RAPID IDENTIFICATION OF A PUTATIVE INTERACTION BETWEEN BETA 2-ADRENORECEPTOR AGONISTS AND ATF4 USING A CHEMICAL GENOMICS APPROACH

Sweta Ladwa2, Suzanne Dilly2, Andrew Clark1, Andrew Marsh1, Paul Taylor1
1University of Warwick, Coventry, United Kingdom, 2a2sp Limited, Warwick, United Kingdom

Beta2-Adrenoreceptor agonists, such as salbutamol, are widely used to treat asthma. However, the use of beta 2-agonists is controversial, since it appears likely that their mechanisms of action are complex and far from fully understood [A Panebra, M R Schwarb, C B Glinka, S B Liggett Am J Physiol, Lung Cell Mol Physiol, 2007, L453 and references therein].

We have developed new technology that allows us to uncover unknown interactions between small, biologically active molecules and polypeptide “targets”, in a chemical genomics approach. As we communicated recently [S J Dilly, M J Bell, A J Clark, A Marsh, R M Napier, M J Sergeant, A J Thompson P C Taylor Chem Commun 2007, 2808], our method combines the benefits of (i) phage display of a set of polypeptides representing a proteome, (ii) a range of different photochemical reactions that immobilise the bioactive molecule in different orientations to maximise the opportunity of specific binding to polypeptides in the library and (iii) the use of a “protein-resistant” surface to minimise non-specific binding of phage-displayed polypeptides.

Following the protocols we have described previously, we photoimmobilised salbutamol from aqueous solution by simple irradiation with 254nm light. We then exposed the surface-bound drug to a library of polypeptides derived from human lung mRNA displayed on T7 bacteriophage in three rounds of “biopanning”, amplifying the phage between rounds in E. coli before sequencing. In each round, members of the library that bind to the immobilised drug are retained, while non-binders are washed away. After the three rounds 42 clones of interest emerged.

Translation of the nucleotide sequences for the 42 selected clones in all six reading frames, to allow for the possibility of frame slippage, revealed five polypeptide sequences of reasonable length (> ca. 50 amino acids) that had significant similarity with the human proteome as judged by NCBI BLASTP.

These five clones proceeded to a competitive elution experiment wherein they were placed again in contact with salbutamol that had been immobilised by the Magic Tag® chemistry that led to each clone’s isolation and then eluted using salbutamol in solution. Of the five, one clone clearly displayed significant selectivity in its binding of salbutamol, namely a sequence corresponding to the known [L M Podust, A M Krezel Y Kim J Biol Chem 2001, 276, 505] DNA binding domain of Activating Transcription Factor 4 (ATF4, previously known as CREB2).

Intriguingly, the recent study from Panebra et al. [see above] has shown, in vitro, that the β2-agonist isoprenaline, a close analogue of salbutamol, affects the expression of ATF4 in airway smooth muscle cells that have been stressed to simulate asthmatic conditions. When IL-13, IL-4, TGF-beta, LTD4 were added to the cells to simulate asthma, production of ATF4 ceased. Addition of isoprenaline restored ATF expression to normal levels. Isoprenaline had no affect on ATF4 production under basal conditions.

Clearly, neither our study nor the recent work of Panebra et al lead to the conclusion that β2-agonists affect ATF4 levels in vivo. Nevertheless, our study shows how a chemical genomics approach can lead very rapidly to information about drug-protein interactions and the combined results from ourselves and Panebra et al. suggest that the interaction of beta 2-agonists with ATF4 is ripe for further investigation.