Lush is a soluble odorant-binding protein of the fruit fly Drosophila. Mutants not expressing this protein have been reported to lack the avoidance behaviour, exhibited by wild type flies, to high concentrations of ethanol. Very recently, the three-dimensional structure of Lush complexed with short-chain alcohols has been resolved, supporting a role for this protein in binding and detecting small alcohols. Lush does not bind ethanol and wild type flies are in fact attracted by high concentrations of ethanol. Lush binds some phthalates and flies are repelled by such compounds. Finally, fluorescence data, interpreted in the light of the three-dimensional structure of Lush, indicate that the protein undergoes a major conformational change, similar to that reported for the pheromone-binding protein of Bombyx mori, but triggered, in the case of Lush, by ligand (Zhou, 2004).

The behaviour was examined of wild type D. melanogaster to different concentrations of ethanol as well as to dibutyl phthalate, which proved to be the best ligand of Lush. Initially the flies were tested with traps containing food (mashed rotten apple) mixed with ethanol at final concentrations of 1% and 50%. The flies clearly preferred the higher ethanol concentration. To avoid effects of possible contaminants, the absence of organic compounds particularly of phthalates, was verified in the sample of ethanol by GC/MS. Since in these conditions no peak of contaminants was detectable, it is estimated that the concentration of any of such compounds, if present, would be lower than 10 ppm (Zhou, 2004).

Similar experiments were performed with pairs of traps containing dibutyl phthalate at the final concentration of 1 mM, diluted in mashed apples or in 50% ethanol in agarose. In each case the same diluent (mashed apple or 50% ethanol in agarose) was used in the control trap. The flies showed a statistically significant (P<0.001) behaviour of avoidance to dibutyl phthalate, regardless of the medium used (Zhou, 2004).

These behaviour results and the binding affinity of Lush to some phthalates provide a new interpretation of the phenomenon previously observed in previous studies. The original hypothesis stated that Lush could modify the perception of ethanol at high concentrations. This requires some specific binding of ethanol to Lush, which in the current study was not detected. It can be argued that a second binding site could be present on the protein, where ethanol could bind, without affecting the binding of 1-NPN in the other pocket. This hypothesis, however, is not supported by the recently published X-ray structure, which shows the presence of a single binding site in Lush, in agreement with previous reports for other insect OBPs. Moreover, the behaviour experiments performed with pure ethanol clearly show that this compound is a potent attractant for wild type flies at high concentrations (Zhou, 2004 and references therein).

A reinterpretation of previous results, assuming that Lush is specific for aromatic compounds structurally similar to phthalates, rather than for ethanol, provides the first and only evidence that insects use OBPs as tools to distinguish different odorants (Zhou, 2004).

This finding also suggests that binding experiments could be used as a first screening to indicate which OBP gene should be deleted or inactivated in order to get mutants with desired types of anosmia (Zhou, 2004).

In the light of these results, it is concluded that the avoidance previously reported for wild type flies could be due to impurities, such as phthalates or structurally related compounds, present as contaminants in the ethanol used. Furthermore, the loss of avoidance in the mutants would result from the inability to detect large aromatic compounds. This conclusion provides new and increased interest for the previously published data, since Lush would be directly and strictly required for the perception of an odorant, rather than being involved only in modulating the response to ethanol. The recently published structure of Lush with a molecule of butanol in its binding pocket does not, in light of the current study, show a true binding role of this protein for small alcohols (Zhou, 2004).

The C-terminus of Lush, which contains the only tryptophan residue of the protein, is located inside the core of the protein. When the effect of ligands on the tryptophan fluorescence was investigated, it was not possible to measure any quenching up to ligand concentrations of 16 microM. This could indicate that when the ligands (1-NPN or phthalate) enter the binding site, the C-terminus of the protein is displaced and the tryptophan residue moves outside the core of the protein. Such a major conformational change has been observed with the pheromone-binding protein of Bombyx mori, as an effect of a pH increase from 4.5 to 6.5. In the current case, the effect would occur at pH 7 and, most interestingly, as a consequence of ligand binding. This hypothesis, if verified in the structure of Lush complexed with a ligand, would represent the first example of a major conformational change of an OBP related to substrate binding and have interesting implications for the mode of action of these proteins and their interactions with membrane-bound receptorswe investigated (Zhou, 2004).