.5–2000 ppb) capable of directly

quantifying melamine in 100% milk without any pretreatment can be applied as a new and simple

method for rapid screening of melamine in milk.

Introduction

Melamine (2,4,6-triamino-1,3,5-triazine), a substance composed

of 66% nitrogen, was synthesized to use as an additive of

industrial products such as plastics, adhesives, countertops,

dishware, and whiteboards. Despite being a non-natural

product, the presence of melamine in various high-protein foods

such as infant formula, pet food, animal feed, and wheat gluten

has been reported since 2007. It is well-known that the contin-

uous intake of edible foods containing melamine cause critical

diseases such as stone in the kidney and bladder and epithelial

hyperplasia of urinary bladder.1 Thus, the addition of melamine

in food products is not approved by the World Health Organi-

zation (WHO), the US Food and Drug Administration (FDA),

and the European Food Safety Authority (EFSA).1 WHO

adopted the tolerable daily intake (TDI) of 0.2 mg kg-1 body

weight a day-1 for melamine. TDI of WHO is lower than that of

US FDA (0.6 mg kg-1 body weight a day-1) and EFSA (0.5 mg

kg-1 body weight a day-1).1

Several analytical methods for quantifying and screening

melamine have been reported.2–15 Melamine existing in various

materials such as tissue, pet food, and protein powders have been

quantified using high performance chromatography (HPLC) or

gas chromatography (GC) with UV1 or mass (MS) spectrom-

etry.14 The limit of detection (LOD) is dependent on the prop-

erties of analytical samples.2–15 Thus, the range of LOD

aLangley High School, 6520 Georgetown Pike, McLean, VA 22101, USAbLuminescent MD, LLC, 20140 Scholar Drive, Hagerstown, MD 21742,USA. E-mail: jhlee@luminescentmd.com; Fax: +1 301393 9092; Tel: +1301 393 9092cDepartment of Packaging Science, Clemson University, Clemson, SC29634, USA

This journal is ª The Royal Society of Chemistry 2010

determined using HPLC (or GC) with UV or MS spectrometry

have been wide (10 ppb � 200 ppm)1. HPLC-MS or GC-MS are

used for the quantification of melamine in foods by US FDA

laboratories.1 Recently, other analytical methods such as

surface-enhanced Raman spectroscopy (SERS),1 competitive

enzyme-linked immunosorbent assay (ELISA),2,3 and competi-

tive chemiluminescent enzyme immunoassay (CLEIA) using

luminol chemiluminescence (CL) detection13 have been devel-

oped for the quantification of melamine. LOD of SERS for the

screening of aqueous solution was 33 ppb.14 LOD of ELISA for

the monitoring of dog food was < 20 ppb.2 Also, LOD of

competitive CLEIA with luminol CL detection for the quantifi-

cation of melamine in milk was as low as 6.3 ppb.14

It is well-known that peroxyoxalate chemiluminescence (PO-

CL) detection is more sensitive and selective than other detec-

tions such as absorbance, electrochemical, fluorescence, and

luminol chemiluminescence widely applied in enzyme immuno-

assay (EIA).16 Unfortunately, oxalate esters such as bis(2,4-

dinitrophenyl) oxalate (DNPO) and bis(2,4,6-trichlorophenyl)

oxalate (TCPO), one of PO-CL reagents, are too unstable in

aqueous solution to be applied as CL reagents for highly sensitive

CLEIA system. Recently, chemical and physical properties of

1,10-oxalyldimidazole (ODI) derivatives formed from the reac-

tion between oxalate esters and imidazole derivatives were

studied.17,18 ODI derivatives are also unstable in aqueous solu-

tion. However, it was confirmed that ODI CL detection system in

aqueous solution is highly sensitive because ODI CL reaction is

much faster than the decomposition rate of ODI in aqueous

solution.19 This result indicated that it was possible to develop

a new CLEIA with ODI CL detection.

Using the advantages of ODI derivative CL, we developed

a rapid and simple CLEIA with ODI CL detection capable of

Analyst, 2010, 135, 2445–2450 | 2445

Fig. 1 Procedures of ELISA, Luminol CLEIA and ODI CLEIA.

Publ

ishe

d on

04

Aug

ust 2

010.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

22/1

0/20

14 1

1:00

:44.

View Article Online

rapidly quantifying and screening melamine in milk. Trace levels

of melamine in diluted and undiluted milk were quantified using

ODI CLEIA. In this paper, we describe in detail the character-

istics and advantages of the new CLEIA.

Experimental

Chemical and materials

Melamine, 4-methylimidazole, and phosphate buffered saline

(PBS) with 0.05% tween-20 were purchased from Sigma. 1 � 8

strip-well coated with rabbit anti-melamine and melamine-

conjugated horseradish peroixdase (HRP) were purchased from

Beacon Analytical Systems, Inc. was purchased from EMD.

Bis(2,4,6-trichlorophenyl) oxalate (TCPO) was purchased from

TCI America. Ethyl acetate (HPLC grade), water (LC/Mass

grade), and Isopropyl alcohol (ACS grade) were purchased from

J. T Baker. 10 � phosphate buffer saline solution (PBS, pH 7.4)

was from EMD. 3.0% hydrogen peroxide was purchased from

VWR. Amplex red and resorufin were purchased from AnsSpec,

Inc. Dimethyl Sulfoxide (DMSO) was purchased from Calbio-

chem. Fat free milk, 2% fat milk, and whole milk were purchased

from a local food market.

Methods

Preparation of standard solutions. Stock solution of melamine

(2000 ppm) was prepared with deionized water. It was stored

under ambient condition. Using the stock solution, 9 melamine

working solutions (0.00, 0.98, 3.90, 7.81, 15.65, 31.25, 125.00,

250.00, and 500.00 ppm) were prepared daily in PBS solution.

Preparation of diluted standard solution. 20 ml of melamine

working solution was spiked into a 1.5 ml - microcentrifuge tube

containing 980 ml of milk. Some of the microcentrifuge tubes

containing different concentrations of melamine were centri-

fuged under 2000 rpm for 10 min at room temperature (21.0 �2.0 �C). The rest were used without any centrifuging. 200 ml of fat

free milk serum obtained from the centrifugation was mixed with

800 ml of PBS buffer of pH 7.4 in a microcentrifuge tube. Also,

200 ml of milk, the mixture of protein and fat, was added in

a microcentifuge tube and mixed with 800 ml of PBS solution.

Based on the procedure described above, 10 standard solutions

containing different concentrations of melamine (0, 3.9, 15.7,

31.3, 62.5, 125, 250, 500, 1000, and 2000 ppb) were prepared.

Preparation of undiluted standard solutions. 20 ml of melamine

working solution was spiked into three 1.5 ml - microcentrifuge

tubes, each containing 980 ml of milk with different fat content

(fat free milk, 2% milk, and whole milk). 200 ml of milk con-

taining melamine was diluted with 800 ml of milk instead of PBS

buffer of pH 7.4.

Preparation of the mixture of Amplex Red and H2O2 as

a substrate of ODI CLEIA. 5.0 mg of Amplex Red was dissolved

in 5.0 ml of DMSO to prepare a stock solution. 500 ml of the

stock solution was added in a 2ml-glass vial. 10 vials containing

the stock solution were stored in a freezer (- 20.0 �C). Using 3%

H2O2 solution and PBS, 20 mM H2O2 stock solution was

prepared daily. In order to prepare the substrate used in ODI

2446 | Analyst, 2010, 135, 2445–2450

CLEIA, 25 ml of Amplex Red stock solution, 100 ml of H2O2

stock solution, and 4875 ml of PBS buffer were added in a 20 ml-

amber glass vial.

Design of ODI CLEIA. As shown in Fig. 1, the procedure of

ODI CLEIA for the quantification of melamin is similar to that

of other EIAs such as ELISA and Luminol CLEIA. However,

the incubation time for the competitive reaction between mela-

mine and melamine-conjugated HRP with anti-melamine coated

on the surface of the well in ODI CLEIA is shorter than those of

other EIAs (ELISA and Luminol CLEIA) because ODI CLEIA

is more sensitive. Also, the incubation time of the substrate

(Amplex Red) added in the well for ODI CLEIA is also shorter

than that for ELISA or same as that for Luminol CLEIA.

Measurement of light emitted from ODI CLEIA. 1.0 mM

TCPO was prepared daily in ethyl acetate as a stock solution.

Also, 10.0 mM 4MImH was prepared in ethyl acetate as a stock

solution. They were stored under ambient condition. 200 ml of

TCPO stock solution and 100 ml of 4MImH stock solution were

mixed with 39.7 ml of ethyl acetate in a 40 ml - amber glass vial.

ODI was rapidly formed from the reaction of TCPO and

4MImH in the vial at room temperature. 80 mM H2O2 was

prepared in isopropyl alcohol as a working solution. Two 40 ml-

glass vials containing ODI or H2O2 solution were placed in the

This journal is ª The Royal Society of Chemistry 2010

Fig. 2 Quantification of resorufin using ODI CL detection. Condition:

1.0 mM TCPO and 4.0 mM 4-methylimidazole (4MImH) dissolved in

ethyl acetate were used to produce ODI. 0.08 M H2O2 was prepared in

isopropyl alcohol.

Scheme 1 Quantification of resorufin formed from the reaction of

Amplex Red and H2O2 in the presence of HRP using ODI CL detection.

1. TCPO, 2. 4MImH, 3. ODI, 4. Amplex Red, 5. Resorufin in ground

state, 6. Resorufin in excited state, X: high-energy intermediate capable of

transferring energy to resorufin

Publ

ishe

d on

04

Aug

ust 2

010.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

22/1

0/20

14 1

1:00

:44.

View Article Online

reagent holder of Lumat LB Luminometer 9507 to inject the two

solutions into a detection cell (12 � 75 mm test tube) through

two dispensers.

As shown in Fig. 2, 100 ml of the substrate used in ODI CLEIA

was added into the washed well. 10 ml of resorufin formed from

the reaction between Amplex Red and H2O2 in the presence of

melamine-conjugated HRP bound with rabbit anti-melamine in

the well for 5.0 min was transferred into a 12 � 75 mm test tube.

The test tube was placed in the sample holder of Lumat LB

Luminometer 9507.

When the start button of the luminometer was pressed, the

sample holder of Lumat LB 9507 Luminometer turned to the

detection area. Then, 25.0 ml of H2O2 and 25.0 ml of ODI were

injected into the detection cell through two dispensers at 0.7 s

intervals. When ODI was inserted into the detection cell, relative

CL intensities emitted from the detection cell were integrated

immediately for 0.5 s.

Fig. 3 CL spectrum obtained in competitive ODI CLEIA for the

quantification of 62.5 ppb melamine. Condition: [TCPO] ¼ 5.0 mM,

[4MImH] ¼ 25.0 mM, [H2O2] ¼ 80 mM.

Results and discussion

Quantificationi of resorufin

Amplex Red (10-acetyl-3,7-dihydroxyphenoxazine) is widely

used to determine the concentration of horseradish peroxidase

(HRP) in fluorescence enzyme assay.20 Amplex Red is converted

to resorufin having high quantum efficiency when Amplex Red

reacts with H2O2 in the presence of HRP. In other words, the

yield of resorufin formed in this reaction depends on the activity

of HRP when the concentrations of Amplex Red and H2O2 are

constant. As shown in Fig. 2, we confirmed that low concen-

trations of resorufin dissolved in water are quantified in ODI CL

reaction. The detection limit (signal/noise ¼ 3.0) of ODI CL

analytical system for the quantification of resorufin was as low as

0.73 nM. The dynamic range of linear calibration curve (R2 ¼0.9952) was wide (3.3 � 141 nM). However, the relative CL

intensity of higher concentration of resorufin than 141 nM was

self-quenched. Fig. 1 indicates that trace levels of resorufin

This journal is ª The Royal Society of Chemistry 2010

formed from the reaction between Amplex Red and H2O2 in the

presence of HRP can be quantified with ODI CL detection

system developed based on the reaction mechanism shown in

Scheme 1.

Concentrations of CL reagents for competitive ODI CLEIA

As shown in Fig. 3, the maximum intensity measured in ODI CL

reaction was reached as quickly as 0.4 s. Thus, light emitted in

ODI CLEIA was integrated for 0.5 s. When higher concentrations

of TCPO and 4MImH than those used for obtaining Fig. 3 were

prepared, relative CL intensities measured in the absence of

melamine and in the presence of low concentrations of melamine

in ODI CLEIA were too high to measure. On the other hand,

Analyst, 2010, 135, 2445–2450 | 2447

Publ

ishe

d on

04

Aug

ust 2

010.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

22/1

0/20

14 1

1:00

:44.

View Article Online

relative CL intensities measured in the presence of high concen-

trations of melamine were too low when lower concentrations of

TCPO and 4MImH than those shown in Fig. 3 were used in

competitive ODI CLEIA.

Relative CL intensity was increased with the increase of H2O2

concentration. However, relative CL intensities measured in the

presence of higher (up to 400 mM) concentrations than 80 mM

H2O2 were slightly higher than that shown in Fig. 3. Thus, 80

mM H2O2 was used to develop competitive ODI CLEIA capable

of detecting trace levels of melamine in milk.

Effect of incubation time on the competitive binding of melamine

and melamine-conjugated HRP with anti-melamine

As shown in Fig. 4, the sensitivity of competitive ODI CLEIA

depends on the incubation time for the competitive binding of

melamine and melamine-conjugated HRP with anti-melamine.

Fig. 4 indicates that the binding between melamine-conjugated

HRP and anti-melamine in the well is faster and more predomi-

nant than that between melamine and anti-melamine. Thus, the

dynamic range (3.9� 125 ppb) of calibration curves obtained after

20 and 30 min of incubation was wider than that (3.9 � 31.3 ppb)

determined after 10 min of incubation because additional mela-

mine molecules were able to bind with anti-melamine with the

increase of incubation time. Based on the results shown in Fig.4,

we selected 20 min as an appropriate incubation time in compet-

itive ODI CLEIA for the quantification of melamine in milk.

Effect of incubation time of Amplex Red and H2O2 in the well

As shown in Scheme 1 and Fig. 3, resorufin is formed from the

reaction of Amplex Red and H2O2 in the well containing mela-

mine-conjugated HRP bound with rabbit anti-melamine. The

yield of resorufin in this reaction depends on the incubation (or

reaction) time of substrates and melamine-conjugated HRP

bound with rabbit anti-melamine in the well. In other words,

relative CL intensity of resorufin measured after 20 min of

incubation was about 7 times higher than that observed after

Fig. 4 Three calibration curves obtained with different incubation time

on the competitive binding of melamine and melamine-conjugated HRP

with anti-melamine.

2448 | Analyst, 2010, 135, 2445–2450

5 min of incubation. However, the dynamic range (7.8�125 ppb)

of linear calibration curve obtained after the 20-minute incuba-

tion was shorter than that (3.9�125 ppb) after the 5-minute

incubation. This is because, during the 20-minute incubation, the

relative CL intensity of saturated resorufin formed from the

reaction of Amplex Red and H2O2 in the presence of lower

concentration of melamine than 7.8 ppb is self-quenched. This

result is consistent with that in Fig. 1, which shows that the

relative CL intensity of higher concentration of resorufin than

140 nM is self-quenched. Based on the results, the 5-minute

incubation of Amplex Red and H2O2 for the quantification of

melamine in milk was selected. The limit of detection (LOD¼ I0 -

3s) determined in this condition was 1.12 ppb. I0 is the relative

CL intensity measured in the absence of melamine in milk. s is

the standard deviation for the average of I0 (n ¼ 20). LOD of

competitive ODI CLEIA is lower than those (6.3 and 9.0 ppb) of

competitive luminol CLEIA13 and competitive ELISA.2

Recovery test

In order to study the effect of fat existing in commercial milk in

competitive ODI CLEIA for the quantification of melamine, 62.5

ppb melamine was spiked in fat free milk, 2% milk, and whole

milk, respectively, purchased from a local food market. Then,

melamine in each of the three types of milk was quantified using

competitive ODI CLEIA. As shown in Table 1, the recovery of

melamine spiked in milk is dependent on the amount of fat

existing in milk. With the increase of fat in milk, the recovery of

melamine was reduced because fat acts as an inhibitor while

melamine interacts with rabbit anti-melamine in the presence of

melamine-conjugated HRP. Thus, each of the three types of milk

containing 62.5 ppb melamine was centrifuged at 2000 rpm for 10

min to remove fat in the milk. Unfortunately, the recovery of

melamine in milk centrifuged wasn’t as good as that not centri-

fuged. This result indicates that parts of melamine spiked in milk

are removed with fat during the centrifuging. Thus, the three

types of milk containing different concentrations of fat should

not be centrifuged for the detection of trace levels of melamine.

Also, the results shown in Table 1 indicate that three different

linear calibration curves obtained with fat free milk, 2% milk,

and whole milk can accurately quantify melamine existing in

each type of milk.

Table 1 Recovery test of melamine (62.6 ppb) spiked in fat free milk, 2%milk, and whole milk not centrifuged and centrifuged at 2000 rpm for 10min

Centrifuging

Fat Free milka 2% milka Whole milka

Cb R (%)c Cb R (%)c Cb R (%)c

No 62.54 100.0 56.1 89.6 53.5 85.5Yes 60.5 96.6 44.1 70.4 41.2 65.8

a Each milk sample was diluted 5 times with PBS (pH 7.4). b Theconcentration of melamine (ppb) for each of milk sample wasquantified with a calibration curve obtained using melamine standards(3.9�125 ppb) spiked in fat free milk diluted 5-fold in PBS (pH 7.4)and not centrifuged. c Recovery. d Standard used to obtaina calibration curve in this research.

This journal is ª The Royal Society of Chemistry 2010

Publ

ishe

d on

04

Aug

ust 2

010.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

22/1

0/20

14 1

1:00

:44.

View Article Online

Intra-assay and inter-assay

The results shown in Table 2 indicate that competitive ODI

CLEIA for the quantification of melamine existing in fat free

milk is precise within acceptable statistical error range (confi-

dence variable, CV). Also, the average values obtained in the

intra-assay and inter-assay shown in Table 2 indicate that

competitive ODI CLEIA has good quantitative accuracy. In

order to study the inter-assay of competitive ODI CLEIA, a total

of 63 milk samples were prepared with each set of 21 samples

containing 15.6, 62.5, or 125 ppb of melamine, respectively. All

the samples were prepared with fat free milk diluted 5-fold with

PBS buffer of pH 7.4. The prepared samples were stored in

a freezer (- 20.0 �C). Three sets (n ¼ 3) of the three milk samples

containing different concentrations of melamine were quantified

every day for a week.

Competitive binding of melamine and melamine-conjugated HRP

with anti-melamine

The results shown in Table 3 indicate that the binding between

melamine and anti-melamine in 40.0% milk sample diluted

2.5-fold with PBS is faster than that in lower percentage milk

sample than 40.0%. Thus, the concentration of melamine in

40.0% milk sample determined using competitive ODI CLEIA

looks higher than that in the lower percentage milk sample even

though each of the three types of milk samples shown in Table 3

equally contains 62.5 ppb of melamine. In other words, the

concentration of melamine existing in 5.0% milk sample deter-

mined using competitive ODI CLEIA looked to be lower than

that in a higher percentage milk sample even though all four milk

samples containing 62.5 ppb of melamine were incubated in the

well for 20 min to bind with anti-melamine in the presence of the

Table 2 Intra-assay (n ¼ 7) and inter-assay (n ¼ 21) of competitive ODICLEIA for the quantification of melamine in fat free milk not centrifuged

Spiked (ppb)

Intra-assay Inter-assay

Measured (ppb) CV (%) Measured (ppb) CV (%)

15.6 15.4 2.55 15.3 4.6862.5 62.3 1.58 62.7 5.13125.0 125.5 1.77 125.4 4.38

Table 3 Quantification of melamine (62.5 ppb) spiked in milk dilutedfrom 2.5 to 20 times with PBS (pH 7.4)

Milk (%) a CL b Concentration c Recovery (%)

5.0 1.61 19.98 31.9110.0 1.35 32.23 51.4820.0 1.00 62.50d 100.0040.0 0.79 91.62 145.31

a Fat free milk was diluted from 2.5 to 20 times with PBS (pH 7.4).b Relative CL intensity measured for each type of milk sample wasnormalized with that measured with the 5-fold diluted milk sample.c The concentration of melamine (ppb) existing in each of milk samplewas quantified with a linear calibration curve obtained using melaminestandards (3.9�125 ppb) spiked in fat free milk diluted 5-fold in PBS(pH 7.4) and not centrifuged. d Standard used to obtain a linearcalibration curve in this research.

This journal is ª The Royal Society of Chemistry 2010

same concentration of melamine-conjugated HRP. Thus,

recovery of melamine spiked in 5.0 and 10.0% milks was less than

that of melamine spiked in 20% milk. Inversely, the recovery of

melamine spiked in 40.0% milk was 145.31%.

In conclusion, the results shown in Table 3 indicate that trace

levels of melamine existing in undiluted milk can be quantified

rapidly with the reduction of incubation time necessary for

binding melamine with anti-melamine.

Quantification of melamine in undiluted milk

Based on the results shown in Table 3, we were able to rapidly

quantify relatively high concentrations of melamine spiked in

undiluted milk with short incubation time (10.0 min) for binding

melamine with anti-melamine in competitive ODI CLEIA. The

incubation time necessary for binding melamine with rabbit anti-

melamine in competitive ODI CLEIA for the quantification of

melamine in undiluted milk was a half of that in 20% milk. In

addition, the three different calibration curves shown in Fig. 5

indicate that the amount of fat existing in milk is an important

factor in determining the sensitivity of competitive ODI CLEIA

using undiluted milk. This result is the same as that observed

with diluted milk with PBS (see Table 1). Thus, Table 4 shows

that the linear dynamic range of calibration curve capable of

quantifying melamine in fat free milk is wider than that in 2.0%

milk and in whole milk, because the fat in the milk acts as an

interference in competitive ODI CLEIA. The competitive ODI

CLEIA for the quantification of melamine in undiluted milk is

faster, more sensitive, and simpler than other conventional

analytical methods1,14 such as HPLC/MS/MS and GC/MS. Thus,

we expect that the competitive ODI CLEIA can be applied for

Fig. 5 Calibration curve for the quantification of melamine in undiluted

milk containing three different % concentrations of fat.

Table 4 Dynamic range of calibration curve for the quantification ofmelamine existing in undiluted fat free milk, 2.0% milk, and whole milk

Milk Dynamic range (ppb) Equation R2

Fat free 62.5–2000 y ¼ �6169ln(x) + 60619 0.9962.0% 125–1500 y ¼ �9669ln(x) + 88803 0.996Whole 250–1000 y ¼ �13878ln(x) + 127439 0.998

Analyst, 2010, 135, 2445–2450 | 2449

Publ

ishe

d on

04

Aug

ust 2

010.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

22/1

0/20

14 1

1:00

:44.

View Article Online

the rapid screening of melamine in undiluted milk. As shown in

Fig. 5, however, the sensitivity of competitive ODI CLEIA for

the quantification of melamine in undiluted milk isn’t as good as

that for the quantification of melamine in 20% milk with PBS.

Conclusion

Rapid and simple competitive ODI CLEIA capable of detecting

trace levels of melamine in milk was developed. The novel

competitive ODI CLEIA is more sensitive than other EIAs such as

competitive ELISA and competitive luminol CLEIA. Also, we

reported for the first time that melamine in milk sample containing

a certain amount of fat should be quantified and monitored with

a calibration curve obtained with milk containing the same

amount of fat existing in the milk sample. Using competitive ODI

CLEIA, we also confirmed that melamine in milk sample can be

directly quantified without the centrifuging of milk sample and the

dilution of milk sample with a buffer solution. Finally, we expect

that the novel CLEIA with ODI CL detection can be applied in

a wide variety of research fields such as biochemistry, clinical

chemistry, environmental science and engineering, pathology and

toxicology.

Acknowledgements

This research was performed based on the intern program (LST-

2009-5) of Luminescent MD, LLC.

2450 | Analyst, 2010, 135, 2445–2450

Notes and references

1 Y.-C. Tyan, M.-H. Yang, S.-B. Jong, C.-K. Wang and J. Shiea, Anal.Bioanal. Chem., 2009, 395, 729.

2 E. A. E. Garber, J. Food Prot., 2008, 71, 590.3 H. Lei, Y. Shen, L. Song, J. Yang, O. P. Chevallier, S. A. Haughey,

H. Wang, Y. Sun and C. T. Elliott, Anal. Chim. Acta, 2010, 665, 84.4 I.-L. Tsai, S.-W. Sun, H.-W. Liao, S.-C. Lin and C.-H. Kuo, J.

Chromatogr., A, 2009, 1216, 8296.5 J. Rima, M. Abourida, T. Xu, I. K. Cho and S. Kyriacos, J. Food

Compos. Anal., 2009, 22, 689.6 J. Yu, C. Zhang, P. Dai and S. Ge, Anal. Chim. Acta, 2009, 651, 209.7 B. Kim, L. B. Perkins, R. J. Bushway, S. Nesbit, T. Fan, R. Sheridain

and V. Greene, J. AOAC Int., 2008, 91, 408.8 L. Chen, Q. Zeng, X. Du, X. Sun, X. Zhang, Y. Xu, A. Yu, H. Zhang

and L. Ding, Anal. Bioanal. Chem., 2009, 395, 1533.9 Q. Cao, H. Zhao, L. Zeng, J. Wang, R. Wang, X. Qiu and Y. Hea,

Talanta, 2009, 80, 484.10 J. Wang, L. Jianga, Q. Chu and J. Ye, Food Chem., 2010, 121, 215.11 Z. Wang, D. Chen, X. Gao and Z. Song, J. Agric. Food Chem., 2009,

57, 3464.12 X. Wang and Y. Chen, J. Chromatogr., A, 2009, 1216, 7324.13 C. Zhai, W. Qiang, J. Sheng, J. Lei and H. Ju, J. Chromatogr., A,

2010, 1217, 785.14 J. Choi, Y.-T. Kim, J. H. Lee, Anal. Chim. Acta., submitted.15 H. Miao, S. Fan, Y.-N. Wu, L. Zhang, P.-P. Zhou, J.-G. LI, H.-

J. Chen and Y.-F. Zhao, Biomed. Environ. Sci., 2009, 22, 87.16 M. Tsunoda and K. Imai, Anal. Chim. Acta, 2005, 541, 13.17 J. H. Lee, J. C. Rock, S. B. Park, M. A. Schlautman and

E. R. Carraway, J. Chem. Soc., Perkin Trans. 2, 2002, 802.18 J. H. Lee, J. T. Je, M. A. Schlautman and E. R. Carraway, Chem.

Commun., 2003, 270.19 J. H. Lee, J. Je, J. Hur, M. A. Schlautman and E. R. Carraway,

Analyst, 2003, 128, 1257.20 V. Towne, M. Will, B. Oswald and O. J. Zhao, Anal. Biochem., 2004,

334, 290.

This journal is ª The Royal Society of Chemistry 2010