Production of monoclonal antibody and development of an icELISA for the detection of paxilline in grain

ABSTRACT

Paxilline (PAX) is a tremorgenic mycotoxin that is thought to be associated with various syndromes in animals. To facilitate the determination of this toxin, a hapten was chemically synthesized and then respectively conjugated to Bovine Thyroglobulin (BTG) and Bovine serum albumin (BSA) for antibody production. Monoclonal antibody (mAb) 4F8 was finally obtained and used to establish an accurate and sensitive indirect competitive ELISA (icELISA). Various parameters that could impact the performance of icELISA were optimized, including the ion concentration, pH value and organic solvent. Under optimal condition, this assay exhibits an IC₅₀ value as low as 0.84 µg L⁻¹. The limit of detection (LOD) of the established icELISA method for PAX was 0.03–0.3 µg L⁻¹_. In conclusion, this study provided a rapid, sensitive, and high-throughput immunoassay for the detection of PAX in grains.

KEYWORDS:

• Paxilline
• monoclonal antibody
• icELISA
• immunoassay

1. Introduction

Paxilline (Figure 1), a tremor fungal alkaloid, could inhibit large conductance Ca²⁺- and voltage-activated K⁺(BK) – type channels, which causes “ryegrass stagger”, a neurological disorder that impairs motor function in livestock. It is produced by the endophytic fungus Acremonium lolii, which is also a precursor to lolitrem B, often infects perennial ryegrass, which is then eaten by grazing livestock, causing neurotoxic symptoms, such as incoordination in their movement (Ball et al., 1995; Krauss et al., 2020). Reports have shown that endophytic fungi had spread infection widely in Europe, Australia, New Zealand, the United States, and many other parts of the world (Ball et al., 1995; Miles et al., 1998; Zhou & Lingle, 2014). The inhibitory effect of paxilline on BK channels involves the strong stabilization of the closed conformation, and the affinity of paxilline for closed channels is 1000 times greater than that for open channels (DeFarias et al., 1996). Paxilline-a fungal metabolites, blockages the high conductivity of Calcium-Activated K⁺ Channels, which would increase the spontaneous contractility of the bladder and duodenum in rats and guinea pigs (Rowan, 1993). Paxilline has been shown to induce intermittent tremor in mice at 35 mg kg⁻¹ level. Fungal alkaloids have long been studied for causing grain diseases, and feeding endophytic fungi-contaminated cereals can result in ischaemic acreolysis, reduced feed intake, slowed growth, and reproductive disorders (Menna et al., 2012; Pichova et al., 2018). Therefore, to control the influence of paxilline exposure to livestock, contamination must be determined by accurate, sensitive, and high-throughput assays.

Figure 1. Structures of paxilline and PAX-CMO hapten.

Although the rapid detection technology for mycotoxins has been one of the hotspots in the world, few studies on paxilline have been reported (Menna et al., 2012). Garthwaite et al. (1994) established an ELISA method with LOD of paxilline in mycelial extracts and dried mycelium was 5 ng mL⁻¹ and 10 ng g⁻¹, respectively. They also combined immunoassays with thin-layer chromatography to detect paxilline, and using this technique, detection levels were much more sensitive than those obtained using thin-layer chromatography alone. In the study by Andersen and Frisvad (2004), commercially purchased or home-grown mouldy tomatoes were tested for residues of paxilline and its metabolite 3-O-acetoxypaxilline using high-performance liquid chromatography (HPLC). In 2015 (Maragos, 2015), a murine monoclonal antibody was prepared and an icELISA method was established to determine Paxilline in maize silage with an IC₅₀ of 1.2–2.5 ng mL⁻¹. It is worth noting that they used a special clean-up column to purify the extract after sample pretreatment. The results show that this sample pretreatment method can indeed eliminate the matrix effect of maize silage samples, and is well applied in paxilline analysis afterward. Bauer et al. (2017) used PAX-CMO (Figure 1), a paxilline derivative, as hapten to conjugate with KLH to prepare complete antigen, and finally obtained monoclonal antibodies that could specifically detect paxilline in the samples of rye and barley. The EIA method was established to ensure food and feed safety. The hapten synthesis method in this paper is based on this method, only with slight changes. The same authors developed an EIA using ryegrass in the following year using the same monoclonal antibody, and then HPLC-MS/MS was used as confirmative assay (Bauer et al., 2018). The main advantages of this method are ease to use and time saving, thereby facilitating large-scale screening of 30 endophytic toxins in pasture.

Paxilline is a biosynthetic precursor of lolitrem B, causing lighter tremor effect in mice than lolitrem B (Menna et al., 2012); however, the detection of paxilline is helpful for assessing the level of lolitrem B contamination. Analysis of mycotoxins in grain samples by instrumental methods such as liquid chromatography (LC), gas chromatography-mass spectrometry (GC-MS) and high-performance liquid chromatography (HPLC), is often high professional, time-consuming and expensive (Garthwaite et al., 1994). Despite extensive extraction and purification prior to analysis being widely used to analyse grains, it still constantly suffer from great matrix effects. In fact, ELISA analysis was proposed as the first paxilline screening method to find non-tremor-producing endophytes in samples screening (Zhou & Lingle C, 2014). In the present study, a sensitive monoclonal antibody was prepared and an immunoassay was developed to realize the analysis of Paxilline in PBS for rapid quantification. However, due to the complex matrix effects in grain samples, the matrix effects in actual samples could not be eliminated by the 10 attempted pretreatment methods, so further exploration is needed in this aspect.

2. Materials and methods

2.1. Chemicals and reagents

Paxilline was acquired from Thermo Fisher Scientific, Inc. (Waltham, MA, USA). Bovine serum albumin (BSA), polyethylene glycol (PEG) 1500, 1-(3-dimethylaminopropyl)-3-ethyl carbodiimide (EDC), N-hydroxysuccinimide (NHS), hypoxanthine aminopterin thymidine (HAT), incomplete Freund’s adjuvant (IFA) and complete Freund’s adjuvant (CFA) were purchased from Sigma-Aldrich. Horse radish peroxidase (HRP)-labelled goat anti-mouse antibody was obtained from Jackson Immuno Reaserch Laboratories. Cell culture medium (DMEM), fetal calf serum and red blood cell lysis buffer were bought from Gibco BRL (Carlsbad, CA, USA). Casein sodium salt, Tween^® 20, 3,3,5,5-Tetramethylbenzidine (TMB) substrate, N, N-dicyclohexylcarbodiimide (DCC) were obtained from Aladdin Reagent (Shanghai, China). Dimethylformamide (DMF) was purchased from Sinopharm Chemical Reagent (Beijing, China). Cell culture plates and 96-well microtitre plates were supplied by Corning Life Sciences. Female BALB/c mice were obtained from Vital River Laboratory Animal Technology (Beijing China). Paxilline-free maize and wheat samples were obtained from the National Reference Laboratory for Veterinary Drug Residues (Beijing, China).

High-performance liquid chromatography–tandem mass spectrometry (HPLC MS/MS) was used to confirm the chemical structure of the hapten. Matrix-assisted laser desorption/ionization time of flight mass spectrometry (MALDI–TOF-MS) (Bruker, Daltonics, Billerica, MA, USA) was used to evaluate the protein conjugates.

2.2. Hapten synthesis and conjugate preparation

The hapten, named PAX-CMO, was synthesized in the following steps. Eight milligram of PAX was dissolved in 1 mL pyridine. Then 12 mg CMO hydrochloride was added. After that, the solution above was then mixed and stirred with 400 µL of sodium acetate solution (20 mg mL⁻¹) and incubated at 70°C for 12 h. The reaction product was evaporated by rotary evaporators, and then 1 mL of pyridine and 1 mL of hydrochloric acid (0.1 mol L⁻¹) were used to redissolve the residue. An equal volume of ethyl acetate was used for liquid extraction twice, separation was then carried out by centrifugation at 4000×g for 10 min. The two batches of ethyl acetate solutions were mixed and evaporated in a rotary evaporator. Then, 1 mL of pyridine was used to redissolve the residue containing the derivatives (PAX-CMO). The product was characterized by mass spectrum, MS: m/z calculated for C₂₉H₃₆N₂O₆, molecular weight, 508.26, found [M + H]-509.3 (ESM Figure S1). The mass spectrum results showed that hapten was synthesized successfully.

The immunogens and coating antigens were respectively prepared by conjugating PAX-CMO to BTG and BSA by an active-ester method. Add PAX-CMO (0.05 mmol), EDC (0.2 mmol) and NHS (0.1 mmol) sequentially to 500 μL DMF and reacted for 5 h at 25°C. Add PAX-CMO to BTG and BSA solutions to prepare conjugates hapten-EDC-BTG and hapten-EDC-BSA. To remove the unbound PAX-CMO, dialysis was conducted in 0.01 mol L⁻¹ PBS for 3 days and then stored at −20°C.

2.3. Preparation of monoclonal antibody

All animal treatments were conducted in strict accordance with Chinese laws and guidelines approved by the Animal Ethics Committee of China Agricultural University. Eight female BALB/c mice were immunized subcutaneously with 100 µg of PAX-BTG. The first dose was given through injection of 100 μg of immunogen in 0.1 mL of Freund’s complete adjuvant. Mice were injected with 100 μg of immunogen in Freund’s incomplete adjuvant at weeks 3 and 6 after the first injection. The sensitivity and specificity of mouse orbital antiserum samples (approximately 50 μL) were detected by icELISA method one week after the third immunization. The optical density (OD) value of the icELISA method was determined by PerkinElmer Envision plate reader (Waltham, MA, USA). For subsequent cell fusion, mice with a lower 50% inhibition concentration (IC50) of PAX were chosen. The procedure for cell fusion was conducted as described earlier (Zhang et al., 2017, Wang et al., 2015). The selected mice were intraperitoneally injected with three times the immunization dose (300 μg antigen, 500 μL PBS buffer). Three days later, SP2/0 myeloma cells and spleen cells from the selected mice were mixed and fused with the help of PEG 1500. Cells that were successfully fused were grown in HAT medium for a period of 7 days, and the cell supernatant was screened using the icELISA technique. Positive hybridomas with high sensitivity were then subcloned using the limiting dilution method. After five subcloning cycles, the optimal single-cell quality was collected and grown. In the end, a portion of the hybridomas were frozen for preservation, while another portion was collected by intraperitoneal injection to produce ascites. Mouse ascites were extracted and purified using saturated ammonium sulphate. Then, the collected ascites were purified using saturated ammonium sulphate (Zhang et al., 2016). The isotype of the purified mAb was identified using a mouse mAb isotyping kit.

2.4. Development of icELISA

The execution of the icELISA protocol is similar to what was previously described (Ceballos-Alcantarilla et al., 2019; Mercader et al., 2017). Briefly, Add the coating antigen (hapten-BSA, 100 μL/well) diluted with carbonate buffer (0.05 mol L⁻¹, pH 9.6) to the microplate, then incubate the microplate in a 37°C for 2 h. Next is the washing step, blocking with 150 μL/well blocking buffer and incubating at 37°C for 1 h in an incubator. The blocking buffer was discarded, and the microplate is added with 50 μL of standard solution (diluted PAX) and 50 μL of antibody solution (diluted antiserum or mAb) in sequence. After incubation at 37°C for 30 min, the microplate is washed three times and then added with HRP-labeled goat anti-mouse IgG secondary antibody (1:5000, 100 μL/well). The microplate is then washed four times before adding TMB substrate (100 μL/well) and incubating at room temperature for 15 min. Afterwards, the stop solution (2 mol L⁻¹ H₂SO₄, 50 μL/well) is added. The OD value is then measured at 450 nm using an enzyme-linked immunosorbent assay. The OD value is plotted on a logarithmic scale against the concentration of the analyte, and the resulting S-shaped curve is mathematically fitted to the following four-parameter logistic equation using OriginPro 7.5 software (OriginLab Corporation, Northampton, MA, USA): (1) $\mathrm{Y}={\displaystyle \frac{(A-D)}{[1+{(X/C)}^B+D]}}$(1) where A = response at high asymptote, B = the slope factor, C = concentration corresponding to 50% specific binding (IC₅₀), D = response at low asymptote, and X = the calibration concentration.

2.5. Optimization of ELISA

In order to improve the performance of the ELISA, the concentration and pairing combinations of coating antigen and mAb, as well as physicochemical parameters such as organic solvent content, ionic strength, and pH value were optimized. The impact of these parameters was assessed using the OD_max/IC₅₀ ratio, where OD_max represents the maximum absorbance and the half-maximal inhibitory concentration was fitted using Equation (1).

2.6. Sample pretreatment and spiking/recovery assay

Corn, wheat and barley are three kinds of important grain both for human and animal husbandry. Furthermore, they are also often contaminated by paxilline (Bauer et al., 2017). The paxilline-negative ground corn, wheat, barley samples were provided by the National Reference Laboratory for Veterinary Drug Residues (Beijing, China). The procedure for preparing the samples was described in detail as follows. As described previously, paxilline was typically extracted by using ethyl acetate (Bauer et al., 2017), Acetonitrile/PBS (Bauer et al., 2017; Liesener et al., 2010), halogenated solvents and methyl alcohol (Bauer et al., 2017; Garthwaite et al., 1994; Maragos, 2015). Therefore, 10 different sample extract methods were chosen to eliminate matrix effects (Table 1).

Table 1. The sample pretreatment method used in this research.

Number	Extraction solvent	Cleanup column ^a	Nitrogen Concentration ^b	Redissolved solvent	Dilution factor
①	Ethyl acetate	–	+	Methyl alcohol	400
②	Dichloromethane: methyl alcohol = 2:1	–	+	Methyl alcohol	400
③	Acetonitrile	–	+	Methyl alcohol	400
④	Methyl alcohol	–	–	–^c	40
⑤	Ethyl acetate	–	+	PBS	10
⑥	Dichloromethane: methyl alcohol = 2:1	–	+	PBS	10
⑦	Dichloromethane: methyl alcohol = 2:1	–	+	PBS	100
⑧	Acetonitrile	–	+	PBS	500
⑨	PBS	–	–	–^c	500
⑩	Acetonitrile:water = 2:1	+	+	PBS	40

Notes: a. Whether extraction column was used to purify the sample extract in the method, “+” means that it is used, “−” means that it is not used; b. Whether the sample extract was evaporated under nitrogen flow in the method, “+” means that it was used, “−” means that it was not used; c. This method did not require redissolution, and the supernatant was directly taken after centrifugation for detection.

For the sample analysis, the ground sample was weighed and extracted using organic solvent for 5 min with vortex. Then the samples were centrifuged at 4000×g for 20 min. In the 10 methods, the supernatant was purified by extraction column. And the supernatant was dried using nitrogen gas in the remaining method. It was detected directly in Methods 4 and 9. The residues were redissolved with phosphate buffer or methyl alcohol, and then the liquid was used for further analytical experiments.

In order to test the accuracy and precision of the developed icELISA detection, recovery studies were validated by a total three spiking blank maize and wheat and barley samples with PAX in method 7 and method 10. Negative samples were spiked with three levels of PAX 2, 5, and 10 ng mL⁻¹. Each spiked sample was analysed in triplicate, and a blank control group was analysed simultaneously. The recovery rate of PAX in grain samples was calculated based on the standard curve in PBS and the following formula: (2) $\mathrm{Re}\mathrm{cov}\mathrm{ery}(\text{\%})={\displaystyle \frac{\mathrm{measuredvalue}(\mathrm{ng}/\mathrm{ml})}{\mathrm{spikedvalue}(\mathrm{ng}/\mathrm{ml})}\times 100\text{\%}}$(2)

3. Results and discussion

3.1. Synthesis of conjugates

The same as other small molecule compounds, paxilline also needed to be conjugated to a carrier protein to elicit an immune response. Before it can be coupled to a macromolecule carrier protein, however, the target must first be modified to produce a group that can be coupled. The most important modification principle is to preserve its characteristic structure as much as possible. In previous reports, several strategies for modifying and conjugating paxilline were proposed. Bauer et al. (2017) and Garthwaite et al. (1994), derivatized PAX to carboxymethyl oxime (CMO), and then coupled with KLH as immunogen. The haptens derivatization method in this paper is similar to that of the former with a slight change, and we replaced the carrier protein KLH with BTG. The hapten was identified by mass spectrometry, the ESM Figure (Figure S1) indicated that PAX-CMO hapten was successfully synthesized. The successful synthesis of the coating antigen (PAX-EDC-BSA) was confirmed using MALDI-TOF/MS analysis (Singh et al., 2004), and the coupling ratio was 11.4 (Figure 2).

Figure 2. (a) MALDI-TOF/MS spectrum of BSA; (b) MALDI-TOF/MS spectrum of PAX-EDC-BSA. The number of modifying groups (11.4) was determined by dividing the mass increase upon modification by the mass added by each modifying group (508.26 Da).

3.2. Preparation of monoclonal antibodies

The PAX-CMO–BTG conjugate was used to immunize eight mice. The effectiveness of immunogen on immune response was evaluated by antibody titre and IC₅₀ value. The parameters of the antisera obtained from the selected mouse at the last bleeding are summarized in Table 2. At the stage of cell cloning, the cell clone 4F8, showed the best inhibition rate, the monoclonal antibody was prepared as ascites without further purification and then characterized (Table 2). As is shown, the optimum antibody dilution for 4F8 remained the same as antisera, and the OD_max was 1.86, while the IC₅₀ of the ascites decreased by 4.8 times. The typical standard curve for paxilline, established with 4F8, was shown in Figure 3. The IC₅₀ was 0.84 µg L⁻¹_, which was lower than that (1.2–2.5 µg L⁻¹) of previously reported paxilline antibodies (Maragos, 2015, Bauer et al., 2017). The LODs achieved were 0.03–0.3 µg L⁻¹ for 4F8. After antibody typing, 4F8 monoclonal antibody was identified as IgG1 (κ) subtype, as shown in Table S1.

Figure 3. Standard curve for paxilline with 4F8.

Table 2. icELISA assay parameters of serum and mAb 4F8.

Antibodies	PAX-CMO-BSA
	Antisera	4F8
Antibody dilution	1/9000	1/9000
Antigen dilution	1/27,000	1/12,000
ODmax	1.67	1.86
IC50 (µg L^−1)	4.04	0.84

3.3. Optimization and development of icELISA for paxilline

Solution conditions may affect the structure or properties of the antibody and thus affect the sensitivity of icELISA (Xu et al., 2016). To achieve a high sensitivity, a series of assay conditions were optimized, such as pH values of the PBS buffer, ionic strength and concentration of acetonitrile and methanol. There are three main parameters used to evaluate whether the assay has a high sensitivity, OD_max, IC₅₀ and OD_max/IC₅₀. Among these parameters, OD_max/IC₅₀ value, as the primary criteria, were used to evaluate the icELISA performance and the higher OD_max/IC₅₀ value indicated the higher performance.

As shown in Figure 4(a), OD_max/IC₅₀ ratio vary not so much over the pH5∼pH9 range, and reached to peak at pH 7.0, decreased sharply when the pH values become 10. However, OD_max values decreased sharply when the ionic strength ranged from 0.111 to 6.0 mol L⁻¹, when the ionic strength reaches 1 M, OD value is less than 1.5 (Figure 4(b)). Therefore, in order to ensure the accuracy of the test, the maximum OD_max/IC₅₀ ratio was selected at 0.111 mM, when the ionic strength was less than 1 M. Since the log P of paxilline is 3.78, it is very fat-soluble. Therefore, the addition of organic reagents such as methanol or acetonitrile during sample pre-treatment may facilitate the extraction of the analyte. So, it is necessary to evaluate the resistance of the antibody to organic solvents. However, the antibody 4F8 in this study was extremely sensitive to methanol and acetonitrile (Figure 4(c,d)). In Figure 3(c), OD_max/IC₅₀ values were dramatically reduced with the tiny change of methanol concentration from 0 to 2%, and then remained flat within the concentration range of 2–20%. When the concentration is greater than 20%, IC₅₀ increases sharply, reducing the performance of icELISA. Similarly, the analysis trend of acetonitrile for paxilline was the same. With the increase of acetonitrile concentration (from 0 to 20%), OD_max/IC₅₀ kept a sharp downward trend. Therefore, the less volume of organic solvent used in the sample pretreatment process, the sensitivity will be better. To sum-up, based on above data, the optimum conditions are pH 7.0 and 0.1 mol L⁻¹ NaCl in PBS buffer. In view of the fact that OD_max/IC₅₀ can reach its maximum in normal PBS (PH = 7.2, ionic strength  = 0.145 M, control group, data not shown), PBS prepared normally in the laboratory was used afterward.

Figure 4. Effects of a pH value of PBS buffer, b ionic strength, c concentration of acetonitrile, and d concentration of methanol on the performance of icELISA (N = 3).

3.4. Matrix effects and recovery

We used 10 different pretreatment methods to access the grain matrix effect on the antibody–antigen reaction as described in Table 1. The influence of matrix effects on OD_max values and IC₅₀ were mainly evaluated (as shown in Figure 4). In this study, we believe that the effect on OD_max should be below ±0.5, and the effect on IC₅₀ should be under ±50% to be considered as the elimination of matrix effect. Unfortunately, only method 7 was able to eliminate the effects of three matrix effects simultaneously, and method 10 also was able to remove maize matrix effect, but neither of them had a good performance in subsequent recovery experiments.

For maize sample (Figure 4(a)), according to the results, as long as methanol exists in the system at last (Methods 1–4), it will have a great impact on OD_max and IC₅₀. This is consistent with the results of the previous organic solvent tolerance experiment, monoclonal antibody 4F8 is extremely sensitive to methanol. After we changed the redissolution solvent (Methods 5–8) to PBS, we found that only method 7 could reduce the impact on the experimental performance. Unfortunately, even with this method, the dilution factor finally reached 100. In Method 9, we changed both the extraction reagent and the redissolution reagent to PBS, but the matrix effect could not be eliminated. Finally, after extracting the matrix, we used clean-up columns to adsorb the impurities in the matrix as much as possible, and finally eliminated the matrix effect.

For wheat sample (Figure 4(b)), we assessed whether pretreatment Methods 4–9 can minimize the effect of matrix on antigen–antibody reaction. The results showed that only methods 6 and 7 had the lowest effects on OD_max and IC₅₀. Same regular works the same for the barley sample (Figure 4(c)), nevertheless, method 10, which previously worked well for corn samples, did not work well for barley samples.

Due to the sixth method is the same as the seventh method, but the seventh method increases the dilution factor by increasing the actual volume of re-dissolution. We finally evaluated the recovery rate of the seventh method on three-grain samples (Figure 5).

Figure 5. Comparison of different pretreatment methods for maize (a), wheat (b) and barley (c) samples.

Each kind of sample was spiked with three concentrations of paxilline (2, 5, 10 ng mL⁻¹), and recoveries were estimated by comparing the detection values with the spiked values. As shown in Table 3, the recoveries for Paxilline in water samples ranged from 21.9 to 69.5%, with coefficient of variation (CV) less than 9.9%. Bauer et al. (2017) used ethyl acetate as the extraction reagent, after extraction, centrifuge, rotary evaporation, etc., and finally using methanol /PBS redissolved residue, the matrix effect was basically eliminated when the dilution factor was 200, and the paxilline residue in barley samples was successfully detected. The same lab (Bauer et al., 2017) tested beer samples the following year with ultrasonic degassing, pH adjustment, cellulose filter, and finally methanol /PBS heavy dissolved residue. Different from the results of these two studies, the intolerance of the antibodies 4F88 prepared in this study to methanol resulted in the inability to use methanol for paxilline detection in this experiment.

Table 3. Mean recoveries and coefficients of variation of Paxilline in maize, wheat and barley samples (n = 3) MR mean recovery, CV coefficient of variation.

Spiked levels (ng mL^−1)		Maize		Wheat		Barley
		MR (%)	CV (%)	MR (%)	CV (%)	MR (%)	CV (%)
Paxilline	2	69.5%	5.13%	60.5%	4.80%	60.0%	5.67%
	5	43.2%	7.55%	45.6%	1.81%	49%	6.68%
	10	21.9%	9.21%	30%	4.26%	29.1%	9.90%

3. Conclusions

The present study synthesized a paxilline immunogen and coating antigen, acquired a mAb (4F8) with an IC₅₀ of 0.84 ng mL⁻¹ for Paxilline. A specific and sensitive icELISA method was established to determine paxilline in grain. The limit of detection (LOD) of the established icELISA method for PAX was 0.03–0.3 µg L⁻¹. In conclusion, this study provided a rapid, sensitive, and high-throughput immunoassay for the detection of PAX in grains.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the National Key Research and Development Program of China [grant number 2022YFF1101001].