Background: Prior estimation of toxicity of each and every, existing and yet to be synthesized
chemicals is a must to elude their adverse effect on the environment. Experimental determination
of such parameters is time consuming, cost effective and above all, it demands the sacrifice of many
vertebrates. At this end, the REACH regulations advocate for the use of non-testing predictive methods
such as read-across, weight-of-evidence and QSAR (quantitative structure–activity relationship) techniques.
Among these methods, QSAR is found to be the best as it is based on molecular structure only.
The descriptors used in deriving the model in QSAR vary according to the nature of the narcotics as
well as the species used for. The success of a model in predicting the toxicity of a narcotic purely depends
on the type of descriptors selected that explains the structural features closely related to the
property under study. In this review, we have focused on the different types of descriptors and QSAR
models used to explain the narcosis phenomenon.
Methods: Literature was scanned for acute toxicity of chemicals on species like tadpoles, protozoa,
planktonic crustaceans, and small fishes like million fish, rainbow fish etc. from different sources. The
toxicity and toxicants were classified considering their polarity and specific interactions of the compounds.
Due to complex nature of the substrate, the mechanism of action of toxicant is uncertain.
However, the overall results obtained from the biological study have been subjected to QSAR studies
to obtain various models, which can provide some ideas on the mode of toxicological action. Different
types of molecular descriptors derived both experimentally and theoretically have been used in the
Results: Mostly biochemicals have a specific signature on oil/water partition (Ko/w, P), which is the
crux in biological activity. Accordingly, the toxicological activities have good correlations with log P.
Addition of some more structural descriptors improves the structure-toxicity relationship. Among these,
electronic descriptors like EHOMO, ELUMO and E derived from molecular orbitals have been used in
the QSAR. ELUMO describing the energy of excited species of the molecule is found to be the most suitable
one. Other molecular descriptors used in the QSAR include constitutional, topological and Abraham’s
solute descriptors. The models derived from the QSAR studies were found to be highly significant
to predict the toxicology as well as to throw light on the mechanism.
Conclusion: The best descriptor for aquatic narcosis is the KO/W or P. Addition of an electronic parameter
(ELUMO) improves the QSAR to some extent. However, substitution of ELUMO by other class of molecular
descriptors has also some statistical significance. To have a global QSAR model, in addition to
P, some more appropriate descriptors are to be derived either experimentally or theoretically, latter being
the more cost effective and easy in derivation.