Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Draize Rabbit Eye Test Compatibility with Eye Irritation Thresholds in Humans: A Quantitative Structure-Activity Relationship Analysis

Draize Rabbit Eye Test Compatibility with Eye Irritation Thresholds in Humans: A Quantitative... Abstract Draize rabbit eye test scores, as modified maximum average score (MMAS), for 68 pure bulk liquids were adjusted by the liquid-saturated vapor pressure P°. These 68 adjusted scores, as log (MMAS/P°), were shown to be completely equivalent to eye irritation thresholds (EIT), expressed as log (1/EIT), for 23 compounds in humans. Thus, for the first time the Draize eye test in rabbits for pure bulk liquids is shown to be perfectly compatible with eye irritation thresholds in humans. The total data set for 91 compounds was analyzed by the general solvation equation of Abraham. Values of log (MMAS/P°) or log (1/EIT) could be fitted to a five-parameter equation with R2 = 0.936, SD = 0.433, AD = 0.000, and AAD = 0.340 over a range of 9.6 log units. When divided into a training set of 45 compounds, the corresponding equation could be used to predict the remaining 46 compounds in a test set with AD = −0.037 and AAD = 0.345 log units. Thus, the 91-compound equation can now be used to predict further EIT values to around 0.4 log units. It is suggested that the mechanism of action in the Draize test and in the human EIT involves passive transfer of the compound to a biophase that is quite polar, is a strong hydrogen bond base, a moderate hydrogen bond acid, and quite hydrophobic. The biophase does not resemble water or plasma, but resembles an organic solvent such as N-methylformamide. The Draize rabbit eye test (Draize et al., 1944) is the only widely used assay for the effect of substances on the eye. In view of the scientific, ethical, and economic concerns over the Draize test (Wilhelmus, 2001), it is not surprising that alternatives to the Draize test have been examined and that various calculation procedures have been published. An in-depth study (Brantom et al., 1997) of numerous alternative assays has been carried out, but the conclusion was that none of them could be regarded as a valid replacement for the Draize test. On the other hand, it has been suggested (Spielmann et al., 1998) that a combination of two in vitro tests could be used to identify severe irritants. One challenge in finding alternatives to the Draize test is that the available data cover compounds in a variety of physical forms (i.e., liquids, solids, and aqueous solutions). The actual mechanism of irritation may well not be the same across these forms, and this would preclude any general alternative test or any general calculation. Fragmentation schemes for particular chemical or biological effects attempt to relate the effect to structural fragments of molecules. These may be functional groups or just parts of molecules, such as the CH3 or CH2 fragments. Then an effect is assigned to each fragment, and predictions are made by summation of the fragment effects in a given molecule. Such schemes for the estimation of eye irritation have been reported (Enslein, 1988; Klopman et al., 1993), but most of the data used by Enslein were not Draize scores. Although Klopman et al. (1993) used Draize scores, these were used in conjunction with other judgments; unfortunately, the full list of compounds studied is not available. Other workers have restricted their analyses to pure organic compounds. Principal components analysis (PCA) and neural networks (Barratt, 1995, 1997; Chamberlain and Barratt, 1995) have been used to discriminate between irritants and nonirritants with reasonable success. On the other hand, investigation of a similar data set using linear combinations of descriptors and PCA (Cronin et al., 1994) failed to generate any general linear correlation of modified Draize scores and failed to observe any marked distinction between irritants and nonirritants by PCA. The modified Draize scores were defined as MMAS divided by the molarity of the pure liquid; the latter is given by 1000 times the density of the pure liquid divided by the liquid molecular weight. The descriptors of the compounds in the best linear equation were ClogP, where P is a calculated water-octanol partition coefficient, the lowest unoccupied molecular orbital (LUMO), and a connectivity index. Cronin et al. (1994) correctly pointed out that use of a physically heterogeneous set of compounds, i.e., pure liquids, solids, and aqueous solutions, would make it very difficult to obtain any useful structure-activity relationship (SAR) and so restricted the analysis to pure bulk liquids. Kulkarni and Hopfinger (1999) obtained a reasonable relationship, but only for a very limited set of 18 compounds in a training set and five in a test set. Patlewicz et al. (2000) restricted their analysis to cationic surfactants, and for this set of compounds found a very good fit of observed and calculated Draize eye scores using a neural network. What is surprising is that such studies have been made before any substantial connection between results of the Draize test in rabbits and the effect of the corresponding substances in man has been established. In a comprehensive review of the Draize test, it was noted that the anatomy and biochemistry of the rabbit eye are not the same as those of the human eye and that there were numerous physiological reasons, including low tear production, blink frequency, and ocular surface area, that such a test on rabbits might not adequately predict human effects (Wilhelmus, 2001). York and Steiling (1998) stressed the need to validate the Draize test against controlled human eye data, but noted that “there are no adequate human data.” What comparisons have been made between the effects on rabbits and the effects on humans have been confined to consumer products that are a mixture of various chemicals. Freeburg et al. (1986) examined four such products and showed that the low-volume Draize test correlated with effects on the eyes of humans better than did the normal-volume Draize test. Allgood (1989) also matched the low-volume Draize test against human experience for four shampoos, and Griffith (1989) compared Draize data to consumer eye accident data for soaps and detergents. Roggeband et al. (2000) studied the effect of very low volumes (1–3 μl) of a liquid detergent and a dishwashing liquid on the eyes of rabbits and human volunteers. They observed that the irritation responses in rabbits were greater than those in man, and suggested that the low-volume Draize test could be used to assess eye irritation hazards in man. One problem until recently has been the lack of controlled human eye data (York and Steiling, 1998), but this has been rectified by the determination of eye irritation thresholds (EIT) in humans by a rigorous standardized procedure (Cometto-Muñiz and Cain, 1991, 1995, 1998; Cometto-Muñiz et al., 1997, 1998b, Cometto-Muñiz et al., c). It is these EIT values that we shall use. For the substances studied in the Draize test as pure bulk liquids, no adequate connection between the effects on rabbits and the effects on humans has been established, although for a limited number of liquids an indirect connection has been indicated (Abraham et al., 1998a,b). It is the purpose of this work to use the indirect method on a large data set to establish whether or not such a connection exists, and, if successful, to use the connection to obtain a quantitative structure-activity relationship (QSAR) for EIT in man. The Draize test scores, MMAS, that we use are those for pure bulk liquids as recorded in the ECETOC manual (ECETOC, 1998). We used values for all the pure bulk liquids given, except for several high boiling liquids for which vapor pressures at 298 K were either unknown or unreliable. MATERIALS AND METHODS Comparison of Draize scores and EIT. The Draize test scores, MMAS, that we shall use (ECETOC, 1998) refer to the effect of pure bulk liquids, whereas the EIT, in ppm, are established from the effect of the vapor of liquids at some particular partial pressure. Hence, a direct comparison of MMAS and EIT is not possible. Consider the transfer of a compound from the vapor phase to a solvent phase, the equilibrium constant being defined as   \[\mathit{K\ =\ [conc\ of\ compound\ in\ solvent]/[conc\ of\ compound\ in\ vapor\ phase]}\] (1) The compound can also be transferred from the bulk liquid to the solvent phase, the equilibrium constant being just the solubility of the bulk liquid.   \[\mathit{S\ =\ [conc\ of\ compound\ in\ solvent]}\] (2) These two equilibrium constants are related through the saturated vapor pressure, P°, of the pure bulk liquid.   \[\mathit{S\ =\ P{^\circ}\ *K\ or\ log\ (S/P{^\circ})\ =\ log\ K}\] (3) This is illustrated in Figure 1A. Exactly the same relationship can be shown for the transfer to a biophase, such as a rabbit or human eye, rather than to a solvent (see Fig. 1B). The transfer from the bulk liquid to the biophase is proportional to the Draize eye score, MMAS, provided that the mechanism of the Draize test involves passive transport to the site of action. Then, with this assumption, following Equation 3 and Figure 1, the Draize scores can be converted to scores for the effect of vapors through Equation 4, where m is some constant.   \[\mathit{Log\ (MMAS/P{^\circ})\ =\ log\ K\ +\ m}\] (4) If, again, we assume passive transfer of vapors to the biophase in human eye irritation, then log (EIT) will be given by the same type of equation as follows:   \[\mathit{Log}\ (1/\mathit{EIT})\ =\ \mathit{log\ K}{^\prime}\ +\ \mathit{m}{^\prime}\] (5) We prefer to use log (1/EIT) because the greater the value the more potent the compound. Then, the combination of Equations 4 and 5 leads to Equation 6, which can be used as a starting point for any comparison of MMAS values with EIT values. Since EIT values are listed in ppm, we use P° in ppm, and at 298 K in Equation 6.   \[\mathit{Log\ (MMAS/P{^\circ})\ =\ log\ (1/EIT)\ +\ m{^{\prime\prime}}}\] (6) QSAR studies. Our procedure is to use log (MMAS/P°) or log (1/EIT) as the dependent variable, SP, and to construct QSARs through Equation 7, the general equation that we have developed (Abraham, 1993; Abraham and Al-Hussaini, 2002). In Equation 7, the independent variables are compound descriptors as follows: E is the solute excess molar refractivity, in units of (cm3 mol−1)/10; S is the solute dipolarity/polarizability; A and B are the overall or summation hydrogen bond acidity and basicity; and L is the logarithm of the gas-hexadecane partition coefficient. Our rationale in using the particular descriptors in Equation 7 was that we had already used this equation to fit and interpret a number of gas-to-solvent partitions. Since we suggest that (MMAS/P°) or (1/EIT) are related to gas-to-biophase partition (see Fig. 1), it is logical to use Equation 7, at least as a first step, in the analysis of log (MMAS/P°) or log (1/EIT). To investigate whether QSARs based on Equation 7 might be improved by the inclusion of other types of descriptors, we calculated shape descriptors using the HyperChem software (HyperChem, 2000), with conformational energy minimized using the AM1 semiempirical method. VG is the three-dimensional volume of the minimum energy conformation computed using the HyperChem QSAR option. LG is the longest length of the minimum energy conformation. In addition, we investigated a shape descriptor, DPO1, calculated from the Dragon software (Todeschini et al., 2002) after conformational energy minimization by HyperChem. DPO1 is a molecular descriptor derived from the distance distribution moments of the geometry matrix defined as the average row sum of its entries. It can be regarded as a shape descriptor that takes into account branching and the distance of atoms from a center of gravity of a molecule   \[\mathit{SP}\ =\ \mathit{c}\ +\ \mathit{e}.\mathbf{E}\ +\ \mathit{s}.\mathbf{S}\ +\ \mathit{a}.\mathbf{A}\ +\ \mathit{b}.\mathbf{B}\ +\ \mathit{l}.\mathbf{L}\] (7) The coefficients in Equation 7 and other equations are evaluated through multiple linear regression analysis. The statistics that we shall detail are the following: N, the number of data points; R2, the coefficient of variation; SD, the standard deviation; AD, the average deviation; AAD, the absolute average deviation; and F, the Fischer statistic. RESULTS AND DISCUSSION Comparison of MMAS and EIT Values The most satisfactory comparison of log (MMAS/P°) and log (1/EIT) would be a direct comparison using Equation 6. Unfortunately, log (MMAS/P°) and log (1/EIT) values are only available for nine common compounds in Table 1. For these nine compounds, log (MMAS/P°) = log (1/EIT) + 0.89 with AAD = 0.57; but, this is not at all sufficient to demonstrate that MMAS is related to EIT. A much better method, which uses the information on log (MMAS/P°) and log (1/EIT) for all the compounds in Table 1, is to combine the two sets of data and to determine whether or not the same QSAR will fit the two sets. A straightforward method is to include a new descriptor, I, in Equation 7 that takes the value I = 0 for the log (1/EIT) series of compounds and I = 1 for the log (MMAS/P°) series of compounds. This I descriptor takes into account data on all 91 compounds and, within the statistics of Equation 8 below, shows that for any compound there is a constant difference between log (MMAS/P°) and log (1/EIT). In Equation 8, this difference is 0.568 ± 0.106 log units, which is in agreement with the difference of 0.89 ± 0.57 found for nine compounds only. For a total of 91 compounds, we constructed Equation 8, where SP = log (MMAS/P°) or log (1/EIT). There was only one outlier that we omitted, namely dodecane.   \[\mathit{SP}\ =\ {-}7.892\ {-}\ 0.379\mathbf{E}\ +\ 1.872\mathbf{S}\ +\ 3.776\mathbf{A}\ +\ 1.169\mathbf{B}\ +\ 0.785\mathbf{L}\ +\ 0.568\mathbf{I}\] (8) The following statistics are applicable: N = 91, R2 = 0.936, SD = 0.433, AD = 0.000, AAD = 0.340, and F = 204.5. The 23 points for log (1/EIT) all fall on the line of identity, as shown in Figure 2. If the deviations from the line for these 23 points are calculated separately, then AAD = 0.371 and AD = 0.000, the latter showing that there is no bias whatsoever in the fitting of the 23 log (1/EIT) points to Equation 8. We can conclude that for the 68 values of log (MMAS/P°) and the 23 values of log (1/EIT) the two sets of data are compatible. Since the two sets both cover a range of variation of compound type, we suggest that this finding is quite general. The combined equation, Equation 8, includes data on aliphatic and aromatic hydrocarbons and halogenated hydrocarbons, alcohols, ketones, acids, esters, nitro compounds, sulfides, and terpenes, as well as pyridine and 4-fluoroaniline. Therefore, we feel that this equation is of such generality that it can be used to predict EIT in man for a host of chemical vapors. Inspection of Figure 2 shows that the range of log (MMAS/P°) is much wider than that of log (1/EIT), with numerous compounds having very positive values of log (MMAS/P°). These are all very involatile liquids with vapor pressures so low that they will elicit no response in human subjects. This illustrates one advantage of including the log (MMAS/P°) data: compounds can be assigned log (1/EIT) values by this indirect method, even though log (1/EIT) cannot be determined directly. Another advantage is that compounds that cannot be studied on humans for ethical reasons can be included. Thus, benzene and the sulfides can be assigned indirect log (1/EIT) values in this way. If the two data sets are indeed compatible, then equations similar to Equation 8 should be obtained if the two sets are treated separately (although the coefficient of I must then be zero). There is not enough data on log (1/EIT) to obtain an equation with five variables, but for the log (MMAS/P°) set we find the following:   \[\mathit{Log\ (MMAS/P{^\circ})}\ =\ {-}7.355\ {-}\ 0.351\mathbf{E}\ +\ 1.997\mathbf{S}\ +\ 4.380\mathbf{A}\ +\ 1.159\mathbf{B}\ +\ 0.758\mathbf{L}\] (9) The following statistics are applicable: N = 68, R2 = 0.954, SD = 0.398, AD = 0.000, AAD = 0.318, and F = 255.8. Equation 9 is statistically the same as Equation 8: note that the SD on the coefficients in these equations averages around 0.30, except for the l coefficient where it is much smaller (0.03). Thus, on the basis that Equations 8 and 9 are statistically the same and that the data on log (1/EIT) exactly fit the general Equation 8, we conclude that the Draize eye scores, modified by the compound vapor pressure, can be combined with the EIT into one equation. Except for our preliminary report on considerably fewer compounds (Abraham et al., 1998b), this is the first time that any real connection between Draize eye scores and effects on humans has been established for the important group of pure bulk liquids. There is enough data in the two combined sets to test the predictive power of Equation 8. We ranked the data in order of SP and chose every other compound as a training set. For the 45 compounds we found the following:   \[\mathit{SP}\ =\ {-}7.856\ {-}\ 0.034\mathbf{E}\ +\ 1.529\mathbf{S}\ +\ 3.656\mathbf{A}\ +\ 1.555\mathbf{B}\ +\ 0.686\mathbf{L}\] (10) The following statistics are applicable: N = 45, R2 = 0.921, SD = 0.469, AD = 0.000, AAD = 0.354, and F = 74.0. The statistics of Equation 10 are comparable to those of the full Equation 8, bearing in mind the SD values of the various coefficients, as noted above. Then, we can use Equation 10 to predict the remaining 46 compounds in the test set that have not been used to derive Equation 10. Results are given in Table 2. Equation 10, and, hence, the full Equation 8, can predict SP values, where SP = log (MMAS/P°) and log (1/EIT) with almost no bias, since AD = −0.04, AAD = 0.35, and SD = 0.43 log unit. Considering that the 90 values cover a range of 9.7 log units and that the 23 log (1/EIT) values cover a range of 4.0 log units, we suggest that the full Equation 8 can be used to predict log (1/EIT) values to within 0.45 log units, quite generally. Interpretation of the QSAR Results The QSAR, Equation 8, has been constructed on the assumption that the Draize eye scores and the human EIT result from passive transfer from the bulk liquid and the vapor phase, respectively, to a biophase. Such passive transfer is usually nonspecific, in that the position of a substituent has little effect, unless there is some particular interaction between the substituents. We can illustrate this through the examination of water-octanol partition coefficients as log Poct (Leo, 2002; see Table 3). The position of substituents has almost no effect except for o-hydroxybenzamide where the substituents interact. The statistics of Equation 8 show that 94% of the information on log (MMAS/P°) and log (1/EIT) can be accounted for on the basis of passive transfer. There is still the possibility that the remaining 6% information applies to transfer from the biophase to some receptor. We would expect such transfer to be more specific; in particular, the shape of a compound might be crucial. To examine this point, we included the three shape descriptors in a QSAR and found the following:   \begin{eqnarray*}&&\mathit{SP}\ =\ {-}6.944\ {-}\ 1.201\mathbf{E}\ +\ 1.187\mathbf{S}\ +\ 3.268\mathbf{A}\ +\ 1.445\mathbf{B}\ +\ 1.445\mathbf{L}\ +\ 0.554\mathbf{I}\ \\&&\ +\ 0.084\mathbf{DPO1}\ {-}\ 0.007\mathbf{VG}\ +\ 0.031\mathbf{LG}\end{eqnarray*} (11) The following statistics are applicable: N = 91, R2 = 0.941, SD = 0.422, AD = 0.000, AAD = 0.334, and F = 144.0. The additional shape descriptors have almost no effect; so, we considered that the main mechanistic step in human eye irritation and in rabbit eye irritation is a simple passive transfer of a compound as a bulk liquid or as a vapor to a biophase. This partly explains the comment (Cronin et al., 1994) that construction of a QSAR for a physically heterogeneous set of compounds is very difficult. For compounds that are tested on rabbits as solids or as aqueous solutions, the passive transfer mechanism shown in Figure 1 will not hold. It has been noted (York and Steiling, 1998) that solids can cause irritancy through abrasive, mechanical effects, and that Draize scores on solutions of liquids cannot be used to assess the irritancy of the pure bulk liquids. Nevertheless, molecular size can be crucial in a fundamental aspect. Although Table 1 includes some quite large compounds such as iso-propyl iso-stearate, studies of eye and nose irritation from members of homologous series have indicated the existence of a cut-off effect in homologous series beyond which larger homologs fail to evoke irritation (Cometto-Muñiz et al., 1998c). If, as suggested (Cometto-Muñiz et al., 1998a), such cut-off rests on a biological restriction (i.e., a receptor-related effect) rather than a physical restriction (i.e., just a low vapor pressure effect), then modeling the dimensional commonalities among cut-off molecules will serve to define the maximum molecular dimensions that the receptor can fit. We are, at present, exploring this issue. Some information about the nature of the biophase can be obtained by a comparison of the coefficients in Equation 8 with those for passive transfer from the vapor to various phases that might be considered models for the biophase. Table 4 gives the coefficients of Equation 8 for transfer from the vapor to water (Abraham et al., 1994) and numerous organic solvents (Abraham et al., 2000, 2001). These coefficients are also known for transfer from the vapor to many biological tissues (Abraham and Weathersby, 1994) and from the vapor to a plant cuticular polymeric membrane matrix (Platts and Abraham, 2000). Inspection of the data in Table 4 shows that the biophase cannot resemble either water or a largely aqueous phase such as plasma. The aqueous phases have very small l coefficients (−0.213 and 0.157), whereas the biophase is relatively hydrophobic with l = 0.787, quite close to that for many organic solvents. By comparison to organic solvents, the biophase is dipolar with s = 2.02, which is almost the same as that for N-methylformamide. The biophase is a strong hydrogen bond base with a = 4.02, nearly as strong as the amides. It is also a moderately strong hydrogen bond acid, with b = 1.15, compared to 1.43 for wet 1-octanol. The nearest organic solvent as regards these chemical properties is the secondary amide, N-methylformamide. Interestingly, the plant polymeric membrane matrix also has properties that are quite similar to the biophase, but it is less polar, less acidic, and less basic. The biophase cannot be situated near an aqueous mucus layer, and it is more likely to be composed of amides (peptides) in a poorly aqueous environment. Conclusions The success of our QSAR to integrate, for bulk liquids, eye irritation data from the Draize test in rabbits, as MMAS values, with EIT in humans obtained by a standardized procedure has several very important implications. Firstly, for bulk liquids there is a proven statistical and physicochemical basis to support the Draize test as an indirect measure of EIT in humans. From this, we show that EIT in humans can be obtained by this indirect method for a large number of compounds that cannot be studied by the standardized procedure. These compounds include those that have too low a vapor pressure to elicit a response in humans, and those that cannot be studied on humans because of ethical considerations. Secondly, the resulting QSAR represents the first time that such a scale having statistical significance, chemical diversity, and physicochemical basis is available specifically for the eye irritation effect of vapors on humans. A test such as the Draize test is certainly needed for eye irritancy of bulk liquids that might come into contact with the eye. However, we feel that consideration should be given to our scale as a new, statistically sound measure of eye irritation of vapors on humans that could be used to assess the environmental impact of vapors. We stress that, although our QSAR covers a wide variety of compounds, there is still considerable scope for extending the range of compounds; this is part of an ongoing project to determine further values of EIT. The general QSAR, Equation 7, has also proved to be valuable in the analysis of odor thresholds in humans and nasal pungency (nasal irritation) thresholds in humans (Abraham et al., 2001, 2002). Further work to extend the range of compounds in these two areas is ongoing. We hope to be able to present numerous very general QSARs, not just the present one for eye irritation, that can be used for the prediction of environmental effects of vapors on humans. TABLE 1 Compounds, Values of SP, and the Compound Descriptorsa Name  SPa  I  E  S  A  B  L  VG  LG  DP01  aSP = log (MMAS/P°) when I = 1 and SP = log (1/EIT) when I = 0.  bThese are the nine common compounds in the two data sets.  2-Bromobutane  −5.16  1  0.344  0.35  0.00  0.14  2.933  381.12  5.578  2.377  1-Bromo-4-chlorobutane  −3.34  1  0.571  0.82  0.00  0.25  4.007  436.15  6.906  2.839  Dichlorotoluenes (3,4)  −2.77  1  0.900  0.80  0.00  0.03  5.089  467.09  6.453  3.173  3,3-Dimethylpentane  −5.24  1  0.000  0.00  0.00  0.00  2.946  450.50  6.771  2.803  1-Bromooctane  −2.90  1  0.339  0.40  0.00  0.12  5.143  606.11  11.300  3.574  iso-Stearyl alcohol  2.37  1  0.140  0.39  0.37  0.48  9.500  1093.42  22.807  4.997  Methylisobutylketone  −3.73  1  0.111  0.65  0.00  0.51  3.089  421.84  6.221  2.833  3-Methylhexane  −5.06  1  0.000  0.00  0.00  0.00  3.044  470.70  8.009  2.938  4-Bromophenetole  −1.89  1  0.967  0.90  0.00  0.23  5.520  529.22  9.018  3.431  Di-n-propyl disulfide  −2.75  1  0.653  0.52  0.00  0.27  4.984  555.76  11.195  3.419  Heptyl methacrylate  −2.15  1  0.445  0.49  0.00  0.45  5.697  718.89  13.840  4.114  1-Bromohexane  −3.60  1  0.349  0.40  0.00  0.12  4.138  498.51  8.813  3.101  iso-Propyl iso-stearate  2.26  1  −0.020  0.53  0.00  0.47  10.250  1259.97  25.318  5.319  1-Bromopentane  −3.92  1  0.356  0.40  0.00  0.12  3.611  444.66  7.530  2.819  1,9-Decadiene  −3.17  1  0.184  0.20  0.00  0.10  4.380  617.13  13.018  3.753  1,6-Dibromohexane  −0.91  1  0.711  0.80  0.00  0.26  5.328  563.59  9.560  3.374  1,3-Diisopropylbenzene  −2.32  1  0.605  0.46  0.00  0.20  5.170  628.84  8.868  3.666  2-Methylpentane  −5.15  1  0.000  0.00  0.00  0.00  2.503  422.99  6.820  2.669  s-Butylbenzene  −3.08  1  0.603  0.48  0.00  0.16  4.506  528.60  8.173  3.369  3-Ethyltoluene  −3.23  1  0.630  0.51  0.00  0.18  4.275  484.64  7.785  3.187  Methyl trimethylacetate  −4.16  1  0.049  0.54  0.00  0.45  2.932  448.41  6.328  2.964  2-Bromopropane  −5.02  1  0.332  0.35  0.00  0.00  2.390  332.65  4.329  2.027  1,5-Dimethylcyclooctadiene  −2.90  1  0.604  0.30  0.00  0.18  4.812  533.10  7.843  3.330  cis-Cyclooctene  −3.48  1  0.460  0.24  0.00  0.10  4.119  446.17  5.699  2.949  iso-Stearic acid  4.38  1  0.015  0.57  0.60  0.49  9.600  1093.73  22.075  5.077  Methylcyclopentane  −4.69  1  0.225  0.10  0.00  0.00  2.907  385.83  5.271  2.509  Ethyl trimethylacetate  −3.55  1  -0.010  0.52  0.00  0.45  3.481  504.71  7.638  3.201  1,4-Dibromobutane  −2.00  1  0.733  0.80  0.00  0.27  4.353  455.93  7.084  2.856  1,5-Dibromopentane  −1.41  1  0.723  0.80  0.00  0.27  4.848  509.83  8.238  3.130  1,3-Dibromopropane  −2.89  1  0.723  0.80  0.00  0.27  3.872  402.12  5.741  2.542  iso-Myristyl alcohol  1.26  1  0.155  0.39  0.37  0.48  7.480  877.65  17.809  4.517  2,4-Difluoronitrobenzene  −1.83  1  0.677  1.20  0.00  0.25  4.350  408.66  6.315  3.446  1,5-Hexadiene  −4.80  1  0.191  0.15  0.00  0.10  2.450  402.30  8.039  2.762  4-Methylpentan-2-one  −3.73  1  0.111  0.65  0.00  0.51  3.089  426.79  6.831  2.845  Allyl methacrylate  −3.24  1  0.290  0.57  0.00  0.54  3.741  484.32  8.803  3.300  Styrene  −3.11  1  0.849  0.65  0.00  0.16  3.856  416.77  7.331  2.978  Butyl acetateb  −3.30  1  0.071  0.60  0.00  0.45  3.353  470.17  8.831  3.196  2,2-Dimethylpentan-3-ol  −2.70  1  0.227  0.27  0.31  0.60  3.400  474.04  6.855  2.989  Tolueneb  −3.62  1  0.601  0.52  0.00  0.14  3.325  384.83  5.889  2.719  m-Xylene  −3.11  1  0.623  0.52  0.00  0.16  3.839  436.84  6.742  2.956  Heptan-2-oneb  −2.68  1  0.123  0.68  0.00  0.51  3.760  489.92  9.321  3.229  2-Methylpentan-1-ol  −2.26  1  0.211  0.39  0.37  0.48  3.530  445.14  7.798  2.925  3-Choroproprionitrile  −2.65  1  0.387  1.22  0.02  0.40  3.070  321.06  5.105  2.440  Cellosolve acetate  −2.20  1  0.099  0.79  0.00  0.79  3.747  504.30  9.859  3.418  Ethyl acetateb  −3.91  1  0.106  0.62  0.00  0.45  2.314  361.36  6.375  2.606  Heptan-2-oneb  −2.49  1  0.123  0.68  0.00  0.51  3.760  489.92  9.321  3.229  Ethyl 2-methylacetoacetate  −1.58  1  0.156  0.85  0.00  0.85  4.214  515.21  7.072  3.358  Cyclopentanol  −2.12  1  0.427  0.54  0.32  0.36  3.241  357.56  5.078  2.493  Ethanolb  −3.51  1  0.246  0.42  0.37  0.48  1.485  242.38  4.054  1.510  Methyl cyanoacetate  −0.78  1  0.291  1.34  0.00  0.64  3.367  353.43  6.454  2.849  Propan-2-ol  −3.27  1  0.212  0.36  0.33  0.56  1.764  293.20  4.326  1.929  Methyl acetate  −3.85  1  0.142  0.64  0.00  0.45  1.911  303.83  5.081  2.252  Octan-1-olb  −0.39  1  0.199  0.42  0.37  0.48  4.619  565.90  11.552  3.548  g-Butyrolactone  −0.94  1  0.366  1.30  0.00  0.58  3.600  314.67  4.355  2.465  Furfuryl alcohol  −1.34  1  0.554  0.73  0.50  0.63  3.357  357.58  6.245  2.751  2,2-Dimethylbutanoic acid  −0.65  1  0.186  0.55  0.60  0.51  3.681  431.19  5.838  2.960  Methoxyethyl acrylate  −2.10  1  0.249  0.80  0.00  0.80  3.876  481.45  9.204  3.357  Pyridine  −2.75  1  0.613  0.84  0.00  0.52  3.022  318.53  4.941  2.428  Butanone  −3.38  1  0.166  0.70  0.00  0.51  2.287  328.03  5.595  2.291  2-Ethylhexan-1-ol  −0.57  1  0.209  0.39  0.37  0.48  4.433  546.11  8.886  3.344  iso-Butanol  −2.36  1  0.217  0.39  0.37  0.48  2.413  342.97  5.314  2.301  Butan-1-olb  −2.13  1  0.224  0.42  0.37  0.48  2.601  350.58  6.554  2.423  Diethylaminopropriontrile  −0.83  1  0.267  0.89  0.00  0.86  4.479  515.76  7.571  3.337  Hexan-1-olb  −1.13  1  0.210  0.42  0.37  0.48  3.610  457.78  9.053  3.061  Propanoneb  −3.66  1  0.179  0.70  0.04  0.49  1.696  275.29  4.316  1.912  Ethyleneglycolmonobutyl ether  −1.32  1  0.201  0.50  0.30  0.83  3.806  491.96  10.074  3.289  4-Fluoroaniline  −1.06  1  0.760  1.09  0.20  0.40  4.007  376.72  6.105  2.950  Cyclohexanol  −1.00  1  0.460  0.54  0.32  0.57  3.758  398.55  4.889  2.704  Propanoneb  −5.27  0  0.179  0.70  0.04  0.49  1.696  275.29  4.316  1.912  Pentan-2-one  −4.05  0  0.143  0.68  0.00  0.51  2.755  382.08  6.819  2.648  Heptan-2-oneb  −2.49  0  0.123  0.68  0.00  0.51  3.760  489.92  9.321  3.229  Nonan-2-one  −2.35  0  0.119  0.68  0.00  0.51  4.735  597.48  11.813  3.686  Ethyl acetateb  −4.69  0  0.106  0.62  0.00  0.45  2.314  361.36  6.375  2.606  Butyl acetateb  −2.87  0  0.071  0.60  0.00  0.45  3.353  470.17  8.831  3.196  Hexyl acetate  −2.41  0  0.056  0.60  0.00  0.45  4.351  577.05  11.307  3.663  Octyl acetate  −2.02  0  0.029  0.60  0.00  0.45  5.364  684.18  13.795  4.043  Decyl acetate  −1.30  0  0.033  0.60  0.00  0.45  6.373  793.03  16.298  4.362  Ethanolb  −4.76  0  0.246  0.42  0.37  0.48  1.485  242.38  4.054  1.510  Propan-1-ol  −3.74  0  0.236  0.42  0.37  0.48  2.031  296.30  5.313  2.016  Butan-1-olb  −3.37  0  0.224  0.42  0.37  0.48  2.601  350.58  6.554  2.423  Hexan-1-olb  −2.60  0  0.210  0.42  0.37  0.48  3.610  457.78  9.053  3.061  Octan-1-olb  −1.71  0  0.199  0.42  0.37  0.48  4.619  565.90  11.552  3.548  Tolueneb  −4.41  0  0.601  0.52  0.00  0.14  3.325  384.83  5.889  2.719  Ethylbenzene  −3.87  0  0.613  0.51  0.00  0.15  3.778  432.04  7.284  2.981  Propylbenzene  −3.43  0  0.604  0.50  0.00  0.15  4.230  486.32  8.368  3.242  Cumene  −3.39  0  0.602  0.49  0.00  0.16  4.084  480.18  7.200  3.160  p-Cymene  −3.11  0  0.607  0.49  0.00  0.19  4.590  532.53  8.190  3.360  d-3-Carene  −3.30  0  0.511  0.22  0.00  0.10  4.649  529.88  7.470  3.280  Linalool  −2.55  0  0.398  0.55  0.20  0.67  4.794  613.77  10.000  3.680  1,8-Cineole  −2.15  0  0.383  0.33  0.00  0.76  4.688  543.92  6.400  3.330  Geraniol  −1.35  0  0.513  0.63  0.39  0.66  5.479  619.88  11.470  3.760  Dodecane  1.00  0  0.000  0.00  0.00  0.00  5.695        Name  SPa  I  E  S  A  B  L  VG  LG  DP01  aSP = log (MMAS/P°) when I = 1 and SP = log (1/EIT) when I = 0.  bThese are the nine common compounds in the two data sets.  2-Bromobutane  −5.16  1  0.344  0.35  0.00  0.14  2.933  381.12  5.578  2.377  1-Bromo-4-chlorobutane  −3.34  1  0.571  0.82  0.00  0.25  4.007  436.15  6.906  2.839  Dichlorotoluenes (3,4)  −2.77  1  0.900  0.80  0.00  0.03  5.089  467.09  6.453  3.173  3,3-Dimethylpentane  −5.24  1  0.000  0.00  0.00  0.00  2.946  450.50  6.771  2.803  1-Bromooctane  −2.90  1  0.339  0.40  0.00  0.12  5.143  606.11  11.300  3.574  iso-Stearyl alcohol  2.37  1  0.140  0.39  0.37  0.48  9.500  1093.42  22.807  4.997  Methylisobutylketone  −3.73  1  0.111  0.65  0.00  0.51  3.089  421.84  6.221  2.833  3-Methylhexane  −5.06  1  0.000  0.00  0.00  0.00  3.044  470.70  8.009  2.938  4-Bromophenetole  −1.89  1  0.967  0.90  0.00  0.23  5.520  529.22  9.018  3.431  Di-n-propyl disulfide  −2.75  1  0.653  0.52  0.00  0.27  4.984  555.76  11.195  3.419  Heptyl methacrylate  −2.15  1  0.445  0.49  0.00  0.45  5.697  718.89  13.840  4.114  1-Bromohexane  −3.60  1  0.349  0.40  0.00  0.12  4.138  498.51  8.813  3.101  iso-Propyl iso-stearate  2.26  1  −0.020  0.53  0.00  0.47  10.250  1259.97  25.318  5.319  1-Bromopentane  −3.92  1  0.356  0.40  0.00  0.12  3.611  444.66  7.530  2.819  1,9-Decadiene  −3.17  1  0.184  0.20  0.00  0.10  4.380  617.13  13.018  3.753  1,6-Dibromohexane  −0.91  1  0.711  0.80  0.00  0.26  5.328  563.59  9.560  3.374  1,3-Diisopropylbenzene  −2.32  1  0.605  0.46  0.00  0.20  5.170  628.84  8.868  3.666  2-Methylpentane  −5.15  1  0.000  0.00  0.00  0.00  2.503  422.99  6.820  2.669  s-Butylbenzene  −3.08  1  0.603  0.48  0.00  0.16  4.506  528.60  8.173  3.369  3-Ethyltoluene  −3.23  1  0.630  0.51  0.00  0.18  4.275  484.64  7.785  3.187  Methyl trimethylacetate  −4.16  1  0.049  0.54  0.00  0.45  2.932  448.41  6.328  2.964  2-Bromopropane  −5.02  1  0.332  0.35  0.00  0.00  2.390  332.65  4.329  2.027  1,5-Dimethylcyclooctadiene  −2.90  1  0.604  0.30  0.00  0.18  4.812  533.10  7.843  3.330  cis-Cyclooctene  −3.48  1  0.460  0.24  0.00  0.10  4.119  446.17  5.699  2.949  iso-Stearic acid  4.38  1  0.015  0.57  0.60  0.49  9.600  1093.73  22.075  5.077  Methylcyclopentane  −4.69  1  0.225  0.10  0.00  0.00  2.907  385.83  5.271  2.509  Ethyl trimethylacetate  −3.55  1  -0.010  0.52  0.00  0.45  3.481  504.71  7.638  3.201  1,4-Dibromobutane  −2.00  1  0.733  0.80  0.00  0.27  4.353  455.93  7.084  2.856  1,5-Dibromopentane  −1.41  1  0.723  0.80  0.00  0.27  4.848  509.83  8.238  3.130  1,3-Dibromopropane  −2.89  1  0.723  0.80  0.00  0.27  3.872  402.12  5.741  2.542  iso-Myristyl alcohol  1.26  1  0.155  0.39  0.37  0.48  7.480  877.65  17.809  4.517  2,4-Difluoronitrobenzene  −1.83  1  0.677  1.20  0.00  0.25  4.350  408.66  6.315  3.446  1,5-Hexadiene  −4.80  1  0.191  0.15  0.00  0.10  2.450  402.30  8.039  2.762  4-Methylpentan-2-one  −3.73  1  0.111  0.65  0.00  0.51  3.089  426.79  6.831  2.845  Allyl methacrylate  −3.24  1  0.290  0.57  0.00  0.54  3.741  484.32  8.803  3.300  Styrene  −3.11  1  0.849  0.65  0.00  0.16  3.856  416.77  7.331  2.978  Butyl acetateb  −3.30  1  0.071  0.60  0.00  0.45  3.353  470.17  8.831  3.196  2,2-Dimethylpentan-3-ol  −2.70  1  0.227  0.27  0.31  0.60  3.400  474.04  6.855  2.989  Tolueneb  −3.62  1  0.601  0.52  0.00  0.14  3.325  384.83  5.889  2.719  m-Xylene  −3.11  1  0.623  0.52  0.00  0.16  3.839  436.84  6.742  2.956  Heptan-2-oneb  −2.68  1  0.123  0.68  0.00  0.51  3.760  489.92  9.321  3.229  2-Methylpentan-1-ol  −2.26  1  0.211  0.39  0.37  0.48  3.530  445.14  7.798  2.925  3-Choroproprionitrile  −2.65  1  0.387  1.22  0.02  0.40  3.070  321.06  5.105  2.440  Cellosolve acetate  −2.20  1  0.099  0.79  0.00  0.79  3.747  504.30  9.859  3.418  Ethyl acetateb  −3.91  1  0.106  0.62  0.00  0.45  2.314  361.36  6.375  2.606  Heptan-2-oneb  −2.49  1  0.123  0.68  0.00  0.51  3.760  489.92  9.321  3.229  Ethyl 2-methylacetoacetate  −1.58  1  0.156  0.85  0.00  0.85  4.214  515.21  7.072  3.358  Cyclopentanol  −2.12  1  0.427  0.54  0.32  0.36  3.241  357.56  5.078  2.493  Ethanolb  −3.51  1  0.246  0.42  0.37  0.48  1.485  242.38  4.054  1.510  Methyl cyanoacetate  −0.78  1  0.291  1.34  0.00  0.64  3.367  353.43  6.454  2.849  Propan-2-ol  −3.27  1  0.212  0.36  0.33  0.56  1.764  293.20  4.326  1.929  Methyl acetate  −3.85  1  0.142  0.64  0.00  0.45  1.911  303.83  5.081  2.252  Octan-1-olb  −0.39  1  0.199  0.42  0.37  0.48  4.619  565.90  11.552  3.548  g-Butyrolactone  −0.94  1  0.366  1.30  0.00  0.58  3.600  314.67  4.355  2.465  Furfuryl alcohol  −1.34  1  0.554  0.73  0.50  0.63  3.357  357.58  6.245  2.751  2,2-Dimethylbutanoic acid  −0.65  1  0.186  0.55  0.60  0.51  3.681  431.19  5.838  2.960  Methoxyethyl acrylate  −2.10  1  0.249  0.80  0.00  0.80  3.876  481.45  9.204  3.357  Pyridine  −2.75  1  0.613  0.84  0.00  0.52  3.022  318.53  4.941  2.428  Butanone  −3.38  1  0.166  0.70  0.00  0.51  2.287  328.03  5.595  2.291  2-Ethylhexan-1-ol  −0.57  1  0.209  0.39  0.37  0.48  4.433  546.11  8.886  3.344  iso-Butanol  −2.36  1  0.217  0.39  0.37  0.48  2.413  342.97  5.314  2.301  Butan-1-olb  −2.13  1  0.224  0.42  0.37  0.48  2.601  350.58  6.554  2.423  Diethylaminopropriontrile  −0.83  1  0.267  0.89  0.00  0.86  4.479  515.76  7.571  3.337  Hexan-1-olb  −1.13  1  0.210  0.42  0.37  0.48  3.610  457.78  9.053  3.061  Propanoneb  −3.66  1  0.179  0.70  0.04  0.49  1.696  275.29  4.316  1.912  Ethyleneglycolmonobutyl ether  −1.32  1  0.201  0.50  0.30  0.83  3.806  491.96  10.074  3.289  4-Fluoroaniline  −1.06  1  0.760  1.09  0.20  0.40  4.007  376.72  6.105  2.950  Cyclohexanol  −1.00  1  0.460  0.54  0.32  0.57  3.758  398.55  4.889  2.704  Propanoneb  −5.27  0  0.179  0.70  0.04  0.49  1.696  275.29  4.316  1.912  Pentan-2-one  −4.05  0  0.143  0.68  0.00  0.51  2.755  382.08  6.819  2.648  Heptan-2-oneb  −2.49  0  0.123  0.68  0.00  0.51  3.760  489.92  9.321  3.229  Nonan-2-one  −2.35  0  0.119  0.68  0.00  0.51  4.735  597.48  11.813  3.686  Ethyl acetateb  −4.69  0  0.106  0.62  0.00  0.45  2.314  361.36  6.375  2.606  Butyl acetateb  −2.87  0  0.071  0.60  0.00  0.45  3.353  470.17  8.831  3.196  Hexyl acetate  −2.41  0  0.056  0.60  0.00  0.45  4.351  577.05  11.307  3.663  Octyl acetate  −2.02  0  0.029  0.60  0.00  0.45  5.364  684.18  13.795  4.043  Decyl acetate  −1.30  0  0.033  0.60  0.00  0.45  6.373  793.03  16.298  4.362  Ethanolb  −4.76  0  0.246  0.42  0.37  0.48  1.485  242.38  4.054  1.510  Propan-1-ol  −3.74  0  0.236  0.42  0.37  0.48  2.031  296.30  5.313  2.016  Butan-1-olb  −3.37  0  0.224  0.42  0.37  0.48  2.601  350.58  6.554  2.423  Hexan-1-olb  −2.60  0  0.210  0.42  0.37  0.48  3.610  457.78  9.053  3.061  Octan-1-olb  −1.71  0  0.199  0.42  0.37  0.48  4.619  565.90  11.552  3.548  Tolueneb  −4.41  0  0.601  0.52  0.00  0.14  3.325  384.83  5.889  2.719  Ethylbenzene  −3.87  0  0.613  0.51  0.00  0.15  3.778  432.04  7.284  2.981  Propylbenzene  −3.43  0  0.604  0.50  0.00  0.15  4.230  486.32  8.368  3.242  Cumene  −3.39  0  0.602  0.49  0.00  0.16  4.084  480.18  7.200  3.160  p-Cymene  −3.11  0  0.607  0.49  0.00  0.19  4.590  532.53  8.190  3.360  d-3-Carene  −3.30  0  0.511  0.22  0.00  0.10  4.649  529.88  7.470  3.280  Linalool  −2.55  0  0.398  0.55  0.20  0.67  4.794  613.77  10.000  3.680  1,8-Cineole  −2.15  0  0.383  0.33  0.00  0.76  4.688  543.92  6.400  3.330  Geraniol  −1.35  0  0.513  0.63  0.39  0.66  5.479  619.88  11.470  3.760  Dodecane  1.00  0  0.000  0.00  0.00  0.00  5.695        View Large TABLE 2 Prediction of SP Values for the 46 Compound Test Set from Equation 10 Statistic  Value  Average deviation (AD)  −0.037  Average absolute deviation (AAD)  0.345  Standard deviation  0.430  Statistic  Value  Average deviation (AD)  −0.037  Average absolute deviation (AAD)  0.345  Standard deviation  0.430  View Large TABLE 3 Values of log Poct for Some Isomeric Compounds (Leo, 2002) Compound  log Poct  Hexane  3.90  2-Methylpentane  3.73  2,2-Dimethylbutane  3.82  Hexan-1-ol  3.23  Hexan-2-ol  3.07  Hexan-3-ol  2.98  o-Methylphenol  1.97  m-Methylphenol  1.98  p-Methylphenol  1.97  o-Hydroxybenzamide  1.28  m-Hydroxybenzamide  0.39  p-Hydroxybenzamide  0.33  Compound  log Poct  Hexane  3.90  2-Methylpentane  3.73  2,2-Dimethylbutane  3.82  Hexan-1-ol  3.23  Hexan-2-ol  3.07  Hexan-3-ol  2.98  o-Methylphenol  1.97  m-Methylphenol  1.98  p-Methylphenol  1.97  o-Hydroxybenzamide  1.28  m-Hydroxybenzamide  0.39  p-Hydroxybenzamide  0.33  View Large TABLE 4 Regression Coefficients in Equation 7 for Gas-Solvent (Phase) Partitions at 298 K Phase  e  s  a  b  l  aAt 310 K.  Biophase, for EIT  −0.44  2.02  4.02  1.15  0.787  Water  0.82  2.73  3.90  4.81  −0.213  Wet 1-octanol  0.00  0.71  3.52  1.43  0.858  Wet chloroform  −0.47  1.20  0.14  1.43  0.994  Dry acetone  −0.27  1.52  3.26  0.08  0.863  Dry N,N′-dimethyl formamide  −0.19  2.33  4.76  0.00  0.808  Dry N-methylformamide  −0.26  2.06  4.56  0.43  0.706  Dry tetraethylene glycol  0.21  1.88  4.64  0.31  0.584  Plant matrix  0.08  1.28  3.12  0.82  0.860  Braina  0.43  0.29  2.78  2.79  0.609  Musclea  0.54  0.22  3.47  2.92  0.578  Plasmaa  0.49  2.05  3.51  3.91  0.157  Phase  e  s  a  b  l  aAt 310 K.  Biophase, for EIT  −0.44  2.02  4.02  1.15  0.787  Water  0.82  2.73  3.90  4.81  −0.213  Wet 1-octanol  0.00  0.71  3.52  1.43  0.858  Wet chloroform  −0.47  1.20  0.14  1.43  0.994  Dry acetone  −0.27  1.52  3.26  0.08  0.863  Dry N,N′-dimethyl formamide  −0.19  2.33  4.76  0.00  0.808  Dry N-methylformamide  −0.26  2.06  4.56  0.43  0.706  Dry tetraethylene glycol  0.21  1.88  4.64  0.31  0.584  Plant matrix  0.08  1.28  3.12  0.82  0.860  Braina  0.43  0.29  2.78  2.79  0.609  Musclea  0.54  0.22  3.47  2.92  0.578  Plasmaa  0.49  2.05  3.51  3.91  0.157  View Large FIG. 1. View largeDownload slide The relationship between (A) liquid to vapor and liquid to solvent transfer and (B) liquid to vapor and liquid to biophase transfer. S is the solubility of a pure liquid in a solvent, P° is the pure liquid saturated vapor pressure, K is the gas-to-solvent equilibrium constant, and MMAS is the Draize eye score. FIG. 1. View largeDownload slide The relationship between (A) liquid to vapor and liquid to solvent transfer and (B) liquid to vapor and liquid to biophase transfer. S is the solubility of a pure liquid in a solvent, P° is the pure liquid saturated vapor pressure, K is the gas-to-solvent equilibrium constant, and MMAS is the Draize eye score. FIG. 2. View largeDownload slide Plot of observed SP versus calculated SP on Equation 8; SP = log (MMAS/P°) or log (1/EIT). ▴ log (MMAS/P°); 0 log (1/EIT). FIG. 2. View largeDownload slide Plot of observed SP versus calculated SP on Equation 8; SP = log (MMAS/P°) or log (1/EIT). ▴ log (MMAS/P°); 0 log (1/EIT). 1 To whom correspondence should be addressed at University College London, Department of Chemistry, 20 Gordon Street, London WC1H 0AJ, UK. Fax: +44 (20) 7679-7463. E-mail: m.h.abraham@ucl.ac.uk. This study was supported by research grants R01 DC 02741 and DC 005003 from the National Institute on Deafness and other Communication Disorders, National Institutes of Health. M. H. thanks the British Council for a scholarship. REFERENCES Abraham, M. H. ( 1993). Scales of hydrogen bonding: Their construction and application to physicochemical and biochemical processes. Chem. Soc. Rev.  22, 73–83. Google Scholar Abraham, M. H., and Al-Hussaini, A. J. M. ( 2002). Solvation descriptors for N-nitroso dialkylamines: Calculation of some of their properties of environmental significance. J. Environ. Monit.  4, 743–746. Google Scholar Abraham, M. H., Andonian-Haftvan, J., Whiting, G. S., Leo, A., and Taft, R. W. ( 1994). Hydrogen bonding. Part 34. The factors that influence solubility of gases and vapors in water at 298 K, and a new method for its determination. J. Chem. Soc. Perkin Trans.  2, 1777–1791. Google Scholar Abraham, M. H., Gola, J. M. R., Cometto-Muñiz, J. E., and Cain, W. S ( 2000). Connection between chromatographic data and biological data. J. Chromatogr. B  745, 103–115. Google Scholar Abraham, M. H., Gola, J. M. R., Cometto-Muñiz, J. E., and Cain, W. S ( 2001). The correlation and prediction of VOC thresholds for nasal pungency, eye irritation, and odor in humans. Indoor Built Environ.  10, 252–257. Google Scholar Abraham, M. H., Gola, J. M. R., Cometto-Muñiz, J. E., and Cain, W. S. ( 2002). A model for odor thresholds. Chem. Senses  27, 95–104. Google Scholar Abraham, M. H., Kumarsingh, R., Cometto-Muñiz, J. E., and Cain, W. S. ( 1998). A quantitative structure-activity relationship (QSAR) for a Draize eye irritation data base. Toxicol. In Vitro  12, 201–207. Google Scholar Abraham, M. H., Kumarsingh, R., Cometto-Muñiz, J. E., and Cain, W. S. ( 1998). Draize eye scores and eye irritation thresholds in man can be combined into one quantitative structure-activity relationship. Toxicol. In Vitro  12, 403–408. Google Scholar Abraham, M. H., and Weathersby, P. K. ( 1994). Hydrogen bonding. 30. Solubility of gases and vapors in biological liquids and tissues. J. Pharm. Sci.  83, 1450–1456. Google Scholar Allgood, G. S. ( 1989). Use of animal eye test data and human experience for determining the ocular irritation potential of shampoos. J. Toxicol. Cutaneous Ocul. Toxicol.  8, 321–326. Google Scholar Barratt, M. D. ( 1995). A quantitative structure-activity relationship for the eye irritation potential of neutral organic chemicals. Toxicol. Lett.  80, 69–74. Google Scholar Barratt, M. D. ( 1997). QSARs for the eye irritation potential of neutral organic molecules. Toxicol. In Vitro  11, 1–8. Google Scholar Brantom, P. G., Bruner, L. H., Chamberlain, M., De Silva, O., Depuis, J., Earl, L. K., Lovell, D. P., Pape, W. J. W., Uttley, M., Bagley, D. M., et al. ( 1997). A summary report of the COLIPA international validation study on alternatives to the Draize rabbit eye irritation test. Toxicol. In Vitro  11, 141–179. Google Scholar Chamberlain, M., and Barratt, M. D. ( 1995). Practical applications of QSAR to in vitro toxicology illustrated by consideration of eye irritation. Toxicol. In Vitro  9, 543–547. Google Scholar Cometto-Muñiz, J. E., and Cain, W. S. ( 1991). Nasal pungency, odor, and eye irritation thresholds for homologous acetates. Pharmacol. Biochem. Behav.  39, 983–989. Google Scholar Cometto-Muñiz, J. E., and Cain, W. S. ( 1995). Relative sensitivity of the ocular trigeminal, nasal trigeminal, and olfactory systems to airborne chemicals. Chem. Senses  20, 191–198. Google Scholar Cometto-Muñiz, J. E., and Cain, W. S. ( 1998). Trigeminal and olfactory sensitivity: Comparison of modalities and methods of measurement. Int. Arch. Occup. Environ. Health  71, 105–110. Google Scholar Cometto-Muñiz, J. E., Cain, W. S., and Abraham, M. H. ( 1998). Nasal pungency and odor of homologous aldehydes and carboxylic acids. Exp. Brain Res.  118, 180–188. Google Scholar Cometto-Muñiz, J. E., Cain, W. S., Abraham, M. H., and Kumarsingh, R. ( 1998). Sensory properties of selected terpenes. Thresholds for odor, nasal pungency, nasal localization, and eye irritation. Ann. NY Acad. Sci.  855, 648–651. Google Scholar Cometto-Muñiz, J. E., Cain, W. S., Abraham, M. H., and Kumarsingh, R. ( 1998). Trigeminal and olfactory chemosensory impact of selected terpenes. Pharmacol. Biochem. Behav.  60, 765–770. Google Scholar Cometto-Muñiz, J. E., Cain, W. S., and Hudnell, H. K. ( 1997). Agonistic sensory effects of airborne chemicals in mixtures: Odor, nasal pungency, and eye irritation. Percept. Psychophys.  59, 665–674. Google Scholar Cronin, M. T., Basketter, D. A., and York, M. ( 1994). A quantitative structure-activity relationship (QSAR) investigation of a Draize eye irritation database. Toxicol. In Vitro  8, 21–28. Google Scholar Draize, J. H., Woodward, G., and Calvery, H. O. ( 1944). Methods for the study of irritation and toxicity of substances applied topically to the skin and mucous membranes. J. Pharmacol. Exp. Ther.  82, 377–390. Google Scholar ECETOC manual No. 48(2) ( 1998). Eye Irritation Reference Chemicals Data Bank, 2nd ed. ECETOC, Brussels. Google Scholar Enslein, K. ( 1988). An overview of structure-activity relationships as an alternative to testing in animals for carcinogenicity, mutagenicity, dermal and eye irritation, and acute oral toxicity. Toxicol. Ind. Health  4, 479–498. Google Scholar Freeburg, F. E., Nixon, G. A., Reer, P. J., Weaver, J. E., Bruce, R. D., Griffith, J. F., and Sanders, L. W. ( 1986). Human and rabbit eye responses to chemical insult. Fundam. Appl. Toxicol.  7, 626–634. Google Scholar Griffith, J. F. ( 1989). Use of human experience to calibrate Draize and in vitro eye test data. J. Toxicol. Cutaneous Ocul. Toxicol.  8, 23–34. Google Scholar HyperChem software, version 6.01 ( 2000). Hypercube Inc., Gainsville, FL. Google Scholar Klopman, G., Ptchelintsev, D., Frierson, M., Pennisi, S., Renssskers, K., and Dickens, M. ( 1993). Multiple computer automated structure evaluation methodology as an alternative to in vivo eye irritation testing. Altern. Lab. Anim.  21, 14–27. Google Scholar Kulkarni, A. S., and Hopfinger, A. J. ( 1999). Membrane-interaction QSAR analysis: Application to the estimation of eye irritation by organic compounds. Pharm. Res.  16, 1245–1253. Google Scholar Leo, A. J. ( 2002). The MedChem database, BioByte Corp. and Pomona College. Daylight Chemical Information Systems, Mission Viejo, CA. Google Scholar Patlewicz, G. Y., Rodford, R. A., Ellis, G., and Barratt, M. D. ( 2000). A QSAR model for the eye irritation of cationic surfactants. Toxicol. In Vitro  14, 79–84. Google Scholar Platts, J. A., and Abraham, M. H. ( 2000). Partition of volatile organic compounds from air and from water into plant cuticular matrix: An LFER analysis. Environ. Sci. Technol.  34, 318–323. Google Scholar Roggeband, R., York, M., Pericoi, M., and Braun, W. ( 2000). Eye irritation response in rabbit and man after single applications of equal volumes of undiluted model liquid detergent products. Food Chem. Toxicol.  39, 727–734. Google Scholar Spielmann, H., Liebsch, M., Kalweit, S., Moldenhaur, F., Wirnsberger, T., Holzhutter, H.-G., Schneider, B., Glaser, S., Gerner, I., Pape, W. J. W., et al. ( 1998). Results of a validation study in Germany on two in vitro alternatives to the Draize eye irritation test, the HET-CAM test and the 3T3 NRU cytotoxicity test. Altern. Lab. Anim.  24, 741–858. Google Scholar Todeschini, R., Consonni, V., and Pavan, M. ( 2002) Dragon software, version 2.1. Milan, Italy. Google Scholar Wilhelmus, K. R. ( 2001). The Draize eye test. Surv. Ophthalmol.  45, 493–515. Google Scholar York, M., and Steiling, W. ( 1998). A critical review of the assessment of eye irritation potential using the Draize rabbit eye test. J. Appl. Toxicol.  18, 233–240. Google Scholar © 2003 Society of Toxicology http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Toxicological Sciences Oxford University Press

Draize Rabbit Eye Test Compatibility with Eye Irritation Thresholds in Humans: A Quantitative Structure-Activity Relationship Analysis

Loading next page...
 
/lp/oxford-university-press/draize-rabbit-eye-test-compatibility-with-eye-irritation-thresholds-in-HVpEGPvYwN

References (37)

Publisher
Oxford University Press
Copyright
© 2003 Society of Toxicology
ISSN
1096-6080
eISSN
1096-0929
DOI
10.1093/toxsci/kfg242
pmid
14514959
Publisher site
See Article on Publisher Site

Abstract

Abstract Draize rabbit eye test scores, as modified maximum average score (MMAS), for 68 pure bulk liquids were adjusted by the liquid-saturated vapor pressure P°. These 68 adjusted scores, as log (MMAS/P°), were shown to be completely equivalent to eye irritation thresholds (EIT), expressed as log (1/EIT), for 23 compounds in humans. Thus, for the first time the Draize eye test in rabbits for pure bulk liquids is shown to be perfectly compatible with eye irritation thresholds in humans. The total data set for 91 compounds was analyzed by the general solvation equation of Abraham. Values of log (MMAS/P°) or log (1/EIT) could be fitted to a five-parameter equation with R2 = 0.936, SD = 0.433, AD = 0.000, and AAD = 0.340 over a range of 9.6 log units. When divided into a training set of 45 compounds, the corresponding equation could be used to predict the remaining 46 compounds in a test set with AD = −0.037 and AAD = 0.345 log units. Thus, the 91-compound equation can now be used to predict further EIT values to around 0.4 log units. It is suggested that the mechanism of action in the Draize test and in the human EIT involves passive transfer of the compound to a biophase that is quite polar, is a strong hydrogen bond base, a moderate hydrogen bond acid, and quite hydrophobic. The biophase does not resemble water or plasma, but resembles an organic solvent such as N-methylformamide. The Draize rabbit eye test (Draize et al., 1944) is the only widely used assay for the effect of substances on the eye. In view of the scientific, ethical, and economic concerns over the Draize test (Wilhelmus, 2001), it is not surprising that alternatives to the Draize test have been examined and that various calculation procedures have been published. An in-depth study (Brantom et al., 1997) of numerous alternative assays has been carried out, but the conclusion was that none of them could be regarded as a valid replacement for the Draize test. On the other hand, it has been suggested (Spielmann et al., 1998) that a combination of two in vitro tests could be used to identify severe irritants. One challenge in finding alternatives to the Draize test is that the available data cover compounds in a variety of physical forms (i.e., liquids, solids, and aqueous solutions). The actual mechanism of irritation may well not be the same across these forms, and this would preclude any general alternative test or any general calculation. Fragmentation schemes for particular chemical or biological effects attempt to relate the effect to structural fragments of molecules. These may be functional groups or just parts of molecules, such as the CH3 or CH2 fragments. Then an effect is assigned to each fragment, and predictions are made by summation of the fragment effects in a given molecule. Such schemes for the estimation of eye irritation have been reported (Enslein, 1988; Klopman et al., 1993), but most of the data used by Enslein were not Draize scores. Although Klopman et al. (1993) used Draize scores, these were used in conjunction with other judgments; unfortunately, the full list of compounds studied is not available. Other workers have restricted their analyses to pure organic compounds. Principal components analysis (PCA) and neural networks (Barratt, 1995, 1997; Chamberlain and Barratt, 1995) have been used to discriminate between irritants and nonirritants with reasonable success. On the other hand, investigation of a similar data set using linear combinations of descriptors and PCA (Cronin et al., 1994) failed to generate any general linear correlation of modified Draize scores and failed to observe any marked distinction between irritants and nonirritants by PCA. The modified Draize scores were defined as MMAS divided by the molarity of the pure liquid; the latter is given by 1000 times the density of the pure liquid divided by the liquid molecular weight. The descriptors of the compounds in the best linear equation were ClogP, where P is a calculated water-octanol partition coefficient, the lowest unoccupied molecular orbital (LUMO), and a connectivity index. Cronin et al. (1994) correctly pointed out that use of a physically heterogeneous set of compounds, i.e., pure liquids, solids, and aqueous solutions, would make it very difficult to obtain any useful structure-activity relationship (SAR) and so restricted the analysis to pure bulk liquids. Kulkarni and Hopfinger (1999) obtained a reasonable relationship, but only for a very limited set of 18 compounds in a training set and five in a test set. Patlewicz et al. (2000) restricted their analysis to cationic surfactants, and for this set of compounds found a very good fit of observed and calculated Draize eye scores using a neural network. What is surprising is that such studies have been made before any substantial connection between results of the Draize test in rabbits and the effect of the corresponding substances in man has been established. In a comprehensive review of the Draize test, it was noted that the anatomy and biochemistry of the rabbit eye are not the same as those of the human eye and that there were numerous physiological reasons, including low tear production, blink frequency, and ocular surface area, that such a test on rabbits might not adequately predict human effects (Wilhelmus, 2001). York and Steiling (1998) stressed the need to validate the Draize test against controlled human eye data, but noted that “there are no adequate human data.” What comparisons have been made between the effects on rabbits and the effects on humans have been confined to consumer products that are a mixture of various chemicals. Freeburg et al. (1986) examined four such products and showed that the low-volume Draize test correlated with effects on the eyes of humans better than did the normal-volume Draize test. Allgood (1989) also matched the low-volume Draize test against human experience for four shampoos, and Griffith (1989) compared Draize data to consumer eye accident data for soaps and detergents. Roggeband et al. (2000) studied the effect of very low volumes (1–3 μl) of a liquid detergent and a dishwashing liquid on the eyes of rabbits and human volunteers. They observed that the irritation responses in rabbits were greater than those in man, and suggested that the low-volume Draize test could be used to assess eye irritation hazards in man. One problem until recently has been the lack of controlled human eye data (York and Steiling, 1998), but this has been rectified by the determination of eye irritation thresholds (EIT) in humans by a rigorous standardized procedure (Cometto-Muñiz and Cain, 1991, 1995, 1998; Cometto-Muñiz et al., 1997, 1998b, Cometto-Muñiz et al., c). It is these EIT values that we shall use. For the substances studied in the Draize test as pure bulk liquids, no adequate connection between the effects on rabbits and the effects on humans has been established, although for a limited number of liquids an indirect connection has been indicated (Abraham et al., 1998a,b). It is the purpose of this work to use the indirect method on a large data set to establish whether or not such a connection exists, and, if successful, to use the connection to obtain a quantitative structure-activity relationship (QSAR) for EIT in man. The Draize test scores, MMAS, that we use are those for pure bulk liquids as recorded in the ECETOC manual (ECETOC, 1998). We used values for all the pure bulk liquids given, except for several high boiling liquids for which vapor pressures at 298 K were either unknown or unreliable. MATERIALS AND METHODS Comparison of Draize scores and EIT. The Draize test scores, MMAS, that we shall use (ECETOC, 1998) refer to the effect of pure bulk liquids, whereas the EIT, in ppm, are established from the effect of the vapor of liquids at some particular partial pressure. Hence, a direct comparison of MMAS and EIT is not possible. Consider the transfer of a compound from the vapor phase to a solvent phase, the equilibrium constant being defined as   \[\mathit{K\ =\ [conc\ of\ compound\ in\ solvent]/[conc\ of\ compound\ in\ vapor\ phase]}\] (1) The compound can also be transferred from the bulk liquid to the solvent phase, the equilibrium constant being just the solubility of the bulk liquid.   \[\mathit{S\ =\ [conc\ of\ compound\ in\ solvent]}\] (2) These two equilibrium constants are related through the saturated vapor pressure, P°, of the pure bulk liquid.   \[\mathit{S\ =\ P{^\circ}\ *K\ or\ log\ (S/P{^\circ})\ =\ log\ K}\] (3) This is illustrated in Figure 1A. Exactly the same relationship can be shown for the transfer to a biophase, such as a rabbit or human eye, rather than to a solvent (see Fig. 1B). The transfer from the bulk liquid to the biophase is proportional to the Draize eye score, MMAS, provided that the mechanism of the Draize test involves passive transport to the site of action. Then, with this assumption, following Equation 3 and Figure 1, the Draize scores can be converted to scores for the effect of vapors through Equation 4, where m is some constant.   \[\mathit{Log\ (MMAS/P{^\circ})\ =\ log\ K\ +\ m}\] (4) If, again, we assume passive transfer of vapors to the biophase in human eye irritation, then log (EIT) will be given by the same type of equation as follows:   \[\mathit{Log}\ (1/\mathit{EIT})\ =\ \mathit{log\ K}{^\prime}\ +\ \mathit{m}{^\prime}\] (5) We prefer to use log (1/EIT) because the greater the value the more potent the compound. Then, the combination of Equations 4 and 5 leads to Equation 6, which can be used as a starting point for any comparison of MMAS values with EIT values. Since EIT values are listed in ppm, we use P° in ppm, and at 298 K in Equation 6.   \[\mathit{Log\ (MMAS/P{^\circ})\ =\ log\ (1/EIT)\ +\ m{^{\prime\prime}}}\] (6) QSAR studies. Our procedure is to use log (MMAS/P°) or log (1/EIT) as the dependent variable, SP, and to construct QSARs through Equation 7, the general equation that we have developed (Abraham, 1993; Abraham and Al-Hussaini, 2002). In Equation 7, the independent variables are compound descriptors as follows: E is the solute excess molar refractivity, in units of (cm3 mol−1)/10; S is the solute dipolarity/polarizability; A and B are the overall or summation hydrogen bond acidity and basicity; and L is the logarithm of the gas-hexadecane partition coefficient. Our rationale in using the particular descriptors in Equation 7 was that we had already used this equation to fit and interpret a number of gas-to-solvent partitions. Since we suggest that (MMAS/P°) or (1/EIT) are related to gas-to-biophase partition (see Fig. 1), it is logical to use Equation 7, at least as a first step, in the analysis of log (MMAS/P°) or log (1/EIT). To investigate whether QSARs based on Equation 7 might be improved by the inclusion of other types of descriptors, we calculated shape descriptors using the HyperChem software (HyperChem, 2000), with conformational energy minimized using the AM1 semiempirical method. VG is the three-dimensional volume of the minimum energy conformation computed using the HyperChem QSAR option. LG is the longest length of the minimum energy conformation. In addition, we investigated a shape descriptor, DPO1, calculated from the Dragon software (Todeschini et al., 2002) after conformational energy minimization by HyperChem. DPO1 is a molecular descriptor derived from the distance distribution moments of the geometry matrix defined as the average row sum of its entries. It can be regarded as a shape descriptor that takes into account branching and the distance of atoms from a center of gravity of a molecule   \[\mathit{SP}\ =\ \mathit{c}\ +\ \mathit{e}.\mathbf{E}\ +\ \mathit{s}.\mathbf{S}\ +\ \mathit{a}.\mathbf{A}\ +\ \mathit{b}.\mathbf{B}\ +\ \mathit{l}.\mathbf{L}\] (7) The coefficients in Equation 7 and other equations are evaluated through multiple linear regression analysis. The statistics that we shall detail are the following: N, the number of data points; R2, the coefficient of variation; SD, the standard deviation; AD, the average deviation; AAD, the absolute average deviation; and F, the Fischer statistic. RESULTS AND DISCUSSION Comparison of MMAS and EIT Values The most satisfactory comparison of log (MMAS/P°) and log (1/EIT) would be a direct comparison using Equation 6. Unfortunately, log (MMAS/P°) and log (1/EIT) values are only available for nine common compounds in Table 1. For these nine compounds, log (MMAS/P°) = log (1/EIT) + 0.89 with AAD = 0.57; but, this is not at all sufficient to demonstrate that MMAS is related to EIT. A much better method, which uses the information on log (MMAS/P°) and log (1/EIT) for all the compounds in Table 1, is to combine the two sets of data and to determine whether or not the same QSAR will fit the two sets. A straightforward method is to include a new descriptor, I, in Equation 7 that takes the value I = 0 for the log (1/EIT) series of compounds and I = 1 for the log (MMAS/P°) series of compounds. This I descriptor takes into account data on all 91 compounds and, within the statistics of Equation 8 below, shows that for any compound there is a constant difference between log (MMAS/P°) and log (1/EIT). In Equation 8, this difference is 0.568 ± 0.106 log units, which is in agreement with the difference of 0.89 ± 0.57 found for nine compounds only. For a total of 91 compounds, we constructed Equation 8, where SP = log (MMAS/P°) or log (1/EIT). There was only one outlier that we omitted, namely dodecane.   \[\mathit{SP}\ =\ {-}7.892\ {-}\ 0.379\mathbf{E}\ +\ 1.872\mathbf{S}\ +\ 3.776\mathbf{A}\ +\ 1.169\mathbf{B}\ +\ 0.785\mathbf{L}\ +\ 0.568\mathbf{I}\] (8) The following statistics are applicable: N = 91, R2 = 0.936, SD = 0.433, AD = 0.000, AAD = 0.340, and F = 204.5. The 23 points for log (1/EIT) all fall on the line of identity, as shown in Figure 2. If the deviations from the line for these 23 points are calculated separately, then AAD = 0.371 and AD = 0.000, the latter showing that there is no bias whatsoever in the fitting of the 23 log (1/EIT) points to Equation 8. We can conclude that for the 68 values of log (MMAS/P°) and the 23 values of log (1/EIT) the two sets of data are compatible. Since the two sets both cover a range of variation of compound type, we suggest that this finding is quite general. The combined equation, Equation 8, includes data on aliphatic and aromatic hydrocarbons and halogenated hydrocarbons, alcohols, ketones, acids, esters, nitro compounds, sulfides, and terpenes, as well as pyridine and 4-fluoroaniline. Therefore, we feel that this equation is of such generality that it can be used to predict EIT in man for a host of chemical vapors. Inspection of Figure 2 shows that the range of log (MMAS/P°) is much wider than that of log (1/EIT), with numerous compounds having very positive values of log (MMAS/P°). These are all very involatile liquids with vapor pressures so low that they will elicit no response in human subjects. This illustrates one advantage of including the log (MMAS/P°) data: compounds can be assigned log (1/EIT) values by this indirect method, even though log (1/EIT) cannot be determined directly. Another advantage is that compounds that cannot be studied on humans for ethical reasons can be included. Thus, benzene and the sulfides can be assigned indirect log (1/EIT) values in this way. If the two data sets are indeed compatible, then equations similar to Equation 8 should be obtained if the two sets are treated separately (although the coefficient of I must then be zero). There is not enough data on log (1/EIT) to obtain an equation with five variables, but for the log (MMAS/P°) set we find the following:   \[\mathit{Log\ (MMAS/P{^\circ})}\ =\ {-}7.355\ {-}\ 0.351\mathbf{E}\ +\ 1.997\mathbf{S}\ +\ 4.380\mathbf{A}\ +\ 1.159\mathbf{B}\ +\ 0.758\mathbf{L}\] (9) The following statistics are applicable: N = 68, R2 = 0.954, SD = 0.398, AD = 0.000, AAD = 0.318, and F = 255.8. Equation 9 is statistically the same as Equation 8: note that the SD on the coefficients in these equations averages around 0.30, except for the l coefficient where it is much smaller (0.03). Thus, on the basis that Equations 8 and 9 are statistically the same and that the data on log (1/EIT) exactly fit the general Equation 8, we conclude that the Draize eye scores, modified by the compound vapor pressure, can be combined with the EIT into one equation. Except for our preliminary report on considerably fewer compounds (Abraham et al., 1998b), this is the first time that any real connection between Draize eye scores and effects on humans has been established for the important group of pure bulk liquids. There is enough data in the two combined sets to test the predictive power of Equation 8. We ranked the data in order of SP and chose every other compound as a training set. For the 45 compounds we found the following:   \[\mathit{SP}\ =\ {-}7.856\ {-}\ 0.034\mathbf{E}\ +\ 1.529\mathbf{S}\ +\ 3.656\mathbf{A}\ +\ 1.555\mathbf{B}\ +\ 0.686\mathbf{L}\] (10) The following statistics are applicable: N = 45, R2 = 0.921, SD = 0.469, AD = 0.000, AAD = 0.354, and F = 74.0. The statistics of Equation 10 are comparable to those of the full Equation 8, bearing in mind the SD values of the various coefficients, as noted above. Then, we can use Equation 10 to predict the remaining 46 compounds in the test set that have not been used to derive Equation 10. Results are given in Table 2. Equation 10, and, hence, the full Equation 8, can predict SP values, where SP = log (MMAS/P°) and log (1/EIT) with almost no bias, since AD = −0.04, AAD = 0.35, and SD = 0.43 log unit. Considering that the 90 values cover a range of 9.7 log units and that the 23 log (1/EIT) values cover a range of 4.0 log units, we suggest that the full Equation 8 can be used to predict log (1/EIT) values to within 0.45 log units, quite generally. Interpretation of the QSAR Results The QSAR, Equation 8, has been constructed on the assumption that the Draize eye scores and the human EIT result from passive transfer from the bulk liquid and the vapor phase, respectively, to a biophase. Such passive transfer is usually nonspecific, in that the position of a substituent has little effect, unless there is some particular interaction between the substituents. We can illustrate this through the examination of water-octanol partition coefficients as log Poct (Leo, 2002; see Table 3). The position of substituents has almost no effect except for o-hydroxybenzamide where the substituents interact. The statistics of Equation 8 show that 94% of the information on log (MMAS/P°) and log (1/EIT) can be accounted for on the basis of passive transfer. There is still the possibility that the remaining 6% information applies to transfer from the biophase to some receptor. We would expect such transfer to be more specific; in particular, the shape of a compound might be crucial. To examine this point, we included the three shape descriptors in a QSAR and found the following:   \begin{eqnarray*}&&\mathit{SP}\ =\ {-}6.944\ {-}\ 1.201\mathbf{E}\ +\ 1.187\mathbf{S}\ +\ 3.268\mathbf{A}\ +\ 1.445\mathbf{B}\ +\ 1.445\mathbf{L}\ +\ 0.554\mathbf{I}\ \\&&\ +\ 0.084\mathbf{DPO1}\ {-}\ 0.007\mathbf{VG}\ +\ 0.031\mathbf{LG}\end{eqnarray*} (11) The following statistics are applicable: N = 91, R2 = 0.941, SD = 0.422, AD = 0.000, AAD = 0.334, and F = 144.0. The additional shape descriptors have almost no effect; so, we considered that the main mechanistic step in human eye irritation and in rabbit eye irritation is a simple passive transfer of a compound as a bulk liquid or as a vapor to a biophase. This partly explains the comment (Cronin et al., 1994) that construction of a QSAR for a physically heterogeneous set of compounds is very difficult. For compounds that are tested on rabbits as solids or as aqueous solutions, the passive transfer mechanism shown in Figure 1 will not hold. It has been noted (York and Steiling, 1998) that solids can cause irritancy through abrasive, mechanical effects, and that Draize scores on solutions of liquids cannot be used to assess the irritancy of the pure bulk liquids. Nevertheless, molecular size can be crucial in a fundamental aspect. Although Table 1 includes some quite large compounds such as iso-propyl iso-stearate, studies of eye and nose irritation from members of homologous series have indicated the existence of a cut-off effect in homologous series beyond which larger homologs fail to evoke irritation (Cometto-Muñiz et al., 1998c). If, as suggested (Cometto-Muñiz et al., 1998a), such cut-off rests on a biological restriction (i.e., a receptor-related effect) rather than a physical restriction (i.e., just a low vapor pressure effect), then modeling the dimensional commonalities among cut-off molecules will serve to define the maximum molecular dimensions that the receptor can fit. We are, at present, exploring this issue. Some information about the nature of the biophase can be obtained by a comparison of the coefficients in Equation 8 with those for passive transfer from the vapor to various phases that might be considered models for the biophase. Table 4 gives the coefficients of Equation 8 for transfer from the vapor to water (Abraham et al., 1994) and numerous organic solvents (Abraham et al., 2000, 2001). These coefficients are also known for transfer from the vapor to many biological tissues (Abraham and Weathersby, 1994) and from the vapor to a plant cuticular polymeric membrane matrix (Platts and Abraham, 2000). Inspection of the data in Table 4 shows that the biophase cannot resemble either water or a largely aqueous phase such as plasma. The aqueous phases have very small l coefficients (−0.213 and 0.157), whereas the biophase is relatively hydrophobic with l = 0.787, quite close to that for many organic solvents. By comparison to organic solvents, the biophase is dipolar with s = 2.02, which is almost the same as that for N-methylformamide. The biophase is a strong hydrogen bond base with a = 4.02, nearly as strong as the amides. It is also a moderately strong hydrogen bond acid, with b = 1.15, compared to 1.43 for wet 1-octanol. The nearest organic solvent as regards these chemical properties is the secondary amide, N-methylformamide. Interestingly, the plant polymeric membrane matrix also has properties that are quite similar to the biophase, but it is less polar, less acidic, and less basic. The biophase cannot be situated near an aqueous mucus layer, and it is more likely to be composed of amides (peptides) in a poorly aqueous environment. Conclusions The success of our QSAR to integrate, for bulk liquids, eye irritation data from the Draize test in rabbits, as MMAS values, with EIT in humans obtained by a standardized procedure has several very important implications. Firstly, for bulk liquids there is a proven statistical and physicochemical basis to support the Draize test as an indirect measure of EIT in humans. From this, we show that EIT in humans can be obtained by this indirect method for a large number of compounds that cannot be studied by the standardized procedure. These compounds include those that have too low a vapor pressure to elicit a response in humans, and those that cannot be studied on humans because of ethical considerations. Secondly, the resulting QSAR represents the first time that such a scale having statistical significance, chemical diversity, and physicochemical basis is available specifically for the eye irritation effect of vapors on humans. A test such as the Draize test is certainly needed for eye irritancy of bulk liquids that might come into contact with the eye. However, we feel that consideration should be given to our scale as a new, statistically sound measure of eye irritation of vapors on humans that could be used to assess the environmental impact of vapors. We stress that, although our QSAR covers a wide variety of compounds, there is still considerable scope for extending the range of compounds; this is part of an ongoing project to determine further values of EIT. The general QSAR, Equation 7, has also proved to be valuable in the analysis of odor thresholds in humans and nasal pungency (nasal irritation) thresholds in humans (Abraham et al., 2001, 2002). Further work to extend the range of compounds in these two areas is ongoing. We hope to be able to present numerous very general QSARs, not just the present one for eye irritation, that can be used for the prediction of environmental effects of vapors on humans. TABLE 1 Compounds, Values of SP, and the Compound Descriptorsa Name  SPa  I  E  S  A  B  L  VG  LG  DP01  aSP = log (MMAS/P°) when I = 1 and SP = log (1/EIT) when I = 0.  bThese are the nine common compounds in the two data sets.  2-Bromobutane  −5.16  1  0.344  0.35  0.00  0.14  2.933  381.12  5.578  2.377  1-Bromo-4-chlorobutane  −3.34  1  0.571  0.82  0.00  0.25  4.007  436.15  6.906  2.839  Dichlorotoluenes (3,4)  −2.77  1  0.900  0.80  0.00  0.03  5.089  467.09  6.453  3.173  3,3-Dimethylpentane  −5.24  1  0.000  0.00  0.00  0.00  2.946  450.50  6.771  2.803  1-Bromooctane  −2.90  1  0.339  0.40  0.00  0.12  5.143  606.11  11.300  3.574  iso-Stearyl alcohol  2.37  1  0.140  0.39  0.37  0.48  9.500  1093.42  22.807  4.997  Methylisobutylketone  −3.73  1  0.111  0.65  0.00  0.51  3.089  421.84  6.221  2.833  3-Methylhexane  −5.06  1  0.000  0.00  0.00  0.00  3.044  470.70  8.009  2.938  4-Bromophenetole  −1.89  1  0.967  0.90  0.00  0.23  5.520  529.22  9.018  3.431  Di-n-propyl disulfide  −2.75  1  0.653  0.52  0.00  0.27  4.984  555.76  11.195  3.419  Heptyl methacrylate  −2.15  1  0.445  0.49  0.00  0.45  5.697  718.89  13.840  4.114  1-Bromohexane  −3.60  1  0.349  0.40  0.00  0.12  4.138  498.51  8.813  3.101  iso-Propyl iso-stearate  2.26  1  −0.020  0.53  0.00  0.47  10.250  1259.97  25.318  5.319  1-Bromopentane  −3.92  1  0.356  0.40  0.00  0.12  3.611  444.66  7.530  2.819  1,9-Decadiene  −3.17  1  0.184  0.20  0.00  0.10  4.380  617.13  13.018  3.753  1,6-Dibromohexane  −0.91  1  0.711  0.80  0.00  0.26  5.328  563.59  9.560  3.374  1,3-Diisopropylbenzene  −2.32  1  0.605  0.46  0.00  0.20  5.170  628.84  8.868  3.666  2-Methylpentane  −5.15  1  0.000  0.00  0.00  0.00  2.503  422.99  6.820  2.669  s-Butylbenzene  −3.08  1  0.603  0.48  0.00  0.16  4.506  528.60  8.173  3.369  3-Ethyltoluene  −3.23  1  0.630  0.51  0.00  0.18  4.275  484.64  7.785  3.187  Methyl trimethylacetate  −4.16  1  0.049  0.54  0.00  0.45  2.932  448.41  6.328  2.964  2-Bromopropane  −5.02  1  0.332  0.35  0.00  0.00  2.390  332.65  4.329  2.027  1,5-Dimethylcyclooctadiene  −2.90  1  0.604  0.30  0.00  0.18  4.812  533.10  7.843  3.330  cis-Cyclooctene  −3.48  1  0.460  0.24  0.00  0.10  4.119  446.17  5.699  2.949  iso-Stearic acid  4.38  1  0.015  0.57  0.60  0.49  9.600  1093.73  22.075  5.077  Methylcyclopentane  −4.69  1  0.225  0.10  0.00  0.00  2.907  385.83  5.271  2.509  Ethyl trimethylacetate  −3.55  1  -0.010  0.52  0.00  0.45  3.481  504.71  7.638  3.201  1,4-Dibromobutane  −2.00  1  0.733  0.80  0.00  0.27  4.353  455.93  7.084  2.856  1,5-Dibromopentane  −1.41  1  0.723  0.80  0.00  0.27  4.848  509.83  8.238  3.130  1,3-Dibromopropane  −2.89  1  0.723  0.80  0.00  0.27  3.872  402.12  5.741  2.542  iso-Myristyl alcohol  1.26  1  0.155  0.39  0.37  0.48  7.480  877.65  17.809  4.517  2,4-Difluoronitrobenzene  −1.83  1  0.677  1.20  0.00  0.25  4.350  408.66  6.315  3.446  1,5-Hexadiene  −4.80  1  0.191  0.15  0.00  0.10  2.450  402.30  8.039  2.762  4-Methylpentan-2-one  −3.73  1  0.111  0.65  0.00  0.51  3.089  426.79  6.831  2.845  Allyl methacrylate  −3.24  1  0.290  0.57  0.00  0.54  3.741  484.32  8.803  3.300  Styrene  −3.11  1  0.849  0.65  0.00  0.16  3.856  416.77  7.331  2.978  Butyl acetateb  −3.30  1  0.071  0.60  0.00  0.45  3.353  470.17  8.831  3.196  2,2-Dimethylpentan-3-ol  −2.70  1  0.227  0.27  0.31  0.60  3.400  474.04  6.855  2.989  Tolueneb  −3.62  1  0.601  0.52  0.00  0.14  3.325  384.83  5.889  2.719  m-Xylene  −3.11  1  0.623  0.52  0.00  0.16  3.839  436.84  6.742  2.956  Heptan-2-oneb  −2.68  1  0.123  0.68  0.00  0.51  3.760  489.92  9.321  3.229  2-Methylpentan-1-ol  −2.26  1  0.211  0.39  0.37  0.48  3.530  445.14  7.798  2.925  3-Choroproprionitrile  −2.65  1  0.387  1.22  0.02  0.40  3.070  321.06  5.105  2.440  Cellosolve acetate  −2.20  1  0.099  0.79  0.00  0.79  3.747  504.30  9.859  3.418  Ethyl acetateb  −3.91  1  0.106  0.62  0.00  0.45  2.314  361.36  6.375  2.606  Heptan-2-oneb  −2.49  1  0.123  0.68  0.00  0.51  3.760  489.92  9.321  3.229  Ethyl 2-methylacetoacetate  −1.58  1  0.156  0.85  0.00  0.85  4.214  515.21  7.072  3.358  Cyclopentanol  −2.12  1  0.427  0.54  0.32  0.36  3.241  357.56  5.078  2.493  Ethanolb  −3.51  1  0.246  0.42  0.37  0.48  1.485  242.38  4.054  1.510  Methyl cyanoacetate  −0.78  1  0.291  1.34  0.00  0.64  3.367  353.43  6.454  2.849  Propan-2-ol  −3.27  1  0.212  0.36  0.33  0.56  1.764  293.20  4.326  1.929  Methyl acetate  −3.85  1  0.142  0.64  0.00  0.45  1.911  303.83  5.081  2.252  Octan-1-olb  −0.39  1  0.199  0.42  0.37  0.48  4.619  565.90  11.552  3.548  g-Butyrolactone  −0.94  1  0.366  1.30  0.00  0.58  3.600  314.67  4.355  2.465  Furfuryl alcohol  −1.34  1  0.554  0.73  0.50  0.63  3.357  357.58  6.245  2.751  2,2-Dimethylbutanoic acid  −0.65  1  0.186  0.55  0.60  0.51  3.681  431.19  5.838  2.960  Methoxyethyl acrylate  −2.10  1  0.249  0.80  0.00  0.80  3.876  481.45  9.204  3.357  Pyridine  −2.75  1  0.613  0.84  0.00  0.52  3.022  318.53  4.941  2.428  Butanone  −3.38  1  0.166  0.70  0.00  0.51  2.287  328.03  5.595  2.291  2-Ethylhexan-1-ol  −0.57  1  0.209  0.39  0.37  0.48  4.433  546.11  8.886  3.344  iso-Butanol  −2.36  1  0.217  0.39  0.37  0.48  2.413  342.97  5.314  2.301  Butan-1-olb  −2.13  1  0.224  0.42  0.37  0.48  2.601  350.58  6.554  2.423  Diethylaminopropriontrile  −0.83  1  0.267  0.89  0.00  0.86  4.479  515.76  7.571  3.337  Hexan-1-olb  −1.13  1  0.210  0.42  0.37  0.48  3.610  457.78  9.053  3.061  Propanoneb  −3.66  1  0.179  0.70  0.04  0.49  1.696  275.29  4.316  1.912  Ethyleneglycolmonobutyl ether  −1.32  1  0.201  0.50  0.30  0.83  3.806  491.96  10.074  3.289  4-Fluoroaniline  −1.06  1  0.760  1.09  0.20  0.40  4.007  376.72  6.105  2.950  Cyclohexanol  −1.00  1  0.460  0.54  0.32  0.57  3.758  398.55  4.889  2.704  Propanoneb  −5.27  0  0.179  0.70  0.04  0.49  1.696  275.29  4.316  1.912  Pentan-2-one  −4.05  0  0.143  0.68  0.00  0.51  2.755  382.08  6.819  2.648  Heptan-2-oneb  −2.49  0  0.123  0.68  0.00  0.51  3.760  489.92  9.321  3.229  Nonan-2-one  −2.35  0  0.119  0.68  0.00  0.51  4.735  597.48  11.813  3.686  Ethyl acetateb  −4.69  0  0.106  0.62  0.00  0.45  2.314  361.36  6.375  2.606  Butyl acetateb  −2.87  0  0.071  0.60  0.00  0.45  3.353  470.17  8.831  3.196  Hexyl acetate  −2.41  0  0.056  0.60  0.00  0.45  4.351  577.05  11.307  3.663  Octyl acetate  −2.02  0  0.029  0.60  0.00  0.45  5.364  684.18  13.795  4.043  Decyl acetate  −1.30  0  0.033  0.60  0.00  0.45  6.373  793.03  16.298  4.362  Ethanolb  −4.76  0  0.246  0.42  0.37  0.48  1.485  242.38  4.054  1.510  Propan-1-ol  −3.74  0  0.236  0.42  0.37  0.48  2.031  296.30  5.313  2.016  Butan-1-olb  −3.37  0  0.224  0.42  0.37  0.48  2.601  350.58  6.554  2.423  Hexan-1-olb  −2.60  0  0.210  0.42  0.37  0.48  3.610  457.78  9.053  3.061  Octan-1-olb  −1.71  0  0.199  0.42  0.37  0.48  4.619  565.90  11.552  3.548  Tolueneb  −4.41  0  0.601  0.52  0.00  0.14  3.325  384.83  5.889  2.719  Ethylbenzene  −3.87  0  0.613  0.51  0.00  0.15  3.778  432.04  7.284  2.981  Propylbenzene  −3.43  0  0.604  0.50  0.00  0.15  4.230  486.32  8.368  3.242  Cumene  −3.39  0  0.602  0.49  0.00  0.16  4.084  480.18  7.200  3.160  p-Cymene  −3.11  0  0.607  0.49  0.00  0.19  4.590  532.53  8.190  3.360  d-3-Carene  −3.30  0  0.511  0.22  0.00  0.10  4.649  529.88  7.470  3.280  Linalool  −2.55  0  0.398  0.55  0.20  0.67  4.794  613.77  10.000  3.680  1,8-Cineole  −2.15  0  0.383  0.33  0.00  0.76  4.688  543.92  6.400  3.330  Geraniol  −1.35  0  0.513  0.63  0.39  0.66  5.479  619.88  11.470  3.760  Dodecane  1.00  0  0.000  0.00  0.00  0.00  5.695        Name  SPa  I  E  S  A  B  L  VG  LG  DP01  aSP = log (MMAS/P°) when I = 1 and SP = log (1/EIT) when I = 0.  bThese are the nine common compounds in the two data sets.  2-Bromobutane  −5.16  1  0.344  0.35  0.00  0.14  2.933  381.12  5.578  2.377  1-Bromo-4-chlorobutane  −3.34  1  0.571  0.82  0.00  0.25  4.007  436.15  6.906  2.839  Dichlorotoluenes (3,4)  −2.77  1  0.900  0.80  0.00  0.03  5.089  467.09  6.453  3.173  3,3-Dimethylpentane  −5.24  1  0.000  0.00  0.00  0.00  2.946  450.50  6.771  2.803  1-Bromooctane  −2.90  1  0.339  0.40  0.00  0.12  5.143  606.11  11.300  3.574  iso-Stearyl alcohol  2.37  1  0.140  0.39  0.37  0.48  9.500  1093.42  22.807  4.997  Methylisobutylketone  −3.73  1  0.111  0.65  0.00  0.51  3.089  421.84  6.221  2.833  3-Methylhexane  −5.06  1  0.000  0.00  0.00  0.00  3.044  470.70  8.009  2.938  4-Bromophenetole  −1.89  1  0.967  0.90  0.00  0.23  5.520  529.22  9.018  3.431  Di-n-propyl disulfide  −2.75  1  0.653  0.52  0.00  0.27  4.984  555.76  11.195  3.419  Heptyl methacrylate  −2.15  1  0.445  0.49  0.00  0.45  5.697  718.89  13.840  4.114  1-Bromohexane  −3.60  1  0.349  0.40  0.00  0.12  4.138  498.51  8.813  3.101  iso-Propyl iso-stearate  2.26  1  −0.020  0.53  0.00  0.47  10.250  1259.97  25.318  5.319  1-Bromopentane  −3.92  1  0.356  0.40  0.00  0.12  3.611  444.66  7.530  2.819  1,9-Decadiene  −3.17  1  0.184  0.20  0.00  0.10  4.380  617.13  13.018  3.753  1,6-Dibromohexane  −0.91  1  0.711  0.80  0.00  0.26  5.328  563.59  9.560  3.374  1,3-Diisopropylbenzene  −2.32  1  0.605  0.46  0.00  0.20  5.170  628.84  8.868  3.666  2-Methylpentane  −5.15  1  0.000  0.00  0.00  0.00  2.503  422.99  6.820  2.669  s-Butylbenzene  −3.08  1  0.603  0.48  0.00  0.16  4.506  528.60  8.173  3.369  3-Ethyltoluene  −3.23  1  0.630  0.51  0.00  0.18  4.275  484.64  7.785  3.187  Methyl trimethylacetate  −4.16  1  0.049  0.54  0.00  0.45  2.932  448.41  6.328  2.964  2-Bromopropane  −5.02  1  0.332  0.35  0.00  0.00  2.390  332.65  4.329  2.027  1,5-Dimethylcyclooctadiene  −2.90  1  0.604  0.30  0.00  0.18  4.812  533.10  7.843  3.330  cis-Cyclooctene  −3.48  1  0.460  0.24  0.00  0.10  4.119  446.17  5.699  2.949  iso-Stearic acid  4.38  1  0.015  0.57  0.60  0.49  9.600  1093.73  22.075  5.077  Methylcyclopentane  −4.69  1  0.225  0.10  0.00  0.00  2.907  385.83  5.271  2.509  Ethyl trimethylacetate  −3.55  1  -0.010  0.52  0.00  0.45  3.481  504.71  7.638  3.201  1,4-Dibromobutane  −2.00  1  0.733  0.80  0.00  0.27  4.353  455.93  7.084  2.856  1,5-Dibromopentane  −1.41  1  0.723  0.80  0.00  0.27  4.848  509.83  8.238  3.130  1,3-Dibromopropane  −2.89  1  0.723  0.80  0.00  0.27  3.872  402.12  5.741  2.542  iso-Myristyl alcohol  1.26  1  0.155  0.39  0.37  0.48  7.480  877.65  17.809  4.517  2,4-Difluoronitrobenzene  −1.83  1  0.677  1.20  0.00  0.25  4.350  408.66  6.315  3.446  1,5-Hexadiene  −4.80  1  0.191  0.15  0.00  0.10  2.450  402.30  8.039  2.762  4-Methylpentan-2-one  −3.73  1  0.111  0.65  0.00  0.51  3.089  426.79  6.831  2.845  Allyl methacrylate  −3.24  1  0.290  0.57  0.00  0.54  3.741  484.32  8.803  3.300  Styrene  −3.11  1  0.849  0.65  0.00  0.16  3.856  416.77  7.331  2.978  Butyl acetateb  −3.30  1  0.071  0.60  0.00  0.45  3.353  470.17  8.831  3.196  2,2-Dimethylpentan-3-ol  −2.70  1  0.227  0.27  0.31  0.60  3.400  474.04  6.855  2.989  Tolueneb  −3.62  1  0.601  0.52  0.00  0.14  3.325  384.83  5.889  2.719  m-Xylene  −3.11  1  0.623  0.52  0.00  0.16  3.839  436.84  6.742  2.956  Heptan-2-oneb  −2.68  1  0.123  0.68  0.00  0.51  3.760  489.92  9.321  3.229  2-Methylpentan-1-ol  −2.26  1  0.211  0.39  0.37  0.48  3.530  445.14  7.798  2.925  3-Choroproprionitrile  −2.65  1  0.387  1.22  0.02  0.40  3.070  321.06  5.105  2.440  Cellosolve acetate  −2.20  1  0.099  0.79  0.00  0.79  3.747  504.30  9.859  3.418  Ethyl acetateb  −3.91  1  0.106  0.62  0.00  0.45  2.314  361.36  6.375  2.606  Heptan-2-oneb  −2.49  1  0.123  0.68  0.00  0.51  3.760  489.92  9.321  3.229  Ethyl 2-methylacetoacetate  −1.58  1  0.156  0.85  0.00  0.85  4.214  515.21  7.072  3.358  Cyclopentanol  −2.12  1  0.427  0.54  0.32  0.36  3.241  357.56  5.078  2.493  Ethanolb  −3.51  1  0.246  0.42  0.37  0.48  1.485  242.38  4.054  1.510  Methyl cyanoacetate  −0.78  1  0.291  1.34  0.00  0.64  3.367  353.43  6.454  2.849  Propan-2-ol  −3.27  1  0.212  0.36  0.33  0.56  1.764  293.20  4.326  1.929  Methyl acetate  −3.85  1  0.142  0.64  0.00  0.45  1.911  303.83  5.081  2.252  Octan-1-olb  −0.39  1  0.199  0.42  0.37  0.48  4.619  565.90  11.552  3.548  g-Butyrolactone  −0.94  1  0.366  1.30  0.00  0.58  3.600  314.67  4.355  2.465  Furfuryl alcohol  −1.34  1  0.554  0.73  0.50  0.63  3.357  357.58  6.245  2.751  2,2-Dimethylbutanoic acid  −0.65  1  0.186  0.55  0.60  0.51  3.681  431.19  5.838  2.960  Methoxyethyl acrylate  −2.10  1  0.249  0.80  0.00  0.80  3.876  481.45  9.204  3.357  Pyridine  −2.75  1  0.613  0.84  0.00  0.52  3.022  318.53  4.941  2.428  Butanone  −3.38  1  0.166  0.70  0.00  0.51  2.287  328.03  5.595  2.291  2-Ethylhexan-1-ol  −0.57  1  0.209  0.39  0.37  0.48  4.433  546.11  8.886  3.344  iso-Butanol  −2.36  1  0.217  0.39  0.37  0.48  2.413  342.97  5.314  2.301  Butan-1-olb  −2.13  1  0.224  0.42  0.37  0.48  2.601  350.58  6.554  2.423  Diethylaminopropriontrile  −0.83  1  0.267  0.89  0.00  0.86  4.479  515.76  7.571  3.337  Hexan-1-olb  −1.13  1  0.210  0.42  0.37  0.48  3.610  457.78  9.053  3.061  Propanoneb  −3.66  1  0.179  0.70  0.04  0.49  1.696  275.29  4.316  1.912  Ethyleneglycolmonobutyl ether  −1.32  1  0.201  0.50  0.30  0.83  3.806  491.96  10.074  3.289  4-Fluoroaniline  −1.06  1  0.760  1.09  0.20  0.40  4.007  376.72  6.105  2.950  Cyclohexanol  −1.00  1  0.460  0.54  0.32  0.57  3.758  398.55  4.889  2.704  Propanoneb  −5.27  0  0.179  0.70  0.04  0.49  1.696  275.29  4.316  1.912  Pentan-2-one  −4.05  0  0.143  0.68  0.00  0.51  2.755  382.08  6.819  2.648  Heptan-2-oneb  −2.49  0  0.123  0.68  0.00  0.51  3.760  489.92  9.321  3.229  Nonan-2-one  −2.35  0  0.119  0.68  0.00  0.51  4.735  597.48  11.813  3.686  Ethyl acetateb  −4.69  0  0.106  0.62  0.00  0.45  2.314  361.36  6.375  2.606  Butyl acetateb  −2.87  0  0.071  0.60  0.00  0.45  3.353  470.17  8.831  3.196  Hexyl acetate  −2.41  0  0.056  0.60  0.00  0.45  4.351  577.05  11.307  3.663  Octyl acetate  −2.02  0  0.029  0.60  0.00  0.45  5.364  684.18  13.795  4.043  Decyl acetate  −1.30  0  0.033  0.60  0.00  0.45  6.373  793.03  16.298  4.362  Ethanolb  −4.76  0  0.246  0.42  0.37  0.48  1.485  242.38  4.054  1.510  Propan-1-ol  −3.74  0  0.236  0.42  0.37  0.48  2.031  296.30  5.313  2.016  Butan-1-olb  −3.37  0  0.224  0.42  0.37  0.48  2.601  350.58  6.554  2.423  Hexan-1-olb  −2.60  0  0.210  0.42  0.37  0.48  3.610  457.78  9.053  3.061  Octan-1-olb  −1.71  0  0.199  0.42  0.37  0.48  4.619  565.90  11.552  3.548  Tolueneb  −4.41  0  0.601  0.52  0.00  0.14  3.325  384.83  5.889  2.719  Ethylbenzene  −3.87  0  0.613  0.51  0.00  0.15  3.778  432.04  7.284  2.981  Propylbenzene  −3.43  0  0.604  0.50  0.00  0.15  4.230  486.32  8.368  3.242  Cumene  −3.39  0  0.602  0.49  0.00  0.16  4.084  480.18  7.200  3.160  p-Cymene  −3.11  0  0.607  0.49  0.00  0.19  4.590  532.53  8.190  3.360  d-3-Carene  −3.30  0  0.511  0.22  0.00  0.10  4.649  529.88  7.470  3.280  Linalool  −2.55  0  0.398  0.55  0.20  0.67  4.794  613.77  10.000  3.680  1,8-Cineole  −2.15  0  0.383  0.33  0.00  0.76  4.688  543.92  6.400  3.330  Geraniol  −1.35  0  0.513  0.63  0.39  0.66  5.479  619.88  11.470  3.760  Dodecane  1.00  0  0.000  0.00  0.00  0.00  5.695        View Large TABLE 2 Prediction of SP Values for the 46 Compound Test Set from Equation 10 Statistic  Value  Average deviation (AD)  −0.037  Average absolute deviation (AAD)  0.345  Standard deviation  0.430  Statistic  Value  Average deviation (AD)  −0.037  Average absolute deviation (AAD)  0.345  Standard deviation  0.430  View Large TABLE 3 Values of log Poct for Some Isomeric Compounds (Leo, 2002) Compound  log Poct  Hexane  3.90  2-Methylpentane  3.73  2,2-Dimethylbutane  3.82  Hexan-1-ol  3.23  Hexan-2-ol  3.07  Hexan-3-ol  2.98  o-Methylphenol  1.97  m-Methylphenol  1.98  p-Methylphenol  1.97  o-Hydroxybenzamide  1.28  m-Hydroxybenzamide  0.39  p-Hydroxybenzamide  0.33  Compound  log Poct  Hexane  3.90  2-Methylpentane  3.73  2,2-Dimethylbutane  3.82  Hexan-1-ol  3.23  Hexan-2-ol  3.07  Hexan-3-ol  2.98  o-Methylphenol  1.97  m-Methylphenol  1.98  p-Methylphenol  1.97  o-Hydroxybenzamide  1.28  m-Hydroxybenzamide  0.39  p-Hydroxybenzamide  0.33  View Large TABLE 4 Regression Coefficients in Equation 7 for Gas-Solvent (Phase) Partitions at 298 K Phase  e  s  a  b  l  aAt 310 K.  Biophase, for EIT  −0.44  2.02  4.02  1.15  0.787  Water  0.82  2.73  3.90  4.81  −0.213  Wet 1-octanol  0.00  0.71  3.52  1.43  0.858  Wet chloroform  −0.47  1.20  0.14  1.43  0.994  Dry acetone  −0.27  1.52  3.26  0.08  0.863  Dry N,N′-dimethyl formamide  −0.19  2.33  4.76  0.00  0.808  Dry N-methylformamide  −0.26  2.06  4.56  0.43  0.706  Dry tetraethylene glycol  0.21  1.88  4.64  0.31  0.584  Plant matrix  0.08  1.28  3.12  0.82  0.860  Braina  0.43  0.29  2.78  2.79  0.609  Musclea  0.54  0.22  3.47  2.92  0.578  Plasmaa  0.49  2.05  3.51  3.91  0.157  Phase  e  s  a  b  l  aAt 310 K.  Biophase, for EIT  −0.44  2.02  4.02  1.15  0.787  Water  0.82  2.73  3.90  4.81  −0.213  Wet 1-octanol  0.00  0.71  3.52  1.43  0.858  Wet chloroform  −0.47  1.20  0.14  1.43  0.994  Dry acetone  −0.27  1.52  3.26  0.08  0.863  Dry N,N′-dimethyl formamide  −0.19  2.33  4.76  0.00  0.808  Dry N-methylformamide  −0.26  2.06  4.56  0.43  0.706  Dry tetraethylene glycol  0.21  1.88  4.64  0.31  0.584  Plant matrix  0.08  1.28  3.12  0.82  0.860  Braina  0.43  0.29  2.78  2.79  0.609  Musclea  0.54  0.22  3.47  2.92  0.578  Plasmaa  0.49  2.05  3.51  3.91  0.157  View Large FIG. 1. View largeDownload slide The relationship between (A) liquid to vapor and liquid to solvent transfer and (B) liquid to vapor and liquid to biophase transfer. S is the solubility of a pure liquid in a solvent, P° is the pure liquid saturated vapor pressure, K is the gas-to-solvent equilibrium constant, and MMAS is the Draize eye score. FIG. 1. View largeDownload slide The relationship between (A) liquid to vapor and liquid to solvent transfer and (B) liquid to vapor and liquid to biophase transfer. S is the solubility of a pure liquid in a solvent, P° is the pure liquid saturated vapor pressure, K is the gas-to-solvent equilibrium constant, and MMAS is the Draize eye score. FIG. 2. View largeDownload slide Plot of observed SP versus calculated SP on Equation 8; SP = log (MMAS/P°) or log (1/EIT). ▴ log (MMAS/P°); 0 log (1/EIT). FIG. 2. View largeDownload slide Plot of observed SP versus calculated SP on Equation 8; SP = log (MMAS/P°) or log (1/EIT). ▴ log (MMAS/P°); 0 log (1/EIT). 1 To whom correspondence should be addressed at University College London, Department of Chemistry, 20 Gordon Street, London WC1H 0AJ, UK. Fax: +44 (20) 7679-7463. E-mail: m.h.abraham@ucl.ac.uk. This study was supported by research grants R01 DC 02741 and DC 005003 from the National Institute on Deafness and other Communication Disorders, National Institutes of Health. M. H. thanks the British Council for a scholarship. REFERENCES Abraham, M. H. ( 1993). Scales of hydrogen bonding: Their construction and application to physicochemical and biochemical processes. Chem. Soc. Rev.  22, 73–83. Google Scholar Abraham, M. H., and Al-Hussaini, A. J. M. ( 2002). Solvation descriptors for N-nitroso dialkylamines: Calculation of some of their properties of environmental significance. J. Environ. Monit.  4, 743–746. Google Scholar Abraham, M. H., Andonian-Haftvan, J., Whiting, G. S., Leo, A., and Taft, R. W. ( 1994). Hydrogen bonding. Part 34. The factors that influence solubility of gases and vapors in water at 298 K, and a new method for its determination. J. Chem. Soc. Perkin Trans.  2, 1777–1791. Google Scholar Abraham, M. H., Gola, J. M. R., Cometto-Muñiz, J. E., and Cain, W. S ( 2000). Connection between chromatographic data and biological data. J. Chromatogr. B  745, 103–115. Google Scholar Abraham, M. H., Gola, J. M. R., Cometto-Muñiz, J. E., and Cain, W. S ( 2001). The correlation and prediction of VOC thresholds for nasal pungency, eye irritation, and odor in humans. Indoor Built Environ.  10, 252–257. Google Scholar Abraham, M. H., Gola, J. M. R., Cometto-Muñiz, J. E., and Cain, W. S. ( 2002). A model for odor thresholds. Chem. Senses  27, 95–104. Google Scholar Abraham, M. H., Kumarsingh, R., Cometto-Muñiz, J. E., and Cain, W. S. ( 1998). A quantitative structure-activity relationship (QSAR) for a Draize eye irritation data base. Toxicol. In Vitro  12, 201–207. Google Scholar Abraham, M. H., Kumarsingh, R., Cometto-Muñiz, J. E., and Cain, W. S. ( 1998). Draize eye scores and eye irritation thresholds in man can be combined into one quantitative structure-activity relationship. Toxicol. In Vitro  12, 403–408. Google Scholar Abraham, M. H., and Weathersby, P. K. ( 1994). Hydrogen bonding. 30. Solubility of gases and vapors in biological liquids and tissues. J. Pharm. Sci.  83, 1450–1456. Google Scholar Allgood, G. S. ( 1989). Use of animal eye test data and human experience for determining the ocular irritation potential of shampoos. J. Toxicol. Cutaneous Ocul. Toxicol.  8, 321–326. Google Scholar Barratt, M. D. ( 1995). A quantitative structure-activity relationship for the eye irritation potential of neutral organic chemicals. Toxicol. Lett.  80, 69–74. Google Scholar Barratt, M. D. ( 1997). QSARs for the eye irritation potential of neutral organic molecules. Toxicol. In Vitro  11, 1–8. Google Scholar Brantom, P. G., Bruner, L. H., Chamberlain, M., De Silva, O., Depuis, J., Earl, L. K., Lovell, D. P., Pape, W. J. W., Uttley, M., Bagley, D. M., et al. ( 1997). A summary report of the COLIPA international validation study on alternatives to the Draize rabbit eye irritation test. Toxicol. In Vitro  11, 141–179. Google Scholar Chamberlain, M., and Barratt, M. D. ( 1995). Practical applications of QSAR to in vitro toxicology illustrated by consideration of eye irritation. Toxicol. In Vitro  9, 543–547. Google Scholar Cometto-Muñiz, J. E., and Cain, W. S. ( 1991). Nasal pungency, odor, and eye irritation thresholds for homologous acetates. Pharmacol. Biochem. Behav.  39, 983–989. Google Scholar Cometto-Muñiz, J. E., and Cain, W. S. ( 1995). Relative sensitivity of the ocular trigeminal, nasal trigeminal, and olfactory systems to airborne chemicals. Chem. Senses  20, 191–198. Google Scholar Cometto-Muñiz, J. E., and Cain, W. S. ( 1998). Trigeminal and olfactory sensitivity: Comparison of modalities and methods of measurement. Int. Arch. Occup. Environ. Health  71, 105–110. Google Scholar Cometto-Muñiz, J. E., Cain, W. S., and Abraham, M. H. ( 1998). Nasal pungency and odor of homologous aldehydes and carboxylic acids. Exp. Brain Res.  118, 180–188. Google Scholar Cometto-Muñiz, J. E., Cain, W. S., Abraham, M. H., and Kumarsingh, R. ( 1998). Sensory properties of selected terpenes. Thresholds for odor, nasal pungency, nasal localization, and eye irritation. Ann. NY Acad. Sci.  855, 648–651. Google Scholar Cometto-Muñiz, J. E., Cain, W. S., Abraham, M. H., and Kumarsingh, R. ( 1998). Trigeminal and olfactory chemosensory impact of selected terpenes. Pharmacol. Biochem. Behav.  60, 765–770. Google Scholar Cometto-Muñiz, J. E., Cain, W. S., and Hudnell, H. K. ( 1997). Agonistic sensory effects of airborne chemicals in mixtures: Odor, nasal pungency, and eye irritation. Percept. Psychophys.  59, 665–674. Google Scholar Cronin, M. T., Basketter, D. A., and York, M. ( 1994). A quantitative structure-activity relationship (QSAR) investigation of a Draize eye irritation database. Toxicol. In Vitro  8, 21–28. Google Scholar Draize, J. H., Woodward, G., and Calvery, H. O. ( 1944). Methods for the study of irritation and toxicity of substances applied topically to the skin and mucous membranes. J. Pharmacol. Exp. Ther.  82, 377–390. Google Scholar ECETOC manual No. 48(2) ( 1998). Eye Irritation Reference Chemicals Data Bank, 2nd ed. ECETOC, Brussels. Google Scholar Enslein, K. ( 1988). An overview of structure-activity relationships as an alternative to testing in animals for carcinogenicity, mutagenicity, dermal and eye irritation, and acute oral toxicity. Toxicol. Ind. Health  4, 479–498. Google Scholar Freeburg, F. E., Nixon, G. A., Reer, P. J., Weaver, J. E., Bruce, R. D., Griffith, J. F., and Sanders, L. W. ( 1986). Human and rabbit eye responses to chemical insult. Fundam. Appl. Toxicol.  7, 626–634. Google Scholar Griffith, J. F. ( 1989). Use of human experience to calibrate Draize and in vitro eye test data. J. Toxicol. Cutaneous Ocul. Toxicol.  8, 23–34. Google Scholar HyperChem software, version 6.01 ( 2000). Hypercube Inc., Gainsville, FL. Google Scholar Klopman, G., Ptchelintsev, D., Frierson, M., Pennisi, S., Renssskers, K., and Dickens, M. ( 1993). Multiple computer automated structure evaluation methodology as an alternative to in vivo eye irritation testing. Altern. Lab. Anim.  21, 14–27. Google Scholar Kulkarni, A. S., and Hopfinger, A. J. ( 1999). Membrane-interaction QSAR analysis: Application to the estimation of eye irritation by organic compounds. Pharm. Res.  16, 1245–1253. Google Scholar Leo, A. J. ( 2002). The MedChem database, BioByte Corp. and Pomona College. Daylight Chemical Information Systems, Mission Viejo, CA. Google Scholar Patlewicz, G. Y., Rodford, R. A., Ellis, G., and Barratt, M. D. ( 2000). A QSAR model for the eye irritation of cationic surfactants. Toxicol. In Vitro  14, 79–84. Google Scholar Platts, J. A., and Abraham, M. H. ( 2000). Partition of volatile organic compounds from air and from water into plant cuticular matrix: An LFER analysis. Environ. Sci. Technol.  34, 318–323. Google Scholar Roggeband, R., York, M., Pericoi, M., and Braun, W. ( 2000). Eye irritation response in rabbit and man after single applications of equal volumes of undiluted model liquid detergent products. Food Chem. Toxicol.  39, 727–734. Google Scholar Spielmann, H., Liebsch, M., Kalweit, S., Moldenhaur, F., Wirnsberger, T., Holzhutter, H.-G., Schneider, B., Glaser, S., Gerner, I., Pape, W. J. W., et al. ( 1998). Results of a validation study in Germany on two in vitro alternatives to the Draize eye irritation test, the HET-CAM test and the 3T3 NRU cytotoxicity test. Altern. Lab. Anim.  24, 741–858. Google Scholar Todeschini, R., Consonni, V., and Pavan, M. ( 2002) Dragon software, version 2.1. Milan, Italy. Google Scholar Wilhelmus, K. R. ( 2001). The Draize eye test. Surv. Ophthalmol.  45, 493–515. Google Scholar York, M., and Steiling, W. ( 1998). A critical review of the assessment of eye irritation potential using the Draize rabbit eye test. J. Appl. Toxicol.  18, 233–240. Google Scholar © 2003 Society of Toxicology

Journal

Toxicological SciencesOxford University Press

Published: Dec 1, 2003

There are no references for this article.