This website uses cookies so that we can provide the best service possible. For more information, please visit our privacy policy.
ok

sign in
|
register

scientific challenge

hall of fame
winners of the past scientific challenges

summer 2017

Julius Pollinger
Goethe-University
Frankfurt am Main, Germany

Metabolic disorders

Computer-assisted selective optimization of side-activities
Julius summarizes:
The discovery of bioactive small molecules is an expensive and time consuming yet central task in drug discovery. A potentially superior way to identify new drug candidates is the selective optimization of side activities (SOSA), which employs known drugs for their side-activities as lead compounds. Virtually every small molecule drug interacts with more than a single molecular target and, thus, has side-activities. Sometimes, these side-activities may be of therapeutic value and structural optimization to turn the side-activity into the main activity can generate a new drug. As a key advantage of this strategy, analogues of approved drugs inherit their previously optimized favourable characteristics concerning toxicity, solubility, bioavailability and metabolic stability, and are drug-like by definition. We intended to apply the SOSA strategy to the CysLT1 inhibitor cinalukast for which we have observed a previously unknown side-activity on the peroxisome-proliferator activated receptors (PPAR) alpha and gamma. These ligand-activated transcription factors play a crucial role in metabolic disorders. To speed up the SOSA-focussed structural variation of cinalukast towards potent PPAR modulators, we intended to employ in silico techniques to support the SOSA concept. We started this endeavour using SeeSAR and its editing mode to visualize and identify key ligand-receptor interactions. Moreover, FlexX and Hyde served to analyse the binding mode of a small combinatorial set of cinalukast derivatives and allowed an in silico estimation of its structure-activity relationship (SAR). This enabled us to select structural elements whose variation promised to improve potency on PPARs. As a proof-of-principle, two cinalukast analogues were prepared and tested in vitro. Their experimentally determined modulatory activity on PPARs nicely correlated with the predictions of Hyde confirming the suitability of our computational approach to support SOSA. Moreover, these cinalukast analogues revealed remarkable improvements in toxicity over the lead compound. Next, we generated a combinatorial library of approx. 8000 cinalukast derivatives using adaptions of the KNIME workflows STORM and MedChemWizard. The BioSolveIT KNIME nodes were then employed for automated docking and screening to identify the most promising candidate compounds in the library. With the implementation of this automated and time-saving workflow for a computationally guided SOSA approach, we are speeding up the structural optimization of cinalukast towards potent PPAR modulators and several predicted analogues are in preparation and characterization. This successful combination of computational tools in compound optimization highlights the potential of computer-assisted SOSA for future drug discovery.
The following goals have been achieved:
  1. First, we intended to identify variations of our lead structure cinalukast to counter the compound’s high toxicity and improve key characteristics such as selectivity or solubility. We employed SeeSAR to study a manually defined library of 40 analogues comprising variations which we chose as a starting SAR. We observed that the core scaffold of cinalukast provides an ideal angle for the ligand-receptor interaction and therefore our initially intended changes in this part were discarded. Thereby, the computational approach significantly reduced the number of analogues that had to be prepared. The most promising analogues according to SeeSAR were further investigated with FlexX and Hyde scoring to analyse ligand-target interactions and binding contributions. Two analogues were selected for synthesis and in vitro characterization taking into account all computational results to confirm a correlation between predicted and experimentally determined values.
  2. As a second goal, we aimed for broader variations of the lead structure to improve potency on the new target. We tried to employ the KNIME workflows MedChemWizard and STORM and successfully adapted them to our needs. With the basic concept of these workflows and using KNIME nodes provided by BioSolveIT, we built and screened a large combinatorial library of approx. 8000 analogues of the lead compound. Predicted information on compound binding obtained from SeeSAR helped the library design. After the automated library generation and screening workflow, we manually analysed top-ranking compounds by docking with the SeeSAR and using the “Interactive BioSolveIT table” for various features. The three most potent derivatives from the automated screening workflow turned out very favourable and were selected for synthesis and in vitro characterization.
  3. As a third objective, we aimed to confirm the predicted data (goal 1 and 2) in vitro by synthesizing and testing the selected analogues. To confirm the suitability of our computational approach, we selected two compounds for synthesis to study the correlation between predicted and experimentally determined value. The in vitro activities of these analogues nicely correlated with their predicted Hyde scores thus characterizing this ranking approach as suitable for our optimization strategy. Moreover, we selected compounds for synthesis to reduce the problematic toxicity of the compound class which was achieved by minor structural changes deduced from SeeSAR calculations. Finally, we employed the previously validated Hyde scoring as decision method for selecting compounds from a large combinatorial library to optimize potency of the compound class with as few analogues to be prepared as possible and the three top-ranking candidates are in synthesis/in vitro characterization.

spring 2017

 
Epigenetics Drug Discovery of natural product libraries to be anticancer/antidiabetic
Veera Chandra Sekhar Reddy Chittepu
Escuela Nacional de Ciencias Biológicas, Instituto Politecnico Nacional, Mexico City, Mexico
Veera Chandra Sekhar Reddy Chittepu
Escuela Nacional de Ciencias Biológicas, Instituto Politecnico Nacional
Mexico City, Mexico

Cancer and Diabetes

Epigenetics Drug Discovery of natural product libraries to be anticancer/antidiabetic
Veera Chandra Sekhar Reddy summarizes:
Natural products have proven to inhibit Lysine-Specific Demethylase-1 (LSD1) were retrieved from literature, and structure-activity relationship (SAR) is depicted using See SAR software. See SAR helped us to visualize hydrophobic, electrostatic and other interactions responsible for inhibition of LSD1. There were no natural products proven to inhibit Peptidyl Arginine Deiminase 4 (PADI4). We had docked manually curated natural product library retrieved from PubChem library, and four natural products were chosen to validate experimentally. We had purchased the natural products from commercial vendors and inbuilt LSD1, and PADI4 inhibitor screening assays were carried out. It resulted in universal inhibitor to inhibit both proteins LSD1 and PADI4. We had utilized ReCore module from LeadIT to develop the derivatives of common inhibitors for the potent inhibitor. We utilized FlexX docking tool to predict the binding affinity and estimated the Gibbs free energy using SeeSAR and as well FlexX score. We repeated the process till we identified inhibitors possesing free energy -kCal/mol. Use of See SAR, FLexX, and ReCore modules resulted in best scoring leads to inhibit PADI4 and LSD1
The following goals have been achieved:
  1. The primary objective of goal 1 was to developed Structure-Activity Relationship (SAR) on natural products to inhibit LSD1 and PADI4 based on literature using SeeSAR and FLexX docking methodology. Proven natural products to inhibit LSD1 are resveratrol, Curcumin, quercetin, myricetin, luteolin, apigenin, genistein, and TCP. The respective three-dimensional structures(natural products) was retrieved from PubChem, and the poses were calculated using FlexX score and the same are used to predict the structure-activity relationship model. To complete the goal to build the model for PADI4, as there were no proven chemicals to inhibit PADI4, we had screened natural products from library using FlexX score and based on experimental studies new model was built on chemicals to inhibit PADI4
  2. The main objective of goal 2 is to perform flexible docking and chose the best scoring chemical based on the rank, score and interacting residues. This goal is to address the optimization of leads to LSD1 and PADI4. ReCore module is used to design, develop derivatives of top scoring leads to inhibit our drug targets. Briefly, natural product library was retrieved from National Cancer Institute (NCI) and the library is used in the process of flexible Docking. FlexX tool is used to perform protein-ligand interactions using Flexx score and the top scoring natural products were studied for their binding interactions The top ranking pose is used as input for ReCore module to develop derivatives. having visual on binding pose view, interacting residues and the scope of expanding the natural product had helped us to develop derivatives.
  3. The main objective of goal 3 is to validate initial leads also called as hits to validate experimentally and use the same tools to optimize the proposed leads. Five natural products were purchased from commercial vendors, and they are validated experimentally using inbuilt experimental assays. SeeSAR, FlexX and ReCore modules were used to develop their derivatives, and a library of 100 chemicals was designed and flexX is used to score the derivatives. We are in the process of synthesis and validating the new derivatives to inhibit LSD1 and PADI4.

winter 2016

 
Small-molecule glycosyltransferase (GTs) inhibitors based on pyranose monosaccharides
Marko Jukič
Department of Medicinal Chemistry, Faculty of Pharmacy, Universiy of Ljubljana, Ljubljana, Slovenia
Marko Jukič
Department of Medicinal Chemistry, Faculty of Pharmacy, Universiy of Ljubljana
Ljubljana, Slovenia

Bacterial infections - Antibacterial drug discovery

Small-molecule glycosyltransferase (GTs) inhibitors based on pyranose monosaccharides
Marko summarizes:
Peptidoglycan transglycosylases or alternatively called glycosyltransferases (GTs) are membrane-bound enzymes that catalyze the polymerization step in cell-wall biosynthesis. They represent a prominent target in antibacterial drug discovery and only a few inhibitors are described in the literature. Moreover, native substrates Lipid II/Lipid IV display considerable structural complexity while previously described inhibitors often closely follow their complex structural motifs. Moenomycins for example, are structured as a tetrasaccharides with an isoprenoid chain connected through a phosphoglyceric acid moiety. Recently, monosaccharide disubstituted with two aromatic moieties, comparatively smaller, inhibitos of glycosyltransferases, have been described by Zuegg et al. Theese compounds could facilitate development of novel small molecule GT inhibitors. As the binding mode of described inhibitors is not known, four ligand based drug design campaigns have been performed with the help of BioSolveIT KNIME tools ammounting to the library of 83 purchased compounds that were all biologically evaluated for their relative inhibition of the transglycosilase reaction of PBP1b from E. coli using a fluorescence assay. Three structurally similar hits were obtained where one compound displayed residual activity of PBP1b at 10 % and is expected to possess IC50 in the low micromolar range. In order to unequivocally describe the PBP1b inhibition, further Lipid II preparation is under way for the support of biological testing and a synthetic campaign has been started to resynthesize purchased hits with preparation of relevant analogues, especially lipophilic-chain conjugates. Latter compounds could shed further light on Lipid II mimicry if present and be used as molecular probes for further inhibitor design/optimisation.
The following goals have been achieved:
  1. First and most important phase 1 goal (all phase 1 goals accomplished). Establish and set up bological evaluation support. From the start of campaign, biological evaluation support has been established where purchased compounds can be evaluated for PBP1b inhibition with a fluorescence test. Substrate consumption by PBP1b can be monitored through fluorescamine labeled Lipid II and HPLC analysis. Test system consists of E. coli enzyme, Lipid II substrate in a suitable buffer/detergent system. Furthermore, in vitro antimicrobial testing and cytotoxicity testing using human embryonic kidney [HEK] 293 cells awaits for (re-)synthesized hit compounds and analogues. Here the emphasis was on the complexity of test system and availability of Lipid II substrate.
  2. Milestone 1 (phase 2) – Identify a hit compound (or compounds) on the basis of 1st, 2nd or multiple iterations of ligand based drug design. With the help of BioSolveIT KNIME tools, 4 iterations of ligand based drug design have been performed. Final library of purchased compounds consists of 83 evaluated compounds and three identified hits. Hit compounds are all structurally related and bear a central 2-(piperazin-1-yl)-1,4,5,6-tetrahydropyrimidine moiety. We hope we will be able to prepare suitable structurally decorated analogues that would display even greater activity on isolated PBP1b and maybe even posess antimicrobial activity. Both outcomes are favourable as molecular probes for the study of glycosyltransferase inhibition are more than welcome.
  3. Milestone 2 (phase 2) – Perform further in silico studies extending from binding mode studies to supporting in silico studies in hit identification and design. With the KNIME – BioSolveIT synergy, multiple workflows have been developed that span from library filtering, PAINS identification, similarity search and visualisation of results. First three ligand based design iterations were focused on monosaccharide disubstituted query molecule, while the fourth campaign has been tailored to expand the chemical space of a first hit compound identified. With the fourth ligand based design iteration, we identified additional two hits with further enhanced activity on E. coli PBP1b. Following resynthesis of hits and relevant analogues, additional structure-based studies will be conducted using BioSolveIT-KNIME integration. The emphasis of the whole project resides on lean and transparent hypotheses in design phase where ligand based workflow is used and extensive experimental support.

fall 2016

 
Drug-Design of 3-Amino Derivatives of (-)-Cytisine- Potential Inhibitors of COX-2
Sophia Borisevich
Ufa Institute of Chemistry, Ufa, Russian Federation
Sophia Borisevich
Ufa Institute of Chemistry
Ufa, Russian Federation

Stroke, all types of taupathies, multiple sclerosis

Drug-Design of 3-Amino Derivatives of (-)-Cytisine- Potential Inhibitors of COX-2
Sophia summarizes:
Usually the process of inflammation is accompanied by neurodegenerative diseases. It may be either a nonspecific response of cells to the nervous system disease or it causes the disease. It is known that if patients take non-steroidal anti-inflammatory drugs, they are unlikely to develop these diseases. Some derivatives of the (-)-cytisine are known for high neuropharmacological activity. These compounds possess pronounced mnestic activity, comparable with the activity of piracetam, and also demonstrate anti-inflammatory activity. Within this context, some of (-)-cytosine derivatives can be considered as substances with two therapeutic indications: anti-inflammatory and neuropharmacological. In the first case the COX-1/2 receptors were considered as biological targets responsible for the inflammatory process. Secondly we examined ionotropic glutamate receptors, involved in learning and memory formation. According to biological experiments and in silico studies we selected two lead-compounds among these derivatives of 12-N-methylcytisine (see figure). These compounds have demonstrated ability to inhibit the oedema comparable with our reference drug diclofenac in carrageenan-induced model of inflammatory in vivo and had a moderate ability to inhibit COX-2 (in vitro test). According to the molecular docking results we suppose that the mnestic activity of derivatives of 12-N-methylcytisine can be explained by its effect on the work of AMPA and KA-receptors. The molecular structures of these compounds were modified saving the molecular core (cytosine fragment). We pursued two goals: on the one hand, to increase the affinity for the tyrosine site of COX-2, on the other hand save the high value of mnestic activity. Based on the calculation results, we synthesized two compounds for the in vivo test. New derivatives of (-)-cytisine demonstrated pronounced anti-inflammatory activity exceeding the activity of diclofenac and had a moderate mnestic activity (see figure). Theoretical calculations correlate with the results of biological tests. However, all of the above is just the first step towards creating new compounds with two biological activities. A number of additional studies are needed.
The following goals have been achieved:
  1. We used a bioisosterism strategy for rational design of new derivatives of (-)-cytisine and created a database of new compounds. We have chosen biological targets and studied their activity. We estimated ADME parameters of new compounds and have performed the docking process into the tyrosine active site of COX-2 and binding sites of AMPA and KA-receptors.
  2. Accordind to the calculation results we selected lead-compounds and synthesized new compounds with anti-inflammatory activity. They have demonstrated ability to inhibit the oedema comparable with diclofenac in carrageenan-induced model of inflammation in vivo.
  3. We suppose that mnestic activities of 12-N-methylcytisine derivatives can be explained by their effect on the work of AMPA and KA-receptors.

summer 2016

 
Inhibitors of dextransucrase enzymes and the decrease of dextran content in the sugar
Reinaldo Fraga
Cuban Research Institute on Sugarcane By-Products (ICIDCA), Havana, Cuba
Reinaldo Fraga
Cuban Research Institute on Sugarcane By-Products (ICIDCA)
Havana, Cuba

Industry, Sugar production, Anti-caries active compounds

Inhibitors of dextransucrase enzymes and the decrease of dextran content in the sugar
Reinaldo summarizes:
A structure-based drug discovery (SBDD) strategy was followed by using the quinoxaline derivative (ZINC 08382282, previously identified as a glucansucrase inhibitor by other authors) as lead molecule to perform a search for similar compounds with FTrees Fragment Space. The Fragment Space Package used was KnowledgeSpace_2.0_20160614 and all the procedure was run on EVOspace-search workflow for KNIME (KNIME Analytics Platform 2.11.3). The node Search FTrees Fragment Space rendered the best results when it was configurated as follow: Number of results _300, Target similarity _0.9, Target diversity _1. The screening of all similars was done by docking them into the active site of the glucansucrase GTF 180-substrate complex pdb code: 3HZ3, using the LeadIT node. The binding modes of compounds were visualized and evaluated with the node Interactive SeeSAR Viewer. Two compounds showed the higher binding affinity, ligand efficiency and torsion. One of this is a very fresh or new compound, not previously synthesized, at least up to where I know. Currently I’m about to get some hundred milligrams of it in collaboration with a research group from Spain. So, probably by the next month I will be trying its inhibitory effects in vitro. If all goes well new variants of it will be synthesized to further potentiate its inhibitory effects. On the other hand, I also did the screening of a library of 248 furanic compounds. The furanic compounds rendered several candidates which could be further optimized.
The following goals have been achieved:
  1. Goal 1. Identify possible inhibitor candidates by similarity screening of Quinoxaline. The Quinoxaline derivative (ZINC 08382282 saved as sdf file) was used as query molecule to perform a search for similar compounds with FTrees Fragment Space. The Fragment Space Package used was KnowledgeSpace_2.0_20160614 and all the procedure was run on EVOspace-search workflow for KNIME (KNIME Analytics Platform 2.11.3). The node Search FTrees Fragment Space was configurated for several conditions, for example: Number of results were varied from_200 to 1000, and Target similarity was moved from_0.9 to 0.8, Target diversity always was kept in_1.
  2. Goal 2. Identify plausible binding modes of compounds from goal 1 by docking. For the docking procedure, the receptor node (Prepare Receptor with LeadIT) was charged with the receptor molecule from the enzyme GTF 180 (PDB: 3HZ3). The receptor was prepared according to the default parameters, with a file 3HZ3.fxx previously prepared. In the Receptor Components, only Chain A was included (the unique), then chose the Reference Ligand SUC-1779-A, Chemical Ambiguities were kept as by default. The binding modes of compounds were evaluated with the node Interactive SeeSAR Viewer.
  3. Goal 3. Identify possible inhibitor candidates from a library of 248 furanic compounds by docking. A library of 248 furanic compounds was inspected in the previous receptor molecule (PDB: 3HZ3), prepared as for goal 2. Around 100 poses of each molecule were analyzed with LeadIT and later visualized with SeeSAR.

spring 2016

 
Reversible inhibitors against drug resistant Epidermal Growth Factor Receptor (EGFR) mutant
Dr. Subhash Mohan Agarwal and Dr Prajwal Nandekar .
Heidelberg Institute for Theoretical Studies (PN), Germany AND Institute of Cytology and Preventive Oncology (SMA), Noida, India
Dr. Subhash Mohan Agarwal and Dr Prajwal Nandekar .
Heidelberg Institute for Theoretical Studies (PN), Germany AND Institute of Cytology and Preventive Oncology (SMA)
Noida, India

Cancer

Reversible inhibitors against drug resistant Epidermal Growth Factor Receptor (EGFR) mutant
Dr. Subhash Mohan Agarwal and Dr Prajwal Nandekar summarizes:
Epidermal growth factor receptor is one of the most studied cancer drug target of pharmaceutical relevance. Several first generation inhibitors of EGFR (Gefitinib and Erlotinib) are in clinical use targeting L858R mutant EGFR. However, acquisition of a secondary point mutation (T790M) leads to acquired resistance to these inhibitors, rendering gefitinib and erlotinib ineffective. Consequently, the search for mutant specific inhibitors led to identification of few covalent molecules which have shown limited clinical efficacy and toxicity. Recently, due to toxicity issues of covalent inhibitors, there has been a shift in the focus of the researchers which is now inclined to identify reversible inhibitors which target both EGFR activating mutation (L858R) as well as drug resistant mutant (T790M) i.e. TMLR inhibitors. Thus, in this project we proposed to identify reversible inhibitors against TMLR mutant EGFR, using ligand and structure based drug design approach via tools available in BioSolveIT software package. Although it is well known that a hit molecule must not only have high affinity against the drug target but also possess good ADME properties yet most drug discovery programs show high attrition rate during early phase of drug development. The reason being that most of the inhibitor identification programs concentrate on identifying highly active ligands but neglect other parameters which make these ligands unattractive for drug development. Therefore, it is important that drug discovery studies are not only based on affinity and potency alone but also aims at identifying molecules which have optimized ADME properties. Thus during the entire period of the Scientific Challenge our aim was to identify/design a few potential new ligands/hits which have not only high affinity against EGFR-TMLR mutant but also favorable ADME properties like low MW, LLE, LE, logP and TPSA. As hits that serve as starting points for lead discovery are invariably made available from either of the 3 resources i.e. (i) known drug like compounds i.e drug repurposing (ii) natural products and (iii) scientific literature; we had designed the computational experiments in a manner which allows identification of putative hits from each of these respective approaches enabling detection of various scaffolds that could later be optimized by experimentalist. The range of modules available in BioSolveIT software including (FlexX-Pharm, FTree, MONA, SeeSAR, HYDE and ADME prediction) along with integration of different approaches (ligand, structure, QSAR) have enabled us immensely to achieve the below mentioned goals.
The following goals have been achieved:
  1. Discovery of TMLR inhibitors by drug repurposing For discovering structurally diverse novel compounds from DrugBank we used BiosolveIT's FTrees, which searches for similar molecules against the query. We then used FlexX-Pharm to screen the constructed library followed by HYDE assessment in SeeSAR. It allowed us to identify a few low molecular weight scaffolds which occupy the critical sub-pockets and have desired interaction. Using one of the scaffold (MW= 295, LE +, LLE 0, logP 3.1) in SeeSAR, we then evolved new molecules having higher interaction affinity and desired ADME properties. The designed compounds have favorable LLE (+), LE (++), low MW (<400), logP (3.9-4.7) and TPSA (~70) and low nM predicted activity as indicated in SeeSAR. Also, we undertook molecular dynamics simulations to check stability of ligand in active site. Thus, we employed integrated ligand and structure based approach using BiosolveIT for identifying inhibitors from known drug molecules.
  2. Screening of natural product libraries for identification of TMLR inhibitors We have screened 12 NP-databases for their oral bioavailability and drug-likeliness using Lipinski and Ghoose filter available in MONA. We then predict the anticancer activity using a published QSAR model recently developed by me. We next examined interactions of these molecules with the EGFR-TMLR protein through guided molecular docking using FLexX-Pharm. The docking results were then imported in SeeSAR for HYDE assessment & visual analysis. We identified a few molecules that have low molecular weight and favorable LLE, LE, logP and TPSA as well as low nM predicted activity as indicated by HYDE affinity parameter. We also evaluated the stability of interactions using MD. Thus in the present work we have integrated MONA based physiochemical filtering, QSAR approach, FlexX-Pharm based molecular docking, SeeSAR hyde assessment and MD for identifying inhibitors against mutant TMLR protein from NP-libraries.
  3. Exploring EGFR targeted synthetic library for prioritizing TMLR inhibitors We have used literature curated synthetic inhibitors of EGFR taken from our published central repository EGFRIndb for virtual screening. We selected the library as it contains inhibitors that act against the wild isoform along with IC50 values, are sure about their synthetic feasibility and can be used as reference for understanding selectivity. We have screened these inhibitors using FLexX-Pharm followed by HYDE assessment in SeeSAR. Our initial result indicate 3 molecules have low nM binding affinity however, the molecules do not comply with all the desired ADME properties (MW, LLE, LE, logP and TPSA).

winter 2015

 
Identification of pharmacological chaperones for cellular prion protein
Maria Letizia Barreca
University of Perugia, Perugia, Italy
Maria Letizia Barreca
University of Perugia
Perugia, Italy

Neurodegenerative diseases

Identification of pharmacological chaperones for cellular prion protein
Maria Letizia summarizes:
Prion disorders are fatal neurodegenerative conditions caused by the conformational conversion of the normal, cellular prion protein (PrPC) into a misfolded isoform (PrPSc) that accumulates in the brain of affected individuals. A possible strategy for tackling prions is to stabilize the monomeric native conformation of PrPC, thus inhibiting its misfolding. This goal could be achieved with PrPC ligands capable of acting as pharmacological chaperones. Interestingly, the impact of such ligands would not be limited to prion diseases. In fact, recent evidence suggests that PrPC could transduce the toxic effects of a variety of misfolded proteins, such as oligomers of the amyloid β peptide (Aβ), and other disease-associated aggregates. Therefore, the identification of PrPC binders might possibly be relevant for several neurodegenerative conditions. The main goal of this project was to rationally identify PrPC ligands with the help of BioSolveIT softwares. Available conformations of human PrPC were analyzed by DoGSiteScorer for protein druggability assessment. The analysis suggested three PrPC protein conformations as the best targets for virtual screenings. At this point, we found really helpful the possibility to integrate BioSolveIT tools into KNIME. In fact, to assist the achievement of our goal we built an ad-hoc KNIME workflow for the identification of potential PrPC ligands. Two different libraries (fragments and in-house molecules) were thus screened against the three protein conformations, and 50 potential PrPC binders were selected for experimental evaluation. Next, we employed two complementary biophysical techniques (i.e. dynamic mass redistribution, DMR; and surface plasmon resonance, SPR) to characterize the binding of the virtual hits to PrPC. Five compounds were confirmed as PrPC ligands with affinity constant (Kd) ranging from 50 to 500 microM; of note, four of them belong to the same chemical family and showed high IP-value. The encouraging finding prompted us to focus our efforts on the identified chemical scaffold and to use the potentiality of CoLibri tool to design a focused set of new compounds structurally related to the hits. The chemical space exploration approach was integrated into the previously developed KNIME workflow and a pilot study was performed by using the in-house chemistry and our own building blocks. Among the in silico generated derivatives suggested by FlexX/HYDE as potentially able to bind PrPC with stronger affinity compared to the parent compounds, the four simplest structures were selected for chemical synthesis. The synthesized derivatives were evaluated for their binding to PrPC by DMR and SPR, and two improved PrPC ligands (MW around 380 Da) were identified and are now under extensive biological evaluation.
The following goals have been achieved:
  1. Development of an ad-hoc BioSolveIT/KNIME workflow to identify potential PrPC ligands. Fourteen protein conformations of human PrPC were retrieved from the RCSB PDB and processed by DogSiteScorer to assess protein druggability. Three PrPC conformations were thus selected as targets for structure-based virtual screening. We have then designed, built and configured an ad-hoc KNIME workflow, which included 1) Receptor preparation and virtual screening of two compound libraries using LeadIT; 2) Assessment of the ligand affinity with HYDE in SeeSAR; 3) Filtering out of compounds with low estimated affinity, unfavorable predicted value of LE and classified as “-“ in the BBB category. The results obtained for each target were concatenated in such a way to easily identify potential PrPC binders conserved across the three different conformations. Next, visual analysis and chemical structure clustering allowed us to identify 52 virtual hits for experimental validation.
  2. Identification of a new chemical class of PrPc ligands. In order to evaluate the binding of the in silico selected compounds to PrPC, we developed a novel screening method based on DMR, a label-free, fully automated biophysical technique capable of detecting molecular interactions at the equilibrium. Positive hits were then further characterized by SPR, a complementary biophysical technique. Five virtual hits were experimentally confirmed as PrPC binders by both techniques (Kd range: 50-500 microM). Notably, four of such hits were in-house developed compounds sharing a common chemical scaffold and possessing high IP-value.
  3. Successful pilot study for chemical space exploration. The binding affinity data prompted us to conduct a pilot study where a restricted chemical space around the identified PrPC ligands was designed and built. For this purpose, the in-house developed KNIME workflow was integrated with appropriate nodes (e.g. Reaction Library synthesizer). The chemical reactions were defined based on our chemistry experience, and in order to speed up the synthesis of possible virtual products, only reagents available in our labs were used as building blocks. Thousands of virtual products were generated and then submitted to virtual screening. HYDE predicted several derivatives as higher-affinity PrPC binders compared to the starting hits. In order to validate the approach, the four simplest structures were synthesized and experimentally evaluated. Encouragingly, two derivatives showed improved binding profile, and their extensive biological evaluation using assays for prion propagation is undergoing.