May 15, 2014 (Vol. 34, No. 10)
Consistent with their central roles in cellular function and disease pathology, kinases are well represented in the human genome. In fact, more than 500 kinases have been identified. Kinases also reflect the complexity of signaling networks. It is well known, for example, that individual kinases can have stimulatory or inhibitory effects on multiple targets.
Many investigators are now querying this entire collection of kinases, or kinome, through large-scale genomic, proteomic, and chemical biological approaches. Through such analysis, investigators hope to better understand kinase function, identify new kinase inhibitors for cancer therapy, and screen candidates for side-effects on other pathways. Kinomic analysis is also well suited for understanding acquired drug resistance in cancer.
The vast amount of information that is now available from kinomic studies makes a global approach to evaluating kinase activity necessary in vitro and through functional cellular assays. Technologies also now permit the structural analysis of inhibitor-kinase complexes to present a clearer picture of inhibitor specificity.
Therapeutic Kinase Inhibitors
Kinases are required for cell cycle progression and cellular responses to growth factors. Given the regulatory role of kinases in cell growth and division, it is not surprising that mutations in kinase genes or misregulation of kinase expression occurs in many cancers. Indeed, nearly 75% of all human kinases are expressed in many breast cancer cell lines. Therefore, developing specific kinase inhibitors for the treatment of cancer and other diseases is an important goal for drug development.
Vichem Chemie, a Hungarian-German biotechnology company, specializes in the development of kinase inhibitors as therapeutics in collaboration with Axel Ullrich, Ph.D., director of molecular biology at the Max Planck Institute of Biochemistry. When Vichem screens for clinically relevant kinase inhibitors, it uses a knowledge-base approach, consulting its Nested Chemical Library, a collection of kinase inhibitory compounds of about 110 core structures and more than 600 scaffolds. The library encompasses both published and proprietary compounds.
According to Vichem’s CSO, György Kéri, Ph.D., D.Sc., the company’s knowledge-base approach is best suited for addressing the complex kinome changes in disease. “According to the Catalogue of Somatic Mutations in Cancer (COSMIC) database, there are 138 driver genes, of which 64 are oncogenic, and 74 are tumor suppressor genes,” notes Dr. Kéri. “Most of these driver genes are associated directly or indirectly with kinases of the 12 major cancer pathways.”
Relying on statistics generated by the BioSeeker Group, Dr. Kéri states that drug development efforts are focused on 224 different targets, 202 of which have been recorded with somatic mutations. Industry-wide, there are 563 protein kinase inhibitor drugs in development for 155 indications, of which 90 are cancer indications.
“At Vichem, we have developed reference compounds for 160 of these kinases, and we focus on compounds that can block the cancer driver pathways,” confides Dr. Kéri. “We try to develop multiple target compounds, which will kill cancer stem cells because these are the really bad guys, and they also happen to have more so-called survival factor kinases than other tumor cells.”
To facilitate the development of these inhibitors, Vichem and its collaborators culture tumor cells from cancer patients and identify mutations in the driver genes through genomic sequencing. According to Dr. Kéri, this approach “allows us to see if there are mutations in known oncogenes and tumor suppressor genes.” In addition, the approach makes it possible “to check for mutations in the downstream targets of these suppressors and to see if other pathways are activated in the patient.”
“Despite the heterogeneity of the patient samples, the current data indicate that we should expect at least two mutations in driver genes. but not more than eight,” explains Dr. Kéri. “This should give us a good chance for proof-of-concept studies for using our Nested Chemical Library to develop targeted inhibitors for kinases that are cancer drivers.”
Vichem is also developing a clonogenic assay to identify biomarkers for cancer stem cells. The company also intends to develop kinase inhibitors to specifically target and kill them. It wants to use these genomic or proteomic biomarkers for the survival cultures from the patient material to utilize so called “stemkill” compounds for killing the cancer stem cells from the patient.
Dr. Kéri believes that this approach—the use of the driver-hit combinations or stemkill compounds derived from phenotypical screening—offers the most promise. “I’ve been working on this for 30 years and can finally see light at the end of the tunnel,” he exclaims. “[The cancer driver and cancer stem cell approaches for developing targeted inhibitors] should provide a very dramatic and synergistic effect.”
Dr. Kéri also sees the development of kinase inhibitors as a community effort. Accordingly, he has made the Vichem resources available for various research collaborations.
Kinases assays are obviously important for validating the effectiveness of the kinase inhibitors that Vichem is developing. Dr. Kéri notes that the company uses a variety of approaches for kinase assays. The immobilized metal assay for phosphochemicals (IMAP) technology, the Transcreener kinase assay, and the fluorescence polarization assay are among the preferred assay systems.
Through partners, Vichem also uses binding assays for 400 kinases for selected inhibitors; high-throughput, homogeneous, time-resolved fluorescence (HTRF) assays for its allosteric kinase inhibitor library; and an assay that resolves fluorescence-resonance energy transfer over time (time-resolved FRET, or TR-FRET). Vichem outsources much of its in vivo work, including ADMET assays for drug development, but conducts in vitro kinases assay as well as various cellular assays in house. The company also uses classic Western Blotting and kits from R&D Systems in addition to functional assays.
“We need to determine the IC50 values for an inhibitor for a series of kinases,” Dr. Kéri states. “We also need to determine whether it is an ATP binding site inhibitor (Type I) or a non-ATP binding inhibitor (Type III or Type IV).” Vichem also measures the inhibitor’s residence time, which reflects the inhibitor’s effectiveness in vivo. To take such measurements, Vichem works with colleagues at Proteros biostructures in Martinsreid, Germany.
BRAF is a member of the RAS/RAF/MEK/ERK pathway, and it plays an essential role in the regulation of cell growth, differentiation, and survival. The most common mutation of the BRAF protein is V600E, which can be found in approximately 60% of melanomas. It is an important driver oncogene. In recent years, several new BRAF inhibitors have been reported. One of them, the pyrazolo[1,5-a]pyrimidine derivative, was described as a non-hinge binder type II BRAF inhibitor (PDB ID: 3II5). Here the compound is shown bound to the cleft between BRAF’s N- and C-lobes (ATP binding site): the hinge region is colored yellow; the aC helix, orange; and the activation loop (containing the DFG motif), green. The kinase is in the inactive (DFG-out) conformation. [Vichem Chemie]
Structure-Kinetic Relationships
Not all kinase inhibitors that are identified will have an optimal residence time for maximum physiological effectiveness. Thus, a better understanding of how the structural features of inhibitors contribute to their kinetic properties is essential for their efficacy as therapeutics.
Structure-kinetic relationships have received less attention that structure-activity relationships, notes Elisabeth Schneider, Ph.D., a Robert Huber post-doc stipendium fellow (Max-Planck-Institute of Biochemistry) at Proteros. Dr. Schneider and colleagues have worked to characterize the structural features that may switch the binding properties of compounds from fast binding to slow binding kinetics and extend residence times for the CycC/CDK8 complex, a potent oncogene that is involved in transcriptional activity and epigenetic processes.
Small molecule inhibitors, including BAY-43006 and imatinib, bind to the deep pockets of kinases (adjacent to the ATP binding site), and their residence time is extended if the DMG (Asp-Met-Gly) motif near the target kinase’s activation loop is maintained in an “out” conformation. Dr. Schneider and colleagues use a high-throughput binding assay to identify core fragments that could bind to the deep pocket of CycC/CDK8, then test a diverse series of functional groups to build out from these core fragments: a “back to front approach,” to identify which groups will favor slow binding kinetics and extended residence time for CycC/CDK8.
“We wanted to take a structural approach for optimizing the residence time of kinase inhibitor compounds,” Dr. Schneider says. “Since I had just published the crystal structure of the CycC/CDK8 complex, this was an opportunity for using the structure to develop a molecular understanding of binding kinetics.” In her approach, Dr. Schneider used a library of 4,000 compounds to select for slow binders.
In order to identify such compounds, Dr. Schneider and her colleagues used the Proteros Reporter Displacement Assay, where probes that emit a specific optical signal bind to the protein target and are displaced by the addition of competing compounds. This assay provides relevant biophysical data including Kd, kon, koff, and residence time.
According to Dr. Schneider, the advantage of this assay is its very high throughput. “You do not need very much protein when compared to other assays,” she notes. “This assay [also made it possible to] directly investigate the structure-kinetic relationship and address a key challenge—the compounds need to be efficient inhibitors in vivo. In addition, you need the right assay to identify promising candidates and screen out the false positives.”
Reading Kinase Assays
The actual output of kinase assays must, of course, be converted into something quantifiable and interpretable. Like all other high-throughput systems, kinase assays are performed in 96- or 384-well microplates, where signals produced by luminescent or fluorescent substrates are read and quantified by a machine to generate data.
One such machine is the CLARIOstar® from BMG Labtech. According to Carl Peters, Ph.D., a senior application scientist at the company, CLARIOstar is a multimode microplate reader for the detection of fluorescent signals. It can determine quantitative fluorescent intensity (FI), far red fluorescence polarization (FP), and TR-FRET. “CLARIOstar uses linear variable filter technology, it doesn’t just move across the spectrum, but defines the leading and trailing edge for variable wavelength and bandpass selection.”
The microplate reader was tested with the Transcreener HTS (high-throughput screening) platform from BellBrook Labs. Transcreener allows direct detection of ADP, AMP, GDP, or UDP. According to Dr. Peters, “The main advantage of direct detection of ADP is the flexibility to compare data for any kinase and any substrate, which is very important for anyone developing inhibitors for multiple specific targets. It also allows the detection of more native targets and not just peptide substrates.”
Stripping Down
One of the biggest challenges facing those who develop kinase inhibitors is that individual assays must be developed for each kinase to be tested for a panel of compounds. In order to create a standardized and reproducible panel of assays for comparison, Promega has developed the ADP-Glo™ Kinase Assay, a selectivity profiling systems that consists of a series of eight-well strips. Each strip contains eight enzymes and their corresponding substrates. These strips are standardized for optimal kinase activity for inhibitor profiling when used with ADP-Glo, a universal platform that measures kinase activity by quantifying the amount of ADP produced during the reaction.
The main advantage of these kinase profiling strips, according to Hicham Zegzouti, Ph.D., senior research scientist at Promega, is their easy of use. “By standardizing the kinase concentration in each well of the strip, the user simply dilutes the kinase and substrate stock solutions and carries on with kinase reaction—no optimization required.”
“Many of the current kinase profiling assays are available as a service from CROs,” adds Dr. Zegzouti. “By creating a simple kinase profiling technology, we are enabling scientists to do kinase profiling in-house in a more timely and cost-effective manner.”
Dr. Zegzouti explains that the kinases were selected for inclusion in these strip panels based on their importance in human biology and what researchers and drug developers are using: “To create a broad coverage of the kinome that also includes various liability targets, we looked at the literature and discussed the panel with our collaborators. We created a ‘one of a kind’ kinase profiling panel, as it has flexibility toward allowing the choice of the kinases to profile either inter- or intra-kinase families.
“The scientists can choose the general panel, which is a mixture of commonly profiled kinases to start with. Then, based on the selectivity requirement of the inhibitor, they can use the family-based panels to see if there is any promiscuity toward the members of a certain family.”
Dr. Zegzouti and his colleagues at Promega have tested eight of these strips representing six kinase families against eight kinase inhibitor compounds. They determined that their results are in good agreement with established in vitro radioactive assays.
As knowledge of the kinome and interactions between kinases and different pathways increases, there is more data (and data for more kinases) to consider than ever before when designing kinase assays to test the growing number of targets and inhibitory compounds. Fortunately, the development of new assay platforms and plate readers has kept pace, so that planning and performing kinase assays will not take time away from the science of designing kinase inhibitors.
Promega kinase profiling strips were used to create either single-dose or dose-response kinase inhibitor profiles for eight kinases at once. Both applications can be used to identify the extent of kinase inhibitor selectivity. In the tyrosine kinase strip (TK-3), tofacitinib inhibited only JAK3 kinase (Selectivity19) whereas PF-477736 inhibited various members of the same kinase subfamily (Promiscuity).
