Do Amino-Oxetanes Resemble Amides? A Matched Molecular Pairs Property and Structural Comparison
J. Med. Chem. ASAP
Oxetanes have long been used in medicinal chemistry campaigns, yet have remained a controversial choice in design due to the lack of this chemotype amongst marketed drugs, until now. This controversy has largely been settled, thanks to the recent FDA-approval of the oral BTK inhibitor: rilzabrutinib, offering a degree of “proof-of-concept” that not even the most sceptical among us can deny.
Figure 1: A – The Oxetane ring, with calculated σp and π values indicating mild electron donation and hydrophobicity-addition, respectively (https://peter-ertl.com/molecular/substituents/craigplot.html). B – Proposed isosteric replacement – the topic of this paper.
Oxetanes represent a convenient R group to try in certain circumstances: a low-profile, hydrophilic, saturated heterocycle with great potential for hydrogen-bonding interactions and only a very small molecular polarisation effect (calculated Hammett σp = -0.03).
Seminal publications from Roche ((1) doi.org/10.1002/anie.200602343, (2) doi.org/10.1021/jm9018788) have revealed the potential for oxetanes to serve as isosteres for a number of functional groups and linker moieties, such as gem-dimethyl groups and amides. As an isostere for amides, 3-amino-oxetanes offer an attractive profile with several key similarities and differences. Most notably: the amide conjugation between the nitrogen and the oxygen is broken, this leaves sp3-hybridised oxygen lone pairs rather than sp2-hybridised, but they are oriented similarly if the structures are aligned. This change also eliminates a potential proteolytic soft-spot and introduces a mild base where a neutral amide once was.
The authors of this paper noted this potential and realised that a systematic appraisal of all the important factors involved in assessing whether one group can act as an “isostere” for another, did not exist. Moreover, synthetic access to amino-oxetanes has drastically improved since the amide isostere question first appeared. This has allowed a curated matched molecular pair (MMP) dataset to be made and tested back-to-back.
So, they set about measuring physicochemical (logD, solubility, stability), metabolic (Human Hepatocyte clearance) and conformational (bulk CSD analysis, DFT/NBO analysis) properties of each. The latter data (conformational properties) is arguably the most important, as it predicts how stable the bioactive conformation is in either case. After all, an isostere that can’t happily adopt the same conformation in the required setting can’t really be called an isostere, can it?
Their findings suggest that, compared to amides, amino-oxetanes are broadly similar in lipophilicity and also chemical and metabolic stability. The solubility of amino-oxetanes is much greater than amides at low pH, due to amine protonation. Importantly however, conformations of substituted amides cannot easily be attained by the equivalent amino-oxetanes, with the latter showing high distortion energies when placed in the bioactive amide conformation from a solvated minimum. This led to the suggestion that amino-oxetanes might be better isosteres for sulfonamides, particularly where a less polar equivalent is required for CNS entry.
Studies like these produce data that is highly valuable to the medicinal chemistry community, informing key decisions on whether to embark on potentially long synthetic routes, or not. Perhaps we will soon see another MMP study on the implications of sulfonamide→amino-oxetane, rather than amide→amino-oxetane.