A Chemical Synthesis of Benzo [ a ] pyrene Hapten and Its Application in a Streptavidin-Horseradish Peroxidase-Based Enzyme-Linked Immunological Analysis

A specific BaP haptenwas synthesized andBaP antigens (immunogen, coating antigen)were prepared using diversemethods. Based on these works, a streptavidin-horseradish peroxidase-based enzyme-linked immunological analysis was developed and firstly used to detect BaP. Several physiochemical factors that may influence the assay performance were optimized in the assay. Under optimal conditions, a detection limit of 0.0094 ngmL was obtained and good linearity was achieved within a range of 0.03–35.40 ngmL. Satisfactory recovery of spiked samples (91.12–109.23%) was obtained and the coefficient of variation was acceptable. Finally, the detection results of BaP in environmental and food samples were consistent with those obtained using high-performance liquid chromatography. The proposed immunoassay is reliable and has great potential for detecting trace amounts of BaP.


Introduction
Polycyclic aromatic hydrocarbons (PAHs) are persistent organic pollutants that contain 2-8 aromatic rings in linear, angular, or cluster arrangements [1].Due to their toxicity and carcinogenicity, PAHs have attracted the attention of environmental chemists and regulatory agencies for several decades [2,3].Among the PAHs, benzo[a]pyrene (BaP) has been widely studied.In recent years, much attention has been devoted to the existence of BaP in the environment.BaP is considered the most dangerous of PAHs because of its chemical and physical characteristics and ease of accumulation in the food chain [4,5].BaP is usually found in exhaust gas produced by the incomplete combustion of fuels, in all smoke resulting from the combustion of organic material and in charbroiled food [6][7][8].BaP can enter into water bodies mainly via atmospheric fallout, municipal/industrial effluents, urban runoff, and oil spillage [9,10]; it is emitted to the atmosphere from different sources, including industrial processes, combustion of fossils and fuels, and irrational application of pesticide [11].
BaP is a great potential hazard to human health.It is a type of carcinogenic compound, which can cause lung, gastric, bladder, and gastrointestinal cancers [12,13].Recent studies of BaP have also demonstrated its involvement in the development of cardiovascular diseases [14].BaP is mutagenic and teratogenic; it can even pass through the placenta to endanger the embryogenesis [15].Dietary intake of BaP is a major source of human exposure [16].Related studies have shown that people in the general population without substantial environmental and occupational exposure had a higher exposure to BaP by food ingestion [17].Probable sources of BaP in food are environmental contamination from soil, water, and the thermal treatment used for the preparation of food, such as roasting.
Up to now, many different methods have been described for the determination of BaP residues.The main methods are gas chromatography and mass spectrometry (GC-MS) and high-performance liquid chromatography (HPLC); however, these chromatographic methods usually require long periods of detection processes that are time-consuming and the follow-up maintenance is expensive.Compared BaP coating antigens which are covered on the surface of the tubes competed with BaP analytes to combine with the bio-pAb-BaP.The horseradish peroxidase-conjugated streptavidin (SA-HRP) is fixed on the wall through the reaction with bio-pAb-BaP.As the increase of BaP concentration, less bio-pAb-BaP can be attached to the solid carrier.In this case, the amount of SA-HRP is less, and then quantitative analysis could be performed after sufficient colour development.
with chromatographic techniques, immunoassays are costeffective, with adequate sensitivity and high selectivity [18].Therefore, immunochemical techniques have been increasingly considered as complementary methods for residue analysis.Among these immunoassays, enzymelinked immunosorbent assay (ELISA) is well suited for detecting trace amounts of pollutants.Some improved immunoassays were researched based on the conventional ELISA, such as fluorescence-enzyme immunoassay (FL-ELISA) and chemiluminescence immunoassay (CL-ELISA) [19,20].Accompanying the improved immunoassays above, the streptavidin-enzyme based ELISA can reduce nonspecific reactions with reagents due to the higher specificity and affinity between streptavidin and biotin, and it has great potential for detecting trace amounts of BaP.However, as we know, no information on detecting BaP by streptavidinenzyme based ELISA has been reported.
In this research, a specific BaP hapten was synthesized.Diverse BaP antigens and biotinylated polyclonal BaP antibody (bio-pAb-BaP) were prepared.Based on these works, a streptavidin-horseradish peroxidase-based enzyme-linked immunoassay was developed (SA-ELISA) and firstly used for the sensitive detection of BaP.Some environmental and food samples were collected in Shanghai, China.Several physiochemical factors that may influence the assay performance were optimized, and the accuracy and reliability of the immunoassay were validated.Figure 1 shows the detection process of SA-ELISA method.

Synthesis of BaP
Hapten.BaP molecule does not contain functional groups that can connect with proteins.Therefore, a specific BaP hapten was synthesized (Figure 2).Then, the BaP antigens and bio-pAb-BaP were prepared.For the preparation of BaP hapten, antigens, and pAb-BaP, we followed the methods of Ma and Zhuang 2017 [21].
BaP (200 mg), n-iodosuccinimid (NIS, 210 mg), and acidic alumina (5 g) were dissolved in toluene (15 mL); the mixture was allowed to react with stirring for 4 days at 20 ∘ C. The residual acidic alumina was washed with dichloromethane (25 mL × 4).Afterwards, the organic phase was collected and extracted by saturated sodium sulphite, saturated salt water, and double-distilled water (ddH 2 O) successively.After being dried with MgSO 4 , the mixture was concentrated down to near dryness; then, the yellow powder (6-I-BaP, 225 mg, yield: 75.3%) was obtained by recrystallization with dichloromethane and petroleum ether.Next, 6-I-BaP (135 mg), palladium acetate (5 mg), and LiCl (10 mg) were dissolved in DMF (10 mL).Then, 4pentene-2-alcohol (35 L) and triethylamine (100 L) were added.The mixture was allowed to react for 15 h at 120 ∘ C.After a completed reaction, toluene (70 mL) was added.The reaction mixture was extracted by saturated salt water and ddH 2 O, respectively.After being dried with MgSO 4 , the volatile was removed under reduced pressure.The crude product was purified through chromatography column filled with silica gel (leaching solution: tert-butyl methyl ether/petroleum ether = 1/2 (v/v)), and then the yellow powder BaP hapten (BaP = O 102 mg, yield: 87.1%) was obtained.

Immunogen and Coating Antigen.
As a contact portion between carrier protein and BaP hapten, the linking arm can become an antigenic determinant to affect the specificity for conjugating with antibodies [22].In order to reduce specific binding caused by the linking arm, BSA-BaP (immunogen) and OVA-BaP (coating antigen) were prepared by the activated ester method and mixed anhydride method, respectively (Figure 3).
BaP hapten (8 mg) and carboxymethyl amine hydrochloride (16 mg) were dissolved in pyridine (5 mL); the mixture was allowed to react in absence of light.After a completed reaction, sodium borohydride (36 mg) dissolved in ddH 2 O (300 L) was added to the reaction mixture and allowed to react for 5 h at ambient temperature.After the organic phase was evaporated to dryness, the crude product was washed with HCl solution (pH 2.0) and ddH 2 O successively.After being dried, the yellow product BaP-COOH was obtained.
BaP-COOH (0.155 g) was dissolved in DMF (1 mL).N-Hydroxysuccinimide (NHS) and N, N  -dicyclohexylcarbodiimide (DCC) were dissolved in DMF (1 mL), respectively and added to the solution above successively.The mixture was allowed to react with stirring for 8 h at room temperature.After centrifugal separation, the obtained supernate was added dropwise into 10 mL of bovine serum albumin (BSA) solution (12 mg mL −1 , in PBS) and then stirred for 6 h at 4 ∘ C. The suspension was dialyzed against PBS for 3 days.After another centrifugal separation, the supernate (immunogen, BSA-BaP) was stored at -20 ∘ C.
BaP-COOH (0.155 g) was dissolved in DMF (1 mL), and then n-butylamine (50 L) and isobutyl chloroformate (62.5 L) were added sequentially.The mixture was stirred magnetically for 3 h at 4 ∘ C. Afterwards, the mixture was added dropwise into 10 mL of egg albumin (OVA) solution (12 mg mL −1 , in PBS) and allowed to react with stirring for 5 h at 4 ∘ C. The suspension was dialyzed against PBS for 3 days.After centrifugal separation, the supernate (coating antigen, OVA-BaP) was stored at -20 ∘ C.

Preparation of Bio-pAb-BaP.
In this research, polyclonal anti-BaP antibody (pAb-BaP) was prepared by taking healthy rabbits as immune objects.Two white rabbits were immunized with BSA-BaP through subcutaneous and intramuscular injections.The immunization lasted for 4 months.pAb-BaP was separated from rabbit serum through ammonium sulphate precipitation method, dialyzed against PBS for 5 days, and stored at -20 ∘ C.
The prepared pAb-BaP was diluted with CBS buffer at a concentration of 2.0 mg mL −1 .Then, the solution was mixed with 1.0 mg mL −1 BNHS (in DMSO) in the mass ratio of 1 : 10.The mixture was stirred for 4 h, and dialyzed against PBS for 3 days.The obtained bio-pAb-BaP was stored at -20 ∘ C.
The specificity of pAb-BaP was evaluated by testing crossreactivity (CR) using other types of PAHs.The CR values were calculated according to the formula CR (%) = (IC 50 of BaP)/(IC 50 of analogues) × 100%.The chemical structures of BaP analogues and CR results are shown in Table 1.In all cases, there was a low CR (below 7%) between BaP and other structurally similar compounds, indicating that the prepared pAb-BaP was suitable for detecting trace amount of BaP.
2.6.Preparation of BaP Samples.PM 2.5 (particles in the air, the aerodynamic diameter of which is less than or equal to 2.5 m) and agricultural soil samples were collected in Shanghai, China; the locations of sampling sites are shown in Figure 4. Vegetables and cereals were purchased from a local food market; roast lamb and duck were obtained from a barbecue shop.The treatment of environmental samples was similar to the description of the standard of People's Republic of China HJ 784-2016.First, PM 2.5 sampling membranes were broken down into pieces; agricultural soil samples were ground into powder after being freeze-dried.Then, each PM 2.5 sampling membrane (in pieces) and agricultural soil (10 g) were extracted for 16 h by soxhlet extraction with acetone/hexane (1 : 1, v/v, 50 mL) and centrifuged at 4500 rpm for 10 min.The organic layer was dried with Na 2 SO 4 and then concentrated to nearly 2 mL.
The food samples were treated as described in the standard of People's Republic of China GB/T 5009.27-2003.The vegetable samples (100 g) were grinded into pieces; the cereal samples (50 g) and barbecue food (50 g) were crushed.Next, all the samples were extracted by ultrasonic processing with hexane/acetone (1 : 1, v/v, 100 mL × 2) for 0.5 h.For the treatment of vegetables, the extract was washed with cyclohexane (100 mL × 2) and ddH 2 O (100 mL × 2) successively.After being dried with Na 2 SO 4 , the cyclohexane layer was condensed to about 2 mL.In the case of the barbecue food and cereals, the extract mixed with CH 3 OH-H 2 O (9 : 1, v/v, 150 mL) was heated at 80 ∘ C for 6 h.Then, the mixture was extracted by cyclohexane (100 mL × 2); the cyclohexane layer was washed by CH 3 OH-H 2 O (1 : 1, v/v, 100 mL × 3) and ddH 2 O (200 mL × 2) successively.Next, the collected organic layer was condensed to about 40 mL and then washed with H 2 SO 4 (60%, 100 mL) and ddH 2 O (25 mL × 4) successively.The organic phase was dried by Na 2 SO 4 and then condensed to about 2 mL.
All the extract was purified with a florisil column which was treated with cyclohexane/dichloromethane (1 : 1, v/v, 20 mL).Benzene (10 mL) was used as the leaching solution, and then the eluent was collected and condensed to near dryness.Afterwards, the residue was redissolved in acetonitrile (0.5 mL).The treated BaP samples above were detected by SA-ELISA after being diluted by 1000 times with PBS containing 5% DMSO as well as the HPLC method.

Establishment of the Immunoassay.
The SA-ELISA determinations were performed in a 96-well microtitre plate.First, microplates were covered with 100 L/tube of coating antigen at 4 ∘ C overnight.After washing, the unbound active sites were blocked with 200 L/tube of blocking reagent.Then, the mixture was incubated for 1 h at 37 ∘ C. Next, 50 L/tube of BaP standard and 50 L/tube of bio-pAb-BaP dilution were added after additional washing and incubated for 1 h at 37 ∘ C.After another washing step, 100 L/tube of SA-HRP dilution was added, and the mixture was incubated for 1 h at 37 ∘ C.After a final washing step, 100 L/tube of TMB + H 2 O 2 substrate solution was added.The enzymatic reaction was stopped by adding 50 L terminating solution (2 mol L −1 H 2 SO 4 ) after sufficient colour development.The absorbance of each well was immediately recorded in dual-wavelength mode (absorbance in 450 nm as test and 630 nm as reference).
Under optimal conditions, the standard curve was constructed by plotting average inhibition (%) values of six replicates against the logarithm of BaP concentrations.Limit of detection (LOD) was evaluated in terms of IC 10 (concentration at which a compound inhibits a particular phenomenon by 10%).Inhibition was calculated as follows ( max is the absorbance in absence of BaP;  min is the absorbance of blank sample; and   is the absorbance of BaP standard solutions): (1)   ) Ionic strength ( According to the characteristic results above, the BaP hapten was synthesized successfully. The prepared BaP-COOH-protein conjugates BSA-BaP and OVA-BaP were identified by the ultraviolet-visible (UVvis) spectrophotometer, respectively.As shown in 7, there were relationships and differences among the characteristic peaks of BaP-COOH, proteins, and conjugates.Moreover, new absorption peaks appeared at 361 nm (for BSA-BaP) and 373 nm (for OVA-BaP) in the UV-vis spectrum of BSA-BaP and OVA-BaP.The results revealed that the conjugates were successfully prepared.Measured by Coomassie Brilliant Blue assay, concentrations of BSA-BaP and OVA-BaP were 5.02 and 2.83 mg mL −1 , respectively.

Optimization of Reaction Conditions.
Coating antigens prepared in different carrier proteins have different structures and molecular weights, which can lead to different abilities to react with the prepared antibody specifically.We selected KLH and PLL as carrier proteins and prepared another two types of coating antigens using the mixed anhydride method described above.Different coating antigens (OVA-BaP, KLH-BaP, and PLL-BaP) were compared.As shown in Table 2, there were no obvious differences between the IC 50 values as well as  0max /IC 50 .Considering the background interferences, OVA-BaP, which obtained the minimum absorbance value (Abs) of the negative control group, was more appropriate for the immunoassay.
The interaction between proteins may affect the adsorption of proteins on the surface of polystyrene carrier at high temperature.Moreover, a long time of incubation has influences on the coating effect and the activity of coating antigens.Therefore, the appropriate incubation condition has a signifi- The optimal concentrations of bio-pAb-BaP and coating antigen were used to improve the sensitivity of the immunoassay.The number of immune complexes reduces rapidly if concentrations of the reagents are not appropriate.A chessboard assay was used to optimize the working concentrations of coating antigen and bio-pAb-BaP.The optimal reaction conditions are the lowest bio-pAb-BaP and coating antigen concentrations that resulted in the maximum absorbance ( 0max ) of approximately 1.0.As shown in Table 4, the optimal coating antigen concentration was 5.66 g mL −1 and bio-pAb-BaP was at a dilution to 1 : 2000 (0.493 g mL −1 ).
Blocking is to eliminate the unoccupied sites on microplates; otherwise the unoccupied sites may interact with the components, such as bio-pAb-BaP and SA-HRP during the subsequent steps.The optimum blocking reagent should achieve the minimal background interference.Several blocking reagents (in PBS) were investigated.As shown in Figure 8(a), 1% gelatin, which achieved the minimal background interference (0.079), was used in the immunoassay.
Different incubation temperatures during the blocking step were compared.The optimum incubation temperature should cause the least background interferences, that is, the lowest absorbance value.As shown in Figure 8(b), the absorbance values obtained at 37 ∘ C and 43 ∘ C were obviously lower than others.Considering the effect on the activity of the coating antigen, 37 ∘ C was selected as incubation temperature during the blocking step.As BaP is a type of hydrophobic organic compound, organic solvent needed to be added in the preparation of the BaP solution.Different concentrations of organic solvent may affect the adsorption capacity of the antibody.The effect of DMSO in different concentrations on the assay was investigated.The result is shown in Figure 8(c).PBS containing 5% DMSO, which induced the lowest IC 50 value (0.22), was selected to improve the analyte solubility in the follow-up test.
Dilutions of SA-HRP (500, 1000, 1500, 2000, and 3000) were also investigated.The nonspecific reaction will increase at high SA-HRP concentrations; however the sensitivity of the immunoassay may be influenced at low SA-HRP concentrations.As shown in Figure 8(d),  0max /IC 50 was the maximum and the IC 50 value was the lowest when the SA-HRP was diluted with PBST to 1 : 1000 (IC 50 = 0.49 g mL −1 ,  0max = 1.07).Hence, the SA-HRP was at a dilution of 1 : 1000.
In the immune reaction of antigen and antibody stage, different incubation times (15,30,45,60,75, and 90 min) were investigated.As shown in Figure 8(e), the  0max value increased with the increase of incubation time.The IC 50 value (0.63) was lowest at 60 min and  0max /IC 50 was the maximum.So, an incubation time of 60 min was selected for the reaction system.
In the reaction system, the hydrogen bonding force of antigen and antibody may be influenced by pH value; different ionic strength also has an effect on the specificity of the immune reaction.As shown in Figures 8(f) and 8(g),  0max decreased with the increase of salt concentrations and the lowest IC 50 (1.81)was induced at an ionic strength of 0.1 mol L −1 . 0max values decreased with the increase of pH and the appropriate IC 50 (1.50)was induced at pH 7.40.So, the salt concentration was 0.1 mol L −1 and the pH of buffer was 7.40 in the immunoassay.

Repeatability of the Immunoassay.
The stability of the proposed immunoassay was estimated by the coefficient of variation (CV).Different concentrations of BaP standard samples (0.05 ng mL −1 , 0.1 ng mL −1 , 0.5 ng mL −1 , 1 ng mL −1 , 5 ng mL −1 , 10 ng mL −1 , and 50 ng mL −1 ) were run in the SA-ELISA method (Table 5).The intra-assay repeatability was evaluated after performing six replicates, and the interassay repeatability was evaluated over several weeks.The CV of intra-assay was <10% and the CV of interassay was <12%, indicating that repeatability and stability of the immunoassay were well.

Determination of BaP in Practical
Samples.The proposed SA-ELISA method was used to detect BaP in several environmental and food samples.The results are shown in Table 6.As the combustion of fossil fuels and industrial production process contribute greatly to BaP emissions, BaP concentration of PM 2.5 samples collected at site 3 which is near a small industrial district was the highest.BaP concentration at site 2 was also significantly high, which was mainly caused by BaP emission of automobile exhaust and human daily lives.In addition, BaP concentration in the agricultural soil sample at site 4 was generally higher than that of site 5.This may be mainly caused by the influences of a small gas distribution plant near site 4.These practical samples were also tested by HPLC to validate the immunoassay.As shown in Table 6, compared with HPLC analysis, the results detected by SA-ELISA were slightly higher.This may be caused by nonspecific adsorption of reagents used in the immunoassay and the cross-reactivity of pAb-BaP with other PAHs existed in the samples, which failed to be analyzed by HPLC method and could contribute to SA-ELISA-derived concentrations.However, the results of the immunoassay still have good consistency with those of HPLC.Statistical study showed that the correlation coefficient was 0.9887 (linear regression equation:  = 0.9555 − 0.3447), and  value from the paired sample -test in comparison with the two methods was 0.0025.That means the difference was statistically significant at 0.05 levels.As the immunoassay was performed in a 96-well microtitre plate in a high throughput detecting manner, it took about 6 h to complete the analysis of 17 treated BaP samples, while HPLC spent about 76.5 h for the analysis (analysis for each sample needs about 15 min and the column needs to be washed after each detection (30 min); six reptiles were tested for each practical sample); it can be concluded that this immunoassay is rapid and effective.
3.6.Recovery Tests.The recovery of several practical samples spiked with BaP analyte was calculated to evaluate accuracy and precision of the immunoassay.Three practical samples (A 2 , P 3 , and spinach) were spiked with BaP standard in different concentrations ranging from 1-50 ng g −1 before extraction.The spiked samples were treated as described above and then tested by the SA-ELISA method.As shown in Table 7, the recovery of SA-ELISA was 91.12-109.23%and the

Comparison with the Correlated Immunoassay.
In the highly sensitive BA-IPCR method for detecting BaP (linear range is 5 pg L −1 -10 ng L −1 ; LOD is 2.85 pg L −1 ) [21], the analysis is based on the PCR technique and the fluorescence labeled in DNA is used as detection signal.The BA-IPCR assay requires an additional step to pretreat PCR tubes to improve the adsorption capacity while there is no need for this operation in the SA-ELISA method.Moreover, after the similar coating and blocking processes, it only needs to add bio-pAb-BaP for immune reaction and to add SA-HRP for chromogenic reaction prior to the final instrumental detection in the SA-ELISA method; however, the BA-IPCR needs step-by-step use of bio-pAb-BaP, streptavidin, biotinylated DNA (bio-DNA), and then PCR mixture, which is much more complicated.With the 96-well microtitre plate, the analysis of nearly 100 BaP samples can be achieved within 6 h while the BA-IPCR spends about 12 h (including the pretreatment of PCR tubes).Therefore, the proposed SA-ELISA is more rapid and convenient.Although the BA-IPCR assay is more sensitive, the actual linear working range is limited due to its trace levels, which may lead to a certain limitation for the detection of some practical samples.In addition, the price of bio-DNA used for PCR analysis is very expensive; this increases the economic cost greatly.Therefore, we consider the SA-ELISA method has a better applicability for detecting BaP samples.

Conclusions
In this study, a specific BaP hapten was synthesized, and diverse BaP antigens and relative polyclonal antibody were prepared; then, a streptavidin-horseradish peroxidasebased enzyme-linked immunosorbent assay was developed and used for the sensitive and high throughput monitoring of BaP in environmental and food samples.Under optimal conditions, this proposed method had a low limit of 0.0094 ng mL −1 .Linear working range was from 0.03 ng mL −1 to 35.40 ng mL −1 .The recovery of spiked samples ranged from 91.12% to 109.23%.IC 50 value of the assay

Figure 1 :
Figure1: Principle of the SA-ELISA method.BaP coating antigens which are covered on the surface of the tubes competed with BaP analytes to combine with the bio-pAb-BaP.The horseradish peroxidase-conjugated streptavidin (SA-HRP) is fixed on the wall through the reaction with bio-pAb-BaP.As the increase of BaP concentration, less bio-pAb-BaP can be attached to the solid carrier.In this case, the amount of SA-HRP is less, and then quantitative analysis could be performed after sufficient colour development.

Figure 8 :
Figure 8: The optimal conditions of the immunoassay.(a) Blocking reagent.(b) Incubation temperature during blocking step.(c) DMSO concentration in PBS buffer.(d) SA-HRP concentration.(e) Incubation time.(f) Ion concentration of PBS.(g) pH of PBS.

Table 1 :
Cross-reactivity of pAb-BaP with structural analogues.

Table 2 :
The optimization of coating antigen.

Table 3 :
The optimization of coating condition for OVA-BaP.
∘ C overnight, 25 ∘ C for 1 h and 2 h, 37 ∘ C for 1 h and 2 h, and 43 ∘ C for 1 h and 2 h) during the coating step were investigated.As shown in Table3, IC 50 value is the lowest and  0max /IC 50 is maximum in the condition of incubation at 4 ∘ C overnight.Therefore, the coating step was completed at 4 ∘ C overnight.

Table 4 :
Optimal concentrations of OVA-BaP and bio-pAb-BaP.The initial concentration of the prepared OVA-BaP was 2.83 mg mL −1 ; the OVA-BaP was diluted with PBS buffer.b The initial concentration of the prepared bio-pAb-BaP was 0.985 mg mL −1 ; the bio-pAb-BaP was diluted with CBS buffer. a

Table 5 :
The variability of intra-assay and interassay for BaP.

Table 6 :
Determination of BaP concentration by SA-ELISA and HPLC.

Table 7 :
1etermination of BaP in spiked samples.1.04ng mL −1 .The SA-ELISA could selectively detect BaP with negligible CR values below 7% against several structural BaP analogues.Meanwhile, the results of BaP detection for several practical samples were consistent with those obtained by HPLC, indicating the proposed method was accurate and reliable.Moreover, due to the application of a 96-well microtitre plate, this immunoassay had potential for the batch-testing detection of BaP samples.Therefore, the established SA-ELISA is a useful option for the batch detection of BaP in environmental and food samples with specificity and sensitivity. was