Skip to main content
NCBI home page
Royal Society of Chemistry logoLink to Royal Society of Chemistry
. 2019 Jul 22;10(10):1796–1802. doi: 10.1039/c9md00215d

Set-up and screening of a fragment library targeting the 14-3-3 protein interface†

Dario Valenti a,b,‡, João Filipe Neves c,‡, François-Xavier Cantrelle c, Stanimira Hristeva a, Domenico Lentini Santo d, Tomáš Obšil d,e, Xavier Hanoulle c, Laura M Levy a,‡, Dimitrios Tzalis a,✉, Isabelle Landrieu c,‡,✉, Christian Ottmann b,f,‡,✉
PMCID: PMC6839876  PMID: 31814953

graphic file with name c9md00215d-ga.jpgFragment-based driven discovery of 3 new low molecular weight starting points for the modulation of 14-3-3 protein–protein interactions.

Abstract

Protein–protein interactions (PPIs) are at the core of regulation mechanisms in biological systems and consequently became an attractive target for therapeutic intervention. PPIs involving the adapter protein 14-3-3 are representative examples given the broad range of partner proteins forming a complex with one of its seven human isoforms. Given the challenges represented by the nature of these interactions, fragment-based approaches offer a valid alternative for the development of PPI modulators. After having assembled a fragment set tailored on PPIs' modulation, we started a screening campaign on the sigma isoform of 14-3-3 adapter proteins. Through the use of both mono- and bi-dimensional nuclear magnetic resonance spectroscopy measurements, coupled with differential scanning fluorimetry, three fragment hits were identified. These molecules bind the protein at two different regions distant from the usual binding groove highlighting new possibilities for selective modulation of 14-3-3 complexes.

Introduction

Adapter proteins play a crucial role in the formation of complexes and their stabilization in regulatory pathways. Given their function – to mediate interactions between two proteins and therefore to regulate the activity or localization of their partners – this protein class represents an attractive point for therapeutic intervention. An example of adapter proteins is represented by the 14-3-3 protein family that – through its seven human isoforms1 – is involved in the regulation of numerous biological functions such as cellular signalling,2 protein translocation,3 enzymatic activity modulation,4 cell cycle regulation,5 structural maintenance,6 and many others. 14-3-3 proteins are dimers composed of two 30 kDa monomers constituted by 9 α helices, named αA–αI.1 Nowadays, PPIs attract significant interest in the pharmaceutical industry but are also challenging targets in drug discovery. PPIs involved in regulation are generally weak, ensuring a turn-over and mainly corresponding to superficial interactions between two – or more – proteins. Nonetheless, the multiplicity of hydrogen bonds,7 lipophilic interactions, π-stacking, π-cation stacking8 and ionic interactions9 observed in these interactions constitutes a complex and dynamic network that is hard to emulate. Their modulation has been studied intensively and several approaches have emerged to improve the discovery of small molecules able to offer therapeutic intervention.

In previous contributions,10–13 the rise of fragment-based approaches in this field was discussed as a good opportunity for developing successful modulators. Circa one third of the clinical candidates or approved drugs modulating PPIs have been indeed developed by application of fragment-based campaigns.10 The central idea of the fragment-based approach (FBA) is to develop a ligand for its own target, building it piece-by-piece from a low molecular weight (MW) molecule. Because of this principle, this strategy offers numerous advantages such as the possibility of constantly validating the molecules' evolution and of covering a very broad chemical space using one binding pharmacophore. This multidisciplinary strategy is generally constituted by a multistep workflow that allows scientists to go relatively fast from a fragment – an entity with a MW ≤ 300 Da – to a drug-like molecule ready to enter the clinical phase. The fragments' character – reported by Congreve et al. in 200314 – enhances the possibilities of finding key structural motifs able to address the target's binding site with higher ligand efficiency. However, as fragments do not contain many functional groups, the number of interactions with the target is also limited and therefore these compounds bind weakly (normally in the high μM–mM range). Among the broad range of biophysical techniques involved in the primary screening, nuclear magnetic resonance (NMR) is especially suitable to detect weak and very weak interactions (μM–mM range),15 which is a major reason for its employment in the FBA. NMR fragment screening methods can be divided into two classes: ligand-based methods and protein-based methods. While ligand-based methods allow the fast and sensitive screening of fragment mixtures with little material consumption,16 protein-based methods are more robust and allow, in cases where a protein assignment is available, the determination of the approximate binding site.17

In recent years, considerable effort has been put on the development of modulators of 14-3-3 PPIs, ranging from covalent18 and non-covalent fragments,19 peptidomimetic inhibitors,20,21 semi-synthetic derivatives of natural compounds22 and molecular tweezers.23,24 Considering that 14-3-3 proteins have hundreds of different protein-partners with a variety of binding interfaces, the discovery of small attaching points in their structures is important towards the selective modulation of 14-3-3 PPIs. The results presented here contribute to the enrichment of the portfolio of 14-3-3 binders with three novel fragments binding to two different areas of 14-3-3σ.

Experimental

Set-up of the fragment collection

With the aim to identify novel chemical entities as a starting point in the development of modulators (either disruptive or stabilizing) of 14-3-3σ interactions, we performed a fragment screening campaign. The first step on this pathway is the set-up of a pool of fragments. Currently, Taros proprietorial fragment collection has a subdivision dedicated to the PPIs' modulation with circa 1230 entries comprising both commercially available and novel structures. In this paragraph, we describe the guidelines used for building up this fragments' set. In usual practice, Congreve's “Rule of Three” (Ro3) would be applied. It represents one of the main guidelines for identifying molecules presenting the optimal physico-chemical properties required to be considered as a reliable member of the set. These attributes can be listed as follows: molecular weight ≤ 300 Da and clog P ≤ 3 as well as the number of rotatable bonds and hydrogen-bond acceptors (HBAs) and donors (HBDs), and polar surface area (PSA) ≤ 60 Å2. However, considering the superficial nature of PPIs as well as the dynamic behaviour of a pocket formed by two proteins, we decided to apply a higher tolerance regarding the previously mentioned selection parameters. Thus, we considered as optimal parameters molecular weight ≤ 330 Da, clog P ≤ 3.4, the number of rotatable bonds ≤4 as well as the hydrogen-bond acceptors (HBAs) and donors (HBDs) and finally a polar surface area ≤70 Å2. Nonetheless, in order to include also very attractive and chemically accessible structures that otherwise would not have passed the filtering phase, we decided to perform pre-filtering on the whole Taros internal compound collection (ca. 20 000 entries) fixing a threshold of 350 Da for the MW. This pre-filtering phase was followed by a visual inspection of the molecules (ca. 300) with a mass between 330 Da (our parameters for the final physico-chemical filtering) and 350 Da. As anticipated, the process started from the whole Taros compound collection and after two rounds of physico-chemical property filtering we reached a pool of ca. 4000 entries. Internal verification of eventual intellectual property issues for all novel structures within this set decreased the number of molecules to 3200, which were submitted to additional considerations. The aim of this second step was to remove the entries presenting undesired structural features and to create a diversified collection. Inspecting the collection from a medicinal chemistry point of view led us to exclude extremely reactive functions, such as alkylating or acylating features like aliphatic halides, acyl chlorides, epoxides, imines, oximes and acetals. We also excluded Michael acceptors and isocyanates due to their strong electrophilic character. On the other hand, polycyclic and heterocyclic compounds were always preferred when possible. The presence of sp3-enriched compounds was also encouraged according to the notion that molecules with a pronounced three-dimensional character appear to be more drug-like.25 Moreover, molecules having a 3D-character may cover a wider portion of extended pockets – such as the ones formed in protein complexes – and could also address the optimization toward different spatial coordinates.26 Another consideration applied in this phase was the synthetic accessibility of the molecules and the presence of exit points for further diversification. Moreover, in order to keep the reactivity of the molecules under control and at the same time to enhance the synthetic tractability of any eventual hit, we considered the isosteric members of the collection. The carboxylic acid feature was generally favoured in respect to the corresponding boronic acid or nitro group, but the substitution was not strictly enforced, depending on the in-house availability for a given core. For maintaining a high diversity between the cores composing the set, we picked – after visual inspection – one of the three possible positions for mono substituted aromatic rings keeping the other two as backup molecules. The physical availability of the selected entries was evaluated and compounds available in less than 30 mg were discarded. Quality control was applied using uHPLC-MS integrated systems and – when necessary – purification by preparative HPLC was carried out. Applying this workflow, we built up a collection of circa 900 compounds. This last set was then further enriched by the design and synthesis of ca. 300 additional novel cores based on both natural products and already known scaffolds in medicinal chemistry. Fig. 1 shows the workflow applied for the set-up of the Taros fragment library. Currently, the PPI-dedicated section of the Taros fragment collection counts circa 1230 entries and is constantly growing. For a statistical analysis of the physico-chemical properties, we refer to Fig. S1 and S2.† Finally, for the goals of this project, we selected circa 800 fragments that were first combined in cocktails containing five fragments each and then screened against 14-3-3σ applying a multistep approach.

Fig. 1. Workflow applied for setting up the Taros PPI fragment library.

Fig. 1

Open in a new tab

Protein production and purification

14-3-3σ (for differential scanning fluorimetry experiments) or 14-3-3σΔC17 (cleaved after T231 – used for NMR experiments to improve the quality of the NMR spectra) was expressed in E. coli BL21 (DE3) cells transformed with a pProExHtb vector carrying the cDNA to express an N-terminally His6-tagged human 14-3-3σ or 14-3-3σΔC17. An overnight 20 mL pre-culture in Luria–Bertani (LB) medium was used to inoculate 1 L of culture medium. For the natural abundance proteins, cells were inoculated in 1 L of M9 minimal medium supplemented with 4 g L–1 glucose, 1 g L–1 ammonium chloride, 5% (v/v) LB medium and 100 μg mL–1 ampicillin. For 15N2H labelling, the cells were inoculated in 1 L of deuterated M9 minimal medium supplemented with 2 g L–1 12C62H7 glucose, 1 g L–1 15N ammonium chloride, 0.4 g L–1 Isogro 15N12C2H powder-growth medium (Sigma Aldrich) and 100 μg mL–1 ampicillin. Cells were grown at 37 °C to an OD600 of 0.9 and induced with 0.5 mM IPTG. Incubation was continued for 12 h at 25 °C for the natural abundance proteins and for 15 h at 28 °C for the 15N2H labelled protein. Cells were harvested by centrifugation and lysed with a homogenizer and the proteins were purified using a Ni-NTA column (GE Healthcare). The N-terminal His6-tag was then cleaved by the TEV protease and the proteins were further dialyzed overnight at 4 °C against the NMR buffer (100 mM sodium phosphate, pH 6.8, 50 mM NaCl), concentrated, aliquoted, flash frozen and stored at –80 °C. Typical yields were in the range of 40 to 80 mg of protein per liter of culture. A detailed protocol can be found in the study of Neves et al.27

Differential scanning fluorimetry

Differential scanning fluorimetry (DSF) thermal melting points were recorded using a LightCycler® 480 (Roche, Switzerland). Measurements were performed in 96-well plates with samples containing 7 μM 14-3-3σ in the presence or absence of each cocktail (2500 μM per fragment). The buffer contained 100 mM Hepes, pH 7.5, 150 mM NaCl and 4% (v/v) DMSO. The fluorescent probe SYPRO Orange (Sigma Aldrich) was added at a dilution of 1 : 600. The excitation filter for SYPRO Orange was set to 465 nm and the emission filter was set to 580 nm. The temperature was increased from 20 °C to 95 °C at a rate of 0.6 °C per minute. Melting temperature values were determined using the LightCycler® 480 software version 1.5.1.62 by plotting the first derivative of the melting curves (Fig. S3†). All the measurements were performed in triplicate and the deviations of melting temperatures (ΔTm) presented for each cocktail were calculated relatively to three control samples (7 μM 14-3-3σ in the absence of the compound) measured in the same 96-well plate.

WaterLOGSY experiments

WaterLOGSY spectra were acquired in 5 mm tubes (sample volume 530 μL) using a 600 MHz Bruker Avance III HD spectrometer equipped either with a CPQCI cryogenic probe or with a TXI non-cryogenic probe. The spectra were recorded with 32 768 complex data points and with a mixing time of 1.7 s. The number of scans per increment was 384 when the spectrometer was equipped with a cryogenic probe (acquisition time of 35 minutes per experiment) and 1280 when the spectrometer was equipped with a non-cryogenic probe (acquisition time of 118 minutes). The spectra were acquired at 16 °C, in a buffer containing 100 mM sodium phosphate, 50 mM NaCl, pH 6.8 and 10% (v/v) D2O. The final concentration of DMSO-d6 was 2% and was kept constant for all the experiments. The spectra were obtained with samples containing 25 μM 14-3-3σΔC17 in the presence or in the absence of each cocktail (500 μM per fragment) or each fragment (at 500 μM). Additional control experiments were performed in the presence of the cocktail/fragment and in the absence of protein. Spectra were collected and processed with Topspin 3.5 (Bruker Biospin, Karlsruhe, Germany).

2D nuclear magnetic resonance experiments

1H–15N TROSY-HSQC spectra were acquired in 3 mm tubes (sample volume 200 μL) using a 900 MHz Bruker Avance-NEO spectrometer, equipped with a cryoprobe. The spectra were recorded with 3072 complex data points in the direct dimension and 120 complex data points in the indirect dimension, with 128 scans per increment (acquisition time of 4 hours per experiment), at 32 °C, in a buffer containing 100 mM sodium phosphate, 50 mM NaCl, pH 6.8, 1 mM DTT, EDTA-free protease inhibitor cocktail (Roche, Basel, Switzerland) and 10% (v/v) D2O. The final concentration of DMSO-d6 was 2% and was kept constant for all the experiments. The spectra were obtained with samples containing 75 μM 15N2H labelled 14-3-3σΔC17 in the presence or in the absence of each cocktail (2000 μM per fragment). Backbone assignments of 15N2H labelled 14-3-3σΔC17 were previously reported.27 The reference for the 1H chemical shift was relative to trimethylsilyl propionate (TMSP) or 4,4-dimethyl-4-silapentane-1-sulfonic acid (DSS). 15N chemical shifts were referenced indirectly. Spectra were collected and processed with Topspin 4.0 (Bruker Biospin, Karlsruhe, Germany) and analyzed with Sparky 3.12 (T. D. Goddard and D. G. Kneller, SPARKY 3, University of California, San Francisco). CSPs in the form of chemical shift value modifications (in ppm) on the 1H–15N TROSY-HSQC were calculated using the following equation:Inline graphic

Results and discussion

Fragment screening strategy

A library of 785 fragments grouped in 157 cocktails of 5 compounds each was screened for binding to 14-3-3σ with the objective of finding low MW starting points for the development of PPI modulators. Initially, 1H–15N TROSY-HSQC spectra were directly recorded for 55 out of the 157 cocktails. For the screening of the rest of the library, DSF and WaterLOGSY were introduced in order to allow a primary screening of the cocktails before confirmation by 1H–15N TROSY-HSQC – which is costly and time consuming but offers the additional advantage of identifying the binding pocket. DSF screening was used for 156 out of the 157 cocktails and WaterLOGSY was applied for the 102 cocktails that were not directly screened by 1H–15N TROSY-HSQC. The cocktails which showed binding by either WaterLOGSY or DSF were further screened by 1H–15N TROSY-HSQC (secondary screening). The fragments contained in the cocktails showing binding by 1H–15N TROSY-HSQC were further individually screened initially by WaterLOGSY and finally by 1H–15N TROSY-HSQC for the determination of the binding site. A diagram illustrating this screening strategy is presented in Fig. 2.

Fig. 2. Workflow of the fragment screening campaign. DSF, WaterLOGSY and 1H–15N TROSY-HSQC were employed for primary screening. Secondary screening by 1H–15N TROSY-HSQC was performed for the cocktails that showed binding by either DSF or WaterLOGSY. WaterLOGSY and 1H–15N TROSY-HSQC were performed for singleton validation.

Fig. 2

Open in a new tab

Primary screening by WaterLOGSY

WaterLOGSY is a ligand-based NMR screening method that allows the fast screening of fragment libraries based on the nuclear Overhauser effect (NOE).28 WaterLOGSY requires a lower amount of protein in comparison with protein-based NMR methods and does not require isotopic labelling of the target. Moreover, this method allows a straightforward hit identification since the NMR signals of binders and non-binders have a different phase in the WaterLOGSY spectrum (Fig. S4†). WaterLOGSY was performed for the 102 cocktails that were not directly screened by 1H–15N TROSY-HSQC. Binding was detected for 43 out of the 102 cocktails, which were selected for secondary screening by 1H–15N TROSY-HSQC.

Primary screening by differential scanning fluorimetry

Differential scanning fluorimetry (DSF) is a fast method to screen a fragment library with little material consumption. The principle relies on monitoring the thermal melting curve of the protein upon gradual heating of the protein solution and on the determination of the melting temperature (Tm). A protein-ligand interaction is therefore detected when the binder induces a shift in the melting temperature of the protein.29 Unlike most proteins, the thermal denaturation profile of 14-3-3σ shows two sigmoidal curves, suggesting that there are two unfolding transitions (Fig. S3†). Both melting temperatures were determined for all the samples and the binding effect was considered significant upon the induction of a thermal shift higher than 1 °C in the sum of both melting temperatures (ΔTm1 + ΔTm2 > 1 °C). After the primary screening, 22 out of the screened 156 cocktails showed a significant thermal shift effect (Table S1†). These cocktails were therefore selected for secondary screening. 15 cocktails out of these 22 also showed binding by WaterLOGSY.

Primary/secondary screening by 1H–15N TROSY-HSQC

1H–15N HSQC is the most popular method for protein-based NMR screening. In spite of being more expensive and time-consuming when compared to ligand-based NMR methods, 1H–15N HSQC is a more robust method and has the important advantage of identifying the binding site of the ligands.301H–15N TROSY-HSQC was used for the screening of 55 out of 157 cocktails and as a secondary screening technique for the 50 cocktails that showed binding by either WaterLOGSY (43), DSF (22) or both (15). From these, 6 cocktails showed a significant effect on the 1H–15N TROSY-HSQC spectrum of 14-3-3σ.

All 6 cocktails were found to be positive by WaterLOGSY and 4 were positive by DSF as well.

False positives in the DSF screening could be due to aggregates of compounds since a considerable part of the hits showed solubility problems at the tested concentration.

Compared to 1H–15N TROSY-HSQC, WaterLOGSY also provided a very high hit rate. Probably some of the hits detected were just binding transiently to the protein and were too weak to be detected by 1H–15N TROSY-HSQC. Although negative control experiments were performed for all the WaterLOGSY tested cocktails, aggregation cannot be excluded as a source of false positives in this method. It is also possible that some of the WaterLOGSY hits were not confirmed due to the fact that around 15% of the amide resonances of 14-3-3σΔC17 are undetected in the 1H–15N TROSY-HSQC.

Confirmation of hits from the cocktails and binding site identification

After the identification and orthogonal validation of 6 hit cocktails, their components were tested as singletons. Thirty singletons were further screened individually by WaterLOGSY. Six out of these 30 singletons (1 per each cocktail) showed a positive effect and were submitted to the final confirmation by 1H–15N TROSY-HSQC. Two out of these 6 singletons failed to cause a significant effect on the 2D spectrum of 14-3-3σΔC17. One singleton caused chemical shift perturbations but also induced protein precipitation and was, for this reason, excluded from the screening. Three out of the 6 singletons were successfully confirmed as binders by 1H–15N TROSY-HSQC because each produced the same effect by itself on the spectrum as its corresponding cocktail of five. These 3 fragments were therefore validated by both a ligand-based method (Fig. S5–S7†) and a protein-based method (Fig. S8–S10†). Fragments 1 and 2 were seen to bind at the top of 14-3-3 helices αH and αI (Fig. S8, S9,† and 3) and fragment 3 was seen to bind at the dimer interface (Fig. S10† and 3). Interestingly, the cocktails containing the fragments that were seen to bind at the top of helices αH and αI did not produce a significant effect by DSF while the cocktail containing fragment 3 induced a significant thermal shift. This fact suggests that binding at the upper region of the protein does not have an influence on its denaturation. NMR was used for the screening because of its sensitivity to even very weak binders, as expected for fragments. The discrete number of resonances in the 1H–15N TROSY-HSQC spectrum affected by the binding, together with the observed fast exchange regime regarding the NMR time scale, gives an estimate of the Kd in the high μM–mM range.31 Later stage optimization of the selected fragments should focus on improving the binding affinity.

Fig. 3. Identification of the binding sites of the hits by 1H–15N TROSY-HSQC. The structures of the hits are represented and their binding sites are shown in the crystal structure of 14-3-3σ (gray surface - PDB ID: ; 1YZ5). Fragments 1 and 2 induce chemical shift perturbations in resonances corresponding to amino acid residues located on the top of helices αH and αI. Fragment 3 induces chemical shift perturbations in resonances corresponding to amino acid residues located at the dimer interface of 14-3-3σ.

Fig. 3

Open in a new tab

Conclusions

Here, we discussed a strategy applied in the set-up of a fragment collection dedicated to PPI modulation. The specificity of this type of contacts, involving superficial interactions (no deep pocket) and flexibility at the interface, was taken into account to build the library by loosening the physico-chemical restrictions imposed by the general Rule of Three guidelines. This approach was also derived from the considerations made by Chen and Hubbard32 about the slight increment of both molecular weight and hydrophobicity in fragment hits modulating protein–protein complexes. The resulting distribution of fragment physico-chemical properties (Fig. S1 and S2†) demonstrates that the described fragment collection indeed meets these criteria and will be valuable for further applications targeting PPI modulation.

In this work, we screened the fragment library on the adapter protein 14-3-3σ using a multi-technique approach. Application of mono- and bi-dimensional NMR coupled with DSF provided us a robust and orthogonally validated dataset of molecules, rapidly excluding false positives and reintegrating false negative hits.

Eventually, we identified three fragment hits binding two different sites from the classic amphipathic groove of 14-3-3σ, where most of the protein partners bind. Nevertheless, some examples in which the protein partners establish contacts with different regions of 14-3-3 are reported.33–38 In these cases, fragments that bind remotely from the classic amphipathic binding groove represent a unique opportunity for selective modulation of the PPIs.

The hits disclosed by this integrated screening campaign represent a new starting point for modulating the 14-3-3 PPIs and highlight an important opportunity for selectivity improvement towards a specific protein complex.

Conflicts of interest

There are no conflicts to declare.

Supplementary Material

Click here for additional data file. (2.3MB, pdf)

Acknowledgments

This research is supported by funding from the European Union through the TASPPI project (H2020-MSCA-ITN-2015, grant number 675179). D. V. acknowledges Dr. Martin Corpet for the fruitful scientific discussion. J. F. N., F.-X. C., X. H. and I. L. acknowledge LabEx (Laboratory of Excellence) for financial support on the scope of the DISTALZ consortium (ANR, ANR-11-LABX-009). The NMR facilities were funded by the Nord Region Council, CNRS, Institut Pasteur de Lille, the European Community (ERDF), the French Ministry of Research and the University of Lille and by the CTRL CPER cofunded by the European Union with the European Regional Development Fund (ERDF), by the Hauts de France Regional Council (contract no. 17003781), Métropole Européenne de Lille (contract no. 2016_ESR_05), and French State (contract no. 2017-R3-CTRL-Phase 1). We acknowledge support for the NMR facilities from TGE RMN THC (CNRS, FR-3050) and FRABio (Univ. Lille, CNRS, FR-3688).

Footnotes

†Electronic supplementary information (ESI) available: Statistical analysis of the fragment collection, DSF melting curves, DSF compound-induced thermal shifts, WaterLOGSY NMR spectra, 1H–15N TROSY-HSQC NMR spectra and chemical shift perturbation plots. See DOI: 10.1039/c9md00215d

References

  1. Obsil T., Obsilova V. Semin. Cell Dev. Biol. 2011;22:663–672. doi: 10.1016/j.semcdb.2011.09.001. [DOI] [PubMed] [Google Scholar]
  2. Daly R. J., Gu H., Parmar J., Malaney S., Lyons R. J., Kairouz R., Head D. R., Henshall S. M., Neel B. G., Sutherland R. L. Oncogene. 2002;21:5175–5181. doi: 10.1038/sj.onc.1205522. [DOI] [PubMed] [Google Scholar]
  3. Aguilera C., Fernández-Majada V., Inglés-Esteve J., Rodilla V., Bigas A., Espinosa L. J. Cell Sci. 2006;119:3695–3704. doi: 10.1242/jcs.03086. [DOI] [PubMed] [Google Scholar]
  4. Masters S. C., Yang H., Datta S. R., Greenberg M. E., Fu H. Mol. Pharmacol. 2001;60:1325–1331. doi: 10.1124/mol.60.6.1325. [DOI] [PubMed] [Google Scholar]
  5. Vousden K. H., Prives C. Cell. 2009;137:413–431. doi: 10.1016/j.cell.2009.04.037. [DOI] [PubMed] [Google Scholar]
  6. Sluchanko N. N., Gusev N. B. J. Alzheimer's Dis. 2011;27:467–476. doi: 10.3233/JAD-2011-110692. [DOI] [PubMed] [Google Scholar]
  7. Milroy L.-G., Grossmann T. N., Hennig S., Brunsveld L., Ottmann C. Chem. Rev. 2014;114:4695–4748. doi: 10.1021/cr400698c. [DOI] [PubMed] [Google Scholar]
  8. Wuo M. G., Mahon A. B., Arora P. S. J. Am. Chem. Soc. 2015;137:11618–11621. doi: 10.1021/jacs.5b05525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Schumacher B., Mondry J., Thiel P., Weyand M., Ottmann C. FEBS Lett. 2010;584:1443–1448. doi: 10.1016/j.febslet.2010.02.065. [DOI] [PubMed] [Google Scholar]
  10. Valenti D., Hristeva S., Tzalis D., Ottmann C. Eur. J. Med. Chem. 2019;167:76–95. doi: 10.1016/j.ejmech.2019.01.084. [DOI] [PubMed] [Google Scholar]
  11. Turnbull A. P., Boyd S. M., Walse B. Res. Rep. Biochem. 2014;4:13–26. [Google Scholar]
  12. Valkov E., Sharpe T., Marsh M., Greive S., Hyvönen M. Top. Curr. Chem. 2012;317:145–179. doi: 10.1007/128_2011_265. [DOI] [PubMed] [Google Scholar]
  13. Magee T. V. Bioorg. Med. Chem. Lett. 2015;25:2461–2468. doi: 10.1016/j.bmcl.2015.04.089. [DOI] [PubMed] [Google Scholar]
  14. Congreve M., Carr R., Murray C., Jhoti H. Drug Discovery Today. 2003;8:876–877. doi: 10.1016/s1359-6446(03)02831-9. [DOI] [PubMed] [Google Scholar]
  15. Williamson M. P. Prog. Nucl. Magn. Reson. Spectrosc. 2013;73:1–16. doi: 10.1016/j.pnmrs.2013.02.001. [DOI] [PubMed] [Google Scholar]
  16. Unione L., Galante S., Díaz D., Cañada F. J., Jiménez-Barbero J. MedChemComm. 2014;5:1280–1289. [Google Scholar]
  17. Gossert A. D., Jahnke W. Prog. Nucl. Magn. Reson. Spectrosc. 2016;97:82–125. doi: 10.1016/j.pnmrs.2016.09.001. [DOI] [PubMed] [Google Scholar]
  18. Sijbesma E., Hallenbeck K. K., Leysen S., de Vink P. J., Skóra L., Jahnke W., Brunsveld L., Arkin M. R., Ottmann C. J. Am. Chem. Soc. 2019;141:3524–3531. doi: 10.1021/jacs.8b11658. [DOI] [PubMed] [Google Scholar]
  19. Sijbesma E., Skora L., Leysen S., Brunsveld L., Koch U., Nussbaumer P., Jahnke W., Ottmann C. Biochemistry. 2017;56:3972–3982. doi: 10.1021/acs.biochem.7b00153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Andrei S. A., Meijer F. A., Neves J. F., Brunsveld L., Landrieu I., Ottmann C., Milroy L.-G. ACS Chem. Neurosci. 2018;9:2639–2654. doi: 10.1021/acschemneuro.8b00118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Milroy L.-G., Bartel M., Henen M. A., Leysen S., Adriaans J. M. C., Brunsveld L., Landrieu I., Ottmann C. Angew. Chem., Int. Ed. 2015;54:15720–15724. doi: 10.1002/anie.201507976. [DOI] [PubMed] [Google Scholar]
  22. Andrei S. A., de Vink P., Sijbesma E., Han L., Brunsveld L., Kato N., Ottmann C., Higuchi Y. Angew. Chem., Int. Ed. 2018;57:13470–13474. doi: 10.1002/anie.201806584. [DOI] [PubMed] [Google Scholar]
  23. Bier D., Rose R., Bravo-Rodriguez K., Bartel M., Ramirez-Anguita J. M., Dutt S., Wilch C., Klärner F.-G., Sanchez-Garcia E., Schrader T., Ottmann C. Nat. Chem. 2013;5:234–239. doi: 10.1038/nchem.1570. [DOI] [PubMed] [Google Scholar]
  24. Bier D., Mittal S., Bravo-Rodriguez K., Sowislok A., Guillory X., Briels J., Heid C., Bartel M., Wettig B., Brunsveld L., Sanchez-Garcia E., Schrader T., Ottmann C. J. Am. Chem. Soc. 2017;139:16256–16263. doi: 10.1021/jacs.7b07939. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lovering F., Bikker J., Humblet C. J. Med. Chem. 2009;52:6752–6756. doi: 10.1021/jm901241e. [DOI] [PubMed] [Google Scholar]
  26. Morley A. D., Pugliese A., Birchall K., Bower J., Brennan P., Brown N., Chapman T., Drysdale M., Gilbert I. H., Hoelder S., Jordan A., Ley S. V., Merritt A., Miller D., Swarbrick M. E., Wyatt P. G. Drug Discovery Today. 2013;18:1221–1227. doi: 10.1016/j.drudis.2013.07.011. [DOI] [PubMed] [Google Scholar]
  27. Neves J. F., Landrieu I., Merzougui H., Boll E., Hanoulle X., Cantrelle F.-X. Biomol. NMR Assignments. 2019;13:103–107. doi: 10.1007/s12104-018-9860-1. [DOI] [PubMed] [Google Scholar]
  28. Dalvit C., Fogliatto G., Stewart A., Veronesi M., Stockman B. J. Biomol. NMR. 2001;21:349–359. doi: 10.1023/a:1013302231549. [DOI] [PubMed] [Google Scholar]
  29. Ciulli A., in Protein-Ligand Interactions: Methods and Applications, ed. M. A. Williams and T. Daviter, Humana Press, Totowa, NJ, 2013, pp. 357–388. [Google Scholar]
  30. Ma R., Wang P., Wu J., Ruan K. Molecules. 2016;21:854. doi: 10.3390/molecules21070854. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Latham M. P., Zimmermann G. R., Pardi A. J. Am. Chem. Soc. 2009;131:5052–5053. doi: 10.1021/ja900695m. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Chen J., Hubbard R. E. J. Comput.-Aided Mol. Des. 2009;23:603–620. doi: 10.1007/s10822-009-9280-5. [DOI] [PubMed] [Google Scholar]
  33. Alblova M., Smidova A., Docekal V., Vesely J., Herman P., Obsilova V., Obsil T. Proc. Natl. Acad. Sci. U. S. A. 2017;114:E9811–E9820. doi: 10.1073/pnas.1714491114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Obsil T., Ghirlando R., Klein D. C., Ganguly S., Dyda F. Cell. 2001;105:257–267. doi: 10.1016/s0092-8674(01)00316-6. [DOI] [PubMed] [Google Scholar]
  35. Ottmann C., Marco S., Jaspert N., Marcon C., Schauer N., Weyand M., Vandermeeren C., Duby G., Boutry M., Wittinghofer A., Rigaud J.-L., Oecking C. Mol. Cell. 2007;25:427–440. doi: 10.1016/j.molcel.2006.12.017. [DOI] [PubMed] [Google Scholar]
  36. Psenakova K., Petrvalska O., Kylarova S., Lentini Santo D., Kalabova D., Herman P., Obsilova V., Obsil T. Biochim. Biophys. Acta, Gen. Subj. 2018;1862:1612–1625. doi: 10.1016/j.bbagen.2018.04.006. [DOI] [PubMed] [Google Scholar]
  37. Sluchanko N. N., Beelen S., Kulikova A. A., Weeks S. D., Antson A. A., Gusev N. B., Strelkov S. V. Structure. 2017;25:305–316. doi: 10.1016/j.str.2016.12.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Taoka K., Ohki I., Tsuji H., Furuita K., Hayashi K., Yanase T., Yamaguchi M., Nakashima C., Purwestri Y. A., Tamaki S., Ogaki Y., Shimada C., Nakagawa A., Kojima C., Shimamoto K. Nature. 2011;476:332–335. doi: 10.1038/nature10272. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Click here for additional data file. (2.3MB, pdf)

Articles from MedChemComm are provided here courtesy of Royal Society of Chemistry

RESOURCES