Supplemental Materials: Urinary Tetrabromobenzoic Acid (TBBA) as a Biomarker of Exposure to the Flame Retardant Mixture Firemaster® 550

Background: Firemaster® 550 (FM550) is commonly added to residential furniture to reduce its flammability. Recent toxicological evidence suggests that FM550 may be endocrine disrupting and obesogenic. Objectives: Our objectives were to develop methods to assess exposure to FM550 in human populations and to identify potential routes of exposure. Methods: Using mass spectrometry methods, we developed a method to measure 2,3,4,5-tetrabromobenzoic acid (TBBA), a urinary metabolite of the major brominated FM550 component 2-ethylhexyl-2,3,4,5-tetrabromobenzoate (TBB). The method was applied to a cohort of adult volunteers (n = 64). Participants completed questionnaires, provided urine and handwipe samples, and collected dust samples from their homes. We measured TBB and the other major brominated FM550 component, bis(2-ethylhexyl)-2,3,4,5-tetrabromophthalate (TBPH), in paired dust and handwipe samples. Results: TBBA was detected in 72.4% of urine samples. Although TBBA is a rapidly formed metabolite, analyses indicated moderate temporal reliability (interclass correlation coefficient = 0.56; 95% confidence interval: 0.46, 0.66). TBB and TBPH were detected frequently in dust samples [geometric mean (GM) = 315.1 and 364.7 ng/g, respectively] and in handwipes (GM = 31.4 and 23.4 ng, respectively). Levels of TBB and TBPH in dust were positively correlated with levels in handwipes. In addition, levels of TBB in handwipes were positively correlated with urinary TBBA. Results suggest frequent hand washing may reduce the mass of TBB on participants’ hands and reduce urinary TBBA levels. Conclusions: Cumulatively, our data indicate that exposures to FM550 are widespread and that the home environment may be an important source of exposure. Urinary TBBA provides a potentially useful biomarker of FM550 exposure for epidemiologic studies. Citation: Hoffman K, Fang M, Horman B, Patisaul HB, Garantziotis S, Birnbaum LS, Stapleton HM. 2014. Urinary tetrabromobenzoic acid (TBBA) as a biomarker of exposure to the flame retardant mixture Firemaster® 550. Environ Health Perspect 122:963–969; http://dx.doi.org/10.1289/ehp.1308028


Table of Contents Page
Dust and handwipe analyses 2 Method development of TBBA analysis in human urine 3 Figure S1. Response of surrogate standard TIBA (5 ng) in 10 mL urine extracts performed by Agilent SampliQ OPT (n = 7) and liquid-liquid extraction (n = 10) with hexane relative to 5 ng TIBA standard in pure solvent with the same volume. 4 Figure S2. Mean cumulative mass of TBBA in urine samples (ng) collected from 6 treated and 6 control adult rats. 5 Figure S3. Instructions for the collection of a house dust sample.

Dust and handwipe analyses
Handwipe and dust samples were extracted in the laboratory and analyzed for TBB and TBPH. fraction was then concentrated to approximately 1 mL using a nitrogen evaporation system and transferred to an autosampler vial (ASV) for GC/MS analysis. Dust samples (~100 mg) were first spiked with FBDE-69 as an internal standard and extracted with 10 mL of 50:50 DCM:hexane using sonication. This process was repeated three times and the combined extract (~30 mL) was concentrated using an automated nitrogen evaporation system and transferred to a 4.0 mL amber vial, stored in a -20ºC freezer. The dust extracts were cleaned using the same method as described for the handwipe samples above. To measure recovery of F-BDE-69 (Chiron, Norway) extracts were spiked with 13 C-CDE 141 (Wellington Laboratories). Analysis of laboratory blanks (n=5) and an indoor dust Standard Reference Materials (SRM 2585, NIST, Gaithersburg, MD) were employed for quality assurance and quality control.

Method development of TBBA analysis in human urine
In our previous in vitro study on TBBA (Roberts et al. 2012), 6 mL Agilent SampliQ OPT SPE columns were used to concentrate and clean microsomal samples. Therefore, SPE method was first tested for efficiency in terms of extracting TBBA in urine samples. However, a significant matrix effect, i.e., ion suppression of TBBA and TIBA was observed, which may be due to the large volume of urine (~ 10 mL) used in each sample. As shown in Figure S1, the response of TIBA in urine samples was approximately 20% of the TIBA standard in pure solvent with the same volume. A similar matrix effect was also observed with other reverse-phase SPE columns (e.g., StrataX-AW Phenomenex, data not shown). Liquid-liquid extraction has been suggested as an effective method to reduce the matrix effect, yielding rather clean extracts. In this study, hexane, dichloromethane, ethyl acetate, and MTBE were tested individually to extract TBBA from urine. Sulfuric acid was added to neutralize TBBA and facilitate the extraction, which might also help hydrolyze possible TBBA conjugates, if they existed, and minimize the matrix.
In a phosphate buffer (PBS, pH: 7.4, 0.1 M) spiked with 1, 5, and 10 ng TBBA, > 80% (n = 3 for each) of the spiked TBBA was recovered. Though other solvents might have better extraction efficacy, hexane extracts showed the least matrix and was used as the extraction solvent. The matrix effect of TBBA in urine was further investigated using a matrix spike approach. A set of urine samples from one pooled urine (in which TBBA was undetected), were spiked with three different levels of TBBA (i.e., 1, 5, and 10 ng, in triplicate) and extracted with hexane using the above-mentioned liquid-liquid extraction technique. The average recovery of the TBBA was 103%, 98% and 102% of the spiked amount, respectively. In another 10 different urine samples, ion suppression was significantly reduced and the response of TIBA in those samples was ~ 80% of the pure standard ( Figure S1), suggesting the current liquid-liquid extraction method could effectively eliminate a majority of the matrix. Since a labeled recovery standard was not available in this study, recovery of TBBA in urine was further examined by spiking 1 ng of TBBA into 10 urine samples in which no TBBA was detected. Recovery of TBBA was estimated by adding TIBA post extraction and cleanup instead of as the surrogate standard. The calculated recovery averaged ± SD of 79 ± 9 % (n = 10).
PBS blanks were run throughout the process as controls. TBBA was not detected in either field or laboratory blanks. Therefore, the method detection limit (MDL) was estimated by measuring the instrumental detection limit (IDL), which was calculated by using a signal-to-noise (S/N) ratio of 3 and a volume of 10 mL. In this study, the MDL of TBBA in urine samples was calculated to be 5 pg/mL urine. Relative to pure standard Figure S1. Response of surrogate standard TIBA (5 ng) in 10 mL urine extracts performed by Agilent SampliQ OPT (n = 7) and liquid-liquid extraction (n = 10) with hexane relative to 5 ng TIBA standard in pure solvent with the same volume. control adult rats. Exposed animals were administered a one-time dose of 1mg Firemaster 550 ® (via treat pellet) at time zero. TBBA concentrations were measure at 1, 3, 6, and 24 hours in urine sample from exposed animals and controls.