Environ. Sci. Technol. 2009, 43, 142–147
Analysis of Perchlorate in Human Saliva by Liquid Chromatography-Tandem Mass Spectrometry JOHN F. OLDI AND KURUNTHACHALAM KANNAN* Wadsworth Center, New York State Department of Health, and Department of Environmental Health Sciences, School of Public Health, State University of New York at Albany, Empire State Plaza, P.O. Box 509, Albany, New York 12201-0509
Received September 12, 2008. Revised manuscript received October 23, 2008. Accepted October 29, 2008.
Perchlorate is both a naturally occurring anion and the disassociated anion of manufactured perchlorate salts. Because perchlorate has the ability to block the uptake of iodide by the thyroid gland, it is considered a potent thyroid hormone disruptor in humans. Methods for the analysis of perchlorate in biological matrices are needed to enable assessment of exposures and to elucidate adverse health outcomes. This study describes a method for the analysis of perchlorate in human saliva samples, using a simple dilution and ultrafiltration technique. Quantification of perchlorate in saliva samples using isotopically labeled standards (Cl18O4) and liquid chromatography-tandem mass spectrometry (LC-MS/MS) offers great selectivity and sensitivity. Matrix effects in perchlorate analysis are compensated by spiking of saliva sampleswithanisotopicallylabeledinternalstandardforperchlorate. The LC-MS/MS calibration was found to be linear over the range from 0.01 to 50 ng/mL, for 100 µL injections (i.e., 1-5000 pg injection). Fortified blank and matrix spike recoveries were between 93% and 97%, when spiked at a 2 ng/mL level. Relative standard deviations (RSDs) of daily calibration checks and fortified blanks were e10%. The relative percent difference, in laboratory duplicate analysis of original samples, was less than 1%. The method quantitation limit (LOQ) was determined to be 0.4 ng/ mL,whichincludesasampledilutionfactor.Salivaryconcentrations of a convenience sample of 83 persons working and/or living in Albany County of New York State ranged from 0.4 to 37 ng/mL, with a mean concentration of 5.3 ng/mL. Including sample preparation steps, 25 samples can be analyzed within 8 h. This selective and rapid method for analysis of perchlorate in human saliva will enable investigators and scientists to determine the extent of anindividual’sperchlorateexposureand,potentially,thecompound’s effects on human health. Analysis of perchlorate in saliva from a population (n ) 86) with no major sources of exposures, using the method developed in this study, suggests the ubiquitous occurrence of this compound in saliva.
Introduction From the mid 1990s on, the perchlorate anion (ClO4-, MW 99) has been found in matrices associated with human exposures, including drinking water and food. Exposure to * Corresponding author phone: 518-474-0015; fax: 518-473-2895; e-mail:
[email protected]. 142
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 1, 2009
perchlorate is of concern, due to the compound’s ability to disrupt the sodium iodide symporter (NIS) function of the thyroid gland, thereby reducing the uptake of iodide by the thyroid gland and, subsequently, the production of the thyroid hormones thyroxin (T4) and triiodothyronine (T3) (1). The presence of perchlorate in our environment has a multitude of causes, and the compound appears to be ubiquitous. Perchlorate production and use in the U.S. have predominantly been in the form of an oxidizer and explosive for military and civilian applications. Perchlorate for these uses is principally produced as sodium perchlorate or ammonium perchlorate. The production of these chemicals in the U.S. to date, since the early 1950s, is estimated to be on the order of 8 × 108 lbs (4 × 108 kg), or an average production rate of 700 000 kg perchlorate per year (2). Ammonium perchlorate is used to propel intercontinental ballistic missiles, the space shuttle, and both Atlas and Minuteman missiles (3). Perchlorate salts are used in various consumer products, including fireworks, road flares, matches, and vehicle airbag inflators. Potassium perchlorate has also been used since 1954 as a clinical treatment for thyrotoxicosis, because of its potent action on the thyroid gland (1). The presence of perchlorate in the environment is of both anthropogenic originsvia improper disposal of wastes associated with production and usesand natural origin. Soils in the Attacama Desert of northern Chile, historically mined as a sodium nitrate fertilizer, have been found to contain perchlorate (4). Soils from semiarid and arid regions of the U.S. Southwest have been shown to contain detectable levels of perchlorate with no known anthropogenic source (5, 6). Natural perchlorate is theorized to be generated through atmospheric mechanisms (7). Despite the range of various sources, the improper disposal of solid-fuel oxidizers probably accounts for the bulk of perchlorate in the environment (2). Perchlorate is present in food, drinking water, natural waters, and other consumables (8-15). The U.S. Food and Drug Administration’s (FDA) Total Diet Study (TDS) of perchlorate in foods indicated its presence in a wide array of foods and beverages expected to be typical of a U.S. diet (12). The FDA study provided a measurement of perchlorate intake from food and beverages in the range from 0.08 to 0.39 µg per kilogram body weight per day (µg/kg bw/day) for the U.S. population. The National Health and Nutrition Examination Survey (NHANES) of the U.S. Centers for Disease Control and Prevention (CDC) indicated widespread occurrence of perchlorate in human urine; an estimate of total daily perchlorate dose for U.S. adults was calculated, and the 95th percentile value of 0.234 µg/kg bw/day was reported (16). Over the last several years, methods have been developed for the analysis of perchlorate at parts per billion (ppb; ng/ mL) and parts per trillion (ppt; pg/mL) levels in various matrices, both human and environmental (17-25). In an effort to assess the ubiquitous exposures reported by the FDA and the CDC and to offer a matrix that is easily and noninvasively obtained, we have used human saliva to assess the presence of and exposure to perchlorate in a group of 86 individuals. A noninvasive sampling technique is especially useful for work with children or through distance sampling, where facilities for more invasive methods (e.g., venipuncture) are not readily available. Saliva has been successfully used for the measurement of cotinine, a marker of exposure to tobacco smoke (26). Methods for measuring the concentrations of anions other than perchlorate in saliva have also been published (27-29). Knowledge of perchlorate’s presence 10.1021/es802575c CCC: $40.75
2009 American Chemical Society
Published on Web 12/03/2008
in our environment and development of methods to assess exposure are of importance to national and international environmental agencies and public health organizations as they make decisions regarding the regulation of perchlorate (30).
Materials and Methods Native perchlorate (Cl16O4-) standard was obtained from Environmental Resource Associates (Arvada, CO). Ammonium perchlorate (99.999%) and methylamine (40 wt. percent solution in water) were obtained from Sigma Aldrich (St. Louis, MO). Stable isotope-labeled perchloric acid sodium salt (18O4 > 90%) was obtained from Cambridge Isotope Laboratories, Inc. (Andover, MA). The Salivette saliva sampling device was obtained from Sarstedt (Numbrecht, Germany), and Vivaspin 2 centrifugal filtration devices (CFDs) were obtained from Sartorius Stedim Biotech (Goettingen, Germany). Equipment. Sample injection was performed using a Gilson 215 Liquid Handler (Gilson, Middleton, WI) and Gilson 819 Injection Module (Gilson) equipped with a 100 µL injection loop. Pumping of isocratic mobile-phase solution was done with an Agilent 1100 HPLC system (Agilent Technologies, Santa Clara, CA). A Dionex IonPac AS-21 (250 mm × 2 mm) (Dionex, Sunnyvale, CA) ion exchange column was used for chromatographic separation. The mass spectrometric system used was a Micromass Quattro LC (Waters Corp., Milford, MA) with an electrospray ionization interface. Control of instrumentation and data acquisition were accomplished with Micromass MassLynx v. 3.5 software (Waters Corp.). Reagents and Standards. At least weekly, a new batch of mobile phase solution was made from 40 wt % methylamine solution. Batches were made in class A volumetric flasks and were then transferred to glass bottles for use with the liquid handling system described above. The concentration of methylamine mobile phase solution was 200 mM. Perchlorate calibration standards were prepared from Environmental Resource Associates source standard, supplied at 100 mg/L. Dilutions of the source standard were used to construct a nine-point calibration curve for a range from 0.01 to 50 ng/ mL. Stable isotope-labeled perchlorate was included in all calibration standards at 2 ng/mL. Calibration standards were prepared by pipetting of the standards into 15-mL polypropylene (PP) tubes (BD, Franklin Lakes, NJ) and were stored at room temperature. An Eppendorf (Hamburg, Germany) pipet (1000 µL), Thermo Fisher Scientific (Waltham, MA) pipettes (200, 20 µL), or 10 mL glass pipettes were used to dispense standards and dilution water to PP tubes. Standards were delivered to the Salivette sampling device and the Vivaspin 2 CFD in the same manner as were the saliva samples (described below) prior to analysis for the measurement of recoveries and instrument calibration. The deionized water used for standard preparation, mobile-phase preparation, and dilutions was generated with a NANOpure Diamond ultrapure water system (Barnstead International, Dubuque, IA) having a resistance of 18.2 MΩ cm. Perchlorate was not detected in the deionized water. Sample Collection. The New York State Department of Health (NYSDOH) Institutional Review Board (IRB) approved the study protocol for the collection of saliva from adult volunteers. A convenience sample composed mainly of employees and students working and studying in the Wadsworth Center, NYSDOH, were asked to donate saliva for the study. A total of 87 adults provided a sample for the study (one sample did not have sufficient volume for analysis, leaving 86 samples for analysis). After signing the informed consent document, volunteers were asked to rinse out their mouths with bottled water (Aquafina, Pepsico) a minimum of three times. After spitting out rinsewater (the rinsewater
contained very trace (
