March 15, 2007 (Vol. 27, No. 6)
Identifying Drug Resistance Patterns and Expanding the Druggable Target Space
Researchers are combining a range of techniques for analyzing key cell signaling pathways in drug discovery and therapeutic development, according to presentations at Cambridge Healthtech’s “Molecular Medicine Tri Conference” in San Francisco and a Keystone Symposium on “PI 3-Kinase Signaling Pathways in Disease”, held in Santa Fe.
Investigators described using cell-signaling pathway analysis to characterize drug-resistance patterns in cancer cells, identify key signaling pathways impacted by specific drug treatments, and expand the druggable target space.
A common theme in the presentations was the application of molecular analysis approaches combining several technologies, such as gene array studies coupled with protein expression and the use of computer programs and computational databases to analyze the results.
Several investigators stressed the need to go beyond single-target-based drug discovery, such as cell-surface-receptor drug targets, and identify key downstream proteins involved in the pathway responses to drug-target interactions.
Platinum-resistant Ovarian Tumors
William Ricketts, Ph.D., director of research for Oncotech (www.oncotech.com), described studies aimed at identifying specific pathways in ovarian cancer cells that were altered in cisplatin-resistant tumor cells. He cited the ongoing need for the development of sophisticated diagnostics to predict the responses of individual tumors to chemotherapy, pointing out that the initial clinical response to platinum is a major determinant of clinical outcome for patients with ovarian cancer.
“Clinical investigators have shown that patients with tumors determined to be resistant in vitro to platinum compounds are at a greatly increased risk for disease progression and death when treated with standard platinum-based regimens. The determination of drug-resistance patterns for a given tumor is important for selecting the most effective therapy and minimizing side effects from drugs to which tumors are unresponsive,” said Dr. Ricketts.
The Oncotech research team has access to primary human tumors (both cryopreserved and paraffin-embedded) and the drug-resistance data based on the company’s Extreme Drug Resistance™ assay, which has been offered clinically for almost 20 years. Researchers selected tumor cells for this study that had been identified as cisplatin resistant using the Extreme Drug Resistance Assay, a clinically validated assay that Oncotech says is 99% accurate in determining tumor-cell drug resistance.
The assay separates cancer cells into one of three groups—extreme drug resistant (EDR), intermediate drug resistant (IDR), or low drug resistant (LDR). The ovarian tumors studied by Dr. Ricketts were from previously untreated patients and had an LDR response to all other drugs tested. The scientists analyzed cell-signaling pathways using applied gene array technology, PathArt analysis (a software analysis tool that classifies genes into pathways developed by Jubilant Biosys (www.jubilantbiosys.com), and In Cell Westerns (a high-throughput detection system developed by Li-Cor Biosciences (www.licor.com) for simultaneous analyses of protein expression and phosphorylation state to conclude that differences in signal-transduction pathways are present at the gene and protein level in human tumors with differing cisplatin responses.
They further found that proteins with altered expression tend to be “intermediate signaling molecules”, such as Erk, Akt, b-catenin, and STAT3, possibly indicating that the significant determinant in drug resistance is not how the cell “sees” the signal but how it “interprets” it and responds through signal amplification or altered gene expression.
Oncotech’s ultimate goal, Dr. Ricketts, explained, is to develop molecular diagnostics that go beyond the EDR soft agar assay and reach into the signaling pathways that contribute to drug resistance by their activation and underlying genetics of resistance. He pointed out, for example, that diagnostics to determine treatment can detect overexpression of a drug or antibody target such as the EGF receptor, but yield no information about whether pathways regulated by the target that may also influence drug response are activated.
“Activated pathways that are altered in resistant tumors may provide biomarkers of drug resistance but may also provide second-line indication for many targeted therapies,” Dr. Ricketts said.
Mab-therapy Pathway Analysis
Chris Huang, Ph.D., principal research scientist, discovery research, Centocor Research and Development (www.centocor.com), used global gene-expression profiling in xenograph tumor models that had responded to Centocor’s antitumor therapeutic monoclonal antibodies. These investigators were able to identify key signaling pathways involved in treatment responsiveness and antiproliferation mechanisms.
“We use microarray technology to capture global gene expression in tumors in a ‘snapshot’ fashion. Using pathway analysis then gives us an understanding of what kinds of changes we are looking at over time,” Dr. Huang said. “We found out that cell-cycle control pathways were negatively impacted even at 24 hours after a single-dose treatment. Cell-cycle controls do not function normally in cancer cells. If we see inhibition, we know our antibodies are doing something. However, these changes are often too subtle to be identified this early with other technologies. We have found global gene-expression profiling powerful in identifying pathway changes. If you have seen 20 genes up- or down-regulated along a particular pathway, even though each change is relatively small, with statistical support you can put them together and get a fuller picture of the treatment response at molecular level.”
Other examples Dr. Huang showed were RAS-signaling pathway, complement pathway, and ERBB family pathways. All had significant changes in response to the treatment in their in vivo model. Together, theses pathways paint a comprehensive picture of mechanism of action of their antibody therapy.
P13-kinase Isoform Specificity
TargeGen (www.targen.com) is fo-cused on developing small molecule kinase inhibitors that address diseases marked by changes in vascular permeability and angiogenesis. John Doukas, Ph.D., senior director of pharmacology, described the company’s progress in identifying an inhibitor of two phosphoinositide 3-kinase PI3K isoforms, PI3K gamma and delta. According to Dr. Doukas, while some PI 3Ks can promote cell survival during tissue ischemia, these isoforms contribute to the inflammation and subsequent tissue damage that occur upon reperfusion.
Dr. Doukas pointed out the functional differences of these kinase isoforms, explaining that “in knockout animal models that don’t express the gamma or delta enzyme isoforms, animals are resistant to inflammation. At the same time, other isoform knockouts, such as alpha and beta, are lethal because these are involved in cell proliferation and survival.
“We needed to come up with small molecules that were specific isoform inhibitors targeted at the gamma and delta forms.” The company identified TG100-115 as a potential selective inhibitor through screening a novel chemical family and molecular modeling studies to define isoform specificity.
This compound was selected based on its potential to inhibit edema and inflammation in response to multiple mediators that participate in myocardial infarction (MI), including VEGF and platelet-activating factor, without disrupting endothelial cell mitogenesis, a repair process required for post-MI tissue survival.
Dr. Doukas and his team tested the compound in rigorous animal myocardial infarction models to determine whether, if administered up to three hours-post infarction, it could produce cardioprotection and prevent post-ischemic damage. “We used extremely aggressive animal models that mimic clinical presentation of MI in humans, dosing animals after both ischemia and reperfusion injury,” said Dr. Doukas.
When animals were dosed up to three hours after myocardial reperfusion, the time period in which therapeutic intervention is possible in patients, the drug candidate reduced infarct development and preserved myocardial function. The company is testing TG100-115 in a double-blind, dose-escalating Phase I/II clinical study.
The Reactome
The Reactome is an open-source, curated online database of human biological pathways developed by scientists at The European Bioinformatics Institute, Cold Spring Harbor Laboratory, and The Gene Ontology Consortium.
Peter G. D’Eustachio, Ph.D., editor-in-chief of Reactome and associate professor of biochemistry and medicine at New York University, says that the Reactome provides both an online database and an infrastructure for computation across the biological reaction network, such as the ability to link large data sets with pathway-specific signaling patterns.
“Our goal is to represent as much human and cellular biology as possible,” said Dr. D’Eustachio. “The Reactome now describes the functions of 8% of known human proteins. Our goal for the next four years is to get to 50% of proteins with a significant depth of coverage for each one.”
The Reactome takes human biological processes and pathways and by expert manual curation reduces them to the “kind of classic type of metabolic chart that we are used to seeing,” said Dr. D’Eustachio. He said that the database now describes pathway input and output molecules, catalysts, intracellular sites where specific reactions occur, and the reactions that both precede and result from these.
An Online Textbook
Dr. D’Eustachio envisions that scientists will use the Reactome as an on-line textbook and as a particularly useful tool for analyzing large data sets in the context of specific pathways.
“If you have a high-throughput data set, such as a cellular RNA expression profile that resulted from treating cells with a kinase inhibitor, the profile data can be superimposed on our chart and each reaction in the chart will be colored to indicate whether the RNAs encoding proteins involved in that reaction are up-regulated or down-regulated.”
The Reactome’s Skypainter utility is of particular value to pathway analyzers as it highlights points on a reaction map based on sequences and other identifiers.
In the domain of kinase pathways, signaling via the insulin receptor is annotated and linked via its downstream effects on transcription to a number of targets within cellular energy metabolism. Work now under way will extend this annotation to EGFR, PDGF, NGF, FGFR, and other receptor tyrosine kinase-signaling pathways.
The database also contains preassembled data sets, such as all proteins associated with genetic diseases related to Mendelian inheritance. The Reactome is accessible at www.reactome.com.