No abstract

A classification model based on 36 compounds inducing allergic contact dermatitis and 36 non‐allergens was developed by the discriminant technique of two‐value multiple regression analysis. One outlier was observed and several of the compounds fell in the indeterminate region between the classes. Based on variable results of clinical data (human and animal), some overlap in the model was considered more reasonable than perfect separation. The descriptors in this model clearly imply intercellular transport (LogP derived) and various forms of binding (London forces [MR] and hydrogen‐bonding [HBA, HBD]) as part of the allergic contact dermatitis (ACD) response mechanism. In addition, substructure descriptors related to the epidermal reactivity of the ACD inducers were statistically important. These include directly reactive functional groups from haptens and metabolizable precursor groups of prohaptens. The model was developed from carefully selected compounds of diverse structure, representing a broad range of hapten and prohapten substructures without undue molecular complexity. Validation of the model is in progress on a broad range of molecular structures far beyond those used in construction.

The electrophilic nature of some contact sensitizers, that is, chemicals that cause allergic contact dermatitis (ACD), is also characteristic of genotoxic tumorigens. Electrophiles can adduct protein, which is the basis for ACD, as well as DNA, which is the basis for mutagenicity and carcinogenicity. This suggests that some electrophilic contact sensitizers may be genotoxic tumorigens. To further investigate this matter, we evaluated 146 chemicals that had been bioassayed for tumorigenicity and mutagenicity in the National Toxicology Program, with an analysis of structure-activity relationships for contact sensitization. Using the data from this analysis and from other sources, the proportion of the contact sensitizers that were both mutagenic and tumorigenic was found to range from 20% to 28%. This finding suggests that there may be in the order of 90 genotoxic tumorigens for rodents among the approximately 384 chemicals that have been validated as contact sensitizers for humans.

The binding of a series of phenylpiperazines (3) and benzoylpiperazines (4) to central serotonin (5-HT) sites was investigated. Several derivatives of 3 displayed nanomolar affinities for 5-HT1 sites, whereas derivatives of 4 were essentially inactive both at 5-HT1 and 5-HT2 sites. 1-(2-Methoxyphenyl)piperazine (2-MPP, 3a) was found to possess an affinity (Ki = 35 nM) for 5-HT1 sites comparable to that of the recognized 5-HT agonist 1-[3-(trifluoromethyl)phenyl]piperazine (TFMPP) (Ki = 20 nM); 3a also displayed a 100-fold selectivity for 5-HT1 sites (as compared to 8-fold for TFMPP). In tests of stimulus generalization using rats trained to discriminate TFMPP (ED50 = 0.17 mg/kg) from saline, 3a was found to be nearly equipotent (ED50 = 0.22 mg/kg) with the training drug. These results suggest that 3a may be a novel and more selective 5-HT1 agonist than TFMPP.

No abstract