1.1.1. Nomenclature

• Chem. Abstr. Serv. Reg. No.: 126-72-7
• Chem. Abstr. Name: 2,3-Dibromo-1-propanol phosphate (3:1)
• IUPAC Systematic Name: 2,3-Dibromo-1-propanol phosphate
• Synonyms: Phosphoric acid, tris(2,3-dibromopropyl) ester; Tris

1.1.2. Structural and molecular formulae and relative molecular mass

1.1.3. Chemical and physical properties of the pure substance

• (a) Description: Viscous liquid (Budavari, 1996)
• (b) Boiling-point: 390°C (WHO, 1995)
• (c) Melting-point: 5.5°C (WHO, 1995)
• (d) Solubility: Slightly soluble in water (0.8 mg/L at 24°C); miscible with carbon tetrachloride, chloroform and dichloromethane (Verschueren, 1996; United States National Library of Medicine, 1997)
• (e) Vapour pressure: 0.03 Pa at 25°C (WHO, 1995)
• (f) Octanol/water partition coefficient (P): log P, 3.02 (WHO, 1995)
• (g) Conversion factor: mg/m³ = 28.54 × ppm

1.3.1. Occupational exposure

Occupational exposures to tris(2,3-dibromopropyl) phosphate may have occurred during its production in the textile and polyurethane foam industries (IARC, 1979).

1.3.2. Environmental occurrence

Environmental release in the past has been shown to result from textile finishing plants and laundering of the finished product (United States National Library of Medicine, 1997).

Tris(2,3-dibromopropyl) phosphate was found in the air and soil in the United States in the 1970s. None was found in samples taken from various water and soil sources in Japan at this time. General population exposures may have occurred from the use of clothing treated with the compound (WHO, 1995).

3.2.1. 2,3-Dibromo-1-propanol

Mouse: Groups of 50 male and 50 female B6C3F₁ mice, eight weeks of age, were administered skin applications of 0, 88 or 177 mg/kg bw 2,3-dibromo-1-propanol (98% pure) in 95% ethanol on five days per week for 36–39 weeks (males) or 39–42 weeks (females). The study was terminated at 36–39 weeks (males) and 39–42 weeks (females) because sera from sentinel mice housed in the same room as the study animals were found to be positive for antibodies to lymphocytic choriomeningitis virus. As shown in Table 1, increased incidences of skin papillomas, forestomach papillomas and fore-stomach carcinomas were observed in both sexes. Hepatocellular adenomas were seen in 1/50 control, 2/50 low-dose and 9/50 high-dose (p < 0.05) male mice; no data for liver were reported in females (Eustis et al., 1995).

Table 1

Increased tumour incidences in mice administered 2,3-dibromo-1-propanol by skin application.

Rat: Groups of 50 male and 50 female Fischer 344/N rats, eight weeks of age, were administered skin applications of 0, 188 or 375 mg/kg bw 2,3-dibromo-1-propanol (98% pure) in 95% ethanol on five days per week for 48–51 weeks (males) or 52–55 weeks (females). The study was terminated at 48–51 weeks for males and 52–55 weeks for females because of reduced survival of the high-dose groups and because sentinel mice housed in the same room as the rats tested positive for lymphocytic choriomeningitis virus. As shown in Table 2, there were increased incidences of skin neoplasms (all types), squamous-cell carcinomas of the skin, basal-cell tumours [not further specified] of the skin, squamous-cell carcinomas of the oral mucosa, squamous-cell papillomas of the oesophagus, squamous-cell papillomas of the forestomach, adenocarcinomas of the small intestine, adenomatous polyps of the large intestine, adenomas of the nasal mucosa, Zymbal gland adenocarcinomas and liver carcinomas (Eustis et al., 1995).

Table 2

Increased tumour incidences in rats administered 2,3-dibromo-1- propanol by skin application.

3.2.2. Bis(2,3-dibromopropyl) phosphate

Rat: Groups of 40 male and 40 female Wistar rats, five weeks of age, were administered the magnesium salt of bis(2,3-dibromopropyl) phosphate [purity not specified] mixed in the diet at concentrations of 0 (control), 80 (low-dose), 400 (mid-dose) or 2000 (high-dose) mg/kg diet (ppm) for 24 months. Oesophageal papillomas were observed in 0/40 control, 0/40 low-dose, 6/40 mid-dose (p < 0.05) and 2/40 high-dose males and in 0/40 control, 0/40 low-dose, 0/40 mid-dose and 6/40 high-dose (p < 0.05) females. Papillomas of the forestomach were seen in 0/40 control, 0/40 low-dose, 8/40 mid-dose (p < 0.05) and 17/40 high-dose (p < 0.01) males and in 0/40 control, 0/40 low-dose, 4/40 mid-dose and 20/40 high-dose (p < 0.01) females. Adenocarcinomas of the small intestine were observed in 0/40 control, 0/40 low-dose, 2/40 mid-dose and 14/40 high-dose (p < 0.01) males and in 0/40 control, 0/40 low-dose, 0/40 mid-dose and 9/40 high-dose (p < 0.01) females. Hepatocellular carcinomas were observed in 0/40 control, 1/40 low-dose, 7/40 mid-dose (p < 0.05) and 24/40 high-dose (p < 0.01) female rats (Takada et al., 1991).

4.1.1. Humans

No data were available to the Working Group.

4.1.2. Experimental systems

The excretion balance and tissue distribution of radiolabelled tris(2,3-dibromopropyl) phosphate in rats were examined by Lynn et al. (1980, 1982) and Nomeir and Matthews (1983). After intravenous administration of 1.76 mg/rat, Lynn et al. (1980) recovered 57% of the dose in the urine in five days and identified the diester bis(2,3-dibromopropyl) phosphate as a minor urinary metabolite (7.8% of urinary ¹⁴C). In further work, Lynn et al. (1982) recovered a total of 86% of the dose in the excreta (58% urine, 9% faeces, 19% as expired ¹⁴CO₂) with a further 9% in the carcass. Bile-duct-cannulated rats excreted 34% of the dose in the bile in 24 h, 20% being eliminated in the first hour after dosing. No unchanged tris(2,3-dibromopropyl) phosphate was detected in the urine, but dibromopropanol was present in addition to the diester previously reported. On high-performance liquid chromatography, numerous ¹⁴C peaks remained unidentified.

Nomeir and Matthews (1983) compared the disposition of tris(2,3-dibromo[1-¹⁴C]propyl) phosphate given intravenously and orally to rats at a dose of 1.4 mg/kg bw (2 µmol/kg bw). Absorption from the gastrointestinal tract was extensive and rapid, the tissue levels of ¹⁴C being essentially identical after both routes of administration. After 24 h, 24% of an oral dose and 17% of an intravenous dose were present in urine and 11% (oral dose), 7% (intravenous dose) in faeces, with 21% (oral dose) and 26% (intravenous dose) present in seven tissues examined. A further 20% was recovered as exhaled ¹⁴CO₂ after intravenous dosing; no data were reported on exhalation after oral dosing. The tissue distribution of ¹⁴C was widespread, with an elimination half-life of 60 h from most tissues and 91 h from liver and kidney.

Lynn et al. (1982) detected three major ¹⁴C-containing compounds in plasma and the pattern was dominated by bis(2,3-dibromopropyl) phosphate as early as 5 min after intravenous dosing. 2,3-Dibromopropanol was also detected up to 8 h after dosing. The elimination of bis(2,3-dibromopropyl) phosphate was biphasic, with half-lives of 6 and 36 h and it was detected up to five days after dosing.

Tissue distribution was examined at five time points, with separate determinations of total ¹⁴C, tris(2,3-dibromopropyl) phosphate and bis(2,3,-dibromopropyl) phosphate. The results confirmed the rapid disappearance of tris(2,3-dibromopropyl) phosphate, this being detected only at 5 and 30 min. Bis(2,3-dibromopropyl) phosphate was the major component in blood, lung, muscle and fat and had a long elimination period. At five days after dosing, there was significant retention of ¹⁴C in the kidney, this comprising various polar components with some bis(2,3-dibromopropyl) phosphate also detected. The extensive biliary excretion of tris(2,3-dibromopropyl) phosphate-related radioactivity and low faecal elimination of the radiolabel indicate that enterohepatic circulation contributes to the retention of ¹⁴C in the body (Lynn et al., 1982).

In addition to the previously reported bis(2,3-dibromopropyl) phosphate and 2,3-dibromopropanol, Nomeir and Matthews (1983) characterized four additional metabolites by mass spectrometry, that arose from further hydrolysis and dehydrobromination of the 2,3-dibromopropane moiety, namely 2-bromo-2-propenyl-2,3-dibromopropyl phosphate, bis(2-bromo-2-propenyl) phosphate, 2,3-dibromopropyl phosphate and 2-bromo-2-propenyl phosphate. All six metabolites were found in 24-h urine and 3-h bile, with bis(2-bromo-2-propenyl) phosphate and 2-bromo-2-propenyl phosphate predominating in urine, while bis(2,3-dibromopropyl) phosphate and 2-bromo-2-propenyl-2,3-dibromopropyl phosphate were the major metabolites identified in bile; 67% of urinary and 47% of biliary ¹⁴C were accounted for by a variety of unidentified metabolites.

In vitro, liver microsomes from rat, mouse, hamster and guinea-pig all activate tris(2,3-dibromopropyl) phosphate resulting in covalent binding to protein. This binding was cytochrome P450-dependent and was inhibited by glutathione. In vivo in rats, the kidney was the principal target organ for covalent binding to protein and, at high doses, to DNA. This binding was enhanced by pretreatment with polychlorinated biphenyls but not sodium phenobarbital and was partly prevented by cobalt chloride (CoCl₂) pretreatment. There was much less binding to liver protein and DNA and minimal binding to muscle (Søderlund et al., 1981, 1982a).

Nelson et al. (1984) and Søderlund et al. (1984) identified the proximate mutagenic metabolite of tris(2,3-dibromopropyl) phosphate as 2-bromoacrolein and used stable isotope techniques in microsomal incubations to show that it is formed by cytochrome P450-dependent oxidative debromination at C-3 of one of the 2,3-dibromopropyl groups followed by β-elimination to break the phosphoester bond. Søderlund et al. (1984) also showed the evolution of bromide ion release in microsomes in a glutathione-dependent reaction.

These studies have been extended by Pearson et al. (1993a,b), who showed that a number of metabolites contribute to protein binding in addition to 2-bromoacrolein. The major metabolic pathway leading to protein binding is C-2 oxidation of the 2,3-dibromopropyl groups, giving a reactive α-bromoketone which might either alkylate proteins directly or be hydrolysed to bis(2,3,-dibromopropyl) phosphate and an α-bromo-α′-hydroxyketone which could mediate the alkylation of protein.

4.2.1. Humans

In a cohort of 3579 white male chemical workers with potential exposures to brominated compounds including tris(2,3-dibromopropyl) phosphate, no significant overall or cause-specific mortality excess was detected (Wong et al., 1984).

4.2.2. Experimental systems

Tris(2,3-dibromopropyl) phosphate caused extensive acute renal tubule necrosis at doses of 175 mg/kg bw and higher in male rats. Treatment also resulted in hepatotoxicity, but this effect was less pronounced and occurred at higher doses (Søderlund et al., 1980). Administration of radioactively labelled tris(2,3-dibromopropyl) phosphate to rats as a single intraperitoneal dose of 250 mg/kg bw led to pronounced binding of radioactivity to kidney but not to liver protein 9 h later (Dybing et al., 1980). Morales and Matthews (1980) and Lynn et al. (1982) also showed that the kidney accumulated the highest rate of radioactivity after injection of [¹⁴C]tris(2,3-dibromopropyl) phosphate, compared with other organs. Dybing and Søderlund (1980) treated rats intraperitoneally with 250 mg/kg bw unlabelled tris(2,3-dibromopropyl) phosphate and determined parameters of kidney and liver toxicity 24 h later. The treatment resulted in increased plasma urea and creatinine levels. Kluwe et al. (1981) reported that tris(2,3-dibromopropyl) phosphate treatment of rodents resulted in decreased non-protein sulfhydryl content in the liver, but not in the kidney as the major target organ.

In young male Fischer 344 rats treated with 100 mg/kg bw tris(2,3-dibromopropyl) phosphate by gavage, severe tubular nephrosis was observed, starting from the corticomedullar junction and spreading to the peripheral cortex (Reznik et al., 1981). A single intraperitoneal dose of 154 mg/kg bw tris(2,3-dibromopropyl) phosphate given to male Sprague-Dawley rats caused cortical damage, significant increases in serum creatinine level and a depression of para-aminohippurate uptake in cortical slices (Elliott et al., 1982).

Cunningham et al. (1993, 1994) fed male Fischer 344 rats a diet containing 0, 50 or 100 ppm (mg/kg) tris(2,3-dibromopropyl) phosphate for 14 days. In the kidney, the treatment induced significant cell proliferation that was localized in the renal outer medulla region. The proliferation rates of the inner medulla, the cortex and the liver were not increased.

In mice, hamsters and guinea-pigs, no clear evidence of renal damage was found at doses of 500–1000 mg/kg bw tris(2,3-dibromopropyl) phosphate (Søderlund et al., 1982a), which were clearly nephrotoxic to rats. Analysis of protein binding of radiolabelled tris(2,3-dibromopropyl) phosphate showed that binding to kidney protein also was much higher in rats than in the other species investigated.

4.3.1. Humans

No data were available to the Working Group.

4.3.2. Experimental systems

In a study by Seabaugh et al. (1981), pregnant Sprague-Dawley rats received 0, 5, 25 or 125 mg/kg bw tris(2,3-dibromopropyl) phosphate by gavage on days 6–15 of gestation. Weight gain during gestation was significantly decreased in the animals treated with 125 mg/kg bw per day, but no other compound-related toxic or teratogenic effect was observed.

4.4.1. Humans

No data were available to the Working Group.

4.4.2. Experimental systems (see Table 3 for references)

Table 3

Genetic and related effects of tris(2,3-dibromopropyl) phosphate.

Tris(2,3-dibromopropyl) phosphate is mutagenic in Salmonella typhimurium in the presence of a metabolic activation system and in V79 Chinese hamster lung cells. With or without metabolic activation, it produces sister chromatid exchanges in the latter system and morphological transformation in C3H 10T½ and Syrian hamster embryo cells. It binds covalently to proteins and DNA, and causes DNA single strand breaks in mammalian cells in vitro and in vivo. It is mutagenic (in somatic and germ cells), clastogenic and recombinogenic in Drosophila melanogaster and induces bone-marrow micronuclei in mice and hamsters, liver micronuclei in rats and gene mutations in mouse kidney in vivo.

4.4.3. Mechanistic aspects

Mechanistic and metabolic studies have suggested that the genotoxicity of tris(2,3-dibromopropyl) phosphate may be mediated by its conversion to reactive metabolites, the most important of which may be 2-bromoacrolein (Nelson et al., 1984; Søderlund et al., 1984). 2-Bromoacrolein and 2,3-dibromopropanal are mutagenic in Salmonella typhimurium TA100, with or without metabolic activation, cause single-strand breaks in DNA of a rat hepatoma cell line and morphological transformation of Syrian hamster embryo cells (Gordon et al., 1985). Furthermore, 2-bromoacrolein forms adducts with DNA which block DNA replication in vitro. It also induces DNA–protein cross-links in Drosophila melanogaster (van Beerendonk, 1992, 1994c).

An equimolar dose of the metabolite bis(2,3-dibromopropyl) phosphate was markedly more nephrotoxic and led also to damage of the descending loop of Henle. Intraperitoneal injection of bis(2,3-dibromopropyl) phosphate to Sprague-Dawley rats resulted in necrosis of the renal cortex, which was less severe in female than in male rats (Elliott et al., 1983). Renal dysfunction, as indexed by serum creatinine level and in-vitro renal cortical uptake of para-aminohippurate and N-methylnicotinamide, was similar in males and females. Evidence for the role of bis(2,3-dibromopropyl) phosphate and mono(2,3-dibromopropyl) phosphate as nephrotoxic metabolites of tris(2,3-dibromopropyl) phosphate was provided by Lynn et al. (1982), Søderlund et al. (1982b) and Fukuoka et al. (1988). In isolated proximal tubule cells from rat kidney, 100 µM bis(2,3-dibromopropyl) phosphate inhibited the uptake of α-methylglucose, a parameter that the authors used to assess cytotoxicity (Boogard et al., 1989).