Click Chemistry: Novel Applications in Cell Biology and Drug Discovery
The contribution of modern synthetic methods to medicinal chemistry has been marginal over the past couple of decades. Among the top 20 reactions used, only two new ones have made it to the standard arsenal of medicinal chemists within the last 20 years.[1] In contrast, modern synthetic methods have been widely applied to chemical biology, which uses chemical probes to answer biological questions. An impres- sive example is click reactions,[2] which do not play a signifi- cant role in medicinal chemistry but have had a massive impact on chemical biology.[3] Labeling proteins using click chemistry has become a standard method, and commercial click chemistry kits for diverse labeling applications are now available from standard suppliers.[4]
In their article “Click chemistry enables preclinical evalua- tion of targeted epigenetic therapies”, Dean S. Tyler et al. recently demonstrated the increasing potential of click chemis- try for labeling experiments in a cellular environment.[5] Classical functional tags such as biotin or fluorophores and the linkers (e.g., PEG) used to attach these labels have a major impact on the physicochemical, pharmacokinetic, and phar- macodynamic properties of small molecules. Data obtained with tagged compounds might therefore be biased, which hampers accurate interpretation. Moreover, these molecules are of limited general utility for in vivo applications. Tyler et al. describe a distinct approach employing clickable inhibitors that were covalently labeled with various tags only after reaching their biological
target, BRD4, in vitro and in vivo.
Bromodomain-containing protein 4 (BRD4) belongs to the family of Bromodomain and extraterminal domain (BET) proteins, which are important regulators of transcription. As key epigenetic readers, BET proteins bind chromatin, where they recognize and respond to the acetylation state of lysines on the surface of histone proteins. BET inhibitors have been suggested as therapeutics for the treatment of inflammatory and oncological disorders.[6]
The clickable probe molecules used in this study were derived from the known BRD4 inhibitors JQ1[7] and IBET-762,[8] which were supplemented with click linkers of small (i.e., propargyl) to moderate (i.e., trans-cyclooctenyl[9]) size (Figure 1). The modified inhibitors were shown to be biologically equivalent to their parent molecules by pheno- typic assays and comparative analysis of gene expression patterns. The click linkers enabled copper-promoted azide– alkyne cycloaddition[10] and copper-free inverse electron- demand Diels–Alder click reactions[9] with fluorescent and affinity labels inside cells, which were used to unravel the molecular and physiological fate of the probes.
The authors found that both JQ1 and IBET-762 share an identical mode of action with the structurally distinct inhibitor IBET-151,[11] which is the reason for known cross-resistances between these inhibitor classes.[11b] The enrichment of JQ1 at genes that are immediately downregulated upon exposure was demonstrated in different cell lines by a technique termed “click sequencing” (click-seq). These experiments suggested different binding modes of BRD4 at genes that respond to BET inhibitors compared to those that do not respond. Furthermore, it was shown that JQ1 preferably binds to the BD2 domain of BRD4 while the BD1 domain remains engaged in interactions with chromatin. Consequently, tar- geting the BD1 domain, which might cause more efficient displacement of BRD4 from chromatin, has been suggested as a novel therapeutic strategy.Tyler et al. used fluorescent tags to elucidate the physio- logical behavior of the probes through fluorescence microsco- py and flow cytometry. In conjunction with quantitative mass spectrometry, they demonstrated that the concentration of both JQ1 and JQ1-TCO was significantly higher in spleen cells compared to cells obtained from the bone marrow of AML mice. This finding is consistent with the previous observation that cells in the latter compartment display significant resist- ance to BET inhibitors and become refractory to treatment[11], an issue that could potentially be addressed by designing inhibitors with improved bone marrow penetration. Moreover, the experiments revealed that JQ1-TCO accumulates in malignant hematopoietic cells, while concentrations in normal blood cells remain significantly lower. This might explain why therapeutic doses of BET inhibitors do not cause substantial cytopenia in heathy animals.[11]
The work of Tyler et al. impressively shows the potential of bioorthogonal chemistry for in vitro, ex vivo, and poten- tially even in vivo applications. In comparison to regular pre- labeled probes, the clickable inhibitors offer a significant benefit since the structural modifications are comparably small and biological equivalence needs to be shown only for few compounds to eventually enable the introduction of a much bigger variety of tags after distribution and target binding. Another key advantage of the method is the potential for probe labeling at defined time points. Thus it can be exploited to map (off-)target binding as well as biological distribution at the tissue, cellular, and subcellular level in a time-resolved manner, thereby providing detailed insight into the dynamic biological properties of candidate molecules. It should not remain unmentioned that the technique also offers an experimental framework for the development of tissue- or compartment-specific drugs.
In principle, this approach could be easily transferred to other target classes. However, it should be kept in mind that the introduction of tags to enable click chemistry, despite their moderate size, can affect biological properties (e.g., target affinity, off-target binding, distribution, and metabo- lism) in an unpredictable manner. Kinases in particular, but also other target classes are very susceptible to even minor structural modifications to their inhibitors, which can have a drastic influence on potency and selectivity (e.g., the flag- methyl group of imatinib[12]). The dramatic effect of minor modifications is impressively highlighted by “magic methyl” effects, where the replacement of a single hydrogen atom with a methyl group can impact potency by several orders of magnitude.[13] Moreover, the unreacted functional label itself might show unwanted behavior such as unspecific or off-target binding, which could corrupt the obtained data. Therefore, the validity of this concept needs to be shown on a case by case basis and control experiments and validation studies using orthogonal methods (e.g., radiolabeling, mass spectrometry, Raman microscopy, or classical antibody-based techniques) will be essential to avoid incorrect conclusions. However, if these considerations are carefully implemented, the approach from Tyler et al. may have the potential to significantly enhance our knowledge of drug action and distribution in living systems even in early stages of preclinical drug development. The gained knowledge could guide decision making in candidate selection, which might ultimately decrease the probability of Molibresib clinical failure.