Design Of Drug-Like Protein-Protein Interaction Stabilizers Guided By Chelation-Controlled Bioactive Conformation Stabilization
Francesco Bosica, Sebastian Andrei, João Filipe Neves, Peter Brandt, Anders Gunnarsson, Isabelle Landrieu, Christian Ottmann, Gavin O'Mahony
1 Design Of Drug-Like Protein-Protein Interaction Stabilizers Guided By Chelation-Controlled Bioactive Conformation Stabilization
Francesco Bosica [a,b] , Sebastian Andrei [b] , João Filipe Neves [c,d] , Peter Brandt [a] , Anders Gunnarsson [e] , Isabelle Landrieu [c,d] , Christian Ottmann [b,f] and Gavin O’Mahony* ,[a] [a] F. Bosica, Dr. P. Brandt, Dr. G. O’Mahony Research and Early Development, Cardiovascular, Renal and Metabolism, BioPharmaceuticals R&D AstraZeneca Gothenburg, Sweden E-mail: [email protected] [b] F. Bosica, Dr. S. Andrei, Prof. C. Ottmann Laboratory of Chemical Biology, Department of Biomedical Engineering and Institute for Complex Molecular Systems (ICMS) Eindhoven University of Technology Den Dolech 2, 5612 AZ Eindhoven, The Netherlands [c] Dr. J. F. Neves, Dr. I. Landrieu ERL9002 Integrative Structural Biology CNRS 50 Avenue de Halley, 59658 Villeneuve d'Ascq, Lille, France [d] Dr. J. F. Neves, Dr. I. Landrieu Department U1167 RID-AGE Risk Factors and Molecular Determinants of Aging-Related Diseases Univ. Lille, Inserm, CHU Lille, Institut Pasteur de Lille 1 Rue du Professeur Calmette 59800 Lille, France [e] Dr. A. Gunnarsson Discovery Sciences, BioPharmaceuticals R&D AstraZeneca Gothenburg, Sweden [f] Prof. C. Ottmann Department of Chemistry University of Duisburg-Essen Universitätsstrasse 7, 45117 Essen, Germany Supporting information for this article is given via a link at the end of the document.
Abstract:
The protein - protein interactions (PPIs) of 14 - - - - - - - ray crystal structure of a PPI in complex with an extremely low potency stabilizer uncovered an unexpected non-protein interacting, ligand-chelated Mg leading to the discovery of metal ion - dependent 14 - -
3 PPI stabilization potency. This originates from a novel chelation - controlled bioactive conformation stabilization effect. Metal chelation has been associated with pan-assay interference compounds (PAINS) and frequent hitter behavior, but chelation can evidently also lead to true potency gains and find use as a medicinal chemistry strategy to guide compound optimization. To demonstrate this, we exploited the effect to design the first potent, selective and drug-like 14 - - Introduction
Protein-protein interactions (PPIs) are a major class of macromolecular interactions which control the function of many proteins. [1]
The protein interactome is crucial to many biological processes, is implicated in many diseases [2-4] and contains many promising intervention points for development of therapeutics. [5-6]
Inhibition is the most advanced mode of action for PPI modulation, with several rationally-designed PPI inhibitors in clinical testing or approved as drugs. [7]
In contrast, PPI stabilization is comparatively unexplored. [8]
Several natural products have been shown post hoc to operate by stabilization of PPIs, [9] but there are few examples of rationally designed small-molecule, drug-like PPI stabilizers. This is partly due to incomplete understanding of the structural and kinetic principles driving stabilization, making robust screen design difficult and resulting in a paucity of tractable starting points for PPI stabilizer development. [10-12] [13-14]
They are involved in many cellular processes such as cell cycle regulation and apoptosis, [15] subcellular localization [16] and enzymatic activity regulation. [17]
There are 7 human 14-3-3 isoforms, structurally composed of 9 helices forming an amphipathic binding groove which binds 14-3-3 consensus motifs on phosphorylated client proteins. [13]
The natural product Fusicoccin A ( , Figure 1a) and related fusicoccanes [18] stabilize 14-3-3 PPIs with many phosphorylated partners. [19-20] Ternary complex crystal structures show that they occupy a well-defined binding pocket (“FC pocket”) at the interface between 14-3-3 and the binding partner (Figure 1b). The fusicoccanes’ synthetic complexity makes them unsuitable for a molecular matched pair approach [21] to generate structure-activity relationship (SAR) principles or to study PPI stabilization selectivity. We therefore aimed to identify synthetically tractable, drug-like small-molecule starting points for systematic Figure 1.
Identification of 2 as a low potency 14-3-3/ER PPI stabilizer. (a) Chemical structures of Fusicoccin A and Pyrrolidone1 . (b) Crystal structure of the ternary complex between 14-3-3 (white surface), ER pT phosphopeptide (green sticks) and (blue sticks) rendered from PDB 4JDD and highlighting the “FC pocket”. (c) Comparison of 14-3-3 /ER (pT ) PPI stabilization activity of (purple inverse triangles) and rac- (blue circles) in an FP assay, using 50 nM 14-3-3 and 10 nM ER (pT ) phosphopeptide FITC- . “Relative stabilization” (y-axis) is the mean fold-increase of FP signal over baseline (i.e. interaction between 14-3-3 and FITC- alone). The error bars in all plots indicate +/- SD (n =3). (d) Comparison of PPI stabilization activity of rac - , (-)- (green squares) and (+)- (red triangles) in 14-3-3/ER (pT ) FP assay. (e) The H- N TROSY-HSQC peak intensities most affected by binding of (-)- correspond to 14-3-3 residues lining the FC pocket (mapped on the crystal structure of 14-3-3 rendered from PDB 1YZ5). The 5 most affected residues are colored red, the next 5 orange and the next 5 yellow. I/I o values and colour coding is provided in Fig S6. investigation of 14-3-3 PPI stabilization. In the course of this, we also discovered a novel chelation-controlled ligand conformational stabilization effect which had profound effects on compound potency, resulting in metal ion-assisted small molecule PPI stabilization. Chelation of metals by ligands of interest has been associated with assay interference and frequent hitter behavior, especially in PPI inhibition assays based on AlphaScreen technology. [22] Based on this observation, compounds containing potential chelating moieties tend to be filtered out from screening collections either prior to screening or in order to triage large screening data sets prior to data analysis, as an extension to the originally-reported PAINS filters. [23]
In this report, we show that metal ion chelation may in some cases lead to true potency gains and allow identification of hits that otherwise would be discarded as of insufficient potency. In our case, this manifested itself as a metal-ion assisted PPI stabilization, where the addition of bivalent metal ions led to up to two orders of magnitude increase in compound potency. We furthermore exploited this effect to design metal-independent PPI stabilizers by mimicking the chelation with intramolecular hydrogen bonds, leading to compounds with high PPI stabilization potency and which are insensitive to metal ion concentration. The resulting compounds enable the systematic investigation of 14-3-3 PPI stabilization SAR and selectivity and further development of small-molecule 14-3-3 PPI stabilizers with potency rivaling that of the natural product . We also propose that this type of ligand-specific conformational effect is a potential source of false negatives and should be considered when analyzing screening data and interpreting SAR for chelation-competent ligands as well as during the analysis and triage of high-throughput screen output. Results
Racemic Pyrrolidone1 ( rac -2) is a weak stabilizer of the canonical 14-3-3/ER (pT ) PPI Binding of 14-3-3 to the ER pT phosphosite has been shown to prevent estradiol-induced activation of ER by preventing ER dimerisation [24] By stabilizing the 14-3-3/ER interaction, was shown to enhance the 14-3-3-mediated inhibition of ER activity and therefore the 14-3-3/ER interaction is of interest as a potential target for development of ER dependent breast cancer therapies. To test compounds for 14-3-3 PPI stabilization, we selected the ER pT phosphosite (ER (pT )) as the phosphorylated 14-3-3 binding partner, due to the availability of a ternary X-ray crystal structure with . Pyrrolidone1 ( , Figure 1a) has previously been reported as a low potency stabilizer of the PPI between 14-3-3 and the plant plasma membrane Figure 2. X-ray crystal structure of 14-3-3/ER (pT )/( R )-2 ternary complex and identification of ( R )-2-chelated Mg ion. (a) Comparison of binding mode of from 3.25 Å 14-3-3/PMA2 complex structure (magenta sticks, assigned as ( S ), rendered from PDB 3M51) with revised binding mode derived from the 1.85 Å 14-3-3/ER (pT ) complex structure (yellow sticks, assigned ( R ) absolute configuration, PDB 6TJM). (b) Binding mode of ( R )- in the FC pocket, showing surface contributions from 14-3-3 (grey surface) and ER (pT ) phosphopeptide (green surface). (c) Details of interactions of salicylate moiety of ( R )- . (d) Details of interactions of nitrophenyl moiety of ( R )- , showing the nitro group acting as H-bond acceptor from 14-3-3 Lys122 and water-mediated interaction with the C-terminus of ER (pT ) phosphopeptide (green). (e) Additional ligand-associated electron density (2F obs -F cal map contoured at 1 ) corresponding to a fully-hydrated Mg ion (shown in green) chelated by the vinylogous carboxylate moiety of ( R )- H + -ATPase 2 (PMA2). [25] Since stabilizes both the 14-3-3/PMA2 and 14-3-3/ER (pT ) PPIs [24, 26] and binds to the FC pocket of the non-canonical 14-3-3/PMA2 complex (the PMA2 peptide used was non-phosphorylated), we decided to investigate as a potential stabilizer of the 14-3-3/ER (pT ) complex. We developed fluorescence polarization (FP) and surface plasmon resonance (SPR) assays based on 14-3-3 and ER pT phosphopeptides ( or FITC- , sequence AEGFPApT V-COOH). The K d of the 14-3-3/ interaction was 227 ± 14 nM (Figure S1) as measured by SPR. In a 14-3-3/ER (pT ) FP assay (based on a labelled derivative of , FITC- ), showed very low activity (1.4-fold stabilization at 200 µM, Fig. 1c) compared to (EC M, 4-fold stabilization at 200 µM). In the 14-3-3/ER (pT ) SPR assay, a comparable EC was determined for (EC M), while the PPI stabilization effect of was undetectable, with only binding of to 14-3-3 protein being observed (Figure S2, S3). The enantiomers of were then separated by chiral HPLC. In the 14-3-3/ER (pT ) FP and SPR assays, (-)- was found to be more active than rac - and (+)- was shown to be inactive (Figure 1d and S9). (-)- was shown to be stable with respect to epimerization under basic conditions (see Supporting Information). NMR experiments were conducted using 14-3-3 C [27] and to determine if (-)- binds to the FC pocket of the 14-3-3/ER pT ) complex. WaterLOGSY experiments [28] were performed with (-)- in the presence and absence of either 14-3-3 C or the 14-3-3 C/ complex. Positively phased H signals for (-)- confirmed binding to 14-3-3 C alone (Figure S4, red spectrum) and the 14-3-3 C/ complex (Figure S4, green spectrum). Protein-based NMR experiments were then carried out to determine the binding site of (-)- on 14-3-3 C. By observing changes in the intensity ratio (I/I ) of resonances in the fully-assigned H- N TROSY-HSQC spectrum of 14-3-3 C [29] in the presence (I) or absence (I ) of (-)- (Figure S5, S6), a mapping of the residues most affected by (-)- binding (Figure 1e) was generated, indicating that the binding site of (-)- corresponds to the FC pocket. [25] Phosphopeptide induced significant chemical shift perturbations in the 14-3-3 C H- N TROSY-HSQC spectrum, with some resonances showing characteristics of a slow exchange regime on the NMR time scale (Figure S7), which is Figure 3. Metal ion-dependent stabilization of 14-3-3/ERa(pT ) and 14-3-3/CaMKK2(pS ) PPIs by ( R )-2. (a) Stabilization of the 14-3-3/CaMKK2(pS ) PPI measured in FP assay. Rac - (blue circles), ( R )- (green squares) are stabilizers while ( S )- (red triangles) and (purple inverse triangles) are not. (b) The addition of 10 mM MgCl increases the apparent potency of ( R )- for 14-3-3/ER (pT ) (light orange circles vs dark orange squares) and 14-3-3/CaMKK2(pS ) (light blue triangles vs dark blue inverse triangles) PPI stabilization. “Relative stabilization” (y-axis, panels a, b and c) is defined in Figure 1. (c) Mg increases PPI stabilization efficacy of 10 M rac - (blue dots) and ( R ) - (green squares) whereas and ( S )- are unaffected (50 nM 14-3-3 /10 nM FITC- or 30 M 14-3-3 /10 nM FAM- ). “Relative stabilization” (y-axis) is mean fold-increase of signal at a given [Mg ] over baseline (no added Mg ). (d) Increase in apparent K d of 14-3-3/ER (pT ) (10 nM FITC- , left panel) or 14-3-3/CaMKK2(pS ) (10 nM FAM- , right panel) PPIs by FP with increasing ( R )- concentration, in the presence of 10 mM Mg . “Fold increase” (y-axis) is the mean fold-increase of signal over baseline (no ( R )- ). (e) Ca (red triangles) and Mn (purple squares) also increase apparent efficacy of ( R )- in 14-3-3/ER (pT ) FP assay. compatible with the high affinity observed by SPR. These strong chemical shift perturbations prevented a comprehensive determination of the binding site of (-)- to the 14-3-3 C/ complex by NMR. Nonetheless, two resonances with chemical shifts minimally but specifically affected by the presence of (-)- were identified. These signals, corresponding to G171 and I219 (Figure S8), lie in the FC pocket and were already affected by the presence of (-)- alone, indicating that (-)- binds to the FC pocket of both 14-3-3 C alone and in the 14-3-3 C/ complex. ( R )-2 is the active enantiomer of 2: reassignment of binding mode and absolute configuration by X-ray crystallography Having confirmed by NMR that (-)- binds to the FC pocket of the 14-3-3/ER (pT ) complex, we set out to obtain an X-ray crystal structure of the 14-3-3 /ER (pT ) phosphopeptide /(-)- ternary complex. By co-crystallization, we obtained crystals which diffracted reproducibly at a resolution of 1.85 Å. The 3.25 Å X-ray crystal structure of the ternary complex of ( S )- bound to the Nicotiana tabacum [25] (Figure 2a, magenta sticks) was initially used to model the binding orientation of . However, it was not possible to model ( S )- into the observed ligand-derived electron density of the 14-3-3 C/ /(-)- ternary complex. The 1.85 Å resolution was sufficient to model the ligand ab initio , confirming that (-)- binds to the FC pocket as well as unambiguously assigning the ( R ) absolute configuration to (-)- (Figure 2a, yellow sticks). This assignment agreed with that obtained by vibrational circular dichroism on (+)- (assigned ( S ), see Figure S23). Testing of (-)- and (+)- by SPR on the 14-3-3 /PMA2 PPI (Figure S10) showed that only (-)- stabilized this complex, as was the case for 14-3-3/ER (pT ). The fact that only (-)- stabilizes both PPIs, together with the higher resolution of the 14-3-3 C/ /(-)- ternary complex crystal structure compared to the published 14-3-3 /PMA2/ structure supports our revised binding mode of (Figure 2a) and reassignment of the absolute configuration for the active 14-3-3 PPI stabilizing enantiomer of as ( R ) (from this point in the text, (-)- and (+)- will be referred to as ( R )- and ( S )- respectively). Overall, ( R )- binds in a T-shaped conformation, with each of its phenyl rings pointing into separate sub-pockets (Figure 2b) and with all three phenyl rings oriented orthogonally to the pyrrolidone ring. The salicylate moiety of ( R )- occupies a narrow, largely negatively charged cleft (Figure 2c) not exploited by (Figure S11). The carboxylate group participates in a bidentate interaction with Arg41 and accepts a H-bond from Asn42. The carbonyl of the pyrrolidone ring participates in an attractive antiparallel interaction with the sidechain of Asn42. [30] The nitro group accepts a H-bond from Lys122 (Figure 2d) which is also in direct contact with the C-terminus of , thus ( R )- bridges 14-3-3 and via polar interactions. Unexpectedly, additional electron density was observed in conjunction with the ligand (Figure 2e), corresponding to a fully hydrated Mg ion (presumably derived from the Mg -containing crystallization buffer) chelated by the vinylogous carboxylate moiety of ( R )- . A p K a of 3.2 was measured (see Supporting Information) for the vinylogous carboxylate moiety of rac - , supporting the observation of the magnesium vinylogous carboxylate salt in the ternary complex. ( R )-2 but not 1 stabilizes the 14-3-3/CaMKK2(pS ) PPI Given the relatively high affinity of the 14-3-3/ interaction, we sought a 14-3-3 binding partner with a binary complex of similar overall structure to the 14-3-3/ER (pT ) structure (i.e. with a vacant FC pocket) but with lower 14-3-3 affinity, in order to investigate the dynamic range of PPI stabilization. We reasoned that such an interaction would also be useful to evaluate the 14-3-3 PPI stabilization specificity of ( R )- . pS of calcium/calmodulin-dependent kinase kinase 2 (CaMKK2) [31] was identified as a suitable candidate. Unlike the pT ER phosphosite from which phosphopeptide is derived, the N-terminal domain of CaMKK2 pS phosphosite is not a canonical mode III 14-3-3 binder. [32] However, the crystal structure of 14-3-3 in complex with a phosphopeptide derived from the CaMKK2 pS site (sequence RKLpS LQER, PDB ID: 6EWW) indicates that it mimics a canonical mode III binder and that the FC pocket in this complex is largely unoccupied and available for potential binding of stabilizers such as ( R )- . [31] We determined the binding affinity of CaMKK2 pS [31] to 14-3-3 by SPR (estimated K d
112 ± 14 M, Figure S12) and found it significantly lower than for 14-3-3 / ( K d
227 ± 14 nM). An FP assay was then developed based on a FAM-labelled derivative of (FAM- ). The stabilization of the 14-3-3/CaMKK2(pS ) PPI by rac -, ( R )- and ( S )- and natural product was then determined using by FP and SPR assays (Figure 3a and S13). Interestingly, despite an apparently vacant FC pocket, exhibited only borderline statistically significant stabilization of the 14-3-3/CaMKK2(pS ) PPI by FP (p 0.060) and SPR (p 0.038) at 200 M (purple inverted triangles). Following the same stereochemical dependence as for the 14-3-3/ER (pT ) PPI, ( R )- was a more effective stabilizer of the 14-3-3/CaMKK2(pS ) PPI than (green squares), while ( S )- showed no significant stabilization activity (red triangles). ( R )-2 exhibits metal ion-dependent 14-3-3 PPI stabilization potency With stabilization by ( R )- confirmed in two 14-3-3 PPIs, we investigated a potential role for the crystallographically-observed Mg chelation by ( R )- . Performing the FP and SPR assays in the presence of Mg led to an apparent increase in stabilization potency and maximum efficacy of ( R )- in both the 14-3-3/ER (pT ) and 14-3-3/CaMKK2(pS ) PPIs (Figure 3b and S14, S15). We then carried out Mg concentration-response experiments in the 14-3-3/ER (pT ) and 14-3-3/CaMKK2(pS ) FP assays at a fixed concentration (10 M) of rac - , ( R )- , ( S )- and (Figure 3c). The 14-3-3/ER (pT ) PPI stabilization efficacy of ( R )- and rac- (green squares and blue circles respectively) and 14-3-3/CaMKK2(pS ) PPI stabilization efficacy of ( R )- (brown diamonds) was shown to be magnesium concentration-dependent. ( S )- is not a PPI stabilizer and hence was unaffected by Mg concentration (red triangles). The 14-3-3/ER (pT ) PPI stabilization effect of (purple inverse triangles) was also unaffected by varying Mg concentration, indicating that the Mg effect is ligand-specific and not due to effects on 14-3-3 protein or the phosphopeptide (see also Figure S16, S17). The addition of 10 mM MgCl afforded full concentration-response curves for ( R )- , allowing the determination of apparent EC values. In the 14-3-3/CaMKK2(pS ) FP assay, adding 10 mM MgCl led to an apparent EC of 3.2 ± 0.3 M and 2.9-fold stabilization (Figure 3b). The PPI stabilizing potency and efficacy of the Mg salt of ( R )- on the 14-3-3/CaMKK2(pS ) PPI is comparable to that of on the 14-3-3/ER (pT ) PPI (EC M, 3.9-fold stabilization, Figure 1c). Taking the relative stabilization effect of ( R )- at 200 µM in the absence of added Mg as a baseline (1.5- and 1.7-fold for 14-3-3/ER (pT ) and 14-3-3/CaMKK2(pS ) PPIs respectively), the addition of 10 mM Mg leads to 7-fold and 99-fold increases in apparent potency (Figure 3b). Exploiting the potentiating effect of Mg on the PPI stabilization potency of ( R )- , we investigated the apparent affinity increase between 14-3-3 and phosphorylated binding partners induced by increasing concentrations of (R)-2 in the 14-3-3/ER (p T594 ) and 14-3-3/CaMKK2(pS ) FP assays (Figure 3d). At the highest concentration of (R)-2 tested (333 M), the observed affinity of the 14-3-3/ER (pT ) PPI (left panel) was increased 57-fold and 3,700-fold for the 14-3-3/CaMKK2(pS ) PPI (right panel). To assess whether this metal ion-assisted 14-3-3 PPI stabilization effect was specific to Mg , we tested by FP the effect of adding increasing concentrations of MnCl , CaCl and ZnCl to a fixed concentration of 14-3-3, FITC- and ( R )- on the overall 14-3-3/ER (pT ) PPI stabilization efficacy (Figure 3e and S17). The potency of the assistance effect of Mg and Ca were comparable, while Mn was approximately 15-fold more effective, with its maximum assistance effect achieved at less than 1 mM MnCl . [33] All three metals show similar overall stabilization efficacy in this setting, affording approximately 2-fold stabilization of the interaction. Addition of Zn caused precipitation of phosphopeptide FITC- precluding measurement of its effects. Mimicry of the chelate by cyclization or intramolecular H-bonds affords metal independent 14-3-3 PPI stabilizers
Based on the 14-3-3/ER (pT )/( R )- ternary complex crystal structure, the origin of the metal ion potentiation effect on the PPI stabilization potency of was assumed to be due to solution phase stabilization of the crystallographically observed binding conformation of ( R )- . To test this hypothesis, we decided to mimic the conformational restriction of the Mg chelate of ( R )- by moieties that would not depend on the chelation of metal ions. Cyclisation of rac - and its analogues to afford pyrazole derivatives such as has previously been reported as a strategy to improve their 14-3-3 PPI stabilization activity. [34] However, this was based on a binding mode (Figure 2b, magenta sticks) determined from a low-resolution crystal structure that we have now shown to be erroneous.
Rac- was reported to be a better stabilizer of the 14-3-3/PMA2 complex than rac -2 . A crystal structure of a similar bicyclic analogue of in the 14-3-3/PMA2 complex was reported, however ( S )- (the inactive absolute configuration) was used to model into the observed electron density, leading to difficulties in interpreting SAR due to incorrect stereochemistry and binding mode assumptions. In addition, based on the 6-membered Mg -containing chelate ring observed (Figure 2e), we reasoned that an intramolecular hydrogen bond contained in a 6-membered pseudo-ring, such as in the regioisomeric vinylogous amides and , would be a better chelate mimic than the 5-membered pyrazole ring. Reaction of rac - with aqueous ammonia in acetic acid under microwave heating afforded only vinylogous amide rac - in 20% yield, with no rac - observed (Scheme 1). This contrasts with literature reports that 4-aroyl substituents favor regiospecific endocyclic nucleophilic attack by sterically unencumbered amines to afford the regioisomers corresponding to . [35-40] Reaction of rac - with p-methoxybenzylamine under microwave heating afforded rac - (6%) and rac - (26%). Removal of the PMB group of rac - using TFA afforded rac - in 53% yield. The single enantiomers ( R )- and ( R )- were prepared analogously to rac - . Treatment of ( R )- with p-methoxybenzylamine yielded the two regioisomeric PMB-protected vinylogous amides ( R )- (10%) and ( R )- (16%), which on deprotection with TFA afforded ( R )- and ( R )- in 10% and 51% yield, respectively. ( R )- was prepared in 47% yield from ( R )- by reaction with hydrazine. N O ONHROHO N COOH N NHROO OHO N COOHN OOHO N COOHNHN rac - or ( R )- rac - (10%) rac - (84%)( R )- (47%) + rac - (6%)( R )- (10%) rac - (26%)( R )- (16%)( R )- (10%) rac - (53%)( R )- (51%) and R = PMB and R = Hiii iii iv iv
Scheme 1.
Synthesis of conformationally-restricted analogues of 2 . i) H NNH , AcOH, 2 h, 120 °C. ii) aq. NH , AcOH, 2h, 120 °C. iii) PMBNH , 2 h, 120 °C. iv) TFA, 30 min, 120 °C. ( R )- , ( R )- and ( R )- were then tested by FP and SPR for their 14-3-3/ER (pT ) and 14-3-3/CaMKK2(pS ) PPI stabilization effect (Figure 4a,b and S19). Gratifyingly, ( R )- (inverted blue triangles) and ( R )- (pink circles) were more potent and efficacious stabilizers of both PPIs than ( R )- (without added Mg , green squares). Pyrazole ( R )- (orange triangles) was somewhat more potent and efficacious than ( R )- alone but was less active than either ( R )- or ( R )- , suggesting that a 6-membered ring is preferable to a 5-membered ring. In the 14-3-3/CaMKK2(pS ) PPI, ( R )- +10 mM Mg is still the most potent, suggesting that the ordered water molecules of the Mg -( R )- chelate may contribute to binding affinity/PPI stabilization in this case. As expected, increasing concentrations of Mg did not affect the 14-3-3/ER (pT ) PPI stabilization efficacy of rac - , or (Figure 4c, see Figure S18 for PPI stabilization data in absence of Mg ) as they lack the chelation-competent vinylogous carboxylate moiety. The salicylate moiety of , , and can potentially chelate metals, but the fact that , or did not show metal-dependent potency (Figure 4c) suggests that the salicylate moiety does not affect the potency of these compounds in our system. Therefore the effect of chelation by the salicylate moiety can be discounted for our purposes. Potency increase upon metal chelation is due to stabilization of bioactive conformation
In the 14-3-3/ER (pT )-bound conformation of (Figure 2a), the O - and carbonyl oxygen of the vinylogous carboxylate moiety are in a syn relationship ( syn -( R )- , Figure 5a). In solution, and in the absence of chelatable metal ions, repulsive electrostatic allylic strain was expected to disfavor this conformation relative to the anti conformation ( anti -( R )- , Figure 5a). Comparing the DFT calculated gas phase free energies of syn - and anti -( R )- (see Supporting Information) shows the protein-bound and Figure 4. Metal ion-independent 14-3-3 PPI stabilizers. (a,b) Stabilization of the 14-3-3/ER (pT ) and 14-3-3/CaMKK2(pS ) PPIs by ( R )- and - as measured by FP. “Relative stabilization” (y-axes) is defined in Figure 1. (c) Mg concentration-response in 14-3-3/ER (pT ) FP assay (50 nM 14-3-3 /10 nM FITC- and 100 M rac - , - , - or - ), showing no effect of Mg on activity of rac - , - or - (orange squares, pale blue triangles and inverted pink triangles respectively). “Relative stabilization” (y-axis) refers to mean fold increase of FP signal of a given [Mg ] over signal observed in absence of added [Mg ]. chelation-stabilized syn -( R )- conformation to be significantly higher energy. Accounting for solvation effects by using two different implicit solvation models for water reduced the energy difference between syn- and anti- ( R )- (Table 1). Both confirm that the syn -( R )- conformation is not favored in water. Therefore, the anti -( R )- conformation is predicted to be present in significant amounts in the solution phase in the absence of Mg or other metals. For the vinylogous amides ( R )- and ( R )- , both solvation models favor the corresponding syn - Table 1.
Calculated relative energies for syn and anti -conformations of ( R )- , - and - . Cpd Conformation of exocyclic carbonyl a Free energies in water B3LYP-D3 PBF b M06-2X-D3 SM6 c ( R )- syn
0 2.2 ( R )- anti R )- syn R )- anti R )- syn R )- anti R )- (Figure 5). [b] PBF = Poisson Boltzmann Finite element method; a solvation model. [c] SM6 = Solvation Model 6. conformations (i.e. corresponding to the protein bound conformation of ) as the lowest energy solution conformation, as expected and as borne out by their increased, metal ion-independent 14-3-3 PPI stabilization compared to ( R )- . To confirm the calculations, 1D ROE difference NMR experiments were performed on rac - , rac - and rac - (Figures S20-S22) to determine their preferred solution conformations. Irradiation of protons of either the benzoyl or nitrophenyl rings of rac - showed only a small resonance transfer between these rings, suggesting that they prefer not to be in close spatial proximity and corresponding to the anti conformation (Figure 5a). On the other hand, strong resonance transfer was observed between the protons of the benzoyl and nitrophenyl rings for both rac - and rac - , indicating close spatial proximity suggestive of the syn conformation induced by intramolecular hydrogen bonding. Precipitation precluded determination by NMR of the conformational effects of Mg on rac - , but these may be inferred from the X-ray crystal structure (Figure 5b). Figure 5 . Conformational analysis of 2 and binding mode of ( R )-6 in 14-3-3/ER (pT ) complex. (a) Comparison of syn and anti conformations of the vinylogous carboxylate moiety of ( R )- . Calculated free energies of ( R )- conformations show that protein-bound conformation syn -( R )- is not favored in solution. (b) and (c), Binding modes and 2F obs -F cal electron density (contoured at 1 ) of the magnesium complex of ( R )- (panel b, PDB 6TJM) and ( R )- (panel c, PDB 6TL3), shown in the same orientation of the 14-3-3/ER (pT ) complex, indicating absence of chelated metal ion in ( R )- structure and intramolecular hydrogen bond-stabilized conformation of ( R )- that mimics syn -( R )- conformation. Rac - was then selected for crystallization studies in the 14-3-3/ER (pT ) complex. A 2.45 Å resolution X-ray crystal structure of the ternary 14-3-3 C/ /( R )- complex was obtained (Figure 5c), confirming the absolute configuration of the active enantiomer as ( R ) and a binding mode analogous to ( R )- . No additional ligand-associated electron density attributable to a chelated metal ion was observed, indicating that the intramolecular hydrogen bond successfully stabilizes the syn conformation as demonstrated computationally.
Discussion
Ligand conformational restriction is a widely used strategy to increase potency, and minimizes the entropic loss observed on ligand binding if the preferred solution and bound conformations differ. [41-43]
This is typically achieved by cyclisation, introducing intramolecular hydrogen-bonding or by exploiting steric/stereoelectronic effects. Ligand chelation by solution phase metal ions can now be added to this list of strategies. However, given the lack of control over metal ion concentration in medicinally-relevant settings ( e.g. intracellularly) and the requirement that the binding site accommodate a metal ion and associated ordered water molecules, it may prove to be difficult to easily and reliably exploit this effect in drug substances. On the other hand, metal contamination has been reported to be a cause of false positives due to assay interference [44] or by ligands complexing to bioactive metal centers. [45]
To our knowledge, no incidences of such ligand-specific conformational effects due to metal ions have been reported. Given that many assays require the presence of metal ions (such as Ca , Mg or Zn ) and high intracellular concentrations of some of these metals, such conformational effects should be considered as a potential cause of difficult to interpret assay data for chelation competent ligands. In principle, such behavior could lead to false negatives in screening campaigns and should be considered when analyzing screening data, especially in the case of conflicting results from samples that may have been subjected to more or less rigorous purification procedures which could result in different metal ion content. This effect could also be at play if unexpected discrepancies are observed between potencies in biochemical and cellular assays, where there may be differences in metal ion concentration. It has been proposed that chelating moieties should be used in substructure alerts to filter out compounds with potential to cause assay interference and frequent hitters. [22] However, the application of such filters needs to be considered on a case by case basis, with the assay technology and particular target in question taken into account. As we have shown, in some cases metal chelation can act as a ligand conformational probe and provide valuable hints as to the optimizability of low potency hits. Conclusion
We have shown that by leveraging the serendipitously discovered Mg -chelate and resulting chelation-controlled stabilization of bioactive ligand conformation, ( R )- can be optimized to afford stabilizers of canonical 14-3-3 PPIs with potency rivaling natural product . We also show that this metal-assisted PPI stabilization effect is due to solution phase stabilization of the bioactive conformation of and can be mimicked by an intramolecular H-bond to afford metal-independent PPI stabilizers such as ( R )- and ( R )- . Compared to , these compounds have the advantage of higher synthetic tractability and stabilize a 14-3-3 PPI that is refractive to stabilization by and they should be useful tool compounds for the study of 14-3-3 PPI stabilization. This work gives new insights into the SAR of small molecule, non-natural product-derived 14-3-3 PPI stabilization and provides opportunities for structure-based drug design to identify new, small molecule 14-3-3 PPI stabilizers. More broadly, this work indicates that ligand-specific conformational effects due to metal ion chelation should be considered during the interpretation of assay and screening data, especially for chelation-competent ligands. Acknowledgements
The crystallographic data collection was performed at the Deutsches Elektronen-Synchrotron (DESY, PETRA III) of Hamburg (Germany). The authors acknowledge Anna Jonson and Kristina Öhlén of the Separation Science Group at AstraZeneca Gothenburg for help with chiral HPLC, Richard Lewis of the NMR Group at AstraZeneca Gothenburg for the VCD analysis and the interpretation of the 1D selective ROE NMR spectra, Frederik Wågberg and Johan Wernevik for assistance with HRMS determination for compounds ( R )- and ( R )- . Anaïs Noisier is thanked for her support in the peptide synthesis. François-Xavier Cantrelle is also thanked for NMR data acquisition and technical advice. This work is supported by the Initial Training Network TASPPI, funded by the H2020 Marie Curie Actions of the European Commission under Grant Agreement 675179. The NMR facilities at Univ. Lille 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 co-funded by the European Union with the European Regional Development Fund (ERDF), by the Hauts de France Regional Council (contract n° 17003781), Métropole Européenne de Lille (contract n° 2016_ESR_05), and French State (contract n° 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). Keywords:
Chelates • Drug design • Medicinal chemistry • Protein-protein interaction stabilization • 14-3-3 [1] J. A. Wells, C. L. McClendon,
Nature , , 1001-1009. [2] J. Menche, A. Sharma, M. Kitsak, S. D. Ghiassian, M. Vidal, J. Loscalzo, A. L. Barabasi, Science , , 1257601. [3] M. Vidal, M. E. Cusick, A. L. Barabasi, Cell , , 986-998. [4] T. Ideker, R. Sharan, Genome Res. , , 644-652. [5] G. Zinzalla, D. E. Thurston, Future Med. Chem. , , 65-93. [6] S. Jaeger, P. Aloy, IUBMB Life , , 529-537. [7] D. E. Scott, A. R. Bayly, C. Abell, J. Skidmore, Nat. Rev. Drug Discovery , , 533-550. [8] P. Thiel, M. Kaiser, C. Ottmann, Angew. Chem. Int. Ed. , , 2012-2018. [9] F. Giordanetto, A. Schafer, C. Ottmann, Drug Discov Today , , 1812-1821. [10] V. Azzarito, K. Long, N. S. Murphy, A. J. Wilson, Nat. Chem. , , 161-173. [11] E. Sijbesma, K. K. Hallenbeck, S. Leysen, P. J. de Vink, L. Skora, W. Jahnke, L. Brunsveld, M. R. Arkin, C. Ottmann, J. Am. Chem. Soc. , , 3524-3531. [12] S. Surade, T. L. Blundell, Chem. Biol. , , 42-50. [13] A. Aitken, Semin. Cancer Biol. , , 162-172. [14] H. Fu, R. R. Subramanian, S. C. Masters, Annu. Rev. Pharmacool. Toxicol. , , 617-647. [15] H. Hermeking, A. Benzinger, Semin. Cancer Biol. , , 183-192. [16] A. J. Muslin, H. Xing, Cell. Signalling , , 703-709. [17] T. Obsil, R. Ghirlando, D. C. Klein, S. Ganguly, F. Dyda, Cell , , 257-267. [18] L. M. Stevers, E. Sijbesma, M. Botta, C. MacKintosh, T. Obsil, I. Landrieu, Y. Cau, A. J. Wilson, A. Karawajczyk, J. Eickhoff, J. Davis, M. Hann, G. O'Mahony, R. G. Doveston, L. Brunsveld, C. Ottmann, J. Med. Chem. , , 3755-3778. [19] M. B. Yaffe, K. Rittinger, S. Volinia, P. R. Caron, A. Aitken, H. Leffers, S. J. Gamblin, S. J. Smerdon, L. C. Cantley, Cell , , 961-971. [20] B. Coblitz, M. Wu, S. Shikano, M. Li, FEBS Lett. , , 1531-1535. [21] P. W. Kenny, J. Sadowski, in Chemoinformatics in Drug Discovery (Eds.: R. Mannhold, H. Kubinyi, G. Folkers, T. I. Oprea), Wiley VCH, , pp. 271-285. [22] K. Schorpp, I. Rothenaigner, E. Salmina, J. Reinshagen, T. Low, J. K. Brenke, J. Gopalakrishnan, I. V. Tetko, S. Gul, K. Hadian,
J. Biomol. Screening , , 715-726. [23] J. B. Baell, G. A. Holloway, J. Med. Chem. , , 2719-2740. [24] I. J. De Vries-van Leeuwen, D. da Costa Pereira, K. D. Flach, S. R. Piersma, C. Haase, D. Bier, Z. Yalcin, R. Michalides, K. A. Feenstra, C. R. Jimenez, T. F. de Greef, L. Brunsveld, C. Ottmann, W. Zwart, A. H. de Boer, Proc. Natl. Acad. Sci. U. S. A. , , 8894-8899. [25] R. Rose, S. Erdmann, S. Bovens, A. Wolf, M. Rose, S. Hennig, H. Waldmann, C. Ottmann, Angew. Chem. Int. Ed. , , 4129-4132. [26] M. Wurtele, C. Jelich-Ottmann, A. Wittinghofer, C. Oecking, EMBO J. , , 987-994. [27] V. Obsilova, P. Herman, J. Vecer, M. Sulc, J. Teisinger, T. Obsil, J. Biol. Chem. , , 4531-4540. [28] C. Dalvit, G. Fogliatto, A. Stewart, M. Veronesi, B. Stockman, J. Biomol. NMR , , 349-359. [29] J. F. Neves, I. Landrieu, H. Merzougui, E. Boll, X. Hanoulle, F. X. Cantrelle, Biomol. NMR Assignments , , 103-107. [30] F. H. Allen, C. A. Baalham, J. P. M. Lommerse, P. R. Raithby, Acta Crystallogr. Sect. B: Struct. Sci. , , 320-329. [31] K. Psenakova, O. Petrvalska, S. Kylarova, D. Lentini Santo, D. Kalabova, P. Herman, V. Obsilova, T. Obsil, Biochim. Biophys. Acta Gen. Subj. , , 1612-1625. [32] S. Ganguly, J. L. Weller, A. Ho, P. Chemineau, B. Malpaux, D. C. Klein, Proc. Natl. Acad. Sci. U. S. A. , , 1222-1227. [33] C. W. Bock, A. K. Katz, G. D. Markham, J. P. Glusker, J. Am. Chem. Soc. , , 7360-7372. [34] A. Richter, R. Rose, C. Hedberg, H. Waldmann, C. Ottmann, Chem. – Eur. J. , , 6520-6527. [35] M. N. Armisheva, N. A. Kornienko, V. L. Gein, M. I. Vakhrin, Russ. J. Gen. Chem. , , 1893-1895. [36] L. F. Gein, V. L. Gein, I. A. Kylosova, Z. G. Aliev, Russ. J. Org. Chem. , , 252-254. [37] V. L. Gein, M. N. Armisheva, N. A. Kornienko, L. F. Gein, Russ. J. Gen. Chem. , , 2270-2272. [38] V. L. Gein, N. L. Fedorova, E. B. Levandovskaya, M. I. Vakhrin, Russ. J. Org. Chem. , , 95-99. [39] V. L. Gein, N. N. Kasimova, Russ. J. Gen. Chem. , , 254-260. [40] V. L. Gein, N. N. Kasimova, Z. G. Aliev, M. I. Vakhrin, Russ. J. Org. Chem. , , 875-883. [41] Z. Fang, Y. Song, P. Zhan, Q. Zhang, X. Liu, Future Med. Chem. , , 885-901. [42] A. D. G. Lawson, M. MacCoss, J. P. Heer, J. Med. Chem. , , 4283-4289. [43] Y. Zheng, C. M. Tice, S. B. Singh, Bioorg. Med. Chem. Lett. , , 2825-2837. [44] J. C. Hermann, Y. Chen, C. Wartchow, J. Menke, L. Gao, S. K. Gleason, N. E. Haynes, N. Scott, A. Petersen, S. Gabriel, B. Vu, K. M. George, A. Narayanan, S. H. Li, H. Qian, N. Beatini, L. Niu, Q. F. Gan, ACS Med. Chem. Lett. , , 197-200. [45] F. E. Morreale, A. Testa, V. K. Chaugule, A. Bortoluzzi, A. Ciulli, H. Walden, J. Med. Chem. , , 8183-8191. S10
Entry for the Table of Contents
No PAINS, No Gain.
Chelation is associated with pan-assay interference compounds (PAINS) but can also lead to true potency gains and be exploited to guide medicinal chemistry optimization. Bivalent metal ions increased the potency of a marginally active 14-3-3 PPI stabilizer up to 100-fold via stabilization of the bioactive ligand conformation. Mimicry of this by intramolecular H-bonds lead to the first potent, drug-like 14-3-3 PPI stabilizers.
S11
SUPPLEMENTARY INFORMATION
METAL-ASSISTED SMALL MOLECULE PROTEIN-PROTEIN INTERACTION STABILIZATION BY CHELATION-CONTROLLED BIOACTIVE LIGAND CONFORMATION STABILIZATION AND RATIONAL DESIGN OF 14-3-3 PROTEIN-PROTEIN INTERACTION STABILIZERS
Francesco Bosica, †,
Sebastian Andrei,
João Filipe Neves, ‡ Peter Brandt, † Anders Gunnarsson, ┴ Isabelle Landrieu, ‡ Christian Ottmann, and Gavin O’Mahony †, * TABLE OF CONTENTS I. SUPPLEMENTARY
FIGURES ........................................................................
S12
II.
NMR
DATA ................................................................................................
S43 II A ) P RODUCTION OF N H LABELED ‐ ‐ σ Δ C FOR H ‐ N TROSY ‐ HSQC
NMR
SPECTROSCOPY . ...................................................................................................... S43 II B ) H ‐ N TROSY ‐ HSQC
NMR
EXPERIMENTS ..........................................................
S43 II C ) W ATER
LOGSY
NMR
EXPERIMENTS .....................................................................
S43
III.
PROTEIN
PURIFICATION
AND X ‐ RAY
CRYSTALLOGRAPHY ..........................
S44
IV.
CHEMISTRY
SECTION ..............................................................................
S46 IV A ) G ENERAL
INFORMATION ....................................................................................
S46 IV B ) S YNTHETIC
PROCEDURES
AND
COMPOUND
CHARACTERIZATION ..................................
S46 IV C ) C HIRAL
SEPARATION , VCD
ANALYSIS
AND
RACEMISATION
STUDIES OF ....................... S54 IV D ) FITC ‐ PEPTIDE
SYNTHESIS ................................................................................
S67 IV E ) p K a DETERMINATION ........................................................................................
S70 IV F ) S PECTROSCOPIC
DATA OF PRODUCTS ....................................................................
S71 V. COMPUTATIONAL
DETAILS ......................................................................
S110
VI.
SUPPORTING
REFERENCES ....................................................................
S120
S12 I. Supplementary Figures
Supplementary Figure 1.
Determination of affinity ER -derived phosphopeptide for 14-3-3 by SPR. a) Representative example of an SPR sensorgram in which increasing concentrations of were flowed over immobilized 14-3-3 . The response units (RUs) achieved (y axis) are presented as a function of time in seconds (x axis). For each curve, the RU values at equilibrium response were extracted and fitted in a dose-response curve using a four-parameter logistic model (4PL), shown in (b), against the log of the molar concentration of (mean ± SD, n = 3). To account for the variations in protein immobilization between runs, the equilibrium RU values were normalized to a 0-100 % interval, where 0% is baseline response and 100% is the mean curve plateau value. Experiment was performed in 1:3 dilution series from an initial concentration of of 500 μ M. S13
Supplementary Figure 2
Stabilization effect of Fusicoccin A for the 14-3-3 / complex, measured by SPR. was titrated (1:3 dilution series, initial concentration 100 μ M) in the presence of 50 nM Er α -derived phosphopeptide and surface-immobilized 14-3-3 . Binding affinity was estimated to be EC = 1.8 ± 0.3 μ M by curve fitting using a four-parameter logistic model (4PL). Error bars show standard deviation from the mean for each data point (n = 3). To account for the variations in protein immobilization between runs, the equilibrium RU values were normalized to a 0-100 % interval, where 0% is baseline response and 100% is the mean curve plateau value. S14
Supplementary Figure 3 . Comparison of the stabilization effect of rac - and for the 14-3-3 /ER complex measured by SPR (in the absence of added Mg ). Compound ( rac - or ) concentration (x axis) is plotted as the log of the compound concentration in molar. (a) “Total effect” is the total RUs afforded by adding compound ( rac- or ) to immobilised 14-3-3 in the presence of 50 nM (“ only”, indicated by dashed line, 9.4 ± 1.6 RUs), i.e. (affinity for 14-3-3 affinity for 14-3-3 / complex + stabilization of 14-4-4 / interaction), (b) Determination of contribution of compound affinity for immobilized 14-3-3 protein in the absence of peptide , c) “Net stabilization effect” ( Δ RUs) is defined as (total RUs from 14-3-3 / /compound)-((RUs from 14-3-3 / interaction) + (RUs from 14-3-3 /compound interaction)). S15 Supplementary Figure 4 . WaterLOGSY NMR experiments indicate that (-)- binds to 14-3-3 σΔ C both in the absence and presence of . H spectrum (blue) and WaterLOGSY spectra of 500 µM (-)- in the presence of either 25 µM 14-3-3 σΔ C (red) or 25 µM 14-3-3 σΔ C+35 µM phosphopeptide (green). S16 Supplementary Figure 5. H- N TROSY-HSQC spectra of 100 µM N H labeled 14-3-3 σΔ C alone (black), or in the presence of 2000 µM (-)- (superimposed in blue). S17 Supplementary Figure 6 . Binding site of (-)- on 14-3-3 σΔ C can be identified by reporting the intensity ratio (I/I ) for each pair of corresponding resonances in the analyzed spectrum (I) compared to the control spectrum (I ). Plot of the I/I values of H- N correlation peak intensities in the spectrum of 100 µM N H labeled 14-3-3 σΔ C in the presence of 2000 µM (-)- (I), compared to corresponding resonances in the reference spectrum of 100 µM 14-3-3 σΔ C (I ) (y axis) versus σΔ C amino acid sequence (x axis, not proportional to sequence length). A total of 131 correlation peak intensity ratios are shown. The helices of 14-3-3 σΔ C are identified below the x axis as blue cartoons, while disordered regions are represented by red lines. S18
Supplementary Figure 7. H- N TROSY-HSQC spectra of 100 µM N H labeled 14-3-3 σΔ C alone (black), or in the presence of 60 µM ER α phosphopeptide (superimposed in red). Note that for some resonances, both the free and the bound form are observed, suggesting a slow exchange regime on the NMR time scale. S19 Supplementary Figure 8. H- N TROSY-HSQC experiments show that (-)-2 binds in the FC pocket. a) H- N TROSY-HSQC spectra of 100 µM N H labeled 14-3-3 σΔ C alone (black), or in the presence of: 60 µM ER α phosphopeptide (superimposed in red); 60 µM ER α phosphopeptide + 1000 µM (-)- (superimposed in blue) or 60 µM ER α phosphopeptide + 2000 µM (-)- (superimposed in green). The spectral regions delimited by orange dashes are enlarged in b and c. (b,c) Overlaid enlarged spectral regions showing the resonances corresponding to G171 (b) and I219 (c). 1000 µM (-)- induced broadening of the resonance of G171 (blue spectrum), while 2000 µM (-)- induced broadening beyond detection (green spectrum). For I219, broadening beyond detection of the resonance (blue spectrum) was observed already at 1000 µM (-)- . S20 Supplementary Figure 9 . Comparison of the stabilization effect of rac - , ( R )- and ( S )- for the 14-3-3 ζ :ER α phosphopeptide complex, measured by SPR (in the absence of added Mg ). Compound ( rac -, ( R )- or ( S )- ) concentration (x axis) is plotted as the log of the compound concentration in molar. (a) “Total effect” is the total RUs afforded by adding compound ( rac -, ( R )- or ( S )- ) to immobilized 14-3-3 in the presence of 50 nM ER α phosphopeptide (“ only”, indicated by dashed line, 10.4 ± 1.2 RUs), i.e. (affinity for 14-3-3 + affinity for 14-3-3 / complex + stabilization of 14-4-4 / interaction), (b) Determination of compound ( rac -, ( R )- or ( S )- ) affinity for immobilized 14-3-3 protein in the absence of peptide , (c) “Net stabilization effect” ( Δ RUs) is defined as (total RUs from 14-3-3 / /compound)-((RUs from 14-3-3 / interaction) + (RUs from 14-3-3 /compound interaction)). S21 Supplementary Figure 10 . Measurement by SPR of the stabilization of the 14-3-3 η :PMA2 complex by (blue dots), ( R )- (green squares), ( S )- (red triangles) and (purple inverse triangles) in the presence of 10 mM MgCl . Response units (RUs) are plotted vs the log of compound concentration (M). In the experiment, PMA2 peptide was immobilised on a CMD200M chip via EDC/NHS coupling chemistry, at 8000 RUs. 14-3-3 η (10 μ M) and increasing concentrations of compound (1:2 dilution series, 100 μ M initial concentration, n = 1) were premixed and injected at a flowrate of 20 μ L/min and 20 °C for or 120 s in running buffer, followed by a single injection of 0.5% SDS for 60 s, as regeneration step. S22
Supplementary Figure 11 . Comparison of the binding modes in the FC pocket of Fusicoccin A (from PDB 4JDD, dark blue sticks) and ( R )- (yellow sticks) in complex with 14-3-3 σΔ C/ER α phosphopeptide (14-3-3 σΔ C protein surface coloured according to electrostatic potential, ER α phosphopeptide surface coloured green) showing that the salicylate moiety of ( R )- occupies a subpocket not exploited by Fusicoccin A . S23 Supplementary Figure 12 . Determination of affinity of CaMKK2 phosphopeptide for 14-3-3 by SPR (a) Representative example of an SPR sensorgram in which increasing concentrations of were flowed over immobilized 14-3-3 . The Response Units (RUs) achieved (y axis) are presented as a function of time in seconds (x axis). For each curve, the values at equilibrium response (i.e. binding coverage) where extrapolated and fitted in a dose-response curve using a four-parameter logistic model (4PL), shown in (b), against the log of the molar concentration of (mean ± SD, n = 3). Titration was performed in 1:3 dilution series from an initial concentration of of 667 μ M. S24
Supplementary Figure 13
Comparison of the stabilization effect of and rac -, ( R )- or ( S )- for the 14-3-3 /CaMKK2 complex measured by SPR (in the absence of added Mg ). Compound concentration (x axis) is plotted as the log of the compound concentration in molar. (a) “Total effect” is the total RUs afforded by adding compound ( , rac -, ( R )- or ( S )- ) to immobilised 14-3-3 in the presence of 30 µM (“ only”, indicated by dashed line, 25.1 ± 4.9 RUs), i.e. (affinity for 14-3-3 + affinity for 14-3-3 / complex + stabilization of 14-3-3 / interaction), (b) Determination of contribution of compound affinity for immobilized 14-3-3 protein in the absence of peptide 3, (c) “Net stabilization effect” ( Δ RUs) is defined as (total RUs from 14-3-3 / /compound)-((RUs from 14-3-3 / interaction)+(RUs from 14-3-3 /compound interaction)). S25 Supplementary Figure 14 . Determination of the effect of MgCl on the stabilization of rac - , ( R )- ( S )- and towards the 14-3-3 /ER α complex measured by FP (a) and SPR (b). (a) Concentration-response of compound in FP assay (10 nM FITC- and 50 nM 14-3-3 in the absence of MgCl (orange circles), 1 mM MgCl (light blue squares), 5 mM MgCl (red triangles) or 10 mM MgCl (gold inverse triangles). “Relative stabilization” (y-axis) is the mean fold-increase of FP signal over baseline (i.e. FP signal from interaction between 14-3-3 and ER α phosphopeptide FITC- alone), (b) Concentration-response of compound in SPR assay in the absence of MgCl (orange circles) or with 10 mM MgCl (gold squares). Compound titration was performed in the presence of 50 nM ER α phosphopeptide “Relative compound effect” is the mean-fold-increase of SPR signal over baseline (without correcting for compound binding to 14-3-3 alone). The error bars indicate +/- SD (n =3). S26 Supplementary Figure 15 . Determination of the effect of MgCl on the stabilization of , ( S )- and towards the 14-3-3 /CaMKK2 complex measured by FP (a) and SPR (b). a) Concentration-response of compound in FP assay (10 nM FAM- and 30 µM 14-3-3 in the absence of MgCl (orange circles), 1 mM MgCl (light blue squares), 5 mM MgCl (red triangles) or 10 mM MgCl (gold inverse triangles). “Relative stabilization” (y-axis) is the mean fold-increase of FP signal over baseline (i.e. FP signal from interaction between 14-3-3 and CaMKK2 phosphopeptide FAM- alone), b) Concentration-response of compound in SPR assay in the absence of MgCl (orange circles) or with 10 mM MgCl (gold squares). Compound titration was performed in the presence of 30 µM CaMKK2 phosphopeptide “Relative compound effect” is the mean-fold-increase of SPR signal over baseline (without correcting for compound binding to 14-3-3 alone). The error bars indicate +/- SD (n =3). S27 Supplementary Figure 16 . Test for assay interference induced by aggregation. Determination of effect of different detergent (Tween20 – P20) concentrations on the affinity of the ER α phosphopeptide FITC- for 14-3-3 (panel a) and on the stabilization effect promoted by ( R )- for the 14-3-3 :ER α complex (panel b) measured by FP and in the presence of 10 mM Mg . (A) The FP response (mP, y axis) achieved is plotted against the log of the 14-3-3 concentration (in molar) at different detergent concentrations: 0.05% (black dots), 0.10% (pink squares) and 0.20% (teal triangles). (B) “Relative stabilization” (expressed as mean fold-increase of FP signal over baseline, i.e. interaction between 14-3-3 and ER α phosphopeptide FITC- alone) is plotted versus increasing concentrations of ( R )- . The error bars indicate +/- SD (n =3). S28 Supplementary Figure 17 . Effect of bivalent metal ions on stabilisation of 14-3-3/ER and 14-3-3/CaMKK2 PPIs by ( R )- in FP assay and various counter-screens. “Relative stabilization” (y-axis) is the mean fold-increase of FP signal over baseline (i.e. interaction between 14-3-3 ζ and phosphopeptides FITC- or FAM- alone). Relative stabilization (black circles) is plotted versus metal ion concentration (1:3 dilution series, 50 mM initial concentration) at fixed concentrations of ( R )- (10 μ M), phosphopeptide (10 nM FITC- in panels a-c and FAM- in panel d) and 14-3-3 (50 nM in panels a-c, 30 μ M in panel d). For the 14-3-3 /ER α complex magnesium (panel a), calcium (panel b) and manganese (panel c) were tested, while for the 14-3-3 :CaMKK2 complex only magnesium was tested (panel d). To rule out protein independent effects, and/or unspecific binding, counter-screens with compound + peptide only (pink squares) and compound alone (teal triangles) were performed. S29 Supplementary Figure 18 . Stabilization effects of rac - , rac - and rac - (conformationally restricted analogues of rac - ) measured in FP assays in the absence of added MgCl . (a) Stabilization of the 14-3-3/ER α PPI (10 nM FITC- , 50 nM 14-3-3 ), (b) Stabilization of the 14-3-3 /CaMKK2 PPI (10 nM FAM- , 30 µM 14-3-3 ζ ). “Relative stabilization” (y-axis) refers to mean fold increase of FP signal at a given [compound] over signal observed in absence of compound. The error bars indicate +/- SD (n =3). S30
Supplementary Figure 19a . H NMR spectrum of rac -2 in D O, with assignments.
Rac- (2.5 mg, 5.43 µmol) was dissolved in 1 mM NaOH in D O (109 µL, 10.86 µmol) transferred into a 3 mm NMR tube and D O added to bring final volume to 160 µL. S31
Supplementary Figure 19b . COSY NMR of rac -2.
S32
Supplementary Figure 19c . 1D selective ROESY of rac -2 (excitation of H3, 6.06 ppm). S33 Supplementary Figure 19d . 1D selective ROESY of rac -2 (excitation of H18-H22, 7.50 ppm). S34 Supplementary Figure 19e . 1D selective ROESY of rac -2 (excitation of H13-H17, 7.61 ppm). S35 Supplementary Figure 20a . H NMR of rac - in D O solvent, with assignments.
Rac - (3.0 mg, 6.53 µmol) was dissolved in 1 mM NaOH in D O (131 µL, 13.06 µmol) transferred into a 3 mm NMR tube and D O was added to bring final volume to 160 µL. S36
Supplementary Figure 20b . COSY NMR of rac - S37
Supplementary Figure 20c . 1D selective ROESY of rac - (excitation of H2, 6.22 ppm). S38 Supplementary Figure 20d . 1D selective ROESY of rac - (excitation of H17-H21, 7.05 ppm). S39 Supplementary Figure 20e . 1D selective ROESY of rac - (excitation of H12-H16, H13-H15 overlapping, 7.34 ppm). S40 Supplementary Figure 21a . H NMR of rac - with in DMSO-d6 solvent, with assignments. Rac - (5.2 mg, 11 µmol) was dissolved in DMSO-d6 (160 µL) and transferred into a 3 mm NMR tube. For full spectrum assignments, refer to NMR spectra of in the spectroscopic data section (pages S70-S73). S41 Supplementary Figure 21b . 1D selective ROESY of rac - (excitation of H15, 6.36 ppm). S42 Supplementary Figure 21c . 1D selective ROESY of rac - (excitation of H17-H21, 6.93 ppm). S43 II.
NMR data
IIa) Production of N H labeled 14-3-3 σΔ C for H- N TROSY-HSQC NMR spectroscopy.
The N H labeled ‐ ‐ σΔ C ( Δ C17, cleaved after
T231) for
NMR studies was expressed in E. coli BL21 (DE3) cells transformed with a pProExHtb vector carrying the cDNA to express an N ‐ terminally His ‐ tagged human ‐ ‐ σΔ C. Bacterial cells were grown in L of deuterated M9 minimal medium supplemented with g/L C H Glucose, g/L N Ammonium
Chloride, g/L
Isogro N C H Powder – Growth
Medium (Sigma
Aldrich) and µg/mL ampicillin.
The recombinant protein was then purified from the bacterial extract by affinity chromatography using a Ni ‐ NTA column (GE
Healthcare).
The
His ‐ tag was further cleaved by the TEV protease.
The protein was finally dialyzed overnight at against NMR buffer (100 mM Sodium
Phosphate, pH mM NaCl), concentrated, aliquoted, flash frozen and stored at ‐ °C. A detailed protocol was previously published can be found at ref . IIb) H- N TROSY-HSQC NMR experiments H ‐ N TROSY ‐ HSQC (Transverse
Relaxation
Optimized
Spectroscopy ‐ Heteronuclear
Single
Quantum
Coherence
Spectroscopy) spectra were acquired at the temperature of °C in mm tubes (sample volume μ L) using a MHz
Bruker
Avance
Neo spectrometer, equipped with a cryoprobe. All samples were prepared in a buffer containing mM sodium phosphate, mM NaCl, pH (v/v) DMSO ‐ d6, mM DTT,
EDTA ‐ free protease inhibitor cocktail (Roche, Basel,
Switzerland) and (v/v) D O. The experiments were recorded with complex data points in the direct dimension and complex data points in the indirect dimension, with scans per increment. For the evaluation of the binding of ( ‐ ) ‐ to ‐ ‐ σΔ C, spectra of N H labeled ‐ ‐ σ µM were recorded in the presence and absence of µM ( ‐ ) ‐ . For the evaluation of the binding of ( ‐ ) ‐ to the ‐ ‐ σΔ C/ER α complex, spectra of N H labeled ‐ ‐ σΔ C µM with ER α peptide were recorded in the presence and absence of µM and µM ( ‐ ) ‐ . For the evaluation of the binding of (+) ‐ to ‐ ‐ σΔ C, spectra of N H labeled ‐ ‐ σ µM were recorded in the presence and absence of µM (+) ‐ . The reference for the H chemical shift was relative to TMSP (trimethylsilylpropanoic acid) while N chemical shift values were referenced indirectly. Assignments of the backbone resonances of N H labeled ‐ ‐ σΔ C were previously reported. Spectra were collected and processed with
Topspin (Bruker
Biospin,
Karlsruhe,
Germany) and analyzed with
Sparky (T. D. Goddard and D. G. Kneller,
SPARKY University of California,
San
Francisco).
IIc) WaterLOGSY NMR experiments
WaterLOGSY spectra were acquired at the temperature of °C in mm tubes (sample volume µL) using a MHz
Bruker
Avance
III HD spectrometer equipped with a CPQCI cryogenic probe.
The spectra were recorded with complex data points, with scans per increment and with a mixing time of (acquisition time of minutes). All samples were prepared in a buffer containing mM sodium phosphate, mM NaCl, pH and (v/v) D O. The final concentration of DMSO ‐ d6 was (v/v) and was kept constant for all experiments. To examine the binding of both ( ‐ ) ‐ and (+) ‐ to either ‐ ‐ σ alone or in complex to the ER α peptide, WaterLOGSY spectra were recorded on solutions containing each of the enantiomers of at µM in the presence of µM ‐ ‐ σΔ C alone or together with µM ER α peptide , respectively. A H spectrum with water ‐ suppression was additionally recorded for each S44 sample.
The reference for the H chemical shift was relative to TMSP (Trimethylsilylpropanoic acid).
Spectra were collected, processed and analyzed with
Topspin (Bruker
Biospin,
Karlsruhe,
Germany).
III.
Protein purification and X-ray crystallography
Protein purification
Prior to purification, the cell pellets were thawed and resuspended in 10 mL/g pellet lysis buffer (25 mM Tris, pH = 8.0, 150 mM NaCl, 5% v/v glycerol, 10 mM imidazole, 4 mM BME and 1 mM PMSF). The cells were then lysed twice by homogenization using an EmulsiFlex-C3 homogenizer. The lysate was incubated with benzonase (Merck Millipore) for 15 minutes and then centrifuged at 20000 g for 15 minutes. The supernatant was applied in overnight circulation at 4 °C to a 5 mL HisTrap column pre-equilibrated with 20 column volumes (CV) lysis buffer. The column was then washed with 20 CV wash buffer (25 mM Tris, pH = 8.0, 300 mM NaCl, 5% v/v glycerol, 25 mM imidazole and 4 mM BME) and the protein eluted with 40 mL elution buffer (20 mM HEPES, pH 8.0, 100 mM NaCl, 5% v/v glycerol, 250 mM imidazole and 4 mM BME). The protein was then pipetted into a SpectrumLabs Spectra/Por 10000 Da MWCO dialysis bag and dialysed overnight at 4 °C against dialysis buffer (25 mM HEPES, pH = 8.0, 100 mM NaCl, 4 mM BME, 2 mM MgCl2). For the σΔ C protein, 1:500 mg/mg TEV protease was added to the dialysis bag. The full length proteins were then concentrated to ~50 mg/mL using 10000 Da MWCO Amicon spinfilters, aliquoted, flash frozen in liquid nitrogen and stored at -80 °C until further usage. The σΔ C protein was instead applied to a 5 mL HisTrap column pre-equilibrated with 20 CV dialysis buffer. The flowtrough was captured and concentrated to ~50 mg/mL using 10000 Da MWCO Amicon spinfilters. The concentrated σΔ C protein was then applied to a HiLoad superdex 75 16/60 SEC column using an Äkta FPLC apparatus. The fractions containing protein were then pooled, concentrated to ~50 mg/mL protein, flash frozen in liquid nitrogen and stored at -80 °C until further use.
Crystallography – data collection and analysis
X-ray diffraction data for the 14-3-3 σΔ C/ER α /( R )- complex was collected at 100 K at the p11 beamline of the PETRA-III synchrotron of the DESY facility in Hamburg, Germany using a Pilatus 6M-F detector. X-ray diffraction data for the 14-3-3 σΔ C/ER α /( R )- complex was collected at 100 K on a Rigaku Micromax-003 sealed tube X-ray source and a Dectris Pilatus 200K detector. The data was indexed, integrated, scaled and merged using xia2 DIALS. Phasing was done by molecular replacement using Phaser and 4JC3 as a starting model and was followed by iterative rounds of refinement and manual model building using Phenix.Refine and Coot respectively. Model validation was performed using MolProbity. Figures were created using PyMol. ‐ ‐ σΔ C/ER α /( R ) ‐ ‐ ‐ σΔ C/ER α /( R ) ‐ PDB code
S45 a Number in parentheses is for the highest resolution shell b As reported by xia2 DIALS. c CC = Pearson's intradataset correlation coefficient, as described by Karplus and Diederichs. Data collection
Resolution (Å) a (1.85 – (2.50 – Space group
C222
C222
Cell parameters (Å) b a = b = c = α = β = γ = ° a = b = c = α = β = γ = ° R mergea,b (1.17) (0.64) Average I/ σ (I)a,b (1.0) (2.03) CC (%) a,b,c (80.3) (86.3) Completeness (%) a,b (99.1) (100)
Redundancy a,b (11.6) (6.6)
Refinement
Number of protein/solvent/ligand atoms R work /R free (%) Unique reflections used in refinement R.m.s. deviations from ideal values bond lengths (Å) / bond angles (°) Average protein/solvent/ligand B ‐ factor (Å ) Ramachandran favored (%)
Ramachandran allowed (%)
Ramachandran outliers (%) S46
IV.
Chemistry section
IVa) General information
All solvents and reagents were obtained from commercially available sources and used without further purification. The microwave syntheses were performed in a Biotage Initiator with an external surface IR probe. Flash column chromatography was carried out on prepacked silica gel columns supplied by Biotage and using Biotage automated flash systems with UV detection. UHPLC-MS experiments were performed using a Waters Acquity UHPLC system combined with a SQD mass spectrometer. The UHPLC system was equipped with both a BEH C18 column 1.7 μ m 2.1×50 mm in combination with a 46 mM (NH ) CO /NH buffer at pH 10 and a HSS C18 column 1.8 μ m 2.1×50 mm in combination with 10 mM formic acid or 1 mM ammonium formate buffer at pH 3. The mass spectrometer used ESI+/- as ion source. UPLC was also carried out using a Waters UPLC fitted with Waters QDa mass spectrometer (Column temp 40°C, UV = 190–400 nm, MS = ESI with pos/neg switching) equipped with a Waters Acquity BEH 1.7 μ m 2.1×100 mm in combination with either 0.1% formic acid in water, 0.05% TFA in water or 0.04% NH in water. The flow rate was 1 mL/min. Preparative HPLC was performed by Waters Fraction Lynx with ZQ MS detector on either a Waters Xbridge C18 OBD 5 μ m column (19×150 mm, flow rate 30 mL/min or 30×150 mm, flow rate 60 mL/min) using a gradient of 5–95% MeCN with 0.2% NH at pH 10 or a Waters SunFire C18 OBD 5 μ m column (19×150 mm, flow rate 30 mL/min or 30×150 mm, flow rate 60 mL/min) using a gradient of 5–95% MeCN with 0.1 M formic acid or on a Gilson Preparative HPLC with a UV/VIS detector 155 on a Kromasil C8 10 μ m column (20 × 250 mm, flow rate 19 mL/min, or 50 × 250 mm, flow rate 100 mL/min) using a varying gradient of ACN with 0.1% formic acid (FA) in water or 0.2% trifluoroacetic acid (TFA) in water or 0.2% acetic acid (AcOH) in water or 0.2% ammonia (NH ) in water. Molecular mass (HR-ESI-MS) was recorded using a Shimadzu LCMS-2020 instrument (ESI+). Purity of all test compounds was determined by LCMS. All screening compounds had a purity >95%.
General H NMR spectra were recorded on a Bruker Avance II, III, AV300, AV400 or AVIII500 spectrometer at a proton frequency of 400, 500 or 600 MHz at 25 °C or at a temperature and frequency stated in each experiment. C NMR spectra were recorded at 101 MHz or 126 MHz. The chemical shifts ( δ ) are reported in parts per million (ppm) with residual solvent signal used as a reference (CD Cl at 5.32 ppm for H NMR and 53.84 ppm for C NMR, (CD ) SO at 2.50 ppm for H NMR and 39.52 ppm for C NMR, CDCl at 7.26 ppm for H NMR and 77.16 ppm for C NMR). Coupling constants (J) are reported as Hz. NMR abbreviations are used as follows: br = broad, s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet. Protons on heteroatoms such as COOH protons are only reported when detected in NMR and can therefore be missing. The PMA2 peptide (C-terminal, 52 mer, amino acids 905-956) expression and purification protocol is reported at ref . IVb) Synthetic procedures and compound characterization
S47
Ethyl 2,4-dioxo-4-phenyl-butanoate.
In a 250 mL RBF, acetophenone (2.92 mL, 24.97 mmol) was dissolved in THF (100 mL) and the resulting solution cooled down to 0 °C. Sodium ethanolate (13.98 mL, 37.45 mmol) was then added dropwise and the reaction allowed to stir for 15min at 0 °C. Diethyl oxalate (3.73 mL, 27.47 mmol) was finally added dropwise, the cooling bath removed, and the reaction allowed to stir overnight at rt. The reaction was quenched with 1M HCl (50 mL). The resulting suspension was poured into a separatory funnel and the crude product was extracted with DCM (3x). The combined organic layers were dried using a phase separator and solvent was removed under reduced pressure. The crude product was purified by preparative HPLC (40-80% ACN in H O/ACN/AcOH 95/5/0.2 buffer over 20 minutes). Collected fractions were freeze-dried, to give ethyl 2,4-dioxo-4-phenyl-butanoate (4.25 g, 77%) as a yellow solid. H NMR (400 MHz, CDCl ) δ C NMR (101 MHz, CDCl ) δ In a 20 mL vial, to a solution of ethyl 2,4-dioxo-4-phenylbutanoate (603 mg, 2.74 mmol) in AcOH (7 mL), 4-nitrobenzaldehyde (422 mg, 2.74 mmol) and 5-amino-2-hydroxybenzoic acid (441 mg, 2.74 mmol) were added. The vial was capped and heated at 120 °C for 180 min in a single node microwave reactor. The pressure monitored was 1 bar. The mixture was diluted with diethyl ether and filtered. The residue was washed with diethyl ether, dried under reduced pressure, to give compound (543 mg, 43%) as a pale yellow solid. H NMR (400 MHz, DMSO) δ C NMR (101 MHz, DMSO) δ m/z [M + H] + calcd for C H N O : 461.0985, found: 461.0974. Matches with a previously reported characterization. S48
In 20 mL vial, compound (98 mg, 0.21 mmol) was suspended in AcOH (6 mL). Hydrazine (35% in water) (0.025 mL, 0.28 mmol) was added, while stirring at room temperature. The vial was capped and heated at 120 °C for 120 min in a single node microwave reactor. The pressure monitored was 1 bar. After solvent removal, the crude mixture was dissolved in EtOAc and washed with HCl aq 1M (3x). The organic layer was dried using a phase separator and concentrated under reduced pressure. The residue was purified by preparative HPLC (15-35% acetonitrile in H O/ACN/NH (82 mg, 84.0%) as a off-white solid. H NMR (500
MHz,
DMSO) δ (d, J = Hz, (d, J = Hz, – (m, (d, J = Hz, (t, J = Hz, (t, J = Hz, (s, (d, J = Hz, C NMR (126
MHz,
DMSO) δ HRMS (ESI) m/z [M + H] + calcd for C H N O : found: Matches with a previously reported characterization. ( R )-2-hydroxy-5-[4-(4-nitrophenyl)-6-oxo-3-phenyl-1,4-dihydropyrrolo[3,4-c]pyrazol-5-yl]benzoic acid (( R )-5). In 5 mL vial, compound ( R )-2 (110 mg, 0.24 mmol) was suspended in AcOH (3 mL). Hydrazine (35% in water) (0.028 mL, 0.31 mmol) was added, while stirring at room temperature. The vial was capped and heated at 120 °C for 120 min in a single node microwave reactor. The pressure monitored was 1 bar. After solvent removal, the residue was purified by preparative HPLC (25-65% acetonitrile in H O/ACN/FA 95/5/0.2 buffer over 20 minutes), to give compound ( R )-5 (51 mg, 46.8 %) as a off-white solid. H NMR (500
MHz,
DMSO) δ – (m, (d, J = Hz, – (m, (d, J = Hz, (br, – (br, (s, (d, J = Hz, C NMR (126
MHz,
S49
DMSO) δ HRMS (ESI) m/z [M + H] + calcd for C H N O : found: [ α ] D20 : +20.0 ( c MeOH).
S50 ‐ [3 ‐ [amino(phenyl)methylene] ‐ ‐ (4 ‐ nitrophenyl) ‐ ‐ dioxo ‐ pyrrolidin ‐ ‐ yl] ‐ ‐ hydroxy ‐ benzoic acid (6). In a mL vial, to a suspension of (100 mg, mmol) in AcOH (3 mL), ammonium hydroxide (0.328 mL, mmol) was added. The vial was capped and heated at °C for min in a single node microwave reactor. The pressure monitored was bar. After solvent removal, the residue was purified by preparative HPLC (10 ‐ acetonitrile in H O/ACN/FA buffer over minutes), to give compound (20 mg, %) as a yellow solid. H NMR (500
MHz,
DMSO) δ (d, J = Hz,
H36), (d, J = Hz,
H35), (d, J = Hz,
H10), (d, J = Hz,
H18 and
H20), (dd, J = Hz,
H6), – (m, H31), – (m, H29,
H30,
H32 and
H33), (d, J = Hz,
H17 and
H21), (d, J = Hz,
H7), (s,
H15). C NMR (151
MHz,
DMSO) δ HRMS (ESI) m/z [M + H] + calcd for C H N O : found: ‐ hydroxy ‐ ‐ [3 ‐ [[(4 ‐ methoxyphenyl)methylamino] ‐ phenyl ‐ methylene] ‐ ‐ (4 ‐ nitrophenyl) ‐ ‐ dioxo ‐ pyrrolidin ‐ ‐ yl]benzoic acid (7) and ‐ [3 ‐ benzoyl ‐ ‐ [(4 ‐ methoxyphenyl)methylamino] ‐ ‐ (4 ‐ nitrophenyl) ‐ ‐ oxo ‐ ‐ pyrrol ‐ ‐ yl] ‐ ‐ hydroxy ‐ benzoic acid (8) In a mL vial, to compound (536 mg, mmol) in AcOH (5 mL), (4 ‐ methoxyphenyl)methanamine (1.086 mL, mmol) was added. The vial was capped and heated at °C for min in a single node microwave reactor. The pressure monitored was bar. After solvent removal, the residue was purified by preparative HPLC (45 ‐ acetonitrile in H O/ACN/TFA buffer over minutes), to give compound (40 mg, and compound (174 mg, both as yellow solids. S51 : H NMR (500
MHz,
DMSO) δ (t, J = Hz, (d, J = Hz, (d, J = Hz, (br, (dd, J = Hz, (t, J = Hz, (d, J = Hz, (d, J = Hz, – (m, (br, (s, (dd, J = Hz, (dd, J = Hz, (s, C NMR (126
MHz,
DMSO) δ HRMS (ESI) m/z [M + H] + calcd for C H N O : found: : H NMR (600
MHz,
DMSO) δ (br, (d, J = Hz, (d, J = Hz, (dd, J = Hz, – (m, (t, J = Hz, (br, (d, J = Hz, (d, J = Hz, (d, J = Hz, (s, ‐ (br d, (s, C NMR (126
MHz,
DMSO) δ HRMS (ESI) m/z [M ‐ H] ‐ calcd for C H N O : found: ‐ [4 ‐ amino ‐ ‐ benzoyl ‐ ‐ (4 ‐ nitrophenyl) ‐ ‐ oxo ‐ ‐ pyrrol ‐ ‐ yl] ‐ ‐ hydroxy ‐ benzoic acid (9). In a mL vial, compound (50 mg, mmol) was dissolved in trifluoroacetic acid (1 mL, mmol). The vial was capped and heated at °C for min in a single node microwave reactor. The pressure monitored was bar. After solvent removal, the residue was purified by automated flash chromatography on a Biotage ® KP ‐ SIL g column (40 ‐ of heptane in EtOAc + FA over to give compound (21 mg, as a pale yellow solid. H NMR (500
MHz,
DMSO) δ (br, H32), (d, J = Hz,
H26), (d, J = Hz,
H18 and
H20), (br,
H34), (dd, J = Hz,
H22), (d, J = Hz,
H12 and
H16), (t, J = Hz,
H14), (t, J = Hz,
H13 and
H15), (d, J = Hz,
H17 and
H21), (d, J = Hz,
H23), (s,
H2). C NMR (126
MHz,
DMSO) δ HRMS (ESI) m/z [M + H] + calcd for C H N O : found: S52 ( R ) ‐ ‐ hydroxy ‐ ‐ [3 ‐ [[(4 ‐ methoxyphenyl)methylamino] ‐ phenyl ‐ methylene] ‐ ‐ (4 ‐ nitrophenyl) ‐ ‐ dioxo ‐ pyrrolidin ‐ ‐ yl]benzoic acid (( R ) ‐ and ( R ) ‐ ‐ [3 ‐ benzoyl ‐ ‐ [(4 ‐ methoxyphenyl)methylamino] ‐ ‐ (4 ‐ nitrophenyl) ‐ ‐ oxo ‐ ‐ pyrrol ‐ ‐ yl] ‐ ‐ hydroxy ‐ benzoic acid (( R ) ‐ In a mL vial, to compound ( R ) ‐ (78 mg, mmol) in AcOH (2 mL), (4 ‐ methoxyphenyl)methanamine (0.158 mL, mmol) was added. The vial was capped and heated at °C for min in a single node microwave reactor. The pressure monitored was bar. After solvent removal, the residue was purified by preparative HPLC (45 ‐ acetonitrile in H O/ACN/TFA buffer over minutes). Collected fractions were freeze ‐ dried, to give compounds ( R ) ‐ (10 mg, and ( R ) ‐ (16 mg, both as yellow solids. ( R ) ‐ : m/z (ESI ‐ MS): [M+H + ] + calculated mass = observed = ( R ) ‐ : m/z (ESI ‐ MS): [M ‐ H + ] ‐ calculated mass = observed = Compounds were not further characterized but used directly for the next step. ( R ) ‐ ‐ [3 ‐ [amino(phenyl)methylene] ‐ ‐ (4 ‐ nitrophenyl) ‐ ‐ dioxo ‐ pyrrolidin ‐ ‐ yl] ‐ ‐ hydroxy ‐ benzoic acid (( R ) ‐ In a mL vial, compound ( R ) ‐ (10 mg, mmol) was dissolved in trifluoroacetic acid (1 mL, mmol). The vial was capped and heated at °C for min in a single node microwave reactor. The pressure monitored was bar. After solvent removal, the residue was purified by preparative HPLC (25 ‐ acetonitrile in H O/ACN/TFA buffer over minutes). Collected fractions were freeze ‐ dried, to give compound ( R ) ‐ (3.2 mg, ee) as a pale yellow solid. H NMR (600
MHz,
DMSO) δ (br, (br, (br, (d, J = Hz, (d, J = Hz, (dd, J = Hz, – (m, – (m, (br, J = Hz,
S53 (d, J = Hz, (s, C NMR (151
MHz,
DMSO) δ HRMS (ESI) m/z [M + H] + calcd for C H N O : found: ( R ) ‐ ‐ [4 ‐ amino ‐ ‐ benzoyl ‐ ‐ (4 ‐ nitrophenyl) ‐ ‐ oxo ‐ ‐ pyrrol ‐ ‐ yl] ‐ ‐ hydroxy ‐ benzoic acid (( R ) ‐ In a mL vial, compound ( R ) ‐ (16 mg, mmol) was dissolved in trifluoroacetic acid (1 mL, mmol). The vial was capped and heated at °C for min in a single node microwave reactor. The pressure monitored was bar. After solvent removal, the residue was purified by preparative HPLC (35 ‐ acetonitrile in H O/ACN/TFA buffer over minutes). Collected fractions were freeze ‐ dried, to give compound ( R ) ‐ (6.4 mg, ee) as a pale yellow solid. H NMR (600
MHz,
DMSO) δ (br, (d, J = Hz, (d, J = Hz, (br, (dd, J = Hz, – (m, (t, J = Hz, (t, J = Hz, (d, J = Hz, (d, J = Hz, (s, C NMR (151
MHz,
DMSO) δ HRMS (ESI) m/z [M + H] + calcd for C H N O : found: S54
IVc)
Chiral separation,
VCD analysis and racemisation studies of Chiral separation ( R and S )-5-[3-benzoyl-4-hydroxy-2-(4-nitrophenyl)-5-oxo-2H-pyrrol-1-yl]-2-hydroxy-benzoic acid (-)-2 and (+)-2). The enantiomers of compound (1.2 g, 2.61 mmol) were separated by chiral column chromatography on a Chiralpak IC (250x20 mm, 5 μ m) column. 50 mg (50 mg/mL in EtOH/TEA 10:0.1) were injected and eluted with 100% EtOH/TEA (100:0.1), 120 bar at 25 °C, a flow rate of 12 mL/min and detected at 270 nm. The first eluted compound was collected and evaporated to give compound (-)- (759 mg, 97.3% ee) as a yellow solid. [ α ] D20 : -96.8 ( c H NMR (500 MHz, DMSO) δ J = 7.7 Hz, 2H), 6.92 (d, J = 8.9 Hz, 1H), 6.47 (s, 1H). C NMR (126 MHz, DMSO) δ m/z [M + H] + calcd for C H N O : 461.0985, found: 461.0966. The second eluted compound was collected and evaporated to give compound (+)- (644 mg, 99.6% ee) as a yellow solid. [ α ] D20 : +103.6 ( c δ δ m/z [M + H] + calcd for C H N O : 461.0985, found: 461.0987. S55 S56 S57 S58 S59 S60 S61 Vibrational Circular Dichroism (VCD)
Summary
VCD analysis was performed on (+)- only. Although not conclusive, the data collected suggest that (+)-xx1 is likely to be the (S) enantiomer. All spectra, experimental and simulated, are shown in Figure 1. Experimental (+)- (5.0 mg) was dissolved in 110 μ l CDCl . Approximately 90 μ l of the solution was transferred to a 0.100 mm BaF cell and VCD spectra acquired for 12 hours in a Biotools ChiralIR2X instrument. The resolution was 4 cm -1 . A VCD spectrum was collected on a CDCl blank in the same cell to act as a baseline reference and subtracted from the experimental spectrum. Computational Spectral Simulation
A Monte Carlo molecular mechanics search for low energy geometries was conducted for the S enantiomer. MacroModel within the Maestro graphical interface (Schrödinger Inc.) was used to generate 123 starting coordinates for conformers. All conformers within 5 kcal/mole of the lowest energy conformer were used as starting point for density functional theory (DFT) minimizations within Gaussian09. Optimized structures, harmonic vibrational frequencies/intensities, VCD rotational strengths, and free energies at STP (including zero-point energies) were determined at B3LYP/6-31G* level of theory. Three conformations were found that contributed over 10% to the Boltzmann distribution. An in-house built program was used to fit Lorentzian line shapes (12 cm -1 line width) to the computed spectrum of a Boltzmann distributed average thereby allowing direct comparisons between simulated and experimental spectra. S62 Supplementary Figure 22.
Comparison of calculated and experimental spectra: the calculated spectrum for the (S) enantiomer is shown in pink and the (R) enantiomer (by inversion) in green. There is a reasonable, but not high certainty match between the experimental spectrum of (+)- and the calculated spectrum for the (S) enantiomer, particularly in the region 1200-1300 cm -1 . S63 Racemisation studies
Sample was dissolved in EtOH, then triethylamine (TEA, 10 eq) was added and the resulting solution was stirred overnight at room temperature. The reaction mixture was then transferred into a vial and analyzed directly. S64 S65 S66 S67
IVd) FITC-3 peptide synthesis
General information
Fmoc-amino acids were purchased from Chem-Impex International, Inc., with the following side-chain protection: Fmoc-Ala-OH, Fmoc-Glu(OtBu)-OH, Fmoc-Gly-OH, Fmoc-Phe-OH, Fmoc-Pro-OH, Fmoc-Thr(PO(Obzl)OH)-OH, Fmoc-Val-OH. L -Amino acids were used in every case. Fluorescein 5-isothiocyanate (5-FITC) was purchased from Sigma-Aldrich. Diisopropylethylamine (DIPEA) was purchased from Sigma-Aldrich, 1-[bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxid hexafluorophosphate (HATU) was purchased from Chem-Impex International, Inc. 2-chlorotrityl chloride resin (200-400 mesh) from Chem-Impex International, Inc. Peptide synthesis was monitored by reversed-phase (RP) ultra-performance liquid chromatography tandem mass spectrometer (UPLC-MS). Analytical RP-UPLC-MS was performed on a Waters Acquity UPLC system (PDA, sample manager, sample organiser, column oven modules) and Waters SQD2 mass spectrometer using the following column: Waters Acquity CSH C18 column, 130 Å, 1.7 µm, 50 x 2.1 mm at a flow rate of 0.5 ml/min at 45 °C. A linear gradient of mobile phase: A = H O + 10 mM formic acid (FA), 1 mM ammonia and 0.03% trifluoroacetic acid (TFA) and B = ACN/H O 95/5 (vol/vol-%) + 10 mM FA, 1 mM ammonia and 0.03% TFA was used with detection at 220 nm.
Resin loading
In a peptide reactor 2-chlorotrityl chloride resin 0.8 mmol/g (0.2 mmol, 250 mg) was swollen in CH Cl for 10 min, then the solvent was drained. Fmoc- L -Val-OH (1 eq) was dissolved in CH Cl and DIPEA (3 eq) was added. The clear solution was added to the resin which was agitated for 10 min. Additional DIPEA (7 eq) was added and the resin was agitated for further 45 min. The remaining trityl groups were capped adding MeOH (0.8 μ L/mg of resin), the resin was agitated for 10 min. The mixture was drained and the resin beads washed with DMF (3 x) and CH Cl (3 x). Fmoc cleavage
The N-terminal Fmoc protecting group was removed with a 20% solution of piperidine in DMF (2 x 5 min). The mixture was drained and the resin beads washed with DMF (3 x) and CH Cl (3 x). Peptide elongation
Fmoc- L -AA-OH (4 eq) and HATU (4 eq) were dissolved in DMF, then DIPEA (6 eq) was added. After a pre-activation period of 2 min, the mixture was added to the resin, which was agitated for 45-60 min. The mixture was drained and the resin beads washed with DMF (3 x) and CH Cl (3 x). Successful coupling was indicated by ninhydrin test. Coupling of the spacer Fmoc-6-aminoexanoic acid (Fmoc-6-Ahx-OH, 4 eq) was performed using the same procedure as above. The peptide was finally reacted with Fluorescein-5-isothiocyanate (5-FITC, 4 eq) and DIPEA (6 eq) for 2h in the dark to yield fluorescent-immobilized peptidyl-resin. Cleavage from resin
S68 The resin-bound peptide was cleaved from resin using a solution of TFA/TIS/H O/EDT (94:1:2.5:2.5). After shaking for 60 min, the TFA solution was filtered and the reaction mixture was poured into cold diethyl ether. Upon precipitation, the suspension was centrifuged, the precipitate dissolved in ACN/H O and lyophilized.
Purification
The compound was purified by preparative HPLC on a Kromasil C8 column (10 μ m 250x20 ID mm) using a gradient of 43-63% ACN in H O/ACN/FA 95/5/0.2 buffer, over 20 minutes with a flow of 19 mL/min. The compounds were detected by UV at 220nm. Collected fractions were lyophilized, to yield the purified peptide FITC- (26 mg, 9.47%) as a yellow solid, in ca. 95% purity as estimated by analytical UPLC. R t = 7.03 min (3-60% B over 10 min, 0.5 mL/min). m/z (ESI-MS): [M+2H + ] calculated mass = 687.2, observed = 697.7 S69 Final sequence: 5-FITC-Ahx- AEGFPApTV-OH (FITC- ) FAM- sequence Sequence: 5-FAM-GSLSARKLpSLQER-OH S70
IVe) pK a determination Protocol
Summary pK a s are determined using the SiriusT3 instrument from
Sirius
Analytical by performing an acid/base titration. The technique uses an in situ UV probe to measure the UV absorbance profile of the compound at each pH point during the titration. Measured pKa values are reported as mean ± SEM.
Experimental procedure
The sample pK a s are investigated using the fast UV ‐ metric method. This involves measuring the UV absorbance profile at each pH point during an acid/base titration using an in situ UV probe in the titration cell of a SiriusT3 instrument.
Each sample is titrated in a triple titration over a nominal pH range of to in approximately % methanol. Sample concentrations are typically in the range ‐ μ M. All titrations are carried out at °C. The pK a (s) are determined by monitoring the change in UV absorbance with pH as the compound undergoes ionisation. This information is used to produce a matrix of pH vs. Wavelength vs.
Absorbance data. A mathematical technique called Target
Factor
Analysis is applied to the matrix to produce molar absorbance profiles for the different light absorbing species present in solution and also a “Distribution of Species” plot showing how the proportion of each species varies with pH. Sample pK a s are extrapolated to aqueous conditions using the Yasuda ‐ Shedlovsky method.
The method requires that the sample compound possesses a chromophore and that changes in ionisation influence the absorbance spectrum of the compound. This procedure will measure pK a values in the range ‐ S71
IVf) Spectroscopic data of products
S72 S73 S74 S75 S76 S77 S78 S79 S80 S81 S82 S83 S84 S85 S86 S87 S88 S89 S90 S91 S92 S93 S94 S95 S96 S97 S98 S99 S100 S101 S102 S103 S104 Enantiomeric purity for ( R )-5-[3-[amino(phenyl)methylene]-2-(4-nitrophenyl)-4,5-dioxo-pyrrolidin-1-yl]-2-hydroxy-benzoic acid (( R )-6). S105 S106 S107
Enantiomeric purity for ( R )-5-[4-amino-3-benzoyl-2-(4-nitrophenyl)-5-oxo-2H-pyrrol-1-yl]-2-hydroxy-benzoic acid (( R )-9). S108 S109 S110 V. Computational details
All calculations were performed within the Schrödinger Small-Molecule Drug Discovery Suite 2019-2. Initial geometries were derived from MCMM conformational searches in Macromodel version 12.4 using the OPLS3e force field in combination with the GB/SA continuum solvation model for water. Density functional theory calculations were performed using Jaguar version 10.4. Structures were optimized using the B3LYP-D3 a posteriori -corrected hybrid functional with the 6-31G**+ basis set and the PBF solvation model for water. Normal-mode analysis were used to estimate the Gibbs free energies. Final energies were calculated using B3LYP-D3/6-311G**+ with the PBF solvation model for water and with M06-2X-D3/6-31**+ together with SM6. M06-2X-D3/6-311**+ together with PBF (water) energies were calculated to compare the two functionals. For comparative purposes, some combinations were also included. For each compound, six different conformations were examined in detail starting from the conformation as found for ( R )- in the crystal structure. The torsion changed next was then the bond to the exocyclic carbonyl in combination with the attached phenyl group resulting in two different anti-conformations. For these three variants the salicylate was also rotated 180°. Structures of the optimized conformations are illustrated in Supplementary Figure 23. S111 ( R )- Syn Anti -conformation 1
Anti -conformation 2 Flipped Salicylic acid ( R )- -sF ( R )- -aF’ ( R )- -aF Conformation of salicylic acid as found in X-ray structure ( R )- -sX ( R )- -aX’ ( R )- -aX Supplementary Figure 23 . Optimized geometries of the 6 conformations of ( R )- studied in detail illustrating the naming scheme also employed for compounds ( R )- and ( R )- . S112 Supplementary Table 2 . Conformational energies in Hartrees for ligands ( R )- , ( R )- , ( R )- calculated using different functionals, basis sets and solvation models. B3LYP-D3 M06-2X_D3 6-31+G** 6-311+G** 6-31+G** 6-311+G** Name of conformation Compound Exocyclic carbonyl Salicylic acid Gas Phase Energy PBF Solution Phase Energy Total Free Energy (au) 298.15K 1.0atm Gas Phase Energy PBF Solution Phase Energy Gas Phase Energy SM6 Solution Phase Energy Gas Phase Energy PBF Solution Phase Energy ( R )- -sF ( R )- syn Flipped -1634.453750 -1634.738383 -1634.452014 -1634.816852 -1635.102297 -1633.751313 -1634.022958 -1634.139214 -1634.427444 ( R )-2-sX ( R )- syn X-ray -1634.447095 -1634.736970 -1634.451019 -1634.809929 -1635.100626 -1633.744478 -1634.023912 -1634.131973 -1634.425593 ( R )- -aF’ ( R )- anti Flipped -1634.462425 -1634.735341 -1634.449469 -1634.825293 -1635.099054 -1633.759245 -1634.025238 -1634.146631 -1634.423957 ( R )- -aX’ ( R )- anti X-ray -1634.457409 -1634.734818 -1634.448884 -1634.820089 -1635.098392 -1633.754053 -1634.026338 -1634.141105 -1634.422973 ( R )- -aF ( R )- anti Flipped -1634.461160 -1634.735939 -1634.449564 -1634.823698 -1635.099807 -1633.758193 -1634.026363 -1634.145333 -1634.424888 ( R )-2-aX ( R )- anti X-ray -1634.457272 -1634.736693 -1634.450466 -1634.819795 -1635.100285 -1633.754083 -1634.027758 -1634.141087 -1634.425030 ( R )- -sF ( R )- syn Flipped -1615.185874 -1615.323308 -1615.012390 -1615.540117 -1615.678202 -1614.483202 -1614.618400 -1614.861152 -1615.001336 ( R )-6-sX ( R )- syn X-ray -1615.177281 -1615.322817 -1615.012520 -1615.531434 -1615.677671 -1614.474501 -1614.619681 -1614.852279 -1615.000525 ( R )- -aF’ ( R )- anti Flipped -1615.167250 -1615.312714 -1615.002754 -1615.521362 -1615.667778 -1614.463804 -1614.604592 -1614.841471 -1614.990533 ( R )- -aX’ ( R )- anti X-ray -1615.164794 -1615.312751 -1615.003840 -1615.518988 -1615.667632 -1614.461861 -1614.608765 -1614.839500 -1614.990146 ( R )- -aF ( R )- anti Flipped -1615.168031 -1615.312706 -1615.002505 -1615.522149 -1615.667602 -1614.464640 -1614.604511 -1614.842350 -1614.990532 ( R )-6-aX ( R )- anti X-ray -1615.161596 -1615.313195 -1615.003560 -1615.515683 -1615.667678 -1614.458256 -1614.607089 -1614.835745 -1614.990340 ( R )- -sF ( R )- syn Flipped -1615.199092 -1615.323821 -1615.012974 -1615.553356 -1615.678701 -1614.496256 -1614.617066 -1614.874292 -1615.002241 ( R )-9-sX ( R )- syn X-ray -1615.193073 -1615.323567 -1615.013153 -1615.547163 -1615.678433 -1614.490318 -1614.617754 -1614.868032 -1615.001661 (
R)- -aF’ ( R )- anti Flipped -1615.190414 -1615.318138 -1615.007994 -1615.544177 -1615.672666 -1614.487504 -1614.611874 -1614.864807 -1614.995877 ( R )- -aX’ ( R )- anti X-ray -1615.183424 -1615.315676 -1615.005847 -1615.537414 -1615.670454 -1614.480460 -1614.610151 -1614.857859 -1614.993418 ( R )- -aF ( R )- anti Flipped -1615.189880 -1615.317617 -1615.007372 -1615.543811 -1615.672397 -1614.487248 -1614.612164 -1614.864834 -1614.995873 ( R )-9-aX ( R )-9 anti X-ray -1615.187412 -1615.318914 -1615.008795 -1615.541159 -1615.673487 -1614.484769 -1614.613301 -1614.862003 -1614.996798 a) Entries in bold are discussed in the main manuscript and corresponds to the conformation of the salicylate and benzoyl as found in the crystal structure.
S113
Supplementary Table 3 . Relative conformational energies in kcal mol -1 for ligands ( R )- , ( R )- , ( R )- calculated using different functionals, basis sets and solvation models. B3LYP-D3 M06-2X_D3 6-31+G** 6-311+G** 6-31+G** 6-311+G** Gas Phase Energy Solution Phase Energy Total Free Energy (au) 298.15K 1.0atm Gas Phase Energy Solution Phase Energy Total Free Energy (au) 298.15K 1.0atm Gas Phase Energy SM6 Solution Phase Energy Total Free Energy (au) 298.15K 1.0atm Gas Phase Energy PBF Solution Phase Energy Total Free Energy (au) 298.15K 1.0atm ( R )- syn Flipped 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 ( R )- syn X-ray 4.2 0.9 0.6 4.3 1.0 -0.9 R )- anti Flipped -5.4 1.9 1.6 -5.3 2.0 1.7 -5.0 -1.4 -1.7 -4.7 2.2 2.2 ( R )- anti X-ray -2.3 2.2 2.0 -2.0 2.5 2.2 -1.7 -2.1 -2.4 -1.2 2.8 2.8 ( R )- anti Flipped -4.6 1.5 1.5 -4.3 1.6 1.6 -4.3 -2.1 -2.1 -3.8 1.6 1.6 ( R )- anti X-ray -2.2 1.1 1.0 -1.8 1.3 -1.7 -3.0 -3.1 -1.2 1.5 1.5 ( R )- syn Flipped -5.4 -0.3 0.1 -5.4 -0.3 0.1 -5.5 0.8 1.2 -5.6 -0.5 -0.5 ( R )- syn X-ray 0.0 0.0 0.0 0.0 0.0 R )- anti Flipped 6.3 6.3 6.1 6.3 6.2 6.0 6.7 9.5 9.3 6.8 6.3 6.3 ( R )- anti X-ray 7.8 6.3 5.4 7.8 6.3 5.4 7.9 6.8 6.0 8.0 6.5 6.5 ( R )- anti Flipped 5.8 6.3 6.3 5.8 6.3 6.3 6.2 9.5 9.5 6.2 6.3 6.3 ( R )- anti X-ray 9.8 6.0 5.6 9.9 6.3 R )- syn Flipped -3.8 -0.2 0.1 -3.9 -0.2 0.1 -3.7 0.4 0.7 -3.9 -0.4 -0.4 ( R )- syn X-ray 0.0 0.0 0.0 0.0 0.0 R )- anti Flipped 1.7 3.4 3.2 1.9 3.6 3.4 1.8 3.7 3.5 2.0 3.6 3.6 ( R )- anti X-ray 6.1 5.0 4.6 6.1 5.0 4.6 6.2 4.8 4.4 6.4 5.2 5.2 ( R )- anti Flipped 2.0 3.7 3.6 2.1 3.8 3.7 1.9 3.5 3.4 2.0 3.6 3.6 ( R )- anti X-ray 3.6 2.9 2.7 3.8 3.1 a) Values in bold are included in the main manuscript and corresponds to the conformation of the salicylate and benzoyl moities as found in the crystal structure. S114
Coordinates in Angstrom of the optimized structures optimized by B3LYP-D3/6-31G**+ in combination with the PBF model for water.
48 (R)-2-sF C 0.17700 -63.60880 -26.05090 C 0.73540 -62.43230 -25.30240 C 1.10920 -62.57440 -23.95780 C 1.59890 -61.49520 -23.22580 C 2.93590 -61.10800 -21.16950 C 2.76240 -61.44840 -19.68970 C 1.58870 -62.20850 -19.57250 C 0.99370 -62.66890 -18.35410 C -0.23380 -63.54020 -18.44890 C -0.20500 -64.76910 -19.12460 C -1.36040 -65.55120 -19.21610 C -2.55800 -65.09950 -18.65160 C -2.59020 -63.87590 -17.97160 C -1.43010 -63.10680 -17.85730 C 0.96880 -62.40660 -20.94840 C -0.46160 -61.89730 -21.09610 C -1.46610 -62.74140 -21.58380 C -2.77570 -62.28590 -21.72140 C -3.06110 -60.96370 -21.37540 C -2.07370 -60.09240 -20.89960 C -0.77610 -60.57200 -20.76110 C 1.73210 -60.24090 -23.84840 C 1.38880 -60.08570 -25.18610 C 0.88440 -61.16980 -25.92190 N 1.88110 -61.64830 -21.83710 N -4.43410 -60.47430 -21.51430 O -0.16740 -63.41400 -27.27410 O 0.06650 -64.71940 -25.46480 O 3.87220 -60.45830 -21.65830 O 3.60850 -61.05570 -18.83730 O 1.41470 -62.35030 -17.21110 O -5.30310 -61.25950 -21.91240 O -4.67080 -59.29470 -21.22700 O 0.54640 -60.96060 -27.22240 H 0.99340 -63.55070 -23.50020 H 0.72050 -65.11810 -19.57540 H -1.32810 -66.50530 -19.73590 H -3.46170 -65.69720 -18.73980 H -3.51930 -63.52070 -17.53350 H -1.45400 -62.15510 -17.33310 H 0.99810 -63.45990 -21.25240 H -1.23340 -63.76830 -21.84830 H -3.55740 -62.93960 -22.09180 H -2.32220 -59.06860 -20.64390 H 0.00000 -59.91130 -20.38440 H 2.09280 -59.39020 -23.28050 H 1.48550 -59.12020 -25.67470 H 0.20640 -61.84080 -27.56450 48 (R)-2-sX C 0.71200 -58.89140 -25.77060 C 1.00660 -60.23620 -25.16740 C 1.35780 -60.33460 -23.81360 C 1.62500 -61.57100 -23.23400 C 2.93690 -61.10710 -21.16880 C 2.76100 -61.45020 -19.68160 C 1.58980 -62.21890 -19.57340 C 0.98920 -62.70840 -18.36950 C -0.25320 -63.55520 -18.51280 C -0.23370 -64.76640 -19.22160 C -1.40570 -65.51260 -19.37710 C -2.61050 -65.04370 -18.84310 C -2.63370 -63.84060 -18.12730 C -1.45800 -63.10670 -17.95030 C 0.95910 -62.38710 -20.94700 C -0.44980 -61.81590 -21.08830 C -1.46640 -62.58380 -21.66810 C -2.75860 -62.07810 -21.79710 C -3.01230 -60.78160 -21.34400 C -2.01070 -59.98390 -20.77660 C -0.73080 -60.51200 -20.65290 C 1.53920 -62.74080 -24.00600 C 1.18030 -62.66690 -25.34720 C 0.90980 -61.42210 -25.93600 N 1.89450 -61.66570 -21.83500 N -4.36810 -60.24120 -21.45970 O 0.81220 -57.85760 -25.05660 O 0.36630 -58.86990 -27.00830 O 3.86660 -60.45100 -21.66150 O 3.60150 -61.05780 -18.82430 O 1.40920 -62.44150 -17.21310 O -5.24770 -60.95690 -21.95410 O -4.58050 -59.09170 -21.05610 O 0.55520 -61.40120 -27.24770 H 1.40250 -59.43170 -23.21390 H 0.69810 -65.12730 -19.64950 H -1.38060 -66.45190 -19.92380 H -3.52650 -65.61230 -18.98280 H -3.56850 -63.47260 -17.71240 H -1.47640 -62.16920 -17.40100 H 0.93880 -63.43540 -21.26470 H -1.25580 -63.59280 -22.00890 H -3.55160 -62.67350 -22.23560 H -2.23630 -58.97920 -20.43780 H 0.05390 -59.90980 -20.20380 H 1.74760 -63.70690 -23.55370 H 1.10290 -63.56370 -25.95570 H 0.39590 -60.43610 -27.47630 48 (R)-2-aF' C 0.25050 -63.61490 -26.08730 C 0.77960 -62.43780 -25.32030 C 1.14180 -62.58950 -23.97380 C 1.61710 -61.51200 -23.23150 C 2.92900 -61.10300 -21.16880 C 2.74380 -61.41170 -19.68060 C 1.59760 -62.22000 -19.57750 C 0.98130 -62.83090 -18.44280 C 1.60930 -62.74060 -17.08150 C 2.94280 -63.11630 -16.86700 C 3.47980 -63.10130 -15.57730 C 2.69610 -62.68260 -14.49510 C 1.36590 -62.29970 -14.70480 C 0.81930 -62.34490 -15.99080 C 0.95670 -62.37940 -20.94130 C -0.43100 -61.75660 -21.06160 C -1.53240 -62.53050 -21.44670 C -2.79030 -61.95000 -21.60080 C -2.92590 -60.57850 -21.36730 C -1.84240 -59.78100 -20.98030 C -0.59880 -60.38290 -20.82820 C 1.74460 -60.25150 -23.84210 C 1.40940 -60.08620 -25.17990 C 0.91990 -61.16830 -25.92710 N 1.89660 -61.67420 -21.84390 N -4.23790 -59.95430 -21.53870 O -0.07600 -63.41260 -27.31360 O 0.14710 -64.73330 -25.51480 O 3.86690 -60.45310 -21.65440 O 3.55990 -60.94580 -18.83910 O -0.10280 -63.46730 -18.54930 O -5.19240 -60.66750 -21.87040 O -4.34060 -58.73630 -21.34560 O 0.58730 -60.94900 -27.22700 H 1.03340 -63.56980 -23.52210 H 3.55340 -63.42790 -17.70840 H 4.50930 -63.41100 -15.41790 H 3.11860 -62.65740 -13.49390 H 0.75460 -61.97230 -13.86800 H -0.21810 -62.06570 -16.15540 H 0.89900 -63.43740 -21.22650 H -1.40880 -63.59470 -21.62770 H -3.64640 -62.54280 -21.90230 H -1.97520 -58.71880 -20.80910 H 0.25270 -59.77680 -20.53180 H 2.09570 -59.40300 -23.26510 H 1.50220 -59.11550 -25.65940 H 0.25610 -61.82890 -27.57920 48 (R)-2-aX' C 1.24190 -59.03530 -25.97890 C 1.34130 -60.36570 -25.28750 C 1.59820 -60.42260 -23.90960 C 1.68980 -61.64530 -23.25060 C 2.92830 -61.10020 -21.16890 C 2.73960 -61.41250 -19.67700 C 1.60130 -62.23300 -19.57990 C 0.99920 -62.90030 -18.46750 C 1.68660 -62.97540 -17.13510 C 3.04790 -63.29900 -17.03070 C 3.64020 -63.45970 -15.77540 C 2.88250 -63.27500 -14.61230 C 1.52440 -62.94650 -14.70950 C 0.92620 -62.81280 -15.96550 C 0.94370 -62.35360 -20.93810 C -0.40690 -61.64790 -21.05020 C -1.53280 -62.33100 -21.52590 C -2.75650 -61.67720 -21.66680 C -2.83400 -60.32430 -21.32500 C -1.72390 -59.61480 -20.85090 C -0.51480 -60.28800 -20.71750 C 1.51320 -62.84030 -23.96740 C 1.24450 -62.80560 -25.33060 C 1.15930 -61.57560 -26.00030 N 1.91270 -61.69850 -21.84320 N -4.11250 -59.62620 -21.46890 O 0.99270 -59.05020 -27.23980 O 1.40220 -57.97730 -25.31350 O 3.86080 -60.44450 -21.65570 O 3.54790 -60.94330 -18.82970 O -0.11620 -63.47890 -18.58800 O -5.08620 -60.25920 -21.89520 O -4.16950 -58.43030 -21.15800 O 0.89950 -61.59170 -27.33420 H 1.71610 -59.49540 -23.35940 H 3.63690 -63.43860 -17.93120 H 4.69120 -63.72790 -15.70520 H 3.34630 -63.39100 -13.63590 H 0.93250 -62.80300 -13.80930 H -0.13200 -62.57920 -16.04770 H 0.82110 -63.40380 -21.22900 H -1.45560 -63.38310 -21.78660 H -3.63140 -62.20060 -22.03590 H -1.81070 -58.56450 -20.59690 H 0.35390 -59.74810 -20.35120 H 1.58860 -63.79640 -23.45670 H 1.10400 -63.72290 -25.89580 H 0.87910 -60.62910 -27.62240 48 (R)-2-aF C -0.13090 -63.34290 -26.02850 C 0.52670 -62.22690 -25.26630 C 0.96310 -62.44380 -23.95090 C 1.54950 -61.42280 -23.20640 C 2.93560 -61.11380 -21.16820 C 2.77770 -61.46900 -19.68940 C 1.58870 -62.20230 -19.57660 C 0.97720 -62.82610 -18.43470 C 1.20420 -62.27080 -17.06110 C 1.35310 -60.88950 -16.85410 C 1.46950 -60.37710 -15.56060 C 1.45470 -61.24350 -14.46090 C 1.30810 -62.62250 -14.65910 C 1.17110 -63.13210 -15.95250 C 0.97710 -62.41130 -20.94780 C -0.45480 -61.90240 -21.05390 C -1.50760 -62.78870 -21.31270
S115
C -2.81880 -62.32440 -21.41850 C -3.05670 -60.95590 -21.26590 C -2.02320 -60.04700 -21.00620 C -0.72540 -60.53320 -20.89810 C 1.72280 -60.15420 -23.78990 C 1.31660 -59.92630 -25.09900 C 0.70800 -60.94850 -25.84350 N 1.87430 -61.64190 -21.83730 N -4.42710 -60.45420 -21.38680 O -0.55070 -63.07070 -27.21330 O -0.24440 -64.47700 -25.49170 O 3.87110 -60.46050 -21.65580 O 3.66710 -61.12530 -18.85640 O 0.19380 -63.80260 -18.57190 O -5.33470 -61.26350 -21.61170 O -4.62280 -59.23910 -21.25940 O 0.30180 -60.66320 -27.11040 H 0.81420 -63.42880 -23.52200 H 1.35860 -60.21720 -17.70560 H 1.56900 -59.30520 -15.41070 H 1.55170 -60.84540 -13.45400 H 1.29500 -63.29600 -13.80630 H 1.03850 -64.19810 -16.11310 H 1.00390 -63.47100 -21.23320 H -1.30460 -63.84890 -21.43700 H -3.63770 -63.00520 -21.62350 H -2.23670 -58.98970 -20.89640 H 0.08830 -59.84090 -20.69960 H 2.16020 -59.34670 -23.21360 H 1.44230 -58.94800 -25.55530 H -0.11920 -61.50330 -27.46100 48 (R)-2-aX C 1.00690 -58.61620 -25.64900 C 1.09950 -60.00830 -25.09120 C 1.49330 -60.20190 -23.75970 C 1.57540 -61.48210 -23.21790 C 2.93710 -61.11350 -21.16880 C 2.78100 -61.47610 -19.68480 C 1.59160 -62.21530 -19.57680 C 0.99530 -62.89070 -18.45800 C 1.29680 -62.45400 -17.05730 C 1.47330 -61.09750 -16.74060 C 1.66240 -60.70120 -15.41500 C 1.69770 -61.65930 -14.39490 C 1.52650 -63.01490 -14.70350 C 1.31380 -63.40830 -16.02670 C 0.96260 -62.38620 -20.94400 C -0.43690 -61.78870 -21.04050 C -1.53710 -62.59310 -21.35930 C -2.81600 -62.04420 -21.46080 C -2.97200 -60.67360 -21.23980 C -1.88910 -59.84420 -20.92160 C -0.62530 -60.41380 -20.82340 C 1.25480 -62.59690 -24.00970 C 0.84860 -62.42540 -25.32850 C 0.76540 -61.13750 -25.87830 N 1.89090 -61.66500 -21.83910 N -4.30690 -60.08290 -21.34820 O 1.32350 -57.63620 -24.92240 O 0.60100 -58.49850 -26.86240 O 3.86460 -60.44980 -21.65710 O 3.67130 -61.13580 -18.85320 O 0.17280 -63.83100 -18.63370 O -5.25420 -60.81870 -21.65230 O -4.43500 -58.87230 -21.13160 O 0.35570 -61.01990 -27.16990 H 1.71700 -59.33420 -23.14910 H 1.44430 -60.35400 -17.53000 H 1.78110 -59.64710 -15.17770 H 1.85270 -61.35070 -13.36410 H 1.55280 -63.76070 -13.91330 H 1.16070 -64.45550 -16.27210 H 0.92230 -63.44340 -21.23400 H -1.39560 -63.65610 -21.53280 H -3.67240 -62.66150 -21.71020 H -2.03940 -58.78300 -20.76040 H 0.22460 -59.78260 -20.58060 H 1.31650 -63.59740 -23.59000 H 0.59040 -63.27980 -25.94820 H 0.34630 -60.03640 -27.37100 50 (R)-6-sF C -0.03490 -63.47270 -26.01580 C 0.58580 -62.32560 -25.26740 C 1.02940 -62.51260 -23.95060 C 1.56790 -61.46010 -23.21370 C 2.94360 -61.10810 -21.17410 C 2.77590 -61.45540 -19.69310 C 1.58340 -62.20990 -19.57060 C 1.02860 -62.66080 -18.36630 C -0.26300 -63.40680 -18.38890 C -0.36110 -64.64710 -19.03900 C -1.60080 -65.28250 -19.14380 C -2.74740 -64.67590 -18.61920 C -2.64980 -63.44340 -17.96410 C -1.40970 -62.81220 -17.83840 C 0.96370 -62.41190 -20.94730 C -0.46680 -61.91470 -21.12660 C -1.44740 -62.76920 -21.64530 C -2.74650 -62.31390 -21.86050 C -3.04210 -60.98320 -21.55880 C -2.08020 -60.10430 -21.04660 C -0.79240 -60.58260 -20.83100 C 1.69340 -60.19110 -23.80740 C 1.28140 -59.99280 -25.11950 C 0.71350 -61.04500 -25.85490 N 1.88150 -61.64590 -21.83310 N -4.40200 -60.48930 -21.78830 O -0.46550 -63.22770 -27.20200 O -0.11140 -64.60310 -25.46320 O 3.87460 -60.45400 -21.66070 O 3.58860 -61.08370 -18.81690 O -5.25180 -61.28010 -22.21310 O -4.64490 -59.30120 -21.54630 O 0.29330 -60.78510 -27.12140 H 0.92180 -63.50020 -23.51590 H 0.52590 -65.10890 -19.46480 H -1.67460 -66.24100 -19.65060 H -3.71500 -65.15970 -18.72520 H -3.53980 -62.96700 -17.56170 H -1.33600 -61.84270 -17.35300 H 1.00850 -63.46670 -21.24420 H -1.20170 -63.79890 -21.88620 H -3.50930 -62.97160 -22.26130 H -2.33930 -59.07470 -20.82750 H -0.03300 -59.91380 -20.43480 H 2.09590 -59.36120 -23.23690 H 1.36850 -59.01490 -25.58520 H -0.09460 -61.64220 -27.46890 H 1.15210 -62.76600 -16.32110 H 2.43050 -61.85890 -17.08400 N 1.57300 -62.40210 -17.17110 50 (R)-6-sX C 0.61790 -58.85370 -25.70920 C 0.95220 -60.20230 -25.13390 C 1.31840 -60.31700 -23.78520 C 1.61480 -61.55780 -23.22970 C 2.94250 -61.10440 -21.17360 C 2.76930 -61.45040 -19.68700 C 1.58680 -62.22260 -19.57290 C 1.03530 -62.71130 -18.38160 C -0.23750 -63.48760 -18.44150 C -0.30130 -64.69320 -19.15940 C -1.52230 -65.35490 -19.30670 C -2.68630 -64.81080 -18.75200 C -2.62320 -63.61630 -18.02640 C -1.40110 -62.95810 -17.86160 C 0.95110 -62.38730 -20.94500 C -0.44960 -61.80230 -21.10480 C -1.45330 -62.55000 -21.73150 C -2.72640 -62.01520 -21.92350 C -2.97250 -60.71420 -21.48110 C -1.98580 -59.94030 -20.85890 C -0.72440 -60.49530 -20.67500 C 1.54200 -62.71720 -24.01920 C 1.17330 -62.62630 -25.35690 C 0.87580 -61.37630 -25.92320 N 1.89770 -61.66900 -21.83140 N -4.30590 -60.13790 -21.67520 O 0.27170 -58.81650 -26.94580 O 0.68660 -57.83380 -24.97190 O 3.86870 -60.44650 -21.66290 O 3.57540 -61.07400 -18.80770 O -5.16890 -60.82620 -22.23110 O -4.51270 -58.98690 -21.27380 O 0.51490 -61.34050 -27.23200 H 1.35040 -59.42270 -23.17180 H 0.60050 -65.11010 -19.60020 H -1.56740 -66.28640 -19.86480 H -3.63980 -65.31480 -18.88760 H -3.52670 -63.18980 -17.59920 H -1.35630 -62.01630 -17.32120 H 0.92800 -63.43540 -21.26160 H -1.24590 -63.55980 -22.07280 H -3.50740 -62.59050 -22.40750 H -2.20630 -58.93080 -20.53130 H 0.05090 -59.90560 -20.19340 H 1.77100 -63.68700 -23.58520 H 1.10960 -63.51370 -25.98080 H 0.33780 -60.37610 -27.44860 H 1.15460 -62.87790 -16.34070 H 2.40730 -61.90950 -17.06820 N 1.56660 -62.47460 -17.17760 50 (R)-6-aF' C -0.12350 -63.18310 -26.08380 C 0.50380 -62.08610 -25.27100 C 0.96610 -62.35860 -23.97550 C 1.52160 -61.35720 -23.18560 C 2.94530 -61.11570 -21.17380 C 2.79750 -61.47320 -19.69230 C 1.58010 -62.20320 -19.57000 C 0.97990 -62.63300 -18.37720 C 1.21460 -61.88720 -17.11460 C 1.42870 -62.55380 -15.89650 C 1.63490 -61.81800 -14.72510 C 1.60650 -60.41890 -14.75700 C 1.37020 -59.75120 -15.96510 C 1.18200 -60.48040 -17.13740 C 0.97500 -62.42000 -20.95040 C -0.45680 -61.90710 -21.05150 C -1.51110 -62.78890 -21.31520 C -2.82290 -62.32050 -21.40730 C -3.05420 -60.95420 -21.24060 C -2.01790 -60.04860 -20.97980 C -0.72040 -60.53780 -20.88240 C 1.63640 -60.05080 -23.69620 C 1.20330 -59.76630 -24.98500 C 0.62690 -60.77140 -25.77860 N 1.87400 -61.63400 -21.83100 N -4.42660 -60.44700 -21.35010 O -0.55570 -62.86870 -27.25200 O -0.20260 -64.34390 -25.59920 O 3.87240 -60.45690 -21.66050 O 3.67360 -61.15440 -18.86530 O -5.33610 -61.25460 -21.56810 O -4.61440 -59.23220 -21.22010 O 0.19490 -60.43480 -27.02160 H 0.86470 -63.37080 -23.59960 H 1.46620 -63.64110 -15.86870 H 1.82310 -62.33660 -13.78890 H 1.76410 -59.84900 -13.84450 H 1.33420 -58.66490 -15.99320 H 0.99900 -59.96280 -18.07430 H 1.01180 -63.47450 -21.25930 H -1.31270 -63.84770 -21.46320
S116
H -3.64490 -62.99680 -21.61420 H -2.22770 -58.99140 -20.86140 H 0.09500 -59.84510 -20.68880 H 2.04940 -59.25900 -23.08020 H 1.28250 -58.75950 -25.38670 H -0.19150 -61.27110 -27.41830 H -0.44140 -63.84510 -17.50280 H 0.00390 -64.31750 -19.09760 N 0.11240 -63.65540 -18.33420 50 (R)-6-aX' C 1.26680 -58.87280 -25.86470 C 1.34490 -60.23220 -25.22670 C 1.59950 -60.34700 -23.85280 C 1.67770 -61.59710 -23.24410 C 2.93420 -61.09780 -21.17440 C 2.74400 -61.39100 -19.67890 C 1.59450 -62.23920 -19.57430 C 1.10290 -62.84860 -18.42200 C 1.83100 -62.73640 -17.12640 C 3.13960 -63.22660 -17.00770 C 3.82550 -63.10180 -15.79930 C 3.21430 -62.47250 -14.70640 C 1.90760 -61.98640 -14.82150 C 1.21040 -62.12590 -16.02660 C 0.93900 -62.35260 -20.93780 C -0.40040 -61.62050 -21.03520 C -1.54070 -62.27580 -21.51390 C -2.75390 -61.59320 -21.63150 C -2.79830 -60.24660 -21.26800 C -1.67230 -59.56490 -20.79050 C -0.47710 -60.26420 -20.67530 C 1.48770 -62.76170 -24.00460 C 1.22050 -62.66830 -25.36560 C 1.15090 -61.41120 -25.98680 N 1.91710 -61.69990 -21.83980 N -4.06600 -59.51680 -21.39260 O 1.01960 -58.83340 -27.12470 O 1.44290 -57.84600 -25.15580 O 3.86200 -60.44030 -21.65950 O 3.51690 -60.90560 -18.83840 O -5.05320 -60.12190 -21.82250 O -4.09120 -58.32670 -21.06100 O 0.89540 -61.37170 -27.31970 H 1.73160 -59.44290 -23.26830 H 3.61820 -63.69370 -17.86410 H 4.83900 -63.48510 -15.71270 H 3.75580 -62.36160 -13.77030 H 1.43120 -61.49470 -13.97720 H 0.19940 -61.73660 -16.12030 H 0.80970 -63.40010 -21.24170 H -1.48590 -63.32160 -21.80740 H -3.64160 -62.09200 -22.00340 H -1.73540 -58.51680 -20.52020 H 0.40300 -59.74390 -20.30720 H 1.55530 -63.73770 -23.53130 H 1.07090 -63.56020 -25.96750 H 0.88600 -60.39930 -27.57350 H -0.30780 -64.05600 -17.53980 H -0.63790 -63.70860 -19.18920 N -0.03090 -63.56970 -18.38780 50 (R)-6-aF C -0.35040 -63.11200 -25.99290 C 0.37900 -62.04600 -25.22420 C 0.86550 -62.32760 -23.94050 C 1.50770 -61.35180 -23.18280 C 2.94570 -61.11650 -21.17460 C 2.79890 -61.47550 -19.69170 C 1.57820 -62.20270 -19.56820 C 0.96860 -62.62420 -18.37670 C 1.21160 -61.88070 -17.11390 C 1.20040 -60.47370 -17.13780 C 1.41000 -59.74660 -15.96810 C 1.64350 -60.41720 -14.76070 C 1.64920 -61.81630 -14.72780 C 1.42450 -62.54970 -15.89730 C 0.97770 -62.42220 -20.95070 C -0.45810 -61.92160 -21.05230 C -1.50500 -62.81480 -21.30300 C -2.82140 -62.35990 -21.38960 C -3.06450 -60.99490 -21.23070 C -2.03520 -60.07740 -20.98520 C -0.73290 -60.55330 -20.89230 C 1.68810 -60.06330 -23.71950 C 1.23390 -59.77350 -25.00010 C 0.56950 -60.75020 -25.75830 N 1.87190 -61.63120 -21.83160 N -4.44190 -60.50270 -21.32930 O -0.79140 -62.78960 -27.15700 O -0.50190 -64.25170 -25.47790 O 3.87330 -60.45730 -21.66070 O 3.67650 -61.15740 -18.86650 O -5.34620 -61.32010 -21.53030 O -4.64000 -59.28880 -21.20610 O 0.11870 -60.40380 -26.99320 H 0.70910 -63.32550 -23.54640 H 1.01940 -59.95460 -18.07360 H 1.39330 -58.66020 -15.99760 H 1.81900 -59.84960 -13.85040 H 1.83510 -62.33690 -13.79250 H 1.44760 -63.63620 -15.87100 H 1.02190 -63.47700 -21.25760 H -1.29750 -63.87270 -21.44550 H -3.63820 -63.04530 -21.58550 H -2.25420 -59.02130 -20.87390 H 0.07690 -59.85230 -20.70820 H 2.16780 -59.28860 -23.13210 H 1.36430 -58.78030 -25.42140 H -0.33730 -61.21710 -27.36270 H -0.47860 -63.80650 -17.50410 H -0.03490 -64.29710 -19.09410 N 0.08920 -63.63820 -18.33020 50 (R)-6-aX C 0.63180 -58.53360 -25.45930 C 0.93200 -59.93180 -24.99530 C 1.31770 -60.15720 -23.66770 C 1.57430 -61.44590 -23.20880 C 2.94720 -61.11570 -21.17430 C 2.80170 -61.48280 -19.68880 C 1.58420 -62.21910 -19.57080 C 0.99380 -62.69370 -18.39120 C 1.29110 -62.04440 -17.08930 C 1.28380 -60.64020 -17.00780 C 1.54000 -60.00490 -15.79450 C 1.82120 -60.76570 -14.65220 C 1.82740 -62.16330 -14.72610 C 1.55270 -62.80470 -15.93800 C 0.95680 -62.38800 -20.94610 C -0.43320 -61.76440 -21.02740 C -1.54950 -62.54470 -21.34620 C -2.82150 -61.97190 -21.41280 C -2.94810 -60.60390 -21.16540 C -1.84660 -59.79730 -20.85360 C -0.59230 -60.38980 -20.78270 C 1.44400 -62.54030 -24.07650 C 1.05890 -62.33920 -25.39770 C 0.80170 -61.04120 -25.86680 N 1.89480 -61.66350 -21.83260 N -4.27660 -59.98680 -21.23820 O 0.27900 -58.39010 -26.68650 O 0.72990 -57.57900 -24.64370 O 3.86370 -60.44350 -21.66200 O 3.68080 -61.16760 -18.86480 O -5.24550 -60.70670 -21.50350 O -4.37030 -58.77230 -21.03030 O 0.42630 -60.89490 -27.16380 H 1.39940 -59.30930 -22.99610 H 1.07080 -60.05140 -17.89470 H 1.52390 -58.91950 -15.73920 H 2.03380 -60.26900 -13.70870 H 2.05150 -62.75360 -13.84180 H 1.57350 -63.89000 -15.99640 H 0.90980 -63.43810 -21.26740 H -1.43130 -63.60570 -21.55230 H -3.69250 -62.56980 -21.65660 H -1.97460 -58.73550 -20.67500 H 0.26980 -59.77260 -20.54450 H 1.64320 -63.54570 -23.71480 H 0.95230 -63.17600 -26.08240 H 0.28360 -59.91100 -27.30470 H -0.44730 -63.90460 -17.54980 H -0.07940 -64.28690 -19.18680 N 0.09140 -63.68760 -18.38390 50 (R)-9-sF C -0.14330 -63.42680 -25.98720 C 0.50430 -62.28930 -25.24650 C 0.98520 -62.49160 -23.94550 C 1.54420 -61.44670 -23.21260 C 2.93580 -61.11410 -21.17470 C 2.75940 -61.46330 -19.72210 C 1.59790 -62.19930 -19.57760 C 1.02010 -62.61430 -18.32570 C -0.26590 -63.39190 -18.36290 C -0.35670 -64.62530 -19.02730 C -1.58010 -65.29820 -19.09260 C -2.72380 -64.73150 -18.52030 C -2.63600 -63.50400 -17.85310 C -1.40890 -62.84320 -17.75980 C 0.96740 -62.41010 -20.94780 C -0.47070 -61.92340 -21.09900 C -1.45280 -62.78320 -21.60650 C -2.76500 -62.34560 -21.77220 C -3.07300 -61.02720 -21.43180 C -2.11000 -60.14270 -20.93210 C -0.80780 -60.60300 -20.76700 C 1.65660 -60.17170 -23.79500 C 1.20980 -59.95910 -25.09330 C 0.61790 -61.00240 -25.82320 N 1.87370 -61.64890 -21.83930 N -4.44900 -60.55380 -21.60300 O -0.60650 -63.16990 -27.15800 O -0.20940 -64.56120 -25.44160 O 3.87210 -60.45740 -21.64630 O 1.52920 -62.28590 -17.22620 O -5.29490 -61.34650 -22.03250 O -4.70860 -59.38110 -21.30990 O 0.16080 -60.72820 -27.07340 H 0.88770 -63.48340 -23.51750 H 0.52490 -65.06020 -19.49110 H -1.64470 -66.25410 -19.60580 H -3.68050 -65.24190 -18.59560 H -3.52410 -63.05960 -17.41160 H -1.34030 -61.88560 -17.25070 H 1.01370 -63.46770 -21.23650 H -1.19950 -63.80520 -21.87050 H -3.52990 -63.00870 -22.15980 H -2.37980 -59.12370 -20.67920 H -0.04920 -59.93090 -20.37520 H 2.07370 -59.34870 -23.22460 H 1.28510 -58.97650 -25.55050 H -0.24210 -61.57920 -27.41840 H 3.56240 -61.25030 -17.84340 N 3.66240 -61.04320 -18.83300 H 4.46880 -60.50750 -19.14520 50 (R)-9-sX C 0.42320 -58.64990 -25.47560 C 0.78140 -60.02970 -24.99830 C 1.24120 -60.21420 -23.68830 C 1.55490 -61.48420 -23.21530 C 2.93630 -61.11330 -21.17470 C 2.75670 -61.46250 -19.71850 C 1.60000 -62.20840 -19.57970 C 1.01790 -62.63740 -18.33500 C -0.27120 -63.40920 -18.38970
S117
C -0.37170 -64.61770 -19.09740 C -1.59790 -65.28330 -19.17960 C -2.73550 -64.73390 -18.57880 C -2.63820 -63.53210 -17.86740 C -1.40810 -62.87890 -17.75950 C 0.95720 -62.39190 -20.94610 C -0.45750 -61.83490 -21.08240 C -1.46660 -62.61740 -21.65640 C -2.76040 -62.11840 -21.79800 C -3.02170 -60.81780 -21.36250 C -2.02860 -60.00750 -20.79990 C -0.74610 -60.52730 -20.66380 C 1.40740 -62.60280 -24.05010 C 0.94450 -62.44240 -25.35180 C 0.62440 -61.16310 -25.83360 N 1.88640 -61.66560 -21.83690 N -4.38040 -60.28500 -21.49380 O -0.03120 -58.54700 -26.67250 O 0.57680 -57.67050 -24.69730 O 3.86690 -60.45060 -21.64830 O 1.52490 -62.32900 -17.22970 O -5.25140 -61.00890 -21.98900 O -4.60070 -59.13380 -21.10060 O 0.15980 -61.05850 -27.10530 H 1.32540 -59.35250 -23.03470 H 0.50570 -65.04050 -19.57970 H -1.66870 -66.22070 -19.72520 H -3.69450 -65.23850 -18.66510 H -3.52150 -63.10160 -17.40300 H -1.33230 -61.94070 -17.21630 H 0.95170 -63.44520 -21.25210 H -1.24820 -63.62780 -21.98800 H -3.54810 -62.72240 -22.23410 H -2.26100 -58.99950 -20.47570 H 0.03310 -59.91240 -20.22200 H 1.64110 -63.59620 -23.67570 H 0.81510 -63.29860 -26.00820 H -0.02800 -60.08400 -27.25630 H 3.54530 -61.23730 -17.83490 N 3.65090 -61.03330 -18.82480 H 4.45260 -60.48700 -19.13160 50 (R)-9-aF' C -0.01230 -63.55010 -26.00840 C 0.62810 -62.40320 -25.27930 C 1.03490 -62.56760 -23.94800 C 1.59930 -61.51260 -23.23430 C 2.92510 -61.10800 -21.17470 C 2.73820 -61.42720 -19.71380 C 1.61040 -62.21240 -19.58170 C 1.00260 -62.84610 -18.43200 C 1.77120 -63.02450 -17.15920 C 3.07720 -63.54160 -17.18060 C 3.73770 -63.82630 -15.98130 C 3.10500 -63.57570 -14.75700 C 1.80460 -63.05240 -14.73200 C 1.13290 -62.79150 -15.92890 C 0.95400 -62.37690 -20.94100 C -0.44000 -61.77050 -21.07150 C -1.51710 -62.55750 -21.49620 C -2.77760 -61.99320 -21.68540 C -2.93880 -60.62600 -21.44440 C -1.88030 -59.81710 -21.01640 C -0.63180 -60.40200 -20.83180 C 1.78120 -60.26670 -23.86220 C 1.40600 -60.09270 -25.18850 C 0.81910 -61.14790 -25.90450 N 1.89170 -61.68000 -21.84840 N -4.25430 -60.01640 -21.65350 O -0.39660 -63.32980 -27.21420 O -0.14720 -64.65770 -25.42350 O 3.86580 -60.45250 -21.63990 O -0.15850 -63.30260 -18.50100 O -5.18330 -60.73720 -22.03560 O -4.38230 -58.80520 -21.44050 O 0.44430 -60.91720 -27.18980 H 0.88320 -63.53800 -23.48560 H 3.56120 -63.74850 -18.13240 H 4.73850 -64.24990 -16.00380 H 3.62020 -63.79450 -13.82470 H 1.31300 -62.85940 -13.78150 H 0.11530 -62.40900 -15.91960 H 0.90930 -63.44100 -21.20580 H -1.37400 -63.61830 -21.68290 H -3.61470 -62.59500 -22.02050 H -2.03390 -58.75840 -20.84070 H 0.20140 -59.78470 -20.50790 H 2.20320 -59.43510 -23.30850 H 1.53910 -59.13270 -25.68110 H 0.03680 -61.77470 -27.51820 H 3.56840 -61.00850 -17.85580 N 3.63190 -60.90660 -18.86150 H 4.39020 -60.33780 -19.22990 50 (R)-9-aX' C 1.16750 -59.04800 -25.97420 C 1.27650 -60.37830 -25.28290 C 1.57620 -60.43270 -23.91340 C 1.67470 -61.65430 -23.25370 C 2.92460 -61.10540 -21.17350 C 2.73370 -61.43000 -19.71120 C 1.61540 -62.22570 -19.58650 C 1.04380 -62.93660 -18.45770 C 1.90940 -63.38540 -17.32290 C 3.22090 -63.83510 -17.55390 C 3.96630 -64.39070 -16.51160 C 3.41360 -64.48540 -15.22970 C 2.10880 -64.03120 -14.99130 C 1.35520 -63.49500 -16.03610 C 0.93890 -62.34880 -20.93760 C -0.41120 -61.64610 -21.06600 C -1.50760 -62.32080 -21.61560 C -2.72170 -61.66290 -21.81170 C -2.81810 -60.31810 -21.44770 C -1.73870 -59.61990 -20.89410 C -0.53700 -60.29540 -20.70890 C 1.46580 -62.85110 -23.95960 C 1.15410 -62.81760 -25.31350 C 1.05640 -61.58840 -25.98440 N 1.90930 -61.70700 -21.84670 N -4.08660 -59.61390 -21.65610 O 0.85690 -59.06180 -27.22050 O 1.38040 -57.99280 -25.31930 O 3.85860 -60.44300 -21.64150 O -0.15860 -63.27330 -18.47300 O -5.02140 -60.22740 -22.18480 O -4.17020 -58.43420 -21.29570 O 0.74680 -61.60760 -27.30660 H 1.71640 -59.50410 -23.37060 H 3.64220 -63.78690 -18.55350 H 4.97070 -64.75950 -16.70330 H 3.99380 -64.91970 -14.42000 H 1.67840 -64.10800 -13.99640 H 0.33340 -63.16500 -15.86830 H 0.82100 -63.40450 -21.21050 H -1.41570 -63.36640 -21.89570 H -3.57370 -62.17720 -22.24130 H -1.84100 -58.57510 -20.62360 H 0.31070 -59.76260 -20.28740 H 1.54410 -63.80690 -23.44850 H 0.98430 -63.73510 -25.87000 H 0.71970 -60.64810 -27.60120 H 3.52020 -61.01830 -17.83630 N 3.61670 -60.91810 -18.84020 H 4.36590 -60.32310 -19.18480 50 (R)-9-aF C -0.33560 -63.22030 -25.97250 C 0.38140 -62.13170 -25.22230 C 0.87730 -62.38490 -23.93640 C 1.50990 -61.38730 -23.19780 C 2.93260 -61.12010 -21.17350 C 2.77360 -61.48920 -19.72320 C 1.59920 -62.19100 -19.57940 C 0.95840 -62.77550 -18.40870 C 1.03930 -62.07300 -17.09180 C 1.15100 -60.67250 -17.02350 C 1.11570 -60.02210 -15.78850 C 0.97770 -60.76510 -14.61070 C 0.86540 -62.16100 -14.67090 C 0.88690 -62.81150 -15.90500 C 0.97950 -62.41650 -20.95000 C -0.46790 -61.95380 -21.06520 C -1.48300 -62.87980 -21.33430 C -2.80700 -62.46310 -21.47430 C -3.09340 -61.10260 -21.34410 C -2.09970 -60.15570 -21.06960 C -0.78720 -60.59350 -20.92810 C 1.67660 -60.10820 -23.75830 C 1.21330 -59.84660 -25.04210 C 0.55250 -60.84240 -25.77820 N 1.86450 -61.63960 -21.84040 N -4.47800 -60.64880 -21.50860 O -0.81600 -62.91240 -27.12470 O -0.43860 -64.36360 -25.45470 O 3.87100 -60.46260 -21.64240 O 0.27390 -63.81310 -18.52200 O -5.34910 -61.49190 -21.74930 O -4.71640 -59.44060 -21.40080 O 0.08740 -60.52130 -27.01460 H 0.73360 -63.37730 -23.52360 H 1.23290 -60.08960 -17.93530 H 1.18560 -58.93850 -15.74680 H 0.94930 -60.25820 -13.64950 H 0.75310 -62.73710 -13.75650 H 0.78100 -63.89080 -15.96250 H 1.03830 -63.47980 -21.21740 H -1.24130 -63.93320 -21.44480 H -3.59700 -63.17350 -21.69070 H -2.35230 -59.10560 -20.97670 H -0.00420 -59.86800 -20.72540 H 2.14990 -59.31920 -23.18440 H 1.33120 -58.86010 -25.48220 H -0.36790 -61.34390 -27.36410 H 3.76600 -61.40240 -17.89950 N 3.75860 -61.13750 -18.87790 H 4.55750 -60.61980 -19.23410 50 (R)-9-aX C 0.73540 -58.54650 -25.51690 C 0.94570 -59.94920 -25.01950 C 1.36950 -60.16500 -23.70190 C 1.54920 -61.45560 -23.21370 C 2.93310 -61.11900 -21.17420 C 2.77440 -61.49270 -19.72060 C 1.60300 -62.20350 -19.58120 C 0.97850 -62.84870 -18.43570 C 1.12220 -62.26700 -17.06760 C 1.25020 -60.87970 -16.87800 C 1.26630 -60.34510 -15.58770 C 1.16740 -61.19220 -14.47820 C 1.04270 -62.57630 -14.66030 C 1.01030 -63.11060 -15.94830 C 0.96470 -62.39160 -20.94630 C -0.45010 -61.83200 -21.05260 C -1.51590 -62.66690 -21.40730 C -2.80770 -62.15530 -21.53890 C -3.00990 -60.79240 -21.31160 C -1.96280 -59.93420 -20.95510 C -0.68440 -60.46610 -20.82590 C 1.29210 -62.56200 -24.03880 C 0.86030 -62.36900 -25.34640 C 0.68500 -61.06940 -25.84680 N 1.88130 -61.66300 -21.84060 N -4.36060 -60.24000 -21.45720 O 0.33920 -58.41130 -26.73040 O 0.94850 -57.57850 -24.73860 O 3.86470 -60.45260 -21.64350
S118
O 0.25770 -63.85430 -18.60790 O -5.27650 -61.00270 -21.78480 O -4.52690 -59.03350 -21.24730 O 0.25690 -60.93150 -27.12830 H 1.53620 -59.30790 -23.05840 H 1.30240 -60.21680 -17.73650 H 1.34510 -59.27020 -15.44950 H 1.17950 -60.77540 -13.47440 H 0.96200 -63.23360 -13.79880 H 0.89230 -64.17950 -16.10090 H 0.95340 -63.45360 -21.22140 H -1.33930 -63.72340 -21.58860 H -3.63770 -62.79500 -21.81780 H -2.14930 -58.87900 -20.78990 H 0.13820 -59.80870 -20.55710 H 1.42200 -63.57010 -23.65410 H 0.64930 -63.21460 -25.99540 H 0.18290 -59.94410 -27.29500 H 3.77840 -61.40590 -17.90310 N 3.76170 -61.13940 -18.88080 H 4.55890 -60.62180 -19.24220
S119
S120
VI.
Supplementary References
1 Neves, J. F. et al.
Backbone chemical shift assignments of human 14-3-3 σ . , 103-107, doi:10.1007/s12104-018-9860-1 (2019). 2 Burkhardt, A. et al. Status of the crystallography beamlines at PETRA III.
The European Physical Journal Plus , 56, doi:10.1140/epjp/i2016-16056-0 (2016). 3 Meents, A. et al.
Development of an in-vacuum x-ray microscope with cryogenic sample cooling for beamline P11 at PETRA III . Vol. 8851 OPO (SPIE, 2013). 4 Winter, G. xia2: an expert system for macromolecular crystallography data reduction.
Journal of Applied Crystallography , 186-190, doi:doi:10.1107/S0021889809045701 (2010). 5 McCoy, A. J. et al. Phaser crystallographic software.
Journal of Applied Crystallography , 658-674, doi:doi:10.1107/S0021889807021206 (2007). 6 Adams, P. D. et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution.
Acta Crystallographica Section D , 213-221, doi:doi:10.1107/S0907444909052925 (2010). 7 Emsley, P. & Cowtan, K. Coot: model-building tools for molecular graphics. Acta Crystallographica Section D , 2126-2132, doi:doi:10.1107/S0907444904019158 (2004). 8 Chen, V. B. et al. MolProbity: all-atom structure validation for macromolecular crystallography.
Acta Crystallographica Section D , 12-21, doi:doi:10.1107/S0907444909042073 (2010). 9 Karplus, P. A. & Diederichs, K. Linking Crystallographic Model and Data Quality. Science , 1030-1033, doi:10.1126/science.1218231 (2012). 10 Rose, R. et al.
Identification and Structure of Small-Molecule Stabilizers of 14–3–3 Protein–Protein Interactions. , 4129-4132, doi:10.1002/anie.200907203 (2010). 11 Richter, A., Rose, R., Hedberg, C., Waldmann, H. & Ottmann, C. An Optimised Small-Molecule Stabiliser of the 14-3-3–PMA2 Protein–Protein Interaction. Chemistry – A European Journal , 6520-6527, doi:10.1002/chem.201103761 (2012). 12 Small-Molecule Drug Discovery Suite v. 2019-2 (Schrödinger, LLC, New York, NY, 2019). 13 Chang, G., Guida, W. C. & Still, W. C. An internal-coordinate Monte Carlo method for searching conformational space. Journal of the American Chemical Society , 4379-4386, doi:10.1021/ja00194a035 (1989). 14 Roos, K. et al.
OPLS3e: Extending Force Field Coverage for Drug-Like Small Molecules.
Journal of Chemical Theory and Computation , 1863-1874, doi:10.1021/acs.jctc.8b01026 (2019). 15 Still, W. C., Tempczyk, A., Hawley, R. C. & Hendrickson, T. Semianalytical treatment of solvation for molecular mechanics and dynamics. Journal of the American Chemical Society , 6127-6129, doi:10.1021/ja00172a038 (1990). 16 Bochevarov, A. D. et al.
Jaguar: A high-performance quantum chemistry software program with strengths in life and materials sciences.
International Journal of Quantum Chemistry , 2110-2142, doi:10.1002/qua.24481 (2013). 17 Grimme, S., Antony, J., Ehrlich, S. & Krieg, H. A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu.
The Journal of Chemical Physics , 154104, doi:10.1063/1.3382344 (2010). 18 Becke, A. D. Density - functional thermochemistry. III. The role of exact exchange. The Journal of Chemical Physics , 5648-5652, doi:10.1063/1.464913 (1993). 19 Marten, B. et al. New Model for Calculation of Solvation Free Energies:
Correction of Self-Consistent Reaction Field Continuum Dielectric Theory for Short-Range Hydrogen-Bonding Effects.
The Journal of Physical Chemistry , 11775-11788, doi:10.1021/jp953087x (1996). 20 Kelly, C. P., Cramer, C. J. & Truhlar, D. G. SM6:
A Density Functional Theory Continuum Solvation Model for Calculating Aqueous Solvation Free Energies of Neutrals, Ions, and Solute − Water Clusters.
Journal of Chemical Theory and Computation , 1133-1152, doi:10.1021/ct050164b (2005). 21 Zhao, Y. & Truhlar, D. G. The M06 suite of density functionals for main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: two new functionals and systematic testing of four M06-class functionals and 12 other functionals. Theoretical Chemistry Accounts , 215-241, doi:10.1007/s00214-007-0310-x (2008)., 215-241, doi:10.1007/s00214-007-0310-x (2008).