Development of Hollow-Fiber Liquid-Phase Microextraction Method for Determination of Urinary trans,trans-Muconic Acid as a Biomarker of Benzene Exposure.

For the first time, hollow-fiber liquid-phase microextraction combined with high-performance liquid chromatography-ultraviolet was used to extract trans,trans-muconic acid, in urine samples of workers who had been exposed to benzene. The parameters affecting the metabolite extraction were optimized as follows: the volume of sample solution was 11 mL with pH 2, liquid membrane containing dihexyl ether as the supporter, 15% (w/v) of trioctylphosphine oxide as the carrier, the time of extraction was 120 minutes, and stirring rate was 500 rpm. Organic phase impregnated in the pores of a hollow fiber was extracted into 24 µL solution of 0.05 mol L(-1) Na2CO3 located inside the lumen of the fiber. Under optimized conditions, a high enrichment factor of 153-182 folds, relative recovery of 83%-92%, and detection limit of 0.001 µg mL(-1) were obtained. The method was successfully applied to the analysis of ttMA in real urine samples.

Introduction

Benzene is an important industrial chemical that has been used in numerous production and synthesis processes in the chemical, petrochemical, paint, plastic, adhesive, rubber, and dye industries. Mineral oil, products of gasoline combustion, and cigarette smoke are the other main sources of widespread environmental and occupational exposure.1 Exposure to benzene has serious adverse health effects. Long-term exposure to benzene damages the bone marrow and can cause leukemia; benzene exposure is also associated with other blood cancers and precancers of the blood.2,3 The International Agency for Research on Cancer (IARC) classifies benzene as a group I carcinogen.4 ttMA, catechol, phenol and hydroquinone, and S-phenylmercapturic acid are the biomarkers used for biological monitoring of people exposed to low benzene concentrations (<1).5,6 The American Conference of Governmental Industrial Hygienists (ACGIH) has recommended ttMA as a biological exposure index of occupational exposure to benzene.7 It appears that ttMA is a more specific biomarker than other metabolites at high levels of benzene exposure.8 The ttMA has been found to be a sensitive biomarker for low levels of benzene exposure.5 Since urinary concentration of ttMA in many subjects exposed to benzene is very low, a reliable means of quantitative analysis, extraction, purification, and preconcentration of the analyte is essential. Several analytical methods, including high-performance liquid chromatography–ultraviolet (HPLC-UV), High Performance Liquid Chromatography—Mass Spectrometry and Liquid Chromatography—Tandem Mass Spectrometry (HPLC/MS/MS), liquid chromatography-tandem mass spectrometry (LC–MS–MS), or gas chromatography mass spectrometry (GC-MS), have been used for the analysis of ttMA in urine.9–14 When these methods are applied to the determination of ttMA, preconcentration and cleanup steps are required.

Liquid–liquid extraction (LLE) has been used to prepare the samples prior to the analysis of ttMA in biological fluids. In this technique, ethyl acetate or diethyl ether is used as an extraction solvent for sample cleanup. In several studies, solid-phase extraction (SPE) on strong anion-exchange sorbents was used to achieve purification and preconcentration of the ttMA in urine samples.9–11,15–22

However, these conventional extraction techniques have disadvantages such as coextraction of interfering substances and the use of highly toxic and expensive organic solvents. In addition, these techniques are time and labor intensive. In recent years, hollow-fiber liquid-phase microextraction (HF-LPME), a new miniaturized format of LLE, has been developed for separation and preconcentration of target analytes. Its advantages are high selectivity and specificity, low consumption of organic toxic solvents, and low cost. This method is quick, easy, and has no carryover because of the single use of fiber, high preconcentration, and automation.23–29 In order to promote the extraction of hydrophilic compounds from the donor aqueous phase into the organic phase, carrier-mediated liquid-phase microextraction is used as an active transport mode.30 In biological and environmental applications, carrier-mediated HF-LPME in its two- and three-phase modes has been used to identify analytes and contaminants such as volatile organic compounds, aromatic amines, herbicides, phenols, pesticides, inorganic compounds, and polar drugs in different matrices.31–38

However, till now no studies have been based on application of carrier-mediated HF-LPME technique to the extraction and determination of ttMA in human urine. In this study, carrier-mediated three-phase hollow-fiber microextraction method prior to HPLC-UV was applied to develop a method for extraction, preconcentration, and determination of ttMA in human urine as a biological exposure index for evaluation of environmental and occupational exposure to benzene. Important factors such as the type of organic membrane solvent and carrier concentration, stirring rate, sample and donor pH, sample temperature, and sampling time were optimized, and a simple protocol of HF-LPME method for the assessment of exposure to benzene is represented.

Experimental

Chemicals and reagents

TtMA, tri-n-octyl amine (TOA), tributyl phosphate (TBP), trioctylphosphine oxide (TOPO), 1-octanol, o-xylene, methanol, NaOH, HCl, NaCl, Na2CO3, and other reagents with analytical grade were purchased from Merck. Dihexyl ether was obtained from Sigma-Aldrich. Deionized water was prepared with a Milli-Q system from Millipore. The Q3/2 Accurel PP polypropylene hollow-fiber (HF) membranes (200 µm wall thickness, 600 µm i.d., 0.2 µm pore size) were obtained from Membrana GmbH.

HPLC system

Separation and determination of the analyte were accomplished on a Knauer HPLC system equipped with a K-2600 variable wavelength UV detector, a K-1001 piston pump (Knauer), an online D-14163 degasser, a 6-port/3-channel injection valve fitted with a 20 µL sample loop and ChromGate software (EZChrom Elite). All injections were performed manually. Chromatographic separations were performed using a C18 analytical column (Eurospher, 25 cm × 4.6 mm i.d., 5 µm).

The isocratic run was performed at a constant flow rate of 1.0 mL minute−1, the mobile phase was water/methanol/acetic acid (69:30:1 (v/v/v)), and the ultraviolet detection wavelength was adjusted at 259 nm with an injection volume of 20 µL.

Supported liquid membrane preparation and extraction procedure

Hollow fibers were manually cut into 8.8 cm pieces with an internal volume of 24 µL, washed with acetone in an ultrasonic bath for five minutes to remove any contaminants and then dried. For each experiment, 24 µL of receiver phase was drawn in to a 25 µL Hamilton HPLC syringe and its needle was inserted into the channel of the hollow fiber. Then, the end of the hollow fiber was soaked in organic solvent for 15 seconds. The hollow fibers were rinsed with distilled water for 10 seconds to remove excess organic solvent from their surfaces. The acceptor phase in the microsyringe was introduced into the lumen of hollow fiber, and one end was sealed with a segment of aluminum foil. During extraction, the membrane portion was immersed in the 11 mL aqueous sample that was contained in a 12 mL sample vial. The sample vial having a 4 mm × 14 mm magnetic stirring bar was placed on a heating magnetic stirrer with 0–1000 rpm (Italy) in a U-shaped configuration. After removing the hollow fiber from the aqueous sample, the end of the hollow fiber was cut and the receiver phase was drawn into the syringe and injected into the HPLC system for analysis. To minimize the memory effect, each piece was used only once.

Preparation of standard solutions and real samples

Stock standard solution of ttMA (1000 µg mL−1) was prepared in HPLC-grade methanol–water (1:1) and stored in a refrigerator at 4°C until use. The working solutions containing analyte of interest were prepared daily by diluting the standard solutions prior to use.

To perform the experimental study, urine specimens were taken from nonsmoking healthy subjects living in a rural area that was not exposed to benzene. Spot urine samples from nine workers who had been occupationally exposed to benzene (in a petrochemical plant, car painting, and gasoline station) were collected. All of the subjects were nonsmoking males with an average age of 43.2 ± 9.7 years and had an average of 13.9 ± 10 years on the job. Urine samples were collected at the end of the shift and at the end of the work week, refrigerated and immediately sent to the laboratory, and kept frozen until analysis. Spot urine samples were collected using 200 mL plastic containers. To protect the containers and specimens from contamination, leak-resistant cups were utilized and each was used only once.

Spiked urine samples with ttMA were made freshly from the stock solution prior to analysis. All subjects gave their written, informed consent to participate in the research, which was conducted in accordance with the principles of the Declaration of Helsinki. The research was approved by the Research Ethics Committee of Hamadan University of Medical Sciences and Health Services.

Results and Discussion

Effect of the carrier and its concentration

An effective organic solvent as a supported liquid membrane should be compatible with hollow fiber, be able to dissolve target analytes, have a low viscosity to increase mass transfer through the membrane, and low volatility to prevent loss of solvent in the extraction process.

In this paper, three extractants were used as a carrier. Tri-n-butyl phosphate (TBP), TOPO, and tri-n-octyl amine (TOA) were added to the three organic solvents comprising 1-octanol, dihexyl ether, and o-xylene held in the wall pores of the polymer hollow fiber. According to the results, xylene containing 10% TOPO has the highest extraction efficiency (Fig. 1A), but the results of the extractions in different experiments were not the same and its repeatability was not acceptable. Dihexyl ether comprising 10% TOPO was considered for subsequent experiments.

Figure 1

Effects of different carriers (A) and carrier concentration (B) on extraction efficiency of ttMA by HF-LPME: 11 mL aqueous sample solution of 0.2 µg mL−1 metabolite with pH = 3; 24 µL acceptor phase sodium carbonate 0.1 M (pH = 9); stirring rate: 800 rpm; 60 minutes extraction time.

TOPO in xylene showed the best performance efficiency, but not the best extraction efficiency, because the TOPO in xylene was unstable.

In order to improve their extraction efficiency, the solutions of dihexyl ether containing various concentrations of TOPO immobilized in the pores of the porous polypropylene hollow fiber, in the range of 5%–25% (w/v), were investigated.

The best extraction performance was obtained using 15% (w/v) TOPO in dihexyl ether (Fig. 1B). It is assumed that the high dielectric constant or polarity and extremely low solubility in TOPO water are the most important properties and allows for hydrogen bonding between the acidic hydrogen atoms of nonionized ttMA and the phosphoryl oxygen of TOPO to establish efficient extraction of the target analyte into the organic liquid membrane.39

Effect of donor pH and acceptor composition

The pH of the aqueous donor and acceptor phase should be adjusted to a value that preserves the analyte in the proper form for improving the extraction recoveries. The ttMA is a compound with a pKa value of 3.87.40 It is uncharged in an acidic solution and charged in basic and neutral solutions. In order to investigate the effect of the donor pH on the extraction efficiency, the pH of samples in the 1–3 range was studied. The expected pH value was adjusted dropwise by adding concentrated hydrochloric acid into the donor phase. The results revealed that the best extraction was at pH 2.0 (Fig. 2A). Moreover, sodium carbonate in different concentrations of 0.01–2 M was used to determine appropriate stagnant acceptor phase. The best extraction recovery was achieved by using sodium carbonate 0.05 M (pH 8).

Figure 2

Effects of sample pH (A) and stirring speed (B) on extraction efficiency of ttMA by carrier-mediated HF-LPME. Extraction conditions: 11 mL aqueous sample solution of 0.2 µg mL−1 metabolite; organic membrane phase: 15% TOPO in dihexyl ether; 24 µL acceptor phase sodium carbonate 0.05 M (pH = 8); 60 minutes extraction time.

Effect of stirring speed

The sample stirring rate in carrier-mediated HF-LPME can increase extraction efficiency so that it promotes mass transfer and diffusion of the analyte from the donor phase into the receiver phase and reduces the time required to reach thermodynamic equilibrium.41

In this work, stirring speeds from 300 to 1100 rpm for a 60-minute extraction were studied. The extraction performance increased with stirring rate up to 500 rpm, but decreased at higher rates (Fig. 2B) because high speeds produce air bubbles that cause loss of solvent and reduced precision.

Effect of temperature

Temperature affects the kinetics and the thermodynamics of the distribution process of analytes during extraction and have important effects on the performance of extraction process. Moreover, high temperature increases the loss of organic membrane phase and decreases precision.31,34,42 In order to observe the effect of temperature on extraction efficiency, the temperatures in the range of 20–60 were studied and 40°C was observed to be the best temperature for the experiments (Fig. 3A).

Figure 3

Effects of sample temperature (A) and extraction time (B) on extraction efficiency of ttMA by carrier-mediated HF-LPME. Extraction conditions: 11 mL aqueous sample solution of 0.2 µL/mL metabolite with pH = 2; organic membrane phase: 15% TOPO in dihexyl ether; 24 µL acceptor phase sodium carbonate 0.05 M (pH = 8); stirring rate: 500 rpm.

Effect of extraction time

Liquid-phase microextraction is an equilibrium process. At equilibrium time, analytes in the acceptor phase are in the steady state and stand at their maximum concentration. Moreover, extraction at longer times than the balance time can lead to a decrease in extraction effectiveness because of a loss of organic solvent in pores of hollow fiber or receiving phase.26,41,43 The peak area in the chromatogram of target analyte increased with extraction time up to 120 minutes. So, a sample extraction time of 120 minutes was chosen for the next study (Fig. 3B).

Method validation and application

Under optimized experimental conditions (15% (w/v), such as TOPO in dihexyl ether as organic liquid membrane, sodium carbonate 0.05 M (pH 8) as acceptor phase, and donor pH 2.0, stirring rate of 500 rpm, 120 minutes as the extraction time, and donor temperature of 40°C, the performance and practical applicability of the three-phase carrier-mediated HF-LPME configuration for extraction of ttMA were investigated. SPE method prior to HPLC-UV analysis was used to confirm the detected analyte according to the validated method of Shahtaheri et al.9

Method validation was evaluated on the spiked urine samples comprising the desired amount of the metabolite at various concentrations. Calibration curve of the ttMA was plotted using six spiking samples from 0.005 to 1.2 mg mL−1. To investigate figures of merit of the method, limit of detection (LOD), limit of quantification (LOQ), linearity, linear detection range, repeatability, reproducibility, relative recovery, and enrichment factor (EF) were calculated. The spiked urine calibration curve for ttMA was detected to be linear in the range of 0.005–1.2 µg mL−1 with a correlation coefficient (r2) of 0.996. In this study, to determine LOD and LOQ, spiked urine with known low concentration of ttMA and blank urine samples under optimized experimental conditions were analyzed. The peak-to-peak noise around the metabolite retention time was considered, and then, the concentration of the ttMA that would yield a signal equal to definite value of noise-to-signal ratio was measured. A signal-to-noise ratio (S/N) of 3 was used to calculate LOD and signal-to-noise ratio of 10 was accepted for estimating LOQ. These were found at the ttMA concentration of 0.001 and 0.005 µg mL−1, respectively.

Interday precision was carried out on three consecutive days at three concentrations of the analyte with 18 replications in relative standard deviation (RSD). Intraday precision and accuracy were studied at low, medium, and high concentration levels and with six replications. As shown in Table 1, the RSD% of intra-and interdays were 2.7%–7.3% and 3.4%–8.1%, respectively. The relative recovery, calculated as the ratio of the concentration of the ttMA in the sample solution, was determined with the proposed method to initial concentration of the metabolite in the sample solution, which was 83%–92%. Under optimized conditions, the EF was 153–182 according to the following equation.

Table 1

Optimization of carrier-mediated HF-LPME coupled with HPLC-UV for quantitative extraction of ttMA in human urine.

CONCENTRATION (µg mL−1)	INTRA-DAY RSD% N = 6	INTER-DAY RSD% N = 18	LOD (µg mL−1)	DLR (µg mL−1)	EF	RR%	R2
0.01	7.3	8.1	0.001	0.005–1.2	153–182	83–92	0.996
0.2	6.4	5.6
1	2.7	3.4

Abbreviations: DLR, dynamic linear range; EF, enrichment factor; RR, relative recovery; RSD, relative standard deviation.

Ca, final and Ca, initial in the donor sample, respectively. The blank urine matrix spiked at three concentrations of the ttMA was used to assess the matrix effects. The matrix effect was established as the ratio of ttMA concentration found in real sample and its concentration in distilled water spiked with the same amount of the analyte. No considerable matrix effects were observed, so it could not have had a significant role on extraction efficiency.

The results of validation experiments indicated that the carrier-mediated three-phase HF-LPME followed by HPLC-UV are suitable and feasible for extraction and analysis of urinary ttMA.

Applicability of the Method for Analysis of Real Samples and Comparison with Other Methods

The proposed carrier-mediated HF-LPME–HPLC-UV method was applied to the biological monitoring of benzene exposure. Nine urine samples were obtained from volunteers at different work environments in Hamedan and Asaluyeh, Iran (collected at the end of the shift and at the end of the work week). The ttMA in these real samples were determined. All of the workers were nonsmoking males with an average age of 43.2 ± 9.7 years and had an average of 13.9 ± 10 years on the job. It was reported that these workers are directly exposed to petroleum products or organic solvent vapors containing benzene in the course of the workday. Before the extraction, the real urine samples were diluted 10-fold with deionized water and, under optimized conditions, extracted and analyzed by using the proposed technique. The results of three replicate analyses of each real urine sample are presented in Table 2. However, a simple dilution of sample prior to analysis is sufficient and desirable when the concentration of the metabolite in the sample is already very high. As illustrated in Table 3, this method is more sensitive and precise than the other published methods (LOD and RSDs). Figure 4 shows typical chromatograms obtained by carrier-mediated HF-LPME for the spiked (0.1 µg mL−1) and real urine samples.

Table 2

Concentrations of ttMA detected in human urine of exposed workers.

NO OF SAMPLE URINE	WORKPLACE	ttMA CONCENTRATION (µg/mL) MEAN ± SD, N = 3
1	Petrochemical plant	1.5 ± 0.13
2	Petrochemical plant	0.7 ± 0.09
3	Petrochemical plant	1.2 ± 0.11
4	Car painting	1.9 ± 0.10
5	Car painting	3.8 ± 0.12
6	Car painting	5.9 ± 0.15
7	Gasoline station	1.5 ± 0.05
8	Gasoline station	0.7 ± 0.07
9	Gasoline station	2.1 ± 0.11

Table 3

Comparison of the carrier-mediated HF-LPME–HPLC-UV method with other preconcentration techniques for extraction and determination of ttMA in human urine.

EXTRACTION	DETECTION	DETECTION LIMIT (µg mL−1)	RSD%	REFERENCES
SPE-derivatization	GC-MS	0.01	7.4	10
SPE	HPLC-UV	0.01	2.7–10	9
SPE	HPLC/MSMS	0.001	3–10	11
LLE	GC-FID	0.02	9.7	16
Filtration	LC–LC–UV	0.05	0.5–5	15
LPME	HPLC-UV	0.001	2.7–7.3	Proposed method

Abbreviations: SPE, solid-phase extraction; LLE, liquid-liquid extraction; LPME, liquid-phase microextraction.

Figure 4

HPLC chromatograms of ttMA after the extraction under optimized conditions as described in the text. (A) Spiked urine sample of ttMA at a concentration of 0.1 µg mL−1 and (B) real urine sample from a petrol station worker exposed to benzene source.

Conclusion

In this work, we presented a new pretreatment method for the quantitative determination of ttMA in urine, which can be used to evaluate occupational and environmental monitoring of exposure to very low concentration of benzene. The factors influencing HF-LPME based on carrier-mediated transport were optimized through this research, concluding that it is an efficient, reliable precision, simple, accurate, and less expensive procedure than conventional extraction methods like LLE, SPE, and derivatization for sample preparation of ttMA as a biomarker of benzene exposure. This method exhibited an EF of around 182, good relative recoveries of 83–92, a very low LOD of 0.001 µg mL−1 with wide dynamic linear ranges (0.005–1.2 µg mL−1) and good reproducibility (RSD, less than 8.1%). All these are great advantages over other sample preparation procedures. With this simple extraction procedure, multifold extractions can be performed at the same time with a multistirrer.