Professor Sem received his B.S. degree in Chemistry from UW-Milwaukee in 1986 and his Ph.D. degree from UW-Madison in 1990 in the area of mechanistic enzymology. He did post-doctoral work under an American Cancer Society grant at McArdle laboratory for Cancer Research (Madison), followed by The Scripps Research Institute (La Jolla, CA) doing NMR/structural biology of nuclear hormone receptors. He then worked several years in biotechnology before leaving to start a company that used NMR/structural biology to accelerate the design of drug leads, addressing significant opportunities presented by the genomics and proteomics projects. He joined the faculty at Marquette in the Fall of 2002. He is currently a member of the Children's Environmental Health Care Center.
Biophysical characterization of the factors responsible for specificity in protein-ligand interactions, addressed in a proteome-wide manner. Also, development of chemical probes to facilitate such studies. Applications include understanding and minimizing the toxic side effects of drugs as well as environmental toxins like endocrine disruptors. Biophysical methods employed include enzymology, NMR (nuclear magnetic resonance) spectroscopy, fluorescence and chemo/bioinformatics.
Sequencing of the human and microbial genomes has led to a paradigm shift whereby questions can be asked of entire classes of proteins called gene families. Gene families related by common binding site features are called pharmacofamilies (Sem et al., J. Cellular Biochem., 2001). Questions asked across these pharmacofamilies pertaining to how protein-ligand interactions affect specificity as well as enzymatic function fall under the umbrella of Chemical Proteomics (Sem (2004) Expert Review of Proteomics, 1, 165). Our research focuses on the development of methods within Chemical Proteomics and Chemical Biology, with applications to mechanistic enzymology, drug design and toxicology.
Comparative Analysis of Binding Sites
At some level, a proteome can be viewed as the entire collection of protein binding sites, each associated with a different biochemical and biological function. In the human proteome, there are over 29,000 such binding sites. Now that the genome and proteome is defined, one can attempt to understand and even predict the fate of organic molecules such as drugs, nutrients, toxins and xenobiotics, in the context of this milieu of binding sites. In most cases involving drugs, binding to proteins beyond the desired target results in: (1) metabolism of the drug (as with the cytochromes P450) or (2) undesired biological side effects (as with binding to isoforms of the desired target, or other off-target proteins). A better comparative understanding of binding site space will allow for the definition of the factors that define specificity. Our research focuses on understanding the interactions of organic molecules with protein anti-targets.
Protein Families being Studied
Proteins of current interest are: (a) estrogen receptors, (b) cytochromes P450, (c) kinases, and (d) dehydrogenases. Our interest in kinases and dehydrogenases is focused especially on enzymes that are infectious disease drug targets, such as Mycobacterium tuberculosis. Our interest in estrogen receptor is both as a drug target (breast cancer) and as an anti-target for endocrine disruptor pollutants. Of particular interest to us is the estrogen receptor from zebrafish – both as a means of developing zebrafish as a model system, and to study the effects of endocrine disruptor pollutants on fish. Our lab also has a strong interest in characterizing the cyochromes P450, both to understand mechanism of substrate recognition, and to better predict which drugs and xenobiotics act as substrates or inhibitors. From a basic research perspective, we are interested in the role of conformational changes and binding site dynamics in substrate recognition and catalysis.
Chemical Biological Probes
Many of our in vitro studies are enabled by chemical probes that are being developed to study protein-ligand interactions. These tend to be either NMR or fluorescence-based probes. Also of interest are probes of oxidative stress inside living cells – since such changes are associated with various pathological states, including bacterial infection.
Chemoinformatics: Computational Model and Chemical Databases
Studies of protein-ligand interactions will ultimately yield databases of structures and affinities for protein-ligand complexes. Data are being obtained using high throughput fluorescence screens, and complementary NMR-based studies at 600 MHz (cryoprobe) and 400 MHz (microflow probe). Such data will be used to build QSAR models to better predict binding to the protein targets, and will be summarized in databases to assist Medicinal Chemists in drug discovery efforts. To this end, optimized building blocks for drugs are also being engineered.
Our lab is involved in developing and applying methods for the study and manipulation of the specificity of interactions between proteins and their ligands. Methods are applied initially to a small subset of proteins, and later in the context of an entire organism (imaging) such as zebrafish (Costache et al. (2005) J. Molecular Endocrinology, 19, 2979), enabling proteomic analyses. Such studies provide a better understanding of the role of specific proteins in biology and the toxic side-effects of drugs and environmental toxins. They also provide a means for designing more selective drugs with minimal side-effects, and even tailoring drug treatments to the unique genetic makeup of individuals according to the promise of pharmacogenomics and personalized medicine. In terms of basic research goals, we are striving to understand the role of protein dynamics and conformational changes in substrate recognition.
Chemical Proteomics Facility at Marquette
An important focus of our lab is the development and management of the Chemical Proteomics Facility at Marquette (CPFM), with associated tools (see www.marquette.edu/cpfm ). The facility has a mission of enabling the structural characterization of protein-ligand interactions in a broad (proteome-wide) manner. The purpose of such studies is to extract maximum value from recently available genomic information. To accomplish this goal, NMR, computational and related resources are being made available.
The facility currently has chemoinformatic computational resources, as well as capability for screening protein ligand-interactions using both fluorescence and NMR (600 MHz + cryoprobe, 400 MHz, 300 MHz). NMR screening is enabled by automation and microfluidics. The 400 MHz spectrometer has microimaging capability.
Computational tools include those for structure calculation, homology modeling, T1-based docking, ab initio docking into NMR-generated structures, etc. Computational resources for such studies include Sun and SGI workstations and access to a Beowolf cluster. On-site software is available, such as Pipeline Pilot for Chemoinformatics, DOCK and Autodock for docking, MODELLER for homology modeling, DYANA for structure calculation, nmrPipe and XEASY for processing NMR spectra, as well as GRASP, VMD and MOLMOL for visualization. In addition to computational resources, any tools that may be of use in addressing the broader goal of characterizing protein-ligand interactions across protein families will be made available in the facility. Since the development of methods, reagents and databases is a current focus of my lab, the CPFM is serving as a venue for dissemination of these new tools. For example, glassware and laser equipment needed to hyperpolarize 129Xenon (used in studying protein-ligand interactions) will be made available through the CPFM. Under constant development are proteome-wide activity assays, NMR reagents for studying mechanism across protein families and drug metabolism/toxicology databases.
Thai Research Connections
As a participant in the Interfaces of Analytical Sciences Workshop we have collaborations with Thai scientists at Kasetsart University directed towards drug discovery, targeting Mycobacterium tuberculosis. A goal is to jointly screen Thai natural products, as part of an initiative being led by the Thai Research Fund.
Selected Publications (since 2001)
REFEREED RESEARCH PUBLICATIONS:
Pullela, P.K., Chiku, T., Carvan, M.J., III and Sem, D.S. (2006) Fluorescence-Based Detection of Thiols in vitro and in vivo using Dithiol Probes Analytical Biochemistry, in press.
Costache, A.D. Pullela, P.K., Kasha, P., Tomasiewicz, H. and Sem, D.S. (2005) Homology modeled ligand-binding domains of zebrafish estrogen receptors a, b1 and b2: from in silico to in vivo studies of estrogen interactions in Danio rerio as a model system. J. Molecular Endocrinology, 19, 2979-2990.
Yao, H. and Sem, D.S. (2005) Cofactor fingerprinting with STD NMR to characterize proteins of unknown function: identification of a rare cCMP cofactor preference. FEBS Letters 579, 661.
Pullela, P.K. and Sem, D.S. (2005) (chapter in Separation Methods in Proteomics) NMR-Driven Chemical Proteomics: the functional and mechanistic complement to proteomics. CRC Press, Boca Raton, FL, pp. 467-487.
Sem, D.S. (2004) Chemical Proteomics from an NMR Spectroscopy perspective. Expert Review of Proteomics, 1, 165-178.
Yao, H., Costache, A.D. and Sem, D.S. (2004) Chemical proteomic tool for ligand mapping of CYP antitargets: an NMR-compatible 3D QSAR descriptor in the heme-based coordinate system. J. Chem. Inform. Comp. Sci. 44, 1456.
Sem, D.S., Bertolaet, B., Baker, B., Chang, E., Costache, A., Coutts, S., Dong, Q., Hansen, M., Hong, V., Huang, X., Jack, R.M., Kho, R., Lang, H., Meininger, D., Pellecchia, M., Pierre, F., Villar, H. and Yu, L. (2004) Systems-based design of bi-ligand inhibitors of oxidoreductases: filling the chemical proteomic toolbox. Chemistry and Biology 11, 185.
Pellecchia, M., Huang, X., Meininger, D. and Sem, D.S. (2003) NMR-based drug design: applications to very large proteins. In "BioNMR in drug discovery" Zerbe O. Ed.
Kho, R, Baker, BL, Newman, JV, Jack, RM, Sem, DS, Villar, HO and Hansen, MR (2003) A Path from Primary Protein Sequence to Ligand Recognition Proteins: Structure, Function, and Genetics 50, 589-599.
Pellecchia, M., Sem, D.S. and Wuthrich, K. (2002) Nuclear Magnetic Resonance in Drug Discovery. Nature Reviews Drug Discovery. 1, 211-219.
Plesniak, L, Horiuchi, Y, Sem, DS, Meininger, D, Stiles, L, Shaffer, J, Jennings, PA and Adams, JA (2002) Probing the Nucleotide Binding Domain of the Osmoregulator EnvZ Using Fluorescent Nucleotide Derivatives Biochemistry 41, 13876-13882.
Pellecchia, M., Meininger, D., Dong, Q, Chang, E, Jack, R and Sem, D.S. (2002) NMR-based Structural Characterization of Large Protein-Ligand Complexes J Biom. NMR 22, 165.
Sem, D.S. and Pellecchia, M. (2001) NMR in the Acceleration of Drug Discovery. Current Opinion in Drug Discovery and Development 4, 479.
Pellecchia, M, Meininger, D., Shen, A.L., Jack, R., Kasper, C.B. and Sem, D.S. (2001) SEA-TROSY (Solvent Exposed Amides with TROSY): A Method to Resolve the Problem of Spectral Overlap in Very Large Proteins. J. Am. Chem. Soc. 123, 4633.
Jack, R., Sem, D.S., Yu, L. (2001) Real-time Structure-Based Drug Development. Genetic and Engineering News. 21, 34.
Sem, D.S., Yu, L., Coutts, S.M. and Jack, R. (2001) Object-oriented Approach to Drug Design Enabled by NMR SOLVE: First Real-Time Structural Tool for Characterizing Protein-Ligand Interactions. J. Cellular Biochemistry. Suppl. 37, 99-105.