Visit the C&I Facebook page to join the debate
Highlights
Kevin Burgess Texas A&M University, US
Biomedical
Small molecule allosteric inhibitor of the HIF-2α/ARNT interaction Under hypoxic conditions, cells assemble heterodimeric complexes of the HIF-2α transcription factors (1 – 3) and a binding partner called ARNT. Overproduction of the HIF proteins correlates with a variety of cancer states and, conversely, inhibition of the protein-protein interactions (PPIs) that enable these complexes to form can impair progression of tumors. Gardner, Bruik, and co-workers report on the structure 1 identified from an alpha-screen to monitor disruption of the HIF-2α/ARNT complex (Figure 1) (Nature Chem. Bio., doi:10.1038/nchembio.1185). They used gel assays to prove this compound disrupts HIF-2α/ARNT assembly, but not the corresponding one involving HIF-1α. They also looked at the effects of 1 on mRNA levels corresponding to proteins known to be upregulated by HIF-2α and found those protein expression events to be suppressed. In X-ray structural studies compound 1 was found to bind an allosteric site, not a region directly in the ARNT interface region.
Peptidic inhibitor of the HIF-1α / p300 interaction Transcriptional activation via HIF proteins actually proceeds via a complex involving several protein components; there is an interaction between HIF-1α and a coactivator protein called p300. Arora and Olenyuk (Proc. Nat. Acad. Sci., 10.1073/pnas.1312473110) have investigated peptidic mimics of a helix in HIF-1α that interacts with p300. The mimics are based on Arora’s C-terminal cap that involves a metathesis- formed N-to-4-butenylpropeonate alkene strap that replaces the first C-terminal H-bond in the helix with an immutable trans alkene; this constraint promotes helical character in the peptide that totally consists of eight amino acid residues. Fluorescence polarisation assays were used to monitor binding of a labeled segment of HIF-1α to an appropriate
p300 domain. That interaction had a Kd of 38nM, whereas the peptide mimic was shown to binds that same domain with a Kd of 690nM. HSQC-NMR titration experiments proved that the helical mimic binds to a pertinent region of p300 in this interaction. At 50μM concentrations, the test peptide mimic caused a 25% reduction in luciferase production in MDA-MB-231 triple negative breast cancer cells, which do not express estrogen, progesterone, or epidermal growth factor, engineered to couple HIF-1α production of VEGF with generation of luciferase. Gene profiling experiments showed that this peptide reduced expression of a string of proteins that are also associated with HIF-1α activation. In vivo, and unsurprisingly, peptide mimic had better stability than a corresponding control peptide. In an appropriate tumor model, when the featured peptide was injected five times over a month (see SI for that information) at a dose of 13mg/Kg each time, a 53% reduction in the tumor volume was achieved.
O
N N
NO2 CI
N H
1
Figure 1 The HIF-2α/ARNT complex
Selective inhibitors of the Mcl-1/ BH3 peptides interaction Each of the Bcl-2 protein family engages other proteins via binding of their helical BH3 domain regions in long binding cavities. This type of protein-protein interaction (PPI) has emerged as a prime hunting ground for finding small molecule inhibitors, but in the particular case of Mcl-1 (myeloid cell leukemia) selective inhibitors have yet to reach the clinic. This is unfortunate because upregulation of Mcl-1 expression is a shared characteristic in many cancer types, and is apparently a feature of sub- types that are resistant to the most widely used drugs (eg gemcitabine-resistance pancreatic cancer).
Figure 2 Cl 2
N H
Conclusion The three papers above illustrate three different approaches to making synthetic compounds to perturb PPIs. The first is via a high throughput screen, the second is via peptidic molecules, and the third via a combination of a large- scale screen, fragments based-design, and then structure-based design via crystallography. The first and third methods were expensive and time consuming. Work with the peptidic mimic illustrated all the facets of inhibition of a PPI up to a higher level, and it is somewhat surprising that it permeated into cells, but in vivo five doses of material were required over a month to retard the tumor growth by only 53% versus a control. Some protein-protein interactions are tractable targets, but there is certainly room for further innovation in this field to streamline the process and improve the outcome.
Cl O
binds Mcl-1 with 55 nM affinity
Chemistry&Industry • November 2013 53 F
Fesik et al (J. Med. Chem., 2013, 56, 15) used labeled Mcl-1 protein to screen a 13,800 small molecule fragment library in groups of 12 via NMR; an expensive process! They selected 93 initial hits, then triaged and classified these via a fluorescence polarisation assay and intuitive evaluation of the structures involved. Double-labeled (13C and 15N) Mcl-1 in NOE experiments, combined with simulated annealing procedures, showed possible binding modes of molecules in each class. The most difficult part of fragment- based design is joining the fragments with linkers that give an entropic advantage yet do not disrupt binding of the fragments. These researchers used a flexible linker to do this, then resorted to SAR and structure-based design via crystallography for optimisation. Out of this work came many lead compounds, a representative one being structure 2 that has a 55nM Ki for Mcl-1 and bound this target with a 16-fold selectivity factor relative to Bcl2 and a 270-fold more strongly than Bcl-xL.
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64