the biological framework: translational research from bench to

The Biological Framework: Translational Researchfrom Bench to Clinic

YOSEF YARDEN

Department of Biological Regulation, The Weizmann Institute of Science, Rehovot, Israel

Key Words. Signal pathways • Vascular endothelial growth factor • VEGF • Epidermal growth factor • EGF •Immunotherapy • Chemotherapy • Heat shock protein inhibitors

Disclosures: Yosef Yarden: Research funding/contracted research: National Institutes of Health, National Cancer Institute;Intellectual property rights/inventor or patent holder: Soluble ErbB proteins, anti-EGF-like antibodies.

ABSTRACT

The understanding of cellular signaling pathways inmalignant tumors is an important aspect of cancer re-search and modern targeted therapy strategies. Growthfactors and their receptors in particular are critical tomodern cancer therapy research, because these factorscontrol all phases of tumor development and metastasis.Most importantly, growth factors are responsible forcell survival under cytotoxic drugs and radiotherapy.These growth factor signaling pathways are composedof complex networks that have adapted to efficiently re-spond to certain disturbances, such as a single agentthat targets one aspect of the pathway. Meanwhile, mul-

tiple insults to the pathway, such as combination ther-apy regimens, are known to be effective in shutting downthese pathways and, consequently, killing the tumor cell.Research is currently under way to find new ways to ex-ploit fragile aspects of oncogenic networks, such as uncom-mon, multiple perturbations that target essential hubsthrough immunotherapy, combinations of antibodies,heat shock protein inhibitors, or novel drug combinations.Complex growth factor signaling networks and novelmethods to shut down these networks are described withina framework of engineering and mathematical concepts.The Oncologist 2011;16(suppl 1):23–29

INTRODUCTION

The medical community has seen a revolution of moleculartargeted therapy in breast and other types of cancer, a rev-olution that started about a hundred years ago with the ideaof Paul Ehrlich, a chemist who worked in Frankfurt on dyes,which he noted specifically stained some tissues and ig-nored neighboring tissues.

Ehrlich posed the brilliant idea, the “magic bullet” con-cept. Accordingly, it would be possible one day to directdrugs to some tissues and spare other tissues. Indeed, Figure

1 illustrates the development, stepwise, of drugs like tras-tuzumab, rituximab, and imatinib over the last 10–12 years,all of which have been instrumental in improving canceroutcomes while reducing systemic toxicity (Fig. 1) [1].

The “oncogene decades,” two decades between the1970s and the 1990s, allowed for this revolution by en-abling researchers to resolve pathways that relay signalsfrom outside the cell, all the way through the membrane andthe cytoplasm, to the nucleus (Fig. 2).

In retrospect, it turned out that several oncogenes act

Correspondence: Yosef Yarden, Ph.D., Department of Molecular Cell Biology, The Weizmann Institute, Rehovot 76100, Israel. Tele-phone: 972-8-9344502; Fax: 972-8-9342488; e-mail: [email protected]

Reprinted from The Oncologist 2010;15(suppl 5):1–7. ©AlphaMed Press 1083-7159/2010/$30.00/0 doi: 10.1634/theoncolo-gist.2011-S1-23

TheOncologist®

The Oncologist 2011;16(suppl 1):23–29 www.TheOncologist.com

by guest on March 22, 2018

http://theoncologist.alphamedpress.org/

Dow

nloaded from


along the same type of pathways that lead from the plasmamembrane, from growth factors like vascular endothelial

growth factor (VEGF), epidermal growth factor (EGF), andfibroblast growth factor, down to effectors like KRAS and

Figure 1. The revolution of molecular targeted cancer therapy.Abbreviation: mAb, monoclonal antibody.Adapted from Ben-Kasus T, Schechter B, Sela M et al. Cancer therapeutic antibodies come of age: Targeting minimal residual

disease. Mol Oncol 2007;1:42–54, copyright 2007, with permission from Elsevier.

Figure 2. 1970–1990: The oncogene decades.Abbreviations: EGFR, epidermal growth factor receptor; ERK, extracellular signal–related kinase; JNK, c-Jun N-terminal

kinase; MAPK, mitogen-activated protein kinase; MEK, MAPK/ERK kinase; PI3K, phosphoinositide 3-kinase; PKA, proteinkinase A; PTEN, phosphatase and tensin homologue deleted on chromosome ten.

24 The Biological Framework



Dow

nloaded from


BRAF, and phosphoinositide 3-kinase (PI3K), as well astranscription factors, such as Myc and p53.

Why are growth factors and their receptors so importantin cancer and in cancer therapy? The primary answer to thisquestion is that these factors control all phases of tumor de-velopment and metastasis, namely, when cancer cells at-tract a mutation, growth factors are responsible for theinitial expansion of a clone of cells. Later on, growth factorsare responsible for angiogenesis. And later on, they are re-sponsible for metastasis, by enhancing motility, penetra-tion, and invasion into blood vessels and lymph vessels, andcolonization of distant organs.

Perhaps most importantly, growth factors are responsi-ble for cell survival under cytotoxic drugs and radiotherapy.In other words, growth factors allow tumor cells to evadeand resist these therapies. This is the reason, perhaps, whythe combination of cytotoxic drugs and targeted therapieshas been so effective, because growth factors are importantin this final step in tumor progression.

TUMOR SIGNALING COMPLEXITY AND

CLINICAL MANIFESTATIONS

Signaling complexity underlies some clinical observationsin terms of patient response, patient resistance, and second-

ary resistance. Figure 3 approximates a map of the EGF re-ceptor (EGFR) signaling pathway, prepared by the group ofDr. Kitano of Sony Corporation’s Systems Biology Insti-tute in Tokyo, and demonstrates the extreme complexity in-herent in tumor cells [2]. Resolving this complexity ischallenging but important. EGFR and human epidermalgrowth factor receptor (HER)-2 are targets for severaldrugs, including gefitinib, erlotinib, lapatinib, cetuximab,panitumumab, and trastuzumab, and still more under devel-opment. It is interesting to note that these drugs targetEGFR and HER-2 in several different cancer indications—including lung cancer, breast cancer, and colorectal tumorswith cetuximab, and lung and pancreatic cancer with erlo-tinib, a kinase inhibitor.

The origin of biological complexity and the layered ar-chitecture of networks of cell–cell interactions stem fromwhole genome and isolated chromosome duplicationevents. Complexity and rich connectivity are the hallmarksof biological networks, such as the EGFR/HER-2 network.On a simplified level, when one drug is used to inactivatepathway A within a network, often there are several othercomponents and interactions that enable bypass of A. Forinstance, the many growth factors of the EGF family bind to10 different homo- and heterodimers, including het-erodimers of HER-2, to establish a layered signaling net-work [3]. The outcomes of gene expression programsregulated by the ErbB network are primarily cell prolifera-tion and cell migration (Fig. 4) [3].

Basic researchers like to conceptually view these bio-logical complexities and networks within an engineeringframework. As such, these researchers take lessons fromelectronic engineers and chemical engineers, posing ques-tions that attempt to resolve why a network developed in aparticular way, in other words: the logic of a network. Oneway to understand this logic is to determine the robustnessor failsafe function of the network, namely, the ability toovercome perturbations, inhibitors, mutations, or drug ef-fects (Fig. 5) [4].

One aspect of this robustness lies in system control,comprising both positive and negative feedback control. Ingeneral, positive feedback loops, such as transforminggrowth factor �, heparin-binding EGF, and ligands of theVEGF family, are elevated in cancer. Meanwhile, negativefeedback control is often defective in cancer.

One mechanism of negative feedback regulation that isfrequently defective in cancer entails growth factor–in-duced endocytosis and degradation of receptors like EGFRand HER-2. This receptor-mediated endocytosis requiresreceptor ubiquitination by an E3 ligase called CBL, casitasB-lineage lymphoma (Fig. 6) [5]. Cancer cells manipulatereceptor endocytosis in many ways [6]. One way relates to

Figure 3. Signaling complexity: The engineering perspec-tive.

Abbreviations: EGFR, epidermal growth factor receptor;GPCR, G protein-coupled receptor; HER-2, human epidermalgrowth factor receptor; MAPK, mitogen-activated protein ki-nase; PIP, phosphatidylinositol-bisphosphate.

Adapted by permission from Macmillan Publishers Ltd:Molecular Systems Biology. Oda K, Matsuoka Y, Funahashi Aet al. A comprehensive pathway map of epidermal growth fac-tor receptor signaling. Mol Syst Biol 2005;1:8–24, copyright2005.

25Yarden

www.TheOncologist.com



Dow

nloaded from


HER-2–overexpressing tumors of the breast: unlike EGFR,HER-2 is endowed with slow endocytosis and rapid recy-cling. Hence, when HER-2 is overexpressed, heterodimerscontaining both EGFR and HER-2 are shunted into a routethat avoids the lysosomal destination, hence permitting en-hanced signaling (Fig. 7) [7].

One manipulator of the EGFR system is HER-2 overex-pression. When HER-2 is overexpressed, it creates het-erodimers that enhance and prolong signals initiated byEGFR. As a result, EGFR no longer translocates throughthe endosome to degradation in the lysosome, but it recy-cles back to the cell surface. This is an example of a nega-tive feedback control that is defective in cancer.

Another, more subtle defect has to do with mRNAs thatare upregulated within minutes after stimulation with EGFor with related growth factors. This first wave of immediateresponse genes is followed by abrupt downregulation, andit includes oncogenic transcription factors such as c-Fosand c-Jun. Subsequently, there is a second wave of delayedearly genes, which are there to inactivate and suppress theoncogenic transcription factors like c-Fos and c-Jun [8].

Interestingly, not only repressors of transcription un-dergo enhanced expression; the group of delayed earlygenes (DEGs) also includes cytokines, mitogen-activatedprotein kinase phosphatases, and RNA-binding proteins.

Figure 4. HER-2/ErbB-2 and ErbB-3 collaborate within a framework of a layered signalling network.Adapted by permission from Macmillan Publishers Ltd: Nature Reviews Molecular Cell Biology. Yarden Y, Sliwkowski MX.

Untangling the ErbB signalling network. Nat Rev Mol Cell Biol 2001;2:127–137, copyright 2001.

Figure 5. Mechanisms that ensure robustness of biologicalnetworks.

Abbreviations: CBL, casitas B-lineage lymphoma; HB-EGF, heparin-binding epidermal growth factor; HSP, heatshock protein.

Adapted by permission from Macmillan Publishers Ltd:Nature Reviews Molecular Cell Biology. Citri A, Yarden Y.EGF-ERBB signalling: Towards the systems level. Nat RevMol Cell Biol 2006;7:505–516, copyright 2006.




Dow

nloaded from


This switch from activation of transcription to repression isdefective in cancer. As shown in Figure 8, many of theDNA- and RNA-binding proteins of the DEG group displaylow expression levels in a wide spectrum of tumors, relativeto normal tissues of the same type [8]. This represents an-other example of breakdown of negative feedback regula-tion in cancer.

Figure 6. E3 ligases and ubiquitin-binding proteins regulate EGFR endocytosis.Abbreviation: EGFR, epidermal growth factor receptor.Adapted by permission from Macmillan Publishers Ltd: Nature. Oved S, Yarden Y. Signal transduction: Molecular ticket to

enter cells. Nature 2002;416:133–136, copyright 2002.

Figure 7. Why is HER-2 oncogenic? HER-2 recycles EGFR.Abbreviations: EGFR, epidermal growth factor receptor;

HER-2, human epidermal growth factor receptor; Ub, ubiquitin.Adapted with permission from: Worthylake R, Opresko

LK, Wiley HS. ErbB-2 amplification inhibits down-regulationand induces constitutive activation of both ErbB-2 and epider-mal growth factor receptors. J Biol Chem 1999;274:8865–8874.

Figure 8. Delayed early genes are downregulated in carcino-mas. A human cancer compendium (1,975 published microar-ray studies) was analyzed for expression levels of the listedDEGs (left column).

Adapted by permission from Macmillan Publishers Ltd:Nature Genetics. Amit I, Citri A, Shay T et al. A module ofnegative feedback regulators defines growth factor signaling.Nat Genet 2007;39:503–512, copyright 2007.

27Yarden




Dow

nloaded from


MicroRNAs, tiny molecules that bind to mRNAs in aspecific way and are responsible for mRNA degradation,have been another area of interest. We recently found thatmicroRNAs are dynamic: in growth-arrested cells, manymicroRNAs are highly expressed. Once a normal epithelialcell is stimulated by EGF, several microRNAs are down-regulated within 20 minutes. The group of immediatelydownregulated microRNAs (ID-miRs) includes micro-RNAs that regulate c-Fos and c-Jun, along with other on-cogenic transcription factors. Overall, the ID-miRsrepresent yet another physiological mechanism of nega-tive feedback control that was found to be defective incancer [9].

The model illustrates an integration of DEGs and themicroRNAs that control them. Namely, in arrested cells,there is little mRNA made for the transcription factors thatare oncogenic. When cells are stimulated with EGF, withinminutes the group of microRNAs called ID-miRs is down-regulated, to allow the transcription factors to increase.Slightly later, the DEGs upregulate to suppress the onco-genic transcription factors. This is a complex transcriptioncontrol mechanism that is defective in cancer.

NETWORK VULNERABILITY AND

CLINICAL APPLICATIONS

Coming back to the issue of complexity from an engineer-ing standpoint, it is possible to design virtual networks andthen attack them, in an attempt to destroy them. Such stud-ies revealed that networks are trained to overcome commonperturbations, but they show extreme fragility to uncom-mon perturbations [10]. As such, researchers would like toidentify the uncommon perturbations that will fail such on-

cogenic networks. In addition, robust networks are hub ad-dicted, meaning that they develop an extreme reliance onsome hubs, such as energy and metabolism (e.g., glucoseand ATP), PI3K, Akt, or mammalian target of rapamycin.Consequently, uncommon drug interceptors may have theability to collapse a network [11].

One example of a clinical application of an uncommonperturbation is the use of immunotherapy, such as cetux-imab or trastuzumab, both of which bind to cancer cells. Inother words, antigen-dependent cellular cytotoxicity repre-sents a very uncommon reaction against a self antigen, andhas to do with the autoimmune response. Researchers donot currently know of any autoantibodies that target recep-tors like EGFR and HER-2, so these are very uncommonperturbations. Researchers have also found that exposure toheat shock protein (HSP) inhibitors, like geldanamycin or17-allylamino-17-demethoxygeldanamycin, result in rapidHER-2 degradation. Other uncommon perturbations oc-cur with drugs that bind to the ATP-binding site of thereceptor, like erlotinib and other drugs that inhibit kinaseactivity.

Drugs that simultaneously target more than one path-way, such as lapatinib, are relatively effective [12] becausethe network was trained to only resist one perturbation at atime. The combination of two simultaneous perturbations isvery uncommon, and did not appear while the network wastrained to resist these perturbations.

In other research, antibody combinations have dis-played synergistic tumor inhibition, resulting in completeinhibition of tumor growth (Fig. 9) [13]. The presentedstudies were conducted in an animal model of HER-2–overexpressing cells and tumor development. Treatment

Figure 9. Antibody combinations display synergistic tumor inhibition.Reprinted with permission from Ben-Kasus T, Schechter B, Lavi S et al. Persistent elimination of ErbB-2/HER2-overexpress-

ing tumors using combinations of monoclonal antibodies: Relevance of receptor endocytosis. Proc Natl Acad Sci U S A2009;106:3294–3299.




Dow

nloaded from


with one antibody resulted in some tumor inhibition, but thecombination of two antibodies resulted in a high degree oftumor inhibition and degradation of HER-2. The use of twoantibodies synergized, perhaps as a result of formation oflarge complexes at the cell surface that collapse into the cy-toplasm, translocate to the lysosome, and slowly degradeHER-2. It is hoped that this very uncommon perturbationcan someday be translated into clinical application.

CONCLUSIONS

Efficacious cancer therapies could potentially exploitfragile aspects of oncogenic networks, such as uncom-mon perturbations that target essential hubs through immu-

notherapy, combination of antibodies, HSP inhibitors, ornovel drug combinations. Drug resistance reflects systemplasticity, and may be overcome by drug combinations.Through evolution, cellular signaling pathways have beentransformed from relatively simple, linear cascades into ro-bust, multilayer, heavily complex signaling networks. Al-though such networks have been trained to overcome singleperturbations, or monotherapy, they have not been trainedto overcome complex, simultaneous perturbations, or com-bination therapy. System controls, such as receptor endocy-tosis and transcriptional feedback loops, are often defectivein cancer, and may be another source of exploitations withfuture therapies.

REFERENCES

1 Ben-Kasus T, Schechter B, Sela M et al. Cancer therapeutic antibodies

come of age: Targeting minimal residual disease. Mol Oncol 2007;1:

42–54.

2 Oda K, Matsuoka Y, Funahashi A et al. A comprehensive pathway map of

epidermal growth factor receptor signaling. Mol Syst Biol 2005;1:8–24.

3 Yarden Y, Sliwkowski MX. Untangling the ErbB signalling network. Nat

Rev Mol Cell Biol 2001;2:127–137.

4 Citri A, Yarden Y. EGF-ERBB signalling: Towards the systems level. Nat

Rev Mol Cell Biol 2006;7:505–516.

5 Oved S, Yarden Y. Signal transduction: Molecular ticket to enter cells. Na-

ture 2002;416:133–136.

6 Mosesson Y, Mills GB, Yarden Y. Derailed endocytosis: An emerging fea-

ture of cancer. Nat Rev Cancer 2008;8:835–850.

7 Worthylake R, Opresko LK, Wiley HS. ErbB-2 amplification inhibits

down-regulation and induces constitutive activation of both ErbB-2 and

epidermal growth factor receptors. J Biol Chem 1999;274:8865–8874.

8 Amit I, Citri A, Shay T et al. A module of negative feedback regulators

defines growth factor signaling. Nat Genet 2007;39:503–512.

9 Avraham R, Sas-Chen A, Manor O et al. EGF decreases the abundance of

microRNAs that restrain oncogenic transcription factors. Sci Signal 2010;

3:ra43.

10 Doyle J, Carlson JM. Power laws, highly optimized tolerance, and gener-

alized source coding. Phys Rev Lett 2000;84:5656–5659.

11 Weinstein IB. Cancer: Addiction to oncogenes—the Achilles heal of can-

cer. Science 2002;297:63–64.

12 Burris HA 3rd, Hurwitz HI, Dees EC et al. Phase I safety, pharmacokinet-

ics, and clinical activity study of lapatinib (GW572016), a reversible dual

inhibitor of epidermal growth factor receptor tyrosine kinases, in heavily

pretreated patients with metastatic carcinomas. J Clin Oncol 2005;23:

5305–5313.

13 Ben-Kasus T, Schechter B, Lavi S et al. Persistent elimination of ErbB-2/

HER2-overexpressing tumors using combinations of monoclonal antibod-

ies: Relevance of receptor endocytosis. Proc Natl Acad Sci U S A 2009;

106:3294–3299.

29Yarden




Dow

nloaded from


the biological framework: translational research from bench to

Documents