Potassium tetrafluoroaluminate

Endpoint summary

Additional information

No genotoxicity studies are available on potassium tetrafluoroaluminate. However, Article 13 of the REACH legislation states that, in case no appropriate animal studies are available for assessment, information should be generated whenever possible by means other than vertebrate animal tests, i. e., applying alternative methods such as in vitro tests, QSARs, grouping and read-across.

In vitro genotoxicity data are available on a structural analogue, multiconstituent aluminium potassium fluoride. Multiconstituent aluminium potassium fluoride is a multi-constituent substance containing ca. 70% of KAlF₄ and ca. 30% of higher homological penta- and hexafluoroaluminic acids, K₂AlF₅ and K₃AlF₆.The presence of these higher homological salts is, however, not expected to alter significantly the physico-chemical and toxicological properties of multiconstituent aluminium potassium fluoride in comparison to potassium tetrafluoroaluminate.

In vivo genotoxicity data on multiconstituent aluminium potassium fluoride are lacking, but in vivo studies are available for another structural analogue, cryolite (Na₃AlF₆). Systemic effects of both substances are expected to be primarily governed by the presence of fluoride anions formed upon dissociation of fluoroaluminate moieties. The contents of fluoride, expressed as a percentage of molar weight, are almost identical for potassium tetrafluoroaluminate and cryolite. Taking into account the structural similarity of the two substances and their comparable toxicological profiles, it is considered acceptable to derive information on in vivo genotoxicity of potassium tetrafluoroaluminate by read-across from cryolite.

In vitro genotoxicity

Multiconstituent aluminium potassium fluoride was examined for mutagenic activity in the Ames test (OECD guideline 471 and GLP compliant) using Salmonella typhimurium strains TA 1535, TA 1537, TA 98, TA 100 and the tryptophan-requiring Escherichia coli strain WP2 uvrA, in the absence and presence of metabolic activation (TNO, 2010a). One bacterial reverse mutation test was performed. All strains were treated with five concentrations of the test substance, ranging from 62 to 5000 µg/plate. Negative controls and positive controls were run simultaneously with the test substance. The mean number of his+ and trp+ revertant colonies of the negative controls were within the acceptable range and the positive controls gave the expected increase in the mean number of revertant colonies. The test substance was not toxic to any strain, in both the absence and presence of S9-mix, as neither a decrease in the mean number of revertants nor a clearing of the background lawn of bacterial growth compared to the negative controls was observed. In both the absence and presence of S9-mix in all strains, multiconstituent aluminium potassium fluoride did not induce a minimal 2-fold and/or dose related increase in the mean number of revertant colonies compared to the background spontaneous reversion rate observed with the negative control. 

In a study according to OECD guideline 476 and in compliance with GLP, multiconstituent aluminium potassium fluoride was examined for its potential to induce gene mutations at the TK-locus of cultured mouse lymphoma L5178Y cells (TNO, 2010b). One assay was conducted in which 7 duplicate cultures were treated for 24 hours and 4 hours in the absence and presence of S9-mix, respectively. The test substance was suspended in dimethyl sulfoxide (DMSO) prior to testing. The highest concentration of multiconstituent aluminium potassium fluoride tested and evaluated for mutagenicity was 10 mmol/L in both the absence and presence of S9-mix. Multiconstituent aluminium potassium fluoride was cytotoxic in both the absence and presence of S9-mix. In the absence of S9-mix cytotoxicity resulted in a reduction in initial cell yield and suspension growth at and above 1.2 mmol/L. The relative total growth (RTG) value at the highest concentration tested and evaluated for mutagenicity (10 mmol/L) was 29% (mean of duplicate cultures).In the presence of S9-mix cytotoxicity was observed at and above 2.5 mmol/L; the RTG at the highest concentration tested and evaluated for mutagenicity (10 mmol/L) was 61% (mean of duplicate cultures). In both the absence and presence of S9 -mix no increase in mutant frequency was observed at any test substance concentration evaluated. Treatment with the positive controls yielded the expected significant increases in mutant frequency compared to the negative controls.

Multiconstituent aluminium potassium fluoride was also tested in an in vitro micronucleus test (OECD guideline 487 and GLP compliant) with cultured human lymphocytes (TNO, 2010c). Two separate in vitro micronucleus tests were conducted for which blood was obtained from two different donors. Dimethylsulfoxide (DMSO) was used as solvent for the test substance. Dose levels, ranging from 3.13 to 1500 µg/ml were tested in the culture medium. In the first test, in the presence and absence of metabolic activation (S9-mix) the treatment / recovery times was 4/20 hours. In the second test, concentration spacing was modified. In the presence and absence of S9-mix, the treatment / recovery times were 4/20 hours (pulse treatment). In the continuous treatment group, the treatment / recovery times were 20/28 hours. In the first test, in the presence of S9-mix, the two highest dose levels (800 and 1500 µg/ml) were slightly cytotoxic to the cells. The highest dose level (1500 µg/ml) induced a marginal higher increase (p=0.0917) in the number of binucleated cells containing micronuclei when compared to the concurrent negative control. In the first test, in the absence of S9-mix, the test substance was not cytotoxic to the cells at any of the dose levels analysed. The highest dose level (1500 µg/ml) induced a statistically significant increase (p<0.05) in the number of binucleated cells containing micronuclei when compared to the concurrent negative control. In the second test, in the presence of S9-mix (pulse treatment), the highest dose level (1500 µg/ml) was slightly cytotoxic to the cells. The test substance did not show a statistically significant increase in the number of binucleated cells containing micronuclei. In the pulse treatment group, in the absence of S9-mix (pulse treatment), the highest three dose levels (1000, 1200 and 1500 µg/ml) were slightly cytotoxic to the cells. The test substance did not show a statistically significant increase in the number of binucleated cells containing micronuclei. In the continuous treatment group, in the absence of S9-mix, all dose levels used were slightly cytotoxic to the cells with the exception of the lowest dose level. The test substance induced a clear dose-related and statistically significant increase (p<0.001) in the number of binucleated cells containing micronuclei, at all dose levels (200, 800 and 1500 µg/ml) analysed.In the first and second test, the negative controls and positive controls yielded the expected results. This demonstrates the validity of the study. It is concluded that, under the conditions used in this study, the test substance multiconstituent aluminium potassium fluoride was clastogenic and/or aneugenic to cultured human lymphocytes.

In vivo genotoxicity

In accordance with Column 2 of REACH Annex 8, appropriate in vivo genotoxicity studies should be considered in case of a positive result observed in any of in vitro assay. However, Article 13 of REACH states also that lacking information should be generated whenever possible by means other than vertebrate animal tests, i.e. applying alternative methods such as in vitro tests, QSARs, grouping and read-across. In vivo data addressing clastogenicity are available for a structural analogue of potassium tetrafluoroaluminate, cryolite. The two substances are structural analogues of each other, being complex salts of two homological fluoroaluminic acids and differing primarily in the alkali metal cation, namely potassium vs. sodium. Their systemic effects are expected to be mostly governed by the presence of fluoride anions formed upon dissociation of the fluoroaluminate moieties. As sodium and potassium cations are essential constituents and two of the most abundant ions in all humans, as well as in all animal species, and are not genotoxic in vivo, it is considered acceptable to derive the lacking data on in vivo genotoxicity of potassium tetrafluoroaluminate by read-across from cryolite.

Two in vivo cytogenicity tests were available for cryolite. One bone marrow chromosome aberration test, in which male Crl:CD BR Sprague-Dawley rats were exposed by inhalation to 4.6 mg/m³cryolite for 13 weeks, was run within the 13-week study on repeated dose toxicity (Bayer AG, 1997a). Increased inorganic fluoride concentrations in urine, bones, and teeth were evident, indicating that the test substance has reached the bone marrow. No increase in chromosome aberrations was observed. Furthermore, no evidence of clastogenicity was observed in the other in vivo bone marrow chromosomal aberration test, in which male and female Crl:CD BR Sprague-Dawley rats were exposed by snout-only inhalation to 2130 mg/m³ cryolite for 6 hours (Huntingdon Life Sciences Ltd., 1997 / Bayer AG, 1997b). Bone marrow cells were sampled after recovery periods of 16, 24 and 48 hours. No clinical signs or mortalities were induced. All experimental parts were run in compliance with GLP and according to OECD guideline 475.

Discussion

In summary, available in vitro studies with the read-across candidate multiconstituent aluminium potassium fluoride gave contradictory results. The substance was negative in the Ames test with S. typhimurium strains TA 1535, TA 1537, TA 98, TA 100 and E. coli strain WP2 uvrA at dose levels up to 5000 µg/plate, with and without metabolic activation. The substance also did not induce an increase in gene mutation frequency at the TK locus in the mouse lymphoma assay at limit concentrations of 10 mM and increase in the number of micronuclei in human lymphocytes at dose levels up to 1500 µg/ml using pulse treatment with 4 hr exposure duration (both experiments were performed with and without S9 mix). However, in a continuous treatment experiment using longer exposure duration times of 20 hr, multiconstituent aluminium potassium fluoride was found to induce statistically significant increase in the number of micronuclei at concentrations 200 µg/ml and above.

At the same time, in vivo studies on another structural analogue cryolite were negative. Cryolite gave negative results in two in vivo bone marrow chromosomal aberration tests with rats exposed by inhalation. One of the studies, in which rats were exposed to 4.6 mg/m³cryolite in air, was performed as part of a 90-day inhalation toxicity study. In another study a single 6 h exposure to 2130 mg/m³cryolite was used and bone marrow cells were sampled after recovery periods of 16, 24 and 48 hours.

Positive results in in vitro experiments have also been reported for other fluoride-containing substances. For example, sodium fluoride has been shown to induce sister chromatid exchanges, unscheduled DNA synthesis and chromosomal aberrations at dose levels as low as 4.5 µg/ml and above. However, an in vivo assay with sodium fluoride gave negative results regarding chromosomal damage in mice. The European Union Risk Assessment Report of hydrogen fluoride (EU RAR 2009) concludes: “since it is unlikely that F^- binds to DNA covalently, a prerequisite for DNA adducts, the DNA damage observed in in vitro studies is probably not caused by a direct interaction of fluoride with DNA. Inorganic fluoride does not induce chromosomal damage in vivo”. This is supported by the overall conclusion of the available carcinogenicity studies with sodium fluoride (four drinking water and diet studies with rats and mice), which show that sodium fluoride is not a carcinogenic substance.

The main difference between cryolite and potassium tetrafluoroaluminate and multiconstituent aluminium potassium fluoride in terms of constituting elements is the alkali metal cation, namely sodium vs. potassium. Both these elements are essential constituents and two of the most abundant ions in all humans, as well as in all animal species. Both elements are known to be non-genotoxic in vivo (see for example OECD SIDS of sodium carbonate, 2002, and of potassium chloride, 2001), although high concentrations of potassium salts, e.g. KCl, have been known to show positive results in a range of in vitro genotoxic screening assays. It was weakly mutagenic in mouse lymphoma assays at high concentrations (above 7000 mg/ml without metabolic activation and above 4000 mg/ml with metabolic activation) and induced chromosome aberrations in Chinese Hamster Ovary (CHO) cells at concentrations approximately 5500 mg/ml and above.The OECD SIDS of potassium chloride (2001) concluded that “the action of KCl in culture seems to be an indirect effect associated with an increased osmotic pressure and concentration. Therefore, KCl do not have any direct relevance in the intact body were such concentrations can not occur.” Potassium compounds are also widely used therapeutically; usual therapeutic doses of potassium for oral solution in human adults are 1.5-3 g/day to prevent depletion, and 3-7.5 g/day for replacement. Thus there is conclusive evidence that potassium does not present a hazard to humans in terms of genotoxicity.

Overall, taking the above mentioned considerations into account, it can be concluded that the positive results in in vitro clastogenicity tests, as observed for multiconstituent aluminium potassium fluoride, have been known to occur for other fluoride- and potassium-containing substances; however, these effects are considered not to be caused by a direct interaction with DNA and thus do not present a risk to humans in vivo. This is in agreement with the data on a sodium-containing structural analogue cryolite, for which negative in vivo micronucleus assays are reported. Therefore, it is considered acceptable to derive the information on in vivo genotoxicity of potassium tetrafluoroaluminate by read-across from cryolite. As reliable negative in vivo micronucleus assays, as well as negative in vitro gene mutation studies are available for cryolite, classification for genotoxicity is not warranted in accordance to Directive 67/548/EEC and EU Classification, Labelling and Packaging of Substances and Mixtures (CLP) Regulation (EC) No. 1272/2008. Consequently, potassium tetrafluoroaluminate also does not need to be classified for genotoxicity.

In summary, the available data provide conclusive evidence that cryolite does not induce chromosome aberrations in vivo. Taking into account the structural similarity and comparable toxicological profiles of cryolite and potassium tetrafluoroaluminate, it is concluded that potassium tetrafluoroaluminate should also be considered non-clastogenic in vivo. As also no increased gene mutation frequencies were observed in the available in vitro studies with another read-across candidate multiconstituent aluminium potassium fluoride, it is concluded that potassium tetrafluoroaluminate does not induce gene mutations.

Short description of key information:
No genotoxicity data on potassium tetrafluoroaluminate are available. The read-across candidate multiconstituent aluminium potassium fluoride did not induce an increased mutation frequency in two GLP-compliant guideline tests with prokaryotes and eukaryotes, with and without metabolic activations, up to limit concentrations. In a GLP-compliant in vitro micronucleus test, performed according to OECD guideline 487, no significant increase in the number of binucleated cells containing micronuclei was observed using a pulse treatment (exposure duration 4 hr). However, in the continuous treatment test (exposure duration 20 hr) a clear dose-related and statistically significant increase (p<0.001) in the number of binucleated cells containing micronuclei was observed at dose levels of 200 μg/ml and above, with and without metabolic activation. Two negative in vivo bone marrow chromosome aberrations assays are available for another structural analogue of potassium tetrafluoroaluminate, cryolite. Consequently, it is concluded that potassium tetrafluoroaluminate is not genotoxic in vivo.

Endpoint Conclusion: No adverse effect observed (negative)

Justification for classification or non-classification

Based on the available data on the read-across candidates multiconstituent aluminium potassium fluoride and cryolite, and in accordance with Directive 67/548/EEC and EU Classification, Labelling and Packaging of Substances and Mixtures (CLP) Regulation (EC) No. 1272/2008, classification of potassium tetrafluoroaluminate is not necessary for genotoxicity.