[Show abstract][Hide abstract] ABSTRACT:
Protein/protein interactions (PPIs) are essential for all biological processes. Their role as central control switches and checkpoints in signaling and regulation makes their dysfunction an eminent cause of cancer and other diseases. Thus, a target-oriented intervention with PPIs, in particular by small molecules suitable for a pharmacological therapy, is the object of intense research.
In the recent years, this research of PPIs yielded a number of low-molecular protein/protein interaction modulators (PPIMs). However, only a few PPIMs were identified by rational or structure-based considerations. The reason is that it is difficult to directly re-use well-established computational methods to identify ligands of conventional targets, such as enzymes, receptors, or transporters.
Conventional targets naturally bind small molecules via pronounced complementary binding pockets. In contrast, many protein/protein interfaces (PPIfaces) are large and lack pronounced pockets. Thus, it is hard to find small molecules with an affinity and specificity that is adequate to displace one of the binding proteins. Nevertheless, the widespread discovery of (drug-like) PPIMs shows: these challenges can be overcome, and PPIs are not undruggable in general.
The goal of this thesis was to develop a computational strategy for the rational identification of PPIMs, notably of NHR2 inhibitors, starting only with a protein/protein complex structure (PPI structure). This task subdivides into the four core themes of this thesis.
First, I reviewed the knowledge about PPIs, PPIMs, the determinants of their interactions as well as computational methods for the identification of druggable sites and PPIMs. Second, I implemented a strategy that uses a PPI-structure based prediction of hot spots and transient pockets to guide a structure-based virtual screening. As a validation, I retrieved known PPIMs that bind to the PPIface of interleukin-2 from a large set of non-binders. Third, I analyzed the potential of teroxazoles as a new class of hydrophilic α-helix mimetics. These present side chains similar to α-helices and mimic a sequence pattern that has not yet been considered. Fourth, I identified the first micromolar inhibitors of the NHR2-mediated tetramerization of RUNX1-ETO (RE) from dimers, a prerequisite for the onset and maintenance of RE-dependent acute myeloid leukemia (RE-AML). Predicted tetramerization hot spots close to the largest pocket in the PPIface guided the computational identification of drug-like PPIMs. These PPIMs mimic NHR2 hot spots, aim at the PPIface, and inhibit dimer association as well as the proliferation of RE-dependent cells.
These PPIMs are valuable as probes and tools to study the effects of NHR2 tetramerization and are an important step towards a personalized therapy of RE-AML. Most importantly, however, the presented strategy can well be the first step in any comparable structure-based endeavor to identify or design PPIMs, even in cases where only a PPI structure with a rather flat PPIface is known.
10/2014, Degree: Dr. rer. nat. (PhD), Supervisor: Holger Gohlke
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed.
The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual
current impact factor.
Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence
agreement may be applicable.