An analysis of nasal irritation thresholds using a new solvation equation.

In the present paper we have developed a quantitative structure-activity relationship (QSAR) equation for nasal pungency caused by nonreactive volatile organic compounds (VOCs). Our QSAR was developed upon previously published nasal pungency thresholds in anosmics, i.e., patients lacking a sense of smell and thus responding only to sensory irritation evoked by trigeminal nerve stimulation. The reported solvation equation, which fits the data with considerable precision, describes sensory potency in terms of interaction via electron pairs, dipolarity/polarizability, hydrogen bond acidity and basicity, and hydrophobicity. It correspondingly suggests relevant physicochemical properties of the biophase where the sensory response is brought about. The equation implies that in the range of molecular size where nonreactive VOCs can produce any pungency, transport from the air to the biophase strictly determines potency. In this respect, the potency of nasal pungency shares characteristics with the ability of VOCs to cause narcosis and anesthesia.

Sensory irritation mechanisms investigated from model compounds: trifluoroethanol, hexafluoroisopropanol and methyl hexafluoroisopropyl ether.

Quantitative structure-activity relationships (QSAR) have suggested the importance of hydrogen bonding in relation to activation of the sensory irritant receptor by nonreactive volatile organic chemicals. To investigate this possibility further, three model compounds with different hydrogen bond acidity, trifluoroethanol, hexafluoroisopropanol and methyl hexafluoroisopropyl ether, were selected for study. The potency of each chemical is obtained from the concentration necessary to reduce respiratory rate in mice by 50% (RD50). The RD50 values obtained were: methyl hexafluoroisopropyl ether (> or = 160,000 ppm), trifluoroethanol (11,400-23,300 ppm), and hexafluoroisopropanol (165 ppm). QSAR showed that trifluoroethanol and methyl hexafluoroisopropyl ether behaved as predicted as nonreactive sensory irritants, whereas hexafluoroisopropanol was much more potent than predicted. The higher than predicted potency of hexafluoroisopropanol could be due to a coupled reaction, involving both strong hydrogen bonding and weak Brönsted acidity. A concerted reaction could thus be more efficient in activation of the receptor. Hydrogen bonding properties and concerted reactions may be important in the activation of the sensory irritant receptor by nonreactive volatile organic chemicals.

Physicochemical properties of nonreactive volatile organic chemicals to estimate RD50: alternatives to animal studies.

This article presents the correlations obtained between the results on the potency of nonreactive airborne chemicals as sensory irritants and several of their physicochemical properties. The potency of airborne sensory irritants obtained from a reflexively induced decrease in respiratory frequency has been measured in the past using mice. Typically, their potency has been expressed as the exposure concentration necessary to decrease respiratory frequency by 50% (RD50). A large database of RD50 values is now available and such values are highly correlated with occupational exposure guidelines such as threshold limit values (TLVs). We used the nonreactive volatile organic chemicals from this database, for which relevant physicochemical variables are available or can be calculated. These variables were vapor pressure (P) or Ostwald gas-liquid partition coefficients (L). The liquids used for L values were n-hexadecane, octanol, N-formylmorpholine, tri-(2-ethylhexyl)phosphate, and olive oil. Excellent correlations were found between log RD50 and log P, as well as between log RD50 and log L16, log L(Oct), log L(NFM), log L(EHP), or log L(Oil). It follows that as an alternative to the bioassay, these physicochemical variables can be used to estimate RD50 of nonreactive volatile organic chemicals. Appropriate exceptions to general estimation of RD50 values from physicochemical variables are also presented, as well as the most appropriate estimates which can be obtained within homologous series.