The effect of aluminium and sodium impurities on the in vitro toxicity and pro-inflammatory potential of cristobalite

C. Nattrass, C.J. Horwell, D.E. Damby, D. Brown, V. Stone

Exposure to crystalline silica (SiO2), in the form of quartz, tridymite or cristobalite, can cause respiratory diseases, such as silicosis. However, the observed toxicity and pathogenicity of crystalline silica is highly variable. This has been attributed to a number of inherent and external factors, including the presence of impurities. In cristobalite-rich dusts, substitutions of aluminium (Al) for silicon (Si) in the cristobalite structure, and impurities occluding the silica surface, have been hypothesised to decrease its toxicity. This hypothesis is tested here through the characterisation and in vitro toxicological study of synthesised cristobalite with incremental amounts of Al and sodium (Na) dopants. Samples of synthetic cristobalite with incremental amounts of Al and Na impurities, and tridymite, were produced through heating of a silica sol-gel. Samples were characterised for mineralogy, cristobalite purity and abundance, particle size, surface area and surface charge. In vitro assays assessed the ability of the samples to induce cytotoxicity and TNF-α production in J774 macrophages, and haemolysis of red blood cells. Al-only doped or Al+Na co-doped cristobalite contained between 1 and 4 oxide wt% Al and Na within its structure. Co-doped samples also contained Al- and Na-rich phases, such as albite. Doping reduced cytotoxicity to J774 macrophages and haemolytic capacity compared to non-doped samples. Al-only doping was more effective at decreasing cristobalite reactivity than Al+Na co-doping. The reduction in the reactivity of cristobalite is attributed to both structural impurities and a lower abundance of crystalline silica in doped samples. Neither non-doped nor doped crystalline silica induced production of the pro-inflammatory cytokine TNF-α in J774 macrophages. Impurities can reduce the toxic potential of cristobalite and may help explain the low reactivity of some cristobalite-rich dusts. Whilst further work is required to determine if these effects translate to altered pathogenesis, the results have potential implications for the regulation of crystalline silica exposures.