toxins Article
Development of a Sensitive Enzyme-Linked Immunosorbent Assay and Rapid Gold Nanoparticle Immunochromatographic Strip for Detecting Citrinin in Monascus Fermented Food Shih-Wei Wu 1 1 2 3 4
*
ID
, Yao-An Yu 2 , Biing-Hui Liu 3, * and Feng-Yih Yu 2,4, *
ID
Graduate Institute of Medicine, Chung Shan Medical University, Taichung 402, Taiwan;
[email protected] Department of Biomedical Sciences, Chung Shan Medical University, Taichung 402, Taiwan;
[email protected] Graduate Institute of Toxicology, College of Medicine, National Taiwan University, Taipei 10051, Taiwan Department of Medical Research, Chung Shan Medical University Hospital, Taichung 402, Taiwan Correspondence:
[email protected] (B.-H.L.);
[email protected] (F.-Y.Y.); Tel.: +886-4-24730022-11816 (F.-Y.Y.)
Received: 12 August 2018; Accepted: 28 August 2018; Published: 2 September 2018
Abstract: Antibodies against citrinin (CTN) were generated from rabbits, which were injected with CTN-keyhole limpet hemocyanin (KLH). This work involved the development of a sensitive competitive direct enzyme-linked immunosorbent assay (cdELISA) and a rapid gold nanoparticle immunochromatographic strip (immunostrip) method for analyzing CTN in Monascus-fermented food. CTN at a concentration of 5.0 ng/mL caused 50% inhibition (IC50 ) of CTN-horseradish peroxidase (CTN-HRP) binding to the antibodies in the cdELISA. The capable on-site detection of CTN was accomplished by a rapid antibody-gold nanoparticle immunostrip with a detection limit of 20 ng/mL and that was completed within 15 min. A close inspection of 19 Monascus-fermented foods by cdELISA confirmed that 14 were contaminated with citrinin at levels from 28.6–9454 ng/g. Further analysis with the immunostrip is consistent with those results obtained using cdELISA. Both means are sensitive enough for the rapid examination of CTN in Monascus-fermented food products. Keywords: Citrinin; ELISA; gold nanoparticle immunochromatographic strip; Monascus fermented food Key Contribution: A specific polyclonal antibody against CTN was produced and successfully used in a sensitive ELISA and a rapid immunochromatographic strip for analyzing CTN in Monascus-fermented foods.
1. Introduction Citrinin (CTN) is a secondary metabolite synthesized from Aspergillus and Penicillium and contaminates corn, wheat, barley and rice [1,2]. CTN is also a secondary metabolite that is synthesized by Monascus spp. (red yeast), which is a fungal strain that is traditionally used in brewing rice. Monascus-fermented products are widely added to meat and wine as a food additive pigment. In western countries, fermentation products developed by Monascus are considered health food and are widely used to prevent cardiovascular diseases [3]. Recently, the Taiwan Food and Drug Administration (Taiwan-FDA) reported that 58 of 84 commercially available Monascus-fermented rice samples collected in 2009–2012 were contaminated with 0.4–93.5 ppm of citrinin, and the contamination of citrinin in fermented rice is very serious [4]. Therefore, the toxicity and content of CTN in Monascus health Toxins 2018, 10, 354; doi:10.3390/toxins10090354
www.mdpi.com/journal/toxins
Toxins 2018, 10, 354
2 of 14
foods has become a public concern. In 1986, the IARC (International Agency for Research on Cancer) categorized CTN as a group 3 carcinogen but with insufficient evidence of its carcinogenicity [5]. In Taiwan, the maximum allowable level of CTN in red yeast rice, Monascus products, and Monascus colors are 5 ppm, 2 ppm and 0.2 ppm, respectively. Japan set a limit of 0.2 ppm CTN in red fermented products [4]. Currently, high-performance liquid chromatography with fluorescence detection (HPLC-FL) and liquid chromatography with tandem mass spectrometry (LC/MS/MS) are used to measure CTN levels in food [6–10]. The advantages of HPLC and LC/MS/MS methods are their high accuracy and reproducibility, but they have drawbacks, for example, they require expensive equipment, qualified personnel, and time-consuming sample preparation. Thus, analyzing large amounts of samples in a short time using these methods is impossible. Therefore, rapid, easy-to-use, immunochemical methods for detecting CTN in fermented foods have been developed. [11–14]. However, most immunochemical methods cannot be used for on-site testing attributable to their multiplex washing steps, long reaction time and the need for expensive instrumentation. Hence, on-site detection methods that are rapid, stable, reliable and affordable must be developed. The advantages of gold nanoparticles are their high stability and high absorption coefficient. Gold nanoparticles can be synthesized promptly and can be conjugated easily with an antibody [15,16]. Therefore, gold nanoparticles are widely used in biosensor and immunochromatographic strips (immunostrip) for detecting mycotoxins [17,18]. In an immunostrip, the test samples and gold nanoparticle-antibody conjugates migrate by capillary action onto a nitrocellulose membrane where they interact with the test and the control line. The colored gold nanoparticle-antibody conjugates offer a quick on-site testing in less than 15 min without additional equipment and provide visible results [19,20]. Previous reports have shown that many samples from diverse regions are contaminated with various levels of CTN [3,21–23]. In this investigation, a specific polyclonal antibody against CTN was produced and successfully used in a sensitive ELISA and a gold nanoparticle immunostrip for analyzing CTN in Monascus-fermented food. 2. Results 2.1. Production and Characterization of Antibodies New Zealand rabbits received four immunizations with CTN-KLH conjugates as immunogens that would cause them to generate antibodies for citrinin. Both competitive indirect and direct ELISAs were then utilized to ascertain whether the antisera yielded antibodies specific to citrinin. The antiserum titer reached its highest level in the 70th week. Therefore, antiserum from the 70th week was conducted in the following study. The concentration that caused 50% inhibition of the binding of antibodies to the solid-phase citrinin—ovalbumin (CTN-OVA) by free citrinin (IC50 ) was 4.6 ng/mL in the ciELISA (Figure 1A). The concentration that caused 50% inhibition of the binding of CTN-HRP to the antibodies by citrinin (IC50 ) was calculated as 5.0 ng/mL in the cdELISA. The detection limit of citrinin (IC10 ) was calculated to be 0.2 ng/mL (Figure 1B) in accordance of a 90% confidence interval at 10% of inhibition of binding of the CTN-HRP conjugate. This cdELISA was further used to test whether the antibodies exhibited cross-reactivities with other mycotoxins. According to Figure 1B, mycotoxins, for example, ochratoxin A (OTA) and 1-hydroxy-2-naphthoic acid whose structures are similar to that of citrinin, showed weak cross-reactivity with the antibodies. The concentration that caused 50% inhibition of the binding of antibodies to the CTN-HRP by 1-hydroxy-2-naphthoic acid and OTA exceeded 1000 ng/mL (Figure 1B). Other mycotoxins, including patulin, zearalenone, aflatoxin and deoxynivalenol at a concentration 1000 ng/mL, could not inhibit the antibodies bind to the CTN-HRP conjugate in the cdELISA (data not shown).
Toxins 2018, 10, x FOR PEER Toxins 2018, 10, x FOR PEER REVIEW Toxins 2018, 10, REVIEW x FOR PEER REVIEW Toxins 2018, 10, x FOR PEER REVIEW 018, 10, x FOR PEER REVIEW Toxins 2018, 10, 354
20
60 40 20
80 60 40 20
0 0 0.01 0.01
0 0.01
0.1
0 0.10.1 0.011 1 00.110 10 1 100100 500 10500 0.01 0.1 1 1 Citrinin 10concentration, 100 500 ng/mL Citrinin concentration, ng/mL
80 60 40 20 100 10 0 0.01
100
80 80
80
60 60 40 40 20 20
% of Binding
20 20
80
100100
% of Binding
40 40
100
% of Binding % of Binding
40
60 60
100
% of Binding
60
80 80
% of Binding
% of Binding % of Binding
% of Binding
80
100
% of Binding
100100 100
3 of 1414 3 of
3 of 14
3 of 14 3 of 14
3 of 14
60 40 20
0 0 0.01 0.01 500 100 500 0.1 1
100 80 60 40 20
0 0.10.1 0.01 1 1 00.1 10 10 1100100 10 1000 1000 100 0.01 0.1 1 10 10 100 1000
1000 100
1000
Citrinin concentration, ng/mL Citrinin concentration, ng/mL Citrinin concentration, ng/mL Citrinin concentration, ng/mL Citrinin concentration, ng/mL Citrinin concentration, ng/mL Citrinin concentration, ng/mL Citrinin concentration, ng/mL
(A) (A)
(A)
(A)
(A)
(B) (B)
(B)
(B)
(B)
Figure ofofcitrinin () inina aciELISA. Cross-reactivity ofofthe Figure1. 1.(A). (A).The The standard curve citrinin ciELISA. (B). Cross-reactivity theCTN Figure 1.standard (A). Thecurve standard curve of() () ina a(B). ciELISA. Cross-reactivity ofCTN the polyclonal CTN Figure 1. (A). The standard curve ofcitrinin citrinin ( ) in ciELISA. (B).(B). Cross-reactivity ofCTN the Figure 1. (A). The standard curve of citrinin () in a ciELISA. (B). Cross-reactivity of the CTN antibody with CTN (), A AA (■(),■ acid asacid polyclonal antibody with CTN (), and acid() as gure 1. polyclonal (A). The standard curve of citrinin () in a(), ciELISA. (B). Cross-reactivity of the CTN antibody with CTN ( ochratoxin ),ochratoxin ochratoxin ),and and 1-hydroxy-2-naphthoic acid ( () ) as determined polyclonal antibody with CTN ochratoxin A1-hydroxy-2-naphthoic (1-hydroxy-2-naphthoic ■), and 1-hydroxy-2-naphthoic () as by a polyclonal antibody with CTN (), ochratoxin A (■), and 1-hydroxy-2-naphthoic acid () as bybyacdELISA. were calculated by the average ofthree sets The were calculated by the average sets of experiments. The absorbance determined acdELISA. cdELISA. Data were by the average ofthree three sets ofexperiments. experiments. The lyclonaldetermined antibody with CTN (), ochratoxin Acalculated (■were ), and 1-hydroxy-2-naphthoic acid () as of experiments. determined by Data aData cdELISA. Data calculated by of the average ofof three sets Theof the determined by a cdELISA. Data were calculated by the average of three sets of experiments. The control, with no present was 1.5. absorbance ofof the control, A0, with notoxin toxin present was 1.5. absorbance the control, A0, with no toxin present was 1.5. termined by a cdELISA. Data were calculated by the average of threewas sets1.5. of experiments. The absorbance ofA0, the control, A0, with no toxin present absorbance of the control, A0, with no toxin present was 1.5. sorbance of the control, A0, with no toxin present was 1.5.
2.2. Analytical ofRecovery CTN Spiked toofto Red Yeast byby cdELISA 2.2. Analytical Recovery of CTN Spiked Red Yeast Rice Samples cdELISA 2.2. Recovery Analytical of CTN Spiked toRice Red Yeast Rice Samples by cdELISA 2.2. Analytical Recovery CTN Spiked toSamples Red Yeast Rice Samples by cdELISA 2.2. Analytical Recovery of CTN Spiked to Red Yeast Rice Samples by cdELISA alytical Recovery of CTN Spiked to Red Yeast Rice Samples by cdELISA AArecovery was conducted totoexamine cdELISA forfor the ofanalysis CTN recovery study was conducted examine the cdELISA correctness theanalysis analysis of CTN Astudy recovery study was was conducted tothe examine thecorrectness cdELISA correctness for the of CTN A recovery study conducted to examine the cdELISA correctness for the analysis of CTN A recovery study was conducted to examine the cdELISA correctness for the analysis of CTN recovery study was conducted to examine the cdELISA correctness for the analysis of CTN which was spiked in red yeast rice samples. Results from the cdELISA are presented in Table 1. When which was spiked in red yeast rice samples. Results from the cdELISA are presented in Table 1. When which was spiked in red yeast rice samples. Results from the cdELISA are presented in Table 1. When which was spiked in red yeast rice samples. Results from the cdELISA are presented in Table 1. which was spiked in red yeast rice samples. Results from the cdELISA are presented in Table 1. When was spiked in red yeastlevels samples. Results from the cdELISA are presented in Table 1. When the CTN atthe spiked ofof100 ng/g, recovery rates were found totobe 81–86% the CTN at spiked levels 100 to500 500 ng/g, the analytical recovery rates were found be 81–86% CTN atrice spiked levels of 100 tothe 500 ng/g, analytical recovery rates were found to be 81–86% When the CTN attospiked levels ofanalytical 100 tothe 500 ng/g, the analytical recovery rates were found to be the CTN at spiked levels of 100 to 500 ng/g, the analytical recovery rates were found to be 81–86% N with atwith spiked levels ofofvariation 100 to 500 ng/g, the(CV) analytical recovery rates were found toOn be 81–86% coefficient of less than 5%. the other the recovery rate and coefficient variation (CV)values values less than 5%.On On the other hand, the recovery rate andCV CV recovery with81–86% coefficient of(CV) variation values less than 5%. Onhand, the other hand, the recovery rate and CV rate with coefficient of variation (CV) values less than 5%. the other hand, the with coefficient of variation (CV) values less than 5%. On the other hand, the recovery rate and CV oefficient ofofthe variation values less than 5%. On the other hand, the recovery rate and CV value spiked with ng/g CTN was 76% and respectively; this sample was valueof thesample sample spiked 5000 ng/g CTN was 76% and11.6, 11.6, respectively; this sample was value of (CV) the sample spiked with 5000 ng/g CTN was 76% and respectively; this sample and CV value ofwith the5000 sample spiked with 5000 ng/g CTN was11.6, 76% and 11.6, respectively; thiswas sample value of the sample spiked with 5000 ng/g CTN was 76% and 11.6, respectively; this sample was of the sampleto spiked with ng/g CTN was 76% and this was subjected more dilution for incdELISA toin the linear portion oflinear standard subjected to5 5fold fold more dilution forcdELISA cdELISA inorder order tolierespectively; liewithin within linear portion of standard subjected to 5 5000 fold more dilution for order liethe within the linear portion of standard was subjected to 5 fold more dilution for 11.6, cdELISA into order to lie sample within the portion of standard subjected to 5 fold more dilution for cdELISA in order to lie within the linear portion of standard edcurve. tocurve. 5 fold more dilution in recovery order to for lie within the linear portion ofng/g standard The overall average ofcdELISA analytical allall the 100–5000 samples was found toto befound The overall average of analytical recovery for the 100–5000 ng/g samples was found be found curve. The overall average of analytical recovery for all 100–5000 samples was to beto be curve. Thefor overall average of analytical recovery forthe allng/g the 100–5000 ng/g samples was curve. The overall average of analytical recovery for all the 100–5000 ng/g samples was found to be The overall average of analytical recovery for all the 100–5000 ng/g samples was found to be 85% (CV, 7.65). These data suggested that the cdELISA is a feasible method for detecting CTN in red 85% (CV, 7.65). These data suggested that the cdELISA is a feasible method for detecting CTN in red 85% 85% (CV, (CV, 7.65).7.65). These data data suggested that the is a feasible method for detecting CTNCTN in red These suggested thatcdELISA the cdELISA is a feasible method for detecting in red 85% (CV, 7.65). These data suggested that the cdELISA is a feasible method for detecting CTN in red V,yeast 7.65). These data suggested that the cdELISA is a feasible method for detecting CTN in red rice samples. yeast rice samples. yeast rice samples. yeast rice samples. yeast rice samples. ice samples. Table 1. 1. Analytical ofof CTN spiked toof red yeast rice samples byby cdELISA. Table Analytical recovery CTN spiked to red yeast rice samples cdELISA. TableTable 1.recovery Analytical recovery of CTN spiked to red rice samples by cdELISA. 1. Analytical recovery CTN spiked toyeast red yeast rice samples by cdELISA. Table 1. Analytical recovery of CTN spiked to red yeast rice samples by cdELISA. Table 1. Analytical recovery of CTN spiked to red yeast rice samples by cdELISA. a a (ng/g) a (ng/mL) Spiked CTN (ng/g) ELISA ELISA (ng/g) (%) Recovery Spiked CTN (ng/g) ELISA (ng/mL) ELISA (ng/g) CV (%) CV Recovery (%) Spiked CTN (ng/mL) ELISACV (ng/g) (%) (%) Recovery (%) aELISA a Spiked CTN ELISA (ng/g) (%) (%) Spiked CTN(ng/g) (ng/g) ELISA(ng/mL) (ng/mL) ELISA ELISA (ng/g) CV CV (%) Recovery Recovery (%) a
Spiked CTN (ng/g) (%) 100 8.18.1 ± 0.4 ±CV 4± 4(%)81 ±Recovery 4.40 100 ELISA ±ELISA 0.4 8.1 (ng/g) 100 (ng/mL) ± 0.4 8181 44.40 4.40 8181 100 0.4 814± 4 4.40 100 8.18.1 ± ±0.4 81 ± 4.40 100 8.1 81 ± ±4 1.4430 500 ± 1.4 ± 14 500 43.0 ± 1.443.0 430 ±4.40 14 430 ±2.50 500± 0.4 43.0 142.5081 2.50 8686 500 43.0 ± 1.4 430 ± 14 2.50 500 43.0 ± 1.4 430 ± 14 2.50 500 43.0 1.496.5 430 86 12.10 1000 ± 11.8 965 ± 118 9797 1000 96.5 ± 11.8 965 ±2.50 118965965 12.10 1000±1000 96.5±96.5 ±14 11.8 ±12.10 118 ± ± 118 12.10 1000 96.5 ±11.8 11.8 965 ± 118 12.10 1000 96.5 11.8 965 ±381.4 118 12.10 97 11.60 5000 381.4 ± 44.8 3814 ± 448 7676 5000 381.4 ± 44.8 ± 448 11.60 5000± 5000 381.4 ± 44.8 3814 ±11.60 448 ±3814 44.8 3814 ± 448 11.60 5000 381.4 ± 44.8 3814 ± 448 11.60 Overal 5000 ± 44.8 3814 ± 448 11.60 Overal 7.65 8585 Overal 381.4 7.6576 7.657.65 Overal Overal a Each toxin level had two samples and each sample was run in7.65 a Each a Each alevel triplicate. Overal 7.65 85 toxin had two samples and each sample was run inin triplicate. toxin level had two samples and each sample was run triplicate. Each toxin level had two samples and each sample was run in triplicate.
81 86 97 76 85
8181 8686 9797 76 76 85
85
Each toxin level had two samples and each sample was run in triplicate. Each toxin level had two samples and each sample was run in triplicate. a
a
2.3. Assay of CTN inin the Samples Using cdELISA 2.3. Assay of CTN Samples Using cdELISA 2.3. Assay ofthe CTN in the Samples Using cdELISA 2.3.Assay AssayofofCTN CTNininthe theSamples SamplesUsing UsingcdELISA cdELISA 2.3. say of CTN in the Samples Using cdELISA The ofofcdELISA citrinin ininfermented samples Theeffectiveness effectiveness cdELISA formeasuring measuring citrininlevels levels fermented sampleswas wasassessed assessed The effectiveness offor cdELISA for measuring citrinin levels in fermented samples was assessed Theeffectiveness effectivenessof ofcdELISA cdELISAfor formeasuring measuringcitrinin citrininlevels levelsininfermented fermentedsamples sampleswas wasassessed assessed The he by effectiveness cdELISA measuring citrinin levels in fermented samples was assessed 19 brand-name samples one sample (rice) from local markets bycollecting collecting 19 brand-name samplesand and onecontrol control sample (rice) from local markets andtesting testing by of collecting 19for brand-name samples and one control sample (rice) from localand markets and testing by collecting 19 brand-name samples and one control sample (rice) from local markets and testing by collecting 19 brand-name samples one control from local markets and testing ecting 19using brand-name samples and one control sample (rice) from local markets and testing them cdELISA. To the level more accurately and prevent interference them using cdELISA. Tomeasure measure the levelof ofcitrinin citrinin more accurately andto(rice) to prevent interference them using cdELISA. To measure the leveland of citrinin moresample accurately and to prevent interference them using cdELISA. To measure the level of citrinin more accurately and to prevent interference them using cdELISA. To measure theextraction level of citrinin more accurately and to prevent interference using cdELISA. To measure the level of citrinin accurately and to prevent interference that would otherwise be byby the extraction solution, the sample extractions were diluted 10that would otherwise becaused caused the extraction solution, the sample extractions were diluted 10that would otherwise be caused bymore the solution, the sample extractions were diluted 10- that that would otherwise be caused by the extraction solution, the sample extractions were diluted 10otherwise bebuffer caused byby the extraction solution, the1were sample diluted 10-fold ould otherwise bewould caused by thePBS extraction solution, the sample extractions fold with 0.01 M buffer and analyzed Samples high fold withfold 0.01 MPBS PBS buffer and analyzed bythe thecdELISA. cdELISA. Samples 1toto6diluted 6allextractions all1contained contained high with 0.01 M and analyzed by the cdELISA. Samples to106 all were contained high fold with 0.01 M PBS buffer and analyzed by the cdELISA. Samples 1 to 6 all contained high with 0.01 PBS buffer and analyzed by the cdELISA. 1 µtog/g) 62.). all contained ithconcentrations 0.01 M PBS buffer andMof analyzed by the cdELISA. Samples 1 toµSamples 6µ g/g) all contained high ofofcitrinin from 1628 toto 9454 ng/g (1.628 toto9.454 g/g) (Table Moreover, the concentrations citrinin from 1628 9454 ng/g (1.628 9.454 (Table 2.). Moreover, theconcentrations concentrations citrinin from 1628 to 9454 ng/g (1.628 to 9.454 (Table 2.). high Moreover, the concentrations of citrinin from 1628 to to 9454 ng/g (1.628 to 9.454 µ g/g) (Table 2.). Moreover, the of citrinin from 1628 to 9454 ng/g (1.628 9.454 µg/g) (Table 2.). Moreover, the concentrations trations of citrinin from 1628 to 69454 toTaiwan-FDA 9.454 µ g/g) (Table 2.). regulatory Moreover, concentrations inin sample 5inand both exceeded regulatory limit ofof 5 µ5limit g/g (ppm) concentrations sample 5sample and 6 both exceeded the Taiwan-FDA regulatory limit µthe g/g (ppm) for(ppm) for in concentrations 5 ng/g and 6(1.628 boththe exceeded the Taiwan-FDA of 5 µ for g/g concentrations in sample 5 andthe 6 both exceeded the Taiwan-FDAof regulatory limit of 5 µ g/g (ppm) for sample 5both and 6 both Taiwan-FDA 5there µg/g (ppm) for in red trations in in sample 5yeast and 6rice. exceeded Taiwan-FDA regulatory limit of 5limit µfood, g/g (ppm) for citrinin red Samples from Monascus-fermented there are five citrinin in redyeast Samples 7tothe to1919 from other Monascus-fermented food, are fivenot not citrinin in rice. red yeast rice.7exceeded Samples 7 toother 19 from otherregulatory Monascus-fermented food, there arecitrinin five not citrinin in red yeast rice. Samples 7 to 19 from other Monascus-fermented food, there are five not rice. Samples 7 to 19 fromMonascus-fermented other Monascus-fermented food, there are five not detected samples, n indetected red yeast rice.yeast Samples 7 tocontained 19others from other food, there are five not samples, the others citrinin at levels from 28.6 to 64.4 ng/g, did not exceed detected samples, the others contained citrinin at levels from 28.6 to 64.4 ng/g, which did not exceed detected samples, the contained citrinin at levels from 28.6 to which 64.4 ng/g, which did not exceed detected samples, the others contained citrinin at levels from 28.6 to 64.4 ng/g, which did not exceed contained citrinin at levels from 64.4 ng/g, which not exceed the 2 µg/g (ppm) d samples, the others contained at levels from 28.6 to28.6 64.4 ng/g, which did notdid exceed the 2µ (ppm) ofothers Taiwan-FDA regulatory limit forfor citrinin into Monascus-fermented products. the 2 g/g µ g/g (ppm) of Taiwan-FDA regulatory limit citrinin in Monascus-fermented products. the 2the µ g/g (ppm) ofcitrinin Taiwan-FDA regulatory limit for citrinin in Monascus-fermented products. the 2 µ g/g (ppm) of Taiwan-FDA regulatory limit for citrinin in Monascus-fermented products. of Taiwan-FDA regulatory limit for citrinin in Monascus-fermented products. g/g (ppm) of Taiwan-FDA regulatory limit for citrinin in Monascus-fermented products.
Toxins 2018, 10, 354
4 of 14
Table 2. cdELISA and immunostrip analytical results of citrinin in red yeast fermented foods.
a
Samples
Food
ELISA (ng/mL) a
ELISA (ng/g) a
Immunostrip Assay
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Control
Red yeast rice Red yeast rice Red yeast rice Red yeast rice Red yeast rice Red yeast rice Cookie Cookie Cookie Wine Capsule Bean Soy sauce Soy sauce Sauce Sauce Sauce Sauce Sauce Rice c
167.7 ± 4.1 468.3 ± 33.5 306.5 ± 24.7 162.8 ± 7.5 945.4 ± 80.6 698.7 ± 2.3 ND b 6.17 ± 0.84 ND b 4.21 ± 0.29 ND b 3.42 ± 0.13 2.86 ± 0.14 4.36 ± 0.11 ND b 3.35 ± 0.1 4.93 ± 0.2 ND b 6.44 ± 0.53 ND b
1677 ± 41 4683 ± 335 3065 ± 247 1628 ± 75 9454 ± 806 6987 ± 23 ND b 61.7 ± 8.4 ND b 42.1 ± 2.9 ND b 34.2 ± 1.3 28.6 ± 1.4 43.6 ± 1.1 ND b 33.5 ± 1.0 49.3 ± 2.0 ND b 64.4 ± 5.3 ND b
+ + + + + + − − − − − − − − − − − − − −
Each sample was extracted duplicate and analyzed in three repeats. b ND, not detected. c Rice is the control sample.
2.4. Fabrication of the Immunostrip The immunostrip provides a suitable means of detecting citrinin in Monascus-fermented products on-site and is rapid and does not involve cumbersome operations. When the citrinin level in the sample surpasses a particular value, the toxin captures all of the binding sites of the antibody-gold conjugates in the sample solution or in the “conjugated pad”. Therefore, no available antibody-gold conjugate can bind with CTN-OVA in the test zone, and no red line is visible in the test zone, which indicated a positive result. The correctness of the assay is verified using a control zone with a goat-anti-rabbit secondary antibody. With proper operation, a red line should always be visible in the control zone in spite of the presence or absence of citrinin. Therefore, the testing of a citrinin-negative sample should always produce two red lines on the membrane, whereas a toxin-positive sample should only display one (Figure 2).
Toxins 2018, 10, 354 Toxins 2018, 10, x FOR PEER REVIEW
5 of 14 5 of 14
Figure 2. A graphic description of the immunostrip. Ab-gold nanoparticle; C, control line (Goat Figure 2. A graphic of the immunostrip. Ab-gold nanoparticle; C, control line (Goat antianti-rabbit IgG); T, description test line (CTN-OVA); released pad. rabbit IgG); T, test line (CTN-OVA); released pad.
2.5. Visual Detection Limit of an Immunostrip for CTN 2.5. Visual Detection Limit of an Immunostrip for CTN Citrinin certified standard (0–100 ng/mL) were subjected to immunostrip testing where CTN-OVA Citrinin certified standard (0–100 ng/mL) were subjected to immunostrip testing where CTNand goat anti-rabbit were drawn on the test zone and control line, respectively. When the citrinin OVA and goat anti-rabbit were drawn on the test zone and control line, respectively. When the certified analytical standard solution and the antibody-gold nanoparticles conjugates were directly citrinin certified analytical standard solution and the antibody-gold nanoparticles conjugates were mixed in a microplate well and the immunostrip was dipped into the well, the detection limit of the directly mixed in a microplate well and the immunostrip was dipped into the well, the detection limit immunostrip was was 20 ng/mL (Figure 3A). When 20ng/mL, ng/mL, the citrinin of the immunostrip 20 ng/mL (Figure 3A). Whenthe thecitrinin citrininlevel level exceeded exceeded 20 the citrinin completely occupied sitessites of theofconjugates then no available nanoparticles completely occupiedthe thebinding binding the conjugates then no antibody-gold available antibody-gold could bind to the CTN-OVA in the test zone. Therefore, the immunostrip exhibited only one red line in nanoparticles could bind to the CTN-OVA in the test zone. Therefore, the immunostrip exhibited the control zone. The color density on the test line was also analyzed by a strip scan reader this only one red line in the control zone. The color density on the test line was also analyzed by aand strip established curves, the whereas the Tcurves, line color density value is less 25, which scan reader the andstandard this established standard whereas the T line colorthan density valueindicated is less a positive result (Figure 3C). Alternatively, when the antibody-gold nanoparticles were absorbed by than 25, which indicated a positive result (Figure 3C). Alternatively, when the antibody-gold the release pad, theabsorbed visual detection limit ofpad, the immunostrip was 20–50 ng/mL 3B) while nanoparticles were by the release the visual detection limit of the(Figure immunostrip wasthe T line color density value lessthe thanT 25 (Figure Accordingly, the samples 20–50 ng/mL (Figure 3B)was while line color3D). density value was less thanand 25 antibody-coated (Figure 3D). gold nanoparticles were mixed first and then subject to the immunostrip for sample analysis. Accordingly, the samples and antibody-coated gold nanoparticles were mixed first and then subject to the immunostrip for sample analysis.
Toxins 2018, 10, 354 Toxins 2018, 10, x FOR PEER REVIEW
6 of 14 6 of 14
(A)
(B)
R 2 = 0.987
(C)
R 2 = 0.958
(D)
Figure 3. 3. The detection limit of the for CTN. Different concentrations (0–100 Figure Thevisual visual detection limit of immunostrip the immunostrip for CTN. Different concentrations ng/mL) of a CTN certified standard was dissolved in PBS. (A) The visual detection limit is 20 ng/mL (0–100 ng/mL) of a CTN certified standard was dissolved in PBS. (A) The visual detection limit using 8 L ofusing Ab-gold were loaded into sample solution; (B) the visual is 20 ng/mL 8 µLnanoparticle of Ab-gold conjugates nanoparticle conjugates werethe loaded into the sample solution; detection limitdetection is approximately 20~50 ng/mL 20~50 when 8ng/mL µ L of conjugates absorbedwere on the release (B) the visual limit is approximately when 8 µL were of conjugates absorbed pad. (C) The standard curve of the T line color density value without the release pad; when the on the release pad. (C) The standard curve of the T line color density value without the releasevalue pad; is lessthe than 25 it is indicates a positive result. a(D) The standard forstandard the T linecurve color for density when value less than 25 it indicates positive result. curve (D) The the Tvalue line with the release pad; when valuepad; is less than 25value it indicates positive result. Each concentration color density value with thethe release when the is less athan 25 it indicates a positive result. was tested three repeats. Each concentration was tested three repeats.
2.6. Assay Assay of of CTN CTN in in Samples Samples with with the the Immunostrip Immunostrip 2.6. The 19 19 extracted control sample (rice) were added to the The extracted red red yeast yeastfermented fermentedsamples samplesand andone one control sample (rice) were added to microplate well to determine citrinin contamination using immunostrips. According to Table 2, the microplate well to determine citrinin contamination using immunostrips. According to Table 2, samples 11 to to 66 contained contained citrinin citrinin levels levels that that were were much much higher higher than than 20 20ng/mL. ng/mL. Samples Samples 77 to to 19 19 samples contained citrinin levels lower than 20 ng/mL in the extraction solution. Therefore, in Figure 4, sample contained citrinin levels lower than 20 ng/mL in the extraction solution. Therefore, in Figure 4, sample to66yielded yielded positive positive results results with with only only aa reddish reddish line line in in the thecontrol control zone zone of ofthe themembrane. membrane. Further Further 11 to dilutions of of samples samples 1–6 1–6 (1:10; (1:10; 1:20; 1:20; 1:50) 1:50) were were subjected subjected to to the the immunostrip immunostrip test. test. The The results results were were dilutions presented in the Figure S1. All dilution samples examined with immunostrips were positive with presented in the Figure S1. All dilution samples examined with immunostrips were positive with only a reddish line in the control zone which are consistent with the cdELISA results. Samples 7 to 19 only a reddish line in the control zone which are consistent with the cdELISA results. Samples 7 to 19 yielded negative negative results results with with one one clear clear red red line line in in the the test test zone zone and and the the other other in in the the control control zone zone of of yielded the immunostrip. the immunostrip.
Toxins 2018, 10, xx FOR Toxins2018, 2018,10, 10,354 FOR PEER PEER REVIEW REVIEW Toxins
of 14 of14 14 777of
Figure 4. Analysis of CTN with the immunostrip in 19 red yeast fermented samples and one control Figure 4. of CTN with immunostrip 19 and control Figure 4. Analysis Analysis of 1–6 CTN with the the immunostrip in 19 red red yeast yeast fermented samples and one one control rice sample. Samples showed that the red line in vanished in thefermented test zone, samples which verified that they rice sample. Samples 1–6 showed that the red line vanished in the test zone, which verified that they rice sample. Samples 1–6 showed that the red line vanished in the test zone, which verified that they are CTN positive. Samples 7–19 containing CTN less than 20 ng/mL displayed two red lines indicating are CTN positive. Samples 7–19 containing CTN less than 20 ng/mL displayed two red lines indicating are CTN positive. Samples 7–19 containing CTN less than 20 ng/mL displayed two red lines indicating that they are negative. Each sample was tested three repeats. that that they they are are negative. negative. Each Each sample sample was was tested tested three three repeats. repeats.
2.7. Comparison of Gold Nanoparticles Size for Immunostrip 2.7. Comparison of Gold Nanoparticles Size for Immunostrip To test whether the size of the gold nanoparticles will affect the sensitivity of the immunostrip To test whether the size of the gold nanoparticles will affect the sensitivity of the immunostrip or not. Two different size of gold nanoparticles with diameters of approximately 15 or 40 nm or not. Two different size of gold nanoparticles with diameters of approximately 15 or 40 nm were were synthesized and conjugated with antibodies specific for CTN. Figure 5 shows that the results synthesized and conjugated with antibodies specific for CTN. Figure 5 shows that the results using using antibody-gold nanoparticles of 15 nm (Figure 5A) are better than that of using antibody-gold antibody-gold nanoparticles of 15 nm (Figure 5A) are better than that of using antibody-gold nanoparticles of 40 nm (Figure 5B) based on a T line color density by the strip scan reader. nanoparticles of 40 nm (Figure 5B) based on a T line color density by the strip scan reader.
(A)
(B)
Figure 5. nanoparticles were used for conjugation with antibodies. The Figure 5. 5. Two Two different sizes ofofgold gold nanoparticles were used for for conjugation withwith antibodies. The Figure Twodifferent differentsizes sizesof gold nanoparticles were used conjugation antibodies. competitiveness of the antigen on the test line with antibody-gold nanoparticles of 15 nm (A) is better competitiveness of the antigen on the test line with antibody-gold nanoparticles of 15 nm (A) is better The competitiveness of the antigen on the test line with antibody-gold nanoparticles of 15 nm (A) is than the 40 nm particles (B). than that that ofthat theof 40the nm40 particles (B). (B). better thanof nm particles
Toxins 2018, 10, 354
8 of 14
3. Discussion Food that is contaminated with mycotoxins such as citrinin affects human health [24]. Thus, the development of a quick detection means for citrinin is very important, simplifies the preparation of samples and reduces the cost of analysis of the critical points. This study focuses on the ELISA and the development of an immunostrip. Both methods are simple and more rapid than that of the chemical methods. CTN is small molecular weight mycotoxin, which is non-immunogenic but has to be coupled with a protein carrier to exhibit this immunogenicity. Although CTN has a carboxylic acid group that could be conjugated to a protein carrier using a water-soluble carbodiimide method, the hydroxyl group on the eighth carbon forms a hydrogen bond with carboxylic acid on the 14th carbon, and the ketone group on the sixth carbon forms a hydrogen bond with carboxylic acid on the 14th carbon [12,25]. These hydrogen bonds eliminate the capacity of the carboxylic acid to conjugate with a protein carrier with an amine group. Therefore, the carbodiimide method is not suitable for conjugating CTN with proteins. CTN also has a methyl group that could be conjugated to a protein carrier by the Mannich method using formaldehyde [26], and therefore, a formaldehyde method was used herein to conjugate CTN with keyhole limpet hemocyanin (KLH) to make it a valid antigen to produce specific antibodies for citrinin. The high sensitivity of cdELISA and ciELISA depends on the quality of the CTN-HRP conjugates and CTN-OVA conjugates. In this study, the conjugation of citrinin and HRP with a molecular ratio 44:1 with the Mannich method maximized the cdELISA sensitivity, and the conjugation of citrinin and OVA with the molecular ratio 22:1 by the Mannich method maximized the ciELISA sensitivity. In the ciELISA, the concentration that caused 50% inhibition of the binding of antibodies to the solid-phase CTN-OVA by free citrinin (IC50 ) was calculated to be 4.6 ng/mL (Figure 1A). In cdELISA, the concentration that caused 50% inhibition of the binding of CTN-HRP to the antibodies by CTN (IC50 ) was 5.0 ng/mL (Figure 1B). Ochratoxin A and synthetic compound 1-hydroxy-2-naphthoic acid whose portion of structures are similar to that of citrinin, showed 10% weak cross-reactivity with the antibodies at a level of 100 ng/mL. Although Kong, Xie, Liu, Song and Kuang (2017) [27] reported a ciELISA and immunostrip for CTN, they had not developed the cdELISA method and were also short of the real sample analysis. Due to the cdELISA is more simple and rapid than ciELISA, this study devoted to the development of cdELISA and its application to screening a large amount of red yeast fermented samples. In cdELISA, the samples were extracted using 100% methanol and were generally diluted at least 10-fold with 0.01 M PBS to avoid solvent and matrix interference. The cdELISA can tolerate up to 10% methanol in the analyzed samples [28] and its detection limit is determined to be 0.2 ng/mL in accordance with a 90% confidence interval at 10% of inhibition of binding of the CTN-HRP conjugate. (Figure 1B). The Taiwan FDA has set regulatory limits of 200 ppb, 2 ppm and 5 ppm on citrinin in red yeast rice pigments, red yeast food and red yeast rice, respectively. Japan also set a regulatory limit of 200 ppb in red yeast pigment [4]. The results of this study reveal that two out of six red yeast rice samples that were collected in a local supermarket contained more than 5 ppm of citrinin and one contained 9.5 ppm, which is almost two times that of Taiwan’s regulatory limits (Table 2). However, some other red yeast samples that were contained with CTN levels that did not violate the limit. This article involved in using immunochemical method such as ELISA and rapid gold nanoparticle immunostrip for rapid and on-site detection of citrinin. In order to confirm the efficacy and correctness of sample analysis by using cdELISA, we have analyzed the recovery of CTN spiked to red yeast rice samples. The results indicated that the recovery rates were found to be 76–97% which suggested that the cdELISA is a feasible method for detecting CTN in the samples. Chemical analytical method such as HPLC for citrinin analysis is not our main focus. Detecting citrinin in red yeast rice samples using HPLC was carried out and published by Taiwan Food and Drug Administration. Liao, et al. [4] have reported that 58 of 84 commercially available Monascus-fermented rice samples collected in 2009–2012 were contaminated with 0.4–93.5 ppm of citrinin. Therefore, the contamination of citrinin in fermented rice is very serious.
Toxins 2018, 10, 354
9 of 14
There are several advantages of the immunostrip over ELISA include time-savings and the ease of application by the user. Unskilled personnel without using any devices can acquire analytical results using the immunostrip in 15 min, which makes this method very feasible for the on-site testing of CTN. This is the first work to develop an immunostrip for detecting CTN in red yeast fermented samples. Even though large amounts of antibody and CTN-OVA conjugates are required for each sample assay, compared to the ELISA, which results in a slightly higher cost than the ELISA, the obvious advantages warrant its further development. Moreover, the absence of an enzyme reaction in the immunostrip provide its stability during a longer-term storage. Figure 3A shows that the visual detection limit of the immunostrip is 20 ng/mL, which is below the regulatory limits of Taiwan. So that our immunostrip can meet the regulatory limit test requirements, we set the different dilution factors of 10, 100 and 250 for yeast rice pigments, red yeast food and red yeast rice, respectively. With respect to sample analysis, the results in Figure 4 are consistent with the data in Table 1, which proves that this immunostrip can be used to detect CTN in red yeast samples. According to a recent report [29], most immunostrip researchers did not detect toxic bacteria or toxins from real food using their technologies. Many groups of researchers artificially added and spiked toxins before detection. There was a limit to the application. However, our study succeeded in detecting toxins from actual fermented foods using immunostrips. As previous studies have indicated, the intensity of the color and the competition between CTN in the sample and the CTN-OVA in the test line can be affected by the size of the antibody-gold nanoparticle conjugates [30,31]. Therefore, this study follows Frens (1973) [31] in the synthesizing of gold nanoparticles with diameters of approximately 15 or 40 nm and then conjugating these different sizes of gold nanoparticles with antibodies specific for CTN. Figure 5 indicates that the results using antibody-gold nanoparticles of 15 nm are better than that of using antibody-gold nanoparticles of 40 nm based on a T line color density by the strip scan reader. However, to develop a highly sensitive immunostrip, the decision was ultimately made to use the 15 nm gold nanoparticles to conjugate with antibodies specific to CTN and without a release pad. 4. Conclusions This work developed antibodies against citrinin and is the first occurrence of their use in an immunostrip to detect CTN in red yeast samples. The concentration of CTN caused 50% inhibition of the binding of CTN-HRP to the antibodies and was calculated to be 5.0 ng/mL in the cdELISA, and the detection limit was determined to be 0.2 ng/mL. The immunostrip had a visual detection limit of 20 ng/mL in 15 min and it is sensitive enough for analyzing red yeast samples. The methods developed can have a vital role for the rapid detection of CTN in red yeast fermented food. 5. Materials and Methods Citrinin, certified analytical standard, ochratoxin A, 1-Hydroxy-2-naphthoic acid, ovalbumin (OVA), bovine serum albumin (BSA), Keyhole limpet hemocyanin (KLH), Tween 20, Hydrogen tetrachloroaurate (III), and trihydrate (HAuCl4) were obtained from Sigma (St. Louis, MO, USA). Ammonium sulfate and sodium acetate were purchased from J. T. Baker (Phillipsburg, NJ, USA). Methanol and formaldehyde were purchased from Merck (Darmstadt, Germany). Goat anti-rabbit IgG-HRP and Horseradish peroxidase (HRP) were obtained purchased from Roche (Mannheim, Germany). Microplates, Freund’s complete adjuvant, Freund’s incomplete adjuvant and goat anti-rabbit IgG were purchased from Thermo (East Grinstead, UK). HRP substrate solution 3,30 ,5,50 -tetramethylbenzidine (TMB) was purchased from Neogen Corp (Lexington, KY, USA). An Easypack kit was purchased from MDI Membrane Technologies (Ambala, India). The Double Axes Programmable Controller (model P602C) for drawing the test line and the control line of the membrane was purchased from Troy (Taichung, Taiwan). The 0.22 µm syringe filter was provided from Gelman Science (Ann Arbor, MI, USA). All of the organic solvents and other chemicals used were reagent grade or better. The New Zealand Rabbits (8 weeks old, female) were ordered from Da Zong Farm (Changhua, Taiwan). A Vmax ELISA reader was obtained from Molecular Devices Co. (Menlo Park,
Toxins 2018, 10, 354
10 of 14
CA, USA). The immunostrip scan reader was obtained from Taiwan Advance Bio-Pharmaceutical Inc. (Taipei, Taiwan). 5.1. Preparation of the Different CTN Conjugates 5.1.1. Coupling of CTN to KLH CTN was covalent coupled to KLH with formaldehyde to serve as an immunogen for immunizing the rabbits [26]. Briefly, the KLH solution (5 mg of KLH in 0.5 mL of 0.1 mM sodium acetate solution, pH 4.2) was added to CTN (1 mg) in 1 mL methanol with constant stirring at room temperature. Then, 0.32 mL of a 36.5% formaldehyde solution was added to the mixture with steady stirring at room temperature for 72 h and stirred at 4 ◦ C for another 16 h. Finally, the mixture was dialyzed against 0.01 M PBS for 72 h with four exchanges of 0.01 M PBS. The products were stored at −20 ◦ C until used. 5.1.2. Conjugation of CTN to OVA As a coating antigen for ciELISA and a control line for the immunostrip, the CTN was conjugated to OVA using formaldehyde [26]. Generally, 8 mg of OVA in 1.28 mL 0.1 mM sodium acetate solution (pH 4.2) is added to CTN (1 mg) in 1 mL of methanol with constant stirring at room temperature. Next, 0.32 mL of the 36.5% formaldehyde solution was added gradually to the OVA combination with steady stirring at room temperature for 72 h and then stirred at 4 ◦ C for another 16 h. Following the coupling reaction, the combination was dialyzed in 0.01 M PBS with the buffer changed 3 times within 72 h and then stored at −20 ◦ C until used. 5.1.3. Conjugation of CTN to HRP Coupling of CTN to HRP was also completed using formaldehyde [26]. Generally, the HRP solution (0.8 mg of HRP in a 0.8 mL of 0.1 M sodium acetate solution, pH 4.2) was added to CTN (0.2 mg) in 0.1 mL methanol with steady stirring at room temperature. Afterward, 0.04 mL of a 36.5% formaldehyde solution was added slowly to the HRP solution with steady stirring at room temperature for 72 h. After the room temperature reaction, the mixture was stirred at 4 ◦ C for 16 h and dialyzed against 1 L of 0.01 M PBS for 48 h with five buffer exchanges. After dialysis, the reactant was kept for the cdELISA analysis. 5.2. Generation of Polyclonal Antibody The immunization method was similar to those previously reported [28]. New Zealand rabbits (8 weeks old and female) were immunized with 0.5 mg of CTN-KLH in 0.5 mL of 0.01 M PBS that had been emulsified with an equal volume of Freund’s complete adjuvant by intradermal injection to generate the polyclonal antibody against CTN. After an initial injection for four weeks, the booster antigen containing 0.5 mg of CTN-KLH in 0.5 mL of PBS and the 0.5 mL of Freund’s incomplete adjuvant was subcutaneously injected at 4 sites on the thigh. The antiserum was collected from the ears of the rabbit beginning at the sixth week. The antiserum was precipitated 2 times with 35% (NH4 )2 SO4 . The final precipitate in (NH4 )2 SO4 was dissolved in distilled water the same as half of the original volume. Afterwards, the mixture was dialyzed in 1 L of 0.01 M PBS and the PBS buffer was changed 3 times within 24 h at 4 ◦ C. 5.3. ciELISA The procedure of ciELISA was carried out based on the method primarily described [19,32]. CTN-OVA (0.1 mL and 2 µg/mL) was coated onto each well of a plate and kept at 37 ◦ C for 1 h. After 1 h of incubation, the plate was washed four times with PBS-Tween (0.35 mL per well; 0.05% Tween 20 in 0.01 M PBS). Then, 0.2 mL of PBS-BSA (0.2 mL per well; 0.1% BSA in 0.01 M PBS) was added and allowed to incubate at 37 ◦ C for 30 min. The plate was washed again after incubation, followed by adding a CTN standard to each well with a concentration from 0−500 ng/mL and 0.05 mL anti-CTN
Toxins 2018, 10, 354
11 of 14
polyclonal antibody (5 µg/mL in 0.01 M PBS) for reaction at 37 ◦ C for 1 h. After the plate was washed four times with PBS-Tween, 0.1 mL of the goat anti-rabbit IgG-HRP conjugate (1:20,000 dilution) was added and reacted at 37 ◦ C for 1 h. The plate was washed and 0.1 mL of TMB substrate solution was added and the color development was permitted to occur for 15 min. The reaction was terminated by adding 0.1 mL of 1 N HCl. A ELISA reader (Vmax, Molecular Devices Co.. Menlo Park, CA, USA) was used to read the 450 nm absorbance value. 5.4. cdELISA The procedure for the cdELISA was the same as that previously reported [32]. The anti-CTN polyclonal antibody was coated to the plate (0.1 mL of 10 µg/mL in 0.01 M PBS and kept at 37 ◦ C for 1 h. The plate was washed with PBS-Tween followed by blocking with PBS-BSA at 37 ◦ C for 30 min. After these steps, the CTN standards (0−1000 ng/mL; 0.05 mL per well in 0.01 M PBS) or CTN samples (0.05 mL) together with the CTN-HRP conjugate (0.05 mL in 0.01 M PBS) were added and reacted at 37 ◦ C for 1 h. The plate was washed again and then 0.1 mL of TMB substrate solution was added. After reaction at room temperature in the dark for 10 min, the reaction was stopped by adding 0.1 mL of 1 N HCl. The absorbance value at 450 nm was measured by a ELISA reader. 5.5. Analytical Recovery of CTN Spiked to Red Yeast Rice Sample by cdELISA The red yeast rice sample collected from the local market was examined to be CTN negative by cdELISA first, and then one gram of grounded sample was spiked with CTN at concentrations ranging from 100 to 5000 ng/g. A control sample with no toxin added served as the blank. The protocol for sample extraction was similar to that previously described [4]. The spiked samples were kept in a cool room (4 ◦ C) for 1 day and each sample in a 50 mL screwed capped bottle was added 10 mL of methanol for extraction. The samples were vortexed for 1 min and heated in the 70 ◦ C water bath for 30 min. After cooling at room temperature, the supernatant of each sample was concentrated by a rotary evaporator and the final residue was reconstituted in 1 mL of methanol. The methanol solution was filtered through 0.22 µm nylon syringe filters and diluted with 9 mL of 0.01 M PBS for the following cdELISA analysis. Further dilution is needed if the sample concentration exceeds the highest point of the curve or the linear portion of standard curve. 5.6. cdELISA of Samples Contaminated with CTN Nineteen samples, including red yeast rice, Monascus cookie and soy sauce samples, obtained from supermarkets were subjected to analyze the CTN levels. The protocol for sample extraction was similar to that above described [4]. Briefly, each sample (1 g) was extracted in 10 mL methanol and heated for 30 min at 70 ◦ C with the water bath. Then, the supernatant was evaporated by a rotary evaporator and reconstituted in 1 mL of methanol. The methanol solution was filtered through 0.22 µm nylon syringe filters and diluted with 9 mL of 0.01 M PBS or more dilutions for both the cdELISA analysis and the immunostrip assay. 5.7. Construction of the Antibody-Gold Nanoparticle 5.7.1. Synthesis of Different Size Gold Nanoparticles The process of synthesis was operated as the method previously published by Frens (1973) [31]. For the synthesis of 15 nm of gold nanoparticles, the 19.7 mg of HAuCl4 dissolved in 50 mL deionized water was heated with constant stirring. A condenser was adapted to avoid the evaporation of the solvent. Five mL of trisodium citrate (34 mM) was added while the solution was boiling. The color of the solution was changed in order (yellow, transparent, black and red) in 1 min, and the solution was cooled at room temperature. On the other hand, 2.5 mL of trisodium citrate (34 mM) was added to the 1 mM HAuCl4 solution. The resulting gold particle size was 40 nm. Transmission electron microscopy was used to confirm the diameter of the gold nanoparticles.
Toxins 2018, 10, 354
12 of 14
5.7.2. Conjugation of the Antibody-Gold Nanoparticle. The process for the conjugation was related to that previously reported [19,33]. A titration result of 20 µg of CTN antibody dissolved in 0.1 mL of boric acid-borax buffer (pH 6.0) was added gradually to 2 mL of the gold nanoparticle (15 or 40 nm in diameter) solution with gentle stirring. The combination was reacted at room temperature for 30 min and stopped with 200 µL of filtered BSA for 30 min. After the reaction, the combination was centrifuged at 18,000× g for 30 min at 4 ◦ C; the supernatant was then removed, and the antibody-gold pellets were reconstituted by adding 200 µL of 20 mM Tris-buffer saline (pH 8.0) with 1% BSA and 0.1% sodium azide. These CTN polyclonal antibody-gold nanoparticle were kept at 4 ◦ C until use. 5.8. Preparation of the Immunostrip The assembly method of the immunostrip was achieved based on the method previously reported [19] with a slight modification. The immunostrip made up of three pads including sample, conjugate release and absorbent pads and the nitrocellulose membrane (NC membrane; pore size of 15 µm) with a test line and a control line. The test and control line of the NC membrane were drawn with 0.25 µL CTN-OVA (4 mg/mL) and 0.25 µL goat anti-rabbit IgG antibody (1 mg/mL), respectively, and then rested for 10 min at room temperature until the line was dried. Subsequently, the CTN polyclonal antibody-gold nanoparticle conjugate (8 µL per well) was applied to the sample solution or absorbed to conjugate release pad, which air dried at 37 ◦ C for 10 min. The release pad was first pasted on the NC membrane by overcrossing 2 mm. Next, the sample pad was pasted on the release pad by overcrossing 2 mm. Finally, the absorbent pad was pasted on the topmost of the NC membrane for 2 mm crossover. The full assembled sheet was cut lengthways (4 mm) with an automatic cutter [34]. 5.9. Assay of the CTN in Samples with the Immunostrip Analysis of the CTN in samples with the immunostrip, the antibody-gold nanoparticle conjugates (8 µL) were premixed with the sample solution (200 µL) or CTN standard solution. Each of the different concentrations of the CTN standard solution (0–100 ng/mL) or the sample solutions were loaded on the sample well. The immunostrips were dipped into the sample wells and the samples or CTN standard solution moved upward toward the membrane by capillary action. The results on the immunostrip were visual estimated after 15 min and the red line color density on the test line was analyzed by an immunostrip scan reader. 5.10. Data Analysis The samples and standards were tested in triplicate and the mean values were obtained from the ELISA. Standard curves were established by plotting the absorbance value against the logarithm of the analyte concentration and fitted to a four-parameter logistic equation using GraphPad prism 5.0 software (La Jolla, CA, USA): Y = {(A − D) ÷ (1 + (x/C) B )} + D where A is the asymptotic maximum (Absorbance value in the absence of an analyte, Amax ), B is the slope of the curve at the inflection point, C is the x value at the inflection point, and D is the asymptotic minimum (Amin , background signal). Supplementary Materials: The following are available online at http://www.mdpi.com/2072-6651/10/9/354/s1, Figure S1: Various dilutions of extraction (1:10; 1:20; 1:50) of red yeast samples 1–6 were subjected to the immunostrip tests. Author Contributions: Conceptualization, S.-W.W. and Y.-A.Y.; Methodology, S.-W.W. and Y.-A.Y.; Validation, S.-W.W. and F.Y.Y.; Investigation, F.Y.Y.; Writing-Original Draft Preparation, S.-W.W.; Writing-Review & Editing, F.Y.Y.; Visualization, B.-H.L.; Supervision, B.-H.L.; Project Administration, F.Y.Y.; Funding Acquisition, B.-H.L.
Toxins 2018, 10, 354
13 of 14
Funding: This research was funded by Ministry of Science and Technology, Taiwan: 103-2313-B-040-002-MY3 and Taiwan Food and Drug Administration 106TFDA-A-111 and 107TFDA-A-108. Acknowledgments: Ted Knoy is appreciated for his editorial assistance. Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2. 3. 4. 5. 6.
7. 8. 9. 10. 11. 12. 13.
14.
15. 16.
17. 18. 19.
20.
Bennett, J.W.; Klich, M. Mycotoxins. Clin. Microbiol. Rev. 2003, 16, 497–516. [CrossRef] [PubMed] Chu, F.S. Mycotoxins: Food contamination, mechanism, carcinogenic potential and preventive measures. Mutat. Res. 1991, 259, 291–306. [CrossRef] Liu, B.H.; Wu, T.S.; Su, M.C.; Chung, C.P.; Yu, F.Y. Evaluation of citrinin occurrence and cytotoxicity in Monascus fermentation products. J. Agric. Food Chem. 2005, 53, 170–175. [CrossRef] [PubMed] Liao, C.D.; Chen, Y.C.; Lin, H.Y.; Chiueh, L.C.; Shih, D.Y.C. Incidence of citrinin in red yeast rice and various commercial Monascus products in Taiwan from 2009 to 2012. Food Control 2014, 38, 178–183. [CrossRef] IARC. Citrinin. IARC Monograph 40; IARC: Lyons, France, 1986; pp. 67–82. Huertas-Perez, J.F.; Arroyo-Manzanares, N.; Garcia-Campana, A.M.; Gamiz-Gracia, L. High-throughput determination of citrinin in rice by ultra-high-performance liquid chromatography and fluorescence detection (UHPLC-FL). Food Addit. Contam. 2015, 32, 1352–1357. [CrossRef] [PubMed] Lee, C.L.; Wang, J.J.; Pan, T.M. Synchronous analysis method for detection of citrinin and the lactone and acid forms of monacolin K in red mold rice. J. AOAC Int. 2006, 89, 669–677. [PubMed] Marti, L.R.; Wilson, D.M.; Evans, B.D. Determination of citrinin in corn and barley. J. Assoc. Off. Anal. Chem. 1978, 61, 1353–1358. [PubMed] Odhav, B.; Naicker, V. Mycotoxins in South African traditionally brewed beers. Food Addit. Contam. 2002, 19, 55–61. [CrossRef] [PubMed] Vail, R.B.; Homann, M.J. Rapid and sensitive detection of citrinin production during fungal fermentation using high-performance liquid chromatography. J. Chromatogr. 1990, 535, 317–323. [CrossRef] Abramson, D. Determination of citrinin in barley by indirect and direct enzyme immunoassay. J. AOAC Int. 1996, 79, 1325–1329. [PubMed] Abramson, D.; Usleber, E.; Martlbauer, E. Immunochemical method for citrinin. Methods Mol. Biol. 2001, 157, 195–204. [PubMed] Singh, D.K.; Ganbold, E.O.; Cho, E.M.; Cho, K.H.; Kim, D.; Choo, J.; Kim, S.; Lee, C.M.; Yang, S.I.; Joo, S.W. Detection of the mycotoxin citrinin using silver substrates and Raman spectroscopy. J. Hazard. Mater. 2014, 265, 89–95. [CrossRef] [PubMed] Xu, Y.; Xiong, L.; Li, Y.; Xiong, Y.; Tu, Z.; Fu, J.; Chen, B. Anti-idiotypic nanobody as citrinin mimotope from a naive alpaca heavy chain single domain antibody library. Anal. Bioanal. Chem. 2015, 407, 5333–5341. [CrossRef] [PubMed] Bailes, J.; Mayoss, S.; Teale, P.; Soloviev, M. Gold nanoparticle antibody conjugates for use in competitive lateral flow assays. Methods Mol. Biol. 2012, 906, 45–55. [PubMed] Wang, Z.; Zhi, D.; Zhao, Y.; Zhang, H.; Wang, X.; Ru, Y.; Li, H. Lateral flow test strip based on colloidal selenium immunoassay for rapid detection of melamine in milk, milk powder, and animal feed. Int. J. Nanomed. 2014, 9, 1699–1707. [CrossRef] [PubMed] Faulk, W.P.; Taylor, G.M. An immunocolloid method for the electron microscope. Immunochemistry 1971, 8, 1081–1083. [PubMed] Paek, S.H.; Lee, S.H.; Cho, J.H.; Kim, Y.S. Development of rapid one-step immunochromatographic assay. Methods 2000, 22, 53–60. [CrossRef] [PubMed] Liu, B.H.; Hung, C.T.; Lu, C.C.; Chou, H.N.; Yu, F.Y. Production of monoclonal antibody for okadaic acid and its utilization in an ultrasensitive enzyme-linked immunosorbent assay and one-step immunochromatographic strip. J. Agric. Food Chem. 2014, 62, 1254–1260. [CrossRef] [PubMed] Sun, Y.; Xing, G.; Yang, J.; Wang, F.; Deng, R.; Zhang, G.; Hu, X.; Zhang, Y. Development of an immunochromatographic test strip for simultaneous, qualitative and quantitative detection of ochratoxin A and zearalenone in cereal. J. Sci. Food Agr. 2015, 96, 3673–3678. [CrossRef] [PubMed]
Toxins 2018, 10, 354
21.
22. 23. 24.
25. 26.
27.
28. 29. 30.
31. 32. 33.
34.
14 of 14
Abdelhamid, A.M.; Dorra, T.M. Study on effects of feeding laying hens on separate mycotoxins (aflatoxins, patulin, or citrinin)-contaminated diets on the egg quality and tissue constituents. Arch. Anim. Nutr. 1990, 40, 305–316. [CrossRef] Roberts, W.T.; Mora, E.C. Toxicity of Penicillium citrinum AUA-532 contaminated corn and citrinin in broiler chicks. Poult. Sci. 1978, 57, 1221–1226. [CrossRef] [PubMed] Zhu, D.; Zhang, H.; Bing, X. Preparation of an immunoaffinity column for the clean-up of fermented food samples contaminated with citrinin. Food Addit. Contam. 2013, 30, 389–394. [CrossRef] [PubMed] Dey, M.; Mayo, J.A.; Saville, D.; Wolyniak, C.; Klontz, K.C. Recalls of foods due to microbiological contamination classified by the U.S. Food and Drug Administration, fiscal years 2003 through 2011. J. Food Protect. 2013, 76, 932–938. [CrossRef] [PubMed] Schwenk, E.; Alexander, G.J.; Gold, A.M.; Stevens, D.F. Biogenesis of citrinin. J. Biol. Chem. 1958, 233, 1211–1213. [PubMed] Duan, Z.H.; Lin, Z.S.; Yao, H.R.; Gao, Y.H.; Zhang, K.; Zhao, S.Q.; Zhu, Z.Y. Preparation of artificial antigen and egg yolk-derived immunoglobulin (IgY) of citrinin for enzyme-linked immunosorbent assay. Biomed. Environ. Sci. 2009, 22, 237–243. [CrossRef] Kong, D.; Xie, Z.; Liu, L.; Song, S.; Kuang, H. Development of ic-ELISA and lateral-flow immunochromatographic assay strip for the detection of citrinin in cereals. Food Agric. Immunol. 2017, 28, 754–766. [CrossRef] Yu, F.Y.; Chi, T.F.; Liu, B.H.; Su, C.C. Development of a sensitive enzyme-linked immunosorbent assay for the determination of ochratoxin A. J. Agric. Food Chem. 2005, 53, 6947–6953. [CrossRef] [PubMed] Cho, I.H.; Ku, S. Current technical approaches for the early detection of foodborne pathogens: Challenges and opportunities. Int. J. Mol. Sci. 2017, 18, 2078–2100. [CrossRef] [PubMed] Bastus, N.G.; Comenge, J.; Puntes, V. Kinetically controlled seeded growth synthesis of citrate-stabilized gold nanoparticles of up to 200 nm: Size focusing versus Ostwald ripening. Langmuir 2011, 27, 11098–11105. [CrossRef] [PubMed] Frens, G. Controlled nucleation for the regulation of the particle size in monodisperse gold suspensions. Nat. Phys. 1973, 241, 20–22. [CrossRef] Yu, F.Y.; Liu, B.H.; Chou, H.N.; Chu, F.S. Development of a sensitive ELISA for the determination of microcystins in algae. J. Agr. Food Chem. 2002, 50, 4176–4182. [CrossRef] Liu, B.H.; Tsao, Z.J.; Wang, J.J.; Yu, F.Y. Development of a monoclonal antibody against ochratoxin A and its application in enzyme-linked immunosorbent assay and gold nanoparticle immunochromatographic strip. Anal. Chem. 2008, 80, 7029–7035. [CrossRef] [PubMed] Tsao, Z.J.; Liao, Y.C.; Liu, B.H.; Su, C.C.; Yu, F.Y. Development of a monoclonal antibody against domoic acid and its application in enzyme-linked immunosorbent assay and colloidal gold immunostrip. J. Agric. Food Chem. 2007, 55, 4921–4927. [CrossRef] [PubMed] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
