New Earth BioMed - Cancer Research for the Common Good

Activity report for March-June, 2011

JohnBoik — Tue, 09 Aug 2011 00:49:30 +0000

Read about our newest research results at New Earth BioMed. A summary of our work with natural compounds and cancer, as well as our laser bioassay work with CARS is now available. A pdf file of the March-June, 2011 activity report can be found here.

HerbalEGram interviews Dr. Boik

JohnBoik — Thu, 03 Feb 2011 20:38:18 +0000

HerbalGram, the newsletter for the American Botanical Council, interviewed Dr. Boik for its February 2011 issue. The article is entitled: A Botanical Look at the Changing Face of Drug Discovery and Development.

Read more at:
American Botanical Council

Dr. Boik interviewed by Dr. Susan Kolb on Temple of Health Radio

JohnBoik — Thu, 30 Dec 2010 23:01:25 +0000

John Boik, PhD, President of New Earth BioMed was interviewed on the Temple of Health radio show by Dr. Susan Kolb on December 11, 2010. This one-hour show is available by download.

Dr. Boik interviewed on WOR-AM radio, NYC

JohnBoik — Thu, 05 Aug 2010 00:24:00 +0000

John Boik, PhD, President of New Earth BioMed was interviewed on WOR-AM news radio New York City on the August 6 show Health Talk with Dr. Ronald Hoffman. The syndicated one-hour show is available via podcast on the WOR web site.

New Earth BioMed launches natural-product-based cancer drug discovery program

JohnBoik — Thu, 29 Jul 2010 01:22:23 +0000

Irvine, California August 4, 2010 – New Earth BioMed, a new 501(c)3 nonprofit corporation, has launched a long-term drug discovery program to identify mixtures of nontoxic plant compounds useful in treating cancer. An estimated 77 percent of approved cancer and anti-infective drugs in the United States are either natural products or are derived from them. Of the tens of thousands of bioactive natural products that exist, New Earth BioMed is identifying those that act synergistically in mixtures to inhibit cancer.

“This new screening program is representative of a new breed of drug discovery initiatives brought about through an evolution in biological understanding.” said New Earth BioMed President John Boik, Ph.D. “Recent advances in bioinformatics, analytical technology and other related fields are strongly pointing to the concept that activities of a cell are dictated by a complex and dynamic internal web of protein interactions.” Boik said the so-called “silver bullet model” of drug development that reigned supreme over the last 20 years is giving way to a new model that better embraces the complexities of life. For complex chronic diseases such as cancer, this new model, based on systems biology, promises more effective and safer drugs, Boik said. “Our task at New Earth BioMed is to discover mixtures of botanical compounds that affect the protein network as a whole, not just a single protein in the network.” Boik said. “It is like stepping out of the forest to see the trees.”

New Earth BioMed is financed by donations from individuals, corporations, and foundations, as well as public grants. President John Boik is the author of two books on natural products and cancer, has a doctorate from the University of Texas/M.D. Anderson Cancer Center, and has done postdoctoral work at Stanford University. For more information, visit the New Earth BioMed website at www.newearthbiomed.org.

###

Types of Chemotherapy Drugs

JohnBoik — Thu, 29 Jul 2010 01:21:26 +0000

The current arsenal of chemotherapy drugs consists primarily of cytotoxic agents, targeted therapy drugs, sex hormones, and immunostimulants. Each are discussed briefly below. The actions of some cytotoxic drugs are cell-cycle specific, and so we start with a quick review of cell division.

Most cells in adults do not divide frequently. Cell division is only necessary to repair wounds or in a few tissues that have unusually high turnover or replication rates. These tissues include the intestinal lining, bone marrow, and hair follicles. The intestinal lining protects the gut, and the bone marrow produces several cells, including red blood cells and white blood cells (immune cells).

When cells do need to divide, they do so in a tightly regulated process called the cell cycle. Once a cell enters the cycle, it takes about 24 hours to complete division, depending on cell type. Cells of some tissues take much longer than others. The cell cycle, which is illustrated in Figure 1↓, can be broken down into five phases:

G1: The G1 (gap) phase begins when the cell is first stimulated to proliferate. During G1 cell contents other than the DNA are duplicated and the volume of the cell increases. At about 8-10 hours into G1, the cell crosses a restriction point, or a point-of-no-return. If the cell senses that it is capable of division, it passes this point and then cannot stop the replication process.
S: The S phase is where DNA synthesis takes place. Many cell-cycle-specific drugs act only on cells that are in the S phase. These drugs interfere with DNA synthesis in some way.
G2: During the G2 (gap) phase, the cell arranges chromosomes for later division (mitosis). At the G2/M checkpoint, a cell either continues with mitosis or, if DNA was not synthesized properly, it undergoes apoptosis (programmed cell death). It is better for the body if a damaged cell dies rather than having it pass on altered DNA to daughter cells. Cancer cells learn to ignore or override the suicide program, and therefore can divide even with malformed DNA.
Mitosis: The cell physically divides in a process called mitosis. A copy of the DNA from the parent is passed along to the daughter cells.
G0: Following mitosis, the cell can enter the G0 (resting) phase, which is outside the cell cycle proper. In G0 the cell is not actively proliferating. Cells can spend long periods in G0 before they reenter the cell cycle in G1. Cell-cycle-specific drugs cannot kill cancer cells that are in the resting phase. Typically, cells in G0 are termed quiescent.

Figure 1 The cell cycle

1 Cytotoxic drugs

Cytotoxic drugs inhibit the proliferation of cancer cells, usually by interfering directly or indirectly with DNA replication. Most of these drugs were developed during the period 1950 to 1990. They target cells that proliferate frequently, but in so doing can affect normal cells that also proliferate frequently. These include bone marrow cells, cells of the gut lining, and hair follicles, as discussed above. Damage to these cells leads to the common side effects of immune suppression, nausea, and hair loss.

It is common to administer cytotoxic drugs in combinations, or cocktails, containing two to five drugs. Such combinations are generally more effective than single drugs. Note that drugs used in combination chemotherapy were first approved by the US FDA as single agent therapies. It is only after they were approved that researchers examined their effects in combinations. This is in contrast to the potentially more effective approach of designing combinations in the early (preclinical) phase of drug development, as New Earth BioMed is attempting to do. Designing combinations in the preclinical phase allows greater flexibility in choosing components, which can allow greater control of the biologic effects. Note that a large majority of approved cytotoxic drugs (~77 percent) derive either directly or indirectly from natural compounds.

There are several ways to categorize cytotoxic drugs, for example by their chemical structure, method of action, or source. Common categories include alkylating agents, antimetabolites, anti-tumor antibiotics, topoisomerase inhibitors, mitotic inhibitors, corticosteroids, and differentiating agents. All are discussed below.

1.1 Alkylating agents

Alkylating agents directly damage DNA, thereby preventing cell division. For example, some can insert themselves into the DNA double helix, thereby providing a physical block to the replication of DNA. Alkylating agents may be further categorized into cell-cycle-specific and non-cell-cycle-specific drugs. As their title implies, cell-cycle-specific drugs only act during a specific phase of the cell cycle. Alkylating agents include mechlorethamine, cyclophosphamide, melphalan, carmustine, busulfan, dacarbazine, and thiotepa. The platinum drugs (cisplatin, carboplatin, and oxaliplatin) are sometimes grouped with alkylating agents.

1.2 Antimetabolites

Antimetabolites interfere with DNA and RNA synthesis by causing malformed nucleotides, the building-blocks for DNA and RNA. These agents damage cells during the S phase. Examples of antimetabolites include 5-fluorouracil, 6-mercaptopurine, methotrexate, gemcitabine, and cytarabine.

1.3 Anti-tumor antibiotics

One class of anti-tumor antibiotics is called anthracyclines. These interfere with enzymes involved in DNA replication and are not cell-cycle-specific. Heart toxicity is a common adverse effect. Examples include daunorubicin, doxorubicin, bleomycin, and mitomycin-C. The drug mitoxantrone is an anti-tumor antibiotic, but also works by inhibiting the enzyme topoisomerase II, which is used during DNA unwinding (DNA must be unwound prior to replication).

1.4 Topoisomerase inhibitors

As just mentioned, DNA must be unwound prior to replication. Several drugs inhibit the enzymes (topoisomerase I and II) that perform this function. Mitoxantrone has already been mentioned. Other topoisomerase II inhibitors include etoposide and teniposide. Topoisomerase I inhibitors include topotecan and irinotecan.

1.5 Mitotic inhibitors

Many of the mitotic inhibitors derive from plant alkaloids. As their name implies, they interfere with the mitosis phase of the cell cycle. They can also have effects in other phases. A common adverse effect is peripheral nerve damage. Mitotic inhibitors include taxanes such as paclitaxel and docetaxel, and also vinca alkaloids such as vinblastine and vincristine.

1.6 Corticosteroids

Certain steroids (hormones and hormone-like drugs) can be used to inhibit some types of lymphoma, leukemias, and multiple myeloma. Example steroids include prednisone and methylprednisolone.

1.7 Differentiating agents

The last category of cytotoxic drugs to be discussed is differentiating agents. While not formal cytotoxics, differentiating agents cause cancer cells to “mature” into cells that are more normal. Many cancer cells have an immature phenotype, which allows them to proliferate frequently. Example differentiating agents include the retinoids and arsenic trioxide.

2 Targeted therapy drugs

Targeted therapy drugs are designed to inhibit a specific protein involved in cancer progression. Target proteins are identified during the drug discovery process as being different in some way in cancer cells versus normal cells.

Only a handful of targeted therapy drugs have been approved so far. These include imatinib, gefitinib, erlotinib, sunitinib, and bortezomib. Imatinib inhibits an enzyme (BCR-ABL protein, a tyrosine kinase) that plays a role in chronic myelogenous leukemia and gastrointestinal stromal tumors. Both gefitinib and erlotinib are epidermal growth factor receptor (EGFR) inhibitors. EGFR is a type of receptor tyrosine kinase (RTK), which are a family of proteins involved in signal transduction. Sunitinib targets multiple RTKs. Bortezomib is a proteasome inhibitor. The proteasome degrades unneeded proteins, and proteasome inhibition may prevent degradation of pro-apoptotic factors, permitting activation of apoptosis (programmed cell death) in cancer cells.

3 Sex hormones

Certain sex hormones, or hormone-like drugs, can be effective against breast, prostate, and some other hormone-dependent cancers. They prevent a cancer cell from using the hormone as a growth factor, or prevent the body from making the hormone. Examples include estrogens, anti-estrogens, tamoxifen, anastrozole, progestins, and leuprolide.

4 Immunotherapy

Immunostimulants can be effective for some types of cancer. They act by stimulating the body’s own immune system or are antibodies or other immune system components made outside of the body. Examples include interleukin-2 (IL-2) and alemtuzumab. IL-2 is an immune system signaling molecule that plays an important role in the body’s response to microbial infection. Alemtuzumab is a humanized monoclonal antibody indicated for the treatment of Chronic lymphocytic leukemia.

Lasers and CARS

JohnBoik — Thu, 29 Jul 2010 01:20:11 +0000

A core technology in the New Earth BioMed mixture discovery program is a laser-based analytical tool based on choerent anti-Stokes Raman scattering (CARS) [1,2,3]. This article describes some technical concepts behind Raman scattering and CARS.

Scattering is a process by which molecules cause an incident photon to change direction. Scattering can be elastic, where the photon does not gain or loose energy to the molecule, or it can be inelastic, where the photon gains or looses energy. Scattering is most efficient at the shorter (blue) wavelengths of the visible spectrum.

A very common form of elastic scattering is Rayleigh scattering. The sky appears blue to us because of Rayleigh scattering. Sunlight hits water or other molecules in the atmosphere, and some of that light, particularly at the blue wavelengths, is scattered.

Raman scattering is inelastic and occurs when incident photons interact with the molecule, altering its vibrational or rotational energy. Here vibrational refers to vibrations of chemical bonds. Each type of bond (C-H2 for example) has a unique resonant frequency.

There are two ways that Raman scattering can occur. The most common is when photons are absorbed by a molecule that is initially in its low-energy ground state. The photon raises the molecule’s energy to a virtual state. Soon afterward the molecule relaxes down to a lower-energy state, but not all the way to the ground state. Because the molecule contains more energy than it started with, the photon given off during relaxation contains less energy (its frequency is reduced). This shift in frequency from that of the incident photon is called a Stokes shift.

A less common form of Raman scattering produces an anti-Stokes shift. Here, the molecule is initially in a slightly elevated state and the incident photon increases its energy level still higher. Upon relaxation, the molecule returns to the ground state, which is less energetic than its initial state. As such, the photon given off during relaxation contains more energy than the incident photon (its frequency is higher). Spontaneous Raman scattering is far less probable than Rayleigh scattering, and the anti-Stokes shift is far less probable than the Stokes shift.

The Stokes and anti-Stokes shifts are unique to molecular bonds. In this way, Raman scattering, particularly laser-stimulated Raman scattering, can provide a chemical (spectral) fingerprint of bonds in the affected molecules. It can be used to identify molecules within a sample.

Stimulated Raman scattering is a very weak process. As such, it requires strong laser light and long exposure times to generate a quality spectrum. This presents problems for using Raman scattering in a screening assay. Lasers cannot be so strong that they damage live cells, and exposure times cannot be so long that they make screening impractical. CARS offers a faster alternative to Raman scattering.

In CARS, two lasers beams are focused on a tiny area, smaller than the size of a cell. The first laser beam (the pump beam, at high frequency) excites molecules in the focal area from a ground state to a higher energy virtual state. A second laser beam (the Stokes beam, at lower frequency) vibrates the excited molecule in such a way as to increase the probability that it will relax to a specific lower energy state. As a rough analogy, shaking an unsteady pile of books at the right frequency can increase the probability that it will crumble into a pile of a certain height. A probe beam, usually of the same frequency as the pump beam, then excites the molecule to a higher virtual state. From here the molecule relaxes almost instantly to the ground state, giving off a photon at the anti-Stokes frequency.

In practice, the frequency of the Stokes beam is tuned so that the low energy state matches the natural frequency of the molecule (i.e., it is tuned to the natural vibration of a particular bond of the molecule). When tuned in this way, the anti-Stokes signal is greatly amplified, allowing fast scanning and making the process viable as an analytical tool. Energy diagrams for Raman scattering and CARS are illustrated in Figure 1↓.

Figure 1 CARS energy diagram

Bibliography

[1] Chan, J and Fore, S and Wachsmann-Hogiu, S., Huser, T. Raman spectroscopy and microscopy of individual cells and cellular components. Laser \& Photonics Review, 2008;2(5):325-349. URL: http://dx.doi.org/10.1002/lpor.200810012.

[2] Muller, M., Zumbusch, A. Coherent anti-Stokes Raman Scattering Microscopy. Chemphyschem, 2007;8(15):2156-70. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=17768730.

[3] Rodriguez, L. G., Lockett, S. J., Holtom, G. R. Coherent anti-stokes Raman scattering microscopy: a biological review. Cytometry A, 2006;69(8):779-91. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=16752420.

Roadmap Chapter 1

JohnBoik — Thu, 29 Jul 2010 01:09:30 +0000

Executive Overview

Who We Are:

New Earth BioMed (NEBM) is a 501(c)3 tax-exempt nonprofit corporation dedicated to the discovery of useful cancer drugs based on mixtures of safe natural products.

Why We Are Different:

The NEBM mixture discovery program is based on emerging technologies as well as on emerging biological theories. To our knowledge, no other drug discovery company is following a similar approach. In particular, the following characteristics set us apart:

Our focus is on natural products, including semi-purified extracts.
Our focus is on distributed polypharmacology, which employs large mixtures of natural products.
We employ next-generation laser-based bioanalytical equipment and other sophisticated technologies to generate screening data for thousands of natural product mixtures.
We rely on mathematical predictive models to guide mixture screening and make it more efficient.
Our in-vitro screening program employs three-dimensional cell co-cultures, which better mimic actual conditions found in humans
Because we are a 501(c)3 corporation, we have flexibility to study natural compounds in ways that are not practical in industry. That is, we are not forced to focus on high-profit programs.

Why We Need Your Help:

Direct public support for the New Earth BioMed research program is critical for its success. We pursue grant opportunities, but grant funds are limited and awards are by no means assured, even for well designed projects. Your continued support provides us with the security and ability to initiate and sustain innovative projects that are vital to reach our long-term goals.

The Hope:

The hope and goal of our efforts is safer and more effective anticancer drugs based on natural products. Successful discovery, development, and market approval of natural product mixtures could have wide-ranging positive effects on patient care and safety, pharmaceutical science, and medical agriculture. To learn more about our research and how it could help change the way we think about, design, and use drugs for treating cancer, please follow this Roadmap through the chapters below. Additional summaries are also available for the laser bioassay and mixture screening programs.

We sincerely thank you for your interest and support,

John Boik, Ph.D., President, New Earth BioMed

Chapter 1	Introduction
Chapter 2	History of Cancer Drug Discovery
Chapter 3	Overview of Network Pharmacology
Chapter 4	Mixture Discovery Program

1 Introduction

New Earth BioMed (NEBM) is a 501(c)3 nonprofit corporation dedicated to the discovery of useful therapeutic agents from natural products. Contingent on adequate funding, this roadmap describes an ambitious long-term drug discovery program focused on identifying personalized mixtures of therapeutic agents that are effective against cancer. Mixtures will consist primarily or solely of natural products, will contain numerous components, and will be designed for oral administration. The basis for this program is rapidly mounting evidence that suggests:

The emerging field of systems biology offers an approach to drug discovery that could lead to safer and more effective cancer therapies [13]. In systems biology, molecular components of a cancer cell, as well as the cell’s interaction with its local environment, are viewed as a dynamic interconnected system. This requires addressing the complex protein-protein, protein-lipid, and signaling transduction networks within a cell, and also cell-matrix and cell-cell communications between cancer cells and their surrounding environment. In contrast to typical drug discovery programs, which seek to alter the function of a single target protein with a single agent, the goal here is to use multiple agents to therapeutically alter the dynamics of the whole system. In most chronic diseases, including most cancers, a wide variety of genetic and epigenetic cellular abnormalities occur that serve to support the disease state. In this complex setting, targeting a single protein may be insufficient to produce an optimal and lasting therapeutic effect [10,11].
Polypharmacology is a promising approach for affecting biological systems. In polypharmacology, (sometimes called network pharmacology), a single drug or a mixture containing multiple drugs is designed to affect multiple protein targets [22]. Typically, mixtures contain two or a few drugs. At NEBM we focus on an extreme variant of polypharmacology, which we call distributed polypharmacology (DPP). DPP is characterized by the use of large mixtures of agents (dozens of active constituents) to safely shape system dynamics [12]. Effects are mediated in part though modest-affinity binding to a diverse set of protein targets [16,6,20]. The use of large mixtures in DPP offers four main advantages:
1. Therapies can be made more robust to genetic and epigenetic variations in cancer cell populations. Robustness is accomplished through redundancy of chemical structure (several components of a mixture may have a similar structure and bind to similar sites on a protein), as well as redundancy in targeting a selected signaling pathway (several proteins in a pathway may be targeted). Therapeutic robustness is intended to overcome the biologic robustness of cancer cells. Biologic robustness stems from genetic instability, the ability to use alternative signaling pathways, the hijacking of critical cellular processes, and other characteristics of cancer cells [12].
2. Large mixtures contain a sufficient diversity of compounds to allow targeting of multiple signaling pathways in a cancer cell, as well as interactions between a cancer cell and its environment (cell-cell and cell-matrix communications) [5,21]. Targeted pathways can include those directly related to cell proliferation, as well as others that support cancer progression by less direct means. For example, signaling pathways that govern drug resistance could be targeted, which could lead to mixtures that are less subject to drug-induced resistance [24,23,29,18]. Multidrug resistance is the curse of modern day cancer therapy.
3. The flexibility offered by large mixtures, and the additive and synergistic interactions they produce, facilitates fine-tuning of the safety profile. Doses of any individual constituent can be kept low, and by accounting for overlapping toxicities the risk of adverse effects can be minimized. Recent in-vitro studies suggest that synergism between multiple drugs can be used to better focus a therapeutic aim, while avoiding synergistic adverse effects [14,15].
4. In combination with genetic assessment of tumor tissue and monitoring of pharmacokinetic, toxicological, and cancer-related biomarkers, the use of large mixtures allows fine-tuning of a therapeutic effect for a specific patient. A recent clinical study suggests that biomarkers can be sucessfully used to guide cancer therapy [17].
Natural products offer a vast, underutilized library of candidate compounds that has been enriched through evolution for bioactivity. Proteins in plants and animals consist of a limited number of three-dimensional folds (shapes), and nature selects for proteins that bind to small molecules and vise versa. For example, under stress from pathogens, grapevine (Vitis spp.) produces resveratrol, along with a series of structurally-related compounds, which inhibit growth in fungi [1]. In addition, resveratrol binds to a number of proteins in humans, producing beneficial effects [28]. Considering cross-species bioactivity, it is not surprising that roughly 77 percent of approved cancer and anti-infective drugs in the United States are either natural products, are semi-synthetically derived from them, or are synthetically produced based on a natural product structure [19].

Tens of thousands of natural products are likely to inhibit cancer cell proliferation in vitro, although not all of these would be suitable for inclusion in a mixture due to pharmacokinetic, toxicological, or other concerns [3]. When screening for new orally administered cancer drugs, the pharmaceutical industry typically favors compounds that bind with high affinity to target proteins and that exhibit favorable pharmacokinetic characteristics (such as good oral absorption). Unfortunately, many otherwise interesting natural products bind with only modest affinity to proteins and/or exhibit problematic pharmacokinetics.

Curcumin (found in the spice turmeric) is a good example. It inhibits the proliferation of many types of cancer cells in vitro by several different molecular mechanisms, but does so at concentrations of about 2 to 40 μM (i.e., it is moderately potent). This concentration is about 100- to 1,000-fold higher than that of typical anticancer drugs. In addition, like many natural compounds, it undergoes extensive metabolism in the gut and liver. In particular, glucuronide (sugar-like) molecules are attached to the compound to make it more water soluble and therefore easier to excrete. As such, only a small fraction of curcumin in the blood is in its unchanged form [2]. Most is in the glucuronide form, which is less active. Compounds such as curcumin have potential in medicine, except for these problems.

The mixture development program at NEBM addresses both potency and pharmacokinetic issues. Potency is addressed through use of additive and synergistic mixtures. In synergism, the effect of the whole is greater than the sum of the parts. Distributed polypharmacology maximizes the biologic effects of natural compounds, including compounds that bind with modest-affinity [16,6,20]. Mixtures are designed to produce small effects on many protein targets, rather than a large effect on one target.

Pharmacokinetic issues are addressed in part through use of nanoparticle delivery vehicles. These vehicles can improve oral pharmacokinetic parameters, as well as guide drugs to the tumor site [26]. Nanoparticle delivery is a new but intensely studied field in cancer research. Use of DPP with nanoparticle delivery overcomes some of the major problems associated with natural products and greatly expands the library of natural compounds that could beneficially be used.

Nature shows us that mixtures of compounds can increased biological efficacy. For example, in response to pathogens the medicinal plant golden seal (Hydrastis canadensis) secretes the antimicrobial alkaloid berberine, as well as 5’-methoxyhydnocarpin (5-MHC). 5-MHC inhibits a multidrug resistance protein in bacteria that pumps toxins like berberine out of the cell. The compound strongly potentiates the action of berberine by preventing its efflux [25]. Hydrastis spp. produce other antimicrobial compounds in addition to berberine [9], and these too may benefit from 5-MHC activity.

Mixtures are also widely used in Traditional Chinese Medicine (TCM), Ayurveda, and other ancient forms of medicine. In TCM, 4 to 12 herbs are typically used in a formula, with each herb containing perhaps several to several dozens of bioactive constituents. The value of such formulas can be gauged by their wide use. In China herbal medicines account for 30 to 50 percent of total medicinal consumption. Worldwide, the figure is about 20 percent [27]. In Japan, medical insurance covers Kampo, a national form of herbal medicine. A recent survey suggests that 78 percent of Japanese medical doctors prescribe Kampo, up from 70 percent six years ago [8].

While the NEBM mixture discovery program is inspired by herbal medicine, there are several important differences. Chief among these is that the NEBM program is based on use of sophisticated bioanalytical and biomedical technologies in combination with complex mathematical models of cell response. Such assessment and modeling is only now becoming feasible using the newest generation of analytical and computer technologies. In addition, different raw materials are utilized. Most herbal medicines rely on ingestion of crude plant extracts. In contrast, pure agents and refined, well-characterized extracts are studied in the NEBM program.

The mixture discovery program at NEBM is designed to occur in four broad phases listed below:

Development of analytical equipment and ancillary models.
In-vitro screening using a single cancer cell line.
In-vitro screening using multiple cancer cell lines.
Additional preclinical mixture development.

Each of these is discussed in more detail in Chapter 4. Briefly, in Phase 1 a laser-based assay system is developed that will allow rapid, label-free, non-destructive, real-time monitoring of drug-induced effects that occur in three-dimensional co-cultures. The co-cultures contain cancer cells and typical normal cells present at cancer sites (fibroblasts and epithelial cells). A commercial analytical tool with these capabilities does not currently exist, and is greatly needed to advance the field [4]. The three-dimensional culture system is also developed, as well as several mathematical models to predict pharmacokinetic properties (such as oral absorption) and drug-protein binding affinity.

In Phase 2, a large number of mixtures are screened against a selected cancer cell line in co-culture. The screen will identify lead mixtures worthy of additional study. Although the co-culture system is more complex than typically-utilized one-dimensional (flat) monoculture systems, three-dimensional co-culture systems better mimic in-vivo conditions at a cancer site and are expected to produce results that are much more predictive of effects in humans [7].

Phase 3 repeats Phase 2 for multiple cancer cell lines and their genetic variants. This phase is key to developing personalized mixtures.

Phase 4 focuses on additional characterization and refinement of lead mixtures. This includes characterization and optimization of pharmacokinetics, safety, and mechanisms of action. To speed development of lead mixtures, Phase 4 can occur in parallel with Phase 3.

Mixtures deemed potentially safe and effective in Phase 4 will be prioritized for clinical trials. Clinical trials are the most expensive part of drug discovery/development and partners will be needed to complete the clinical trials process. As the mixture discovery program progresses we will be developing relationships to help us move promising mixtures into clinical trials.

Chapter 2 and Chapter 3 provide background material that help explain the NEBM mixture discovery program, and Chapter 4 discusses the program in more detail.

Bibliography

[1] Adrian, M., Rajaei, H., Jeandet, P., Veneau, J., Bessis, R. Resveratrol Oxidation in Botrytis cinerea Conidia. Phytopathology, 1998;88(5):472-6. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=18944929.

[2] Anand, P., Kunnumakkara, A. B., Newman, R. A. and Aggarwal, B. B. Bioavailability of curcumin: problems and promises. Mol Pharm, 2007;4(6):807-18. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=17999464.

[3] Boik, J. C., Newman, R. A. Structure-activity models of oral clearance, cytotoxicity, and LD50: a screen for promising anticancer compounds. BMC Pharmacol, 2008;8:12. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=18554402.

[4] Burdett, E., Kasper, F. K., Mikos, A. G., Ludwig, J. A. Engineering Tumors: A Tissue Engineering Perspective in Cancer Biology. Tissue Eng Part B Rev, 2010 URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=20092396.

[5] Chung, L. W., Baseman, A., Assikis, V., Zhau, H. E. Molecular insights into prostate cancer progression: the missing link of tumor microenvironment. J Urol, 2005;173(1):10-20. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=15592017.

[6] Csermely, P., Agoston, V., Pongor, S. The efficiency of multi-target drugs: the network approach might help drug design. Trends Pharmacol Sci, 2005;26(4):178-82. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=15808341.

[7] Friedrich, J., Seidel, C., Ebner, R., Kunz-Schughart, L. A. Spheroid-based drug screen: considerations and practical approach. Nat Protoc, 2009;4(3):309-24. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19214182.

[8] Fujiwara, K., Imanishi, J., Watanabe, S., Ozasa, K., Sakurada, K. Changes in Attitudes of Japanese Doctors toward Complementary and Alternative Medicine—Comparison of Surveys in 1999 and 2005 in Kyoto. Evid Based Complement Alternat Med, 2009 URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19465404.

[9] Hwang, B. Y., Roberts, S. K., Chadwick, L. R. and Wu, C. D., Kinghorn, A. D. Antimicrobial constituents from goldenseal (the Rhizomes of Hydrastis canadensis) against selected oral pathogens. Planta Med, 2003;69(7):623-7. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=12898417.

[10] Jia, J., Zhu, F., Ma, X., Cao, Z., Li, Y. and Chen, Y. Z. Mechanisms of drug combinations: interaction and network perspectives. Nat Rev Drug Discov, 2009;8(2):111-28. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19180105.

[11] Keith, C. T., Borisy, A. A., Stockwell, B. R. Multicomponent therapeutics for networked systems. Nat Rev Drug Discov, 2005;4(1):71-8. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=15688074.

[12] Kitano, H. A robustness-based approach to systems-oriented drug design. Nat Rev Drug Discov, 2007;6(3):202-10. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=17318209.

[13] Laubenbacher, R., Hower, V., Jarrah, A., Torti, S. V., Shulaev, V., Mendes, P., Torti, F. M., Akman, S. A systems biology view of cancer. Biochim Biophys Acta, 2009;1796(2):129-39. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19505535.

[14] Lehar, J., Krueger, A. S., Avery, W., Heilbut, A. M., Johansen, L. M., Price, E. R., Rickles, R. J., Short, 3rd, G. F., Staunton, J. E., Jin, X., Lee, M. S., Zimmermann, G. R., Borisy, A. A. Synergistic drug combinations tend to improve therapeutically relevant selectivity. Nat Biotechnol, 2009;27(7):659-66. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19581876.

[15] Lehar, J., Krueger, A. S., Zimmermann, G. R., Borisy, A. A. Therapeutic selectivity and the multi-node drug target. Discov Med, 2009;8(43):185-90. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=20040268.

[16] Lipton, S. A. Turning down, but not off. Nature, 2004;428(6982):473. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=15057811.

[17] MD Anderson Cancer Center. BATTLE Links Potential Biomarkers to Drugs for Lung Cancer; 2010. Press Release. URL: http://www.mdanderson.org/newsroom/news-releases/2010/battle-links-potential-biomarkers-to-lung-cancer-drugs.html.

[18] Moserle, L., Ghisi, M., Amadori, A., Indraccolo, S. Side population and cancer stem cells: therapeutic implications. Cancer Lett, 2010;288(1):1-9. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19523754.

[19] Newman, D. J., Cragg, G. M. Natural products as sources of new drugs over the last 25 years. J Nat Prod, 2007;70(3):461-77. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=17309302.

[20] Ohlson, S. Designing transient binding drugs: a new concept for drug discovery. Drug Discov Today, 2008;13(9-10):433-9. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=18468561.

[21] Petrelli, A., Valabrega, G. Multitarget drugs: the present and the future of cancer therapy. Expert Opin Pharmacother, 2009;10(4):589-600. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19284362.

[22] Pujol, A., Mosca, R., Farres, J., Aloy, P. Unveiling the role of network and systems biology in drug discovery. Trends Pharmacol Sci, 2010;31(3):115-23. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=20117850.

[23] Silva, A. S., Gatenby, R. A. A theoretical quantitative model for evolution of cancer chemotherapy resistance. Biol Direct, 2010;5(1):25. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=20406443.

[24] Smalley, K. S., Haass, N. K., Brafford, P. A. and Lioni, M., Flaherty, K. T., Herlyn, M. Multiple signaling pathways must be targeted to overcome drug resistance in cell lines derived from melanoma metastases. Mol Cancer Ther, 2006;5(5):1136-44. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=16731745.

[25] Stermitz, F. R., Lorenz, P., Tawara, J. N., Zenewicz, L. A., Lewis, K. Synergy in a medicinal plant: antimicrobial action of berberine potentiated by 5′-methoxyhydnocarpin, a multidrug pump inhibitor. Proc Natl Acad Sci U S A, 2000;97(4):1433-7. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=10677479.

[26] Wang, J., Sui, M., Fan, W. Nanoparticles for tumor targeted therapies and their pharmacokinetics. Curr Drug Metab, 2010;11(2):129-41. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=20359289.

[27] World Health Organization. WHO issues guidelines for herbal medicines. Bull World Health Organ, 2004;82(3):238. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=15112024.

[28] Xia, E. Q., Deng, G. F., Guo, Y. J., Li, H. B. Biological activities of polyphenols from grapes. Int J Mol Sci, 2010;11(2):622-46. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=20386657.

[29] Zhu, Y., Hu, J., Hu, Y., Liu, W. Targeting DNA repair pathways: a novel approach to reduce cancer therapeutic resistance. Cancer Treat Rev, 2009;35(7):590-6. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19635647.

Roadmap Chapter 2

JohnBoik — Thu, 29 Jul 2010 01:08:14 +0000

Chapter 1	Introduction
Chapter 2	History of Cancer Drug Discovery
Chapter 3	Overview of Network Pharmacology
Chapter 4	Mixture Discovery Program

2 History of cancer drug discovery

This chapter provides a brief review of protein biology and a short history of cancer drug discovery. The background information in this chapter and the next provides a basis for describing the strategy adopted in the NEBM mixture discovery program.

2.1 Living in a time of -omics

The modern field of cancer drug discovery has been in a constant state of evolution since its inception in the 1950s. Starting from humble beginnings in the laboratories of a handful of investigators, it is now a multi-billion dollar business employing hundreds of thousands of skilled workers around the globe. Broad public support, and dramatic advances in biology, medicine, mathematics, and analytical chemistry, have propelled the field forward at an ever-increasing rate.

We now live at a very exciting time—the beginning of the “-omics” period. For the first time, we can make meaningful measurements of thousands of cellular components across multiple cell types and species. Thus we study genomics (genes), proteomics (proteins), lipidomics (lipids), and metabolomics (metabolites). In fact, we are moving past the initial step of creating inventories to the far more difficult step of trying to understand how all the components dynamically interact. The result is that we are just beginning to see something that resembles a “big picture” in cell biology, although that picture is still quite fuzzy.

Studying dynamic interactions between cellular components is the realm of systems biology. The term is loosely used in science to refer to a inter-disciplinary field of study that relies on biology, mathematics, statistics, chemistry, and bioinformatics to describe the dynamic interplay life’s elements. It is contrasted by a reductionist approach, which focuses on each element of a system in isolation. The term systems biology is often used interchangeably with network biology. It is usually applied to the cellular level, but it could also be applied to higher levels of organization, such as tissues, organisms, and society. For our purposes, it refers to study of the dynamic protein-protein, protein-lipid, protein-polysaccharide, protein-nucleic acid, and protein-small molecule interactions that occur both within a cell and between a cell and its local environment.

As a quick review of biology, proteins are the most abundant of the four types of cellular macromolecules, the other three are lipids, nucleic acids (DNA and RNA), and polysaccharides. Lipids serve as energy storage and as components of the barriers (membranes) that separate cellular compartments. For example, the plasma membrane surrounding a cell is made up mostly of lipids. Nucleic acids consist of nucleotides strung together. They store genetic information. Polysaccharides consist of simple sugars strung together. They act as structural components, as well as energy storage. Proteins consist of amino acids strung together. Nearly all activity in a cell is mediated by proteins. Another important group of compounds in a cell is small (signaling) molecules. These are much smaller than the macromolecules, and they bind to proteins, polysaccharides, and nucleic acids, affecting their action.

Figure 1↓ provides a surface representation of a typical protein, cyclin-dependent kinase 2, which is active in the cell cycle (the process of cell division). The structure (2UZO) was obtained from the Protein Data Bank [15]. Notice the small molecule attached to the protein in one of its pockets. Small molecules (such as drugs) affect protein conformation by binding to specific sites on a protein, and this alters the protein’s structure and activity. Due to the central role that proteins play in cellular life, their abundance in a cell, and the fact that their actions can be altered with external compounds, proteins are of prime interest in biology and drug discovery.

Figure 1 CDK2 protein with bound inhibitor

Proteins can interact with small molecules and other proteins in a process called signal transduction. In this process, signals (messages or instructions) are propagated from a receptor on the cell surface or within the cell to some final destination where cell response is affected. Typically, the end result of signal transduction is the binding of transcription factors to DNA, which leads to the production of proteins, and in some cases proliferation. Signal propagation occurs via protein-protein and protein-small molecule interactions. It can also occur through redox (reduction-oxidation) signaling, where free radicals act as messenger compounds.

A common example of signal transduction occurs when you cut your finger. Damaged cells and local immune cells release a wave of signal molecules into the interstitial space (between cells) that instructs various cell types to begin the repair process. For example, local epithelial cells may be instructed to proliferate in order to create new skin and capillaries. In this case, a growth factor outside an epithelial cell binds to a receptor (a protein) attached to the cell’s surface. In response, the receptor changes its conformation (shape), which sets off a chain of events whereby the growth signal is propagated from the cell’s surface to the cell’s nucleus. Once the signal reaches the nucleus, it causes transcription factors to bind to DNA. This sets in place the machinery to initiate gene transcription, production of more proteins, and proliferation. This process is illustrated in Figure 2↓.

Figure 2 Signal transduction

2.2 Historical periods

To understand the directions that cancer drug discovery could take in the next few decades, it is helpful to look back at where it started. Modern cancer therapy is only about 65 years old. Even so, our understanding of biology was much different then. The first papers on the use of a cytotoxic drug (mustard gas) to treat cancer were published in 1946 [4], which was seven years before the structure of DNA was first published. Scientists were working nearly blinded, by today’s standards, in their efforts to develop useful cancer therapies. Clearly, cancer drug discovery has gone through many changes. These changes fall into two main historical periods, and a third period appears to have just begun. These are the cytotoxic period, the targeted therapy period, and the systems biology period.

2.2.1 The cytotoxic period (mid 1940s to mid 1980s)

During the cytotoxic period our biological understanding and technologies were relatively crude and progress was slow. Drug discovery focused on a simple task: identify compounds (such as DNA poisons) that kill the cells that divide most frequently. Because cancer cells tend to proliferate more frequently than most types of normal cells, the hope was that these drugs would act somewhat specifically to cancer cells. It turns out, however, that several types of normal cells proliferate at rates similar to cancer cells, and these too are affected by cytotoxic drugs. These cells include hair follicles, bone marrow cells, and cells of the intestinal lining. Their destruction results in the common chemotherapy side effects of hair loss, immune suppression, and nausea. (See article Types of Chemotherapy Drugs.)

In addition to dose-limiting adverse effects, a second problem that limits the effectiveness of cytotoxic drugs is induced resistance. A small tumor of one cubic centimeter contains about a billion cancer cells. Within this population is a large degree of genetic variation. One effect is that a small percentage of cells will be naturally resistant to any given drug. While administration of a drug might cause the majority of cells to die, a small fraction often survives and over time repopulates the tumor. Commonly, the new population is more aggressive then the original one, and is resistant to a wide variety of drugs. This is called multidrug resistance (MDR). If not for MDR, we would likely already have curative cancer treatments.

To help reduce the multidrug resistance problem, researchers began to administer cytotoxic drugs in mixtures, sometimes called cocktails. The mixtures worked by attacking different parts of the cell, and most contained drugs that were of highly dissimilar structures. The result was that MDR could often be dampened, but not necessarily eliminated. A change in approach was needed if safe and effective drug treatments were to become a reality.

2.2.2 The targeted therapy period (mid 1980s to mid 2000s)

Drug discovery programs began to adopt a new approach, called targeted therapy, in the mid 1980s. By then, technology and biology were starting to make rapid advances, and new tools and ideas had become available. The task in targeted therapy was to identify a key protein that is overactive, overproduced, or somehow altered in cancer cells, and then design a drug that will bind to that protein and inhibit its action. This is a one-drug, one-protein model. The hope was that targeted therapies would act more specific to cancer cells, and thus be safer and more effective.

One wonderful success from this period is the drug Imatinib, which is useful against chronic myelogenous leukemia (CML) and gastrointestinal stromal tumors (GISTs). In spite of a few successes, however, serious drawbacks to targeted therapy have become apparent over time. First, as with cytotoxic drugs, tumors can develop resistance to targeted therapy [17]. Second, while the risk of side effects may be reduced compared to cytotoxic drugs, serious adverse effects can still occur.

Third, it appears that targeted therapy will be somewhat limited to rare forms of cancer. These are cancers that happen to exhibit distinct differences in one or a few key proteins, compared to normal cells. In most common cancers, a large number of proteins are altered in function or production, and no single highly effective target stands out. For example, the cells of some tumors can contain more than one thousand somatic mutations [10].

Lastly, it turns out that the one-drug one-protein model is not plausible for most drugs. Nearly all drugs bind to multiple proteins within a cell [8]. In fact, a recent study suggest that on average, a drug will interact with six protein targets [12]. Some bindings may produce wanted effects, and some may produce unwanted effects. Unexpected off-target binding and attendant adverse effects is one of the key reasons why many targeted therapy drugs fail in the last and most expensive phase of drug development (clinical trials) [9]

The pharmaceutical industry has embraced the targeted-therapy approach not just for cancer but for most diseases. Along with increased regulation and other factors, the unfortunate result has been a highly expensive period of drug research that has resulted in a dwindling number of new drug classes approved for market [11,13,7]. The industry as a whole is currently under pressure, as patents have expired or will soon expire on many of the blockbuster drugs developed in years past. There are few new drugs to replace them. In short, it appears that while targeted therapy will bring some useful drugs to market, the approach is too simplistic to meet the demand for a new generation of safe and highly effective drugs for the majority of chronic diseases, including common cancers [6]. There is room for improvement.

2.2.3 The network pharmacology period (mid 2000s+)

The current drug arsenal for treating cancer consists primarily of cytotoxic chemotherapy cocktails and targeted therapy drugs. Hormonal therapies, immune therapies, and some other classes make up a smaller fraction. While the pharmaceutical industry remains focused on development of new targeted-therapy drugs, a growing number of scientists have turned their sights to the next evolution: systems biology.

In systems biology, the dynamics of protein networks are targeted, rather than the activity of a single protein. This is like stepping away from the trees to see the forest. It is the system and its dynamics that as a whole allows cancer cells to exhibit the complex characteristics and activities that set them apart from normal cells. Of course, understanding how the system works so that it can be optimally manipulated is a very difficult task, one that may take many decades and further advances in technology to perfect. Nevertheless, the information gathered even at this early stage is of sufficient quality to allow meaningful models to be developed [16]. In some cases, such models are already leading to faster drug discovery and initiation of clinical trials [3].

Systems biology, when applied to drug discovery, is often referred to as network pharmacology, or polypharmacology. The later term implies that multiple proteins are targeted. Three strategies of polypharmacology have appeared in the literature so far:

discovery of single drugs that affect multiple proteins,
discovery of small mixtures of drugs that affect multiple proteins, and
discovery of large mixtures of drugs that affect multiple proteins.

Of the clinical studies reported in the literature, all are examples of the first two strategies [14]. Compared with the third strategy, the first two focus on a small number of protein targets. The third strategy has only been mentioned by a few authors as a promising direction [11,5,2,18,1]. It is widely utilized in various forms of herbal medicine, however. NEBM is focused on a particular implementation of the large-mixture strategy, which we term distributed polypharmacology (DPP).

The next two chapters further discusses network pharmacology and provide other background information. Chapter 4 describes the NEBM mixture development program, which is based on DPP.

Bibliography

[1] Araujo, R. P., Liotta, L. A., Petricoin, E. F. Proteins, drug targets and the mechanisms they control: the simple truth about complex networks. Nat Rev Drug Discov, 2007;6(11):871-80. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=17932492.

[2] Csermely, P., Agoston, V., Pongor, S. The efficiency of multi-target drugs: the network approach might help drug design. Trends Pharmacol Sci, 2005;26(4):178-82. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=15808341.

[3] Erler, J. T., Linding, R. Network-based drugs and biomarkers. J Pathol, 2010;220(2):290-6. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19921715.

[4] Goodman, L.S., Wintrobe, M.M., Dameshek, W., Goodman, M., Gilman A., McLennan, MT. Nitrogen mustard therapy. JAMA, 1946;132:126-132.

[5] Hase, T., Tanaka, H., Suzuki, Y., Nakagawa, S., Kitano, H. Structure of protein interaction networks and their implications on drug design. PLoS Comput Biol, 2009;5(10):e1000550. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19876376.

[6] Hellerstein, M. K. A critique of the molecular target-based drug discovery paradigm based on principles of metabolic control: advantages of pathway-based discovery. Metab Eng, 2008;10(1):1-9. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=17962055.

[7] Honig, P., Lalonde, R. The economics of drug development: a grim reality and a role for clinical pharmacology. Clin Pharmacol Ther, 2010;87(3):247-51. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=20160740.

[8] Hopkins, A. L., Mason, J. S., Overington, J. P. Can we rationally design promiscuous drugs?. Curr Opin Struct Biol, 2006;16(1):127-36. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=16442279.

[9] Janga, S. C., Tzakos, A. Structure and organization of drug-target networks: insights from genomic approaches for drug discovery. Mol Biosyst, 2009;5(12):1536-48. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19763339.

[10] Jones, S., Zhang, X., Parsons, D. W., Lin, J. C., Leary, R. J., Angenendt, P., Mankoo, P. and Carter, H., Kamiyama, H., Jimeno, A., Hong, S. M., Fu, B., Lin, M. T., Calhoun, E. S., Kamiyama, M., Walter, K., Nikolskaya, T. and Nikolsky, Y., Hartigan, J., Smith, D. R., Hidalgo, M., Leach, S. D., Klein, A. P., Jaffee, E. M., Goggins, M., Maitra, A., Iacobuzio-Donahue, C., Eshleman, J. R., Kern, S. E., Hruban, R. H., Karchin, R., Papadopoulos, N., Parmigiani, G., Vogelstein, B., Velculescu, V. E., Kinzler, K. W. Core signaling pathways in human pancreatic cancers revealed by global genomic analyses. Science, 2008;321(5897):1801-6. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=18772397.

[11] Kitano, H. A robustness-based approach to systems-oriented drug design. Nat Rev Drug Discov, 2007;6(3):202-10. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=17318209.

[12] Mestres, J., Gregori-Puigjane, E., Valverde, S. and Sole, R. V. The topology of drug-target interaction networks: implicit dependence on drug properties and target families. Mol Biosyst, 2009;5(9):1051-7. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19668871.

[13] Ohlson, S. Designing transient binding drugs: a new concept for drug discovery. Drug Discov Today, 2008;13(9-10):433-9. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=18468561.

[14] Petrelli, A., Valabrega, G. Multitarget drugs: the present and the future of cancer therapy. Expert Opin Pharmacother, 2009;10(4):589-600. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19284362.

[15] Protein Data Bank. http://dx.doi.org/10.2210/pdb2uzo/pdb; .

[16] Pujol, A., Mosca, R., Farres, J., Aloy, P. Unveiling the role of network and systems biology in drug discovery. Trends Pharmacol Sci, 2010;31(3):115-23. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=20117850.

[17] Redner, R. L. Why doesn’t imatinib cure chronic myeloid leukemia?. Oncologist, 2010;15(2):182-6. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=20124443.

[18] Zanzoni, A., Soler-Lopez, M., Aloy, P. A network medicine approach to human disease. FEBS Lett, 2009;583(11):1759-65. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19269289.

Roadmap Chapter 3

JohnBoik — Thu, 29 Jul 2010 01:07:40 +0000

Chapter 1	Introduction
Chapter 2	History of Cancer Drug Discovery
Chapter 3	Overview of Network Pharmacology
Chapter 4	Mixture Discovery Program

3 Overview of network pharmacology

This chapter further discusses background material that will help describe the NEBM mixture discovery program. It provides a brief overview of the structure of biological networks, discusses how drugs can interact to affect the network, and discusses some molecular targets other than classic protein targets.

3.1 Characteristics of biological networks

Although pharmaceutical corporations have been developing targeted therapy drugs for two decades, conservative estimates suggest that only about 50 percent of the proteins that could be targeted with drugs have been intentionally targeted. A cell contains roughly 24,000 genes that code proteins. Less than 10 percent of these (roughly 1,700) have been directly tied to diseases, however. These are termed disease genes. A larger set of genes (roughly 3,000) are believed to be druggable based on their physical structure and other features. The intersection of these two sets, illustrated in Figure 1↓, contains roughly 520 genes that are directly involved in human diseases and which may be susceptible to manipulation by drugs [23]. Of these about 270 have been intentionally targeted with drugs [19]. Thus, many opportunities still exist to target new proteins in the treatment of human diseases.

Figure 1 The intersection of disease genes and druggable genes

Many more useful targets may in fact exist. First, genes that play a secondary or indirect role in disease are not necessarily included in the set of known disease genes. Second, through alternative splicing and other processes, each gene can produce more than one protein isoform. This results in a human proteome that is at least an order of magnitude larger than the human genome [25]. Any given isoform could be a target for therapy.

Proteins within a cell form a network of interactions, and for many interactions the network can be broken down into smaller signaling pathways. Proteins can directly interact in a pathway to form stable or transient complexes, such as may occur during signal transduction. They can also interact via indirect ways. For example, proteins on the same signaling pathway may interact via small molecules or via protein intermediates. Similar mechanisms allow proteins to interact on separate signaling pathways, in a process called cross-talk. The number of distinct signaling pathways in a cell is unknown, as new ones are still being discovered. In cancer, about a dozen have been heavily studied.

Figure 2↓ illustrates an example network of 231 proteins that are likely involved in the G2/M transition of the cell (proliferation) cycle [20]. (See the article Types of Chemotherapy Drugs for an explanation of the cell cycle.) The protein CDK2, illustrated in Chapter 2, Figure 1, plays an important role in the cell cycle and is marked by the arrow. As is apparent from the figure, some proteins are very highly connected to other proteins. These proteins are said to have a very high degree, and are called hubs. An example is marked in the figure. Other proteins have few connections, and are of low degree. Hubs tend to be of special interest in network analysis, because if a hub fails a large number of other proteins can be affected.

Figure 2 Network of proteins involved in G2/M transition of the cell cycle

3.2 Targeting the protein network

Considering network topologies like the one shown in Figure 2↑, one may ask, “How many nodes should be targeted to affect a disease, and how strongly does each one need to be inhibited?”. From a targeted therapy perspective, inhibition of a single vital node is sufficient if its characteristics are different enough from that of normal cells. The goal is to strongly inhibit that node. But systems biology offers a different perspective. Recent papers suggest that multiple weak hits on a network can have a greater overall effect than a single strong one [1,8,9]. Moreover, synergistic actions between multiple drugs can be used to focus the overall effect towards the therapeutic goal, and away from adverse effects [17]. This would suggest that many nodes could be targeted, and that none need be strongly inhibited. Such an approach could have a large impact on the safety and effectiveness of new drug therapies.

Another reasonable question is, “Which nodes should be targeted?” To help answer that question, it is useful to look closer at disease genes. It appears that most disease genes are of intermediate degree in the protein network [10]. Indeed, most successful drugs target nodes that are of higher than average degree, but are not hubs [27]. Cancer drugs, however, tend to target nodes of higher degree than do drugs for other diseases [14]. This is not surprising, as cancer drugs are intended to produce radical effects (cell death), and cancer disease genes tend to be more highly connected [13]. Thus from a conventional perspective, nodes of higher than average degree should be targeted to affect cancer. But this viewpoint derives from the fact that typically only one node is being targeted.

Targets of distributed polypharmacology (DPP) are likely to span a wider degree range, but on average be of lower degree. Although DPP targets could include highly-, moderately-, and lightly-connected nodes, integration between functional modules in a cell tends to occur via middle-degree nodes [14]. Low-degree nodes tend to facilitate specific effects in specific tissues, and in some cases this reduces their value as targets. But they are also the safest ones to affect. In contrast, targeting high-degree nodes has potential to produce the most radical effects in cancer cells, but also the most adverse effects in normal cells. By concentrating on somewhat lower degree nodes relative to conventional therapies, DPP mixtures could be less prone to adverse effects. Furthermore, because DPP employs many components, its effects are potentially more tunable than therapies with few components. Small hits on many targets are used to shape the total effect, including establishment of therapeutic robustness. The intent is to create an orchestra effect, rather than a solo or duet.

As an aside, many cancer disease genes play critical roles in development and growth, and are expressed in a wide variety of tissues. For example, genes controlling proliferation are found in all cell types and under normal circumstances are tightly regulated to prevent aberrant cell division. In cancer, these regulatory controls are hijacked, allowing cells to proliferate at higher rates. The hijacking of common genes and pathways by cancer makes selective targeting of cancer cells very difficult. In general, the difference between cancer and normal cells is that cancer cells exhibit abnormal network wiring and information flow, as opposed to unique proteins. That is why targeting the network, rather than a single protein, is such an appealing concept.

The above discussion provides a rough answer to which proteins should be targeted. To identify the particular proteins that should be targeted for a particular cancer requires far more work. The traditional targets for cancer chemotherapy have been DNA itself or proteins directly involved in the cell cycle. We are quickly learning that many signal transduction pathways, apart from the cell cycle, are involved in tumor progression. And we are learning that the cell cycle itself is more complex than first thought. Thus an expanding range of cellular and extracellular macromolecules present useful targets.

Any protein involved in proliferation, cell repair, drug resistance, invasion, metastasis, or angiogenesis is a target. (Angiogenesis is the growth of new blood vessels towards a tumor.) Within any one of these process, multiple proteins and signaling pathways will be active. For example, subpopulations within a tumor may be drug resistant because they produce proteins that pump the drug out of the cell, they use alternative proteins in signal transduction, they produce enzymes (proteins) that degrade the drug, they are quiescent (not proliferating) during drug exposure but proliferate later when the drug is removed, they exhibit altered cell-repair mechanisms, they exhibit altered apoptotic (programmed cell death) mechanisms, or they exhibit altered epigenetics (discussed below). All proteins and networks involved in these processes are targets for therapy. The goal is to determine which combinations of targets result in optimally effective and safe therapies.

Whole new categories of protein targets are being discovered. One example is proteins involved in the epigenetic process. We know from classical biology that traits can be passed on to new generations via the coded information contained in DNA. But traits can also be inherited in another way, called epigenetics. A cell has roughly two meters of DNA within it, which under normal circumstances is compactly stored, wrapped around spherical proteins called histones. The histone and wrapped DNA is called chromatin. If a cell wants to produce a protein, it has to unwind the DNA from the correct histone. Once the gene sequence is unfolded, transcription enzymes can bind to the DNA and initiate the copying process.

In epigenetics, the DNA sequence is not changed but other modifications are made. For example, bulky methyl groups can be attached to DNA molecules, or methyl or acetyl groups can be attached to histones. The added molecules get in the way of enzymes that perform chromatin remodeling or DNA transcription. Epigenetic processes act as a gateway, determining if a gene is transcribed or silenced. Enzymes (proteins) that attach methyl and acetyl groups are now considered targets for cancer therapy.

In cancer, many abnormal epigenetic changes can occur, resulting in activation of genes that should be silenced, or silencing of genes that should be active. Epigenetic targets provides an attractive opportunity to re-establish proper transcription in a genome that is heavily modified by mutations [4]. In this way, cancer cells could be directed into a more normal phenotype despite their underlying genetic abnormalities. Various flavonoids and other natural compounds have been shown to inhibit the proteins that play a role in epigenetics [15,12].

3.3 How small molecules interact to affect networks

Small molecules, such as many natural products and drugs, can affect network dynamics in a variety of ways. The classic and most studied is by binding to one or more proteins. Binding to a protein changes its conformation, and the change in conformation alters its action. When multiple compounds bind to a protein or proteins, interactions between compounds can occur. These can be antagonistic (subadditive), additive, or synergistic (superadditive). If two or more compounds are inactive alone but active when in combination, the interaction is called coalistic. Various mathematical models have been developed to assess the magnitude of such interactions, including the Mixlow method proposed by Dr. Boik et al. [16,3,6,11,26,7].

Interactions between compounds can occur in a number of ways, including:

Binding to the same active site on a protein. Similarly-shaped compounds can compete for the same binding site, which generally leads to additive interactions. Additive interactions can be useful to improve the safety of a mixture by allowing individual components to be used at a reduced dose. Binding-site competition can also be useful to increase robustness of the mixture against genetic variants of the target protein.
Binding to different active sites on the same protein. Some proteins contain more than one active site, and simultaneous binding has the potential to produce any type of drug interaction.
Binding to active sites on interacting proteins. Binding to different proteins that physically interact, interact in the same pathway, or interact in coupled pathways (via cross-talk) has potential to produce any type of drug interaction.
Binding to protein-protein interfaces. While active sites on proteins are the traditional target in drug discovery, recent evidence suggests that the (mostly flat) interface between physically interacting proteins is also a potential target. Some natural products are known to affect protein-protein binding [2], and this could lead to any type of drug interaction.

Compounds can also interact via distant proteins to produce pharmacokinetically potentiative or reductive effects. Here, one compound affects the absorption, distribution, metabolism, or excretion (ADME) of another compound. This can affect plasma concentrations of the compounds, and thereby affect therapeutic efficacy.

Yet other ways exist for compounds to interact. For example, proteins that span the plasma membrane tend to float around the membrane in small groups, called lipid rafts. The composition of the membrane has a large effect on movement and other characteristics of the rafts, which in turn has a large effect on protein activity [22]. Evidence is mounting that rafts represents viable targets in cancer and other diseases. Various natural compounds, including flavonoids and fatty acids, can affect lipid rafts [24,21,5]. The effects of mixtures on lipid rafts have yet to be studied.

In summary, compounds can interact in a large number of ways to affect network function. Some protein targets within networks are well understood, and others are just being discovered. Thus any given combination of drugs or natural products has potential to act via mechanisms that are novel and unexpected. For this reason, it is important to base a mixture screening program on observation of living cells interacting in a realistic environment. Screening programs based primarily on cell-free biochemical assays, which study isolated proteins, are likely to miss important drug effects, both beneficial and detrimental. Biochemical assays can, however, add useful information to a cell-based screening program. Cell-based screening in realistic three-dimensional environments provides opportunities not only to discover useful drugs and mixtures, but also to probe the workings of both intra- and inter-cellular networks [18]. Much could be learned about the wiring and dynamics of networks via systematic perturbations of the cellular environment.

The next and last chapter in this Roadmap, Chapter 4, describes the NEBM mixture development program.

Bibliography

[1] Agoston, V., Csermely, P., Pongor, S. Multiple weak hits confuse complex systems: a transcriptional regulatory network as an example. Phys Rev E Stat Nonlin Soft Matter Phys, 2005;71(5 Pt 1):051909. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=16089573.

[2] Bauer, R. A., Wurst, J. M., Tan, D. S. Expanding the range of ‘druggable’ targets with natural product-based libraries: an academic perspective. Curr Opin Chem Biol, 2010 URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=20202892.

[3] Berenbaum, M. C. What is synergy?. Pharmacol Rev, 1989;41(2):93-141. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=2692037.

[4] Best, J. D., Carey, N. Epigenetic opportunities and challenges in cancer. Drug Discov Today, 2010;15(1-2):65-70. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19897050.

[5] Biondo, P. D., Brindley, D. N., Sawyer, M. B. and Field, C. J. The potential for treatment with dietary long-chain polyunsaturated n-3 fatty acids during chemotherapy. J Nutr Biochem, 2008;19(12):787-96. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=18602809.

[6] Boik, J. C., Newman, R. A., Boik, R. J. Quantifying synergism/antagonism using nonlinear mixed-effects modeling: a simulation study. Stat Med, 2008;27(7):1040-61. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=17768754.

[7] Chou, T. C., Talalay, P. Quantitative analysis of dose-effect relationships: the combined effects of multiple drugs or enzyme inhibitors. Adv Enzyme Regul, 1984;22:27-55. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=6382953.

[8] Csermely, P., Agoston, V., Pongor, S. The efficiency of multi-target drugs: the network approach might help drug design. Trends Pharmacol Sci, 2005;26(4):178-82. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=15808341.

[9] Csermely, P. Strong links are important, but weak links stabilize them. Trends Biochem Sci, 2004;29(7):331-4. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=15236738.

[10] Feldman, I., Rzhetsky, A., Vitkup, D. Network properties of genes harboring inherited disease mutations. Proc Natl Acad Sci U S A, 2008;105(11):4323-8. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=18326631.

[11] Fidler, M., Kern, S. E. Flexible interaction model for complex interactions of multiple anesthetics. Anesthesiology, 2006;105(2):286-96. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=16871062.

[12] Gilbert, E. R., Liu, D. Flavonoids Influence Epigenetic-Modifying Enzyme Activity: Structure – Function Relationships and the Therapeutic Potential for Cancer. Curr Med Chem, 2010 URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=20345345.

[13] Goh, K. I., Cusick, M. E., Valle, D., Childs, B., Vidal, M., Barabasi, A. L. The human disease network. Proc Natl Acad Sci U S A, 2007;104(21):8685-90. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=17502601.

[14] Hase, T., Tanaka, H., Suzuki, Y., Nakagawa, S., Kitano, H. Structure of protein interaction networks and their implications on drug design. PLoS Comput Biol, 2009;5(10):e1000550. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19876376.

[15] Hauser, A. T., Jung, M. Targeting epigenetic mechanisms: potential of natural products in cancer chemoprevention. Planta Med, 2008;74(13):1593-601. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=18704881.

[16] John C. Boik and Balasubramanian Narasimhan. An R package for assessing drug synergism/antagonism. Journal of Statistical Software, 2010;34(6):1—18. URL: http://www.jstatsoft.org/v34/i06.

[17] Lehar, J., Krueger, A. S., Avery, W., Heilbut, A. M., Johansen, L. M., Price, E. R., Rickles, R. J., Short, 3rd, G. F., Staunton, J. E., Jin, X., Lee, M. S., Zimmermann, G. R., Borisy, A. A. Synergistic drug combinations tend to improve therapeutically relevant selectivity. Nat Biotechnol, 2009;27(7):659-66. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19581876.

[18] Lehar, J., Zimmermann, G. R., Krueger, A. S., Molnar, R. A., Ledell, J. T., Heilbut, A. M., Short, 3rd, G. F., Giusti, L. C., Nolan, G. P. and Magid, O. A., Lee, M. S., Borisy, A. A., Stockwell, B. R., Keith, C. T. Chemical combination effects predict connectivity in biological systems. Mol Syst Biol, 2007;3:80. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=17332758.

[19] Overington, J. P., Al-Lazikani, B., Hopkins, A. L. How many drug targets are there?. Nat Rev Drug Discov, 2006;5(12):993-6. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=17139284.

[20] Pathway Commons/Reactome. http://www.pathwaycommons.org; .

[21] Patra, S. K., Rizzi, F., Silva, A., Rugina, D. O., Bettuzzi, S. Molecular targets of (-)-epigallocatechin-3-gallate (EGCG): specificity and interaction with membrane lipid rafts. J Physiol Pharmacol, 2008;59 Suppl 9:217-35. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=19261982.

[22] Patra, S. K. Dissecting lipid raft facilitated cell signaling pathways in cancer. Biochim Biophys Acta, 2008;1785(2):182-206. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=18166162.

[23] Sakharkar, M. K., Sakharkar, K. R., Pervaiz, S. Druggability of human disease genes. Int J Biochem Cell Biol, 2007;39(6):1156-64. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=17446117.

[24] Tarahovsky, Y. S., Muzafarov, E. N., Kim, Y. A. Rafts making and rafts braking: how plant flavonoids may control membrane heterogeneity. Mol Cell Biochem, 2008;314(1-2):65-71. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=18414995.

[25] Taussig, M. J., Stoevesandt, O., Borrebaeck, C. A., Bradbury, A. R., Cahill, D., Cambillau, C. and de Daruvar, A., Dubel, S., Eichler, J., Frank, R., Gibson, T. J., Gloriam, D., Gold, L., Herberg, F. W., Hermjakob, H., Hoheisel, J. D., Joos, T. O., Kallioniemi, O., Koegl, M. and Konthur, Z., Korn, B., Kremmer, E., Krobitsch, S., Landegren, U., van der Maarel, S. and McCafferty, J., Muyldermans, S., Nygren, P. A., Palcy, S., Pluckthun, A., Polic, B., Przybylski, M., Saviranta, P., Sawyer, A., Sherman, D. J., Skerra, A., Templin, M., Ueffing, M., Uhlen, M. ProteomeBinders: planning a European resource of affinity reagents for analysis of the human proteome. Nat Methods, 2007;4(1):13-7. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=17195019.

[26] White, D. B., Slocum, H. K., Brun, Y., Wrzosek, C., Greco, W. R. A new nonlinear mixture response surface paradigm for the study of synergism: a three drug example. Curr Drug Metab, 2003;4(5):399-409. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=14529372.

[27] Yao, L., Rzhetsky, A. Quantitative systems-level determinants of human genes targeted by successful drugs. Genome Res, 2008;18(2):206-13. URL: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?cmd=prlinks&dbfrom=pubmed&retmode=ref&id=18083776.