Intralaboratory development and evaluation of a high-performance liquid chromatography–fluorescence method for detection and quantitation of aflatoxins M 1,B 1,B 2,G 1,and G 2 in animal liver

Abstract

Aflatoxins are potent mycotoxins with effects that include hepatotoxicity, immunosuppression, and suppression of animal growth and production. The etiologic diagnosis of aflatoxicosis, which is largely based on analysis of contaminated feed matrices, has significant disadvantages given the fact that representative feed samples may not be available and feed-based test methods are not confirmatory of an etiologic diagnosis. A tissue-based analytical method for biomarkers of exposure would be valuable for confirmation of aflatoxicosis. We describe in-house development and evaluation of a high-performance liquid chromatographic method with fluorescence detection and precolumn derivatization for determination of aflatoxins M₁, B₁, B₂, G₁, and G₂ in animal liver. The method demonstrates good selectivity for the tested aflatoxins in the liver matrix. The overall range was 0.03–0.10 ng/g for limit of detection and 0.09–0.18 ng/g for limit of quantitation. The correlation coefficient (R²) of calibration curves was >0.9978 for AFM₁, 0.9995 for AFB₁, 0.9986 for AFB₂, 0.9983 for AFG₁, and 0.9980 for AFG₂. For fortification levels of 0.2–10 ng/g, repeatability was 10–18% for AFM₁, 7–14% for AFB₁, 5–14% for AFB₂, 6–16% for AFG₁, and 10–15% for AFG₂. Recovery was 52–57% for AFM₁, 54–62% for AFB₁, 55–61% for AFB₂, 57–67% for AFG₁, and 61–65% for AFG₂. There was no liver matrix effect found. The method is rugged against minor changes based on the selected factors. The results indicate that the proposed method is suitable for quantitative determination of aflatoxins M₁, B₁, B₂, G₁, and G₂ in liver.

Keywords

Aflatoxin chromatography evaluation fluorescence liver

Introduction

Aflatoxins B₁, B₂, G₁, and G₂ are mycotoxin derivatives of difuranocoumarin produced by many species of molds that grow on high-carbohydrate feeds. The main aflatoxin producers include Aspergillus fungal species, such as A. flavus and A. parasiticus, and Penicillium fungal species, such as P. puberulum.³⁴ Aflatoxins, as well as other mycotoxins, are secondary metabolites of fungi, and their function is not well known.^13,23 Aflatoxin-generating fungi have been found in a wide variety of feeds and feed ingredients including grains⁷ (e.g., barley and corn), rice, brazil nuts, pistachio nuts, peanuts and peanut products, wheat flour, and other food products such as figs, date fruits, and spices.²³ Of the animal feed ingredients, corn, peanuts, cotton seed, and dried distiller grains pose the highest risk of aflatoxin contamination.⁷ In feed commodities, the order of abundance from highest to lowest concentration is aflatoxin B₁ (AFB₁) > aflatoxin G₁ (AFG₁) > aflatoxin B₂ (AFB₂) > aflatoxin G₂ (AFG₂).³⁴ Once ingested, aflatoxins are absorbed and metabolized to other aflatoxin metabolites. These metabolites react with protein and DNA forming adducts that impair animal health. AFM₁ is a major hydroxylated metabolite of AFB₁ in animals, including humans, livestock, pets, and poultry.^13,14

In veterinary medicine, aflatoxicosis has been reported in virtually all species, including pets, livestock, poultry, and fish. Dogs, ducks, and fish are extremely sensitive to aflatoxins. The toxic outcome depends on the amount and frequency of exposure. Chronic aflatoxicosis in pets, especially dogs, is a well-recognized disease.^10,30,34 Chronically intoxicated dogs exhibit lethargy, jaundice, anorexia, weight loss, polyuria and polydipsia, and bleeding disorders^17,34 Feeding aflatoxin-contaminated feeds causes reduced growth rates in animals.³⁴ Aflatoxins are also mutagenic,^23,45 carcinogenic,^15,23 teratogenic,^11,23 and immunosuppressive.³³ Overall, aflatoxins exact significant negative health and economic impacts in livestock. They are a common cause of illness and death in pets that ingest contaminated feeds.^5,17,30,35 Aflatoxins have been confirmed by the International Agency for Research on Cancer as human carcinogens, responsible for causing hepatocellular carcinoma, more frequently in developing countries.^9,22

Considering the significant toxicity and negative health impacts of aflatoxins, various tests and procedures are currently in use for diagnosis of aflatoxicosis in animals. Test methods are currently available for determination of aflatoxins in animal feeds,^{12,16,19,21,38,39,41,43} in aqueous matrices,²⁶ and milk.^{6,10,27,29,37,46} These tests are used in conjunction with serum chemistry profiles showing elevated liver enzyme activity, and in case of deceased animals, liver lesions.³¹ However, there are many challenges in using feed samples for the etiologic diagnosis of aflatoxicosis. For example, in most cases, the causal feed has already been consumed and is no longer available for testing. Hence, there is usually a lack of “representative” feed samples for testing. Also, histopathology findings such as lipidosis and bile duct proliferation or elevated serum liver enzymes, although suggestive, are not pathognomonic of aflatoxicosis. For all these reasons, there is a need for a tissue-based analytical method to detect and quantify aflatoxins and/or their metabolites to aid in the diagnosis of aflatoxicosis in deceased animals. Demonstrated presence of aflatoxins in tissues along with consistent clinical signs, elevated liver enzymes in serum, and/or liver histopathology will provide veterinarians with information needed for a solid diagnosis of aflatoxicosis.

AFB₁, the most prevalent and potent aflatoxin in feeds, is a frequent cause of animal intoxications. It is metabolized in the liver to AFM₁, which is excreted through milk and/or urine in mammals. AFM₁ is recognized as a reliable biomarker of AFB₁ exposure in humans.^24,47,49 AFM₁ is also used as a biomarker of AFB₁ exposure in milk and milk products.^{6,10,27,29,37,46} Currently, there are no established test methods for detection and quantitation of aflatoxins in liver for veterinary applications. Existing methods are feed-based and are not ideal for tissue application. We have extended this observation by developing and evaluating a liver-based method for detection and quantitation of AFM₁, AFB₁, AFB₂, AFG₁, and AFG₂ using a simple, affordable, and more efficient clean-up procedure and high-performance liquid chromatography (HPLC)–fluorescence detection. We chose HPLC–fluorescence because we know from experience that this is the most sensitive means to detect and measure aflatoxins. Parameters evaluated included limit of detection (LOD), limit of quantitation (LOQ), selectivity, repeatability, recovery, matrix effect, and ruggedness. We also investigated liver matrix effects in this study because it is well known that matrix effects could affect the quality of analytical results.

Materials and methods

Instrumentation

Measurements were performed with an HPLC system^a equipped with a computer with an interface for system manipulation, data acquisition, and analysis, as well as a fluorescence detector^a using a C18 column^b and a C18 guard column.^c The mobile phase consisting of 18.2 Mohm·cm water^d and acetonitrile^e was pumped at a flow rate of 1 mL/min. A gradient elution was used to give optimal separation (Table 1). A total run time of 40 min was used. The injection volume was 20 µL. The separation was performed at room temperature. The optimized excitation and emission wavelengths for the fluorescence detector were 360 and 440 nm, respectively.

Table 1.

High-performance liquid chromatography gradient for analyzing aflatoxins in bovine liver.

Time (min)	Water (%)	Acetonitrile (%)
0–7.5	86	14
7.5–8.0	85	15
8.0–24.0	85	15
24–25.5	75	25
25.5–35.0	75	25
35.0–36.0	86	14
36.0–40.0	86	14

Samples

Experiments were performed using bovine and porcine liver samples, in order to determine if the method could perform well in multiple species. Bovine liver was purchased from the local market. Aflatoxin-negative porcine liver samples were collected from regular diagnostic case submissions at the Veterinary Diagnostic Laboratory, Iowa State University (Ames, Iowa). Liver was blended by a single-speed blender.^f

Chemicals and reagents

All aqueous solutions were prepared in 18.2 Mohm·cm water.^d The following reagents were used: acetonitrile, methanol, dichloromethane, glacial acetic acid, trifluoroacetic acid, citric acid monohydrate, sodium sulfate anhydrous, alumina neutral (60–325 mesh)^e; diatomaceous earth^g; octadecyl (C18) 40-μm preparation of LC packing^h; and AFM₁, AFB₁, AFB₂, AFG₁, and AFG₂ standards.ⁱ Citric acid solution (20%) was prepared by dissolving 20.0 g citric acid monohydrate in water and bringing the total volume to 100 mL. Acetonitrile:water (90:10, v/v) was prepared by adding 10 mL of water to 90 mL of acetonitrile and mixing well. The packing material of mixed-phase clean-up columns consisted of C18:alumina neutral (50:50, w/w), which was prepared by mixing 50 g of C18 LC packing and 50 g of alumina neutral and manually shaking for 5 consecutive min. The packaging material was manually shaken for 5 s at a pace of 1 shake per second before each use. C18:alumina neutral (50:50, w/w) clean-up columns were prepared in house by weighing 1.5 g of C18:alumina neutral (50:50, w/w) into 6-mL plastic syringes. The packing was immobilized by filter discsⁱ on both ends. Water:trifluoroacetic acid:glacial acetic acid (35:10:5, v/v) was prepared by mixing 10 mL of trifluoroacetic acid with 5 mL of glacial acetic acid and 35 mL of water and vortexing at maximum speed for 10 sec. It was protected from light by a layer of aluminum foil wrap. A mixed aflatoxin intermediate standard solution (250 ng/mL for each aflatoxin) was prepared by using stock standard solutions, evaporating the solvents to dryness under nitrogen, and diluting to 5 mL with methanol. This mixture of intermediate solution is stable for 1 month when stored at −20°C when not in use. A working standard solution of 25 ng/mL was prepared on the day of use by subsequent dilution of 1 mL of the 250 ng/mL standard solution using methanol and then bringing the volume to 10 mL.

Sample preparation

The method for sample preparation was adapted from published extraction methods,^1,42 with some modifications. Briefly, we used a much smaller sample size (~1.0 g) instead of 50 g⁴² or 100 g³⁰ recommended in the original methods. This was possible because our method showed good sensitivity and low LOD (pg/g level). Another improvement was adding a wash step after filtration, with sodium sulfate anhydrous. This wash step increased recovery significantly for all aflatoxins. Likely this step increased the amount of aflatoxins collected by reducing aflatoxins retained in the filtration device. The mixed-phase clean-up procedure was performed by using a C18:alumina neutral (50:50, w/w) column prepared in-house. We recommend this approach because it provides a more affordable option than the widely used but otherwise relatively expensive immunoaffinity columns.^27,29 A rinse step with 5.0 mL of acetonitrile:water (90:10, v/v) was added prior to loading the mixed-phase clean-up columns with testing solutions. This was performed in order to further clean up polar interferences that could otherwise co-elute with AFM₁.

For evaluation purposes, samples other than liver matrix blanks were fortified with AFM₁, AFB₁, AFB₂, AFG₁, and AFG₂ before extraction, using a composite aflatoxin working standard consisting of 25 ng/mL of each aflatoxin. The fortified samples were vortex-mixed for 10 s for even mixing before sample preparation. Each liver sample was then vortex-mixed with 1,000 µL of 20% citric acid solution, incubated at room temperature for 10 min, and vortex-mixed again. Diatomaceous earth^g (0.20 g) was subsequently vortex-mixed with the sample for 10 s. The mixture was extracted with 20.0 mL of dichloromethane by rotary-mixing for 35 min. The mixture was then centrifuged at 842 × g for 10 min. The bottom clear layer was subsequently aspirated and passed through 7.20 g of sodium sulfate anhydrous housed in a glass fiber filter and a polymer funnel. The filtrate was collected in a clean 50-mL tube. The sodium sulfate was subsequently washed with 6.0 mL of dichloromethane, and the wash was combined with the solution from the previous step. The combined solution was evaporated to dryness under nitrogen at ambient temperature.

The dried extracts were cleaned up using the following procedure (GE Rottinghaus, pers. comm., 2015). Extracts were first reconstituted with 5.0 mL of acetonitrile:water (90/10, v/v) and vortex-mixed for 10 s. An in-house prepared mixed-phase clean-up column (C18:alumina neutral, 50:50, w/w) was rinsed with 5.0 mL of acetonitrile:water (90:10, v/v) as preconditioning. The rinse was discarded. The reconstituted extracts were then loaded onto the mixed-phase clean-up columns and allowed to flow through the column by gravity. The rinse was eluted into 7-mL glass vials. After the solution had drained out of the column, the plungers were pushed down the barrels of the syringes to drive out the residual liquid in the column. The residual eluents were collected in the same vials. The cleaned-up extracts were subsequently concentrated to dryness under nitrogen. The dried extracts were then subjected to a derivatization procedure adapted from an AOAC method.² Briefly, the extracts were reconstituted in 400 μL of water:trifluoroacetic acid:glacial acetic acid (35:10:5, v/v), vortexed for 10 s, and heated at 65°C in heating dry bath for 15 min for derivatization. The major modification was that the mixtures were incubated for 12–16 h at ambient temperature before HPLC injection.

Evaluation protocol

Selectivity was evaluated by testing the liver sample fortified with each standard in triplicate in each run, for a total of 3 runs. When peaks were free of interference, it was an indication of good selectivity. In our case, selectivity was determined by testing liver samples fortified at 0.2 ng/g, which is usually considered more challenging than higher levels. A control liver matrix was tested as well for the purpose of interference examination.

For evaluation of the method LOD and LOQ, 20 negative control liver samples were extracted and analyzed.^4,40 The mean blank response and standard deviation (SD) corresponding to the elution times of each aflatoxin was calculated. LOD was expressed as (mean) + 3 × SD,^4,40 and LOQ was expressed as (mean) + 10 × SD.^4,40 The intensity signals were converted to concentration units using calibration curves established in fortified liver samples.

Repeatability was determined from 3 different runs of analyses, with each run consisting of 3 replicates fortified at 0.2, 2.0, and 10 ng/g aflatoxins, respectively. The relative SD (RSD) was used to evaluate repeatability. It was compared with the Horwitz ratio, which is a reference value for repeatability at various concentrations of interest.^20,25 A Horwitz ratio of 0.5–2 normally indicates satisfactory repeatability.

Recovery was evaluated by calculating the ratio of the peak area of liver samples fortified before extraction to those of liver samples fortified after extraction.^3,8,18 Both types of sample preparations were subsequently subjected to the same derivatization and HPLC procedure. Recovery was evaluated concurrently with repeatability. Therefore, a mean value ± SD was calculated to represent recovery for each level of each run based on 3 replicates. A mean ± SD was used to express the overall recovery for each level over all 3 runs.

The matrix effect of liver was determined by calculating the ratio of the peak area obtained from negative control bovine liver samples fortified with aflatoxins at 2.0 ng/g after sample preparation to the peak area obtained with the same amount of aflatoxin prepared in methanol.¹⁸ The matrix effect was based on 3 replicates of each type. Both types of samples were subsequently subjected to the same derivatization and HPLC procedure. Normally, a lack of matrix is indicated if the ratio falls within 100% ± (15–20%).^18,28,32

The last parameter that we investigated was ruggedness, which is a measure of the resiliency of the method to withstand minor deviations from the prescribed procedure. We evaluated the following factors: extraction time, amount of dichloromethane, amount of packing material for mixed-phase clean-up cartridge, source of trifluoroacetic acid, heating time for derivatization, derivatization temperature, and postderivatization incubation time. For each factor, the nominal approach was conducted along with the corresponding altered approach. A set of 8 samples was processed, with each sample being assigned to either nominal (prescribed) or altered approach for each factor (Table 2).^29,36,48 The nominal and altered level for each factor is shown in Table 3. Liver samples were fortified at 2.0 ng/g level. This level of fortification was chosen because it was in the middle of our calibration curves and was considered to be within the target analytical range.

Table 2.

Design of ruggedness experiment for determination of aflatoxins M₁, B₁, B₂, G₁, and G₂ in bovine liver.*

Selected factor	Experiment no.
Selected factor	1	2	3	4	5	6	7	8
Extraction time	A	A	A	A	a	a	a	a
Amount of C18:alumina neutral (50:50, w/w)	B	B	b	b	B	B	b	b
Trifluoroacetic acid	C	c	C	c	C	c	C	c
Heating time for derivatization	D	D	d	d	d	d	D	D
Derivatization temperature	E	e	E	e	e	E	e	E
Amount of dichloromethane	F	f	f	F	F	f	f	F
Postderivatization incubation time	G	g	g	G	g	G	G	g
Observed results	s	t	u	v	w	x	y	z

Uppercase letters represent nominal levels of the variables, and lowercase letters represent altered levels of the variables.

Table 3.

Nominal and altered levels for variables listed in the ruggedness test design for determination of aflatoxins M₁, B₁, B₂, G₁, and G₂ in bovine liver.*

Selected factor	Unit	Abbreviation	Nominal (A–G)	Altered (a–g)
Extraction time	min	A, a	35	37
Amount of C18:alumina neutral (50:50, w/w)	g	B, b	1.50	1.65
Trifluoroacetic acid	—	C, c	†	‡
Heating time for derivatization	min	D, d	15	20
Derivatization temperature	°C	E, e	65	70
Amount of dichloromethane	mL	F, f	20.0	19.0
Postderivatization incubation time	h	G, g	20	26

Dash (—) = not applicable. Uppercase letters represent nominal levels of the variables, and lowercase letters represent altered levels of the variables.

†

Acros Organic, Thermo Fisher Scientific Inc., Waltham, Massachusetts.

‡

Fisher Scientific USA, Pittsburgh, Pennsylvania.

For each factor, a value D_i was defined as the difference between the mean recovery of all nominal results and that of all altered results. $S_{D_{i}}$ was expressed as in equation (1).

S_{D_{i}} = \sqrt{2 \times \sum (\frac{D_{i}^{2}}{7})}

To determine the effect of all factors on ruggedness, $S_{D_{i}}$ was compared with the standard deviation of recovery at the same fortification level (2.0 ng/g in this case). If the former was greater than the latter, the method was not rugged against all the factors as a whole, and vice versa.

To determine if each factor alone had any effect on ruggedness, a t-value was obtained from equation (2).

t = \frac{\sqrt{n} \times | D_{i} |}{\sqrt{2} \times S D}

where n is the number of replicates processed for each level for each factor, which, in our study, was 4 for either nominal or altered level. SD is the standard deviation of the recovery of the method. This t-value was compared with the 2-tailed t-critical value for n − 1 degrees of freedom at 95% confidence level, where n is the number of replicates used in estimating SD. If the t-value was greater than the t-critical value for the varied factor, then the method was not rugged against such factor, and vice versa.

Results

AFM₁, AFG₁, AFB₁, AFG₂, and AFB₂ were selectively separated from each other and free of interferences in both a negative control bovine liver sample and in a bovine liver sample fortified with AFM₁, AFG₁, AFB₁, AFG₂, and AFB₂ at 0.2 ng/g (Fig. 1). LOD was 76 pg/g for AFM₁, 80 pg/g for AFB₁, 69 pg/g for AFB₂, 103 pg/g for AFG₁, and 31 pg/g for AFG₂ (Table 4). LOQ is 176 pg/g for AFM₁, 91 pg/g for AFB₁, 114 pg/g for AFB₂, 128 pg/g for AFG₁, and 88 pg/g for AFG₂. These results indicate that this method can be used to quantify aflatoxins in the liver even at concentrations much lower than 0.2 ng/g, which we choose as our lowest test concentration of the aflatoxins for this study.

Figure 1.

A representative chromatograph of a control bovine liver matrix (dotted line), and liver matrix fortified with aflatoxins M₁, G₁, B₁, G₂, and B₂ at 0.2 ng/g (solid line).

Table 4.

Limit of detection (LOD)* and limit of quantitation (LOQ)† for aflatoxins M₁, B₁, B₂, G₁, and G₂ in bovine liver.

Parameter	AFM₁	AFB₁	AFB₂	AFG₁	AFG₂
LOD (pg/g)	76	80	69	103	31
LOQ (pg/g)	176	91	114	128	88

LOD = (mean) + 3 × SD (standard deviation) of matrix blanks (control liver), 20 replicates, converted to concentration units using calibration curves established in fortified liver samples.

†

LOQ = (mean) + 10 × SD of matrix blanks (control liver), 20 replicates, converted to concentration units using calibration curves established in fortified liver samples.

Calibration curves were established using 4 series of fortified liver samples (Fig. 2). The peak area reported as fluorescence intensity was plotted against fortified aflatoxin standard in ng, and linearly regressed. The regression equations are y = (118,205 ± 22,378)x + (–2,269 ± 14,426), y = (369,654 ± 113,457)x + (–26,052 ± 36,616), y = (404,584 ± 86,139)x + (–17,825 ± 52,419), y = (91,424 ± 33,387)x + (–6,501 ± 10,311), and y = (204,563 ± 48,991)x + (1,679 ± 35,175) for AFM₁, AFG₁, AFB₁, AFG₂, and AFB₂, respectively. The correlation coefficient (R²) was >0.9978 for AFM₁, >0.9995 for AFB₁, >0.9986 for AFB₂, >0.9983 for AFG₁, and >0.9980 for AFG₂.

Figure 2.

A representative calibration curve of aflatoxin M₁ (AFM₁) (solid line, solid diamond data point), AFB₁ (solid line, empty square data points), AFB₂ (dashed line, solid triangle data points), AFG₁ (dashed line, empty circle data points), and AFG₂ (solid line, solid square data points) in fortified liver samples. The slopes and y-intercepts are reported as mean ± confidence interval at 95% confidence level.

For the fortification levels 0.2–10 ng/g, repeatability (in terms of RSD) was in the range of 10–18% for AFM₁, 7–14% for AFB₁, 5–14% for AFB₂, 6–16% for AFG₁, and 10–15% for AFG₂ (Table 5). The result was compared with the reference value obtained from the Horwitz equation [PRSD_r (%) = c^−0.15],^20,25 where c is the concentration in mass fraction (Table 6).^20,25 The comparison shows that all of the RSD values were well below the upper limit of Horwitz reference level, indicating satisfactory repeatability. For the fortification levels 0.2–10 ng/g, mean recovery was in the range of 52–57% for AFM₁, 54–62% for AFB₁, 55–61% for AFB₂, 57–67% for AFG₁, and 61–65% for AFG₂ (Table 7).

Table 5.

Repeatability results for aflatoxins M₁, B₁, B₂, G₁, and G₂ in bovine liver.*

Fortification level (ng/g)	Testing run	RSD (%)
Fortification level (ng/g)	Testing run	AFM₁	AFB₁	AFB₂	AFG₁	AFG₂
0.2	1	9	10	4	5	10
	2	8	5	6	3	3
	3	18	7	6	5	7
	Overall	16	11	5	6	14
2.0	1	16	11	7	18	12
	2	22	11	11	11	8
	3	11	7	8	9	10
	Overall	18	14	14	16	15
10.0	1	11	4	4	5	5
	2	10	7	7	7	5
	3	8	3	3	4	4
	Overall	10	7	7	11	10

Each testing run was based on 3 replicates for each fortification level. “Overall” results were based on 9 replicates, with 3 replicates each run, a total of 3 runs. RSD = relative standard deviation (intralaboratory), an indication of method repeatability.

Table 6.

Reference values determined by the Horwitz equation.

Concentration level (ng/g)	Lower limit (%)	Higher limit (%)
0.2	14	57
2.0	10	40
10.0	8	32

Table 7.

Mean recovery results for aflatoxins M₁, B₁, B₂, G₁, and G₂ in bovine liver for different fortification levels.*

Fortification level (ng/g)	Testing run	Mean recovery (%)
Fortification level (ng/g)	Testing run	AFM₁	AFB₁	AFB₂	AFG₁	AFG₂
0.2	1	56 ± 6	66 ± 9	68 ± 3	65 ± 4	75 ± 7
	2	53 ± 5	65 ± 4	53 ± 4	67 ± 2	56 ± 1
	3	56 ± 11	57 ± 5	45 ± 3	59 ± 4	65 ± 4
	Overall	55 ± 7	62 ± 7	55 ± 10	64 ± 5	65 ± 10
2.0	1	56 ± 9	64 ± 7	65 ± 5	64 ± 12	68 ± 8
	2	47 ± 10	48 ± 6	49 ± 6	53 ± 6	56 ± 5
	3	54 ± 6	51 ± 4	54 ± 5	54 ± 5	58 ± 6
	Overall	52 ± 9	54 ± 9	56 ± 8	57 ± 9	61 ± 8
10.0	1	66 ± 7	69 ± 2	67 ± 3	73 ± 3	69 ± 4
	2	58 ± 6	63 ± 4	63 ± 4	75 ± 5	72 ± 4
	3	47 ± 4	52 ± 2	53 ± 2	53 ± 2	54 ± 2
	Overall	57 ± 10	62 ± 8	61 ± 7	67 ± 11	65 ± 9

Recovery was evaluated by calculating the ratio of the peak area of liver samples fortified before extraction to those of liver samples fortified after extraction. Each testing run was based on 3 replicates for each fortification level. “Overall” results were based on 9 replicates, with 3 replicates each run, a total of 3 runs.

Results for matrix effect results for all 5 aflatoxins were within the acceptable range (<15–20%),^18,28,32 which indicates the absence of liver matrix effects (Table 8). For all 5 aflatoxins, the calculated $S_{D_{i}}$ value was 3.2–6.1% (Table 9), which was lower than the method precision of recovery (7.7–8.8%). This demonstrates that the method is rugged and robust against all of the altered factors tested in this experiment. The t-values for each studied factor were less than the t-critical value of 2.31 at 95% confidence level (Table 10). This further indicates that the method is rugged enough against each selected factor.

Table 8.

A summary of results of liver matrix effect test for aflatoxins M₁, B₁, B₂, G₁, and G₂ in bovine liver.*

	Normalized peak area (%)
	AFM₁	AFB₁	AFB₂	AFG₁	AFG₂
Standards in methanol	100 ± 12	100 ± 7	100 ± 8	100 ± 8	100 ± 8
Fortified liver samples	109 ± 4	108 ± 2	110 ± 1	106 ± 2	108 ± 1

Matrix effect was determined by calculating the ratio of the peak area of blank bovine liver samples fortified with aflatoxins at 2.0 ng/g after sample preparation to the peak area obtained with the same amount of aflatoxin prepared in methanol, with 3 replicates for each type. Note that results in the rows are similar for each aflatoxin, indicating there is no matrix effect.

Table 9.

Overall ruggedness test results for aflatoxins M₁, B₁, B₂, G₁, and G₂ in bovine liver.*

	AFM₁	AFB₁	AFB₂	AFG₁	AFG₂
$S_{D_{i}}$ (%)	3.2	6.1	5.9	6.0	5.3
Method precision (SD, %)	8.6	8.7	8.1	8.8	7.7

$S_{D_{i}} = \sqrt{2 \times \sum (\frac{D_{i}^{2}}{7})}$ , where D_i was defined as the difference between the mean result of recovery from all the nominal results and that of all the altered results for Factor i. SD = standard deviation of recovery, obtained in liver at 2.0 ng/g, from the 9 replicates analyzed for method recovery and precision.

Table 10.

Factor-specific ruggedness test results for aflatoxins M₁, B₁, B₂, G₁, and G₂ in bovine liver.*

Selected factor	t-value
Selected factor	AFM₁	AFB₁	AFB₂	AFG₁	AFG₂
Extraction time	0.13	0.59	0.65	0.82	0.57
Amount of C18:alumina neutral (50:50, w/w)	0.70	0.79	0.79	0.74	0.79
Trifluoroacetic acid	0.29	0.96	1.05	0.79	0.95
Heating time for derivatization	0.28	1.05	0.98	0.95	0.90
Derivatization temperature	0.08	0.51	0.67	0.44	0.78
Amount of dichloromethane	0.11	0.05	0.07	0.20	0.06
Postderivatization incubation time	0.53	0.37	0.39	0.56	0.05

$t = \frac{\sqrt{n} \times | D_{i} |}{\sqrt{2} \times S D}$ , where n = 4; SD was defined as the standard deviation of recovery, obtained in liver at 2.0 ng/g, from the 9 replicates analyzed for method recovery and precision; D_i was defined as the difference between the mean result of recovery from all the nominal results and that of all the altered results for Factor i. Two-tailed critical value t-critical = 2.31, ν = 9-1, 95% confidence level.

Because we used bovine liver for our method evaluation, questions could arise whether the method can be used to analyze liver of other species. For this reason, extensive testing was also done using porcine liver. Intra-run repeatability results for porcine liver were 4–16% at the fortification level of 0.2 ng/g for all 5 aflatoxins (Table 11). All RSD values met the criteria and were within the reference Horwitz value^20,25 (Table 6). These results from the porcine portion of the study suggest that the method is suitable for analysis of pig liver.

Table 11.

A summary of porcine liver intrarun repeatability test results for aflatoxins M₁, B₁, B₂, G₁, and G₂.*

	Fortification level (ng/g)	RSD (%)
	Fortification level (ng/g)	AFM₁	AFB₁	AFB₂	AFG₁	AFG₂
Porcine	0.2	16	13	9	4	12

RSD = relative standard deviation (intralaboratory), an indication of method repeatability. RSD was evaluated by triplicates for each aflatoxin.

Discussion

A major advantage of this method is that it requires a small sample size, ~1 g of liver. This means the method can also be used for testing liver from small animals such as cats, dogs, and birds.

One of the challenges we faced during the method evaluation was finding an ideal incubation time for derivatization. By lengthening the time for derivatization, we were able to completely derivatize AFM₁ and AFB₁ consistently. The adapted method² gives satisfactory results for AFM₁, AFB₁, AFB₂, AFG₁, and AFG₂. For AFM₁, however, it shows a much slower reaction rate. Therefore, an incubation step was added to allow for thorough derivatization of AFM₁. The time of incubation was optimized based on the peak area of fluorescence response. We determined that an incubation time of 12–16 h is required in order to obtain a stable response. Also, an additional step of preconditioning the mixed-phase clean-up columns was introduced to eliminate possible interference from these columns. This improved the selectivity of the method, especially for AFM₁.

Based on our observations in this study, we recommend that calibration curves be established in fortified liver instead of adding standards in pure solvents. Matrix-matched calibration curves produce much more repeatable results than solvent-matched calibration curves. We suspect that the liver matrix protects the analytes from adsorbing on the walls of the unsilanized glass containers, especially at the low concentrations of interest in this study. The low aflatoxin concentrations used in our study are clinically and diagnostically relevant. From our experience working with diagnostic cases, we expect that the liver aflatoxin concentrations in animals that have died of acute aflatoxicosis to be in the range of 1–10 ng/g.⁴⁴

Overall, the method is very sensitive and selective. The part-per-trillion levels of LOD and LOQ indicate that the method is very sensitive. This is a tremendous improvement to the adapted method⁴² where the instrument LOQ was 0.5 ng (AFB₁ and AFG₂) and 0.3 ng for AFM₁ compared to 2.8 pg (AFB₁), 2.9 pg (AFG₂), and 4.8 pg (AFM₁) from our method. Moreover, the adapted method⁴² did not mention the instrument LOQ for AFB₂ and AFG₁.The ruggedness test indicates that minor deviations from the method as described will not adversely affect results. However, of the selected factors we tested, we found the amount of C18:alumina neutral (50:50, w/w), the source of trifluoroacetic acid, and the heating time for derivatization for aflatoxins to be the most susceptible and most likely to cause erroneous results. We therefore highly recommend using the source of trifluoroacetic acid, heating time for derivatization, and the amount of C18:alumina neutral (50:50, w/w) recommended in our procedure for best performance of this method.

The overall mean recoveries of the different aflatoxins measured were 57–67%. These recoveries are acceptable considering the analytical range targeted is 1–10 ppb in liver tissue. Generally, recoveries are expected to be excellent at higher analytical concentrations such as the ppm range. Although recoveries were not good at the analytical range of the method, the recovery was consistent. This indicates that the method performance is reliable.

We were concerned about the method performance in testing for aflatoxins in livers from different animal species. We therefore also studied porcine liver. Our method shows that parameters tested in porcine liver including selectivity, within-run repeatability, and linearity of calibration curve established with fortified liver samples, are all similar to those obtained using bovine liver. This is a good indication that the method is potentially valid for analyzing livers from a variety of animal species.

We have developed and evaluated a method that is suitable for the quantitative measurement of AFM₁, AFB₁, AFB₂, AFG₁, and AFG₂ in animal liver. The method, which simultaneously detects 5 aflatoxins, is very selective. Selectivity is also confirmed by the absence of interference in the clinically relevant range of 1–10 ng/g. There was no matrix effect for all 5 aflatoxins. The method is also rugged against all factors tested. Detection and quantitation of aflatoxins in liver, when used in conjunction with histopathology and/or serum liver chemistry profiles consistent with liver injury will increase a diagnostician’s confidence in making a diagnosis of aflatoxicosis in deceased animals. Future work will focus on developing and evaluating a urine-based antemortem test for diagnosis of aflatoxicosis in pets and livestock.

Footnotes

Acknowledgements

We thank the reviewers for their suggestions, which have helped improve this article. We greatly appreciate the valuable input from Dr. Renate Reimschuessel, Dr. Andriy Tkachenko, and Sarah Nemser from the Veterinary Laboratory Investigation and Response Network (Vet-LIRN), Center for Veterinary Medicine, U.S. Food and Drug Administration. We are grateful to Dr. Chong Wang for his help on statistical analysis of experimental design and data. We also appreciate the technical assistance of Olha Ivanytska, Raymond Grover Jr., Dr. Larry Wulf, and Daniel Schultz from the Veterinary Diagnostic Laboratory, Iowa State University.

Authors’ contributions

D Shao, PM Imerman, and DE Schrunk contributed to conception and design of the study, and to acquisition, analysis, and interpretation of data. Steve M. Ensley contributed to conception of the study. WK Rumbeiha contributed to conception and design of the study, and to analysis and interpretation of data. D Shao and WK Rumbeiha drafted the manuscript. All authors critically revised the manuscript; gave final approval; and agreed to be accountable for all aspects of the work in ensuring that questions relating to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

a.

HPLC 2695, Waters Corp., Milford, MA.

b.

Brownlee C18 column (5 μm, 100 mm × 4.6 mm i.d.), Perkin-Elmer Inc., Waltham, MA.

c.

Pursuit XRs 3 C18 guard column (2.0 mm i.d.) Agilent Technologies Inc., Santa Clara, CA.

d.

18.2 MΩ·cm water, high purity water system; Aries Filter Network, West Berlin, NJ.

e.

Acetonitrile (HPLC grade), methanol (HPLC grade), dichloromethane (pesticide grade), glacial acetic acid (A.C.S reagent grade), trifluoroacetic acid (bioanalysis grade), citric acid monohydrate (certified A.C.S grade), sodium sulfate anhydrous (certified A.C.S grade), alumina neutral (60–325 mesh); Thermo Fisher Scientific, Waltham, MA.

f.

Waring single speed blender, Conair Corp., Stamford, CT.

g.

Celite 503 poly pail (1 kg, diatomaceous earth), purchased from Midland Scientific Inc. Omaha, NE, manufactured by J.T. Baker, Center Valley, PA.

h.

Octadecyl (C18) 40 μm preparation LC packing, purchased from VWR International, Radnor, PA, manufactured by J.T. Baker, Center Valley, PA.

i.

Aflatoxin standards AFM₁, AFB₁, AFB₂, AFG₁, and AFG₂, Whatman 740-E filter discs; Sigma-Aldrich, St. Louis, MO.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for this project provided by the Veterinary Laboratory Information and Response Network (Vet-LIRN), Center for Veterinary Medicine, U.S. Food and Drug Administration (grant 1U18FD005006-01). Their support is deeply appreciated.

References

AOAC Official Methods 982.24. Aflatoxin B₁ and M₁ in liver. Thin-layer chromatographic method. Gaithersburg, MD: AOAC International, 2005.

AOAC Official Methods 994.08. Aflatoxin in corn, almonds, brazil nuts, peanuts, and pistachio nuts. Multifunctional column (Mycosep) method. Gaithersburg, MD: AOAC International, 2005.

AOAC Official Methods of Analysis. Appendix F: Guidelines for standard method performance requirements. Gaithersburg, MD: AOAC International, 2012.

AOAC Official Methods of Analysis. Appendix K: Guidelines for dietary supplements and botanicals. Gaithersburg, MD: AOAC International, 2012.

Bastianello

. Pathological findings in a natural outbreak of aflatoxicosis in dogs. Onderstepoort J Vet Res 1987;54:635–640.

Biancardi

. A simple and reliable liquid chromatography-tandem mass spectrometry method for the determination of aflatoxin M₁ in milk. Food Addit Contam Part A Chem Anal Control Expo Risk Assess 2013;30:381–388.

Bräse

. Chemistry and biology of mycotoxins and related fungal metabolites. Chem Rev 2009;109:3903–3990.

Burns

. Use of the term “recovery” and “apparent recovery” in analytical procedures (IUPAC Recommendations 2002). Pure Appl Chem 2002;74:2201–2205.

Butler

Neal

. The effect of aflatoxin B₁ on the hepatic structure and RNA synthesis in rats fed a diet marginally deficient in choline. Cancer Res 1973;33:2878–2885.

10.

Campone

. Rapid analysis of aflatoxin M₁ in milk using dispersive liquid–liquid microextraction coupled with ultrahigh pressure liquid chromatography tandem mass spectrometry. Anal Bioanal Chem 2013;405:8645–8652.

11.

Cilievici

. The toxic and teratogenic effect of aflatoxin B₁ on the chick embryo development. Morphol Embryol (Bucur) 1980;26:309–314.

12.

Curticapean

. HPLC method validation for simultaneous determination of three mycotoxins from corn seeds. Croat Chem Acta 2011;84:413–417.

13.

Dhanasekaran

. Aflatoxins and aflatoxicosis in human and animals. In: Guevara-Gonzalez

, ed. Aflatoxins—Biochemistry and Molecular Biology. 1st ed. Rijeka, Croatia: InTech, 2011:221–254.

14.

Diaz

Murcia

. Biotransformation of aflatoxin B₁ and its relationship with the differential toxicological response to aflatoxin in commercial poultry species. In: Guevara-Gonzalez

, ed. Aflatoxins—Biochemistry and Molecular Biology. 1st ed. Rijeka, Croatia: InTech, 2011:1–20.

15.

Eaton

Gallagher

EP.

Mechanisms of aflatoxin carcinogenesis. Annu Rev Pharmacol Toxicol 1994;34:135–172.

16.

Gimeno

. Thin layer chromatographic determination of aflatoxins, ochratoxins, sterigmatocystin, zearalenone, citrinin, T-2 toxin, diacetoxyscirpenol, penicillic acid, patulin, and penitrem A. J Assoc Off Anal Chem 1979;62:579–585.

17.

Greene

. Disseminated intravascular coagulation complicating aflatoxicosis in dogs. Cornell Vet 1977;67:29–49.

18.

Hall

. Identifying and overcoming matrix effects in drug discovery and development. In: Prasain

, ed. Tandem Mass Spectrometry—Applications and Principles. 1st ed. Rijeka, Croatia: InTech, 2012:389–420.

19.

Holcomb

. Determination of aflatoxins in food products by chromatography. J Chromatogr 1992;624:341–352.

20.

Horwitz

Alber

The Horwitz ratio (HorRat): a useful index of method performance with respect to precision. J AOAC Int 2006;89:1095–1109.

21.

Howell

Taylor

PW.

Determination of aflatoxins, ochratoxin A, and zearalenone in mixed feeds, with detection by thin layer chromatography or high performance liquid chromatography. J Assoc Off Anal Chem 1981;64:1356–1363.

22.

IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. Some traditional herbal medicines, some mycotoxins, naphthalene and styrene. IARC Monogr Eval Carcinog Risks Hum 2002;82:1–556.

23.

Jaimez

. Application of the assay of aflatoxins by liquid chromatography with fluorescence detection in food analysis. J Chromatogr A 2000;882:1–10.

24.

Leong

. Exposure measurement of aflatoxins and aflatoxin metabolites in human body fluids. A short review. Mycotoxin Res 2012;28:79–87.

25.

Linsinger

TPJ

Josephs

RD.

Limitations of the application of the Horwitz equation. Trends Analyt Chem 2006;25:1125–1130.

26.

Liu

. Photodegradation kinetics and byproducts identification of the Aflatoxin B₁ in aqueous medium by ultra-performance liquid chromatography–quadrupole time-of-flight mass spectrometry. J Mass Spectrom 2010;45:553–559.

27.

Martins

. Aflatoxin M₁ determination in cheese by immunoaffinity column clean-up coupled to high-performance liquid chromatography. Revista Portuguesa Ciências Veterinárias 2007;102:321–325.

28.

Matuszewski

. Strategies for the assessment of matrix effect in quantitative bioanalytical methods based on HPLC-MS/MS. Anal Chem 2003;75:3019–3030.

29.

Muscarella

. Validation according to European Commission Decision 2002/657/EC of a confirmatory method for aflatoxin M₁ in milk based on immunoaffinity columns and high performance liquid chromatography with fluorescence detection. Anal Chim Acta 2007;594:257–264.

30.

Newberne

PM.

Chronic aflatoxicosis. J Am Vet Med Assoc 1973;163:1262–1267.

31.

Ortatatli

. Evaluation of pathological changes in broilers during chronic aflatoxin (50 and 100 ppb) and clinoptilolite exposure. Res Vet Sci 2005;78:61–68.

32.

Panchal

. Development and validation of GC/MS method for determination of pramipexole in rat plasma. Biomed Chromatogr 2011;25:524–530.

33.

Raisuddin

. Immunosuppressive effects of aflatoxin in growing rats. Mycopathologia 1993;124:189–194.

34.

Rumbeiha

. Clinical implications of mycotoxicosis in companion animals. In: Lyons

Jacques

, eds. Science and Technology in the Feed Industry. 1st ed. Nottingham, UK: Nottingham University Press, 2001:225–233.

35.

Rumbeiha

Morrison

. A review of class I and class II pet food recalls involving chemical contaminants from 1996 to 2008. J Med Toxicol 2011;7:60–66.

36.

Scortichini

. ELISA qualitative screening of chloramphenicol in muscle, eggs, honey and milk: method validation according to the Commission Decision 2002/657/EC criteria. Anal Chim Acta 2005;535:43–48.

37.

Scott

PM.

Methods for determination of aflatoxin M₁ in milk and milk products—a review of performance characteristics. Food Addit Contam 1989;6:283–305.

38.

Sharman

Gilbert

Automated aflatoxin analysis of foods and animal feeds using immunoaffinity column clean-up and high-performance liquid chromatographic determination. J Chromatogr 1991;543:220–225.

39.

Sheijooni-Fumani

. Determination of aflatoxin B1 in cereals by homogeneous liquid-liquid extraction coupled to high performance liquid chromatography-fluorescence detection. J Sep Sci 2011;34:1333–1337.

40.

Shrivastava

Gupta

VB.

Methods for the determination of limit of detection and limit of quantitation of the analytical methods. Chron Young Sci 2011;2:21–25.

41.

Song

. Simultaneous determination of bisphenol A, aflatoxin B₁, ochratoxin A, and patulin in food matrices by liquid chromatography/mass spectrometry. Rapid Commun Mass Spectrom 2013;27:671–680.

42.

Stahr

. Mycotoxin analysis. In: Analytical Methods in Toxicology. 1st ed. New York. NY: Wiley-Interscience, 1991:104–106.

43.

Stroka

Anklam

New strategies for the screening and determination of aflatoxins and the detection of aflatoxin-producing moulds in food and feed. Trends Analyt Chem 2002;21:90–95.

44.

Stubblefield

. Fate of aflatoxins in tissues, fluids, and excrements from cows dosed orally with aflatoxin B₁. Am J Vet Res 1983;44:1750–1752.

45.

Trottier

. Kinds of mutations induced by aflatoxin B₁ in a shuttle vector replicating in human cells transiently expressing cytochrome P4501A2 cDNA. Mol Carcinog 1992;6:140–147.

46.

Wang

. Simultaneous determination of chloramphenicol and aflatoxin M₁ residues in milk by triple quadrupole liquid chromatography-tandem mass spectrometry. J Agric Food Chem 2011;59:3532–3538.

47.

Williams

. Human aflatoxicosis in developing countries: a review of toxicology, exposure, potential health consequences, and interventions. Am J Clin Nutr 2004;80:1106–1122.

48.

Youden

Steiner

. Statistical techniques for collaborative tests. Planning and analysis of results of collaborative tests. In: Statistical Manual of the Association of Official Analytical Chemists. 1st ed. Gaithersburg, MD: AOAC International, 1984:33–36.

49.

Zhu

. Correlation of dietary aflatoxin B₁ levels with excretion of aflatoxin M₁ in human urine. Cancer Res 1987;47:1848–1852.