Modulating the nuclear receptor-cofactor interaction : characterization and inhibition

S. Möcklinghoff

Research output: ThesisPhd Thesis 1 (Research TU/e / Graduation TU/e)

926 Downloads (Pure)

Abstract

The nuclear receptor (NR) – coactivator interaction is one of the key steps in transcription control and concomitantly related to many diseases and an attractive drug target. Various ligands have been synthesized in the past to indirectly modulate this protein-protein interaction. However, despite their therapeutically value, the occurrence of undesired side effects entailed the requirement of novel approaches to address this issue. Direct inhibition of the NR – coactivator interaction by peptides and small molecules that compete with the coactivators for the NR binding groove represents an attractive new approach to increase the physiological knowledge on the complex processes involved in selective cofactor binding and might ultimately pave ways for novel drug development. Additionally, the control mechanisms beyond the ligand binding event, such as post-translational modifications, have come forward as important regulators of the NR – coactivator interaction for which molecular insights are urgently required. Targeting the Nuclear Receptor – Coactivator interaction Potent and selective peptide-based coactivator binding inhibitors (CBIs) have been developedin the past. The development of an on-bead peptide library screen for the identification of novel peptide inhibitor sequences for the AR - coactivator interaction newly described here allowed for the screening of peptides with non-proteinogenic amino acids in linear peptide sequences. Beads amendable to organic synthetic modifications and compatible to protein screening conditions were modified in a combinatorial fashion with a specific peptide library, leading to an One-Bead-One-Compound library. The library was incubated with Texas Red or Qdot labeled AR LBD and bright beads observed under the confocal fluorescence microscope were isolated to analyze their bound peptide sequence. The reliability of the assay as well as the exact binding affinity of the re-synthesized peptide hits for the AR LBD was investigated via fluorescence polarization studies (Figure 1). This On-Bead library screening method generated novel peptide sequences in a rapid manner, including non-natural amino acids that are able to inhibit the AR-cofactor interaction. Further, this methodology allows a screening of complete randomized libraries and other nuclear receptors such as the ER. This will hopefully result in a set of molecular tools that can be used in follow-up studies to help understanding critical issues such as selectivity and recognition motifs. A structural analysis of the AR-cofactor interaction shows that the coactivator peptide has to fold into a short helix to bind between a charge clamp at a fixed position on the AR surface. The length and stability of the helix thus appear to be crucial elements to achieve optimal binding. Control over helix stability and length is typically difficult for regular peptides and therefore stabilization of short peptides in a defined and stable fold is being pursued. Several different miniproteins featuring stabilized helices by means of two or three disulfide bridges were taken as scaffold for the in silico design of helical peptide binders for AR (Figure 2). Competitive fluorescence polarization binding studies for the AR LBD showed that many of the designed and synthesized miniproteins featured remarkable binding affinities around and below 1 µM. The length of the miniprotein helix was shown to be optimal when featuring around two turns. Since the influence of the point mutations on helix stability is significantly less prominent in miniproteins as in linear peptides, the peptides also allowed the evaluation of specific point mutations on binding affinity. Cysteine to methionine mutations of the miniproteins demonstrated the importance of the FXXLF motif on a performed stable helical segment in the miniprotein for high affinity binding. The introduction of an LXXLL motif into the helix enabled one miniprotein to bind the ER LBD. Initial crystallization studies of the miniproteins in complex with NR LBDs provide an entry to evaluate the molecular interactions of the miniproteins with the NR. Future attempts to generate miniproteins, stable under cellular conditions offer the possibility for applications in cell-based studies. As such, other miniprotein libraries can give even more detailed molecular insights into the molecular recognition of NRs by coactivators and provide the molecular requirements to generate new coactivator binding inhibitors for NRs. Chemical Biology Evaluations of Estrogen Receptor Post-translational Modifications NRs undergo a variety of post-translational modifications (PTM). The influence of these PTMs on ligand binding and on the formation of multiple protein complexes is, however, largely unknown. A synthetic entry into NR constructs featuring specifically introduced PTMs, would open up the possibility to study the crosstalk between NRs and their cofactors, as well as other phenomena, on the molecular level. A protein semi-synthesis method to generate correctly folded and active ERa and ERß LBDs was successfully established. This method allowed the generation of the ER LBD with a selectively and homogenously phosphorylated tyrosine, not accessible via the typical biochemical enzymatic approach. Using expressed protein ligation (EPL), the recombinant ER thioesters, lacking helix 12, were generated. Helix 12 of both isoforms was chemically synthesized bearing a phosphorylated tyrosine. The chosen strategy allows the use of a native cysteine in the ER LBD for successful ligation of the synthetic peptide to the protein thioester. Using this approach, both the phosphorylated ERa (pY537) and the phosphorylated ERß LBD (pY488) could be successfully synthesized.For the first time, the defined three- dimensional crystal structure of a phosphorylated ERß LBD in complex with cofactor peptide and the agonist estradiol (E2) could be determined with a maximal resolution of 1.5 Å. The observation that the phosphorylated Y488 sticks out of the global protein complex supports the assumption that the phosphorylated tyrosine can be target by the SH2 domain of cSrc via this phosphate. Further the access to crystal structures of post-translational modified NRs provides an entry to study the structural influence of PTMs on NR – protein interactions. The influence of tyrosine phosphorylation on cofactor binding was investigated using an on-chip- and FRET-based cofactor recruitment studies. Tyrosine phosphorylation is able to decrease the binding efficiency of distinct coactivator peptides to agonist-bound ERa and ERß. This observation strongly suggests that tyrosine phosphorylation of both ER LBDs represents an important control site that is involved in regulating cofactor binding under certain cellular conditions. However, in contrast to ERa, tyrosine phosphorylation of the ERß appears to lead to a higher flexible character of helix 12. This is accompanied by a significant increased affinity to coactivator peptides in the absence of ligand or in the presence of antagonist, as most prominent effect. Interestingly, this agonist-independent increase in cofactor binding was not observed for the ERa, reinforcing the idea that phosphorylation provides the two ER subtypes with distinct cofactor regulatory functions. Fragment-based Design and Structural Elucidation of Estrogen Receptor Agonists Despite the recent advances in directly inhibiting the NR – coactivator interaction with CBIs, the development of novel synthetic hormone analogs, binding the ligand binding pocket of the NR LBD, is of high importance, e.g. to generate sub-type selective ligands. Although many biological active compounds have been identified to bind the ER LBD, the search for compounds that selectively distinguish between the two ER isoforms ERa and ERß is still a challenging issue. Based on a simple two ring tetrahydroisoquinoline scaffold, a series of novel active scaffolds were identified. Using fragment based structure optimization it was demonstrate that certain functional groups incorporated during the optimization were important to enhance ER activity in both, biochemical and cellular transactivation assays up to an EC50 value of 0.59 µM and a 4-fold higher selectivity for ERß. Co-crystallization studies of some active compounds in complex with the ERß LBD revealed that the tetrahydroisoquinoline scaffold mimics the A and B ring of E2 and thus forms hydrogen bonds with Glu305/Arg346 (Figure 4). Further, electronegative groups like trifluorosulfonamide groups enhance the affinity to ER, most likely via electrostatic interactions. Further, the observed slight preference for ERß could be explained by electron-withdrawing groups, like fluoro groups, in close proximity to the Ile373 residue. It seems that in ERa these groups can be less effectively accommodated due to the presence of the more bulky Met421 at the same position. The identified new ER ligands together with the information gained from the x-ray studies present a suitable tool for further investigations of ER modulators. Especially with respect to the Ile373/Met421 interaction, these scaffolds could pave the way for the design of highly selective ERß binders.
Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Biomedical Engineering
Supervisors/Advisors
  • Brunsveld, Luc, Promotor
Award date11 May 2010
Place of PublicationEindhoven
Publisher
Print ISBNs978-90-386-2210-1
DOIs
Publication statusPublished - 2010

Fingerprint

Dive into the research topics of 'Modulating the nuclear receptor-cofactor interaction : characterization and inhibition'. Together they form a unique fingerprint.

Cite this