Areas of Focus


One of the major areas of focus at the UPDDI is the identification of new anti-cancer agents and development of tools to help understand mechanisms of cancer at the molecular and cellular level. By combining the expertise at UPDDI in high content analysis, phenotypic, small organism assay development, and high throughput screening with the expertise of renowned scientists in cancer biology and computational sciences we are able to advance new approaches to drug discovery, facilitate Quantitative Systems Pharmacology, and identify novel therapeutic agents. We have established internal collaborations within the School of Medicine, School of Pharmacy, the University of Pittsburgh Cancer Institute including the Chemical Biology Facility, the University of Pittsburgh Medical Center (UPMC), and the Dietrich School of Arts and Sciences (Departments of Chemistry and Biological Sciences). The Pittsburgh Specialized Application Center (PSAC) is a program within UPDDI led by principal investigator Lans Taylor and together with the University of Pittsburgh Chemical Diversity Center (UPCDC) led by principal investigator Donna Huryn, has partnered with the National Cancer Institute Experimental Therapeutics (NExT) program on two external collaborative projects. Below are some of the projects, as examples, in which UPDDI is engaged in understanding cancer and identifying novel therapeutic agents and approaches to discovery and development.


Discovery and optimization of inhibitors of STAT3 activation for the treatment of squamous cell carcinoma of the head and neck

Project collaborators: Jennifer Grandis (PI, Dept. Otolaryngology), PSAC, UPCDC, Paul Johnston (lead discovery scientist, Dept. Pharmaceutical Sciences, Co-PI of PSAC), Sean Xie (Co-PI and Director of Cheminformatics, Dept. Pharmaceutical Sciences), NCI Division of Cancer Treatment and Diagnosis (DCTD), and SAIC-Frederick (SAIC-F).

The University of Pittsburg Specialized Applications Center (PSAC) and Chemical Diversity Center (UPCDC) are members of the NCI Experimental Therapeutics Program’s (NExT’s) Chemical Biology Consortium and are partnering with the NCI to advance novel therapeutics against STAT3.

STAT3 is one of seven members of the signal transducer and activator of transcription (STAT) family of proteins whose function is to relay signals from the cell surface receptors to the nucleus and initiate transcription. STAT3 responsive genes include those that regulate cell differentiation, proliferation, apoptosis, angiogenesis, metastasis, and immune responses [1,2], and there is compelling evidence indicating that dysregulated STAT3 signaling plays a central role in a variety of human tumors [1,3]. STAT1 has significant sequence and functional similarity to STAT3, however, whereas STAT3 is oncogenic, STAT1 is associated with tumor suppression. Selective inhibition of STAT3 function has potential therapeutic benefit, however, finding selective STAT3 inhibitors using target-based approaches has been challenging, and to date no STAT3 selective inhibitors have been identified [1].

The goal of the STAT3 NExT project is to identify small molecules that selectively inhibit STAT3 over STAT1 signaling and that can be developed into clinical compounds for the treatment of squamous cell carcinoma of the head and neck. The PSAC team has developed a high content, phenotypic screen that quantifies STAT3 and STAT1 translocation to the nucleus and is a direct measure of STAT signaling in the cell. In addition to the direct cellular readout of STAT signaling, the advantages of this approach over target-based approaches are that selectivity of STAT3 versus STAT1 inhibition is measured in the same cell system and that inhibitors identified are cell permeable and active in cells. Further, the phenotypic approach has the potential to uncover inhibitors of STAT3 signaling with novel mechanisms of action (MOA), as well as STAT3-targeting compounds. A number of hits that selectively inhibit STAT3 translocation have been shown to inhibit STAT3 regulated gene expression and cell growth of a variety of HNSCC lines. To further elucidate MOA of lead compounds and to help drive SAR, NCI DTP DCTD and NCI/SAIC-F laboratories are undertaking additional bioassays, including STAT3 binding and reporter assays. In vitro ADMET properties of selected leads will also be characterized by NCI DTP DCTD. Several novel lead series are currently being explored by the UPCDC team using both traditional medicinal chemistry and computational chemistry methods for optimization.


1 Johnston PA, Grandis JR. STAT3 Signaling: Anticancer strategies and challenges. Molecular Interventions, 2011;11:18-26.

2 Leeman RJ, Lui VWY, and Grandis JR. STAT3 as a therapeutic target in head and neck cancer. Expert Opin. Biol. Ther. 2006;6:231-241.

3 Siddiquee, KAZ, Turkson, J. STAT3 as a target for inducing apoptosis in solid and hematological tumors. Cell Research. 2008;18:254-267.

This project has been funded in whole or in part with Federal Funds from the National Cancer Institute, National Institutes of Health, under Contract No. HHSN261200800001E. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.

Discovery of pharmacologic regulators of the c-Myc oncoprotein

Project collaborators: Edward Prochownik (Dept. Oncology Research, UPMC Children's Hospital), Peter Wipf (Dept. Chemistry), Kristina Paris, Ahmet Bakan and Ivet Bahar (Dept. Computational and Systems Biology/UPDDI)

The c-Myc gene is de-regulated in a significant number of human cancers and its over-expression in various animal models leads to the development of a variety of malignancies. The de-regulated expression of c-Myc also leads to abnormalities of cell growth, survival, differentiation and chromosomal stability; several of these effects were first described in the Prochownik laboratory. The c-Myc protein is a transcription factor that regulates the expression of a large number of downstream target genes. DNA microarrays have been used to identify many of these genes and work in the Prochownik lab has recently focused on how several of them mediate the downstream effects of c-Myc. In a collaborative study between the Prochownik, Wipf, and Bahar labs, a number of low molecular weight, drug-like compounds that potentially inhibit the transcriptional activity of c-Myc and may serve as novel chemotherapeutic agents have been developed. Small, drug-like molecules that disrupt protein-protein interactions (PPIs) are of great interest because of their potential therapeutic benefits and commercialization possibilities. However, numerous factors suggest that PPIs may be difficult to target. Current c-Myc inhibitors bind directly to Myc, prevent its association with its obligate partner protein Max and thus lead to a loss of DNA binding and transcriptional regulatory activity.

A new and powerful in silico search algorithm that relies on the crystal structure of the Myc-Max heterodimerization domain to predict novel small molecules with the potential to compete with Max for binding to Myc was recently developed by members of the Bahar laboratory. Relative to the originally described Myc compounds, these so-called "peptidomimetic" compounds (PMCs) were found to have a much higher affinity for Myc and are significantly more potent at inhibiting the growth of tumor cells. We are further evaluating HTS hits by utilizing chemoinformatics analysis to score and refine our c-Myc pharmacophore models for synthesizing chemical libraries based upon tractable lead scaffolds.

An additional novel computational method for assessing protein druggability in which small molecular probes that have differing physicochemical properties are utilized to identify possible binding sites, give an estimation of the maximum achievable binding affinity of a potential site, and allow for design of new pharmacophore models (Bakan, Paris, and Bahar) while other work by the Prochownik lab is using techniques such as X-ray crystallography, NMR spectroscopy, and in vitro mutagenesis to determine the structures of Myc compounds in association with c-Myc. Such structural data should aid in the rational design of even more effective compounds.


Wang H, Hammoudeh DI, Follis AV, Reese BE, Lazo JS, Metallo SJ, Prochownik EV. Improved low molecular weight Myc-Max inhibitors. Mol Cancer Ther. 2007; 6:2399-408.

Hammoudeh DI, Follis AV, Prochownik EV, Metallo SJ. Multiple independent binding sites for small-molecule inhibitors on the oncoprotein c-Myc. J Am Chem Soc. 2009; 131:7390-401.

Mustata G, Follis AV, Hammoudeh D I, Metallo S J, Wang H, Prochownik EV, Lazo J S, Bahar I. Discovery of novel myc-max heterodimer disruptors with a three-dimensional pharmacophore model. J Med Chem. 2009; 52:1247-50.


Evaluation of Rationally-designed small molecules directed against the cMyc oncoprotein

Project collaborators: Ed Prochownik (PI, Dept. Oncology Research, UPMC Children’s Hospital), PSAC, UPCDC, Emory Chemical Biology Discovery Center (ECBDC) led by Haian Fu (ECBDC PI), and Paul Johnston (lead discovery scientist, Dept. Pharmaceutical Sciences, Co-PI PSAC).

The University of Pittsburg Specialized Applications Center (PSAC) and Chemical Diversity Center (UPCDC) are members of the NCI Experimental Therapeutics Program’s (NExT’s) Chemical Biology Consortium, and are partnering with the NCI to advance novel therapeutics against c-Myc.

The Myc oncogene proteins are transcription factors that regulate a number of genes involved in cell growth and metabolism, apoptosis, and cellular transformation (1). Deregulation and over expression of Myc proteins have been demonstrated to play a critical role in the pathogenesis of a variety of human cancers (2), and cMyc is among the most frequently deregulated oncoproteins in human cancer. cMyc functions by heterodimerizing with the transcription factor Max and binding to the E-box DNA recognition site to activate transcription. Disruption of Myc/Max signaling is an attractive approach to inhibit the oncogenic activity of cMyc (1,3).

The goal of the cMyc NExT project is to identify small molecule inhibitors of the Myc/Max protein interaction that can be developed into clinical compounds for the treatment of cancer. The ECBDC team is currently developing and validating multiplexed HTS assays to quantify Myc/Max dimerization and Myc/Max/DNA interaction that will be used in a high throughput screening campaign to find inhibitors of Myc/Max signaling. Hits identified in the screen, and whose activity has been confirmed, will be characterized by the PSAC team for their ability to inhibit growth of a variety of cMyc-dependent cell lines. Once a go-decision is made by the NExT Governing Body, compounds demonstrating cMyc-dependent growth inhibition will be advanced to the UPCDC team who will apply both traditional medicinal chemistry and computational chemistry methods to develop active chemical series into potential clinical candidates.


1 Yin X, Giap C, Lazo JS, Prochownik EV. Low molecular weight inhibitors of Myc-Max interaction and function. Oncogene. 2003;22:6151-6159.

2 Nesbit CE, Tersal JM, Prochownik EV. MYC oncogenes and human neoplastic disease. Oncogene. 1999;18:3004-3016.

3 Berg T, Cohen SB, Desharnais J, Sonderegger C, Maslyar DJ, Goldberg J, Boger DL, Vogt PK. Small-molecule antagonists of My/Max dimerization inhibit Myc-induced transformation of chicken embryo fibroblasts. Proc. Nat. Acad. Sci. 2002;99:3830-3835.

This project has been funded in whole or in part with Federal Funds from the National Cancer Institute, National Institutes of Health, under Contract No. HHSN261200800001E. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.


Discovery of allosteric inhibitors of mitogen activated protein kinase phosphatases by zebrafish chemical genetics

Project collaborators: Andreas Vogt (Dept. Computational and Systems Biology/UPDDI), Michael Tsang (Dept. Developmental Biology), Ahmet Bakan and Ivet Bahar (Dept. Computational and Systems Biology/UPDDI), and Billy Day (Dept. Pharmaceutical Sciences/UPDDI)

The goal of this project is to discover novel inhibitors of mitogen activated protein kinase phosphatases (MKPs) through high-content, high-throughput screening in transgenic zebrafish embryos.

Cancer cells adapt to environmental stresses or antineoplastic therapy by expression of cytoprotective proteins that help maintain the tumorigenic state, a process that has been termed "non-oncogene addiction", or NOA. Mitogen activated protein kinase phosphatase-1 (MKP-1, aka CL100/DUSP1), a member of a family of dual specificity phosphatases (DUSPs) is a potential NOA target that promotes cancer cell survival under conditions of environmental or chemotherapy-induced stress. MKP-1 is overexpressed in many human tumors, inhibits cell migration and invasion, protects cells from apoptosis by DNA damaging agents or radiation, and limits the efficacy of clinically used antineoplastic agents. Fibroblasts from MKP-1 null mice are hypersensitive to gamma-irradiation. Downregulation of MKP-1 enhances cytotoxicity of clinically used antineoplastic agents in cultured cells and prevents the growth of pancreatic tumor cells in mice. Despite the evidence linking MKP-1 to cancer and proof-of-principle studies demonstrating the effects of MKP-1 knockdown can be phenocopied by small molecules, MKPs are underexplored drug targets because potent and specific small molecule inhibitors have been lacking.

The MKP catalytic site is shallow and sensitive to oxidation, and prior high throughput screens with recombinant protein and small molecule artificial substrates have identified largely chemically reactive, promiscuous, redox-active, or biologically inactive compounds. We have recently overcome these impasses through the combined use of a fluorescent zebrafish biosensor, artificial intelligence-based image analysis, and computational modeling, which identified an allosteric inhibitor of MKP-1 and MKP-3 that inhibits cancer cell proliferation and migration but lacks whole embryo toxicity. In collaboration with Dr. Michael Tsang in the Department of Developmental Biology, we have implemented and validated the zebrafish biosensor assay in high-throughput, including a pilot library screen of 1040 agents with known mechanisms of action. An ongoing small molecule screen in the transgenic zebrafish model has identified several novel MKP inhibitory scaffolds, which are being optimized by computational modeling (Ahmet Bakan and Ivet Bahar, Department of Computational and Systems Biology/UPDDI) and medicinal chemistry (Billy Day, Pharmaceutical Sciences/UPDDI).


Vogt A, McDonald PR, Tamewitz A, Sikorski RP, Wipf P, Skoko JJ 3rd, Lazo JS. A cell-active inhibitor of mitogen-activated protein kinase phosphatases restores paclitaxel-induced apoptosis in dexamethasone-protected cancer cells. Mol Cancer Ther. 2008;7:330-340.

Molina G, Vogt A, Bakan A, Dai W, de Oliveira PQ, Znosko W, Smithgall TE, Bahar I, Lazo JS, Day BW, et al.. Zebrafish chemical screening reveals an inhibitor of Dusp6 that expands cardiac cell lineages. Nat Chem Biol. 2009;6:680-687.

Saydmohammed M, Vollmer LL, Onuoha EO, Vogt A, Tsang M. A high-content screening assay in transgenic zebrafish identifies two novel activators of fgf signaling. Birth Defects Res C Embryo Today. 2011;93:281-287.


Activators of PUMA as novel anticancer agents

Project collaborators: Lin Zhang (Dept. Pharmacology and Chemical Biology) and Andreas Vogt Dept. Computational and Systems Biology/UPDDI)

Human tumors are often either inherently insensitive, or develop resistance to conventional anticancer therapies, including radiation and chemotherapy. Apoptosis is a major cytotoxic mechanism of anticancer therapies, and defective apoptosis contributes to therapeutic resistance in cancer. PUMA, (p53-upregulated modulator of apoptosis), a BH3-only Bcl-2 family member discovered in Zhang laboratory, plays a critical role in radio- and chemotherapy-induced, p53-dependent apoptosis. PUMA deficiency renders resistance to p53-dependent apoptosis induced by common chemotherapeutic drugs, gamma-irradiation, hypoxia, and oncogene activation in human cancer cells and mice. PUMA binds to pro-survival Bcl-2 family members such as Bcl-2 and Bcl-XL, relieving their inhibition on the pro-apoptotic members Bax and Bak, and leading to mitochondrial membrane permeabilization, subsequent activation of the caspase cascade, and eventually, cell death. PUMA is unique among Bcl-2 family proteins because it binds to all members of the pro-survival Bcl-2 family proteins;therefore activation of PUMA is can promote apoptosis in cancer cells with various genetic defects.

Until recently, it was thought that transcriptional activation of PUMA by p53 is defective in most cancer cells due to p53 abnormality. Recent studies have demonstrated that several non-genotoxic, targeted anticancer drugs can activate PUMA through p53-independent mechanisms, and that PUMA activation is essential for apoptosis induced by these agents. Our results suggest that PUMA is a useful therapeutic target that is amenable to pharmacological manipulation, which will generate therapeutic effects in p53-deficient tumor cells. We have used PUMA fluorescence reporter cells to develop a high-content, cell-based screening assay that reliably detects PUMA induction in cancer cells. The assay was miniaturized, implemented on robotic liquid handling equipment, and delivered Z-factors above 0.6 in multi-day variability studies. Screening the LOPAC library of 1280 compounds with known mechanisms of action identified a number of novel PUMA-activating small molecules with potential anticancer activities. Efforts are underway to develop these novel PUMA inducers into lead compounds that restore apoptosis in human cancer cells. Many cancers have reduced PUMA expression, and this is often correlated with shorter survival of cancer patients and poor response to therapies. Compounds that activate PUMA expression would therefore be useful under conditions of defective apoptosis and impaired response to curative therapy. In the long run, such efforts may pave the way for the development of more effective anticancer agents and combination therapies.


Yu J, Yue W, Wu B, Zhang L. PUMA sensitizes lung cancer cells to chemotherapeutic agents and irradiation. Clin Cancer Res. 2006;12:2928-2936.

Yu J, Zhang L. PUMA, a potent killer with or without p53. Oncogene. 2008;27 Suppl 1, S71-83.

Dudgeon C, Peng R, Wang P, Sebastiani A, Yu J, Zhang L. Inhibiting oncogenic signaling by sorafenib activates PUMA via GSK3beta and NF-kappaB to suppress tumor cell growth. Oncogene. 2012;


Interactive technologies for leveraging the known chemistry of ANCHOR residues to disrupt protein-protein interactions (PPI's)

Project collaborators: Carlos Camacho and David Koes (Dept. Computational and Systems Biology) and Alex Doemling (University of Groningen)

Success stories in drug discovery are mostly circumscribed to traditional targets that often have natural small molecule substrates. PPI's have been very difficult to target with small molecules due to their diverse structure, topology and flexibility, as well as the lack of suitable chemical scaffolds. In fact, it has been reported that "hit" rates from traditional high throughput screening approaches are «1%. Here, we address this challenge by applying novel computational methods to develop interactive technologies to disrupt protein interactions. As an alternative to the conventional screening of historical or known compounds, our approach leverages the known chemistry of well-defined anchor residues in protein interactions and multi-component reaction (MCR) peptido-mimetic chemotypes to design small-molecule inhibitor starting points. Our ANCHOR-centric method predicts that roughly 50% of PPIs in the protein data bank (PDB) are druggable, enabling a powerful tool for modulating signaling pathways for systems biology.

The ANCHOR approach has been applied to the design of antagonists of the p53/MDM2 and p53/MDM4 PPIs. Both MDM2 and MDM4 bind to P53 and negatively regulate its tumor suppressor activity and inhibition of these interactions offer potential therapeutic benefits in cancer. By combining the ANCHORQuery analysis with the efficient MCR chemistry we have identified several novel, specific and potent scaffolds that inhibit the p53 interactions. Three of these compound classes have been optimized towards low nM antagonists. All compound classes have been extensively characterized by the use of structural biology, in cell based assays and in initial PK/PD studies.Three of the inhibitors have been chosen for further development and are currently undergoing animal PK and efficacy studies.In addition to finding novel p53/MDM2 inhibitors the ANCHOR approach has identified for the first time a small molecule inhibitor of the related MDM4 receptor. This is noteworthy since none of the currently pursued antagonists has any mdm4 activity. This is a direct result of the advanced "out-of-the-box" ANCHOR technology.


Wells JA, McClendon CL. Reaching for high-hanging fruit in drug discovery at protein-protein interfaces. Nature. 2007;450(7172):1001-9.

Domling A. Small molecular weight protein-protein interaction antagonists-an insurmountable challenge? Curr Opin Chem Biol. 2008;12:281-91. PMCID: 18501203.

Czarna A, Beck B, Srivastava S, Popowicz GM, Wolf S, Huang Y, et al. Robust generation of lead compounds for protein-protein interactions by computational and MCR chemistry: p53/Hdm2 antagonists. Angew Chem Int Ed Engl. 2010;49(31):5352-6.

Popowicz GM, Czarna A, Wolf S, Wang K, Wang W, Domling A, et al. Structures of low molecular weight inhibitors bound to MDMX and MDM2 reveal new approaches for p53-MDMX/MDM2 antagonist drug discovery. Cell Cycle. 2010;9(6).


Discovery and development of dictyostatin anticancer agents

Project collaborators: Dennis Curran (co-PI, Dept. Chemistry), Billy Day (co-PI, Dept. Chemistry and Dept. Pharmaceutical Sciences), Andreas Vogt (lead scientist, Dept. Computational and Systems Biology/UPDDI)

The natural product (-)-dictyostatin is a microtubule-stabilizing agent that potently inhibits the growth of human cancer cells, including paclitaxel-resistant clones [1]. Extensive structure-activity relationship studies based on an efficient total synthesis route have revealed several regions of the molecule that can be altered without loss of activity [2]. The most potent synthetic dictyostatin analogue discovered in a these SAR studies is 6-epidictyostatin, which has outstanding properties in in vitro and cell assays and exhibits superior in vivo antitumor activity against human breast cancer xenografts compared with paclitaxel [3].


Most recently, we applied a new, highly convergent synthesis to generate 25,26-dihydrodictyostatin and 25,26-dihydro-6-epidictyostatin [4]. Like 6-epidictyostatin, both compounds are potent microtubule-perturbing agents that induce mitotic arrest and microtubule assembly in vitro and in intact cells [5]. In vitro radioligand binding studies show that 25,26-dihydrodictyostatin and its C6-epimer are capable of displacing [3H]paclitaxel and [14C]epothilone B from microtubules with potencies comparable to (-)-dictyostatin and discodermolide. Both compounds inhibit the growth of paclitaxel- and epothilone - resistant cell lines at low nanomolar concentrations, synergize with paclitaxel in MDA-MB-231 human breast cancer cells, and have antiangiogenic activity in transgenic zebrafish larvae. Thus, we have identified three novel and exceptionally potent dictyostatin analogs as candidates for scale-up synthesis and advanced preclinical development for cancer chemotherapy.


1 Madiraju C, Edler MC, Hamel E, Raccor BS, Balachandran R, Zhu G, Giuliano K, Vogt A, Shin Y, Fournier JH, Fukui Y, Brückner A M, Curran DP, Day BW. Biochemistry. 2005;44:15053-15063.

2 Jung WH, Harrison C, Shin Y,Fournier JH, Balachandran R, Raccor BS, Sikorski RP, Vogt A, Curran DP, Day BW. J. Med. Chem. 2007;50:2951-2966.

3 Eiseman JL, Bai L, Jung W-H, Moura-Letts G, Day BW, Curran DP. J. Med. Chem. 2008;51:6650-6653.

4 Zhu W, Jiménez M, Jung W-H, Camarco DP, Balachandran R, Vogt A, Day BW, Curran DP. J. Am. Chem. Soc. 2010;132:9175-9187.

5 Vollmer LL, Jimenez M, Camarco DP, Zhu W, Daghestani HN, Balachandran R, Reese CE, Lazo JS, Hukriede NA, Curran DP, Day BW, Vogt A. Mol. Cancer Therapeut. 2011;10:994-1006.


Characterizing Cellular Heterogeneity of Drug Responses for the Development of Improved Therapeutics and Diagnostics: Squamous Cell Carcinoma of the Head and Neck

Project collaborators: Tim Lezon, Robert Boltz, Lawrence Vernetti, (Dept. Computational and Systems Biology/UPDDI), Jennifer Grandis (Dept. Otolaryngology), and Lans Taylor and Albert Gough, (Dept. Computational and Systems Biology/UPDDI)

Non-genetic heterogeneity in cellular function, while long recognized as more than just experimental noise [1, 2], has now been established as a fundamental feature of cell biology [3], but methods for large-scale single cell analysis remain a challenge. An understanding of the sources and implications of heterogeneity, both in vivo and in vitro, is critically important in the construction of more effective models for therapeutic discovery, diagnostics and for the treatment of disease. We are developing methods for large-scale acquisition and analysis of hyperplexed data to monitor multiple pathways in single cells.

Our initial project is the evaluation of heterogeneity in multiple pathways in squamous cell carcinoma of the head and neck (SCCHN). While many sources of genetic heterogeneity in cancer are well understood, the sources of non-genetic heterogeneity, especially arising from the complexity of tumor microenvironments, are not. Even a clonal cell line under carefully controlled experimental conditions exhibits a significant degree of intrinsic cellular heterogeneity, which is believed to be an important factor in cell adaptation and survival, and therefore of particular significance in cancer therapies. The goal of this project is to characterize the heterogeneity in SCCHN cells, identify the biologically relevant mechanisms that result in heterogeneity, and to use this knowledge to create a better model for development of new therapies.

From the computational and systems standpoint, a core challenge in quantitative modeling of cellular heterogeneity is to identify its biological and environmental origins. Although variations in cellular response to perturbation may arise from biologically relevant differences in multiple cellular pathways, they might also simply reflect the nonlinear response of a single pathway. We employ a variety of data-driven approaches to confine our model search to a relevant region of the model space.

A systems-based analysis of cellular heterogeneity, includes information on single cells as well as cell populations. When combined with pathway models, heterogeneity analysis has great potential to identify novel targets and compounds. For example, we use supervised and unsupervised machine learning methods to identify maximally informative sets of pathway targets; automated searches of mathematical transformations of biomarkers into useful features; development and comparison of methods for phenotype identification and classification; and constructing and updating the mapping between perturbations and their corresponding population distributions.


1 Huang S. Non-genetic heterogeneity of cells in development: more than just noise. Development. 2009;136(23):3853-62. Epub 2009/11/13. PubMed PMID: 19906852.

2 Bright GR, Whitaker JE, Haugland RP, Taylor DL. Heterogeneity of the changes in cytoplasmic pH upon serum stimulation of quiescent fibroblasts. Journal of cellular physiology. 1989;141(2):410-9. Epub 1989/11/01. PubMed PMID: 2478571.

3 Altschuler SJ, Wu LF. Cellular heterogeneity: do differences make a difference? Cell. 2010;141(4):559-63. PubMed PMID: 20478246.


Manipulating mitochondrial dynamics for the inhibition of cancer cell growth

Project Collaborators: Wei Qian (Postdoctoral fellow, Department of Pharmacology and Chemical Biology), Bennett Van Houten (PI, Dept. Pharmacology and Chemical Biology, University of Pittsburgh Cancer Institute) and Lee McDermott (Dept. Pharmaceutical Sciences/UPDDI).

Mitochondria are dynamic organelles that constantly move, fuse and divide. These dynamic features of mitochondria are achieved by the coordination of several specific regulatory proteins. Dynamin-related protein 1 (Drp1) and mitochondrial fission factor (Mff) are the main proteins responsible for mitochondrial fission, optic atrophy 1 (Opa1) and mitofusin 1/2 (Mfn1/2) are the main proteins responsible for mitochondrial fusion. Miro and milton are the essential components of the machinery for mitochondrial movement. Deficiencies proteins that control mitochondrial dynamics are associated with a number of human pathologies, including autosomal dominant neurodegenerative disease Charcot-Marie-Tooth type 2A, sporadic Alzheimer’s disease, Parkinson’s disease, cardiomyopathy, muscle atrophy, and newborn mortality. Mitochondrial dynamics are essential for several critical cellular events including energy metabolism, cell differentiation, proliferation [1,2], immune response and apoptosis. Since alterations in mitochondrial metabolism and resistance to apoptosis induced by anticancer chemotherapeutic drugs are frequently observed in cancer cells, interruptions of mitochondrial metabolism and mitochondria-initiated apoptotic pathway have been considered attractive anticancer strategies. However, whether mitochondrial dynamics play roles in cancer cell physiology and is a potential target for cancer treatment are not clear. The goal of our studies is to elucidate the roles of mitochondrial dynamics in regulating cancer cell physiology and test the strategies in effectively disrupting mitochondrial dynamics which impact cancer cell growth by either genetic or pharmacological tools. We have discovered that mitochondrial dynamics are not only involved in mediating cancer cell energy metabolism, they also impact genome integrity and cell proliferation in an energy metabolism and reactive oxygen species-independent manner. Our results indicate that mitochondrial dynamics could be a feasible anticancer target. Due to the complex feature of mitochondrial dynamics, we will be testing which process has the most dramatic impact on tumor cell growth by targeted disruption of each single process of mitochondrial dynamics. To assess what types of cancers show the greatest response to the inhibition of mitochondrial dynamics, a variety of cancer cell lines including tumor cell lines from breast, lung, prostate, head and neck, and bone are being studied. The effect of disruption of mitochondrial dynamics on cell growth will also be tested in both normal and tumor cell lines in order to compare the sensitivity and assess the anticancer activity vs. possible side effects.


1 Mitra K, Wunder C, Roysam B, Lin G, Lippincott-Schwartz J. Proc Natl Acad Sci U S A. 2009;106:11960-5

2 Kashatus DF, Lim KH, Brady DC, Pershing NL, Cox AD, Counter CM. Nat Cell Biol. 2011;13:1108-1115.


Breaking tumor cell metabolic symbiosis: a novel treatment paradigm

Project collaborators: Bennett Van Houten (PI, Dept. Pharmacology and Chemical Biology), Lee McDermott (Dept. Pharmaceutical Sciences/UPDDI)

The widespread use of F18-2-deoxyglucose and positron emission tomography (PET) for the visualization of tumor burden in humans clearly demonstrates that tumors show enhanced metabolism of glucose. Furthermore, it is generally accepted that increased glycolytic potential is one hallmark of cancer. This observation has been cited as strong evidence for the Warburg hypothesis, the idea that tumor cells show mitochondrial dysfunction and increased glycolysis even in the presence of oxygen 1. However, over the years, increasing evidence suggests that mitochondrial function is not broken, but just altered to suit the metabolic needs of the tumor, and thus the Warburg hypothesis has been hotly debated. Tumor metabolic symbiosis reconciles both of these points of view and posits that tumors maintain regions of glycolysis and oxidative phosphorylation depending on the carbon sources and levels of oxygenation [1]. During metabolic symbiosis, the hypoxic region of the tumor generates energy primarily by metabolizing glucose through glycolysis producing the waste product lactate. This lactate is converted to pyruvate in well vascularized regions of the tumor to drive oxidative phosphorylation. Our first goal is to develop imaging tools to visualize these two populations of cells in xenograft tumor models. This project will test the hypothesis thattumor cells that have high metabolic flexibility will show the highest ability to exhibit metabolic symbiosis and therefore support the most rapid tumor development. We expect that the poorly oxygenated population of cells will show high glucose uptake, glycolysis and lactate production, whereas the other population of well oxygenated cells will use this lactate to drive OXPHOS. Having achieved direct evidence for metabolic symbiosis in xenograft models, the second phase of this work will examine the effects of specific inhibitors of metabolism and lactate efflux or uptake alone or in combination with standard chemotherapeutic approaches. Cell culture models and direct analysis of xenograft tumors will be examined in this second phase.


1 Nakajima EC, Van Houten B. Metabolic Symbiosis in cancer: refocusing the Warburg lens. Mol. Carcinogenesis [Epub ahead of print] PMID:22228080. 2012.


Identification and HDAC7-specific inhibitors for endocrine resistant breast cancer

Project collaborators: Steffi Oesterreich (Dept. Pharmacology and Chemical Biology)

Targeting the estrogen receptor (ER) in ER+ve breast cancer has been the most successful targeted therapy in oncology over the last decades. Endocrine therapies act by blocking the activation of ERα, thus inhibiting its agonist-dependent transcriptional activation function. These therapies include tamoxifen, a competitive inhibitor for estrogen binding, and aromatase inhibitors, which block conversion of testosterone to estrogen in postmenopausal women. Unfortunately, resistance to endocrine therapy is a frequent event, and thus a major clinical problem. In the case of tamoxifen treatment, resistance is associated with tamoxifen switching from being an antagonist to functioning as an agonist.

We and others have recently demonstrated that greater than 50% of ERα target genes are repressed in response to estradiol, both in cell culture models and in breast tumors. Mechanisms required for ERα-mediated gene repression remain poorly understood. Importantly, many repressed genes are cell cycle inhibitors, pro-apoptotic proteins or tumor suppressor genes. Repression of these targets may be critical in maintaining growth and survival of breast tumor cells.

Our laboratory has recently identified histone deacetylase 7 (HDAC7) as having a unique, critical role in the repression of ERα target genes (Malik et al, MCB 2009). The repression of genes including the tumor suppressor reprimo (RPRM) does not require HDAC1-6 or HDAC8-10, but specifically HDAC7.HDAC7 belongs to the family of class II histone deacetylases, which deacetylates histones but also other non-histone proteins; its activity is regulated via regulation of its expression, and nuclear-cytoplasmic shuttling. Importantly, for many repressed targets, tamoxifen functions as an agonist (i.e. represses gene expression). Continued repression of these genes, for example in the presence of overexpressed HDAC7, may allow for tumor cell growth and contribute to endocrine resistance.

Based on these observations, we hypothesize that HDAC7 represents a novel target for epigenetic therapy to overcome endocrine resistance. HDAC7 inhibition may sensitize breast cancer cells to endocrine therapy. We will therefore collaborate with the UPDDI to screen and characterize HDAC7-targeting compounds.


Malik S, Jiang J, Garee J, Verdin E, Lee AV, O’Malley BW, Zhang W, Belaguli NS, Oesterreich S. Intricate Interplay between HDAC7 and FoxA1 in estrogen-mediated repression of RPRM. Mol Cell Biol. 2010;30:399-412, PMID 19917725; PMC 27948743


Systems modeling to identify preclinical biomarkers of anti-IGF-IR antibodies and small molecular tyrosine kinase inhibitor activity

Project collaborators: Adrian Lee (Dept. Pharmacology and Chemical Biology), Xinghua Lu (Dept. Biomedical Informatics), Takis Benos (Dept. Computational and Systems Biology), Bino John (Dept. Computational and Systems Biology), Lans Taylor (UPDDI)

The insulin-like growth factor (IGF) system consists of two ligands, IGF-I and IGF-II, two receptors, IGF-IR and IGF-IIR, and six high-affinity binding proteins, IGFBP 1-6. IGF-IR is a receptor tyrosine kinase which is activated upon binding of either IGF-I or IGF-II. IGF-IR binds the adaptor proteins termed insulin receptor substrates (IRSs) which have no intrinsic kinase activity but contain multiple potential tyrosine phosphorylation sites that can bind SH2-containing proteins resulting in activation of a number of downstream signaling pathways including PI3K/Akt and ERK1/2 [1].

Insulin-like growth factor-I receptor (IGF-IR) signals to cell proliferation, survival, invasion, and transformation. IGF-IR is hyperactive and overexpressed in many cancers, and preclinical data have validated IGF-IR as a therapeutic target — several IGF-IR inhibitors have recently entered clinical trials. While early results were promising, several Phase 3 trials in unselected cancer populations have recently been stopped. Anti-IGF-IR therapy has clear activity; however, it is only in a small number of patients. Similar to other targeted therapies, there is a desperate need for biomarkers of sensitivity and resistance.

We recently reported that IGF-IR levels correlate with sensitivity to an IGF-IR antibody (Figitumumab) in a small Phase 2 study in non-small cell lung cancer, however, the predictive value was weak [2]. Expanding upon this we have developed a signature of IGF-I regulated genes which indicated poor breast cancer patient prognosis [3] (see Figure to the left), and correlated with response to an anti-IGF-IR inhibitor [3]. We are continuing this approach by applying numerous genome-wide approaches (transcriptomics, proteomics, genomics) to understand how IGF-I regulates tumor cell biology, with the goal of identifying preclinical biomarkers of IGF-IR activity. These assays include the use of 2D and 3D model systems and the use of preclinical models and human tissues for validation.


1 Taniguchi CM, Emanuelli B, Kahn CR. Critical nodes in signalling pathways: insights into insulin action. Nat Rev Mol Cell Biol. 2006;7:85-96.

2 Gualberto A, Dolled-Filhart M, Gustavson M, Christiansen J, Wang YF, Hixon ML, Reynolds J, McDonald S, Ang A, Rimm DL, Langer CJ, Blakely J, Garland L, Paz-Ares LG, Karp DD, Lee AV. Molecular analysis of non-small cell lung cancer identifies subsets with different sensitivity to insulin-like growth factor I receptor inhibition. Clin Cancer Res. 2010;16:4654-65. PMCID:2952544

3 Creighton CJ, Casa A, Lazard Z, Huang S, Tsimelzon A, Hilsenbeck SG, Osborne CK, Lee AV. Insulin-like growth factor-I activates gene transcription programs strongly associated with poor breast cancer prognosis. J Clin Oncol. 2008;26:4078-85. PMCID:2654368


Development of combination vaccines selectively targeting blood vessels in solid cancers

Project collaborators: Walter Storkus (PI, Departments of Dermatology and Immunology) in collaboration with the UPCI CBF Mass Spectrometry Platform (MSP) and the University of Pittsburgh Drug Discovery Institute.

Cancer vaccines based on tumor-associated antigens (TAA) have thus far proven to be immunogenic, but only rarely curative in the clinic. This limitation in efficacy may relate, at least in part, to the heterogeneity of cancer cells found within a given tumor lesion, and the ability of variant tumor cells to escape immune regulation. A theoretical means by which to therapeutically circumvent the “instability” of cancer cells themselves involves the development of vaccines eliciting protective T cells that are capable of selectively recognizing tumor blood vessel-associated cells (TBVC), such as vascular endothelial cells and pericytes. Indeed, we have recently shown that vaccination against TBVC-derived peptides promotes the T cell-dependent regression of colon carcinoma and melanoma tumors in mice bearing a human-like T cell repertoire [1]. Additional work in our group suggests that the efficacy of such vaccines may be improved using combination vaccine approaches that incorporate small molecules that inhibit tumor resident cell populations that functionally suppress the recruitment and protective action of vaccine-activated T cells [2].    

Ongoing collaborative studies are designed to:  1) use LC-MS/MS proteomics-based approaches to define additional, potentially therapeutically-preferred TBVC-associated targets for inclusion in future vaccine for the treatment of solid cancers and 2) identify novel small molecule inhibitors that “condition” the tumor microenvironment for improved therapeutic action mediated by anti-TBVC T cells. Using these strategies, we anticipate developing new immuno-therapeutic regimens that will mitigate the confounding clinical phenomena of tumor-induced immune suppression and tumor cell heterogeneity that has thus far limited the impact of cancer cell-targeted vaccines.


1 Zhao X, Bose A, Komita H, Taylor JL, Chi N, Lowe DB, Okada H, Cao Y, Mukhopadhyay D, Cohen PA, Storkus WJ. J Immunol. 2012;188:1782-1788.

2 Bose A, Taylor JL, Alber S, Watkins SC, Garcia JA, Rini BI, Ko JS, Cohen PA, Finke JH, Storkus WJ. Int J Cancer. 2011;129:2158-2170.


High throughput screening and rational design of small molecule inhibitors of NADPH oxidase 

Project collaborators:  Patrick J. Pagano (PI) and M.Eugenia Cifuentes (Dept. Pharmacology and Chemical Biology and Vascular Medicine Institute), UPDDI, and Peter Wipf and Erin Skoda (Dept. Chemistry and Accelerated Chemical Discovery Center

Reactive oxygen species (ROS) such as superoxide (O2-) and peroxide (H2O2) have been shown to act as signaling molecules for the regulation of various cellular functions  such as signal transduction, cell proliferation, gene expression, angiogenesis. However, an increased ROS generation referred to as oxidative stress has clearly been implicated in a broad range of pathophysiological conditions including hypertension, diabetes, atherosclerosis, Parkinson's disease, Alzheimer's disease, and cancer. NADPH oxidase (Nox) is a major source of ROS practically in every tissue of the body and represents an important therapeutic target for the treatment of diseases involving oxidative stress. NADPH oxidase is a multi-subunit enzyme complex that includes two membrane-spanning subunits, Nox2 and p22phox, and major cytoplasmic subunits, p40phox, p47phox and p67phox and rac.  Upon stimulation, the cytosolic subunits are recruited to the membrane to form an active enzyme. Various isoforms of NADPH oxidase have been described in a variety of tissues which differ from the Nox2 system in unique substitutions of their Nox-subunit amino acid sequence as well as the cytoplasmic components that they require (Nox1- and Nox 4-based oxidase systems).  

Given the plethora of diseases in which NADPH oxidase-derived ROS are implicated as a pathological factor, the quest for specific inhibitors has intensified in recent years. The availability of these inhibitors, however, has been limited despite considerable efforts made in the field. Major deficiencies in available compounds have been the lack of specificity and a lack of understanding of the mechanisms of action. To date, the best characterized and widely-used isoform-specific inhibitor of Nox2, Nox2ds, is a peptidic inhibitor rationally designed in Dr. Pagano’s lab to block the interactions of cytosolic and membrane subunits of the enzyme [1]. This inhibitor has been shown to attenuate angioplasty-induced neointimal proliferation [2] and ameliorate numerous other diseases. Despite their specificity and efficacy, the use of peptides as therapeutics is considered problematic due to their degradation in the gut and limited oral bioavailability. Accordingly, peptidomimetic development strategies are being pursued.

The major goal of this project is to identify isoform-specific small molecules inhibitors of Nox that can be developed as clinically active drugs for the treatment of cardiovascular diseases. In collaboration with the UPDDI, we have established and validated a cell-based platform to perform high throughput screening of Nox2 and other Nox inhibitors and are currently applying rational design and medicinal chemistry to optimize lead compounds. 


1 Rey FE, Cifuentes ME, Kiarash A, Quinn MT, Pagano PJ. Novel competitive inhibitor of NAD(P)H oxidase assembly attenuates vascular O 2  -  and systolic blood pressure in mice. Circ.Res. 2001;89:408-414.

2 Dourron HM, Jacobson GM, Park JL, Liu J, Reddy DJ, Scheel ML, Pagano PJ. Perivascular gene transfer of NADPH oxidase inhibitor suppresses angioplasty-induced neointimal proliferation of rat carotid artery. Am.J.Physiol.Heart Circ.Physiol. 2005;288:H946-H953.