A Part of the University of Maryland Medical Center

Connect with UMGCC
Facebook Twitter YouTube Blog iPhone
Email PageEmail page Print PagePrint page

Research Programs

High Throughput Screening (HTS) Shared Service: Research Foci

David J. Weber, Ph.D. (MSB), has been utilizing the HTS SS to discover inhibitors of S100 proteins in the hope of restoring normal p53 function in tumor cells (CA107331). Elevated S100B protein is a known prognostic marker for malignant melanoma [1, 2], however, the Weber laboratory and others have shown that S100B binds to the p53 tumor suppressor protein in a calcium-dependent manner [3, 4], dissociates the p53 tetramer [3], and down-regulates p53 protein levels and function [3-6] in cells. Therefore, the goal is to inhibit the S100B-p53 interaction to restore wild-type p53 protein levels in these cancer cells. As a proof in principle, it has been demonstrated that inhibiting S100B production (via siRNAS100B) restores wild-type p53 protein levels and activity in primary malignant melanoma [4]. Based on this, Dr. Weber in collaboration with Alexander MacKerell, PhD (ET) and Andrew Coop, PhD (ET) is undertaking a multifaceted approach to finding low molecular weight compounds to inhibit the S100B-p53 interaction, and restore normal tumor suppression function in cancers with elevated S100B and wild type p53 such as malignant melanoma, astrocytomas, renal tumors [7], and some forms of leukemia [8].

Currently, three major HTS assay have been utilized in these studies to find S100B inhibitor: (1) a fluorescent polarization competition assay (FPCA), (2) NMR based screens, and (3) cellular assays. While screening of the HTS's compound libraries is being performed, computer aided drug design (CADD) performed by Dr. MacKerell in the UMB, School of Pharmacy, CADD Center is also being performed in an attempt to increase the chances of finding a candidate inhibitor and decrease the number of compounds that needs to be screened. However, it was still necessary to screen these compounds using the same assays.

  1. Fluorescent polarization competition assay (FPCA) was developed by the HTS manager, Dr. Paul Wilder, utilizing the BMG POLARstar microplate reader to monitor the binding of a fluorescently labeled peptide to S100B via changes in fluorescence polarization after addition of small compounds (figure 9.1.1-2). This type of in vitro screen can be done in the HTS SS on regular bases, and we wish to increase to scope by screening the larger library available to the HTS SS. This type of assay is quick, relatively inexpensive to run, and has produced several candidate inhibitors [9]; however, it does have its limitations. One issue with the use of a small peptide is it defines a small area of the protein-protein interface (PPI), whereas most PPI are formed by large surfaces area forming multiple weak interactions. We therefore may be missing potential inhibitor that may bind outside the area defined by the peptide, but by using NMR based screens that will measure binding throughout the entire protein at least one other distinct site on the surface of S100B has been identified.

  2. NMR based screens have the advantage to be able to detect weaker binding compounds and they will show any interaction with a protein regardless of where it binds. Another goal in this study is to increase the affinity of compounds that bind S100B weakly but in different regions by linking them together chemically thus getting a significant increase in affinity and specificity. NMR based screening is perfect for this application because of its ability to measure binding anywhere on the protein, and it can detect interactions as weak a millimolar affinities. The HTS SS manager, Dr. Wilder, has been utilizing the Bruker Automatic Sample Changer (B-ACS 60, Bruker Inc.) equipped Bruker 800-MHz NMR in collaboration with Dr. Kristen Varney at the NMR Center at UMB to screens pools of 4 compounds to increase the throughput, 125 uM each with 100 uM S100B per tube, using HSQC NMR experiments (figure 9.1.1-3). The binding of ligands to a protein will cause a change in the chemical shift of nearby 1H or 15N labeled amino acid residues allowing the binding to be monitored and mapped onto the protein of interest. Further HSQC titrations and saturation transfer difference spectra (STDs) experiments run on the NMR can be done to better map the S100B-inhibitor interface and provide thermodynamic binding data. The HTS SS manger due to close collaboration with the NMR Center and technical expertise in NMR, will be able to provide this type of screen to all UMGCC members. In addition, NMR of small compounds is performed to check the integrity and insure the identity of compounds.

  3. Cellular based assays are also being run to determine the cellular activity and availability of the compounds to complement the computational and biophysical approaches. A cell proliferation assay using the HTS SS Beckman Coulter Biomek FX and the BMG FLOUstar microplate reader was run by Dr. Wilder. Briefly, two melanoma cell lines with high levels of S100B were used, C8146A with wild type p53, and SK-MEL-28 with mutant inactive p53, are grown in a 384-well plate with enough cells so they will reach ~80% confluence in 4 days. After 12-24 hours to allow the cells to adhere to the plate, compounds are added at 6.25 µM or an equivalent amount of DMSO is added to the wells as a negative control. The cells are allowed to grow for 3 more days when lysis/detection solution containing 1.2% Igepal CA-630 (Sigma) and a 1:1000 dilution of SYBR Green I nucleic acid stain (Invitrogen, Rockville, MD) are added to each well. After an overnight incubation, the fluorescence intensity is determined as the DNA content directly correlates with cell number. Since the hypothesis is that S100B acts on p53, the ratio of the percent C8146A cells killed to the SK-MEL-28 cells killed was determined to find compounds that selectively kill the C8146A cells. Further secondary titrations were performed on these compounds, and any that significantly kill either cell line, on a panel of melanoma cell lines (C8146A, Malme-3M, WM115, and A2058) and a non-tumorgenic melanocyte cell line (Malme-3) to screen out false positives and determine the LD50.

William Morgan, Ph.D., began the search for novel radiation sensitizers and protectors as a member of the UMGCC (CA124635) before moving to the PNNL but the work continues at the HTS SS as a subcontract. All aspects of this multi-year project are being performed by the HTS SS staff including maintenance of the cell lines, liquid handling steps using the Biomek FX Laboratory Automation, data collection using the BMG FluoroStar Microplate Reader, and the data analysis. The goal of this project is to identify new radiation sensitizers and protectors for use in clinical radiation oncology and in the event of radiological terrorism. Given the number of patients undergoing radiation therapy, identifying and characterizing new compounds that radio-sensitize tumors and/or protect normal tissues would have a major clinical impact as there are very few FDA approved drugs available for clinical use. Outside the radiation therapy clinic there is also significant relevance to identifying and characterizing novel compounds that protect cells from radiation induced cell killing. Public concern over the potential for nuclear terrorism or a radiological dispersal device, e.g., a dirty bomb has spurred new interest in developing radio-protectors.

The cell based assay has been performed on MCF-10A cells, a non-tumorigenic, genetically stable, near diploid, mammary epithelial cell line [10], and well-characterized isogenic derivatives of MCF-10A including p53 full deletion, p21 full deletion, PTEN full deletion, BRCA1 knock-in of mutant alleles, MLH1 knock-in of mutant alleles, PIK3CA knock-in of mutant allele, and K-RAS knock-in of mutant allele. Using the Beckman Coulter Biomek FX for all liquid handling, on day 0, 20µl of medium containing ~300 MCF-10A cells is plated per well of a 384-well tissue culture plate. These are incubated overnight at 37 degrees Celsius with 5% CO2 and 90% humidity. Twenty-four hours post-seeding (day 1), compounds are added. As a result, each 384-well plate contains 64 control wells and 160 unique compounds. Triplicate 384-well plates are prepared with one plate serving as the non-irradiated control to assess potential drug cytotoxicity. The second plate is exposed to 2Gy of x-rays to investigate a clinically relevant radiation dose primarily to select for radio-sensitizers. The third plate is exposed to 6Gy of x-rays, a higher radiation dose primarily to select for radio-protectors. Plates were irradiated using a Pantak HF320 x-ray machine (250-kV peak, 13mA; half-value layer 1.65mm copper), at a dose rate of 2Gy/min. Following irradiation one set of plates will have the medium removed and replaced with fresh medium. The rationale for this is to investigate whether a given compound offers sensitization protection at concentrations where continuous exposure would be cytotoxic, as is the case with WR-1065. In a second set of plates the media will not be removed but cells will be cultured with the test compound for the duration of the experiment. In this way we can more realistically mimic the human exposure situation. All plates will be incubated for 4 days (days 1-4), a time at which the control cells (0.5 % DMSO only) had previously been determined to reach ~70-80% confluency. On day 4, 40µl of lysis/detection solution containing 1.2% Igepal CA-630 (Sigma) and a 1:10000 dilution of SYBR Green I nucleic acid stain (Invitrogen, Rockville, MD) are added to each well. Following an overnight incubation at 37 degrees Celsius, total fluorescence is measured using the FLOUstar fluorescence plate reader (BMG, LabTech). By comparing fluorescence intensity in control versus drug +/- irradiation measured the percentage of radiation sensitization/protection can be determined.

Chen-Yong Lin, Ph.D. (HRC), is performing an ELISA based screen for inhibitors of matriptase activation (CA104944). Surface proteolysis plays important roles in cancer development and progression. Proteolytic activities have been considered to be promising targets for drug development. However, major clinical trials of protease inhibitors, such as matrix metalloprotease (MMP) inhibitors, showed limited success [1]. The fruitless results were in part attributed to the selectivity of protease catalytic inhibitors for the desired protease target, due to the high structural homology and overlapping specificity among protease catalytic domains and the complexity of regulatory mechanisms for proteases activation and inhibition, particularly for those proteases which are tightly regulated by their endogenous inhibitors. Matriptase and its endogenous inhibitor HAI-1 may represent such a complicated protease-protease inhibitor system in which the classic, protease catalytic inhibitors may not be effective as therapeutic agents [11]. Therefore, Dr. Lin hypothesized that prevention of zymogen activation, instead of inhibition of matriptase catalytic activity, will be both a novel and an effective way to therapeutically intervene in carcinoma. This hypothesis is based on the following observation. First, matriptase and its endogenous inhibitor HAI-1 are commonly deregulated and ubiquitously expressed in a variety of human carcinomas [12-20]. Second, matriptase possess strong oncogenic and prometastatic activity, when deregulated [21-23]. Third, despite high levels of zymogen activation, functionally active matriptase is scarce due to the highly efficient inhibition of active matriptase by HAI-1 immediately after matriptase activation [24, 25]. Based on these observations, they set out to develop a high throughput assay to screen against compound libraries for small molecule inhibitors targeting zymogen activation of matriptase.

In previous studies, Dr. Lin's lab has developed several model systems in which matriptase activation can be induced in intact cells. Among these model systems, exposure of matriptase-expressing cells to mildly acidic milieu is the most efficient way to induce rapid and robust matriptase activation (figure 9.1.1-5). Like in other model systems, activation of matriptase is immediately followed by HAI-1-mediated inhibition and active matriptase is rapidly sequestered into functionally inactive HAI-1 complex. In order to monitor the proceeding of matriptase activation, they developed and generated three monoclonal antibodies (mAbs) directly against total matriptase (mAb M32), HAI-1 (mAb M19), and activated matriptase (mAb M69). All these three mAbs have been used to detect matriptase-HAI-1 complexes, the end products of matriptase activation by immunoblotting analyses. Among these three mAbs, the activated matriptase mAb M69 is particularly useful for the development of high throughput assay for monitoring the activation of matriptase in intact cells. They have demonstrated that the M69 mAb is able to detect activated matriptase in intact cells by immunofluorescent staining (figure 9.1.1-5). Based on these successful model systems and valuable reagents, they have miniaturized our matriptase activation system into 96-well plate. Two assay systems will be developed: (1) a traditional ELISA-based system by using mAb M69, HRP-labeled 2nd antibody, and chromogen substrates, and (2) a dual fluorescent assay system by using fluorescent dye-labeled mAb M32 (for total matriptase) and M69 (for activated matriptase). The latter will allow simultaneous measurement of both total matriptase and activated matriptase, and the ratio of activated matriptase relative to total matriptase may be a better readout for the activation of matriptase.

They have completed the development of the traditional ELISA-based assay systems, and have been conduct a pilot screening of a set of compounds available from the HTS Center at University of Maryland, Baltimore to validate the assay. To date we have screened 12,160 compounds from the ChemDiv 40,000 compound library as pools or 4 compounds per well, finding two active pools that were tested individually. In this case, the screen had been performed by a postdoctoral fellow who was trained by the HTS manager to use the equipment.

Iris Lindberg, Ph.D. (HRC), has begun a HTS project to develop potent and specific furin inhibitors through screening of chemical libraries. The proprotein convertase (PC) furin is emerging as an important player in the regulation of carcinogenesis, angiogenesis and metastasis. Its physiological role is to activate a variety of precursor proteins carboxy-terminal to specific basic residue-containing motifs; depending on the precursor, it can act within the Golgi complex, the cell surface, or the endosomal system [26]. Substrate precursors include growth and differentiation factors, receptors, adhesion molecules and enzymes such as matrix metalloproteases (MMPs). Since these factors play important roles at different stages of tumor development, progression and metastasis, it is not surprising that tumor aggressiveness has been found to correlate with increased furin expression [27]. Furthermore, inhibition, knock-down and genetic ablation of furin reduce tumorigenesis in both cell lines and in mouse models [28, 29]. The main objectives of the project are (1) to develop potent and specific furin inhibitors through screening of chemical libraries; and (2) to evaluate these furin inhibitors in cellular migration and invasion assays. Accomplishing these goals will validate furin as a drug target in cancer, will facilitate the development of efficacious new compounds, and will begin to define the therapeutic potential of novel furin inhibitors. When completed the screen will be done on the complete Maybridge HitFinder (14,400 compounds) and the ChemDiv collection (40,000 compounds) consisting of over 56,000 organic compounds available at the HTS SS will be screened utilizing the automated liquid handling system and fluorescent microplate reader at the HTS SS. The usual fluorogenic furin assay will be used, in which AMC is cleaved from pERTKR-amc and measured at 380/460 excitation/emission; the assay is currently being performed in a 96 well format but it will be adapted to 384-well by the HTS center. All new inhibitors identified will be enzymatically characterized (for mode of inhibition) and also tested for cross-reactivity with other PCs (we have PC1, PC2, and PACE on hand), as well as with members of other protease families with related substrate specificity, such as trypsin and thrombin.

Future Plans

The HTS SS plans on applying for an shared instrumentation grant to purchase a state of the art microplate reader, BMG PHERAstar, that will allow increase use of 384 and even 1536-well plates, provide a stacker for auto-loading assay plates, and on board injectors for time sensitive application such as chemiluminescence and enzyme kinetic measurement assay. It is equipped with both a high energy xenon flashlamp and a solid state laser allowing for AlphaScreen assays, in addition to fluorescence intensity, fluorescence polarization, luminescence, time-resolved fluorescence, and UV/Vis absorbance. This would increase throughput, decrease sample size requirement, and provide a wider range of assays to be performed.

Plans also include increasing the HTS SS's ability to store, analyze, and organize the large amount of data collected during HTS screens. The purchase of a RAID server to store and backup data properly is a simple but necessary step in protecting data. Currently, the data collected for each compounds in an assay is not directly linked, most of the data handling is done in Microsoft Excel, statistics and curve fitting done manually in Origin (OriginLab, Northampton, MA), and the compound inventory control is done using ChemFinder (CambridgeSoft, Cambridge, MA). The makers of ChemFinder offer a suite of products that work together, ChemBioOffice (CambridgeSoft, Cambridge, MA), that would allow linking of the data, better organization, and automated analysis, and report building that would benefit both the HTS staff and investigators using the SS. More importantly, the HTS SS would like to seek collaborations with the Biostatistics Shared Service to bring in expertise already on campus to help users with data-mining and statistical analysis of their screen.

For More Information

For more information about the research behind the High Throughput Screening (HTS) Shared Service at UMGCC, please review this list of references.

Click here to return to the High Throughput Screening (HTS) Shared Service main page.


This page was last updated on: June 14, 2010.