**2. Lateral flow immunoassays for aflatoxins**

Since the early 2000's, scientific papers and commercial devices aimed at measuring myco‐ toxinsin food and feed have appeared, and recentlya certain amount of literature on this topic has become available, including comprehensive reviews [42-44]. In particular, some LFIAs for the qualitative and semi-quantitative detection of aflatoxins in food and feed have been described and will be discussed below. At the same time, commercial LFDs for the de‐ tection of aflatoxinsin various commodities have become available and some of them have also been validated by USDA-GIPSA [45].

#### **2.1. Principle of the method**

from the 50 ng kg-1 established by the EU to the 500 ng kg-1 established by US FDA [2, 8]. More restrictive MRLs have been decided by the EU for the presence of AFM1 in baby food [2].

A part from safety issue, food contamination caused by AFs also strongly affects economic interests; so much effortis devoted to the development of analytical methods for detecting these contaminants. Newly developed methods of analysis are intended both for screening purposes (rapid, economic and simple methods) and for the accurate, reproducible and sen‐

Numerous chromatographic methods to detect AFs in foods have been developed, coupled to fluorescent or mass spectrometric detection [9-11]. Likewise, several methods for aflatoxin M1 determination in milk based on high-performance liquid chromatography associated to fluorescence or mass spectrometric detection have been developed [12-13]. However, chro‐ matographic techniques are mainly used in confirmatory analyses and are usually not ap‐ plied to routine controls owing to the necessity to use expensive equipment and extensive

The first rapid methods of analysis for AFs were based on Thin Layer Chromatography [14]; this technique is still used today even though in a significant lesser extent compared to methods based on the use of antibodies. Immunochemical methods of analysis are widely employed as screening methods for measuring AFs in food and feed [9, 14-18] and also for AFM1 quantification in milk and dairy products [19-21] thanks to their rapidity, selectivity and sensitivity. Several ELISA kits are commercially available, whose performances are gen‐ erally adequate to meet legal requirements, and are routinely employed for aflatoxin moni‐ toring. Some of these methods have also been validated [17-18]. However, even immunoassays need to be run in a laboratory, use a minimum of equipment and occasional‐ ly require some sample treatments, which may also involve the use of hazardous chemicals. Instead, affordable monitoring of food contaminants requires the highest-through put and more economical methods of detection and, possibly, little or no sample treatment, userfriendliness, employment of non-hazardous chemicals, in situ applicability. Additional requisites in aflatoxin detection would be low detection limits (especially for aflatoxin M1)

Several innovative strategies have been proposed for the rapid, qualitative, semi-quantita‐ tive or quantitative detection of aflatoxins, also based on the use of specific antibodies with‐ out constraints of classical immunoassays [22]. For example, an interesting qualitative approach has been described for the detection of AFM1 in milk [23-24]. The proposed meth‐ od is based on a flow-through immunoassay with visual detection. Main advantages are represented by the high sensitivity and by the on site applicability of the assay which does not require any equipment for the treatment of the sample, norfor the analysis. In addition, it allowed the possibility of obtaining sample pre-concentration and/or clean-up in the same device used for the analysis [25]. Nevertheless, this method implies several subsequent steps to be carried out, thus limiting simplicity and rapidness of use. Very recently, the same ap‐ proach has also been demonstrated for the multi-detection of different mycotoxins, thus in‐

and adaptability to very differing commodities (for aflatoxins B and G).

creasing its potentiality of utilization [26].

sitive quantification by confirmatory methods.

316 Aflatoxins - Recent Advances and Future Prospects

clean-up steps.

As aflatoxins are low-molecular-mass compounds, immunoassays in competitive formats should be conceived to measure them. The same principles and reagents as in the micro‐ well-type immunoassays could be applied, except for the fact that,in LFIA, the separation of bound and unbound antibody sites is obtained by means of the lateral flow on a suitable support (the membrane). The liquid flow transports immunoreagents along the membrane where they encounter their partners in spatially confined zones of the membrane itself where immuno reactions take place.

for example, coloured or fluorescent nanoparticles. However, gold nanoparticles (GNPs) are generally employed, with few exceptions, because of some characteristics which make them particularly suitable for the purpose. First, the conjugation of antibodies with GNPs is very easily obtained by simply mixing the two components at a proper pH (at or above the pI of the antibodies). The preparation and characterisation of stable colloidal solution of GNPs al‐ so follows well-established, easy protocols and a wide literature is available on this topic. The surface plasmon resonance of GNPs determines an intense colour of colloidal gold, which varies from orange to pink depending on particle dimensions and on surface overlay, therefore coloured nanoparticles can be prepared and the colour nuance can be use to moni‐

Lateral Flow Immunoassays for Aflatoxins B and G and for Aflatoxin M1

http://dx.doi.org/10.5772/51777

319

The principles of the indirect competitive LFIAs which exploit GNP-labelled antibodies

Briefly, the labelled specific antibody is suspended in the liquid sample and flows through the membrane where it first encounters the antigen coated in a zone indicated as "Test line" (T-line). In the absence of the target compound (negative sample, Figure 2), labelled antibod‐ ies bind to the coated antigen and are focused on the T-line, so that a visible (detectable) line

Usually, a second so-called "Control line" (C-line) follows and is constituted by secondary

**Figure 2.** A lateral flow immunoassay in the indirect formatwith GNP-labelled antibodies for a negative sample (no AF is present). The Test line is made by a protein conjugate of the target toxin, while the Control line is constituted of anti-species antibodies. Anti-aflatoxin antibodies mixed together with non-specific γ-globulins (both GNP-labelled) move along the membrane. Anti-AF antibodies bind the antigen coated in the Test zone and are focused, thus form‐ ing a visible (detectable) line. Non-specific γ-globulins pass the Test line and are captured by the anti-species antibod‐

The appearance of a C-line can be regarded simply as the confirmation of the correct devel‐ opment of the assay (reagents and materials integrity) or else can be exploited to calculate the T/C signal ratio with the aim of normalizing strip-to-strip variations [46] or can also be regarded as an internal standard to which the intensity of the T-line is compared to deter‐

tor preparation and conjugation to antibodies.

is formed.

have been widely described and are schematized in Figure 2 and 3.

anti-species antibodies which capture any excess of specific antibodies.

ies in the Control line where they are focused and form a second visible (detectable) line.

mine positivity/negativity [47-48].

Besides the porous membrane which assures the flow, lateral flow devices (LFDs) usually include: an absorbent pad (positioned at the top of the membrane to increase the volume of the flowing liquid), a sample pad (to assure contact between the liquid sample and the membrane), and a rigid backing (Figure 1). A release pad can be added, whose role is to ad‐ sorb labelled antibodies in such a way that they are included in the device itself.

The membrane is almost exclusively made of nitrocellulose (NC); sample and adsorbent pads are usually made of cellulose, although sample pads could also be made of glass fibre or other materials and sometimes are soaked with proteins and/or surfactants for special ap‐ plications. Release pads are usually glass fibre pads. Lines are traced on the NC membrane by means of dedicated dispensers which enables the dispensing of small volumes (typically few µl per cm) with high reproducibility.

**Figure 1.** Schematic of a lateral flow device in the dipstick format.

The simplest LFD is a dipstick, which is dipped directly into the sample solution. Labelled antibodies can be added to the sample as a concentrated suspension or provided in a lyophi‐ lized form to be re-suspended by the sample itself. Alternatively, the labelled antibody can be pre-adsorbed onto the releasing pad, which partially overlaps the membrane. The liquid sample itself causes the re-suspension of the pre-adsorbed labelled antibodies during the as‐ say. The sample pad is added in such a way that it overlaps the membrane or the releasing pad. Its role is the reduction of matrix interference by filtration alone or combined with some chemical action by means soaked reagents.

Besides the most popular dipstick format, LFDs exist in which the strip is inserted into a rig‐ id plastic cassette provided with a sample well and a reading window. The main advantage of these housings is the guarantee of a reproducible compression of all components in the overlapping zones, which assures faster and more reproducible flows.

With few exceptions, the indirect competitive format, in which the antigen (a protein conju‐ gate of the target toxin) is coated on the membrane and the antibody is labelled, is strongly preferred in the development of LFIA for AFs. Antibody labelling can be obtain by using virtually whatever nanoparticles that have a spectroscopically detectable property, such as, for example, coloured or fluorescent nanoparticles. However, gold nanoparticles (GNPs) are generally employed, with few exceptions, because of some characteristics which make them particularly suitable for the purpose. First, the conjugation of antibodies with GNPs is very easily obtained by simply mixing the two components at a proper pH (at or above the pI of the antibodies). The preparation and characterisation of stable colloidal solution of GNPs al‐ so follows well-established, easy protocols and a wide literature is available on this topic. The surface plasmon resonance of GNPs determines an intense colour of colloidal gold, which varies from orange to pink depending on particle dimensions and on surface overlay, therefore coloured nanoparticles can be prepared and the colour nuance can be use to moni‐ tor preparation and conjugation to antibodies.

where they encounter their partners in spatially confined zones of the membrane itself

Besides the porous membrane which assures the flow, lateral flow devices (LFDs) usually include: an absorbent pad (positioned at the top of the membrane to increase the volume of the flowing liquid), a sample pad (to assure contact between the liquid sample and the membrane), and a rigid backing (Figure 1). A release pad can be added, whose role is to ad‐

The membrane is almost exclusively made of nitrocellulose (NC); sample and adsorbent pads are usually made of cellulose, although sample pads could also be made of glass fibre or other materials and sometimes are soaked with proteins and/or surfactants for special ap‐ plications. Release pads are usually glass fibre pads. Lines are traced on the NC membrane by means of dedicated dispensers which enables the dispensing of small volumes (typically

The simplest LFD is a dipstick, which is dipped directly into the sample solution. Labelled antibodies can be added to the sample as a concentrated suspension or provided in a lyophi‐ lized form to be re-suspended by the sample itself. Alternatively, the labelled antibody can be pre-adsorbed onto the releasing pad, which partially overlaps the membrane. The liquid sample itself causes the re-suspension of the pre-adsorbed labelled antibodies during the as‐ say. The sample pad is added in such a way that it overlaps the membrane or the releasing pad. Its role is the reduction of matrix interference by filtration alone or combined with

Besides the most popular dipstick format, LFDs exist in which the strip is inserted into a rig‐ id plastic cassette provided with a sample well and a reading window. The main advantage of these housings is the guarantee of a reproducible compression of all components in the

With few exceptions, the indirect competitive format, in which the antigen (a protein conju‐ gate of the target toxin) is coated on the membrane and the antibody is labelled, is strongly preferred in the development of LFIA for AFs. Antibody labelling can be obtain by using virtually whatever nanoparticles that have a spectroscopically detectable property, such as,

overlapping zones, which assures faster and more reproducible flows.

sorb labelled antibodies in such a way that they are included in the device itself.

where immuno reactions take place.

318 Aflatoxins - Recent Advances and Future Prospects

few µl per cm) with high reproducibility.

**Figure 1.** Schematic of a lateral flow device in the dipstick format.

some chemical action by means soaked reagents.

The principles of the indirect competitive LFIAs which exploit GNP-labelled antibodies have been widely described and are schematized in Figure 2 and 3.

Briefly, the labelled specific antibody is suspended in the liquid sample and flows through the membrane where it first encounters the antigen coated in a zone indicated as "Test line" (T-line). In the absence of the target compound (negative sample, Figure 2), labelled antibod‐ ies bind to the coated antigen and are focused on the T-line, so that a visible (detectable) line is formed.

Usually, a second so-called "Control line" (C-line) follows and is constituted by secondary anti-species antibodies which capture any excess of specific antibodies.

**Figure 2.** A lateral flow immunoassay in the indirect formatwith GNP-labelled antibodies for a negative sample (no AF is present). The Test line is made by a protein conjugate of the target toxin, while the Control line is constituted of anti-species antibodies. Anti-aflatoxin antibodies mixed together with non-specific γ-globulins (both GNP-labelled) move along the membrane. Anti-AF antibodies bind the antigen coated in the Test zone and are focused, thus form‐ ing a visible (detectable) line. Non-specific γ-globulins pass the Test line and are captured by the anti-species antibod‐ ies in the Control line where they are focused and form a second visible (detectable) line.

The appearance of a C-line can be regarded simply as the confirmation of the correct devel‐ opment of the assay (reagents and materials integrity) or else can be exploited to calculate the T/C signal ratio with the aim of normalizing strip-to-strip variations [46] or can also be regarded as an internal standard to which the intensity of the T-line is compared to deter‐ mine positivity/negativity [47-48].

When the target is present above the lower detectable concentration level (positive sample, Figure 3), binding of labelled antibodies to the coated antigen is inhibited, resulting in a non-visible (undetectable) T-line.

**2.2. LFIAs for aflatoxins B and G**

in a total of 12 minutes, including sample preparation [49].

60 samples through comparison with ELISA.

and HPLC confirmatory analyses was also reported [55].

The first LFIA aimed at measuring any one of aflatoxins appeared in the scientific literature ten years ago and was one of the first reported lateral flow assays for food contaminants. The authors described a simplified device formed by aNC membrane on which the T-line had been traced upon by dispensing antibodies towards AFB1. The signal reporters were lip‐ osomes, which were tagged with AFB1 and encapsulated a visible dye. The tagged lipo‐ somes flowed along the membrane where encountered the coated anti-AFB1 antibodies and were captured, thus determining the appearance of a coloured T-line due to the focalization of the encapsulated dye. If some AFB1 was present in the sample, the binding of the tagged liposomes to the coated antibodies was inhibited and the colour of the T-line faded. The ab‐ solute limit of detection of such a device was 18 ng of AFB1 and the test could be completed

Lateral Flow Immunoassays for Aflatoxins B and G and for Aflatoxin M1

http://dx.doi.org/10.5772/51777

321

Apart from this early approach, following papers described more usual LFDs based on the use of GNPs as antibody labels. In 2005, Delmulle and co-workers reported the development of a dipstick which allowed authors to detect AFB1in pig feed. The visual detection limit (VDL) was set at 5 µg kg-1 and the analysis could be completed in 10 minutes [50]. In the same year, the group of Xiulianal so described the preparation of GNP-labelled antibodies towards AFB1 and their exploitation in a visual LFIA [51]. The application of the developed dipstick to measure AFB1 in rice, corn, and wheat was reported in a following paper of the same group [52]. The described LFD showed a VDL of 2.5 µg l-1 in buffer, which became 0.05 µg l-1 when the colour intensity of lines was determined by means of a photometric reader. Therefore, a sensitive quantification of the target toxin (limit of detection, LOD, 2 µg kg-1 in food) could be demonstrated; moreover, accuracy of the developed assay was confirmed on

A visual LFIA for detecting AFB1 was also described by papers of Shim et al [53-54]. The de‐ veloped LFD was shown to cross-react to some extent to other major aflatoxins (AFB2, AFG1, and AFG2) but not to differing mycotoxins (such as ochratoxin A, citrinine, patulin, zearale‐ none, and T-2 toxin). Nevertheless, it was applied for selectively measuring the sole AFB1in rice, barley and feed. VDLs of 5-10 µg l-1(rice, barley) and 10-20 µg l-1(feed) were obtained and the proposed method showed agreeing results towards HPLC analysis on up to 172 food and feed samples. The same group also published results obtained with a multi-analyte device aimed at contemporary measuring AFB1 and ochratoxin A in feed. The described method allowed the simultaneous detection of the two toxins which could be completed in 15 minutes and showed a VDL of 10 µg kg-1for AFB1. Method validation by means of ELISA

Although regulations prescribe the simultaneous determination of AFB1, AFB2, AFG1, and AFG2 beside AFB1 quantification, most papers described LFIA selective towards AFB1.To meet the need of measuring all the four major AFs our group developed a quantitative LFIA for total aflatoxin determination in corn samples. The assay could be completed in 10 mi‐ nutes, showed a LOD of 10 µg l-1 and was validated through comparison with HPLC on 25 samples. In addition, an aqueous extracting medium was also optimized and proven to al‐ low reliable quantification of total aflatoxin [56]. Except in this case, AFs were always ex‐

Interpretation of assay results depends on the presence and intensity of both Test and Con‐ trol lines as schematized in Figure 4.

**Figure 3.** A lateral flow immunoassay in the indirect format for a positive sample (AF above the detectable limit). GNPlabelled anti-aflatoxin antibodies and non-specific γ-globulins move along the membrane. Anti-AF antibodies bind the toxin in the sample and the interaction with coated antigen is thus inhibited. Non-specific γ-globulins pass the T-line and are captured by the anti-species antibodies in the Control line where they are focused. Therefore, a single line (Cline) appears on the membrane.

**Figure 4.** Assay result interpretation. Two intense lines: valid test, negative sample (target toxin below the detection limit of the method); intense C-line and fading T-line: valid test, the amount of the target toxin is near to the detection limit; intense C- line: test valid, positive sample (target toxin above the detection limit); intense or fading T-line: invalid test.

## **2.2. LFIAs for aflatoxins B and G**

When the target is present above the lower detectable concentration level (positive sample, Figure 3), binding of labelled antibodies to the coated antigen is inhibited, resulting in a

Interpretation of assay results depends on the presence and intensity of both Test and Con‐

**Figure 3.** A lateral flow immunoassay in the indirect format for a positive sample (AF above the detectable limit). GNPlabelled anti-aflatoxin antibodies and non-specific γ-globulins move along the membrane. Anti-AF antibodies bind the toxin in the sample and the interaction with coated antigen is thus inhibited. Non-specific γ-globulins pass the T-line and are captured by the anti-species antibodies in the Control line where they are focused. Therefore, a single line (C-

**Figure 4.** Assay result interpretation. Two intense lines: valid test, negative sample (target toxin below the detection limit of the method); intense C-line and fading T-line: valid test, the amount of the target toxin is near to the detection limit; intense C- line: test valid, positive sample (target toxin above the detection limit); intense or fading T-line: invalid test.

non-visible (undetectable) T-line.

320 Aflatoxins - Recent Advances and Future Prospects

trol lines as schematized in Figure 4.

line) appears on the membrane.

The first LFIA aimed at measuring any one of aflatoxins appeared in the scientific literature ten years ago and was one of the first reported lateral flow assays for food contaminants. The authors described a simplified device formed by aNC membrane on which the T-line had been traced upon by dispensing antibodies towards AFB1. The signal reporters were lip‐ osomes, which were tagged with AFB1 and encapsulated a visible dye. The tagged lipo‐ somes flowed along the membrane where encountered the coated anti-AFB1 antibodies and were captured, thus determining the appearance of a coloured T-line due to the focalization of the encapsulated dye. If some AFB1 was present in the sample, the binding of the tagged liposomes to the coated antibodies was inhibited and the colour of the T-line faded. The ab‐ solute limit of detection of such a device was 18 ng of AFB1 and the test could be completed in a total of 12 minutes, including sample preparation [49].

Apart from this early approach, following papers described more usual LFDs based on the use of GNPs as antibody labels. In 2005, Delmulle and co-workers reported the development of a dipstick which allowed authors to detect AFB1in pig feed. The visual detection limit (VDL) was set at 5 µg kg-1 and the analysis could be completed in 10 minutes [50]. In the same year, the group of Xiulianal so described the preparation of GNP-labelled antibodies towards AFB1 and their exploitation in a visual LFIA [51]. The application of the developed dipstick to measure AFB1 in rice, corn, and wheat was reported in a following paper of the same group [52]. The described LFD showed a VDL of 2.5 µg l-1 in buffer, which became 0.05 µg l-1 when the colour intensity of lines was determined by means of a photometric reader. Therefore, a sensitive quantification of the target toxin (limit of detection, LOD, 2 µg kg-1 in food) could be demonstrated; moreover, accuracy of the developed assay was confirmed on 60 samples through comparison with ELISA.

A visual LFIA for detecting AFB1 was also described by papers of Shim et al [53-54]. The de‐ veloped LFD was shown to cross-react to some extent to other major aflatoxins (AFB2, AFG1, and AFG2) but not to differing mycotoxins (such as ochratoxin A, citrinine, patulin, zearale‐ none, and T-2 toxin). Nevertheless, it was applied for selectively measuring the sole AFB1in rice, barley and feed. VDLs of 5-10 µg l-1(rice, barley) and 10-20 µg l-1(feed) were obtained and the proposed method showed agreeing results towards HPLC analysis on up to 172 food and feed samples. The same group also published results obtained with a multi-analyte device aimed at contemporary measuring AFB1 and ochratoxin A in feed. The described method allowed the simultaneous detection of the two toxins which could be completed in 15 minutes and showed a VDL of 10 µg kg-1for AFB1. Method validation by means of ELISA and HPLC confirmatory analyses was also reported [55].

Although regulations prescribe the simultaneous determination of AFB1, AFB2, AFG1, and AFG2 beside AFB1 quantification, most papers described LFIA selective towards AFB1.To meet the need of measuring all the four major AFs our group developed a quantitative LFIA for total aflatoxin determination in corn samples. The assay could be completed in 10 mi‐ nutes, showed a LOD of 10 µg l-1 and was validated through comparison with HPLC on 25 samples. In addition, an aqueous extracting medium was also optimized and proven to al‐ low reliable quantification of total aflatoxin [56]. Except in this case, AFs were always ex‐ tracted in methanol/water (typically 70/30 or 80/20 v/v) followed by dilution of the extract before LFIA analysis to reduce the proportion of the organic solvent, which is hardly com‐ patible with materials composing LFDs. However, a methanol amount lower than 15-20% has been demonstrated by most authors to be compatible with LFD materials and further more not to affect immunoassay performance.

with blank samples as calibrators [56, 61-63]. Matrix components not only interfere in defin‐ ing appropriate standards for calibration but also determine requirement of distinct devices

Lateral Flow Immunoassays for Aflatoxins B and G and for Aflatoxin M1

http://dx.doi.org/10.5772/51777

323

Despite the fact that the some authors reported calibration by using standard AFs diluted in buffers (to which methanol is added in limited proportions, as discussed above) and stated no interference from matrix given a limited dilution of sample extracts, the application of LFDs for the effective AF B and G detection in food remains the bottleneck in the develop‐ ment of new LFIAs. This taskis also made particularly complex by the multiplicity andvari‐

The development of LFIAs for AFM1 is one of the most challenging goals in this field of re‐ search because of the extreme sensitivity required by legislation for this contaminant (partic‐

The first paper dealing with the subject reported a validation study on a commercial device which was conceived for meet US regulations and did not described any preparation proto‐ cols and methods. The ROSA Charm Aflatoxin M1™ aimed at quantitatively measuring AFM1 in milk was validated as the result of an inter-laboratory trial, which involved 21 par‐ ticipants, at four levels above and two below the declared LOD of the assay (400 ng l-1) [64]. Less than 5% of false negative (n=83) and no false positive below 300 ng l-1 were found. For

More recently, Wang et al first described the development of a LFD for the detection of AFM1 [65]. The cut-off level (0.5-1 µg l-1) is just above the eligible value required by the US regulation [8] and far beyond the more severe limits imposed by the European Un‐ ion for this contaminant [2]. However, it is an effectively sensitive and rapid assay, provid‐ ed that the whole analytical procedure can be completed in 10 minutes, as no sample

A visual device has also been developed by Zhang et al which showed a VDL for AFM1of 0.3 µg l-1 [66]. Although the sensitivity improvement respect to the work of Wang et al, the obtained VDL remains far away from the detectability demand imposed by EU MRLs for

**3. Development of a highly sensitive LFIA for measuring AFM1 in milk**

With the aim of producingasystem sensitive enough to reach the limits imposed by Europe‐ an regulations, we developed a competitive lateral flow immunoassay which exploited rab‐ bit polyclonal antibodies towards AFM1that had been previously employed in the development of a sensitive ELISA [19]. A classic device, including a NC membrane (onto which the two lines of reagents had been immobilized), cellulose sample and adsorbent

contaminations between 350 and 450 ng l-1 false positivity increased from 21 to 93%.

to be developed for individual foods.

**2.3. LFIAs for aflatoxin M1**

ularly in the European Union).

treatment is required.

this contaminant.

ety of matrices to be considered in aflatoxin B and G analysis.

Most recent contributes to the topic are due to the group of Zhang and co-workers, who de‐ scribed two LFDs, the first highly selective towards AFB1 and the second able to measure total aflatoxins [57-58]. Both devices have been applied to visually detect target toxins in peanuts (the highly selective one could also be exploited to detect AFB1 in pu-erh tea, vege‐ table oil and feed). Both methods allowed reliable results (agreeing with HPLC determina‐ tion) to be obtain in 15 minutes. In addition, the LFIA aimed at measuring total AFs was extremely sensitive, with VDL in peanut extracts as low as 0.03, 0.06, 0.12, and 0.25 µg l-1for AFB1, AFB2, AFG1, and AFG2, respectively.

In addition to papers aimed at describing actually functioning devices for measuring AFs, those targets have often been chosen as system models for the development of original devi‐ ces which exploited non-traditional signal reporters to label antibodies. Besides the above mentioned approach of Ho and Wauchope, based on the use of dye-encapsulating lipo‐ somes, Liao and Li described a visual device which exploited nanoparticles with a silver core and a gold shell as the reporters in the construction of a LFD for AFB1. The toxin was determined in cereals and nuts and performances were compared to those of a GNP-based LFIA and to results obtained through a classic microwell-based immunoassay. The authors demonstrated that the newly developed LFD was comparable to the GNP-LFD in terms of stability of components and reproducibility of signals. On the other hand, it allowed a great enhancement in sensitivity so that values as low as 0.1 µg l-1AFB1 could be measured [59].

With the expectation of increasing the useful signal, therefore being able to reduce immu‐ nore agents for the benefits of the competition, magnetic nanogold microspheres with a Fe2O3 core and a shell of multiple GNPs have also been proposed. The magnetic core of par‐ ticles allowed authors to simplify separation steps during the labelling of antibodies and their micro- dimensions to enhance colour during the test itself. A three-fold increase in sen‐ sitivity was stated for the visual detection of AFB2 compared to the use of simple gold col‐ loid nanoparticles [60].

#### *2.2.1. Application of LFIA for aflatoxins B and G in food analysis*

A major concern in the development of LFDs for aflatoxins is the unpredictable effects due to food components co-extracted from the sample beyond the target and which affect not only the antigen-antibody interaction on which the immunoassay is based, but also the me‐ chanics of the device itself.

Some authors experienced the apparently inexplicable failure of recovery experiments con‐ ducted on fortified materials and the incongruity of results attained for artificially and natu‐ rally contaminated samples, which necessitate matrix-matched calibrations and recommended the use of naturally contaminated samples blended in varying proportions with blank samples as calibrators [56, 61-63]. Matrix components not only interfere in defin‐ ing appropriate standards for calibration but also determine requirement of distinct devices to be developed for individual foods.

Despite the fact that the some authors reported calibration by using standard AFs diluted in buffers (to which methanol is added in limited proportions, as discussed above) and stated no interference from matrix given a limited dilution of sample extracts, the application of LFDs for the effective AF B and G detection in food remains the bottleneck in the develop‐ ment of new LFIAs. This taskis also made particularly complex by the multiplicity andvari‐ ety of matrices to be considered in aflatoxin B and G analysis.
