Verification of Chlorine Exposure via LC-MS/MS Analysis of Base Hydrolyzed Chlorophenols from Chlorotyrosine-Protein Adducts

Abstract

Inhalation of chlorine gas, with subsequent hydrolysis in the airway and lungs to form hydrochloric acid (HCl) and hypochlorous acid (HOCl), can cause pulmonary edema (i.e., fluid build-up in the lungs), pulmonary inflammation (with or without infection), respiratory failure, and death. The HOCl produced from chlorine is known to react with tyrosine to form adducts via electrophilic aromatic substitution, resulting in 3-chlorotyrosine and 3,5-dichlorotyrosine adducts. While several analysis methods are available for determining these adducts, each method has significant disadvantages. Hence, a simple and sensitive ultra-high performance liquid chromatography-tandem mass spectroscopy (UHPLC-MS/MS) method was developed for the determination of chlorotyrosine adducts. The sample preparation involves base hydrolysis of isolated plasma proteins to form 2-chlorophenol (CP) from monochlorotyrosine adducts and 2,6-dichlorophenol (2,6-DCP), from dichlorotyrosine adducts, as markers of chlorine exposure. The chlorophenols are extracted with cyclohexane prior to UHPLC-MS/MS analysis. The method produced excellent sensitivity for 2,6-DCP with a limit of detection of 2.2 μg/kg, calibration curve linearity extending from 0.054 - 54 mg/kg (R² ≥ 0.9997 and %RA >94), and accuracy and precision of 100±14%, and <15% relative standard deviation, respectively. The sensitivity of the method for 2-CP was relatively poor, so it was used only as a secondary marker for severe chlorine exposure. The method successfully detected elevated levels of 2,6-DCP from hypochlorite-spiked plasma protein and plasma protein isolated from chlorine-exposed rats.

1.0. Introduction

Chlorine is a dense yellow-green gas at room temperature. It has a pungent, irritating odor characteristic of bleach [1]. It is slightly water soluble but reacts quickly with moisture to form hypochlorous acid (HOCl) and hydrochloric acid (HCl). Chlorine is used in a variety of industries, such as health, agro-food, textiles, transport, and cosmetics [2, 3]. Chlorine is also used for plastic production, pulp and paper production, water purification, and chemical synthesis [3].

While chlorine has many industrial uses, it is highly toxic. Exposure to chlorine can be historically divided into a number of categories: chlorine gas as a chemical warfare agent in military applications (e.g., Cl₂ was used during the First World War at Ypres in April 1915) [3, 4]; occupational exposure of industrial workers [5-8]; leaks or spills of chlorine including transportation mishaps [9, 10]; and exposure involving commercially available chlorine-containing products, such as pool chemicals or laundry bleach [11-22]. Chlorine has a toxic threshold limit value (TLV) of 0.1 ppm (i.e., the TLV is defined as the concentration of the compound in the air that can be breathed for five consecutive eight-hour working days by people without significant damage to the health) [23, 24]. At approximately 30 - 50 ppm, chlorine causes immediate substernal chest pain, shortness of breath, cough, and toxic pneumonitis with acute pulmonary edema [25]. Higher level exposures (> 400 ppm) result in asphyxia with respiratory failure, pulmonary edema, acute pulmonary hypertension, cardiomegaly, pulmonary vascular congestion, and acute burns of the upper and proximal lower airways and can result in death [25-27]. Although chlorine is toxic, its strong odor can be detected by humans at 0.1-0.3 ppm, and mucus membranes become irritated at >1 ppm. While this results in limited harm during most exposure events, large-scale and/or sudden exposure events are still very concerning [28, 29].

Chlorine toxicity results from the formation of HOCl and HCl following contact with surface liquid on mucosal surfaces and airways [25]. Aside from their inherent toxicity, HCl and especially HOCl may additionally produce highly reactive oxidants via combination with reactive oxygen species and other airway constituents [30]. This causes instant oxidative injury to the epithelium and migration and activation of inflammatory cells such as neutrophils. Chlorine exposure can damage eyes, skin, and upper airways from the nose to the bronchi [26, 31]. Furthermore, subepithelial fibrosis, mucous hyperplasia, and nonspecific airway hyperresponsiveness have been reported following chlorine injury [26, 32-34].

Exposure to chlorine gas results in chlorine adducts with proteins, specifically at tyrosine residues (see Fig 1) [35, 36]. Chlorine modifies these residues via electrophilic aromatic substitution at the ortho position(s), causing the formation of 3-chlorotyrosine and 3,5-dichlorotyrosine adducts [37]. These adducts can be utilized as biomarkers to confirm chlorine exposure. For example, both biomarkers were readily detected in rat nasal tissue following a 90 min exposure to chlorine gas (~3 ppm) [38]. Moreover, the chlorotyrosine adducts have been broadly studied as biomarkers for inflammation and oxidative tissue damage resulting from neutrophil myeloperoxidase in cases of chronic inflammatory disease [36, 39-42]. Specifically, increased levels of monochlorotyrosine adducts have been correlated with renal failure [43], atherosclerosis (ATH) [42], myocardial infarction [44], and cystic fibrosis (CF) [45]. The higher-order dichlorotyrosine adduct is less abundant but may prove to be the more specific biomarker when separating acute and/or relatively high-level chlorine exposure from chronic inflammatory disease [46]. There are other adducts has been utilized as biomarkers to confirm chlorine exposure in human, those are chlorinated lipids, and phosphatidylglycerol chlorohydrin [47-51]. However chlorotyrosine adducts are generally considered more common biomarkers in relevent research.

Fig 1.

Reaction scheme for the formation of chlorophenols from the reaction of tyrosine residues with HOCl (produed via Cl₂ exposure) and subsequent base-catalyzed hydrolysis. Tyrosine residues react with HOCl via electrophilic aromatic substitution to form a chlorotyrosine adduct which can undergo base hydrolysis to produce 2-CP and 2,6-DC

Multiple methods of analysis of chlorotyrosine adducts have been developed. These methods are listed in Table 1 along with comparison of some important parameters. Acid hydrolysis is the earliest and most common technique used to prepare chlorotyrosine adducts. Acid hydrolysis is achieved via hydrolysis of protein with a strong acid (~ 6 N HBr or methane sulfonic acid containing 1% phenol) at high heat (~110 °C for ~ 18 h) to produce isolated chlorotyrosine (Cl-Tyr) and dichlorotyrosine (Cl₂-Tyr) for analysis. Acid hydrolysis can result in the degradation of heat or acid sensitive compounds due to the harsh conditions involved. This can lead to the formation of unwanted byproducts or the loss of desired products [39, 42-46, 52]. Another approach for chlorotyrosine adduct analysis is enzymatic hydrolysis, which is performed by 24 h incubation with pronase at 37 °C to also produce Cl-Tyr and Cl₂-Tyr for analysis. As enzymatic hydrolysis requires 24 h incubation, it is a lengthier sample preparation. It is also more costly, and its efficiency can be influenced by factors like pH, temperature, and substrate concentration [40, 53]. Cl-Tyr and Cl₂ -Tyr have been analyzed by GC-MS [38, 39, 41-46], HPLC–UV [40], and LC-MS/MS [53] (see Table 1). Each of these methods involves more than 20 steps, over 24 h of sample preparation, and requires a relatively large amount of organic solvent and energy.

Table 1.

Comparison of methods for the analysis of chlorotyrosine adducts.

Analyte	Sample preparation (each includes protein precipitation)	Analysis technique	Year	Reference
Cl-Tyr	Acid hydrolysis, SPE, derivatization	NCI GC-MS	1996	Hazen et al. [39]
Cl-Tyr	Acid hydrolysis, SPE, derivatization	NCI GC-MS	1997	Hazen et al. [42]
Cl-Tyr	Acid hydrolysis, SPE, derivatization	NCI GC-MS	2001	Himmelfarb et al. [43]
Cl-Tyr	Acid hydrolysis, SPE, derivatization	GC-MS	2007	Mocatta et al. [44]
Cl-Tyr & Cl2-Tyr	Acid hydrolysis, SPE, derivatization	GC-MS	2004	Kettle et al. [45]
Cl-Tyr & Cl2-Tyr	Acid hydrolysis, RP-HPLC, derivatization	GC-MS	2000	Chapman et al. [46]
Cl-Tyr & Cl2-Tyr	Acid hydrolysis, SPE, derivatization	GC-MS	2008	Sochaski et al. [38]
Tyr & Cl-Tyr	Enzymatic hydrolysis, derivatization	RP-HPLC-UV	1996	Kettle et al. [40]
Cl-Tyr & Cl2-Tyr & CHPA & DCHPA	Acid hydrolysis, derivatization	RP-HPLC-UV	2000	Davies et al. [52]
Cl-Tyr & Cl2-Tyr	Enzymatic hydrolysis, SPE	RP-HPLC-MS/MS	2022	Bruin-Hoegee et al. [53]
Cl-Tyr &Cl2-Tyr	Enzymatic hydrolysis, SPE	UHPLC-MS/MS	2016	Crow et al [37]
Cl-Tyr &Cl2-Tyr	Enzymatic hydrolysis	UHPLC-MS/MS	2021	Pantazides et al. [69]
2-CP & 2,6-DCP	Base hydrolysis, LLE	RP-UHPLC-MS/MS	2024	Current study

Cl-Tyr = 3-chlorotyrosine, Cl₂-Tyr = 3,5-dichlorotyrosine, Tyr = tyrosine, CHPA = 3-chloro-4-hydroxyphenylacetaldehyde, DCHPA = 3,5-dichloro-4-hydroxyphenylacetaldehyde, 2-CP = 2-chlorophenol, 2,6-DCP= 2,6-dichlorophenol, NCI-GC-MS = negative chemical ionizationgas chromatography-mass spectrometry, RP-HPLC-MS/MS = reversed phase high performance liquid chromatography-tandem mass spectrometry, RP-HPLC-UV = reversed phase high performance liquid chromatography-ultraviolet, SPE= solid-phase extraction, LLE= liquid-liquid extraction

Because each of the available methods for chlorotyrosine adduct analysis has significant drawbacks, there is a need to develop a simple and sensitive sample preparation method to analyze protein adducts of chlorotyrosine to verify exposure to chlorine. This study aimed to develop a simple and sensitive sample preparation method utilizing a novel strong base hydrolysis technique for chlorine-adducted tyrosine plasma protein residues to form the phenolic hydrolysis products 2-CP and 2,6-DCP and to analyze them via UHPLC-MS/MS to verify acute, high-level exposure to chlorine.

2.0. Materials and Methods

2.1. Materials

All reagents and solvents were at least HPLC grade unless otherwise stated. Ammonium acetate (LC/MS grade), sodium hydroxide, cyclohexane (Certified ACS), acetonitrile, and ammonium hydroxide were purchased from Fisher Scientific (Fair Lawn, NJ, USA). Water used for this study was purified by reverse osmosis and filtered through a Lab Pro polishing unit from Labconco (Kansas City, KS, USA) and had a resistivity of 18.2 MΩ-cm (referred to as deionized (DI) water). Hydrochloric acid, 2-chlorophenol (2-CP), and 2,6-dichlorophenol (2,6-DCP) were obtained from Sigma-Aldrich (St. Louis, MO, USA). In addition, isotopically labeled 2,4-dichlorophenol ring-d₃ (2,4-DCP-d₃) with 98% purity was purchased from Cambridge Isotope Laboratories, Inc. (Tewksbury, MA, USA). A freshly prepared mixture of 2-CP and 2,6-DCP was prepared as a stock solution in methanol (10 mM) for each experiment, and the stock solution was kept in the dark at ambient temperature. The as-received internal standard solution was diluted in methanol to produce a stock solution of 10 mM, which was further diluted to 1 mM in DI water and stored in a −80 °C freezer (So-Low Ultra-Low Freezer, Ohio, USA). A Parker Balston nitrogen gas generator with a pierce Reacti-Therm heating module was used for solvent evaporation.

2.2. Biological samples

Sprague Dawley rat plasma (3.2% sodium citrate) was purchased for method development and validation from BioIVT (Westbury, NY, USA) and immediately stored at − 80 °C until used. This plasma was used for protein collection, NaOCl spikes, and other procedures necessary to develop and validate the method.

To verify the method’s ability to detect 2,6-DCP as a marker of chlorine exposure, rats were exposed to chlorine, and plasma was sampled for 2,6-DCP analysis. This work was done in collaboration with the University of Colorado Denver. Male Sprague Dawley rats (300 -350 g) were obtained. The strain and vendor of the animals can impact how they react to agents and anesthesia. Sprague-Dawley rats, male, from Envigo were used to reduce variability and consistency of exposure. All procedures were approved by the UCD-AMC Institutional Animal Care and Use Committee.

Many experiments over many years were performed to characterize the injury at various concentrations of chlorine. A certified 600 ppm chlorine tank was used to expose the animals (N=8) for 27 min (LD50 = 6 h) via whole-body exposure in a sealed 5 L cylindrical glass chamber, with a 5-min ramp up and 5-min ramp down, no dilutions, just direct exposure to what comes out of the tank. Rats were anesthetized by IP injection of a mixture of ketamine (75 mg/kg), xylazine (7.5 mg/kg), and acepromazine (1.5 mg/kg), and blood was collected from the descending aorta into blood collection vials with 3.2% sodium citrate. There was a control group (N=5) that had no exposure. Rat plasma was fractionated by centrifugation of whole blood at 3000 x g for 15 min at 4 °C. Plasma was removed, and aliquots (0.5-1 mL) of blood plasma were immediately frozen and stored at −80 °C. The frozen plasma samples were then shipped to South Dakota State University on dry ice and stored at −80 °C until analysis.

2.3. Sample preparation

Plasma protein precipitation was achieved by following the method of Logue et al [54]. Briefly, plasma samples (250 μL) were added to a 2-mL centrifuge tube, and acetonitrile was added to the plasma samples with a 3:1 volume ratio of acetonitrile to plasma. The mixture was vortexed and centrifuged at 20000 x g for 10 min at 5 °C. The supernatant was discarded, the precipitate was collected, the pellet was broken, and the protein was rewashed with the same volume of fresh acetonitrile. The mixture was vortexed and centrifuged, and the supernatant was discarded as above. The residue was dried in a centrifugal evaporator (Labconco, Kansas City, USA) equipped with a rotary vacuum pump (Edwards, Glenwillow, USA) at 30 °C until dry. If necessary, the prepared precipitate was stored at −80 °C. Next, the precipitated plasma protein (15 mg) was added to a 2-mL centrifuge tube, and aqueous NaOH (200 μL of 1 M) was added to the plasma protein for base hydrolysis. The internal standard (2,4-DCP-d₃) was spiked (100 μL, 2 μM) into the solution. The solution was then vortexed and heated at 75 °C for 2 h. The sample was removed from heat and cooled to room temperature. To neutralize the base, HCl (100 μL, 3 M) was added to the solution. Cyclohexane (1.2 mL) was added to the mixture to extract 2,6-DCP, and the resulting mixture was capped, vortexed, and centrifuged at 13000 x rpm for 20 min at 5 °C. An aliquot of the organic layer (1 mL) was then transferred into a 4-mL glass screw-top vial and nitrogen dried for 13 min at room temperature. The dried samples were reconstituted with 100 μL of 50% acetonitrile in ammonium acetate buffer (0.1 mM, pH 10.0), mixed thoroughly, filtered with a 0.22 μm nylon syringe filter, and analyzed using UHPLC-MS/MS.

2.4. UHPLC-MS/MS analysis of 2,6-DCP

UHPLC-MS/MS analysis was performed on a Shimazu UHPLC with an LC-20ADXR controller connecting to a tandem mass spectrometer (Sciex Q-Trap 5500 MS) equipped with an electrospray ionization interface. The column used for chromatography was a Waters reversed-phase column, charged surface hybrid (CSH) C18 (3.0 x 75 mm, 2.5 μm). The chromatographic separation was achieved using isocratic elution at a flow rate of 0.3 mL/min at 90% B for 5 min. Mobile phases A and B were ammonium acetate buffer in water (0.1 mM, pH 10.0) and acetonitrile, respectively. The column was equilibrated for 1 min, and a volume of 10 μL was injected for UHPLC-MS/MS analysis. The oven temperature was 30 °C.

The electrospray ionization (ESI) interface in the negative mode was used to detect 2,6-DCP, 2,4-DCP-d₃, and 2-CP. Mass spectrometric conditions were optimized by directly infusing a standard solution of 2,6-DCP, 2,4-DCP-d₃, and 2-CP into the mass spectrometer at a flow rate of 10 pL/min. After infusion of standard solutions into ESI, molecular ions of m/z 160.9 ([M−H]⁻), m/z 163.8 ([M−H]⁻), and m/z 126.9 ([M−H]⁻), respectively, were identified. Multiple reaction monitoring (MRM) parameters for 2,6-DCP, 2,4-DCP-d₃ and 2-CP were optimized and are outlined in Table 2. Nitrogen (20 psi) was used as the curtain and nebulization gas. The ion spray voltage was 4,500 V, the source temperature was 600 °C, and the nebulizer (GS1) and heater (GS2) gas pressure were 30 psi. The collision cell was operated at a “medium” collision gas flow rate. The total mass spectrometry acquisition time was 5 min. Note: although the sensitivity of the method for 2-CP was low and its method validation parameters were not determined, it was retained in the method to help differentiate extreme exposure events.

Table 2.

MRM transitions, optimized collision energies (CEs), collision cell exit potentials (CXPs), and declustering potentials (DPs) for detection of 2,6-DCP and 2,4-DCP-d₃ and 2-CP by MS/MS analysis

Compounds	Q1 (m/z)	Q3 (m/z)	Dwell Time (msec)	CE (V)	CXP (V)	DP (V)
2,6-DCP (quantification)	160.9	125.0	100	−24.81	−8.34	−85.86
2,6-DCP (identification)	160.9	89.0	100	−38.98	−9.71	−65.02
2,4-DCP-d3 (quantification)	163.8	126.9	100	−24.51	−8.94	−37.27
2,4-DCP-d3 (identification)	163.8	90.0	100	−30.06	−8.23	−39.34
2-CP (quantification)	126.9	91.0	100	−66.67	−24.62	−16.87
2-CP (identification)	126.9	65.0	100	−63.60	−41.28	−30.72

2.5. Limit of detection, calibration, quantification, and sensitivity

Food and Drug Administration guidelines were used to validate the method [55-59]. The limit-of-detection (LOD) is generally defined as the lowest concentration which reproducibly produces a signal-to-noise ratio of at least 3. Since endogenous levels of chlorinated protein adducts in plasma protein are known to be present from natural sources [60, 61], it is difficult to determine an accurate LOD. Therefore, a realistic LOD estimate was calculated using Equation ((1) [62]. The blank signal was estimated by averaging the signal 1 min prior to and 1 min following the analyte retention time (i.e., to avoid signal produced from endogenous concentrations of 2,6-DCP).

$${\overline{L}OD}_{signal}=x_{blank}+3s_{blank}$$ (1)

where ${\text{LOD}}_{\text{signal}}$ is the signal LOD, ${\overline{\mathrm{x}}}_{\text{blank}}$ is the average signal of blank, and ${\mathrm{s}}_{\text{blank}}$ is the standard deviation of blank measurement. The signal LOD was converted to a concentration using the calibration curves produced during validation.

Calibration standards (0.05, 0.1, 0.2, 0.5, 1, 2, 5, 10, 20, and 50 μM) and quality control (QC) standards (0.3, 3, and 30 μM) were prepared from a stock solution of 2,6-DCP in 50% acetonitrile:ammonium acetate buffer. Aliquots (100 μL each) of 2,6-DCP and 2,4-DCP-d₃ were spiked into plasma protein to create calibration standards utilizing the sample preparation method described above. The aqueous standards (0.3, 3, and 30 μM) were also prepared from the same stock solution. The QC standards were not included in the calibration curve. Each calibration standard was prepared in triplicate, with QC standards in quintuplicate. After analysis of the calibration standards, calibration curves were plotted using the average signal ratios of 2,6-DCP and 2,4-DCP-d₃ as a function of 2,6-DCP nominal concentration. Based on the criteria of <15% relative standard deviation (%RSD, as a measure of precision) and a percent error (as a measure of accuracy) of 100±15%, the lower limit of quantitation (LLOQ) and upper limit of quantification (ULOQ) were established. The LLOQ is generally at least 3 times higher than the LOD. The percent error was calculated based on comparing the 2,6-DCP concentration back-calculated from the calibration curve to the nominal 2,6-DCP concentration of the QC standard. The percent residual accuracy (PRA) was used to determine the goodness-of-fit of the calibration curves (i.e., PRA values ≥90% are indicative of a good fit) [63]. Intraassay precision and accuracy were calculated from each day’s analysis, and interassay precision and accuracy were calculated by comparing the data gathered over three separate days.

The ability to differentiate and quantify 2,6-DCP in the presence of other plasma protein constituents (i.e., selectivity) was determined by precipitating blank rat plasma protein and initiating base hydrolysis. Some of these samples were spiked with 2,6-DCP and all samples were prepared by the procedure described above. The spiked samples were compared with the unspiked samples to evaluate method selectivity. Additionally, because some endogenous chlorotyrosine adducts are expected, the sample preparation procedure was also performed without protein.

2.6. Recovery and matrix effect

The recovery of 2,6-DCP was determined by analyzing five QC replicates (low, medium, and high concentrations) prepared in an aqueous solution compared with equivalent concentration QCs in plasma protein. Recovery, evaluated in this manner, is a combination of matrix effect and loss of the analyte during sample preparation. Therefore, recovery was determined by correcting the peak area of the 2,6-DCP signal in plasma protein for the matrix effect and then dividing by the peak area of the equivalent concentration of aqueous QC standards containing 2,6-DCP reconstituted in 50% acetonitrile in ammonium acetate buffer and representing as a percentage. Recovery calculated in this manner is influenced by matrix effects. Matrix effects for the analysis of 2,6-DCP were evaluated by comparing an aqueous calibration curve with a calibration curve in plasma protein. The ratio of calibration curve slopes, using plasma protein matrix versus aqueous calibration standards, was used to estimate the matrix effect. A slope ratio (plasma protein slope/aqueous slope) equal to one indicates no matrix effect, less than 1 represents the suppression of the analyte signal, and greater than 1 represents an enhancement effect by the plasma protein. The effectiveness of the internal standard to compensate for the matrix effect and limited recovery was also evaluated.

2.7. Stability

Both short- and long-term stability of 2,6-DCP in plasma protein matrix was evaluated in triplicate using low and high QCs stored at various temperatures at multiple time periods. For short-term stability, low and high QCs were evaluated in an autosampler. Prepared samples were placed on the auto-sampler (ambient temperature) and analyzed at approximately 0, 1, 2, 4, 8, 12, and 24 h following preparation. The internal standard was not considered when evaluating the autosampler stability, as it would correct for the loss of 2,6-DCP in the plasma protein matrix during the analysis. To evaluate the benchtop stability of the dichlorotyrosine-protein adduct, hypochlorite was initially spiked in rat plasma at concentrations that produced 2,6-DCP signals similar to the low and high QCs: 1.7 μmol/g and 25 μmol/g, respectively. The hypochlorite-spiked rat plasma protein samples were vortexed and placed at room temperature for 2 h. The samples were then stored on the benchtop for 0, 1, 2, 4, 8, 12, and 24 h. The samples were subsequently precipitated, dried, and stored in a −80 °C freezer until all samples were ready for analysis. During analysis, samples were prepared using the sample preparation steps outlined above and analyzed using the developed UHPLC-MS/MS method. Sample preparation was carried out for long-term stability as with the benchtop stability, except that samples were stored at room temperature, 4 °C, −20 °C, and −80 °C for 0, 1, 2, 5, 15, and 30 days. Stability was measured as a percentage of the “time 0” signal for all stability experiments.

3.0. Results and Discussion

3.1. UHPLC-MS MS analysis of 2,6-DCP

Donkor et al. [64] utilized base hydrolysis of methyl isocyanate (MIC)-adducted tyrosine residues to produce phenylmethyl carbamate for simple analysis of MIC-adducted proteins. We hypothesized that chlorotyrosine adducts would behave similarly to produce 2-CP and 2,6-DCP. To confirm the formation of these compounds from chlorine exposure, an analytical method was developed to detect and quantify both 2-CP and 2,6-DCP from the hydrolysis of chlorine-adducted plasma protein. The mass spectra of 2,6-DCP, 2,4-DCP-d₃, and 2-CP produced by ESI (−) MS are shown in Fig 2, with abundant ions identified. For 2,6-DCP and 2,4-DCP-d₃, the proposed major MS fragmentation occurs at two of the carbon─chlorine bonds [65, 66]. Note that while the method was initially developed for both analytes, the sensitivity of 2-CP was severely limited. Therefore, the optimization of the method continued for 2,6-DCP alone. Note that unclear as to if inefficient base hydrolysis of the monochlorotyrosine adduct contributes to the limited sensitivity of the LC-MS/MS method for monochlorotyrosine adducts.

Fig 2.

ESI (−) product ion mass spectra of 2,6-DCP (A), 2,4-DCP-d₃ (B), and 2-CP (C) with suggested structural assignments identification of the abundant ions. Molecular ions of 2,6-DCP, 2,4-DCP-d₃, and 2-CP [M−H]− correspond to 160.9, 163.8, and 126.9, respectively. Insets, the structures of 2,6-DCP (A), 2,4-DCP-d₃ (B), and 2-CP (C) with suggested fragmentation are shown.

This method allowed detection of 2,6-DCP (eluting at about 1.7 min) and 2,4-DCP-d₃ (eluting at about 2 min) from spiked rat plasma protein precipitates (Fig 3). Additionally, 2,6-DCP was detected from NaOCl-spiked rat plasma (Fig 3C). UHPLC-MS/MS analysis of 2,6-DCP and 2,4-DCP-d₃ produced sharp and symmetrical peaks with peak asymmetry factors of <1.1 and <1.3, respectively, and excellent efficiency. Both peaks were resolved from other components in the plasma matrix. The overall analysis required 12 steps and lasted approximately 4 h, with a 5 min chromatographic analysis time. This produced far shorter analysis times and reduced effort compared to previous acid hydrolysis and enzymatic digestion methods (Table 1). Using this method, roughly 90 parallel samples could be processed and analyzed in a 24-h period.

Fig 3.

UHPLC-MS/MS chromatograms of (A) 2,6-DCP-spiked rat plasma, (B) 2,4-DCP-d₃ spiked rat plasma, and (C) 2,6-DCP from NaOCl-spiked rat plasma, following sample preparation are presented. The 160.9→ 125 m/z and 163.8→ 126.9 m/z transitions for 2,6-DCP (A, C) and 2,4-DCP-d₃ (B) are plotted. The insets show the magnified baseline of unspiked rat plasma.

3.2. Dynamic range, the limit of detection, and sensitivity

Initially, the linearity of the method was evaluated in the range of 0.011–109 mg/kg precipitated plasma protein. Multiple calibration curves were constructed to analyze the calibration behavior of 2,6-DCP by plotting the concentration of 2,6-DCP versus the signal ratio (i.e., the peak area of 2,6-DCP at 160.9→125 divided by the corresponding peak area of 2,4-DCP-d₃ at 163.8→126.9) as a linear relationship over multiple calibration ranges. We confirmed that the calibration behavior was best described by non-weighted linear least-squares regression. The 0.011, 0.022, and 109 mg/kg calibration standards fell outside the linear range when the inclusion criteria were 100 ±15% for accuracy and ≤15% for precision. Therefore, the method had a linear range of 0.054–54 mg/kg, corresponding to 0.05-50 μM. The LLOQ was 0.05 μM, which was at least 3 times higher than the LOD (0.0072 μM or 7.8 μg/kg). The LOD was 7.8 μg/kg, as calculated by Equation (1) and converted to concentration using the constructed calibration curves. This LOD corresponds to 7.2 nM in the base hydrolysate. This high sensitivity is very important to accurately quantify 2,6-DCP in plasma protein samples [46].

Calibration curve equations of three separate calibration curves prepared over a 3-day period with their corresponding R² and PRA values shown in Table 3. All three calibration curves were found to be highly reproducible in terms of slope, R², PRA, accuracy, and precision. The %RSD of the slopes for the three calibration curves was <15 %.

Table 3.

Calibration equations, coefficients of determination (R²), and PRA for 2,6-DCP calibration curves created over 3 days.

Day	Calibration Equation	R2	PRA(%)	Accuracy (%)	Precision (%RSD)
1	y = 0.5646x-0.0033	1.00	95.3	100 ± 14.02	<15
2	y = 0.7134x-0.0091	0.9997	94.7	100 ± 9.4	<10
3	y = 0.6953x-0.0058	0.9998	94.8	100 ± 8.6	<8.5

3.3. Accuracy and precision

The accuracy and precision of the method were established by quintuplicate analysis of three QC standards: low, medium, and high (0.3, 3, and 30 μM, corresponding to 0.33, 3.3, and 33 mg/kg) on three days within 7 calendar days. FDA method validation guidelines were followed to evaluate the results [56, 67, 68]. The precision and accuracy of the method were excellent (Table 4), with interassay and intraassay accuracies within 100 ± 6% and 100 ± 13%, respectively (Note: one data point was + 13% and all other were <8%). The interassay and intraassay precisions were also within ≤6.6% and ≤12% of the nominal concentration, respectively.

Table 4.

Intra- and interassay accuracy and precision of 2,6-DCP produced by base hydrolysis of Cl₂-tyrosine protein adducts.

Nominal concentration (μM)	Intraassay accuracy (%)a			Intraassay precision (%RSD)a			Interassay accuracy (%)b	Interassay precision (%RSD)b
	Day 1	Day 2	Day 3	Day 1	Day 2	Day 3
0.3	100± 13	100 ± 5.4	100 ± 1.5	9.2	3.1	6.9	100± 5.4	< 6.6
3	100 ± 7.5	100 ± 2.7	100 ± 4.02	10.9	9.5	8.8	100± 0.25	< 6.3
30	100 ± 1.6	100 ± 6.3	100 ± 0.91	11.6	8.5	6.4	100± 1.3	< 4.5

^aQC method validation (N=5).

^bMean of three different days of QC method validation (N=15).

3.4. Matrix effect and recovery

To evaluate the matrix effect, calibration curves were constructed in plasma protein and aqueous samples. The ratio of slopes (m_plasma/m_aq) was 0.88. Therefore, a minor matrix effect of approximately 12% suppression of 2,6-DCP signals was present in the rat plasma protein matrix. To evaluate the ability of the internal standard to correct for the matrix effect, the IS-corrected slope ratio was calculated as 0.96, revealing the ability of IS to correct for matrix effects.

The recoveries (determined by dividing the peak area of the 2,6-DCP signal in plasma protein by the peak area of the equivalent concentration of aqueous QC standards containing 2,6-DCP reconstituted in mobile phases and accounting for matrix effect) for low, medium, and high QCs were 43.6%, 45.8%, and 46.7%, respectively. The loss of 2,6-DCP during the sample preparation steps is most likely due to the incomplete extraction (part of the analytes remain in the aqueous layer) of 2,6-DCP from the plasma matrix into the organic layer. Comparison of the recovery and matrix effect reveals that the loss of 2,6-DCP during sample preparation is the major contributor to the loss of signal for the method. To investigate the ability of the internal standard to correct for the low recovery and matrix effect, the internal standard was used to correct the signals for the recovery experiment. The corrected recoveries were 98.8%, 96.7%, and 97.6% for the low (LQC), medium (MQC), and high QC (HQC) standard, respectively. Therefore, the internal standard did an excellent job of correcting for recovery and the matrix effects and is essential to ensure accurate quantification of 2,6-DCP.

3.5. Stability of 2,6-DCP and dichloro-tyrosine adducts

The stabilities of 2,6-DCP and the dichlorotyrosine adduct were evaluated under multiple storage conditions. To determine the autosampler stability of 2,6-DCP (at 15 °C), LQC and HQC standards were placed in the autosampler after they were processed using the sample preparation described herein and were subsequently analyzed after 0, 1, 2, 4, 8, 12 and 24 h. 2,6-DCP signals for both LQC and HQC samples were stable for all times tested. We also investigated the stability of the chlorotyrosine adducts on the benchtop. Samples were prepared and analyzed after 0, 1, 2, 4, 8, 12, and 24 h of storage. Both LQC and HQC concentrations of chloro-tyrosine adducted protein were stable for the entire period tested.

For long-term stability, the chloro-tyrosine adducted protein was stored at room temperature, 4 °C, −20 °C, and −80 °C. Samples were prepared and analyzed after 0, 1, 2, 5, 15, and 30 days. At room temperature and 4 °C, samples were stable for 5 and 15 days, respectively. However, samples stored at lower temperatures (−20 °C and −80 °C) were stable for at least 30 days (i.e., the most extended period tested). Based on the aggregate results of the stability studies, we suggest storage of samples at −20 °C or −80 °C. Once thawed, samples should be stable during processing and on an autosampler (at 15 °C) for at least 24 h each.

3.6. Verification of chlorine exposure via 2,6-DCP analysis

The ability of the method to verify exposure to chlorine was confirmed by the detection of 2,6-DCP from base-hydrolyzed plasma proteins of Cl₂-exposed rats compared to non-exposed (naïve) rats. Fig 4 shows the representation of 2,6-DCP chromatograms of these two groups of rats (A) and the mean 2,6-DCP produced (B). Fig 4 B shows the mean amount of 2,6-DCP generated from hydrolysis of plasma protein from non-exposed (N=5) and Cl₂ exposed (600 ppm inhaled Cl₂) rats (N=8). The dose of chlorine was modelled after a large-dose acute exposure which may be produced in a mass-casualty situation. For this study, 2,6-DCP was detected over the concentration range of 0.15 to 0.97 mg/kg (0.14 to 0.9 μM) from the exposed rats and between 0.023 to 0.038 mg/kg (0.021 to 0.035 μM) for the non-exposed rats. It was expected that 2,6-DCP would be detected in naïve rats since endogenous levels of chlorinated protein adducts in plasma protein are known to be present from natural sources (i.e., natural sources leading to chlorinated tyrosine adducts include diet and oxidative environments such as those suffering from inflammatory disease) [60, 61]. While endogenous levels of 2,6-DCP were detected in the non-exposed rats, the maximum concentration detected, 0.038 mg/kg (0.035 μM), was well below the minimum 2,6-DCP detected for exposed rats. Differences in the aggregated groups (non-exposed and 600 ppm) were evaluated using a one-way t-test analysis. The p-value obtained was < 0.0040, indicating a significant difference between the groups. The relationship shown in Fig 4 is very promising for using 2,6-DCP as a biomarker to confirm chlorine exposure utilizing this method.

Fig 4.

UHPLC-MS/MS chromatograms of 2,6-DCP from the plasma of Cl₂-exposed and non-exposed rats are presented (A). The quantification transition (160.9 → 125 m/z) is plotted. Correlation of Cl₂ dose to the concentration of 2,6-DCP from exposed rats (B). The error bars represent the standard error of the mean, ** = p<0.01 (N=5 for 0 ppm, and N=8 for 600 ppm).

4.0. Conclusions

A novel marker of chlorinated tyrosine protein adducts, 2,6-DCP, was discovered on base hydrolysis of these adducts. As base hydrolysis requires comparatively low heat and reaction time (~75 °C for ~ 2 h), which is good for heat or acid sensitive compounds. A comparatively simple and rapid UHPLC–MS/MS method with excellent sensitivity to 2,6-DCP was successfully developed to verify chlorine exposure. The detection of chlorophenol in plasma protein isolated from chlorine-exposed rats indicates that 2,6-DCP is a promising biomarker of chlorine exposure.

Highlights:

• Analysis methods for determination of chlorine exposure have significant drawbacks.

• Hypochlorous acid generated from chlorine forms chlorotyrosine protein adducts.

• Base hydrolysis converts chlorotyrosine adducts into chlorophenols.

• The developed method is simple and sensitive compared to previous studies.

• Elevated levels of chlorophenol from rats exposed to chlorine were detected.