Research Study on Anticancer Drugs Targeting the Cdk2 and Efgr Proteins
Project location: ITALY, Modena
Project start date: September 2013 - Project end date: September 2015
Project number: 2013-025
Beneficiary: Università di Modena e Reggio Emilia
The goal of this project, which received a grant from the Nando Peretti Foundation, is to design and develop anticancer drugs targeting two important proteins, the CDK2 and EGFR kinases, by exploiting a promising and hitherto unexplored allosteric type III mechanism.
Virtual screening (VS) is a straightforward computational tool for the rapid and rational identification of bioactive compounds from large collections of chemical compounds. Typically, VS is able to virtually screen the potential interaction of a large number of molecules into the binding site of a target protein, and to rank them based on the strength of the interaction. The more the interaction is favorable, the more the molecule is potentially active as a drug. This technique is powerful and cost-effective, because it allows to rationally prioritize and select potential drug candidates for biological assays. For example, one can virtually screen hundreds of thousands (even millions) of compounds using a computer cluster and then select only the best 30-50 compounds for biological and pharmacological testing, thus enabling a significant reduction of laboratory costs, man power and resources. MMDDLab has recently developed its in-house software for VS, named "Binding Estimation After Refinement" (BEAR). BEAR has been set up, improved, and extensively validated in drug discovery applications. BEAR proved to be significantly more accurate compared to standard methods and has been able to discover several biologically active compounds. From 2007, we have published 13 peer-reviewed research articles describing BEAR validation and application.
As for the present proposal, MMDDLab has already performed a virtual screening of a huge database of compounds (more than 600.000 molecules) in the binding site of the CDK2 protein (PDB code 1PXF) , using BEAR.
As a result, six of the thirty-four best hit compounds experimentally tested by the group of Massimo Broggini at Istituto di Ricerche Farmacologiche Mario Negri were biologically active.
These molecules proved able to bind to the allosteric site with dose-dependent response, and were also evaluated for their ability to inhibit in vitro kinase activity (unpublished data). Importantly, the compounds were experimentally confirmed to be allosteric, in agreement with the design.
Once established that the molecules interact effectively - and with the expected mechanism - with the purified CDK2 protein, the six compounds were also tested in vitro on selected cancer cell lines in order to verify if they were able to inhibit the growth of cancer cells. This additional test is important to understand if the molecules are able to penetrate the cell membrane and block the iperproliferation of cancer cells. Importantly, preliminary data obtained with two human ovarian cancer cell lines (Ovcar5 and Igrov1) show that the compounds have cytotoxic activity in the micromolar range. Tests on other cancer cell lines are currently on going.
The preliminary data already in the MMDDLab hands confirm that its strategy is able to provide drug candidates with the desired mechanism, and strongly support the further prosecution and optimization of these and other newly identified molecules for their preclinical development.
Owing to their crucial role in the modulation of cell pathways, protein kinases are important targets for anticancer drug discovery. The classic approach of targeting the canonical ATP active site has recently come up against selectivity issues, which can be considerably reduced by following an allosteric (i.e. distinct from the ATP site) modulation approach. Moreover, drugs that target allosteric sites can help circumvent drug resistance, which is another major problem that severely impairs the use of the available anticancer drugs. Therefore, allosteric ligands are expected to have fewer side effects and higher potency on cancer cells resistant to commonly used drugs.
Type III inhibitors are expected to bind exclusively to allosteric pockets located in the proximity of the αC helix of the kinase.15 This project will focus on the rational design, chemical synthesis, and biological and pharmacological evaluation of type III allosteric ligands with anticancer activity that bind in proximity of helix aC of CDK2 and EGFR, two key kinase targets in cancer.16,17
CDK2 and EGFR share important structural and activation mechanism similarities. Therefore, it is possible to exploit these similarities to design the first in class type III allosteric ligands that inactivate these kinases through binding around helix aC. To MMDDLab knowledge, no drug with such a mechanism of action has yet been reported for CDK2 and EGFR on publications or patents. Therefore, the six compounds already discovered using virtual screening with BEAR constitute the first drug candidates with this mechanisms. This result further corroborates the utility and potential of BEAR in discovering biologically active compounds.
CDK2 plays an important role in the G1- to S-phase transition. To date, none of the ATP-competitive CDKs inhibitors have been approved for clinical use, mainly because of selectivity issues.16 The majority of the inhibitors tested so far were able to inhibit several CDKs with comparable IC50, but due to undesired toxicity, their use at effective doses was limited. This is further corroborated by the recent evidence that a relatively more selective (although not as selective as an allosteric inhibitor would be) CDK4/CDK6 inhibitor was reported to triple the Progression Free Survival (PFS) in combination with an aromatase inhibitor in breast cancer.18
EGFR has been strongly implicated in the biology of human epithelial malignancies, with therapeutic applications in cancers of the colon, head and neck, lung, and pancreas. In non small cell lung cancer (NSCLC) patients in particular, the prognosis is still very poor with a percentage of survivors that is lower than 15% for all stages and in metastatic patients show a 5-year survival rate lower than 5%.19 In a subset of NSCLC patients presenting mutations in the EGFR gene, two ATP competitive, small molecule reversible EGFR inhibitors (gefitinib and erlotinib) have significantly improved the outcome and these agents are now approved in first line treatment. Unfortunately, the antitumor activity of these classical EGFR inhibitors is severely limited by the development of drug resistance arising mainly from the development of a secondary mutation (T790M) which alters the conformation of the receptor and renders it insensitive to the inhibitors.20 The use of EGFR irreversible inhibitors (such as afatinib) did not significantly improve the overall survival.21 The availability of allosteric inhibitors could therefore significantly help in solving this critical issue.
Based on these evidences and background, in this projec, MMDDLab plans to:
1) perform hit-to-lead optimization of the six newly identified CDK2 allosteric hits, expand SARs, optimize potency and bioavailability, with the final aim of developing these compounds further in their preclinical evaluation.
2) design and optimize the first type III allosteric EGFR inhibitors using the same research strategy already validated for CDK2.
To accomplish these goals, several derivatives will be designed using computational methods, synthesized and/or purchased from commercial libraries, and tested in vitro on isolated proteins and in cancer cell lines. Target-guided synthesis will also be considered to assist the drug optimization and to probe the binding site. Promising compounds will be selected for further in vivo analyses.
MMDDLab goal is to design and optimize these allosteric ligands. MMDDLab plans to focus its attention on two largely diffused tumors, non-small cell lung cancer (NSCLC), for which EGFR kinase inhibitors have proved clinical activity, and triple negative breast cancer which represents one of the most aggressive breast cancer subtype for which the development of new therapeutic strategies is particularly needed. In the latter tumor very recent clinical evidences suggest that CDK inhibitors could impact on the Progression Free Survival (PFS) of these patients.18
MMDDLab goals are related with the Foundation's goals of improving the general welfare of people at both global and local levels and the Foundation's commitment in help relieving suffering of people. The two types of aggressive cancers the Team is tackling are still leading causes of cancer deaths. Currently, it remains a challenge to offer these cancer patients an effective therapy able to reach the goal of eliminating death and suffering. Moreover, the few available therapies are impaired by the emergence of drug resistance, which dramatically reduces the survival time of these patients. For these reasons, the development of drugs with an allosteric mechanism may offer a more effective, less toxic and more clinically amenable therapeutic strategy.
The project will be focused on the design, synthesis and biological/pharmacological evaluation of allosteric type III inhibitors of CDK2 and EGFR.
The project will be divided into three tasks, pursued during the two-years duration of the requested grant, that involve:
1) The molecular modelling and rational design of the allosteric inhibitors (Prof Rastelli, Molecular Modelling & Drug Design Lab, UNIMORE)
2) The chemical synthesis of the designed molecules (Prof. Passarella, UNIMI)
3) The biological and pharmacological evaluation of the designed candidates (Dott. Broggini, Istituto di Ricerche Farmacologiche Mario Negri)
Task1 - Molecular modelling and rational design of allosteric inhibitors
Task1a: Hit expansion and hit-2-lead optimization of the six CDK2 inhibitors
In this task MMDDLab plans to further optimize the properties of the six ligands already discovered for CDK2 in order to improve their potency, solubility, bioavailability, and cell penetration ability. An optimal balance between all these properties is important to obtain good quality drug candidates. To accomplish this goal, hit analogues of the six compounds will be searched into a larger database of commercial compounds (see Task1b) and then virtually screened with BEAR in order to select additional candidates. This operation will increase the number of potential hits without requiring chemical synthesis. In parallel, analogues with modified scaffolds, substituents, chemical groups will be designed and then synthesized by the UNIMI collaborators (Task 2). Predictions of solubility, bioavailability and cell penetration (ADMET properties) will be performed with QikProp (Schrödinger suite). The best compounds emerging from these analyses will be experimentally tested in Task 3.
Task1b: Preparation of a larger database of commercially-available compounds
The virtual screening already performed on CDK2 was done using the Asinex collection (~600.000 commercially-available compounds). MMDDLab plans to build a larger database starting from the publicly available "Big vendors" subset (~5 million compounds) of the ZINC database, which includes molecules from catalogues of well-known vendors such as Princeton, InterBioScreen, Enamine, Asinex, ChemDiv and others. With this database MMDDLab will be able to cover almost entirely the chemical space available from the main vendors, enlarging the possibility to find a number of molecules able to bind CDK2 and EGFR. Compounds with poor drug-like properties will be filtered out from the database, in order to exclude a priori compounds with unfavorable pharmacokinetic properties
Task1c: Design of type III allosteric ligands of EGFR
In this task MMDDLab plans to identify the first type III allosteric inhibitors of EGFR using a strategy similar to that already validated for CDK2. To this aim, the Team will build EGFR structures with an open allosteric pocket and perform virtual screenings (BEAR) and structure-based designs in this pocket.
The conformations of EGFR with an open CDK2-like allosteric pocket will be generated via homology modelling techniques (Modeller 9.11 software) and classical and enhanced sampling (e.g. umbrella sampling, steered MD etc.) molecular dynamics simulations (Amber12, Gromacs software). Particular attention in modelling structures will be given to the so-called "gatekeeper" resistant mutation (T790M), which is known to affect EGFR conformation and cause drug resistance. In this respect, developing allosteric EGFR inhibitors with high affinity for this resistant mutant is a highly desirable goal. Then, molecular dynamics simulations will be used to refine the resulting homology models and to evaluate their stability.
Extensive virtual screenings of the commercial compounds library prepared in Task 1b will then be performed using BEAR. The best ranked compounds resulting from the screenings will be biologically evaluated. When the first hits will be identified, MMDDLab will proceed with the optimization of their biological/pharmacological properties using the approach already described for CDK2.
Task2 - Chemical synthesis of the designed molecules
Task2a: Synthesis of analogues of the CDK2 allosteric inhibitors
Starting from the six compounds that bind CDK2, the Team will modify their structure and create different libraries of new analogues bearing different substituents, following the designs emerging from Task1a. The UNIMI collaborators are equipped with the appropriate facilities to complete under conventional and advanced methodologies the chemical synthesis of the designed derivatives with high levels of purity and structural characterization. The synthesis will be based on the use of the active compounds as building blocks for the introduction of different substituents or in the complete construction of the molecules in the light of the obtainment of convenient procedures to generate differentiated libraries of products. The six hit compounds (Figure 1) belong to 4 different classes that require different synthetic approaches. Class 1 (tetrahydroquinolines) could be obtained by condensation of anilines with different aldehydes and subsequent aza-Diels-Alder reaction. The addition of 2-nitrobenzenesulfenyl chloride secures the introduction of the thioether and the chloride. As for Class 2 (thienopyrimidinones), the synthesis is based on the reaction of substituted cyclohexanones with cyanoacetamide to generate tetrahydro-benzothiophenones. The final reaction with arylaldehydes permits the obtainment of the desired compounds. The preparation of Class 3 (dihydropyridines) is supposed to be easily accessible by Hantzsch-type cyclocondensation reaction with the possibility to differentiate the three building blocks. The Pd-catalyzed arylation reaction of amines will be useful for the preparation of Class 4 (adamantanamines).
Task2b: Synthesis of the EGFR allosteric inhibitors
The analysis of the structures of the compounds suggested by computational studies during the project will guide the planning of a convenient synthetic strategy that will secure the preparation of libraries of compounds as EGFR allosteric inhibitors.
Task2c: Target-guided synthesis of CDK2 and EGFR ligands
MMDDLab recently obtained encouraging results in the application of target-guided synthesis for finding new tubulin binders (exploiting the disulfide exchange reaction) and for probing the binding site of Abl tyrosine kinase (exploiting azide-alkine 3+2 cycloaddition). This type of approach, that consists in performing the chemical synthesis directly within the binding site of the target protein (CDK2 and EGFR), will be considered whenever one of the hit compounds presents a structure which can be divided in two frameworks connected by an appropriate functional group. In this regard, Classes 1, 2 and 4 of CDK2 ligands can already be considered appropriate for this approach.
Task3 - Biological and pharmacological evaluation of the designed drug candidates
Task 3a: Activity of CDK2 inhibitors
Regarding biological assays, CDK2 putative inhibitors will be initially tested for their ability to displace the extrinsic fluorophore 8-anilino-1-napthalene sulfonate (ANS) from CDK2 and subsequently for their ability to inhibit the kinase activity of CDK2 investigated in vitro using purified CDK2/CyclinA complex incubated with different concentrations of the selected compounds. The most potent compounds will be tested for their ability to exert cytotoxic activity against cancer cells. A panel of 5 different human breast cancer cell lines will be treated or not with five different escalating concentrations of the compounds for 72 hrs and then cell viability tested by MTS. The most active compounds (with IC50 in the sub-mM range) will be further investigated for their ability to cause cell cycle perturbation, modulate phosphorylation of downstream effectors as a read out marker for CDK2 inhibition and induce apoptosis. All these tests will be done treating exponentially growing cells with a fixed IC50 dose for 24 hrs and, at the end of the treatment and at different time point after removal of the drug (2, 6, 24 and 48hr), the cells will be fixed for cell cycle analysis and apoptosis, while cell extracts will be obtained to evaluate the phosphorylation status of proteins affected by the CDK2 inhibition. All these data will test if the selected compounds hit the target they were design against. Selected and promising compounds derived from previous in vitro experiments will be selected to determine their activity against human cancer cells growing in immunodeficient mice. For this purpose a triple negative human breast cancer cell line, MDA-MB231, will be initially used. These cells have been engineered to express luciferase, which allows the use of in vivo imaging analysis. The cancer cells will be implanted subcutaneously or orthotopically in the mammary fat pad and the growth of the tumors followed by tumor diameter measurement or by imaging analysis using the Explore Optix System. The Team has already experience in the use of this instrument which is available in its animal facility. The Mario Negri group has already evidence supporting the robustness of the CDK2-displacement assay for the detection of potential allosteric inhibitors.
Task 3b Activity of EGFR inhibitors
Potential allosteric EGFR inhibitors will be tested for their activity in inhibiting EGFR kinase activity in vitro and to inhibit the growth of human NSCLC cells expressing a wild type EGFR or mutated (hence activated) EGFR. The Team has already available a panel of cell lines with differential response to classical EGFR competitive inhibitors which will be used. In particular the Team plans to use H1299 and A549 as NSCLC cells expressing a wild type EGFR, NCI-H1650 expressing an activating mutation in exon 19 and NCI-H1723 expressing an activating mutation in exon 21 and concomitantly a secondary mutation T790M conferring resistance to erlotinib and gefitinib. Compounds showing activity independently from the presence of mutations in EGFR, and particularly in the cell line expressing the secondary T790M mutation, will be selected for further in vivo analysis. The human cell lines will be injected subcutaneously or intravenously (for lung colonisation) and the growth of the tumors followed as described in task 3a.
The Mario Negri group has already experience with these cell lines in terms of their growth in vitro and in vivo since they have been used for the determination of activity of different compounds. The compounds could also be tested in combination with erlotinib and gefitinib to see whether a synergistic activity of the compounds could be found. This would have anyway an important translation to the clinic as a way to increase the response to treatment.
The goal of this project is to discover potential anticancer drugs targeting CDK2 and EGFR via a novel type III allosteric inhibition mechanism and to optimize their chemical and pharmacological properties for preclinical development. The Team will focus its attention on two largely diffused tumors, non small cell lung cancer (NSCLC), for which EGFR kinase inhibitors have proved clinical activity, and triple negative breast cancer which represents one of the most aggressive breast cancer subtype for which the development of new therapeutic strategies is particularly needed.
Although many anticancer drugs have been used successfully for cancer treatment in humans, many cancers become resistant to these drugs over the course of chemotherapy. In this regard, allosteric modulators are designed to be more selective and can help overcome drug resistance issues. To date, increasing evidences suggest that inhibitors targeting the kinase allosterically could offer safety advantages over targeting the classical ATP site. By developing allosteric inhibitors that target less conserved binding sites, higher selectivity and thus fewer side effects can be obtained. Moreover, the availability of allosteric inhibitors for EGFR and CDK2 can potentially overcome the problem of drug resistance (due to mutations and changes in conformation of the receptor) that is clinically observed with classical ATP competitive inhibitors. In fact, drug resistance severely limit the efficacy of current drugs on cancer patients.
Anticancer drug candidates acting as allosteric ligands of EGFR and CDK2 have not yet been reported. Therefore, the success of this project will open important new ways for obtaining safer and more effective drugs targeting these important anticancer targets.
With the support of the Nando Peretti Foundation, at the end of the two years duration of the requested grant MMDDLab expects to deliver the first-in-class CDK2 and EGFR allosteric inhibitors with appreciable activity in cancer cell-based systems and in in vivo preclinical models of non small cell lung cancer and triple negative breast cancer.
Considering the relevance and timeliness of these goals and standing on the results obtained so far, these investigations have good chances of positioning MMDDLab collaborative team at the forefront of anticancer drug discovery on these kinase drug targets. This will likely attract the interest of additional collaborators that may further expand the network and include additional expertise. By doing so, the Team also aims at building an international collaborative network able to compete for EU funding in the next Horizon 2020 calls.
About this Project
1) The Molecular Modelling & Drug Design Laboratory (MMDDLab) at UNIMORE
The Molecular Modelling & Drug Design Laboratory (MMDDLab) at the University of Modena and Reggio Emilia (UNIMORE) has several years' experience in the structure-based design and development of potential drugs acting on enzyme targets of pharmaceutical interest. In the last years particular interest was focused on the design and development of potential anticancer drugs targeting protein kinases. Moreover, the lab recently developed BEAR, an innovative software tool for virtual screening that will be used for this project. The laboratory is fully equipped with a large number of computational software for structure-based drug design and hardware (linux clusters) for high performance computing.
Prof. Rastelli, associate professor in medicinal chemistry and head of the MMDDLab, is an internationally recognized and productive investigator with a 20-year experience in drug design and discovery using computational methods. He is the author of 81 peer-reviewed publications in international journals and inventor in 2 patents. His bibliometric indexes are H-index=22, total citations=1617.
He collaborates with several academic and private institutions for the design and discovery of targeted drugs. He received grants from EU, MIUR, AIRC, Regione Emilia Romagna, Spinner, bank foundations, private pharma and biotech companies.
2) The collaborators
The present research proposal is based on a tight collaboration between the research groups of Prof Rastelli (UNIMORE), Dr. Massimo Broggini (Istituto di Ricerche Farmacologiche Mario Negri, Milano, Italy) and Prof. Daniele Passarella (University of Milano, Italy). Given their clearly complementary areas of expertise (computational medicinal chemistry, biology and pharmacology, and chemical synthesis), the combined efforts of the three groups will have a very strong likelihood of achieving the objectives of this innovative and synergistic study.
The Mario Negri Institute in Milano is a recognized excellence center for pharmacological research. Dr Broggini is head of the Molecular Pharmacology Lab of the Mario Negri Institute. He has been involved for many years in evaluating the role of key regulatory proteins and determining cellular sensitivity to anticancer agents treatment. The laboratory has all the facilities and equipments necessary for the development of the project, including cell cultures facilities for the maintenance of cells, cellular and molecular biology facilities including gel electrophoresis systems, PCR thermal cyclers, real time PCR apparatus, fluorescent microscopy, flow cytometry facility. The laboratory has also access to the Institute animal facilities to perform the described in vivo experiments. In particular, the Molecular Pharmacology Lab at the Mario Negri Institute has available new instruments for the in vivo imaging analysis either as fluorescence or luminescence imaging which can be coupled with microTAC analysis. These new instruments will allow us to follow the growth of the tumors and the metastatic dissemination with a more precise quantification and with the possibility to reduce the number of animals needed for the experiments since the same animals can be followed throughout the entire experiment without the need of sacrifice them.
Prof Passarella group at UNIMI has been involved for many years in projects aimed at the chemical synthesis of organic compounds. The group has been involved in the preparation of different types of anticancer compounds, including the synthesis of tyrosine kinase inhibitors. His research lab is completely equipped with the instruments required for organic synthesis, purification and structural determination. The lab is excellently equipped with the necessary infrastructures and facilities to carry out advanced research in organic synthesis. The Scientific-Technical Services of the Department of Chemistry offer the researchers the latest high-tech instruments, including high field NMR and MS spectrometry. The presence of the inter-departments Analytical Center (CIGA) offers the use of the ESI High Resolution MS, MALDI TOF-TOF MS, surface plasmon resonance (SPR), X-Ray crystallography equipment.