CURRENT RESEARCH PROJECTS

PPI targeting

Protein-protein interactions (PPIs) regulate virtually every cellular process ranging from cell cycle control, cellular signaling, DNA replication, transcription and translation to apoptosis, and aberrant PPIs have the potential to cause or contribute to human disease. Therefore, inhibition of the direct PPIs that mediate many important biological processes is an emerging and challenging area in drug discovery. Previously, the spotlight in drug discovery has been on a relatively small number of validated therapeutic target classes, such as G-protein coupled receptors and protein kinases, with well characterized enzymatic and cellular activities that are chemically tractable and small molecular weight PPI antagonists were generally regarded as too difficult to be targeted. Discovery of a small molecule to bind to a protein-protein interface and subsequently inhibit the interaction poses several challenges, including the initial identification of suitable protein-protein interactions, the surface area of the interface, and the location of 'hot spots'. However, in recent years, advances in structure-based drug design and the development of an impressive variety of high-throughput screening (HTS) assay formats have yielded an expanding list of new antagonists of diverse PPIs. Along this line, we have designed and constructed our own HTS system to discover lead compounds targeting PPIs and identified new small molecule antagonists of the STAT3, Foxp3 and Survivin.

Discovery of novel STAT3 antagonists
Signal Transducer and Activator of Transcription (STAT) proteins were originally discovered as latent cytoplasmic transcription factors that mediate cytokine and growth factor-related extracellular signals. These signaling pathways involve the activation of receptor tyrosine kinases, such as epidermal growth factor receptor and platelet-derived growth factor receptor, and Janus kinases (JAKs). Following phosphorylation at tyrosine residue, two STAT monomers dimerize through reciprocal phosphotyrosine and Src homology 2 (SH2) domain interactions. A dimeric form of STAT translocates into the nucleus and binds to STAT-specific DNA-response elements of target genes for transcription activation. The STAT family comprises seven members, namely, STAT1-STAT4, STAT6, and the closely related STAT5a and STAT5b proteins. The members play a role in diverse biological functions, including cell proliferation, cell survival, angiogenesis, apoptosis, and inflammation. Several lines of evidence have implicated some of the STAT family members in malignant transformation and tumor cell survival. In particular, STAT3 is constitutively activated in many types of hematopoietic and solid tumors, such as leukemia and breast and prostate cancers. In light of these findings, STAT3 has been considered an attractive target for the development of new anticancer drugs. Because of its central role just downstream of protein tyrosine kinases, aberrant STAT3 activity is often associated with transforming mechanisms induced by oncogenic tyrosine kinases. In addition, recent reports showed that STAT3 was constitutively activated both in tumor cells and in immune cells confined in tumor microenvironments, and that constitutively activated STAT3 inhibited the expression of mediators necessary for immune responses against the tumor cells. Although a large number of small molecules have been reported to inhibit STAT3 signaling, the vast majority of them act on STAT3 by inhibiting the upstream tyrosine kinases responsible for STAT3 activation. In particular, small-molecule inhibitors of JAK2 abolished STAT3 activity and induced tumor cell apoptosis. In another study, peptide-based STAT3 inhibitors designed to target the STAT3 SH2 domain were effective in suppressing the cellular functions of STAT3. In addition, several small molecules such as STA-21, Stattic and S3I-201 have been reported to inhibit STAT3 by targeting to the SH2 domain. S3I-201 exhibited antitumor efficacy in a mouse xenografted with human breast tumor harboring constitutively active STAT3, and thus presenting a proof-of-concept for the potential clinical use of such small molecules.

 

We hypothesized that the AlphaScreen system could be used for detecting the interaction between the SH2 domain and the pTyr-containing peptide. In order to develop the assay, SH2-containing recombinant proteins STAT3, STAT1, and Grb2 were expressed and subsequently biotinylated through avi-tag introduced at the N-terminus of the proteins. Evaluation by immunoblotting using antibodies specific for STAT3, STAT1 and Grb2 clearly showed successful expression and biotinylation of the recombinant proteins. The sequences of the pTyr peptides as SH2 ligands used in this study were the same as those used in previous studies. When biotinylated STAT3 and FITC-GpYLPQTV were combined in the assay mixture, the maximum chemiluminescence signal was generated from the reaction of the streptavidin-coated donor beads and anti-FITC acceptor beads. A non-labeled inhibitor peptide, Ac-GpYLPQTV-NH2, competitively and strongly inhibited the binding of STAT3 to FITC-GpYLPQTV (IC50 = 0.3 μM), whereas its non-phospho counterpart, Ac-GYLPQTV-NH2, or any other peptides used in this study did not. STAT1 and Grb2 with their corresponding peptides showed similar results and no crossover inhibition among those phosphopeptides was observed. These results show that the chemiluminescence signal detected in the assay indicates specific interaction between the SH2-containing protein and its target peptide, and thus suggest that this system is capable of screening test compounds for the discovery of SH2 antagonists.

We then screened our chemical library using the in vitro AlphaScreen assay and identified several hit compounds that had the ability to inhibit the STAT3 binding to phosphopeptides in a concentration-dependent manner. One of them is a compound structurally related to porphyrin, 5,15-DPP. This compound selectively antagonized STAT3-SH2 (IC50 = 0.28 μM) in vitro over the other SH2-containing proteins STAT1 and Grb2. Reduced inhibition of SH2 function in STAT1 and Grb2 induced by 5,15-DPP correlated with the degree of sequence similarity (78% and 65%, respectively) in the SH2 domain among the three proteins. Another biochemical analysis with Surface Plasmon Resonance (SPR) determined the direct binding of 5,15-DPP to the recombinant STAT3. From the kinetic analyses using SPR, the estimated KD values for the STAT3 binding to the Ac-GpYLPQTV-NH2 and 5,15-DPP were 19 nM and 880 nM, respectively. Although the affinity of 5,15-DPP to STAT3 was lower than that of the phosphopeptide ligand, these results suggest that 5,15-DPP directly binds to STAT3 and antagonizes the function of STAT3-SH2 in vitro. In order to analyze the ability of 5,15-DPP to inhibit STAT3 dimerization in cells, a fluorescence resonance energy transfer (FRET) assay was performed. HEK293 cells were co-transfected with both STAT3-CFP and STAT3-YFP. IL-6 stimulation induced a high FRET signal by the interaction between CFP-STAT3 and YFP-STAT3 in the cells. Pretreatment of the cells with 5,15-DPP before IL-6 stimulation resulted in the reduction of FRET signals. We next examined whether 5,15-DPP inhibited STAT3 nuclear translocation and subsequent DNA binding in human breast cancer cells with activated STAT3. An ELISA-based assay with the consensus-binding site on the olignucleotide immobilized in a 96-well microtiter plate was used to analyze the nuclear lysates from MDA-MB-468 cells treated with 5,15-DPP. The DNA binding activity of STAT3 was suppressed by the treatment with 5,15-DPP in a concentration-dependent manner. STAT3 nuclear translocation was also inhibited by 5,15-DPP, suggesting that inhibition of STAT3 activation blocks its nuclear localization. In contrast, no inhibition was observed in the nuclear translocation and DNA binding of STAT1, STAT5a, or STAT5b. In order to elucidate the effect of 5,15-DPP on the STAT3-DNA complex in cells, a chromatin immunoprecipitation (ChIP) assay was conducted. Following stimulation of cells with IL-6 for 1 hr, protein-DNA complexes were fixed and immunoprecipitated with anti-STAT3 antibody. The complex was subjected to PCR amplification for the detection of the c-myc promoter, which is known as one of the targets for transcriptional activation of STAT3. The amplified c-myc promoter was clearly detected in the absence of 5,15-DPP, whereas a reduction in the band intensity was observed in the presence of this compound. These results support the hypothesis that 5,15-DPP inhibits STAT3 dimerization by acting as an antagonist at the SH2 domain and thus blocks its DNA binding events that is required for the transcriptional activity of STAT3 in cells. The expression level of c-myc protein was also decreased in the 5,15-DPP-treated cells. However, the phosphorylation of STAT3 was not affected. Taken together, the inhibition of STAT3 function by 5,15-DPP in MDA-MB-468 cells may be caused by the direct inhibition of STAT3 dimerization.

We identified a N-[2-(1,3,4-oxadiazolyl)]-4-quinolinecarboxamide derivative, STX-0119, as another STAT3-SH2 antagonist by our HTS system combined with the DOCK4-based virtual screening system. In order to examine the ability of STX-0119 to inhibit STAT3 dimerization in cells, a FRET assay was performed. Pre-treatment of the cells with STX-0119 prior to IL-6 stimulation resulted in the reduction of FRET signals. In addition, a ChIP assay of STX-0119-treated MDA-MB-468 cells revealed a reduction in amplification of the c-myc promoter. Inhibitory activities were not exhibited by a truncated analogue lacking the 2-Ph, STX-0872 in those assays. These results suggest that the DNA binding activity of STAT3 was effectively inhibited by STX-0119 presumably due to disruption of STAT3 dimerization in cells. To further evaluate the effects of STX-0119 against downstream transcriptional activation by STAT3, we investigated the expression of its target proteins in human breast cancer MDA-MB-468 cells. Western blotting analysis of lysate from MDA-MB-468 cells treated with STX-0119 showed that STX-0119 reduced the expression of STAT3 target proteins, namely, c-myc, cyclin D1, and survivin in a concentration-dependent manner. STX-0872, however, did not suppress the expression of those STAT3-regulated oncoproteins. It is noteworthy that STX-0119 has no effect on the level of STAT3 or Tyr705-phosphorylated STAT3. This suggests that STX-0119 inhibits STAT3 dimerization through a direct interaction with the STAT3 protein and not via the modulation of upstream regulators such as JAK. We next examined the selectivity of STX-0119 against STAT3. Nuclear lysates from MDA-MB-468 cells treated with STX-0119 were analyzed by quantitative ELISA assay. The DNA binding activity of STAT3 was suppressed by STX-0119 treatment. No inhibition of DNA binding activity of other STAT family members, namely, STAT1, STAT5a, and STAT5b, was observed. Selective induction of apoptosis against cancer cells harboring constitutively activated STAT3 was further examined by measuring caspase-3/7 activity in cells treated with STX-0119. The human breast cancer cell line MDA-MB-468 was shown to harbor the constitutively activated pSTAT3 (Tyr705) but MDA-MB-453 did not. STX-0119 induced apoptosis in MDA-MB-468 cells. However, STX-0119 had a minimal effect on MDA-MB-453 cells. Taken together, the results suggest that STX-0119 selectively disrupted STAT3 dimerization and the subsequent transcriptional activity of STAT3, followed by the induction of apoptosis of cancer cells with constitutively active STAT3. Finally, we evaluated the anti-tumor activity of STX-0119 in vivo. We selected the human lymphoma cell line SCC-3 that had been characterized as a cell line which expresses constitutively activated STAT3 and the most sensitive to STX-0119 in vitro among cell lines used.29 SCC-3 was implanted into the hind flank of male BALB/cA-nu/nu nude mice and allowed to establish sizable tumors. Oral gavages of STX-0119 at 160 mg/kg s.i.d. for 4 days suppressed the growth of SCC-3 cells significantly. Pharmacokinetic analysis showed that the plasma concentration of STX-0119 was maintained at >100 µg/mL (>260 µM) even at 8 hr after administration. No obvious body weight loss or toxicological effects were observed during the evaluation. This is the first demonstration of in vivo efficacy following oral administration of a STAT3 dimerization inhibitor. Further optimization of the STX-0119 is underway.

Bendamustine (BENDA), which bears the bis(2-chloroethyl)amino moiety, is an alkylating agent that stops the growth of cancer cells by binding to DNA and interfering with its replication. However, the mechanism of action underlying its excellent clinical efficacy remains unclear. In this work, we report that BENDA inhibits signal transducer and activator of transcription 3 (STAT3). In an AlphaScreen-based biochemical assay using recombinant human STAT3, binding of STAT3-Src homology 2 (SH2) to the phosphotyrosine (pTyr, pY) peptide was inhibited by BENDA but not by the inactive metabolite dihydroxy bendamustine (HP2). When a single point mutation of C550A or C712A was introduced into recombinant human STAT3, its sensitivity to BENDA was substantially reduced, suggesting that these cysteine residues are important for BENDA to inhibit STAT3. Furthermore, BENDA suppressed the function of cellular STAT3 as a transcriptional activator in a human breast cancer cell line, MDA-MB-468, with constitutively activated STAT3. A competitive pull-down assay using biotinylated BENDA (Bio-BENDA) revealed that BENDA bound tightly to cellular STAT3, presumably through covalent bonds. Therefore, our results suggest that the anticancer effects of BENDA may be associated, at least in part, with its inhibitory effect on the SH2 domain of STAT3.

Discovery of Epirubicin as a Foxp3 inhibitor
Regulatory T cells (Tregs) play a significant role in protection against autoimmune diseases and prevention of rejection of allogenic transplants. Forkhead box protein p3 (Foxp3) is a master transcription factor of Tregs and is crucial to their development and inhibitory function. In humans, an autoimmune syndrome termed IPEX (immune dysregulation, polyendocrinopathy, enteropathy, X-linked) is caused by mutations in Foxp3 . Conversely, the immunosuppressive activity of Tregs may hamper the induction of immune responses against cancer and infectious agents. Indeed, it was shown that Tregs suppress the function of tumor-reactive T cells in vitro and accumulation of Tregs in tumors predicts poor survival in many types of human tumors. Development of small molecule inhibitors of Foxp3 function is therefore considered a promising strategy to enhance anti-tumor immunity. In this study, we developed a novel cell-based assay system in which the NF-κB luciferase reporter signal is suppressed by the co-expressed Foxp3 protein. Using this system, we screened our chemical library consisting of approximately 2,100 compounds and discovered that a cancer chemotherapeutic drug epirubicin restored the Foxp3-inhibited NF-κB activity in a concentration-dependent manner without influencing cell viability. Using immunoprecipitation assay in a Treg-like cell line Karpas-299, we found that epirubicin inhibited the interaction between Foxp3 and p65. In addition, epirubicin inhibited the suppressor function of murine Tregs and thereby improved effector T cell stimulation in vitro. Administration of low dose epirubicin into tumor-bearing mice modulated the function of immune cells at the tumor site and promoted their IFN-γ production without direct cytotoxicity. In summary, we identified the novel action of epirubicin as a Foxp3 inhibitor using a newly established luciferase-based cellular screen. Our work also demonstrated our screen system is useful in accelerating discovery of Foxp3 inhibitors.

 

Discovery of novel Survivin antagonists
Survivin is a member of the inhibitor of apoptosis (IAP) protein family and has been implicated in both cell survival and the regulation of mitosis. It is expressed predominantly and at high levels in the vast majority of tumors and during embryonic development, but is undetectable in terminally differentiated adult tissues. Clinically, a correlation exists between a high level of expression of Survivin in tumors and poor prognosis among patients with various cancers. However, the molecular mechanisms involved in its function are not clear, although several binding partner proteins have been proposed to date. Here, we report the identification of a novel small molecule Survivin antagonist, which disrupts the Survivin-Smac/DIABLO interaction in cells. In order to identify Survivin-directed antagonists, we developed a high-throughput screening system based on AlphaScreen technology, which allows the identification of small molecules with the ability to inhibit the interaction of Survivin with Smac/DIABLO or INCENP in vitro. We screened chemical libraries, generated in-house, using this system and identified a 5-deazaflavin analog (PVHE-136) as a hit compound that selectively inhibited the interaction of Survivin with Smac/DIABLO but not INCENP. In cultured cells, PVHE-136 abrogated the formation of the complex between Survivin and Smac/DIABLO. In addition, this compound was able to sensitize cultured cells to doxorubicin-mediated DNA damage stress and synergistically enhance apoptotic cell death. Thus, the small-molecule inhibitor described here may serve as a proof-of-principle agent for discriminating between the multiple functions of Survivin.

 

Enzyme targeting

Human beings have used natural substances as a medicine from experience in ancient times. After the 20th century came, so many natural products and synthetic compounds have been developed as a drug with evolution of pharmacological, physiological, and biochemical technologies. Furthermore, the human genome sequence was completed in 2003 and paradigm shift to new drug development style symbolized by the word "genome-based drug discovery" took place. In genome-based drug discovery, target validation is the foundation of drug discovery and requires greater attention if we are to reduce the risk of failure after significant investment. On the other hands, researchers have increasingly explored the use of small molecules to modulate and characterize protein function. These methods are generally analogous to classical genetic approaches and have accordingly been termed chemical genetics. Both forward and reverse strategies with HTS are employed for chemical genetics. The potential gains of chemical genetics in both basic science and clinical medicine are immense.

IDO inhibitors
Indoleamine 2,3-dioxygenase (IDO) is an extrahepatic heme-containing dioxygenase that catalyzes the addition of oxygen across the C-2/C-3 bond of the indole ring of tryptophan (Trp). This is the initial and rate-limiting step in the catabolism of the essential amino acid Trp to N-formylkynurenine along the kynurenine pathway, which leads to biologically active metabolites such as the neurotransmitter serotonin, excitoxin quinolinic acid, N-methyl-D-aspartate (NMDA) receptor antagonist kynurenic acid, and nicotinamide adenine dinucleotide (NAD). IDO is expressed ubiquitously but predominately in cells within the immune system where it is specifically induced in dendritic cells and macrophages at the sites of inflammation by cytokines. It is known that IDO is overexpressed in a variety of diseases, including cancer, neurodegenerative disorders (e.g., Alzheimer’s disease), age-related cataract, and HIV encephalitis. Among these, IDO has been shown to play an important role in the process of immune escape by tumors. In environments where the Trp concentration has been depleted by IDO, killer T cells cannot be activated by antigens, and they undergo G1 cell cycle arrest leading to apoptosis and immunosuppression. Consequently, tumor cells escape from the immune response and survive. This is consistent with the observation that increased expression of IDO in tumor cells is correlated with poor prognosis for survival in patients with serious ovarian and colorectal cancers. Several IDO inhibitors have been reported, so far. 1-Methyltryptophan (1-MT) is the most frequently used inhibitor and currently in clinical trials in the US.

In order to identify new chemical entities for IDO inhibitors, we developed a cell-based HTS system, which measures the abilities of test compounds to inhibit the kinurenine production in human epithelial carcinoma A431 cells. We screened our chemical library using this system and identified several hit compounds which directly inhibited the enzymatic activity of IDO. Some of those contain S-benzylisothiourea scaffold in common. In the western blotting analysis, those S-benzylisothiourea derivatives suppressed kunurenine production without reduction of the IDO expression. The Lineweaver-Burk plot was found to be consistent with noncompetitive inhibition mode unlike the competitive inhibitor, 1-MT. Now, further analysis of the mode of action of S-benzylisothiourea derivatives is in progress. We also found that the antihypertensive agent candesartan cilexetil inhibits IDO. The structural modifications provided a >10-fold more potent inhibitor. Structure–activity relationship and docking studies suggested that candesartan analogues uniquely bind to the entrance of the active site in IDO, and not to the haem region. We further screened our chemical libraries using a cell-based Kyn production assay to identify a new type of small molecules that regulate Kyn production, and for the first time identified a benzenesulfonamide derivative as a hit with the ability to inhibit Kyn production in interferon-γ (IFN-γ)-stimulated A431 and HeLa cells. Unlike the previously identified S-benzylisothiourea derivative, benzenesulfonamide derivatives had little effect on the enzymatic activity of recombinant human IDO in vitro but suppressed the expression of IDO at the mRNA level in cells. Furthermore, the compound suppressed STAT1-dependent transcriptional activity and DNA binding, whereas no decrement in either the expression or phosphorylation level of STAT1 was observed. The inhibition of IDO expression by several benzenesulfonamide derivatives is associated with the suppression of STAT1. Thus, benzenesulfonamide derivatives might be useful for analyzing the regulation of IDO activation, and STAT1-targeting could be an alternative to the IDO-directed approach for the regulation of Kyn levels by small molecules in the tumor microenvironment.

KSP inhibitors
Antimitotic agents such as taxol, epothilone, and vinca alkaloids have found clinical utility as cancer chemotherapeutic agents. These agents bind to the tubulin protein a component of the mitotic spindle microtubules. Although microtubules play important roles throughout the cell cycle of cancer cells, disruption of microtubule dynamics also produces undesirable side effects, such as toxicity in non-dividing cells like peripheral neurons, leading to peripheral neuropathy in patients. The microtubule associated kinesin spindle protein KSP (also known as Eg5, Kif11 and kinesin-5) plays an important role in the early stages of mitosis. It is responsible for the formation and maintenance of the bipolar spindle. Inhibition/knockdown of KSP leads to cell cycle arrest during mitosis and causes cells with a monopolar spindle, so called monoasters. Because KSP is not expressed in post-mitotic neurons and is likely to act only in dividing cells, its inhibitors might provide better specificity than microtubule inhibitors in the treatment of human malignancies. In 1999, Monastrol, the first small-molecule inhibitor of KSP, was discovered in a phenotype-based screening. Since then, a number of KSP inhibitors have been discovered and some of them appeared as the therapeutic potential for anticancer drugs.

In order to find a new structural class of KSP inhibitors, we screened our chemical library using in vitro ATPase assay which could measure the test compound’s ability to prevent the ATP hydrolysis by KSP. As the results of our screening campaign, we identified two kinds of inhibitors, S-trityl-L-cycteine (STLC, 1a) and 4-(4-tert-butylphenyl)pyridine (TBPP, 2a). In the KSP ATPase assay, IC50 values of those hits are 1.8 µM and 1.0 µM, respectively. Furthermore, those hits showed selective inhibition against KSP over other kinesins such as CENP-E, Kid, MKLP-1 and KIF-4, and induced typical “monoastral” phenotype in HeLa cells. Thus, those two chemo-types were considered to be important as lead compounds and we decide to explore the chemical space around those leads for discovering more potent anticancer drug candidates.

Our initial exploration studies of STLC (1a) reached to identify several derivatives such as MeO-STLC (1b) and CF3-STLC (1c) those of which were 7-10 fold more potent than STLC in KSP ATPase assay and HeLa cells cytotoxity assay. In order to analyze the molecular interaction between STLC derivatives and cellular KSP, we designed and synthesized compound 1b-immobilized affinity beads (1d). The affinity beads 1d have the ability to capture KSP efficiently from the mixture of proteins in HCT116 cell extract in pull-down experiments. When HCT116 cells were pre-treated with compound 1c, capture of KSP by affinity beads 1d in cell extract was effectively inhibited, although no decline of KSP expression was seen. Thus, our analysis using drug-immobilized affinity beads provide the first direct evidence of the interaction of STLC derivatives with KSP in cells and demonstrated that 1dis a useful tool for analyzing the mode of action of KSP inhibitors.

Furthermore, effect of compound 1c on 18 human hematological cancer cell lines was investigated and K562 human chronic myeloid leukemia cell line was shown to be one of the most sensitive to compound 1c. Treatment of K562 cells with 1c induced mitotic arrest of the cell cycle with the appearance of characteristic monoastral spindles, subsequent apoptotic cell death and cleavage of PARP-1, caspase 3, and 4E-BP1. The wide ranging caspase inhibitor z-VAD fmk prevented the cleavage of caspase-3 and 4E-BP1, but failed to attenuate PARP-1 cleavage or cell death triggered by 1c. These results suggest that 1ccan induce apoptotic cell death in a caspase-independent manner, and could work effectively as an anti-cancer agent for human hematological malignancies.

Another exploration study around TBPP (2a) was also performed to indentify more potent KSP inhibitors. Typically, Suzuki-Miyaura coupling for the corresponding arylhalide/triflate and arylboronic acid/ester was employed to afford desired derivatives. Among diverse of the derivatives, compound 2b was 8.3 fold more potent than initial lead 2a in KSP ATPase assay (IC50: 0.12 µM). In addition, unexpectedly this compound showed unique biological effects on KSP inhibition. In contrast to STLC derivative 1b, compound 2b inhibits KSP in a microtubule-dependent manner in the biochemical ATPase assay in vitro. Furthermore, we investigated the dynamics of chromosomes, microtubules and KSP in HeLa cells by treatment with 2b or STLC derivative 1b. As we expected, compound 2b induced the “monoastral” phenotype, which is typically observed under KSP inhibition. Surprisingly, KSP distribution in mitotic arrested HeLa cells was quite different between compound 1band compound 2b, despite the same localization of chromosomes and microtubules for both KSP inhibitors. Further structural optimization studies and evaluation of the derivatives as anticancer drugs including in vivo studies are ongoing.