xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200409Ka 4500
controlfield tag 001 001970469
007 cr mnu uuuuu
008 081126s2008 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002330
Carie, Adam E.
Tumor suppressive effects of the Beta-2 adrenergic receptor and the small GTPase RhoB
h [electronic resource] /
by Adam E. Carie.
[Tampa, Fla.] :
b University of South Florida,
Dissertation (Ph.D.)--University of South Florida, 2008.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 201 pages.
Adviser: Sad M Sebti, Ph.D.
ABSTRACT: Receptor tyrosine kinases such as ErbB2 contribute greatly to human malignant transformation, but the role that other receptors such as 2 adrenergic receptor (B2 AR)play in cancer is ill defined. Furthermore, while some GTPases such as Ras and RhoA promote oncogenesis, RhoB has been suggested to have tumor suppressive activity. In this thesis the tumor suppressive activity of 2 adrenergic receptors through blockade of the Ras/Raf/Mek/Erk pathway is demonstrated. Furthermore, this thesis provides strong evidence in support of a tumor suppressive activity of RhoB, but not RhoA, in delaying EbB2 mammary oncogenesis in a transgenic mouse model. Chapter 1 describes a chemical biology approach that identifies a beta 2 adrenergic receptor agonist, ARA-211 (also known as pirbuterol) that suppresses the growth of cultured cells and of human tumors grown in nude mice by a mechanism involving stimulation of the 2 AR, cAMP production and activation of PKA, which in turn leads to the inactivation of C-Raf, Mek1/2 and Erk1/2. Chapter 2 describes the translation of these findings by ex-vivo treatment of fresh human tumor biopsies, with the ultimate goal of validating this novel therapeutic approach. Chapter 3 describes the generation of transgenic mice that over express ErbB2 along with either RhoB or RhoA to determine the effects of these two small GTPases on ErbB2-mediated mammary tumorigenesis. The findings indicate that overexpression of RhoB, but not RhoA, results in decreased multiplicity and delay in the tumor onset mediated by ErbB2 overexpression.In summary, this thesis work resulted in the discovery of how crosstalk between the 2 AR/cAMP/PKA circuit with the Raf/Mek/Erk1/2 cascade leads to tumor suppression; and the discovery of the suppression of ErbB2-mediated breast cancer by the GTPase RhoB.
MAP Kinase Signaling System.
Receptors, Adrenergic, beta-2.
Receptor Protein-Tyrosine Kinases.
rhoA GTP-Binding Protein.
rhoB GTP-Binding Protein.
Cyclic AMP-Dependent Protein Kinases.
Proto-Oncogene Proteins c-raf.
Guanine Nucleotide Exchange Factors.
Tumor Cells, Cultured.
Protein kinase A.
x Cancer Biology
t USF Electronic Theses and Dissertations.
Tumor Suppressive Effects of the Beta-2 Adrenergic Receptor and the Small GTPase RhoB by Adam E. Carie A dissertation submitted in partial fulfillment of the requirement s for the degree of Doctor of Philosophy Department of Cancer Biology College of Graduate Studies University of South Florida Major Professor: Sad M Sebti, Ph.D. Srikumar Chellappan, Ph.D. Hong-Gang Wang, Ph.D. Noreen Luetteke, Ph.D. Date of Approval: March 24, 2008 Keywords: MAP Kinase, Protein Kinas e A, cAMP, Rho, Tumor Suppression Copyright 2008, Adam Carie
Dedication This work is dedicated to the love d ones that supported me during my educational career. To my family: my wife, parents, sister and brothers in law, parents in law and grandparent s, your confidence in me and guidance helped me to remain steadfast in the pursuit if my do ctorate. I would al so like to dedicate this work to memory of my late mother Chereen Ann Carie, th e inspiration for my interests in cancer research.
Acknowledgements I would like to thank my major profe ssor and mentor Dr. Sad Sebti for all of the time and effort that was invested in my education and training. You always encouraged me to challenge myself, as we ll as dogma, and to not take anything for granted. You taught me how to think anal ytically and to dissect problems in a rational manner. You gave me an opportuni ty to experience research as an undergraduate student, which pl anted the seed that flour ished over the next seven years. I am forever grateful for inst illing in me characteristics that, in part, have made me who I am today. I would also like to thank my committee members, Dr. Srikumar Chellappan, Dr. Noreen Luetteke, and Dr. Hong-Gang Wang for their guidance and excellent suggestions for my projects. I appreciate your efforts and approachability, your doors were always open and you were always willing to help. Special thanks goes to Dr. Ad rienne Cox, the outside chair for my dissertation defense. Your comments and suggestions help ed to mold this thesis into a more complete work. There are many people that have hel ped me tremendously throughout the years in the Sebti lab, and through intrai nstitutional collaborations. I would especially like to thank Dr. JiaZhi (G eorge) Sun for taking me in as an
undergraduate student and teachi ng me techniques and ski lls I would utilize for the remainder of my graduate career. I would also like to thank my bench buddy Michelle Blaskovich for sharing her technical expertise, as well as the occasional buffer, and always making life interesting in the lab (especia lly when George was around). Likewise, I would like to thank Dr. Aslamuzzaman Kazi and Dr. DeAn Wang for their hard work on our collaborativ e projects, it was a pleasure working with you both. Finally, to the rest of the Sebti lab members past and present, thank you for always treating me like fami ly, we were a tight knit group and our friendship will continue long after my graduate career is over. This thesis work is dedicated to my lo ving family, and to the memory of my late mother, Chereen Ann Carie, who was my inspiration for getting involved in cancer research. I would never have been able to go the distance and complete the Ph.D. program without the loving suppor t of my family and friends, thank you all.
Note to Reader The original of this document contains color that is necessary for understanding the data. The original diss ertation is on file with th e USF library in Tampa, Florida.
i Table of Contents List of Tables iii List of Figures vii Abstract ix Background and Significance 1 Chapter 1: A Chemical Biology Approach Identifies a Beta-2 Adrenergic Receptor Agonist that Causes Human Tumor Suppression by Blocking the CRaf/Mek1/2/Erk1/2 Pathway Abstract 14 Introduction 15 Materials and Methods 26 Cell Lines and Transfection 26 Cytoblot Screening for Small Molecules that Decrease Phospho-Erk1/2 27 Western Blotting 27 In Vitro Kinase Assay 28 In Vivo Kinase Assay 30 MTT Assay 30 Alamar Blue Proliferation A ssay 30 Soft Agar Assay 31 Apoptosis Assay 31 Total Cellular cAMP Assay 32 PKA Kinase Assay 33 Tumor Suppression in Nude Mouse Xenograft Models 33 Immunohistochemistry and Slide Quan titation 34 Results 36 Discussion 75
ii Chapter 2: Validation of ARA-211 Activity Ex-Vivo in Fresh Biopsies from Patient Samples Abstract 84 Introduction 85 Materials and Methods 88 Human Tissue Array 88 Immunohistochemistry 89 Treatment of Patient Samples Ex-vivo 90 Results 91 Discussion 124 Chapter 3: RhoB, but not RhoA Overexpression Delays ErbB2Mediated Mammary Tumor Onset and Reduces Tumor Multiplicity in Transgenic Mice Abstract 129 Introduction 131 Materials and Methods 137 cDNAs and Gene Subloning 137 Southern Blot and Genotyping 156 DNA Preparation and PCR 158 Protein Preparation and Anal ysis 158 Tumor Onset, Growth Rate, and Multiplicity Calculation 159 Results 161 Discussion 172 Conclusions and Future Directions 176 List of References 186 About the Author End Page
iii List of Tables Table1. Chemical structures of the compounds identified as Erk1/2 inhibitors from NCI Diversity Set 39 Table 2. Structure-activity relationship studies identify the ARA family of inhibitors as analogues of epinephrine, and identifies ARA-211 as pirbuter ol 45 Table 3. ARA-211 sensitive and insens itive cell lines tested from the NCI 60 cell line panel do not correlate with C-Raf and B-Raf expression levels 60 Table 4. Physiological effects of ARA-211 treatment on sensitive cell lines versus insensit ive cell lines 71 Table 5 Summary of tissue array results for 2 AR and P-Erk1/2 105 Table 6 Summary of fresh human tumor biopsy staining results for 2 AR and P-Erk1/2 109 Table 7 RhoB, but not RhoA overexpression results in a significant delay in ErbB2-mediated tumor onset and multiplicity, but not tumor gr owth rate 169
iv List of Figures Figure 1. Cancer cells contain numerous, aberrant signal Transduction circuits 2 Figure 2. Aberrant signaling through receptor tyrosine kinases (RTKs) and downstream signaling through Ras is well characterized 4 Figure 3. RhoB, a close family member of RhoA, antagonizes Ras malignant transformation and has been suggested to have tumor suppressive effe cts 6 Figure 4. High throughput screening of the NCI Diversity Set by cytoblot analysis reveals 3 inhibitors of Erk1/2 activation 38 Figure 5. ARA-family of compounds se lectively inhibit P-Erk1/2 in MDA-MB-231 breast cancer cells 40 Figure 6. ARA-211 inhibits the kinase ac tivity of Raf-1, but not B-Raf, by inhibiting upstream acti vation of Raf 42 Figure 7. ARA-211 stimulates cAMP selectively through 2 adrenergic receptor activation 47
v Figure 8. ARA-211 mediated cAMP stimulation through 2 AR activation results in inhibition of P-Erk1/2 48 Figure 9. ARA-211 mediated cAMP stimulation through 2 AR activation results in i nhibition of P-Erk1/2 through PKA activation, not EPAC activation of Rap1 51 Figure 10. siRNA and kinase assays validate that ARA-211 mediated cAMP stimulation re sults in inhibition of P-Erk1/2 through PKA activation 52 Figure 11. ARA-211 inhibition of MDA-MB -231 cell proliferation requires 2 AR and inhibition of P-Er k1/2 54 Figure 12. Direct activation of adenyl cyclase by forskolin inhibits the proliferation of MD A-MB-231 cells 55 Figure 13. ARA-211 mediated inhibi tion of MDA-MB-231 cell proliferation is dependent on the ability to inhibit Mek1/2 and to decrease P-Erk1/2 levels 57 Figure 14. Western blot verification of 2 AR, P-Erk1/2, C-Raf and B-Raf expression levels from the NCI 60 cell line panel 62 Figure 15. ARA-211 induces cAMP formation in MDA-MB-231, SF-539 and ACHN cells but not in A549, SNB-19 and HCT-116 cells 65
vi Figure 16. ARA-211 inhibits anchorage independent growth in MDA-MB-231, SF-539 and ACHN cells but not in A549, SNB-19 and HCT-116 cells 66 Figure 17. ARA-211 induces apoptosis in MDA-MB-231, SF-539 and ACHN cells but not in A549, SNB-19 and HCT-116 cells 67 Figure 18. ARA-211 treatment of MDA -MB-231 xenografts results in tumor regression, and treat ment of AHCN xenografts completely inhibits tumor growth in nude mice 69 Figure 19. ARA-211 treatment of mi ce with A549, HCT-116 and SNB-19 xenografts has no effect on tumor growth 70 Figure 20. ARA-211 suppresses P-Erk1/2 levels, inhibits tumor cell growth and induces apoptosis in human xenografts in nude mice 73 Figure 21. ARA-211 stimulates crosstalk between 2 AR and Raf/MEK/Erk1/2 resulting in inhibition of P-Erk1/2 through PKA, but not EPAC 78 Figure 22. Tissue array identifies human breast tumor samples that express 2 AR and P-Erk1/2 93 Figure 23. Tissue array identifies human liver tumor samples that express 2 AR and P-Erk1/2 95
vii Figure 24. Tissue array identifies human bladder tumor samples that express 2 AR and P-Erk1/2 97 Figure 25. Tissue array identifies human ovarian tumor samples that express 2 AR and P-Erk1/2 99 Figure 26. Tissue array identifies human pancreatic tumor samples that express 2 AR and P-Erk1/2 101 Figure 27. Tissue array identifies human prostate tumor samples that express 2 AR and P-Erk1/2 103 Figure 28. Fresh patient samples from Moffitt Cancer Center express P-Erk1/2 and 2 AR 108 Figure 29. Ex-vivo treatment of fresh hum an tissue demonstrates variable efficacy of ARA-211 to inhibit P-Erk1/2, inhibit proliferation and induce apoptosis 115 Figure 30. Patient treatment with ARA-211 requires predetermination of 2 AR and P-Erk1/2 expression, as well as ex-vivo treatment to determine efficacy to inhibit proliferation and induc e apoptosis 123 Figure 31. MMTV-TGF construct linearized by EcoRI digestion for insertion of human RhoA and R hoB genes 139 Figure 32. Isolation of MMTV, RhoB and RhoA DNA and confirmation of insertion in the correct orientation of RhoB and RhoA sequences in the MMTV vector 142
viii Figure 33. RhoA, but not RhoB DNA insertion is confirmed in MMTV vector after restriction en zyme digestion 146 Figure 34. pcDNA3 vector is linearized by KpnI and XhoI digestion and MMTV-RhoB is inserted into the plasmid 148 Figure 35. MMTV-RhoB insertion into the pcDNA3 vector is verified by the presence of the 2.7 kb MMTV-RhoB ligation band 151 Figure 36. Verification of the in tegration of the 600 bp MMTV-RhoB insert 153 Figure 37. Final isolation of MMTV-RhoA and MMTV-RhoB transgenes 154 Figure 38. Southern blot of DNA fr om transgenic founder mice confirms integration of RhoA or RhoB into the host genome 157 Figure 39. Schematic for generation of MMTV-RhoB and MMTV-RhoA constructs for generation of transgenic mice that over express human HA-RhoB and HA-RhoA under the MMTV prom oter 163 Figure 40. RhoB, but not RhoA overexpression results in a significant delay in ErbB2-medi ated tumor onset 168
ix Figure 41. Tumors from EB and EA transgenic mice express RhoB and RhoA as determined by detection of HA by western blot 171
x Tumor Suppressive Effects of the Beta-2 Adrenergic Receptor and the Small GTPase RhoB Adam Carie ABSTRACT Receptor tyrosine kinases such as ErbB2 contribute greatly to human malignant transformation, but the role that other receptors such as the 2 adrenergic receptor play in cancer is ill defined. Furthermore, while some GTPases such as Ras and RhoA promote oncogenesis RhoB has been suggested to have tumor suppressive acti vity. In this thesis the tumor suppressive activity of 2 adrenergic receptors ( 2 AR) through blockade of the Ras/Raf/Mek/Erk pathway is demonstrated. Furthermore, this thesis provides strong evidence in support of a tumor s uppressive activity of RhoB, but not RhoA, in delaying EbB2 mammary oncogenesis in a transgenic mouse model. Chapter 1 describes a chemical biology approach that identifies a beta 2 adrenergic receptor agonist, ARA-211 (also k nown as pirbuterol) that suppresses the growth of cultured ce lls and of human tumors grown in nude mice by a mechanism involving stimulation of the 2 AR, cAMP production and activation of PKA, which in turn leads to the inactivation of C-Raf, Mek1/2 and Erk1/2. Chapter 2 describes the translation of these findings by ex-vivo treatment of fresh
xi human tumor biopsies, with the ultimate goa l of validating this novel therapeutic approach. Chapter 3 describes the generat ion of transgenic mice that over express ErbB2 along with eit her RhoB or RhoA to determine the effects of these two small GTPases on ErbB2-mediated mammary tumorigenesis. The findings indicate that overexpression of RhoB but not RhoA, results in decreased multiplicity and delay in the tumor onset mediated by ErbB2 overexpression. In summary, this thesis work result ed in the discovery of how crosstalk between the 2 AR/cAMP/PKA circuit with the Ra f/Mek/Erk1/2 cascade leads to tumor suppression; and the discovery of the suppression of ErbB2-mediated breast cancer by the GTPase RhoB.
1 Background and Significance Accumulation of genetic alterations by overexpression or hyperactivation of oncogenes and inactivation or deletion of tumor suppressor genes leads to aberrant signal transduction, which confers the transforming cellular properties that result in hallmarks of cancer. These properties include self-sufficiency in growth signals and resistance to anti-gr owth signals, evasion of apoptosis, the ability to induce and sustain angiogenes is, the capacity to metastasize and invade foreign tissue, and the potential to r eplicate limitlessly. Ultimately, the aberrant signal transduction provides a mechanism for the transformation of normal cells to malignant. In general, this aberrant signaling will be the focus of my thesis, however, there ar e a multitude of irregular signaling pathways found in cancer cells. Figure 1 demonstrates just a few of the many pathways known to be misregulated in cancer.
Figure 1. Cancer cells contain numerous, aberrant signal transduction circuits2AR 2
3 One of the major signal transducti on pathways that bec ome aberrant in cancer is that of recept or tyrosine kinases (RTKs). For example, receptor overexpression or mutation of kinase doma ins leads to persistent activation of downstream signals. The ErbB family of receptor tyro sine kinases is the most intensely studied. There are four mem bers of the ErbB family: ErbB1 (EGFR, Epidermal Growth Factor Receptor), Er bB2, ErbB3 and ErbB4(1). Ligands for these receptors induce homo or heterodime rization leading to activation of intracellular tyrosine kinase domains(2, 3) There is no known ligand for ErbB2, but epidermal growth factor (EGF) can induce heterodimerization of ErbB1 and ErbB2(4, 5). Similarly, neuregulins are known to activate ErbB3 and ErbB4, which can both complex with ErbB2(6). When activated, ErbB2 heterodimers are potent signal transducers, in part, due to th eir relatively slow rates of receptor internalization and ligand diss ociation(4, 7). ErbB2 gene amplification leading to receptor overexpression is common among breast cancer, occurring in approximately 26% of all cases(8). Signal transduction pathw ays downstream of ErbB2, and many other gr owth factor receptors, become aberrant in cancer, including persistent activation or mutation of Ras proteins(3). Ras is a central signaling node found mutated in 30 % of all human tumors(9) that confers transforming signals through activati on of downstream cascades such as Raf/Mek/Erk, PI3K/Akt and RalGDS/RalA and RalB, as shown in Figure 2. These signals have been shown to result in cell proliferation, survival and metastasis when hyperactive(10).
Raf Mek ErkAkt Grb2 P Ras SOS PI3K Ras P RalGDSRalA/B Erk DNANucleus FosMyc RTK PDGF VEGF EGF Figure 2. Aberrant signaling through receptor tyrosine kinases (RTKs) and downstream signaling through Ras is well characterized 4
5 In addition to Raf, PI3K and RalG DS mediating Ras cancer-causing activity, other small G-proteins such as RhoA also mediate Ras oncogenesis. The mechanism by which Ras communicate s with RhoA is unknown, but there is evidence supporting indirect activation of R hoA by Ras resulting in proliferation and increased cell motility(11, 12). There is evidenc e supporting regulation of RhoA and Rac1 activity by the Mek/Er k MAPK pathway(13). Rho kinase (RhoK) is an immediate downstream effector of RhoA that st imulates proliferation and migration, possibly through c-MYC as a dow nstream effector(14). Surprisingly, a close family member of RhoA, RhoB whic h shares 86% amino acid identity with RhoA, has been suggested to have tumor suppressive activity. Our lab and others have demonstrated in-vitro that overexpression of RhoB, but not RhoA, antagonizes the transforming activity of ErbB 2, EGFR, H-Ras, PI3K and Akt, but not c-Myc(15, 16). These studies also provide evidence supporting the antitumor effects of RhoB through inhibition of phospho-Akt as illustrated in Figure 3.
Figure 3. RhoB, a close family member of RhoA, antagonizes Ras malignant transformation and has been suggested to have tumor suppressive activityAkt PI3K Grb2 P Ras SOS DNANucleus PDGF VEGF EGF RhoB RhoB Malignancy, Apoptosis,Uncontrolled growth, Migration, Invasion,Metastasis RTK 6
7 Furthermore, other groups have demonstrated that RhoB is mainly localized to endosomes and can limit cell proliferation, survival and invasion and metastasis(17, 18). RhoB has been sugg ested to target K-Ras, B-Raf, Cdc6, spindle checkpoint assemblies, and Lipocor tin to induce apoptiosis regulated by inhibition of farnesyl transferase for c ancer therapy(19). RhoB knockout mice studies show that RhoB is not essential to normal development, however the mice showed an increased sensitivity to chemical carcinogenesis by 7,12dimethylbenz[ ]anthracene as well as increased efficiency of intraperitonel tumor formation(20). Likewise, mouse embr yonic fibroblasts lacking RhoB showed increased sensitivity to Ras and E1A tr ansformation-mediated cell adhesion, and increased sensitivity to TGF signaling(20). Also, targeted ablation of RhoB in transformed cells was shown to protect ce lls from Taxol-induced apoptosis, but not doxorubicin or UV irr adiation, suggesting an apoptotic checkpoint role of RhoB in actin cytoskeleton remodeling(21). Similarly, IHC studies in paraffinembedded tumor specimens from lung and head and neck cancer patients have demonstrated that RhoB expression is decreased as cancers progress from hyperplastic to deeply invasive carci noma by immunohistochemistry(22, 23). Finally, oncogenic proteins such as ErbB2, EGFR, Ras, PI3K and Akt all repress the expression of RhoB at the transcripti onal level(15). Taken together, these studies suggest that as normal cells acquire initial genetic alterations, RhoB is induced and serves as a gatekeeper to suppress the growth and induce apoptosis of these cells. However, as the cells accumulate more genetic alterations such as ErbB2, EGFR, Ra s, PI3K and Akt, these transcriptionally
8 repress RhoB. Support for this hypothes is was provided by the Sebti lab that showed that forced expressi on of RhoB reverses malignant transformation by these oncogenes(15). However, in-vivo evidence in animal models to support the above hypothesis is lacking. Chapter 3 of this Thesis provides evidence for the role of RhoB, but not its closely rela ted family member R hoA, in suppressing human ErbB2-driven breast cancer in a m ouse model where ErbB2 expression is regulated by the mouse mammary tumor vi rus (MMTV) promoter. This model is highly relevant to human breast cancer wh ere ErbB2 is believed to contribute to at least 26% of the progressi on of this disease(24). In Chapter 3 the RhoB and RhoA transgenic mice were crossed with the ErbB2 overexpressing mice to study the effects of RhoB and RhoA overexpression on ErbB2-mediated tumor onset, multiplicity and growth rate. In Chapter 1 and 2 of this thesis the relevance of the crosstalk between the Ras/C-Raf/Mek1/2/Erk1/2 and the beta 2 adrenergic receptor ( 2 AR) signaling circuits to suppression of tumor growth is studied. The 2 AR is a well-studied and characterized pathway in normal cell physiology. However, we became interested in this signal transduction pathway as we init ially searched for novel inhibitors of the Raf/Mek/Erk ki nase cascade. Adrenergic receptors belong to a large super-family of membrane s panning proteins known as G-protein coupled receptors. This family contains over 500 members, each containing a membrane-spanning domain of seven -helices principally made up of 22-28 hydrophobic amino acids each( 25-27). When a ligand binds its specific receptor a conformational change takes place that allows for the activation of specific
9 heterotrimeric G-proteins. Upon activation, the G-proteins active subunit, such as G s releases GDP and binds GTP, allowi ng the active subunit to dissociate from the other subunits (such as & ) and affect other proteins in its signal transduction pathway. In some cases the -dimer can also transmit signals that mediate receptor response. It is well documented that adrenergic receptors regulate many facets of the sympathetic ne rvous system. One of the first second messenger systems discovered was t he adrenergic receptor-cAMP-Ca ++ pathway. Stimulants such as epi nephrine and isoproterenol bind the adrenergic receptors resulti ng in the release of G s subunits, which stimulate the enzyme adenyl cyclase(28, 29). Adenyl cycl ase then catalyses the conversion of ATP to cAMP, leading to the activation of cAMP-dependant protein kinases, ultimately resulting in the opening of Ca ++ channels and a biological response such as increased heart rate(30, 31). On the other hand, inhi bitory agonists such as adenosine cause the release of inhibitory G i subunits that decrease adenyl cyclase activity. The extensive research in this area of adrenergic receptor function has lead to the discovery of many small molecule drugs that target the different receptors, and lead to physiolog ical responses in heart rate, smooth muscle contraction or dilati on and nerve conduction. Mo re recent research has demonstrated that cAMP signa ling can modulate normal ce ll proliferation through crosstalk with signal transduction pathwa ys such as the Raf/Mek/Erk1/2 kinase cascade. Cell proliferation in co mplex organisms is contro lled by numerous systems involving hundreds of different proteins and signaling mechani sms. One way in
10 which cAMP is involved in cell prolifer ation is through t he modulation of the mitogen activated protein kinase (MAPK) ca scade. The effects of cAMP on the levels of Erk1/2 activation in normal ce lls are widely variab le depending on the type of cells and the intracellular pathway s that crosstalk in these cells. One mechanism in which ERKs are activat ed via cAMP has been shown to involve stimulation of B-Raf, which activates th e Mek/Erk1/2 pathway(32, 33). The means through which cAMP activates B-Raf is still unclear, yet probable mechanisms have been proposed involving PKA, Src, and Rap-1( 32-35). In cells that express B-Raf it is suggested that Rap-1 activation by Src leads to stimulation of B-Raf and ultimately in t he activation of Erk1/2(34). Furthermore, In some leukemia and normal kidney cells studies have shown that cAMP-GEFs, or EPACs (exchange proteins directly activated by cAMP), have an affinity for Rap-1(36-38). This demonstrates ye t another mechanism in which cAMP can stimulate Erk1/2 activity, but in a PKA-independent manner. While in many normal cells, such as rat thyroid ce lls, bone cells, polycystic kidney epithelium, Sertoli cells, cardiac myocytes, granulosa cells, pre-adiposites, pituitary cells and PC12 cells, cAMP has been shown to stimul ate Erk1/2 and prol iferation(39-46), only one study in leukemia cells has show n that EPACs activate Erk1/2(47). There have also been a multitude of mechanisms proposed for the inhibitory consequences of cAMP on the pr oliferation of normal cells, such as in adipocytes, endothelial cells, NIH 3T3 cell s, rat fibroblasts, smooth muscle cells, hepatocytes and pancreatic ac inar cells(40, 48-56).. The mechanisms by which this could occur include inhibition of Erk1/2 through a PKA-dependent activation
11 of Rap-1, up-regulation of Erk phosphatases (MKPs), and inhibition of Raf-1 kinase activity via direct phosphoryl ation by PKA(49, 57-61). A generally accepted mechanism in which cAMP inhibi tion of Erk1/2 takes place is through the ability of PKA to block Raf-1 activity Upon cAMP stimulation the catalytic subunits of PKA are free to phosphorylate downstream tar gets, including Raf-1. Phosphorylation of Raf-1 by PKA on se rine 621 inhibits the Raf-1 kinase domain(50, 52, 62-65). Other inhibito ry phosphorylation sites on Raf-1 have been proposed, including serine 259 and 43, but it is not known if PKA subunits phosphorylate these sites in-vivo An alternative mechanism that has been proposed is that of PKA activating c-Src, in turn activating Rap-1, which in the absence of B-Raf, competes with C-Raf for binding sites to Ras(34, 37, 38, 59). While in normal cells extensive studies have investigated the role of cAMP in cell proliferation, only a few studies were carried out in tumor cells. These studies showed that in breast cancer and neuroblas toma cancer cells, cAMP induction results in inhibition of DNA synthesis and induction of apoptosis, respectively(66, 67). However, the effects of cAMP on inhibition of the Ras/Raf/Mek/Erk pathway has not been investigated in tumors. We believe that there is an unmet need to investigate whether stimulat ion of the adrenergic receptor can be used as a novel approach to cancer chemotherapy. Fundamental questions that need to be addressed include first to determine w hether there is crosstalk between the AR signaling circuit and the Ras/Raf/Mek/Erk circuit. Second, if this crosstalk exists, then to determine whether AR stimulation results in ac tivation or inactivation of the MAPK pathway, and whether this affect s several hallmarks of cancer cells
12 such as uncontrolled prol iferation, anchorage-depe ndent and independent growth, apoptosis and tumor growth in animals. Third, the mechanism of the crosstalk should be investigated as we ll. Therefore, determining how, for example, stimulation of the AR regulates the Raf/Mek/Erk circuits in cancer is of prime importance. Other important questions that need to be addressed are determining the types of AR involved (i.e. 1 or 2). Finally, the prevalence of the AR circuits in fresh tumor biopsi es, especially those that contain persistently activated Ras/Raf/Mek/Erk signaling needs to be investigated. Chapters 1 and 2 of this Thesis will address the important issues above.
13 Chapter 1 A Chemical Biology Approach Identi fies a Beta-2 Adrenergic Receptor Agonist that Causes Human Tumor Suppression by Blocking the CRaf/Mek1/2/Erk1/2 Pathway Data from this Chapter was published by Adam Carie and Said Sebti in the journal Oncogene, i ssue 26, 3777-3788, 2007. All of the work in this Chapter was per formed by Adam Carie except for Figure1 cytoblot, which was performed by Adrian Kenny
14 Abstract A chemical biology approach identifie s a beta 2 adrenergic receptor ( 2AR) agonist ARA-211 (pirbuterol), which causes a poptosis and human tumor regression in animal models. 2AR stimulation of cAMP formation and PKA activation leads to C-Raf (but not B-Raf) kinase inactiva tion, inhibition of Mek1/2 kinase and decreased phosphoErk1/2 levels. ARA-211induced inhibition of the C-Raf/Mek1/2/Erk1/2 pathway is mediated by PKA an d not EPAC. ARA-211 is selective and suppresses P-Erk1/2 but not P-JNK, P-p38, P-Akt or P-STAT3 levels. 2AR stimulation results in inhibition of anchorage-dependent and independent growth, induction of apoptosis in vitro and tu mor regression in vivo. 2AR antagonists and constitutively acti ve Mek-1 rescue from the effects of ARA-211 demonstrating that 2AR stimulation and Mek1/2 kinase inhibition are required for ARA-211 antitumor activity. Furthermore, suppression of growth occurs only in human tumors wher e ARA-211 induces cAMP formation and decreases P-Erk1/2 levels. Thus, 2AR stimulation results in significant suppression of malignant transformation in cancers where it blocks the CRaf/Mek1/2/Erk1/2 pathway by a cAMP -dependent activation of PKA, but not EPAC.
15 Introduction Malignant transformation of normal cells to cancer cells requires the acquisition of several oncogenic traits such as uncontrolled cell division, resistance to programmed cell death (apoptosis), invasion and angiogenesis(68). These malignant transformation traits are believed to be the consequence of the accumulation of genetic alterations that re sult in deregulated signal transduction circuits(68). Such genetic alterations can simply be point mutations that result in constitutive activation of key signal transducers such as Ras, a GTP/GDP (gunosine triphosphate/guanosine diphosphate) binding GTPase that contributes to 30% of all human cancers(9). Alter natively, these alterations can involve entire genes such as the receptor ty rosine kinases epidermal growth factor receptor and ErbB2 that are found over expressed in many major human cancers such as breast tumors(69, 70). Often such genetic alterations result in constitutive activation of common downstr eam signal transduction pathways such as the mitogen activated protein kinase (MAPK) cascades of several serine/threonine kinases including C-Ra f, Mek1/2, Erk1/2, p38 and JNK (Jun kinase)(64, 71-73). Other pathways involved in uncontrolle d proliferation, apoptosis, invasion and angiogenesis also include those mediated by the signal transducer and activator of transcrip tion 3 (STAT3) and the serine/threonine
16 kinase Akt(74, 75). Not only have these oncogenic and tumor survival pathways been found constitutively activated in the great majority of human cancers, but also their hyperactivation has been associated with poor prognosis and resistance to chemotherapy in cancer patients(69, 70) This has prompted drug discovery efforts targeting receptor tyrosine kinases, Ras, C-Raf, Mek, Akt and STAT3 to thwart aberrant signal transduction pathways in tumor cells (65, 76-79). For example, inhibitors of receptor tyrosine kinases (RTKs), Ras, C-Raf and Mek1/2 have all been identified and are at various stages of development(63, 80). Some, like the epidermal growth fact or receptor (EGFR) kinase inhibitor Tarceva, was approved by the FDA for t he treatment of patients with lung cancer(81). Genetic abnormalities that often predispose or lead to cancer, such as overexpression of growth factors, mutations leading to constitutive activation of cellular receptors, deletion or misregulat ion of tumor suppre ssors, or uncontrolled signal transduction pathways often signal directly through the small GTPases Ras. Ras proteins are central nodes in the cellular signaling response from growth factors and receptors. They ar e localized to the cellular membrane regions by a lipid post-trans lational modification known as prenylation(9, 82-87). Ras activation occurs through the exchange of guanosine diphosphate nucleotides for guanosine triphosphate nucleotides catalyzed by exchange factors known as GEFs (guanosine nucleot ide exchange factors) Activated Ras proteins can interact with more than 20 downstream effectors, the bestcharacterized being Raf, PI3K (phos phatidylinositol-3-kinase), and Ral
17 GTPases. These effectors ar e responsible for the prolif erative, prosurvival, and differentiation effects from Ras signaling(9, 82-87). RasGTP-mediated prosurvival signaling occurs at least in part through activation of the regulatory subunit of PI3K, which target s it to the membrane to phosphorylate PIP2 (phosphatidylinositol-4,5 -bisphosphate), converting it to PIP3 (phosphatidylinositol-3,4,5-triphosphate)(74, 88, 89). PIP3 recruits PDK and Akt through their PH domains, which leads to the phosphorylation of Akt by PDK1 and other kinases(74). Akt is a kinase t hat has been shown to phosphorylate and inactivate pro-apoptotic proteins including BAD and FKHR (forkhead homologue 1) transcription factors, leading to survival under circumstances that cells would normally undergo apoptosis(74). Furthermore, PI3K has been shown to activate Rac, a Rho family protei n that is important for cytoskeletal rearrangement leading to transformation, invasiveness and metastasis(90-93). The PI3K/Akt pathway has been the fo cus of major drug discovery and development efforts fo r cancer therapy. Another important cellula r signaling component downstream of Ras is the Ral GDS/Ral pathway. Ral GEFs, such as RalGDS (Ral guanine nucleotidedissociation stimulator), ac tivate the Ras-like Ral GTPase proteins RalA and RalB. Similar to Ras, the Ral proteins are reliant on preny lation, whereby a 20carbon lipid geranylgeranyl group is added to the C-terminus, leading to membrane anchorage for downstream signaling to effectors(94-98). The downstream effectors of RalA and RalB have been suggested to be involved in early transforming events due to aberrant Ras signaling. Recent ly, in cell culture
18 models, the RalGEF-Ral pathway was shown to be required for Ras mediated transformation(94, 95). Likewise, in a knockout mouse model lacking RalGDS, skin tumors from carcinogen-induced Ras showed impaired growth(99). Although RalA and RalB share 85% sequence identity, it has recently been discovered that these proteins play di stinct roles in anc horage-independent growth and cell survival, respecti vely(94). Furthermore, our lab has demonstrated that Ral proteins are ex clusively geranygeranylated, and that inhibition of the geranyl geranylation of RalB mediates GGTI-induction of apoptosis and inhibition of an chorage-independent growth. In contrast, inhibition of RalA geranylgeranylation mediates GGTI-induced inhibition of anchorage independent growth(94). Currently, GGTI compounds are still in preclinical development; however, clinical candidate s are not far off from human trials. Development of GTPase inhibitors is in its infancy compared to the efforts to develop selective kinase inhibitors. One of the first discovered and most targeted kinase pathways is downstream of Ras, and belongs to the mitogen activated protein kinase (MAPK) family. MAPKs are the effectors in a line of kinases that transduce signals from growth factors and mitogens from the cell surface to the nucleus. With a large number of activators the specific physiol ogical message of the external ligand is translated into cellular responses. Depending on the stimuli, these responses include proliferation, differ entiation, and survival. There are four major families of mammalian MAP kinase proteins. They ar e all known to be activated by growth factors, but only the Raf/Mek/Erk pat hway has not been linked to stress
19 activation. The p38, JNK an d Erk5 pathway are activated by cellular stress, as well as by cytokines and growth factor s(100). The signaling circuitry provides three steps at which signaling can be posi tively or negatively regulated. Once activated by upstream signals, the fi rst MAP kinase kinase kinase (MAPKKK) phosphorylates MAP kinase kinase (MAPKK), which in turn, phosphorylates MAPK kinase (MAPK) to target transcription factors to either activate or repress gene expression(100). These kinases ar e evolutionarily conserved among mammals and are important for devel opment as well as homeostasis. The Ras/Raf/Mek/Erk kinase cascade has been scrutinized and targeted in the past 20 years. Ras activation of Raf proteins, most notably C-Raf and BRaf, leads to activation of Mek kinases (87, 101, 102). The pathway bottlenecks at Mek, as Mek1 and Mek2 are the onl y known activators of Erk1 and Erk2. These terminal serine threonine kinas es phosphorylate and re gulate over 150 different proteins(103). Most of the activated Erk1/2 goe s to the nucleus where it activates transcription factors that control gene expression and functions such as cell proliferation and ev asion of apoptosis(104). T hese effectors play an important role in the onc ogenic signaling of Ras. The Raf serine/threonine kinases (A-Raf, B-Raf and C-Raf or Raf-1) are best known for their activation of downs tream Mek kinases. The isoforms are similar in their effectors, but have di fferent expression profiles. C-Raf is ubiquitously expressed, wher e A-Raf is generally found in the urogenital organs, and B-Raf is primarily neuronal (65). Retroviral Raf protein is a potent oncogene whose constitutive activation results in cellular transformation. The B-Raf
20 isoform is found mutated in many hum an tumors, including melanoma, colon, lung and thyroid tumors((105-107). The majo rity of mutations of B-Raf found in cancer are of the V600E variety(106). The most recent molecularly targeted therapies against mutated B-Raf have poor preclinical profiles or off target effects(108). However, a newly devel oped inhibitor, PLX4720, demonstrates selective inhibition of mutated B-Raf ki nase activity and promising antitumor efficacy for melanoma(109). C-Raf and BRaf both mediate signals downstream of Ras. Ras activation brings Raf to the membrane where it interacts with many signaling partners. This interaction resu lts in the phosphorylation of Raf. The activation or inhibition of C-Raf kinas e activity is dependent on the tyrosine, serine or threonine amino acid site wher e the phosphorylation events occur(110, 111). Known activators of C-Raf incl ude Src and PKC, which phosphorylate CRaf on S388, T491, S494 and S499(112). Deactivators include Akt and PKA, which phosphorylate C-Raf on S43, S621 and S259(112). Although many pathways feed into Raf signaling, current ly the only validated physiologically relevant downstream target s of C-Raf are the Mek1/2 MAP kinase proteins(111). C-Raf has been widely targeted for canc er therapy. Both small molecule inhibitors as well as anti-sense oligonuc leotides (oligo) have been developed to inhibit C-Raf signaling. A lipid-encaps ulated C-Raf oligo known as LErafAON finished phase I studies in 2006, and is currently undergoing further clinical development(113). Sorafenib, the most su ccessful small molecule inhibitor of CRaf kinase, was approved by the FDA for renal cell carcinoma. Upon further clinical and preclinical investigation it was f ound that Sorafenib not only inhibits
21 C-Raf and B-Raf kinase activities, but al so VEGF, PDGF, Flt3, c-Kit and FGFR signaling(114). Currently Sora fenib is undergoing further clinical investigation for single and combination therapy in hepatocellular carcinoma, non-small cell lung cancer, pancreatic cancer, breast c ancer, melanoma, and hematological malignancies(115). The success of Sorafe nib as a multi-kinase inhibitor has led to the approval of Sunitinib, an inhibi tor that targets Ra f along with VEGF-R, PDGF-R and c-Kit signaling. Sunitini b is currently approved for renal cell carcinoma and Imatinib-resistant GIST (gastr ointestinal stromal tumors), and is now considered the gold standard of care for these diseases. The success of these compounds in the clinic has led to debate as to whether molecularly targeted therapies for cancer need to be highly specific or broadly specific. Due to the multi-faceted broad spectrum of proteins inhibited by Sorafenib and Sunitinib, it is difficult to determine from this example if blocking Raf kinase activity selectively will be beneficial as a mono-therapy for cancer. However, the number of inhibitors present ly clinically available ha s validated targeting the Raf/Mek/Erk pathway for cancer, whether at the Raf or Mek level. The MAPK cascades continues downstr eam of Raf activation with Mek phosphorylation. Mek1 and Mek2 belong to a gene family comprised of 5 Mek proteins altogether. Mek1 and 2 are highly homologous, differing only in the proline-rich region among the kinase s ub-domains 9 and 10(116). The proline rich domain is necessary for Raf binding and activation, and this difference may confer substrate specificity for the bi nding affinity of the different Raf isoforms(117). Mek1 and 2 are also t he only known activators of Erk1/2,
22 representing a classic bottleneck in cell signaling. Severa l studies have shown that constitutive activation of the Mek pat hway is associated with different cancer types, including hepatocellular carcinoma, renal cell carcinoma, breast cancer, squamous cell carcinoma, AML (acute myelogenous leukemia) and CML (chronic myelogenous leukemia) (reviewed in (65). Therefore, Mek kinase activity has been a sought-after target for cancer therapy. The fi rst inhibitors of Mek activation were developed in the mid-1990s. PD098059 and U0126 both inhibit the activation of Mek, but did not fit t he pharmacological profile necessary for development. The lack of solubility and bioavailability prompted the search for more Mek inhibitors. The first Mek inhibitor to demonstrate antitumor efficacy in mouse models, and subsequently the first to move to clinical trials, was CI-1040. In mouse models CI-1040 in hibited human and mouse colo n tumors by as much as 80 percent(77). CI-1040, like PD98059 and U-0126, is noncompetitive with ATP. It binds an allosteric site on Mek, preventing its activation. Phase I and II studies, however, showed that CI-1040 la cked the sufficient antitumor activity, metabolic stability, and bioavailability for further development. Currently there are 2 promising Mek inhibitors undergoing phase II clinical trials. PD-0325901 and AZD6244 are both orally available, selective Mek inhibitors that are noncompetitive with ATP. Both compounds are well tole rated in the clinic, and have been shown to inhibit P-Erk1/2 levels in biopsies of human tumors as well as in peripheral blood mononucleocytes (PBMCs). Dosing with either of the Mek inhibitors alone have yet to show objecti ve response, however stable disease was reported in 12-28 percent of patients. Future ef forts for developing these
23 Mek inhibitors include combination tria ls to inhibit other known signaling pathways common to cancer progressi on, such as the PI3K pathway. The final step in the MAP kinase pathway is the activation of extra-cellular signal-regulated kinase 1 and 2 (Erk1/ 2), also known as p44 and p42 MAP kinases. There has been a positive correlation with overexpression or constitutive activation of Erk1/2 in both human tumo rs and tumor cell lines. Likewise, activation is also involved in many physi ologically relevant events such as cell motility, proliferation, differentiation, and survival(71, 118, 119). Erk1/2 have been shown to be rapidly phosphorylated downst ream of growth factor signaling. Activating phosphorylation events by Mek1 and Mek2 have been described on threonine 183 and tyrosine 185 (116, 120). These activating phosphorylation events allow Erk1/2 to translocate to the nucleus where they target the transcription factors Elk-1, c-Fos, and c-MYC, among others. These targets are thought to be responsible for proliferative effects of Raf/Mek/Erk signaling, as they are tied to increasing transcrip tion of cell cycle regulators and DNA synthesis machinery. Aside from translo cation to the nucleus, activated Erk1/2 proteins also activate cytoplasmic si gnaling modifiers such as microtubuleassociated proteins, ribosom al protein S6 kinase, SH P2, EGFR, SOS, C-Raf and Mek(121, 122). The turnover of phosphorylated Erk prot eins is mediated by regulatory protein phosphat ase that remove the activating phosphate groups from the proteins. Loss of these phosphatases has been described in neoplasia, and is currently targeted for therapeutic in tervention. The most characterized phosphatases for Erk1/2 is the mitogenactivated protein phosphatase-1 (MKP-
24 1). MKP-1 is a dual-specificity, serum-inducible phosphatase selective for Erk1/2 over Mek1(123). Loss of MKP-1 is correla ted with increasing grades of prostate, colon, and bladder cancer(124, 125). In normal cells, the regulation of t he MAP kinase pathway is complex. While it is well established that the activation of Er k kinases is regulated by receptor tyrosine kinases such as EGFR and PDGFR via activation of Ras (64, 71-73), more recently, adrenergic receptor (AR) stimulation of cAMP has also been shown to regulate the activation of Er k1/2 in normal cells (reviewed in (34, 36, 126, 127)). For example, in some no rmal cells such as cardiac myocytes and bone cells cAMP was shown to activate Erk1/2, whereas in others such as adipocytes and endothelial cells it was shown to inhibit Erk1/2 activation(33, 50, 59, 128, 129). The mechanism for the crosstalk between the heterotrimeric G proteins and MAP kinase pathway in norma l cells is still relatively unclear, with many signaling mechanisms pr oposed depending on cell ty pes. Activation of the 2 AR results in release of the activating subunit Gs, which stimulates the enzyme adenyl cyclase to convert ATP to cAMP. cAMP is a classic second messenger that signals to multiple substrates and controls many physiological events. cAMP activates both protei n kinase A (PKA) as well as guanine nucleotide exchange factor EPAC1 and 2 (Exchange Proteins Activated by cAMP), resulting in two different pathw ays in which communication with C-Raf occurs. In normal cells, PKA is known to phosphorylate C-Raf on multiple serine residues (S43, S259, S233) that inhibit the kinase activity or activation of Raf(112). Likewise, EPAC activation result s in GTP-binding and activation of
25 Rap1, which can compete with C-Raf for bi nding sites on Ras. Ultimately this competition results in decreased activati on of C-Raf and inhi bition of downstream MAP kinase signaling. Alter natively, in astrocytes, thyr oid cells, kidney cells and megakaryocytic cells it has been shown that cAMP-mediated activation of Rap-1 can result in activation of B-Raf thr ough Src signaling, leading to increased Erk1/2 phosphorylation and cell proliferation(34). In tumo r cells, however, little is known about adrenergic receptor regulation of Erk1/2 acti vation. In this Chapter of the thesis we demonstrat e that stimulation of the 2AR induces significant tumor growth suppression in human cancers where this stimulat ion results in the inactivation of the C-Raf/Mek1/2/Er k1/2 pathway by a cAMP-dependent activation of PKA, but not the Rap1 guanine nucleotide exchange factor EPAC.
26 Materials and Methods Cell Lines and Transfections All cell lines used were obtained from American Type Tissue Collection (ATCC), (Manassas, VA), or from the DCTD Tumor Repository, (Frederick, MD). Propagation was carried out according to ATCC protocols. Propagation media was obtained from Invitrogen Corporation (Carlsbad, CA). DNA transfections were carried out using TransIT-LT1 transfection reagent (Takara Mirus, Madison, WI ) according to the manufacture rs protocol. Briefly, 4x10 5 MDA-MB-231 cells were grown to approximately 70% conf luency in medium s upplemented with 10% FBS (Atlanta Biologicals, Atlanta, GA) in the absence of antibiotics. For transfection, LT1 was pre-incubated in 200 l per transfection in Opti-MEM (Invitrogen) for 20 minutes pr ior to addition of plasmid. 2 g of plasmid DNA (pFC-Mek Stratagene, La Jolla, CA) per transfection was incubated with the Opti-MEM/LT1 mixture for 20 minutes afte r which the transfection mixture was added to cells and incubated at 37 C for 48 hours. At that time, cells were then treated with either DMSO or ARA-211 (10M) for 48 hours and subsequently assayed for proliferation via Alamar Blue metabolism and harvested for western blot analysis as described below.
27 Cytoblot Screening for Small Molecu les that Decrease Phospho-Erk1/2 MDA-MB-231 cells were plated into sterile, white opaqu e, 96 well tissue culture plates at a density of 25,000 cells/well, cells were allowed to attach overnight and were treated for 1 hour in the presence of either vehicle control, 20 M PD-98059 (EMD Biosciences, San Diego, CA), or 10 M of NCI Diversity Set of 2000 compounds (1compound/well) (http ://dtp.nci.nih.gov/) as described by us (78). Cells were then washed, fixed for one hour at 4 o C cold 3.7% formaldehyde and permeabilized for 5-min with ice-cold me thanol. Cells were washed, and rocked overnight at 4 o C with anti-phospho-p44/42 MAPK (C ell Signaling Technology, Beverly, MA) and Horse Radish Pero xidase conjugated anti-Rabbit IgG secondary antibody (Jackson ImmunoResearch, West Grove, PA). The plates were washed and chemilumines cence reagent added to t he wells of the plates, then x-ray film directly plac ed on top of the plates. Quantification of the results was done using a GS-700 scanning dens itometer (Bio-Rad Laboratories, Hercules, CA). Western Blotting Treated cell samples were lysed in 30 mM HEPES, 10 mM NaCl, 5 mM MgCl 2 25 mM NaF, 1 mM EGTA, 1% Triton-X100, 10% Glycerol, 2 mM Sodium orthovanidate, 10 g/mL aprotinin, 10 g/m L soybean trypsin inhibitor, 25g/mL leupeptin, 2 mM phenylmethylsulfony lfluoride (PMSF), 6.4 mg/mL p-
28 nitrophenylphosphate for 30 minutes at 4 o C, and proteins r un on SDS-PAGE gels then transferred to nitrocellulose membranes. Membranes were blocked in either 5% milk in PBS, pH 7.4, contai ning 0.1% Tween-20 ( PBS-T) or 1% BSA in TBS, pH 7.5, containing 0.1% Tw een-20 (TBS-T). Phospho-STAT3 and phospho--AKT antibodies were diluted in 1% BSA in TBS-T while phospho-CRaf, phospho-Mek, and phospho-p44/p42 MAPK antib odies (Cell Signaling) were diluted in 5% milk in PBS-T for either 1 hour at room temperat ure or overnight at 4 o C. HRP-conjugated secondary antibod ies (Jackson ImmunoResearch, West Grove, PA) were diluted in 5% milk in PBS-T or TB S-T at a 1:1000 dilution for one hour at room tem perature. Western blots results were visualized using enhanced chemiluminescence. Stripping of the membranes for reblotting was done using stripping buffer (62.5 mM Tris pH 7.6, 2% SDS, 0.7% 2mercaptoethanol) at 50 0 C for 30 minutes. Cells treated with ARA-211, 8CPT-2 O-Me-cAMP, H89 and Forskolin to det ermine effects on P-Erk 1/2 and other signaling events by Western blot were treated for one hour at varying doses. Pre-treatment with H89 was done for 15 mi nutes followed by 45 minutes of ARA211 treatment. In Vitro Kinase Assays MDA-MB-231 cells (2x10 6 ) were lysed for 30 minutes at 4 o C in RIPA-150 buffer (10 mM Tris pH 7.5, 150 mM NaCl, 10% glycerol, 5 mM EDTA, 1% Triton X-100, .1% SDS, 100 M sodium or thovanadate, 1 M aprotinin 10 M leupeptin, 10
29 M antipain). Cell debris was pelle ted at 13000 rpm for 15 minutes at 4 o C and the lysate was rocked overnight with 10 g of either C-Raf or Mek antibodies (Santa Cruz Biotechnologie s, Santa Cruz, CA). 25 l of protein A/G PLUS agarose (Santa Cruz Biotechnology) was then added and rocked for 4 hours at 4 o C. Samples were spun at 2000 rpm fo r 5 minutes to collect the agarose beads, which were then washed four times with lysis buffer. On the final wash each pellet was divided into 5. Pellets were re-suspended in 30 l kinase buffer (30 mM HEPES, pH 7.5, 7 mM MnCl 2 5 mM MgCl 2 1 mM dithiothreitol and 15 M adenosine triphosphate (ATP) with eit her 10 M U-0126 (EMD Biosciences) ARA-211, Raf Kinase Inhibitor (EMD Biosci ences), or DMSO vehicle control. Finally, 0.5 g Mek1 or ERK2 peptide substrate (Upstate Biotechnologies, Lake Placid, NY) was added, followed by 20 Ci [ 32 P]-ATP per sample. Samples were then incubated at room temperature for 30 minutes and the reaction terminated by the addition of 30l 4x SDS-PAGE sample buffer (33% glycerol, 0.3 M DTT, 6.7% SDS, 15% -mercaptoethanol, 0.1% bromphenol blue), which was then boiled at 100 o C for 5 minutes. Samples were run on a 10% polyacrylamide gel to separate proteins ; gels were dried and phosphorylation results visualized via autoradiography.
30 In Vivo Kinase Assay Intact MDA-MB-231 cells (2x10 6 ) were treated either with vehicle control (DMSO), 10 M U-0126, 10 Raf Kinase Inhibitor, or 10 ARA-211. Samples were then processed for kinase assays as described above. MTT Assay 2000 cells per well were plated into 96 we ll tissue culture plates, and the cells were allowed to attach overnight at 37 o C with 10% CO 2 Cells were then treated with increasing concentrations of ARA-21 1, Forskolin (Sigma -Aldrich, St. Louis, MO) or vehicle (DMSO) and incubated for f our days. At that time treatment medium was removed and replaced with 100 l of 1mg/ml MTT (Sigma-Aldrich) in media and incubated at 37 o C, 10% CO 2 for 3 hours. After 3 hours the MTT media was removed and replaced with 100 l/well of DMSO, rocked at room temperature for 5 minutes, then absorba nce read at 540 nm on a micro-plate reader. Alamar Blue Proliferation Assay For proliferation assays using exogenous constitutively activated Mek or the 2AR antagonist ICI-118.551, Al amar Blue metabolism assay used to determine cell viability according to the manufacturers protocol (Biosource, Camarillo, CA). Briefly, cells were trea ted and grown as described above; Alamar Blue was
31 added to the media at a 1:10 dilution and incubated for 3 hours. Media/Alamar Blue was removed from the cells and aliq uoted into 96 well plates (100l per well) and fluorescence read on a Wallac Victor 2 plate reader at excitation 540nM, and emission 615nM. Soft Agar Assay Anchorage independent growth assays we re performed as previously described (85). Briefly, agar was di luted to 0.3% in propagatio n media contai ning 20,000 cells and either vehicle or ARA-211 and 1ml of cell/agar mixture plated on top of 2ml of a 0.6% layer of agar in medium alon e. Cells were grown for three to four weeks at 37 C, 10% Co2 and were fed w eekly with 100l per well of media/treatment. Quantification of co lonies was done using a GS-700 scanning densitometer. Apoptosis Assay Apoptosis was analyzed using the Ce ll Death Detection ELISA (Roche, Indianapolis, IN) measuring cytoplas mic histone-associated monoand oligonucleosome DNA fragments. Procedur es were carried out according to the manufacturers protocol. Briefly, 1x10 5 cells were plated in 60mm dishes and treated with various doses of ARA-211 or v ehicle for 48 hours. Cells were lysed
32 with the provided lysis buffer and lysate was incubated on the ELISA plate for 1 hour to capture the histone associated DNA indicative of apoptotic cells. AntiDNA peroxidase conjugated antibody was a dded to detect the cleaved DNA and the microplate was washed and develo ped with ABTS substrate which was quantitated by micropl ate reader at 405nm. Total cellular cAMP Measurement The measurement of total cellular cA MP was done using a cAMP Biotrak Enzyme immunoassay (EIA) system (Amersham Biosciences, Piscataway, NJ ) according to the manufacturers protocol Briefly, 2000 cells were plated into each well of a 96 well plate. The next day cells were treated with increasing concentrations of ARA-211 or with vehicle for one hour. Alternatively, MDA-MB231 cells were pretreat ed with antagonists to the 1 (Prazosin, Sigma), 2 (Yohimbine, Sigma), 1 (Metoprolol, Sigma), and 2 (ICI 118.551, Sigma) adrenergic receptors at 100nM for 15 minut es followed by treatment with 10M ARA-211 for 45 minutes. Cells were then lysed directly in the wells using the supplied lysis buffer. The lysate was tr ansferred to a 96-well cAMP ELISA plate, and results were quantified using a Wallac Victor 2 plate reader.
33 PKA Kinase Assay PKA kinase assays were performed with the SignaTECT cAMP-Dependent Protein Kinase (PKA) Assay System (P romega, Madison WI) according to the manufacturers protocol. Briefly, cells were plated (5X10 6 ) into 100 mm dishes and treated the next day with ARA-211, H-89 (EMD Biosciences), or vehicle for one hour. When pre-treating cells, H89 was dosed for 15 minutes following ARA211 treatment for 45 minutes. Cells were lysed using a Dounce homogenizor and lysates were incubated in the s upplied kinase reaction buffer, with 32 P-ATP and biotin labeled PKA peptide substrate for 5 min at 37 o C. Once the reactions were quenched they were spotted onto streptavidin membranes and 32 P-ATP transfer was measured via scintillation counter. Antitumor Activity in Nude Mouse Xenograft Models Female athymic nude (nu/nu) mice, 5-6 we eks old, were purchased from Charles River (Wilmington, MA) and allowed to acc limate in the animal facility for one week. After harvesting, A-549, SNB-19, ACHN, and HCT-116 cells were, resuspended in sterile PBS, and injected s. c. into the right and left flanks (10x10 6 cells per flank) of the mice as previ ously described (130). For MDA-MB-231, cells were harvested and resuspended in PBS (10x10 6 cells per 50 l) and an equal volume of Matrigel (BD Biosciences San Jose, CA). 100 l of MDA-MB231 cells in PBS/Matrigel we re injected s.c. into the mammary fat pads between
34 the 2 nd and 3 rd nipples on each side of the mi ce. Once the tumors reached approximately 200-300 mm 3 the mice were randomized and treated i.p. with 0.1ml vehicle (PBS) or ARA-211. Each treatment group consisted of five mice (2 tumors per mouse, a total of 10 tumors). The tumor volumes were determined by measuring the length ( l ) and width ( w ) and calculating the volume (V= w 2 l /2). Statistical significance between the cont rol and treated animals was determined using a students t test. Immunohistochemisty (IHC) and Slide Quantitation Upon completion of the xenograft study, mice were euthanized and the tumors were removed and fixed in 10% buffered fo rmalin for at least 24 hours. Three tumors per treatment group were then prepared and st ained for proliferation marker Ki-67, apoptosis by TUNEL, and phosphor-Erk1/2 as previously described(131). The IHC slides were analyzed at the Anal ytical Microscopy Core facility of the H. Lee Moffitt Cancer Center & Research Institute. The Ariol SL-50 automated image analysis system (Applied Imaging, Santa Clara, CA) was used for quantitation of each slide as prev iously described(132). Briefly, highresolution digital pictures were taken of the entire stained area of each slice of tumor tissue for automatic staining quantitat ion. Two thresholds were used to signify positive staining, one rec ognizing background and one recognizing positive brown staining. The percentage of positive st aining was calculated by dividing the positive staining by the total staining (positive + negative) divided by
35 the total area stained, and the intensit y of the stain was determined by calculating the integrated opt ical density of the positiv e stain with the negative background stain subtracted out. Each slide analyzed consisted of 54,000 to 160,000 cells analyzed depending on the size of the tumor when resected. Significance of the quantit ation was analyzed by a twotailed students t-test.
36 Results Identification of ARA-211, a selective suppressor of P-Erk1/2 levels in MDAMB-231 breast cancer cells. Many common genetic alterations in human cancers result in the hyperactivation of the Mek1/2 kinases le ading to constitutively phosphorylated Erk1/2(64, 71-73). Becaus e aberrant activation of the Mek1/2/Erk1/2 pathway has been implicated in oncogenesis and tumor survival, we sought to identify molecules that thwart this pathway. To this end, we first used a chemical biology approach by screening a 2,000 small molecu le compound library (NCI diversity set, http://dtp.nci.nih.gov/) from the Nati onal Cancer Institute for pharmacological agents capable of suppressing the high levels of P-Er k1/2 in the human breast cancer cell line MDA-MB-231. Treatment of MDA-MB-231 cells with compounds from the NCI diversity set (96 well plate assay (25 pl ates total), 1 compound per well see Figure 4) and s ubsequently processing the plates for a cytoblot using an antibody specific for PErk1/2, resu lted in several com pounds that suppress PErk1/2 levels. PD-98059, a Mek kina se inhibitor was used as a positive control for inhibition of Erk1/2 phosphoryl ation. MDA-MB-231 cells were treated with the hits from the assay to confi rm the anti-P-Erk1/2 activity. Three
37 compounds from the screen were identified as inhibitors of Erk activation (Table 1). The IC50 for each compound was obtained from dose response treatment on 231 cells, resulting in IC50 values of 8, 10, and 500 nM for inhibition of Erk1/2 phosphorylation (Table 2). ARA-211 also suppressed P-Mek1/2 levels but not PC-Raf, P-JNK, P-p38, P-Akt or P-STAT3 (F igure 5). These results demonstrated that ARA-211 is selective for disruptin g the Mek1/2/Erk1/2, but not other oncogenic and tumor survival pathways.
PD-98059 UnconfirmedHit Vehicle ARA-211 PD-98059 UnconfirmedHit Vehicle ARA-211Figure 4. High throughput screening of the NCI Diversity Set bycytoblotanalysis reveals three inhibitors of Erk1/2 activationFigure 4. High throughput screening identifies ARA-211. MDA-MB-231 cells were plated in twenty-five 96-well plates, treated with vehicle (8 wells), PD-98059 (2 wells) or 2000 compounds (1 compound/well) from the NCIdiversity set, and the plates processed for immunocytoblottingwith an anti-phospho-Erk1/2 antibody as described under Materials and Methods. Plate#23 is shown with wells treated with vehicle as well as those treated with the Mekinhibitor PO98059. 38
Table 1. Chemical structures of the compounds identified from the NCI Diversity Set as inhibitors of P-Erk1/2 levelsTable 1. Cytoblotscreen reveals three structurally similar chemical compounds that lower anti-P-Erk1/2 levels.OHNHOHOH NOHNHOHOH HH O O NH S NSC #: 289336NSC #: 355078NSC #: 39215 Plate #: ERKi-2067ARA-67Plate #: ERKi-23211ARA-211Plate #: ERKi-624ARA-624OHNHOHOH OHNHOHOH NOHNHOHOH NOHNHOHOH HH O O NH S HH O O NH S O O NH S NSC #: 289336NSC #: 355078NSC #: 39215 Plate #: ERKi-2067ARA-67Plate #: ERKi-23211ARA-211Plate #: ERKi-624ARA-624 39
ARA-211 (M)0 10P-JNKP-p38P-AKTP-STAT 3 P-Erk1/2P-C-RafP-Mek-1 ARA-211 (M)0 10P-JNKP-p38P-AKTP-STAT 3 P-Erk1/2P-C-RafP-Mek-1 C U .001 .01 .1 1 P-Erk1/2Erk1/2 P-Erk1/2Erk1/2P-Erk1/2Erk1/2C .1 .3 1 3 10 30 C .1 .3 1 3 10 30 MARA-211 (M)ARA-67 (M)ARA-624 (M) C U .001 .01 .1 1 P-Erk1/2Erk1/2 P-Erk1/2Erk1/2P-Erk1/2Erk1/2C .1 .3 1 3 10 30 C .1 .3 1 3 10 30 MARA-211 (M)ARA-67 (M)ARA-624 (M)Figure 5. ARA-family of compounds selectively inhibit P-Erk1/2 levels in MDA-MB-231 breast cancer cellsFigure 5. Validation of ARA-211 as selective suppressor of P-Erk1/2 levels in MDA-MB-231 breast cancer cells.ARA-211 treatment of MDA-MB-231 cells and Western blotting were performed as described under Materialsand Methods, and resulted in suppression of P-Erk1/2 and P-Mek1/2 but not C-Raf, P-JNK, P-p38, P-Aktand P-STAT3 levels. 40
41 The fact that ARA-211 suppressed the levels of P-Mek1/2 and PErk1/2 suggested that it might be a C-Raf kinase inhibitor. In-vitro incubation of C-Raf and Mek1, immunoprecipitated from MDA-MB-231 cells, with ARA-211 did not inhibit C-Raf and Mek1 kinase activities (Figure 6). As expected, the C-Raf kinase inhibitor (Rki, Calbiochem) and the Mek1 inhibitor (U-0126, EMD Biosciences, San Diego, CA) blocked C-Raf and Mek1 kinase activities in vitro, respectively (Figure 6). In contrast to the in vitro studies, C-Raf and Mek1 kinase activities were blocked by ARA-211 when intact MDA-MB-231 cells were first treated with ARA-211, prior to cell lysis and immunoprec ipitation of C-Raf and Mek1 (Figure 6). As expected only Rki but not U-0126 inhibited C-Raf kinase whereas both Rki and U-0126 inhibited Mek1 kinase when whole cells were treated. Contrary to the effects on C-Ra f kinase activity, B-Raf kinase activity was not affected upon stimulat ion of cells with ARA-211.
P-Mek-1C RKi 211P-Erk-2C U 211 P-Mek-1C Rki U 211 IP Raf-1 P-Erk-2C Rki U 211 IP Mek-1 IP Raf-1 IP Mek-1 IP B-Raf P-Mek-1C 211 Iso U Rki Figure 6. Raf-1, but not B-Raf, kinase activity is inhibited upon ARA-211 treatment of whole cells, but not in vitro. A. Raf-1 and Mek-1 were immunoprecipitatedfrom MDA-MB-231 cells and treated in vitro with ARA-211, Rkior U-0126 and kinase assays followed by SDS-PAGE were performed as described in Materials and Methods. B. and C. Intact cells were first treated with ARA-211, then C-Raf and Mek-1 (B.) and B-Raf (C.) were immunoprecipitated and processed as for A. Only in cells treated with ARA-211 is Raf-1 kinase inhibited. In vitro only the Raf kinase inhibitor (RKi) and Mek kinase inhibitor (U-0126) blocked MAPK signaling. Figure 6. C-Raf, but not B-Raf, kinase activity is inhibited upon ARA-211 treatment of whole cells, but not in-vitroA.B.C. 42
43 ARA-211 is a selective 2 adrenergic receptor agonist that inhibits P-Erk 1/2 levels by increasing cAMP The fact that ARA-211 inhibits CRaf and Mek1 kinase activities when intact cells are treated but not in vitro s uggested that the target for this compound is upstream of C-Raf kinase. A chemical similarity search i dentified ARA-211 as pirbuterol, and resulted in many anal ogues of the beta adrenergic receptor agonist family. Table 2 depicts structural analogues of the ARA compounds that were located using SciFi-Scholar, a pr ogram linked to the American Chemistry Society registry of chemical compounds. The commercially available compounds were ordered and first screened at 10M for inhibition of activation of Erk1/2 in MDA-MB-231 cells. Epinephrine, Isoproterenol, and Adrenalone showed inhibition of ERK 1/2 activation; while 2,6-Lutidine,3-diol, 1-(6-(1-Hydroxyethyl)-Pyridin-2-Yl)-Ethanol, and 3-Hydr oxy-2,6-Pyridinedimethanol-HCl did not inhibit the activation of ERK 1/2 in MDA-MB-231 cells. The active compounds were then tested in MDA-MB-231 cells in a dose response manner to better characterize each compound. IC50 va lues of epinephrine and adrenalone were found to be 5 and 10nM respectively, whil e Isoproterenol demonstrated an IC50 of .01 nM (Table 2). Compar ing ARA-211 with 2,6-Lutidine,3-diol & 3-Hydroxy2,6 Pyridinedimethanol-HCl it can be det ermined that the tertiary butyl amine group is critical to ARA-211 activity. Ho wever, the size of this group appears to
44 be important as well, with the isopropyl (isoproter enol) being better than the methyl (epinephrine). It can also be determined, by comparing ARA 211 to epinephrine and isoproterenol that the cyclic nitrogen is not significant. Finally, the aliphatic hydroxyl is critical as its r eplacement with S-methyl results in loss of the ability to lower P-Erk1/2 levels (com pare ARA-624 to epinephrine). The main value from the SAR howeve r, was the crucial information that led us to investigate stim ulation of the 2 AR as an event leading to the inhibition of PErk1/2 in MDA-MB-231 cells. Based on t he chemical structure search we found that the hit that was identified as ARA-211 is pirbuterol, a selective 2 AR agonist.
P-Erk1/2 IC50(nM)StructureName>10,0001-(6-(1-Hydroxy-ethyl)-Pyridin-2-Yl)-Ethanol (HEPE)>10,0003-Hydroxy-2,6-Pyridinedimethanol-HCl (HPDM)>10,0002,6-Lutidine-a,3-diol0.01 .018Isoproterenol5 15Adrenalone10 7Epinephrine500 25ARA-62410 5ARA-678 5ARA-211(Pirbuterol) NOHNHOHOH OHNHOHOH NHOHOH NHOHOH OH OH NH S OH OH NH S OH OH OH NH O OH OH NH OH OH OH H NH N OH OH N OH OH OH N OH OH Table 2. Structure-activity relationship studies identify the ARA family of inhibitors as analogues of epinephrine, and identifies ARA-211 as pirbuterol 45
46 To demonstrate that ARA-211 acts as a adrenergic receptor agonist in MDA-MB-231 cells, and that this results in blockade of the Raf/Mek/Erk pathway, we treated these cells with ARA-211 and first showed that this compound induced the formation of cAMP (Figure 7) We then determined that ARA-211 is highly selective for 2 adrenergic receptors by determining that only 2 but not 1, 1 or 2 receptor antagonists block the abi lity of ARA-211 to induce cAMP formation (Figure 7). Furthermore, we used isopreterenol, another selective adrenergic receptor agonist and showed that it also stimulates the formation of cAMP and decreases the levels of P-Erk 1/2 (data not shown). Finally, to validate the effects of cAMP signaling on P-Erk1/2 activation MDA-MB-231 cells were pretreated with selective adrener gic antagonists followed by ARA-211 treatment. ARA-211 mediat ed inhibition of P-Erk1/2 levels was abrogated only when 2 adrenergic signaling was blocked (Fi gure 8). Treatment of MDA-MB-231 cells with a 1 selective, dobutamine, and a 3 selective agonist CI316243, did not inhibit P-Erk 1/2 levels (data not shown). Furthermore, to demonstrate that the effects of P-Erk1/2 inhibition wa s due to cAMP mediated signaling we stimulated adenyl cyclase to produce cAMP directly by forskolin treatment. Forskolin inhibited P-Erk1/2 levels with an IC50 of 10 nM, similar to the ARA-211 induced inhibition (Figure 8). Taken t ogether these data demon strate that ARA211 is a 2-selective adrenergic receptor agonist that suppresses P-Erk1/2 by increasing cAMP levels.
0100200300400500600cAMP (fMol/Well)Control2111 + 2112 + 2111 + 2112 + 211 0100200300400500600cAMP (fMol/Well)Control2111 + 2112 + 2111 + 2112 + 211Figure 7. ARA-211 stimulates cAMP selectively through 2 adrenergic receptor activationFigure 7. ARA-211 is a selective 2 adrenergic receptor agonist. MDA-MB-231 cells were treated with ARA-211 alone or in the presence of 1, 2, 1 or 2 adrenergic receptor antagonists and the levels of cAMP were determined as described under Materials and Methods. Only pre-treatment with a 2 selective adrenergic receptor antagonist blocked the ability of ARA-211 to stimulate cAMP in MDA-MB-231 cells. 47
ARA-211 (10M) -+ ----+ + + + 2 antagonist (nM) 0 0 1 10 50 100 1 10 50 100 P-Erk1/2 Erk1/2 ARA-211 (10M) -+ ----+ + + + 2 antagonist (nM) 0 0 1 10 50 100 1 10 50 100 P-Erk1/2 Erk1/2 P-Erk1/2Erk1/2Forskolin(nM) 0 0 1 10 50 100 ARA-211 (10M) -+ ---Figure 8. ARA-211-mediated cAMP stimulation through 2 adrenergic receptor activation results in inhibition of P-Erk1/2Figure 8. ARA-211 is a selective 2 adrenergic receptor agonist that inhibits P-Erk1/2 levels by increasing cAMP.A.MDA-MB-231 cells were treated with ARA-211 alone or in the presence of 1, 2, 1 or 2 adrenergic receptor antagonists and the levels of P-Erk1/2 were determined as described under Materials and Methods. B. Similarly, direct activation of cAMP by forskolinalso results in inhibition of P-Erk1/2 activation.B.A. 48
49 ARA-211 inhibits P-Erk 1/2 levels by PKA but not EPAC activation The data from Figures 7 and 8 demons trated that ARA211 stimulates the production of cAMP, which in turn suppre sses the C-Raf/Mek1/2/ Erk1/2 pathway. Previous studies primarily in normal cells have shown that the Ras-Erk pathway can be regulated by the ability of cAMP and PKA to modulate growth factor signaling via C-Raf and B-Raf. PKA, which is activated by cAMP, has been shown in some normal cells to interf ere with Ras-C-Raf binding, downregulating MAPK-Erk signaling. This occurs th rough the PKA-dependant activation of the small G-protein Rap-1. Rap-1 competes with C-Raf for binding to GTP-Ras, resulting in a smaller amount of activated C-Raf available to activate Mek1/2. Alternatively, Rap1 can be activated by cAMP independently of PKA. Exchange Proteins Activated by cAMP (EPACs) are guanosine nucleotide exchange factors (GEFs) that are activated by cAMP and modulate the GTP loading of Rap1. Other studies have shown that if a cell line expresses high levels of B-Raf that PKA can activate B-Raf in a Ras-independent fashion. Active B-Raf is then free to phosphorylate and activate Mek-2, whic h activates Erk1/2. Therefore, the ARA-211 cAMP-dependent decreas e in P-Erk1/2 levels could be modulated by cAMP-activation of either PKA or EPAC. In addition, PKA could either inactivate C-Raf directly, or through activation of Rap-1. Based on thes e possibilities, we next examined whether cAMP suppresses the P-Erk1/2 levels by activating the protein kinase A (PKA) or by activati ng the Rap1 guanine exchange factor EPAC. To this end, we treated MDA-MB-231 cells either with the EPA C activator 8-CPT-
50 2 -O-Me-cAMP, which would mimic the effects of ARA-211 on P-Erk1/2 levels, or the PKA kinase inhibitor H-89, which would block the e ffects of ARA-211. Figure 9 demonstrates that treatment with the EPAC activator did not affect P-Erk1/2 levels. In contrast, treatment with the PKA inhibitor prevented ARA-211 from suppressing the levels of P-Erk1/2. T hese results suggest that ARA-211-induced decrease in p-Erk1/2 levels is mediated through a cAMP-dependent PKA activation, but not a cAMP-activation of EPAC. If this is the mechanism involved then ARA-211 would be expected to acti vate PKA, and PKA siRNA would be expected to inhibit ARA-211 from decreasi ng the P-Erk1/2 levels. Therefore, to further strengthen this suggestion, we first determined if ARA-211 is able to activate PKA. Figure 10 shows that ARA-211 treatment of MDA-MB-231 cells activated PKA kinase activity, and that H-89 prevented ARA-211 from activating PKA. siRNA to the alpha catalytic s ubunit of PKA was then used to further validate the necessity of PKA activation for ARA-211 mediated inhibition of Erk1/2 phosphorylation. Fi gure 10 demonstrates that as PKA protein levels are diminished the effects of AR A-211 on P-Erk1/2 are rescued. A dose response of PKA -cat siRNA of 5, 50 and 100 nM we re used to demonstrate a linear signaling correlation between PKA catalytic subunit protei n levels and P-Erk1/2 phosphorylation.
P-Erk1/2 Erk1/2 ARA-211 (10M) -+ --------8-CPT-2?-O-Me-cAMP(M) 0 0 .001 .01 .03 .1 .3 1 10 30 P-Erk1/2 Erk1/2 ARA-211 (10M) -+ --------8-CPT-2?-O-Me-cAMP(M) 0 0 .001 .01 .03 .1 .3 1 10 30 P-Erk1/2 Erk1/2 ARA-211(10M) -+ -+ + + + + +H-89(M) 0 0 20 .005 .05 .25 .5 5 20 P-Erk1/2 Erk1/2 ARA-211(10M) -+ -+ + + + + +H-89(M) 0 0 20 .005 .05 .25 .5 5 20Figure 9. ARA-211-mediated cAMP stimulation (through 2 adrenergic receptor activation) results in inhibition of P-Erk1/2 through PKA activation, not EPAC activation of Rap1Figure 9. ARA-211 inhibits P-Erk1/2 levels by PKA but not EPAC activation. MDA-MB-231 cells were treated either with ARA-211, the EPAC activator 8CPT-2-O-Me-cAMP (A.), the PKA inhbitorH-89 (B.), or ARA-211 and H-89 (B.). Direct activation of EPAC did not inhibit P-Erk1/2 levels, while blockade of PKA kinase activity by H89 abrogated ARA-211-mediated inhibition of P-Erk1/2A.B. 51
05001000150020002500PKA Activity (pMol/min/g ) Control2112 + 211H-89+211 05001000150020002500PKA Activity (pMol/min/g ) Control2112 + 211H-89+211 ARA-211 (10M) -+ -+ -+ -+ -+ -+ PKA siRNA(nM) ------5 5 50 50 100 100Mock --+ + --------Neg. siRNA(100nM) ----+ + ------P-Erk1/2PKA Cat. ARA-211 (10M) -+ -+ -+ -+ -+ -+ PKA siRNA(nM) ------5 5 50 50 100 100Mock --+ + --------Neg. siRNA(100nM) ----+ + ------P-Erk1/2PKA Cat. Figure 10. ARA-211 inhibits P-Erk1/2 levels by PKA but not EPAC activation. A.PKA kinase assays show ARA-211 stimulates PKA, which is blocked by pre-treatment with a 2 antagonist as well as PKA kinase inhibitor H89. B. siRNAto PKA demonstrates that as PKA expression decreases the ability of ARA-211 to inhibit P-Erk1/2 levels also diminishes.Figure 10. siRNAand kinase assays validate that ARA-211-mediated cAMP stimulation results in inhibition of P-Erk1/2 through PKA activationA.B. 52
53 ARA-211 inhibits tumor cell proliferation by 2AR-stimulation of cAMP and inhibition of Mek1/2 The data from the previous figures demonstrated that st imulation of the 2 AR with ARA-211 suppresses the C-Raf/Mek1/2/Erk1/2 pathway by a cAMPdependent activation of PKA but not EPAC. Whether stimulation of the 2 AR and subsequent inhibition of Raf/Mek/Erk causes tumor suppression is not known. Therefore, we next det ermined whether activation of the 2 adrenergic receptor, which leads to inhibition of t he C-Raf/Mek1/2/ Erk1/2 pathway, results in inhibition of tumor ce ll growth and survival of hu man cancer cells. Figure 11 demonstrates that ARA-211 inhi bits the proliferation of MDA-MB-231 cells in a dose dependent manner, and requires the abi lity of ARA-211 to bind the 2 adrenergic receptor. MTT and Alamar Blue assays were used to determine the effects of ARA-211 on cell pr oliferation. The prolifer ation of cells that were pretreated with 2 AR antagonist was not affected by ARA-211, in contrast to cell treated with ARA-211 al one. Western blotting from cells from the Alamar Blue proliferation assay de monstrated that pr e-treatment with the 2 AR antagonist rescued from the ARA-211 induced inhibiti on of Erk1/2 phosphorylation (Figure 11). Likewise, forskolin was used to directly stimulate adenyl cyclase to determine if 2 AR stimulation resulting in cAMP production is a necessary step in the inhibition of cell pro liferation. Figure 12 shows th at forskolin indeed inhibits MDA-MB-231 cell proliferation wit h an IC50 of approximately 50 nM.
54 0204060801001201400110100ARA-211 (M)Proliferation (% Control)ARA-211 (10M) -+ -+ 2 antagonist (nM) --50 50 P-Erk1/2Erk1/2 ARA-211 (10M) -+ -+ 2 antagonist (nM) --50 50 P-Erk1/2Erk1/2 Figure 11. ARA-211 inhibition of MDA-MB-231 cell proliferation requires AR activation and inhibition of P-Erk1/2Figure 11. ARA-211 inhibition of MDA-MB-231 cell proliferation requires 2 AR activation and inhibition of P-Erk1/2. MDA-MB-231 cells were treated with ARA-211 alone (square) or in the presence of the 2AR antagonist ICI 118.551 (triangle) and the effects of treatment on tumor cell growth (A.) and P-Erk1/2 levels (B.) were determined as described under Materials and Methods. A.B.
0204060801001200.11101001000Forskolin (nM)Proliferation (% Control)Figure 12. Direct activation of adenylyl cyclase by forskolininhibits the proliferation of MDA-MB-231 cellsFigure 12. Forskolininhibits MDA-MB-231 cell proliferation correlating with inhibition of P-Erk1/2. MDA-MB-231 cells were treated with vehicle (square) or Forskolin(triangle) and the effects on tumor cell growth determined. 55
56 The above data clearly demonstrated that th e ability of ARA-211 to disrupt the CRaf/Mek1/2/Erk1/2 pathway and to inhibi t tumor cell growth and survival depends on its ability to stimulate the 2 adrenergic receptor and produce cAMP. However, the ability of ARA-211 to inhibi t tumor cell growth could be due to its ability to affect cellular events other t han those leading to decreased P-Erk1/2. Therefore, we next determi ned whether the inhibition of tumor cell growth by ARA-211 requires its ability to inhibit the C-Raf/Mek1/2/Erk1/2 pathway. We reasoned that if suppression of C-Raf and Mek1/2 were critical to ARA-211 inhibition of tumor growth then ectopically expressing constituti vely active Mek1/2 would rescue human cancer cells from t he effects of ARA-211. Figure 13 shows that MDA-MB-231 cells that ectopically ex press constitutively active Mek1/2 (CAMek) are resistant to ARA-211 inhibition of PErk1/2 and inhibition of tumor cell growth. Mek1/2 levels clearly show a dramatic induction of expression with transfection of CA-Mek plasmid. Likewise, downstream signaling through Erk1/2 activation is also increased. Finally, P-Er k1/2 inhibition mediated by ARA-211 is rescued in the cells over expressing CA-Me k. This demonstrates that ARA-211 inhibits proliferation of MD A-MB-231 cells through inhibition of Mek/Erk signaling.
0204060801001200.00.11.010.0100.0ARA-211 (M)Proliferation (% Control)CA-Mek (g) 0 0 2 2 2 ARA-211 (M) 0 10 0 10 30P-Erk1/2 Mek-1 Erk1/2CA-Mek (g) 0 0 2 2 2 ARA-211 (M) 0 10 0 10 30P-Erk1/2 Mek-1 Erk1/2Figure 13. ARA-211-mediated inhibition of MDA-MB-231 cell proliferation is dependent on the ability to inhibit Mek1/2 and to decrease P-Erk1/2 levelsFigure 13. Inhibition of proliferation by ARA-211 is dependent on Mek1/2/Erk1/2 signaling. MDA-MB-231 cells were treated with ARA-211 alone (open square) or in the presence of CA-Mek(closed square) and the effects of treatment on tumor cell growth (A.) and P-Erk1/2 levels (B.) were determined as described under Materials and Methods.A.B. 57
58 Screening of NCI 60 cell line panel leads to discovery of ARA-211 sensitive and insensitive cell lines So far we have demonstrat ed that ARA-211 binds the 2 AR, stimulates cAMP, and inhibits Erk1/2 phosphoryl ation through a PKA dependent mechanism leading to inhibition of cell proliferation in MDA-MB-231 cells. While these findings have not been report ed previously and ther efore are novel, it is critical that we determine whether this mechanism of tumor suppression is relevant to other human tumors, or if it is unique to the MDA-MB-231 breast cancer cell line. To this end we searched for cancer cell lines that possess the correct signaling circuitry to respond to 2 AR stimulation similarly to the MDA-MB-231 cell line. We hypothesized that ce lls that express the 2 AR and high levels of P-Erk1/2 would be candidates for ARA-211 sensitiv ity based on the necessity of ARA-211 to bind the 2 AR and inhibit P-Erk1/2 to blo ck cell proliferation. Likewise, literature reports suggest that in some no rmal cells the ratio of C-Raf to B-Raf expression levels may dictate whether 2 AR activation inhibits or activates Erk1/2. To reach our goal we determi ned by western blotting the levels of 2 AR, P-Erk1/2, C-Raf and B-Raf in the lysates fr om the NCI 60 cell line panel. Table 3 summarizes the findings from the wester n blot analysis shown in Figure 14. A minus (-) sign indicates little to no expre ssion, a plus (+) sign indicates moderate expression, whereas a double plus (++) sign indicate s high protein expression (Table 3). Based on the expression analysi s, the cell lines t hat were available
59 and expressed both the 2 AR and P-Erk1/2 were tested for ARA-211 sensitivity. Cells insensitive to treatment based on t he ability of ARA-211 to inhibit P-Erk1/2 were denoted by a *, and the cell lines tested that were sensitive to ARA-211 treatment were denoted by a **. Out of 20 lines tested 3, MDA-MB-231, ACHN, and SF-539, were found to be sensitive to ARA-211 treatment. Analysis of the findings from the cell lines tested with ARA-211 compared to C-Raf and B-Raf expression led to the finding that the rati o hypothesis did not hold true. No cell lines tested demonstrat ed an ARA-211 mediated increase in P-Erk1/2 expression, even when expression of BRaf was high and C-Raf was low. Similarly, it was found that even though some cells express 2 AR, P-Erk1/2, and C-Raf but not B-Raf, the cells were still insensitive to ARA-211 treatment (IGROV1, MCF-7, Table 3, Figure 14). It is clear however, that all cell lines that responded to ARA-211 express 2 AR and have high levels of P-Erk1/2. Therefore, the sensitive cell lines MDA-MB-231, SF -539, and ACHN, along with insensitive cell lines A549, HCT-116, and SNB-19 were chosen to provide a panel of cell lines to demonstrate that only cells where ARA-211 stimulates cAMP production (which results in decreased levels of P-Erk1/2) are sensitive to ARA-211-induced apoptosis and human tumor suppression.
Table 3. ARA-211 sensitive and insensitive cell lines tested from the NCI 60 cell line panel do not correlate with C-Raf and B-Raf expression levelsOrigin Gel # Cell Line 2-AR pErk C-Raf B-Raf Colon 1 *Colo-205 + 2 HCC-2998 + 3 HCT-15 + + + 4 *HCT-116 ++ ++ ++ + 5 *HT29 ++ ++ ++ 6 KM 12 ++ ++ + 7 *SW-620 ++ + ++ CNS 8 SF-268 ++ + ++ 9 SF-295 ++ ++ 10 **SF-539 + ++ + + 11 *SNB-19 + + ++ 12 SNB-75 + + + ++ 13 *U251 + ++ + ++ Leukemia 14 CCRF-CEM + 15 HL-60 ++ ++ ++ 16 K562 ++ ++ ++ 17 MOLT4 18 RPMI-8226 19 SR ++ ++ + Lung 20 *A549 + + ++ 21 *EKVX ++ ++ 22 HOP-62 + + 23 HOP-92 ++ + + 24 NCI-H23 ++ + ++ 25 NCI-H226 ++ 26 NCI-H322M ++ 27 NCI-H460 + Mammary 29 *MCF7 + + + 30 MCF7-Adr Res ++ ++ + 31 HS 578T ++ + 32 **MDA-MB-231 + ++ + 33 *MDA-MB-435 ++ + 34 BT-549 + + ++ 35 *T-47D + Origin Gel # Cell Line 2-AR pErk C-Raf B-Raf Colon 1 *Colo-205 + 2 HCC-2998 + 3 HCT-15 + + + 4 *HCT-116 ++ ++ ++ + 5 *HT29 ++ ++ ++ 6 KM 12 ++ ++ + 7 *SW-620 ++ + ++ CNS 8 SF-268 ++ + ++ 9 SF-295 ++ ++ 10 **SF-539 + ++ + + 11 *SNB-19 + + ++ 12 SNB-75 + + + ++ 13 *U251 + ++ + ++ Leukemia 14 CCRF-CEM + 15 HL-60 ++ ++ ++ 16 K562 ++ ++ ++ 17 MOLT4 18 RPMI-8226 19 SR ++ ++ + Lung 20 *A549 + + ++ 21 *EKVX ++ ++ 22 HOP-62 + + 23 HOP-92 ++ + + 24 NCI-H23 ++ + ++ 25 NCI-H226 ++ 26 NCI-H322M ++ 27 NCI-H460 + Mammary 29 *MCF7 + + + 30 MCF7-Adr Res ++ ++ + 31 HS 578T ++ + 32 **MDA-MB-231 + ++ + 33 *MDA-MB-435 ++ + 34 BT-549 + + ++ 35 *T-47D + 60
Ovarian 44 *IGR-OV1 + ++ ++ 45 *OVCAR-3 + 46 OVCAR-4 + + 47 *OVCAR-5 + ++ 48 OVCAR-8 ++ 49 SK-OV-3 Prostate 50 DU-145 51 *PC-3 + + Renal 52 *786-0 + ++ + 53 A498 ++ 54 **ACHN + + + 55 CAKI-1 ++ ++ 56 RXF 393 + + 57 *SN 12C + + ++ 58 TK-10 + + 59 UO-31 Melanoma 36 LOX IMVI-+++ 37 M14 ++ + 38 MALME-3M + + 39 SK-MEL-2 40 SK-MEL-28 ++ 41 UACC-62 + + 42 UACC-257 + + Ovarian 44 *IGR-OV1 + ++ ++ 45 *OVCAR-3 + 46 OVCAR-4 + + 47 *OVCAR-5 + ++ 48 OVCAR-8 ++ 49 SK-OV-3 Prostate 50 DU-145 51 *PC-3 + + Renal 52 *786-0 + ++ + 53 A498 ++ 54 **ACHN + + + 55 CAKI-1 ++ ++ 56 RXF 393 + + 57 *SN 12C + + ++ 58 TK-10 + + 59 UO-31 Melanoma 36 LOX IMVI-+++ 37 M14 ++ + 38 MALME-3M + + 39 SK-MEL-2 40 SK-MEL-28 ++ 41 UACC-62 + + 42 UACC-257 + + -= little to no expression+ = moderate expression++ = high expression* = ARA-211 and Isoproterenol have no effect on pERKlevels** = ARA-211 and Isoproterenol decrease levels of P-Erk1/2-= little to no expression+ = moderate expression++ = high expression* = ARA-211 and Isoproterenol have no effect on pERKlevels** = ARA-211 and Isoproterenol decrease levels of P-Erk1/2Table. 3 Cell lines from the NCI 60 cell line panel were screened for ARA-211 sensitivity based on P-Erk1/2 inhibition. Likewise, the levels of 2 AR, P-Erk1/2, C-Raf and B-Raf were analyzed by western blot to determine if sensitivity is conferred by C-Raf versus B-Raf expression levels (Figure 10).Table 3. ARA-211 sensitive and insensitive cell lines tested from the NCI 60 cell line panel do not correlate with C-Raf and B-Raf expression levels 61
2-ARP-ERK1/2 C-Raf ColonCNSLeukemia B-Raf 2-ARP-ERK1/2 C-Raf ColonCNSLeukemia B-RafFigure 14. Western blot verification of 2 AR, P-Erk1/2, C-Raf and B-Raf expression levels from the NCI 60 cell line panel 2-ARP-ERK1/2 C-Raf LungMammaryMelanoma B-Raf 2-ARP-ERK1/2 C-Raf LungMammaryMelanoma B-Raf 62
63 Figure 14 (continued). Western blot verification of 2 AR, P-Erk1/2, C-Raf and B-Raf expression levels from the NCI 60 cell line panel 2-ARP-ERK1/2 C-Raf MelanomaOvarianProstate B-Raf 2-ARP-ERK1/2 C-Raf MelanomaOvarianProstate B-Raf Renal 2-ARP-ERK1/2 C-RafB-Raf Renal 2-ARP-ERK1/2 C-RafB-RafFigure 14. Western blot analysis of 2 AR, P-Erk1/2, C-Raf and B-Raf expression levels from the NCI 60 cell line panel. Cell lysateswere obtained from NCI, and western blot anaysiswas performed according to Materials and Methods.
64 ARA-211 inhibits anchorage independent proliferation, induces apoptosis, and suppresses the growth of human tumor xenografts in nude mice only in cell lines where it produces cA MP and inhibits P-Erk1/2 levels We next determined if ARA-211 suppre sses tumor growth only in those cells where it inhibits PErk1/2 in a cAMP-dependent m anner. To this end, we treated with ARA-211 the 6 cell lines identified above and processed them for assays for cAMP stimulation (ELISA), anchorage-independent pro liferation (soft agar growth), apoptosis (TUNEL) and in-vivo tumor growth (x enograft). Figure 15 demonstrates the stimulati on of cAMP in MDA-MB-231 (breast), SF-539 (cns), and ACHN (renal) tumor cell lines by 16, 13, and 16 fold with 10 M ARA-211 stimulation. However, ARA-211 failed to stimulate cAMP in A549 (lung), HCT116 (colon) and SNB-19 (cns) (summarized in Table 4). In agreement with Table 3, Table 4 also shows that ARA-211 inhibi ted P-Erk1/2 levels only in cell lines where it induces cAMP formation. Fu rthermore, Table 4 shows the anchoragedependent cell proliferati on is inhibited in a dos e dependent manner with ARA211 treatment of MDA-MB-231, ACHN, and SF-539, but not of A549, HCT-116, or SNB-19 cells. Inhibition of anc horage-dependent growth at 10 M ARA-211 stimulation was 60, 55, and 62 percent for ARA-211 sensit ive cell lines versus 9, 6, and 0 percent for insensitive cell li nes (Table 4). There was even greater inhibition of anchorage-independ ent cell proliferation, with sensitive cell lines inhibited by 82, 94, and 97 percent, whereas the insensitive cell lines showed 0, 7, and 5 percent inhibition at 10 M (Figure 16, Table 4).
65 0102030405060700.0000.0010.0100.1001.00010.000100.0001000.000 MDA-MB-231 SF-539 ACHN 0102030405060700.0000.0010.0100.1001.00010.000100.0001000.0 0 A-549 SNB-19 HCT-116ARA-211ARA-211Fold Induction of cAMPFold Induction of cAMP 0102030405060700.0000.0010.0100.1001.00010.000100.0001000.000 MDA-MB-231 SF-539 ACHN 0102030405060700.0000.0010.0100.1001.00010.000100.0001000.0 0 A-549 SNB-19 HCT-116ARA-211ARA-211Fold Induction of cAMPFold Induction of cAMPFigure 15. ARA-211 induces cAMP formation in MDA-MB-231, SF-539 and ACHN cells but not in A549, SNB-19 and HCT-116 cells Figure 15. cAMP ELISA assays were used, as described under Materials and Methods, to determine the sensitivity of cell lines to ARA-211 mediated stimulation of cAMP.
Figure 16. ARA-211 inhibits anchorage independent growth in MDA-MB-231, SF-539 and ACHN cells but not in A549, SNB-19 and HCT-116 cells 020406080 1 00 1 200.0000.0010.0100.1001.00010.000100.000 MDA-MB-231 SF-539 ACHN 0204060801001200.000.010.101.0010.00100.00 A-549 SNB-19 HCT-116ARA-211ARA-211Anchorage Independent Growth(% Control)Anchorage Independent Growth(% Control) 020406080 1 00 1 200.0000.0010.0100.1001.00010.000100.000 MDA-MB-231 SF-539 ACHN 0204060801001200.000.010.101.0010.00100.00 A-549 SNB-19 HCT-116ARA-211ARA-211Anchorage Independent Growth(% Control)Anchorage Independent Growth(% Control)Figure 16. Anchorage-independent growth on soft agar was used, as described under Materials and Methods, to determine the sensitivity of cell lines to ARA-211 mediated stimulation of cAMP. 66
05101520253035400.000.010.101.0010.00100.00 MDA-MB-231 SF-539 ACHN 05101520253035400.000.010.101.0010.00100.00 A-549 SNB-19 HCT-116ARA-211Figure 17. ARA-211 induces apoptosis in MDA-MB-231, SF-539 and ACHN cells but not in A549, SNB-19 and HCT-116 cells ARA-211Figure 17. A cell death detection apoptosis ELISA kit was used,as described under Materials and Methods, to determine the sensitivity of cell lines to ARA-211 mediated induction of Apoptosis.Apoptosis (%Control)Apoptosis (%Control) 67
68 ARA-211 suppresses tumor growth and causes tumor regression The ability of ARA-211 to inhibit tumo r cell growth and induce apoptosis in cultured cells in a 2 AR and Mek1/2-dependent ma nner suggested that ARA211 might induce tumor regression in human cancers where it can induce cAMP and block the C-Raf/Mek1/2/ Erk1/2 pathway. Theref ore, we determined the ability of ARA-211 to interfere with t he growth and progression of nude mice xenografts of the 6 human cancer cell lines described above. Figure 18-19 and Table 4 shows that ARA-211 was only able to suppress tumor growth in human tumors (MDA-MB-231, and ACHN) where 2AR stimulation occurs but not in those (A-549, SNB-19, and HCT-116) in which it does not. Figure 18 demonstrates that treatment with ARA-211 (i.p.) of mice bearing established MDA-MB-231 breast tumors under the mamma ry fat pads suppressed growth at 100mpk/day and actually caused tumor r egression at 200mpk/day. In the same MDA-MB-231 mammary fat pad model, is opreterenol at 200mpk/day did not cause tumor regression and inhibited tumo r growth by 89% (data not shown). After two weeks of treatment with ARA211 and total tumor regression the mice were taken off of the treatment regiment and monitored for tumor growth for an additional three weeks, dur ing which none of the tumo rs regrew and remained undetectable (data not shown). Similarly, the growth of AC HN renal tumors was inhibited when the mice we re treated i.p. with ARA211 (75 mpk/day) and tumor growth was blocked at 200 mpk/day (Figure 18, 200 mpk day 12 (see ). SF-539 cells did not grow consistently in nude mice, so no in-vivo data was collected.
02004006008000246810121416 050010001500024681012141618ACHNTumor Vol. (mm3)Tumor Vol. (mm3)MDA-MB-231 Figure 18. ARA-211 treatment of MDA-MB-231 xenograftsresults in tumor regression, and treatment of AHCN xenograftscompletely inhibits tumor growth in nude miceFigure 18. ARA-211 suppresses tumor growth and causes tumor regression. MDA-MB-231 cells were implanted under mammary fat pads (A.),and ACHN were implanted s.c. in nude mice (B.), and the mice treated (i.p.) either with vehicle (square )or ARA-211, 100mpk/day (MDA-MB-231, closed triangle), 200 mpk/day (MDA-MB-231, open triangle), 75mpk/day ( ACHN open circle) or 200 mpk/day (ACHN starting day 12, see ?) as described under Materials and Methods. A.B. 69
01002003004000246810121 4 010020030040050060002468101214 06001200180002468101214 A-549HCT-116Tumor Vol. (mm3)Tumor Vol. (mm3)Tumor Vol. (mm3)SNB-19DaysFigure 19. ARA-211 treatment of mice bearing A549, HCT-116 and SNB-19 xenograftshas no effect on tumor growthFigure 19. ARA-211 does not inhibit tumor growth in A549, HCT-116 or SNB-19 cell lines. A-549 (A.), HCT-116 (B.) and SNB-19 (C.) were implanted s.c. in nude mice and the mice treated (i.p.) either with vehicle (square) or 75mpk/day (circle) as described under Materials and Methods. A.B.C. 70
MDA-MB-231SF-539ACHNA-549HCT-116SNB-192-AR ExpressionP-Erk1/2 (% Inhibition)10010091000cAMP (Fold Induction)16.7 5.413.1 3.016.5 3.21.3 0.31.1 0.11.2 0.4Anchorage Dependent Growth (% Inhibition)60.6 6.555 6.962 12.79.6 6.16.5 4.00 (n=3)Anchorage Independent Growth (% Inhibition)82.7 7.994.4 1.597.9 1.20 (n=6)7.3 3.15.1 5.0Apoptosis (% Induction)15 4.09 1.711.4 1.93.1, 2.20, 1.20, 0Tumor Growth in Mice (% Inhibition)100*100000 Table 4. ARA-211 inhibits tumor growth and induces apoptosis only in human cancer cell lines where it induces cAMP and lowers P-Erk1/2 levelsTable 4. Summary of physiological effects of ARA-211 mediated inhibition of P-Erk1/2 compared to vehicle, 2 AR expression, induction of cAMP and apoptosis, inhibition of anchorage dependent and independent cell growth, and inhibition of tumor growth in nude mice xenografts. 71
72 ARA-211 in-vivo treatment suppresses P-Erk1/2 levels, inhibits tumor growth and induces apoptosis in human xenografts in nude mice Treatment with 200 M PK ARA-211 in the MDAMB-231 xenograft model resulted in complete tumor regression, wh ich left no tissue available at the end of the study for Immunohisctochemical (IHC) evaluation. Therefore, we repeated the xenograft study with 75 MPK to elic it a response to AR A-211 but allow for partial inhibition of tumor growth. Figur e 20 shows that treatment of mice with ARA-211 (75 mpk/day) inhibited tumor growth by 63%. Two hours after the final treatment the mice were euthanized, and the tumors were harvested and snap frozen. Tumors from MDA-MB-231 tu mor xenografts treated with ARA-211 (75 mpk/day i.p.) or vehicle were analyzed by IHC for inhibition of P-Erk1/2, inhibition of proliferation (Ki-67), and induction of apoptosis (TUN EL). Tumors from ARA211 treated mice had a significant decrease in cells staining positive for P-Erk1/2 (62%, p = 0.02) and Ki-67 (38% P = 0.007) along with a significant increase in cells staining positive for T UNEL (6-fold, P =0.0045).
73 010020030040004812 ControlARA-211 P-Erk1/2 0510152025303540DaysTumor Volume (mm3)% Positive StainingC 211* Figure 20. In-vivo treatment with ARA-211 suppresses P-Erk1/2 levels, inhibits tumor cell growth and induces apoptosis in human xenograftsin nude miceFigure 20. ARA-211 suppresses P-Erk1/2 levels in human xenograftsin nude mice. A. MDA-MB-231 cells were implanted under mammary fat pads and the mice treated with vehicle (square) or ARA-211 (circle, 75mpk/day) as described previously. B. Tumors were removed on day 12 and processed for P-Erk1/2 levels (IHC) as described under Materials and Methods. p < 0.05. A.B.
74 Apoptosis (TUNEL)% Positive StainingProliferation (Ki-67) 010203040506070 01234567C 211C 211** ControlARA-211 Figure 20 (continued). In-vivo treatment with ARA-211 suppresses P-Erk1/2 levels, inhibits tumor cell growth and induces apoptosis in human xenograftsin nude miceFigure 16. ARA-211 inhibits proliferation and induces apoptosis in human xenograftsin nude mice. MDA-MB-231 cells were implanted under mammary fat pads and the mice treated with vehicle (square) or ARA-211 (circle, 75mpk/day) as described previously, tumors removed on day 12 andprocessed for proliferation (Ki-67) (A.) and apoptosis (Tunel) (B.) as described under Materials and Methods. p < 0.05.A.B.
75 Discussion A chemical biology approach was us ed to interrogate aberrant signal transduction circuits in human cancer cells about the importance of the 2 adrenergic receptor in regul ating cell division and tumor survival. Our studies clearly demonstrated t hat activation of the 2 adrenergic receptor by ARA-211 results in inhibition of anchoragedependent and independent tumor cell growth as well as induction of apoptosis, and tu mor regression in human tumors where stimulation of the 2AR results in blockade of the C-Raf/Mek1/2/Erk1/2 pathway. The mechanism by which ARA-211 inhibits the C-Raf/Mek1/2/Erk1/2 involves a cAMP-dependent PKA but not EPAC activation, as s hown in Figure 21. This result was surprising in that PKA is tr aditionally known as a protooncogene. The fact that PKA can modulate C-Raf activity is known, but the mechanism by which PKA regulates C-Raf is controversial(61). Studies using radiolabeled phosphopeptide mapping for C-Raf demons trated that serine 43 was the major residue directly phosphorylated by PKA in-vivo, but did not affect the kinase activity of C-Raf(133). This suggests that PKA either phosphorylates another inhibitory residue on C-Raf, or that PKA indirectly decr eases the binding affinity of C-Raf for Ras. Other data suggests that inhibitory phosphorylation of serine 259 by PKA on C-Raf mediates the negativ e regulatory effects of PKA on the
76 Raf/Mek/Erk1/2 cascade(112). We investi gated the affects of PKA activation by ARA-211 treatment of MD A-MB-231 cells on the phos phorylation of C-Raf on serine 43 and 259 residues, but did not find any changes compared to control treatments. It is possible that another proposed mechanism by which PKA indirectly inhibits C-Raf through activati on of Rap-1 leading to competition for binding sites on Ras is occurring in MDA-MB-231 cells. However, we analyzed the levels of GTP-loaded Rap-1 from MD A-MB-231 cells treated with or without ARA-211, but did not find a difference in activated Rap-1 (data not shown). Therefore, our data suggests that t he signaling mechanisms by which PKA controls C-Raf kinase activity in these breast cancer cells is more complex and will require more interrogation. Data obtained from our ARA-211 studies clearly demonstrates that ARA211-mediated inhibition of tumor cell growth is rescued by 2 AR antagonists, suggesting that growth inhibition requires 2 AR signaling. This is consistent with other studies that showed that isoprenaline, chol era toxin, forskolin, and 8bromo-cAMP inhibit DNA synthesis and ce ll growth in normal cells(67, 134). However, these studies failed to interr ogate the signaling responsible for these inhibitory effects. Increases in cAMP levels have also been shown to inhibit mitosis and MMP expression in MDA-MB -231 cells(67). T hese data, taken together with others demonstr ating the ability of cAMP to regulate MAPK-Erk signaling and proliferation in normal cells, suggest that cAMP signaling may have tumor suppressive effects. However, this thesis work provides the first comprehensive examination of the tu mor suppressive effects of cAMP
77 stimulation by a 2 AR agonist in multiple cancer cell lines both in-vitro and invivo.
C-RafRap1 Gs ARA-211 Gs Adenylyl Cyclase ATPcAMP PKA EPAC Ras MEK Erk1 Erk2 B-Raf Figure 21. ARA-211 stimulates crosstalk between 2 AR and Raf/MEK/Erk1/2 resulting in inhibition of P-Erk1/2 through a cAMP-dependent activation of PKA but not EPAC.Figure 21. Schematic representation of crosstalk between signaling from ARA-211-stimulated 2 AR to the Raf/MEK/Erk1/2 pathway through PKA or EPAC intermediates. Src 2AR 78
79 Another important finding fr om our studies is that cAMP induction by ARA211 treatment induces apoptosis in cancer cells both in-vitro and in-vivo. Currently we do not have evidence that di rectly relates the inhibition of the Raf/Mek/Erk kinase cascade with the induc tion of apoptosis seen with ARA-211 treatment, however there is data in the literature supporting the proapoptotic functions of cAMP. These data include investigations showing cAMP induction resulting in S and G2/M cell cycle arres t, with cells eventually undergoing apoptosis in neuroblastoma and leukemia cell lines(66). Fu rthermore, another group identified a leukemia ce ll line that is resistant to cAMP-induced apoptosis based on overexpression of Bc l-2(135). This demonstrates that the apoptosis induced by cAMP may be independent of Erk1/2 inhibition, but other groups showed that inhibition of Erk1/2 could also lead to apoptosis. In melanoma cells it was shown that S6 kinase (S6K), an effector of Erk1/2 downstream, phosphorylates the pr oapoptotic protein Bad, leading to sequestration in the cytoplasm(136). Therefore, when P-Erk1/2 is inhibited, Bad is not sequestered and can localize to the mitochondrial membrane where it blocks the prosurvival functions of Bcl-2(136, 137). Similar studi es have demonstrated that inhibition of Mek/Erk1/2 signaling results in accumulati on of Bim, which ultimately associates with Bax, resulting in t he induction of apoptosis(138, 139) Aside from inhibition of Mek/Erk1/2 activation, there have been reports of C-Raf modulating apoptosis, however these mechanisms were independent of the kinase activity of C-Raf. This is the case for Raf-related apoptos is through direct binding and influencing of MST2(140). These mechanisms all represent valid ways by which ARA-211
80 could induce apoptosis. Further work is necessary to elucidate the exact mechanisms causing the ARA-211mediated induction of apoptosis. Our studies also demonstrated that c onstitutively active Mek1/2 rescues from ARA-211 inhibition of tumor cell growth, suggesting that ARA-211 must inhibit C-Raf/Mek1/2/Erk1/2 to induce its antiproliferative effect. This is an important finding as a la rge number of human cancers harbor genetic alterations that ultimately result in hyperactivation of the C-Raf/Mek1/2/Erk1/2 pathway. For example, it is estimated that 30% of all human cancers contain Ras mutations that lead to hyperactivation of this pat hway(9). And yet, some tumors with activated Ras do not have hyperactivacti vated Erk1/2 due to activation of MAP kinase phosphatases. Similarly, a great majority of hu man cancers over express or contain constitutively activated receptor and non-receptor tyrosine kinases and/or growth factor autocri ne loops that also result in hyperactivation of the CRaf/Mek1/2/Erk1/2(141-144). More important ly, hyperactivation of this pathway is critical to the growth and survival of human tumors(72, 77). Finally, thwarting this pathway in human cancers may result in great benefits in the treatment of cancer patients since the genetic aberra tions (i.e. Ras mutations, EGFR and ErbB2 overexpression) that result in hy peractivation of the C-Raf/Mek1/2/Erk1/2 pathway have all been associated with poor prognosis, resistance to therapy and shortened patient life( 69, 70, 145, 146). Using 2 stimulation to cause tumor regression by blocking the CRaf/Mek1/2/Erk1/2 pathway is a novel and powerful approach th at could have a widespread use in cancer therapy. Indeed, tumors activate this pathway with a
81 broad spectrum of growth factors/cytoki nes and their receptors. For example, many growth factors such as EGF, PDGF, FGF, VEGF, IGF-1, heregulin and others all activate this pathway(72, 112, 141, 147). Therefore, tumors that express two or more of these growth factors may require a cocktail of anticancer drugs. In contrast, 2AR stimulation will block activation of the CRaf/Mek1/2/Erk1/2 pathway when it is driven by several growth factors since they all required Mek to activate Erk1/2(48, 53, 61, 62, 76). For example, we have shown that growth factor (i.e. EGF) stim ulation of P-Erk1/2 is blocked by ARA211 in MDA-MB-231 cells (data not shown). This approach, however, is limited to those human tumors that express the 2AR and where stimulation of this re ceptor results in blockade of CRaf/Mek1/2/Erk1/2. In normal cells it has been shown that adrenergic stimulation can either stimulate or inhi bit P-Erk1/2 and proliferation( 34, 35, 48). However, in tumor cells, it is not known if stimulation of the 2AR can stimulate tumorigenesis. In some nor mal cells expressing B-Raf, 2 adrenergic receptors stimulate Erk1/2 by activating B-Raf wh ich in turn activates Mek1/2(33, 148) while in other normal cells, 2 adrenergic receptors inactivate Erk1/2 by blocking C-Raf(50, 57, 59, 60). Therefore, 2 adrenergic receptors may also inhibit or stimulate Erk1/2 and prolif eration in tumors. To translate these important results to clinically relevant findings, fresh human tumor biopsies were used to determine the effects of 2 AR stimulation on P-Erk1/2 levels in 2 AR expressing tumors, as well as to determine if the ability of 2 AR stimulation inhibits prolif eration and/or induces apoptosis.
82 Chapter 2 focuses on the effects of ARA 211 on fresh human tumor biopsies from the operating rooms at Moffi tt Cancer Center. The fact that ARA-211 (pirbuterol) has previously been used ora lly and locally (149-151) in humans to evaluate its potential as an anti-asthmatic and for the treatment for congestive heart failure will facilitate its testing as an anti-cancer drug in hypothes is-driven clinical trials with patients where 2AR stimulation is predicted to block C-Raf/Mek1/2/Erk1/2.
83 Chapter 2 Determining the Prevalence of 2 AR expression and P-Er k1/2 Activation in Fresh Biopsies: ARA-211 Ex-vivo Studies
84 Abstract In Chapter 1 of this thesis we have demonstrated that st imulation of the 2 AR with ARA-211 induced tumor suppressi on by a mechanism involving a cAMPdependent activation of PKA, which in tu rn inhibited C-Raf and its downstream effectors Mek1/2 and Erk1/2. While this is an important fi nding, its benefit to patients will depend on confirming that 2 AR has tumor suppressor activity in patient tumors. Furthermore, based on our data in human cancer cell lines, it is anticipated that only certain patients whose tumors express the appropriate markers will benefit from this novel therap eutic approach. Therefore, it is critical that fresh patient biopsies first be examined for the levels of 2 AR as well as activation of Erk1/2, and that we determi ne whether the treatm ent of these fresh biopsies with ARA-211 ex-vivo leads to tumor growth suppression. In this chapter we determined by tissue array as well as fresh biopsies from Moffitt Cancer Center that appr oximately 25% of human tumors express both P-Erk1/2 and 2 AR. Furthermore, ARA-211 treatment ex-vivo of these fresh patient samples followed by immunohistochemical staining of the tissue for P-Erk1/2, 2 AR, Ki-67 for proliferation, and TUNEL for apoptosis, revealed that the effects are highly variable and need to be confirmed by xenograft studies of fresh biopsies.
85 Introduction Recently the adrenergic receptor/trime ric G-protein/Adenyl Cyclase/cAMP pathway was shown to crosstalk to the RT K/Ras/Raf/Mek/Erk pathway in a tissue specific manner in normal cells. Both and isoform subtypes have been shown to crosstalk with MAP kinase cascade via Gs, Gi, and activation(49, 152, 153). In some normal cell types such as adipocytes, endothelial cells, fibroblasts, smooth muscle cells, hepatocytes pancreatic acinar cells (AR42J), and bone cells (MG63), 2 AR stimulation of cAMP production results in inhibition of the Ras/Raf/Mek/Erk pathw ay (60). The mechanism by which 2 AR-mediated formation of cAMP blocks the Raf/Mek/Erk pathway is not well understood. For example, some reports have suggested a cAMP-dependent activation of PKA, which phos phorylates C-Raf and inactivates it. Other reports described a cAMP/PKA activation of Rap1 which binds Ras and inhibits its binding to C-Raf (60). Fi nally, others have shown that cAMP can activate the guanine nucleotide exchange factors EPAC1/2 that activate Rap1, which binds Ras and inhibits binding and activation of C-Raf (36, 37). In these cells where 2AR stimulation shuts down the Ras/ Raf/Mek/Erk pathway, normal cell proliferation is inhibited. On the contrary, in normal cells such as cardiac myocytes, rat thyroid cells, bone cells (ATDC5, MC4), prostate (PC12) and, sertoli cells, granulosa
86 cells, preadiposities, neuronal, and pi tuitary cells (GH4C1/GH3 cells), 2AR stimulation of cAMP format ion results in activation of the B-Raf/Mek/Erk pathway enhancing cell proliferation, inducing differ entiation, or protec tion from apoptosis (32, 34, 126). Unlike the nor mal cells where cAMP produc tion results in inhibition of C-Raf, in these normal cells cAMP produ ction appears to be st imulating B-Raf. Here again, several mechanisms have been proposed by which the increase of cAMP results in the activation of BRaf. One of these proposes a cAMP activation of EPAC, which activates Rap1 to activate B-Raf by an as yet unknown mechanism. Another reported mechanism involves cAMP activation of PKA, which activates Src, in turn activating Rap1 to activate B-Raf(34). Here PKA directly phosphorylates Src on serine 17, leading to activati on of Rap-1 through exchange factors activated i ndirectly by Src(109). In contrast to normal cells, lit tle is known about the role of 2AR stimulation in tumor growth and surviv al. The previous chapter demonstrated that activation of the 2 AR in tumor cell lines MD A-MB-231, ACHN, and SF-539 resulted in induction of cAMP and subsequ ent stimulation of PKA kinase activity resulting in inhibition of C-Raf, but not B-Raf, kinase activity and inhibition of anchorage-dependent and -independent cell prolif eration, induction of apoptosis, and inhibition of tumor growth in nude mi ce. It was also shown that these physiological effects were dependent on cA MP activation of PKA, but not EPAC, that resulted in the inhibiti on of Erk1/2 phosphor ylation. However, these results relied only on cell lines, and we felt that translating these important findings to the clinic is critical. In order to do th is we first needed to confirm our results and
87 conclusions in fresh human biopsies. There were several questions we needed to address; these are: 1. What is the prevalence of 2AR expression and PErk1/2 levels in fresh human biopsies? 2. Are there certain tumor types that express these important mark ers, whereas others do not express these markers? 3. Are tumors that express 2AR and P-Erk1/2 more sens itive to ARA-211 than those that do not, as we hav e seen in cell lines? 4. Are there any human tumors where ARA-211 may actually stimulate proliferation? In this chapter we describe our work where we examined fresh tumor biopsies from patients from Moffitt C ancer Center for expression of 2AR and PErk1/2, as well as the effects of AR A-211 ex-vivo treatment of these fresh biopsies on P-Erk1/2 levels, tumor cell proliferation, and apoptosis to answer these questions. In this chapter we deter mined by tissue array as well as fresh biopsies from Moffitt Canc er Center that approxim ately 25% of human tumors express both P-Erk1/2 and 2 AR. Furthermore, ARA-211 treatment ex-vivo of these fresh patient samples followed by immunohistochemical staining of the tissue for P-Erk1/2, 2 AR, Ki-67 for proliferation, and TUNEL for apoptosis, revealed that the effects are highly va riable and need to be confirmed by xenograft studies of fresh biopsies.
88 Materials and Methods Human Tissue Array Using stage oriented human cancer ti ssue microarrays (Img enex, Inc., San Diego, CA, catalogue # MB2 60), 56 tumor samples were analyzed for 2 AR and MAPK expression by immunohistochem istry (four cores revealed no evidence of tumor). The tumors included in vasive ductal carcinomas from breast (8); medullary carcinoma of breast (2); hepatocellular carcinomas (10); transitional cell carcinomas of the bladder (9); mucinous adenocarcinoma of urachal remnant (1); papillary serous cystadenocarcinomas of the ovary (9), mucinous cystadenocarcinoma of t he ovary (1); pancreatic duct adenocarcinomas (10); prostatic adenocarcinomas (6). The patients had an average age of 59 years (range 16 to 85). Thirty were male and 26 were female. The tumors were staged according to the TNM system, following the recommendations of the Americ an Joint Committee on Cancer, 5 th Edition. The stage of the tumors was as follow: 4 patients had stage I, 9 stage II, 28 stage III, and 15 stage IV disease.
89 Immunohistochemistry. Anti-2 Adrenergic Receptor ( 2 AR) rabbit polyclonal antibody (H-73; sc-9042, Santa Cruz Biotechnology, Inc., dilu tion: 1: 400) and phospho ERK1/2 rabbit monoclonal antibody (20G11; Cell Signaling Technology Inc., dilution 1:200) were applied to 3M sections from formalin fi xed, paraffin embedded tissue specimens, using the avidin-biotin-perox idase complex method (Vectastatin Elite ABC Kit, Vector, Burlingame, CA), following the manuf acturers instructions. Slides were blocked with normal serum for 20 minutes, followed by incubation with the Anti-2 Adrenergic Receptor and the pERK 1/2 primary antibodies, at the dilution given, for 60 min. After rins ing with PBS for 5 minutes, sections were incubated with a biotinylat ed secondary antibody for 20 min. Following washing with PBS for 5 minutes, slides were incubated with avidin-biotin complex for 30 minutes and washed again. Chromogen was developed with 10 mg of 3,3 diaminobenzidine tetrahydrochloride (Sigma St. Louis, Mo) diluted in 12 mL of Tris buffer, pH 7.6 for 2 minutes. All samples were lightly counterstained with Mayer's hematoxylin for 30 seconds befor e dehydration and mounting. Positive controls (cell line and non-stained ti ssue) and non-immune protein-negative controls were used for each section. No antigen retrieval was performed. The stain was semi-quantitat ively examined by a b oard certified molecular pathologist. The positive reactions of each of the ant ibodies were scored into four grades according to the intensity of the staining: 0, 1+, 2+, and 3+. The percentages of 2 -AR, and p-ERK1/2 positivity in the tumor cells were also
90 scored into four categories: 0 (0%), 1 (1% to 25%), 2 (26% to 50%), 3 (51% to 75 %), 4 (76% to 100%) as determined by pathological scoring. The sum of the intensity and percentage scores was used as the final staining score. The final staining pattern of the tumors is defined as follows: 0 to1, negative; 2 to 3, weak; 4 to 5, moderate; 6 to 7, strong. Treatment of Patient Samples Ex-vivo Fresh human tumor tissue samples were ta ken from biopsies or resections at Moffitt Cancer Center and immediately placed in RPMI media containing 20% calf serum. Samples were cut into th in sections using a surgical scalpel and placed in placed in 24 well pl ates with media completely covering the tissue. Samples were treated with vehicle (0.1% DMSO) or ARA-211 at 100 M for 2 and 18 hours. These time points were selected to minimize the potential for tissue necrosis to occur. Tissue was coll ected at these time points and fixed in 10 % neutral buffered formalin for at least 48 hours. Fixed samples were taken to the histology core facility at the Univ ersity of South Flor ida College of Medicine for immunohistochemical staining for P-Erk1/2, 2 AR, Ki-67 and TUNEL. Staining analysis and sli de quantitation was provid ed by the Moffitt Cancer Center Pathology Core Facility.
91 Results Tissue arrays and fresh human biopsies reveal human tumor samples that express 2 AR and have activated Erk1/2 To determine if patients could benefit from ARA-211 therapy for cancer we first determined if human tu mor tissue expressed the 2 AR, which is necessary for ARA-211 to induce cAMP through adenyl cyclase activation, as well as whether they contained persistently active hyper-phosphorylated Erk1/2. To this end, we first determined the levels of 2 AR in a human tumor tissue array that contained tumor specimens from breast, liver, bladder, ovary, prostate and pancreatic tissue (Figures 22-27, Table 5). Strong to m oderate cytoplasmic 2 AR immunostaining (final score betw een 4 and 7), was observed in 100% of prostate (6/6) (Figure 27) and breas t (10/10) (Figure 22) carcinomas, independently of the tumor subtype; 80% of ovarian carcinomas (8/10) (Figure 25); 70% of pancreatic cancers (7/10) (F igure 26); 55% of bladder tumors (5/9) (Figure 24), and in only 30% of hepatocellu lar carcinomas (3/10) (Figure 23). One of the two ovarian tumors with low final 2 AR score (3) was a mucinous cystadenocarcinoma. Six (60% ) of the hepatocellular carcinomas, 4 (44%) bladder transitional cell carcinomas, and 1 (10%) of the pancreatic cancers
92 demonstrated no 2 AR expression (final score 0) Moderate to strong P-Erk1/2 reactivity (final score between 4 and 7) was observed in 50% of the ovarian tumors (5/5) (Figure 25); 2 of 10 bladd er tumors (20%) (Figure 24); 2 of 6 prostate cancers (33%) (Figure 27); 2 of 10 breast cancers (20%) (Figure 22), and in only 1 of 10 pa ncreatic tumors (10%) (Figure 26 ). All of the other tumors examined were found to be PERK1/2 negative (final score 0). The P-Erk1/2 immunostain localized to both nucleus a nd cytoplasm of the tumor cells, in the majority of the positive cases. In general, 2 AR was more commonly expressed in this array than P-Erk1/2. 2 AR was expressed in all breast and ovarian samples and in 90% and 75% of pancreatic and prostate samples. A significant inverse correlation between 2 AR and P-Erk1/2 staining was observed in 31 of 56 cases (55%). All together the data from the tissue array demonstrated that even with a small sample si ze we were able to id entify human tumors that express both the 2 AR as well as P-Erk1/2. In fact, approximately 20 percent of the tissue analyzed expressed both protei n markers. Furthermore, ovarian cancer appears to have the highest freq uency of expression of both markers (50%), followed by prostate (25%) and br east (20%). On the other hand, liver, bladder and pancreas tissues have the lowest with 0, 10 and 10%, respectively. These findings suggest that there ma y be patient populations whose tumors express both 2 AR and high levels of P-Erk1/2 which may respond favorably to ARA-211 therapy.
93 Breast infiltrating duct carcinoma IIBreast infiltrating duct carcinoma IIIBreast infiltrating duct carcinoma IIIBreast infiltrating duct carcinoma IIIBreast infiltrating papillary carcinoma IIITissue ArraySample P-Erk1/2 2 ARFigure 22. Tissue array identifies human breast tumor samples that express 2 AR and P-Erk1/2
94 Breast infiltrating duct carcinoma IIIBreast mixed infiltrating duct and lobular carcinoma IIIBreast medullarycarcinoma IIBreast atypical medullarycarcinoma IIBreast infiltrating duct carcinoma IIITissue ArraySample P-Erk1/2 2 ARFigure 22 (continued). Tissue array identifies human breast tumor samples that express 2 AR and P-Erk1/2
95 F iure 23. Tissue array identifies human liver tumor samples that express 2 AR and P-Erk1/2 g Liver combined hepatocellularand cholangiocarcinoma IIILiver hepatocellular carcinoma IIILiver hepatocellular carcinoma IILiver hepatocellular carcinoma IVLiver hepatocellular carcinoma IITissue ArraySample P-Erk1/2 2 AR
96 Liver hepatocellular carcinoma IILiver hepatocellular carcinoma IILiver hepatocellular carcinoma IIILiver combined hepatocellularand cholangiocarcinoma IIILiver hepatocellular carcinoma IIITissue ArraySample P-Erk1/2 2 ARsamples that express 2 AR and P-Erk1/2 Figure 23 (continued). Tissue array identifies human liver tumor
Urinary bladder mucinous adenocarcinomaUrinary bladder transitional cellcarcinoma IVUrinary bladder transitional cellcarcinoma IVUrinary bladder transitional cellcarcinoma IVUrinary bladder papillary transitional cell carcinoma IVTissue ArraySample P-Erk1/2 2 ARFigure 24. Tissue array identifies human bladder tumor samples that express 2 AR and P-Erk1/2 97
Urinary bladder papillary transitional cell carcinoma IVUrinary bladder transitional cellcarcinoma IIIUrinary bladder papillary transitional cell carcinoma IUrinary bladder transitional cellcarcinoma IIIUrinary bladder transitional cellcarcinoma IIITissue ArraySample P-Erk1/2 2 ARFigure 24 (continued). Tissue array identifies human bladder tumor samples that express 2 AR and P-Erk1/2 98
Ovary papillary serous cystadenocarcinoma Ovary papillary serous cystadenocarcinomaI Ovary papillary serous cystadenocarcinoma Ovary papillary serous cystadenocarcinoma Ovary papillary serous cystadenocarcinoma Tissue ArraySample P-Erk1/2 2 ARFigure 25. Tissue array identifies human ovarian tumor samples that express 2 AR and P-Erk1/2 99
Ovary papillary serous cystadenocarcinoma Ovary papillary serous cystadenocarcinoma IOvary papillary serous cystadenocarcinoma IIIOvary mucinous cystadenocarcinoma Ovary serous cystadenocarcinoma Tissue ArraySample P-Erk1/2 2 ARFigure 25 (continued). Tissue array identifies human ovarian tumor samples that express 2 AR and P-Erk1/2 100
Pancreatic ductal adenocarcinomaIPancreatic ductal adenocarcinomaIIPancreatic ductal adenocarcinomaIIIPancreatic ductal adenocarcinomaIPancreatic ductal adenocarcinomaITissue ArraySample P-Erk1/2 2 ARFigure 26. Tissue array identifies human pancreatic tumor samples that express 2 AR and P-Erk1/2 101
Pancreatic ductal adenocarcinomaIIIPancreatic ductal adenocarcinomaVPancreatic ductal adenocarcinomaIIIPancreatic ductal adenocarcinomaIIIPancreatic ductal adenocarcinomaIII Tissue ArraySample P-Erk1/2 2 ARFigure 26 (continued). Tissue array identifies human pancreatic tumor samples that express 2 AR and P-Erk1/2 102
Prostate adenocarcinoma Prostate adenocarcinoma Prostate adenocarcinoma Prostate adenocarcinoma Prostate adenocarcinoma Tissue ArraySample P-Erk1/2 2 ARFigure 27. Tissue array identifies human prostate tumor samples that express 2 AR and P-Erk1/2 NDNDND = Not Determined 103
Prostate adenocarcinoma I Prostate adenocarcinoma Prostate adenocarcinoma IIIProstate adenocarcinoma Tissue ArraySample P-Erk1/2 2 ARFigure 27 (continued). Tissue array identifies human prostate tumor samples that express 2 AR and P-Erk1/2 104
25105010020P-Erk1/2 b2575Prostate(n =9)1090Pancreas(n =10)50100Ovary(n =10)1060Bladder(n =10)040Liver(n =10)20100Breast(n =10)2 AR & P-Erk1/2 c2 AR aTissue Type 25105010020P-Erk1/2 b2575Prostate(n =9)1090Pancreas(n =10)50100Ovary(n =10)1060Bladder(n =10)040Liver(n =10)20100Breast(n =10)2 AR & P-Erk1/2 c2 AR aTissue Type a -% of tumors positive for 2 ARb -% of tumors positive for P-Erk1/2c -% of tumors positive for bothTable 5. Summary of tissue array staining results for 2 AR and P-Erk1/2 105
106 Fresh human tumor biopsies from Moffitt C ancer Center were obtained to extend the findings of the tissue array to fresh sa mples. The tissues were then fixed in 10% neutral buffered formalin and stained for expression of 2 AR and P-Erk1/2. Results obtained from these tissues were similar to the tissue array in that approximately 28 percent of the ti ssues stained positive for both 2 AR as well as P-Erk1/2 (Table 6). Figure 28 demonstrates the typical staining patterns of both 2 AR and P-Erk1/2 in the fresh human tu mor biopsies. P-Erk1/2 staining is localized to both the nucleus and cytoplasm, whereas 2 AR staining is seen mostly in the cytoplasm and somewhat around the plasma membrane. Table 6 further characterizes the staining perce ntages seen in the tumor biopsies collected from all tissue types. Result s from the fresh biopsy samples differed from those of the tissue a rray in that tumors from the liver, colon and pancreas demonstrated the highest percent age of co-staining for the 2 AR and P-Erk1/2 at 50%. Interestingly, 100% of the liver, retroperitoneal and pancreatic tumor tissues showed moderate to strong staining for 2 AR, however even in some tissues that expressed high 2 AR levels there were st ill high levels of P-Erk1/2 co-expressed. This suggests that ei ther the activating ligands for the 2 ARare not present, or that 2 AR signaling does not result in inhibition of the CRaf/Mek/Erk pathway in these tissues. Ul timately, 56% of all tumor types tested showed 2 AR staining. In contrast to the tissue array staining results in which 2 AR staining was more prevalent than P-Erk1/2, the samples collected from Moffitt Cancer Center demonstrated similar staining percentages for P-Erk1/2 at
107 57%, compared to 2 AR at 56%. The differences in staining between the two sets of tissues may be due to the wider variety cancer tissues obtained from Moffitt Cancer Center, compared to the 6 tumor types from the array. 100% of the sarcomas, brain tumors, and mediasti num tumors demonstrated staining for P-Erk1/2, although the sample size for thes e tissues were low. Likewise, 50% of the tumors tested for 12 of the 15 tumor types showed expression of P-Erk1/2. These findings confirm that there may be a patient population whose tumors express both 2 AR and high levels of P-Erk1/2 however we do not know if stimulation of the 2 AR in these tumors will result in decreases in P-Erk1/2, inhibition of cell proliferati on or induction of apoptosis.
Pt SampleLiver LungLungMediastinumPancreas+ + ++ +P-Erk1/2 2AR+ + ++ + ++ + ++ + + + +Figure 28. Fresh patient samples from Moffitt Cancer Center express 2 AR and P-Erk1/2 108
Table 6. Summary of fresh human tumor biopsy staining results for 2 AR and P-Erk1/2 3310033Sarcoma(n=3)03333Stomach(n=3)406060Kidney(n=5)3366100Retro-peritonel (n=3)507550Colon(n=4)206060Breast(n=5)5050100Liver(n=6)60P-Erk1/2b4060Lung(n=5)2 Ar & P-Erk1/2c2 ARaTumor Type 3310033Sarcoma(n=3)03333Stomach(n=3)406060Kidney(n=5)3366100Retro-peritonel (n=3)507550Colon(n=4)206060Breast(n=5)5050100Liver(n=6)60P-Erk1/2b4060Lung(n=5)2 Ar & P-Erk1/2c2 ARaTumor Type a -% of tumors positive for 2 ARb -% of tumors positive for P-Erk1/2c -% of tumors positive for both 109
05050Omentum(n=2)100100100Mediastinum(n=1)2 Ar & P-Erk1/2cP-Erk1/2b2 ARaTumor Type000Thyroid(n=1)01000Brain(n=1)0050Lymph Node(n=2)05050Head and Neck(n=2)5050100Pancreas(n=2) 05050Omentum(n=2)100100100Mediastinum(n=1)2 Ar & P-Erk1/2cP-Erk1/2b2 ARaTumor Type000Thyroid(n=1)01000Brain(n=1)0050Lymph Node(n=2)05050Head and Neck(n=2)5050100Pancreas(n=2) Table 6 (continued). Summary of fresh human tumor biopsy staining results for 2 AR and P-Erk1/2 Table 6. Summary of fresh human tumor biopsy staining results for 2 AR and P-Erk1/2 as described under Materials and Methods. a -% of tumors positive for 2 ARb -% of tumors positive for P-Erk1/2c -% of tumors positive for both 110
111 Fresh human tumor samples treated ex-vi vo demonstrate variable efficacy of ArA-211 The results described under Figur es 22-28 and Tables 5 and 6 demonstrated that about 1/4 (~ 25%) of human tumors analyzed express both 2 AR and have activated Erk1/2. We next attempted to determine whether the expression of these 2 markers predicts sensitivity to ARA-211 in fresh human biopsies treated ex-vivo. We therefore hy pothesized that at least some of the human tumors that express both 2 AR and P-Erk1/2 wi ll respond to ARA-211 treatment ex-vivo, as shown by decreased P-Erk1/2 levels (I HC) and proliferation (Ki-67 staining), and increas ed apoptosis (TUNEL staining). To test this hypothesis we prepared the fresh biopsie s in a variety of ways to optimize preserving the human cancer cells as well as to avoid necrosis, independent of drug treatments. The first cohort of samples was processed by mincing the tissue to increase the exposure to ARA211. Treatments we re done in a time course of 0.5, 2, 6, 18, and 24 hours at 20 M to determine the best time point to harvest the tissue and analyze expression of P-Erk1/2, 2 AR, Ki-67 and TUNEL. Prior to ex-vivo treatment a portion of t he sample was taken as a pre-treatment sample and analyzed for 2 AR and P-Erk1/2 expression. Analysis of the treated tumor samples revealed severe necrosis induced by mincing, which rendered the data obtained uninterpretable. The second cohort of samples was subsequently co llected and prepared for ex-vivo treatment by cutting the tiss ue into thin slices, and increasing the
112 concentration of ARA-211 from 20 M to 100 M to increase the tissue exposure to the drug. Vehicle treated samples were exposed for 24 hours, while ARA-211 treated samples were exposed for 2, 6, and 24 hours. Specimens were collected and analyzed for P-Erk1/2 and 2 AR expression, along with proliferation and apoptosis by Ki-67 and TUNEL staining, respectively. Upon further examination of the results it was determined that the 24 hour time point for the vehicle treated sample was too distant fr om the 2 hour time point to analyze P-Erk1/2 and 2 AR expression. In each case, the vehicl e treated samples demonstrated a loss of expression of both P-Erk1/2 and 2 AR compared to the pre-treatment samples collected before ex-vivo treatments. Due to the loss of expression in the 24 hour vehicle sample we were unable to determi ne the efficacy of ARA-211 to inhibit PErk1/2. Finally, 2 hour and 18 hour time points were used for both vehicle and ARA-211 treated samples for the third cohort of patient samples collected. Aside from this difference the samples were treated just as they were in cohort two. In this case 20 fresh biopsies were collected and treated ex-vivo. Pathological analysis determined that interpretable dat a was obtained using this procedure as necrosis was minimized and interfered in onl y 4 of the 20 samples. However, in some samples necrosis was induced in the ARA-211 treated samples, but not the vehicle-treated samples. Necrosis was determined by pathological analysis of HE staining morphology, indicating overall intactness of the tissue before and after ex-vivo treatment. From the rema ining samples in the absence of ARA-211 treatment there were 6 tumors that expressed both 2 AR and P-Erk1/2, 6
113 tumors expressed only the 2 AR, 2 tumors expressed only P-Erk1/2, and 1 tumor expressed neither 2 AR or P-Erk1/2. The majority of the ex-vivo treated samples did not demonstrate changes in P-Erk1/2 or 2 AR upon ARA-211 treatment. In 2 samples, Lung I and Kidney IV, P-Erk1/2 decreased 25% and 3%, respectively. These changes did no t change staining by Ki -67, but resulted in a 15% and 5% increase in TUNEL st aining, respectively. However, pathological analysis dete rmined that the TUNEL bac kground staining was high for several samples and additional apoptosis markers such as caspase 3 are needed. Ex-vivo treatment resulted in increases in P-Erk1/2 in 4 samples (Colon, Lung I, Kidney II, Omentum II), but with no other changes except for a 50% increase in the Ki-67 staining of the Omentum II sample from 15% to 30%. However, in the Lung I sample ARA211 treatment increased TUNEL staining from 5% to 45%, but at 18 hours the samp le also exhibited 90% necrosis. The 2 AR staining remained fairly constant at the 2 hour treatment time point among most of the samples treated ex-vivo. An increase in 2 AR expression was seen in 4 samples and a decrease was noted in only 2 samples, where the expression did not change among the remainder of t he 14 samples. Likewise, the Ki-67 staining for proliferation ex hibited little changes in thes e samples as well. Only 3 samples demonstrated a decrease in Ki-6 7 positive cells, wh ile 1 sample showed an increase. No correlation was noted between decreases in Ki-67 and decreases in P-Erk1/2. In fact, 2 of the samples in which Ki-67 decreased there was no P-Erk1/2 expression, and in 1 sample the P-Erk1/2 staining increased by 15% while the Ki-67 staining decreased by 30%. Taken together, based on the
114 limited sample size of 20 and t he variability of the result s, it is not possible to determine whether expression of 2 AR and P-Erk1/2 predicts response to ARA211. Indeed, only 6 samples expressed both 2 AR and P-Erk1/2. Furthermore, the necrosis observed not only interfered wit h the quality of t he data, but also made it difficult to interpret the data. This led us to question whether slices of tumor biopsies were the best approach to answer the questions we posed. Therefore, our conclusion is that an al ternative approach must be utilized. To this end, perhaps using fresh tumor biopsies that have been passaged in nude mice to select for tumors that grow in-vivo is a better approach(154).
122 Proposed method of selecting patients for evaluating the efficacy of ARA211 clinically Ultimately, the goal of tr eating patient samples ex-vivo is to determine which, if any, patients would benefit from ARA211 therapy for cancer. To this end we have proposed a flow chart for t he identification of patients who may be candidates for ARA-211 therapy (Figure 25). Patient samples that have been archived in the tissue bank at Moffi tt Cancer Center will be screened for expression of P-Erk1/2 and 2 AR expression. Those that do not express both protein markers will be considered off st udy, while those that express both markers will have a tumor biopsy done to confirm the current expression levels. These biopsies will also be used for ex-vivo treatment to determine the sensitivity of the samples to ARA-211 treatment. Those samples in which treatment decreases the levels of PErk1/2, decreases proliferation by Ki-67 staining, or increases apoptosis by TUNE L staining will be consid ered on study for treatment with ARA-211. Those ex-vivo treated samples that ex hibit no change or an increase in P-Erk1/2, Ki-67 or a decrease in TUNEL will be off study.
IHC on archived samples + 2-AR, + P-Erk1/2-2-AR, -P-Erk1/2+ 2-AR, -P-Erk1/2-2-AR, + P-Erk1/2 Off Study Fresh biopsy + 2-AR, + P-Erk1/2 Off Study Ex-vivo treatment-2-AR, -P-Erk1/2+ 2-AR, -P-Erk1/2-2-AR, + P-Erk1/2 P-Erk1/2Ki-67TUNEL On Study P-Erk1/2Ki-67TUNEL No /No /No /Off StudyFigure 30. Patient treatment with ARA-211 requires pre-determination of 2 AR and P-Erk1/2 expression, as well as ex-vivo treatment to determine efficacy to inhibit proliferation and induce apoptosis Figure 30. Flow chart schematic for selection of patients whosetumor may be susceptible to ARA-211 therapy. 123
124 Discussion The use of pirbuterol for cancer therapy is a realistic goal given the preclinical data demonstrated in Chapters 1 and 2 of this thesis as well as the extensive clinical data in the liter ature for the bronchoand vaso-dilatation properties of pirbuter ol as a specific 2 adrenergic receptor agonist. In fact, pirbuterol was approved by the FDA as an anti-asthma tic delivered by nebulizer. Because of this, pirbuterol has been te sted extensively in animals and humans, and the toxicity profiles along with the pharmacoki netics and pharmacodynamics are well defined. In order to effectively treat cancer though, we first need to determine which tumors are most likely to respond to pirbuterol by inhibiting proliferation and in ducing apoptosis. Previ ous data from Chapter 1 demonstrated that the antitumor effect of pirbuterol depends on its ability to bind the 2 AR and to inhibit Erk1/2 phosphorylation. Therefore, we hypothesized that patient samples must express the 2 AR along with high levels of P-Erk1/2 for the treatment to be effective. Un til now there was not much known about expression of the 2 AR in human tumors, and the link between 2 AR and PErk1/2 signaling was only demonstrated in cert ain cells in vitro. This led us to first investigate the expre ssion of these protein mark ers in tissue arrays from human tumors. To our surpri se we discovered that the 2 AR is ubiquitously
125 expressed in all tumor types, with 100% of ovarian and breast tissues staining positive. Furthermore, 90% and 75% of t he pancreas and prostate tumors tested expressed 2 AR, based on the tissue array results. It was also surprising that the localization of the 2 AR was in the cytoplasm and not in the plasma membrane. One possible explanation for th is mislocalization is that the antibody concentration used to stain the tissue was too high. Non-specific staining could account for false positive st aining in areas such as the nucleus and cytoplasm. Better characterization of the anti body, and optimizatio n of staining concentrations should be done for futu re studies using IHC to identify 2 AR expression in human tumor biopsies. In contrast to 2 AR, P-Erk1/2 levels we re found in only 10-50% of pancreas, bladder, breast, prostate and ovar ian cancer samples, and none of the liver cancer samples tested stained pos itive for P-Erk1/2, based on the tissue array results. With regards to the tissue that expressed both 2 AR and PErk1/2, we found tissue from ovarian c ancer having the highest frequency (50%) followed by prostate (25%), breast (20% ), bladder (10%) and pancreas (10%). On average 20% of the tiss ues from all tumor types stained positive for both 2 AR and P-Erk1/2. These data suggest that about 20% of hum an tumors may be susceptible to pirbuterol therapy. Howe ver, as discussed in the introduction to Chapter 2, in some normal cells 2 AR stimulation leads to an increase in PErk1/2 levels resulting in stimulation of cell proliferation thr ough B-Raf activation. In our human cancer cell li ne studies in Chapter 1, 2 AR stimulation did not
126 result in stimulation of pr oliferation in any of the ce ll lines used; whether this occurs in fresh biopsies is not known. It is therefore critical that the studies of Chapter 1 in human cancer cell lines be ex tended to fresh human biopsies. It is anticipated that 2 AR stimulation will result in i nhibition of P-Erk1/2, but clearly any patient whose tumor responds to pirbut erol by increasing P-Erk1/2 would not be a candidate for 2 AR agonist treatment. In order to determine how human tumors respond to pirbuterol we treated patient fresh biopsies ex vivo, and utilized IHC, Ki-67 and TUNEL staining to determine the ability of pirbuterol to inhibit P-Erk1/2, and cell proliferation an d to induce apoptosis. From the studies we have carried out we conclude that tr eating fresh biopsies in tissue culture poses challenges that were not overcome. Primarily the biggest obstacle was the inability to preserve the viability of the tumor tissues and stop the necrosis that led to lack of consistency among the responses to treatment. For example, the expression of P-Erk1/2 did not change much following AR A-211 treatment in most samples, whereas in some the expression increased and in others it decreased. Aside from the technical setb acks, analysis of the limited numbers of samples where the staining was clean did not reveal a correlation between inhibition of P-Erk1/2 and inhibition of proliferation or induction of apoptosis. Furthermore, tumor cell proliferation rates increased in a few sa mples that did not even appear to express the 2 AR. Another issue that was not totally overcome by the third cohort of samples was the ne crosis in some biopsies that occurred after as little as 18 hours of incubation. Even more surprising was the data that showed that there was more necrosis induced in the treated samples than the
127 vehicle samples ex-vivo. ARA-211-induced necrosis was not seen in tissue culture, as the primary cause of cell death was determined to be apoptosis by TUNEL staining (Chapter 1). TUNEL results for apoptosis may have been masked in the ex-vivo studies due to high background staining, most likely due to the necrosis seen in the samples. Further research needs to be done in the area of ex-vivo treatment of patient samples to determine the efficacy of pirbuterol treatment in order to push 2 AR therapy into the clinic. One way to deal with the issue of necrosis is to implant small pieces of the patient fresh tumor biopsies into the flanks of athymic nude mice, and then treat the mice with ARA211. Effects of t he drug in vivo can be determined by measuring the tumor gr owth in the presence and absence of treatment. Furthermore, the tissue ca n be harvested from the mice after treatment and staine d for P-Erk1/2, 2 AR, Ki-67 and TUNEL. Although, generally human tumor fresh biopsy explants in mice have a very low take rate, and it may take as many as 10 passages to get enough tumors for antitumor efficacy studies, recent studies however, demonstrated that these explants retain the majority of their initial characteristics as judged by gene expression profiling(154).
128 Chapter 3 RhoB, but not RhoA Overexpression Delays ErbB2Mediated Mammary Tumor Onset and Reduces Tumor Multiplicity in Transgenic Mice All of the work in this Chapter was performed by Adam Carie except for the transgene construct generation, which wa s performed by Cassandra Martin and Kun Jiang
129 Abstract Ras-homologous (Rho) proteins are GTPa ses that play essential roles in regulating cellular functions such as ce ll cycle progression, actin cytoskeletal rearrangement, cell motility, and have a strong link to oncogenic transformation and metastasis. The RhoA-like family of Rho GTPases are very interesting in that two of the family members, RhoA and RhoC are implicated in cell proliferation, inva sion and metastasis, whereas the third member RhoB has been implicated in tumor suppr ession and induction of tumo r apoptosis. Furthermore, RhoB expression was found to significantly decrease as tumors progress in head and neck cancer as well as tumors from lung cancer patients. Although in cultured cells, ectopic expression of RhoB antagonizes malignant transformation driven by oncogenes such as receptor tyrosine kinases, Ras and Akt, evidence for this in-vivo is lacking. Ther efore, we have developed transgenic mouse models in which wild type RhoB or RhoA is over expressed in the mammary fat pads under the control of the mouse mammary tumor virus (MMTV) promoter. The transgenic mice were then crossed with wild type MMTV-ErbB2 over expressing mice to determine the effe cts of RhoB and R hoA on ErbB2-mediated mammary oncogenesis. In this chap ter of the thesis, we show that overexpression of RhoB, but not RhoA, is sufficient to delay the tumor onset of
130 ErbB2 mediated tumorigenesis. Furthermo re, we demonstrat e that RhoB, but not RhoA overexpression resu lts in a decrease in tumor multiplicity compared to the ErbB2 or ErbB2/RhoA animals. This work provid es validation of the tumor suppressive effects of RhoB in a transgenic mouse model, which is lacking in the literature to date. Furthermore, this provid es us with the means to dissect out the signaling pathways that are affected by tumo r suppressive effects of RhoB. Also, the generation of the RhoA and RhoB transgenic mice will allow for the determination of physiological events t hat RhoA and RhoB may play in the normal proliferation, diffe rentiation and/or apoptoitic ev ents that mammary ductal epithelial cells undergo during the pregnancy cycle.
131 Introduction Ras-related low molecular weight GT Pases of the Rho subfamily act as molecular switches to transduce signals to mediate adhesion, morphology, motility, and cell cycle progre ssion(155, 156). On/off cyc ling of these switches are dependent on guanosine nu cleotide exchange factor s (GEFs) and GTPase activating proteins (GAPs) which control the rate of GTP loading and hydrolysis(157). GEFs can be activated by growth factors and cytokines through stimulation of receptors upstream, l eading to exchange of GDP-bound for GTPbound Rho proteins(156, 158) Under normal physiological conditions Rho GTPases are under tight regul ation by both guanosine dissociation inhibitors (GDIs) and reliance on post-translational modification. Rho-GDI inhibits spontaneous nucleotide exc hange and intrinsic GTPase activity, as well as masking of the lipid modification required for localization(159-162). Sub-cellular localization of activated Rho family members is crucial for the downstream activation of cellular effect ors(163). Membrane localiza tion is achieved through post-translational modification of the Cterminal cysteine with a lipid prenyl group(163, 164). Further r egulation of some Rho family members (such as RhoB) is achieved through additional modi fication of C-terminal cysteine motifs by palmitoylation. This also affects sub-cellular lo calization by preventing
132 association with Rho-GDI resulting in a more robust activation(165). Many of the Rho family members have high sequenc e homology, so localization plays an important role in each me mbers physiological func tion. RhoB is unique among the RhoA-like family members in that it can be modified by both farnesyl (F) and geranylgeranyl (GG) lipid groups(163). Prenylation of endogenous RhoB was found to be approximately equal, with half of the protein farnesylated and localized to the plasma membrane, whil e geranylgeranylated RhoB was found in endosomal compartments(166). Further studies demonstrated that when farnseylated RhoB was inhbited by FTI treatment, the population of geranylgeranylated RhoB increased and localized to late endosomal compartments(167). However, the c hanges in prenylation induced by FTI treatment did not result in differences in the antitumor properties of RhoB. Both RhoB-F and RhoB-GG inhibit anchorage-dependent and -i ndependent growth, induce apoptosis, inhibit constitutive activation of Erk and insulin-like growth factor-1 stimulation of Akt, and suppress tumor growth in nude mice(168). The Rho family of proteins can be divided into six distinct sub-families based on amino acid sequence and biological function. These proteins modulate many similar biological effects, including growth and cell cycle promotion, regulation of gene expression, and i nduction of actin cytoskeleton reorganization(169, 170). Many of these fam ily members, such as RhoA, Rac1 and cdc42, have been implicat ed in malignant transformation, invasion and metastasis, and drug resistance(87, 171, 172) One intriguing family of RhoGTPases is the RhoA-related proteins, wh ich is comprised of RhoA, RhoB, and
133 RhoC. Although the three members of this family share an amino acid sequence homology of 85%, share GEFs and effect ors, and stimulate actin-myosin contractility, there are gl aring functional differences that may be dictated by distinct sub-cellular localization(155, 163, 173, 174). Despite the amino acid similarity, RhoB contrasts with RhoA and RhoC in regulation and in physiological effects. RhoB is tightly regulated at the transcriptional, translational and posttranslational levels. The half-life (30120 minutes) of RhoB mRNA and protein is much shorter than that of other Rho prot eins (18-24 hours). RhoB expression is induced by growth factors, UV-irradiati on and many chemical agents. RhoB is also up-regulated during G1 and S phases of the cell cycle(175) RhoB, but not RhoA or RhoC, has been shown to exist within cells in both farnesylated (F) as well as geranylgeranylated (GG) forms, and is also modified by the fatty-acid palmitoyl group(163). Similarly, RhoB has been shown to be crucial for stressinduced apoptosis and anti-neo plastic activity (22, 176) As opposed to RhoA, RhoB has been suggested to ha ve tumor suppressive activity. This is based on data demonstrating that over-expression l eads to inhibition of cancer cell proliferation, induction of apoptosis, and inhibition of tumor growth in a nude mouse xenograft model(168, 177, 178). Furthermore, RhoB has been shown to negatively regulate NF B gene transcription, leading to apoptosis in response to genotoxic stress(179). Support for the ani tneoplastic role of RhoB has been sustained clinically by st udies demonstrating suppression of RhoB expression in invasive carcinoma from head and neck tu mors, and loss of RhoB expression during human lung cancer progression(22, 23).
134 As opposed to RhoB, RhoA and RhoC expression correlates with tumor progression, invasiveness and metastas is(180-183). RhoA has been implicated in activation of STAT3 and STAT5 leading to cell migration, proliferation, and epithelial-to-mesenchymal transition (EMT)( 184, 185). Similarly, siRNA to RhoA and RhoC inhibits the pro liferation and invasiveness of human gastric carcinoma by modulating the PI3K/Akt pathway(186). Microarray analysis of transformation by RhoA(Q63L) in NIH3T3 cells demons trates up-regulation of AP1, E2F and cMyc through Rock and other effectors(14). Likewise, microarray studies in MCF10A cells demonstrate that over-exp ression of RhoC resu lts in up-regulation of genes involved in cell pr oliferation, invasion/adhes ion, and angiogenesis(183). Similarly, over-expression of RhoA results in up-regulation of proliferative genes such as cyclin D1, cyclin-dependent kina se 8, cyclin A2 an d HMGI-C(183, 187). It is clear from the studies described in t he literature from our lab and others that though RhoA and RhoB are highly homologous, they play opposing roles in oncogenesis, with RhoA having tumor pr omoting whereas RhoB having tumor suppressive activities. However, t hese conclusions are based on correlative studies in cell culture, nude mice, and human paraffi n-embedded biopsies. To gain direct evidence in support of the tumor suppressive and tumor promoting activities of RhoB and RhoA, respectively, we created RhoB and RhoA transgenic mice under the transcription al control of the mammary-specific promoter, MMTV. We t hen crossed these mice with transgenic mice that aberrantly express the recept or tyrosine kinase ErbB2, al so under the control of the MMTV promoter. These mice were generated to determine the effects of
135 RhoA and RhoB on ErbB 2-mediated breast oncogenes is; specifically tumor onset, growth rate and multiplicity. Furt hermore, little is known about the role of RhoA or RhoB on normal mammary gl and morphogenesis. Therefore, we will also use the RhoA and RhoB transgenic mice we have generated to investigate the role of these GTPases on the morphol ogical changes that the mammary fat pads undergo during pregnan cy. The major stages of mammary gland development occur at puberty, from 3 weeks to 3 months of age, and during pregnancy, where epithelial cells proliferat e, differentiate, and eventually undergo involution by apoptosis. Initial development occurs during and just after puberty, where the ductal structures lengthen an d branch out filling the mammary fat pads. The rapid onset of ductal epithelial cell proliferation le ads to glands that are prepared to sustain differentiation and produce milk for the pregnancy stages to come(188). In the early stage of pregnancy the ductal branches undergo another round of substantial cell proliferation and the al veolar buds begin to form for milk production(189). Differentiation begins in the second half of pregnancy, characterized by the lobuloalveolar phase of growth leading to the cleavage of alveolar buds that will become milk-sec reting lobules at la ctation(190). This stage nears completion by D18 pregnancy, where the alveolar epithelial cells produce milk proteins and lipids. The final stages of gland morphological changes occur at involution; afte r weaning the glands undergo massive apoptosis and remodeling(191). The majori ty of epithelial cells will undergo apoptosis between day 2 and 3 of involution, leading to collapsing of alveolar structures and significant remodeling of the glands beginning around day 6 of
136 involution, and lasting until as long as day 21(192, 193). By utilizing these known morphological changes we c an identify the stages at which RhoA or RhoB have physiological effects on proliferation, differentiation or apoptosis. In summary, based on the body of literature supporting contrasting roles for RhoA and RhoB, we set out to validat e in this chapter our hypothesis that RhoB has tumor suppressive activity using transgenic mouse models comparing the effects of RhoA and RhoB expres sion on breast tissue proliferation, differentiation and apoptosis, as well as breast tissue neoplasia. To this end we have created transgenic mice that expres s RhoA or RhoB under the mouse mammary tumor virus (MMT V) promoter to determi ne the physiological and morphological effects of these Rho pr oteins on normal breast tissue. Furthermore, we have created bitransgenic mice that expre ss the ErbB2 protooncogene in the mammary fa t pads in combination wi th RhoA or RhoB to determine the effects of eac h Rho protein on the Er bB2-mediated spontaneous tumor onset, growth ra te and multiplicity.
137 Materials and Methods cDNAs and Gene Subcloning To generate the constructs for human RhoB and RhoA containing plasmids under the MMTV promot er, approximately 2 g MMTV-containing human TGFalpha gene (Figure 31) along with pcDNA3-HA-RhoB and pcDNA3-HA-RhoA plasmids were all digested using Ec oRI restriction endonuclease in its appropriate NEB buffer. The M ouse Mammary Tumor Virus (pMMTVglobin) vector containing the human TGF-alpha gene was supplied by Dr. Richard Joves lab at Moffitt Cancer Center. The gene for human RhoB was a gift from Dr. Gilles Favre. The human RhoA gene was a gift from Dr. C hanning Der at the Lineberger Comprehensive C ancer Center, University of North Carolina at Chapel Hill. This restriction endonucleas e digestion resulted in the removal of the TGF-alpha gene sequence and linearization of all 3 plasmids. Following digestion, the MMTV vector was treated wi th calf intestinal alkaline phosphatase (CIP) in order to increase ligation effi ciency. The RhoB insert, RhoA insert and MMTV vector were then isolated via agar ose gel (1.5%) electrophoresis (Figure 32). The desired bands were then remo ved from the gel and purified with Bio101 Geneclean gel extraction kit following the pr otocol provided in the kit. EcoRI,
138 XhoI, BglI, SgrAI, XbaI, PvuI and KpnI restriction endonuc leases as well as calf intestinal alkaline phosphatase (CIP) and T4 DNA ligase were all purchased from New England Biolabs (Beverly, MA). The NotI, BglII and EcoRV restriction endonucleases were purchas ed from Promega (Madison, WI). The Geneclean gel extraction kit was purchased from Bio 101 (Carlsbad, CA).
Figure 31. MMTV-TGFconstruct linearized by EcoRI digestion for insertion of human RhoA and RhoB genes MMTV Vector ~ 5.8 kbLinearizedMMTV VectorEcoRIEcoRIGlobin Poly AMMTV Promoter EcoRI DigestionFigure 31.Mouse Mammary Tumor Virus vector map. The MMTV vector map shown here indicates the significant restrictionendonucleasesites in thesubcloningof human RhoA and RhoB into this vector as well as the linearized vector following EcoRI digestion. TGFTGF 139
140 The Rho inserts were then ligated to the linearized MMTV vector. To this end, a small volume (<1 l) of MMTV vector along with 16 times more RhoA insert, RhoB insert or no insert was combined wi th T4 DNA ligase in ligation specific buffer. These reactions were incubated at 16 C for a period of 16 hours. The products were once again purified with the gel extraction kit, as noted earlier, in order to remove any salt from the ligat ion buffer that could potentially interfere with the next step. DH10B bacteria cells were then transformed with 1 l MMTVRhoA, MMTV-RhoB or MMTV control purif ied ligation product in electroporation cuvettes. The cuvettes were each placed in the electroporation machine that force current through the sample in order to induce transformation. This apparatus was set at 2.5 kV/resistance, 2.45 kV and 129 for resistance. The bacteria were then transferred into 500 l SOC media and placed in the 37 shaker until sufficient bacterial growth was apparent (approxim ately 1 hour). 100 l from each sample was plated onto LB agarose ampicillin pl ates, and incubated at 37 C overnight. DH10B competent cells and the pcDNA3 vector were purchased from Invitrogen (Carlsbad, CA) along with SOC media and LB broth base. Bacto-agar was obtained from Difc o Laboratories (West Molesey Surrey, UK). Electroporation cuvettes plus were ordered from Fisher (Pittsburgh, PA). Maxi prep and mini prep kits were obt ained from Qiagen (Valencia, CA). The following oligonucleotide primer sequenc es were designed based on the RhoB and RhoA cDNA and MMTV promoter sequences. The following day the MMTV control pl ate had 12 clones, while RhoA had 25 clones and RhoB had 50. 10 clones each were pi cked from the RhoA and the
141 RhoB plates. These 20 clones were gr own in LB overnight and the DNA was prepared using a mini prep kit, following t he protocol provided in the kit. The purified DNA was then digest ed with EcoRI, as noted earlie r, in order to identify clones containing the RhoB or RhoA insert. RhoB clo nes positive for an insert were then digested with NotI and BglII restriction enzymes, while the RhoA plasmids were digested with EcoRV an d BglII to identify the clones that contained the RhoB or RhoA insert in the correct orientation. Correctly orientated MMTV-RhoB clon es result in 6 kb and 463 base pair bands, while clones containing an incorrectly orientat ed RhoB insert results in 6 kb and 288 base pair bands when digested in this manner. Correctly orientated MMTV-RhoA clones would result in 6 kb and 527 base pair bands, wh ile clones containing an incorrectly orientated RhoA insert would result in 6 kb and 215 base pair bands when digested in this manner. Three Rh oA clones, clones 1, 2 and 8, and two RhoB clones, clones 2 and 3, were found to have co rrectly orientated inserts (Figure 31).
8.4kb4.8/4.3kb2.3kb1.3kb1.2kb700bp 1 2 3 4 5 6 7 8 9 10 8.4kb4.8/4.3kb2.3kb1.3kb1.2kb700bp8.4kb4.8/4.3kb2.3kb1.3kb1.2kb700bp8.4kb4.8/4.3kb2.3kb1.3kb1.2kb700bp8.4kb4.8/4.3kb2.3kb1.3kb1.2kb700bp1 2 3 4 5 6 7 8 9 101 2 3A. B. C. Figure 32. Isolation of MMTV, RhoB and RhoA DNA and confirmation of insertion in the correct orientation of RhoB and RhoA sequences in the MMTV vector 142
Figure 32. Isolation of MMTV vector, RhoA and RhoB inserts. MMTV vector, HA-RhoB insert, HA-RhoA insert isolation through 1.5%agarosegel electrophoresis followingEcoRIdigestion. 28 A. illustrates the isolation of the 5.4 kb MMTV vector (Lane 1) from the 2.3 kb insert. This figure also shows the isolation of the 622bpHA-RhoB insert (Lane 2) from its original pcDNA3 vector. Figure 28 A. shows the isolation of the 611bpHA-RhoA insert (Lane 3) from its original pcDNA3 vector. All three isolations were executed usingEcoRIrestriction enzyme.Figure 28 B. demonstratesligationof MMTV vector and RhoA or RhoB insert mini prep. Figure 28 B. shows 10 clones from the MMTV-RhoAligationplate from which plasmids have been isolated and digested withEcoRIin order to locate clones that had incorporated the RhoA insert.Clones 1,2,3,5,6 and 8 in this figure contain the RhoA insert. Figure 28 C. shows 10 clones picked from the MMTV-RhoBligationplate from which plasmids have been isolated and digested withEcoRIin order to locate clones that had incorporated the RhoB insert. Clones 1,2,3,4,5,6,7,8 and 10 in this figure contain the RhoB insert. All digests were run on 1.5%agarosegel. 143
144 One RhoA and one RhoB clone were grown separately in large (500 ml) culture, producing the desired DNA plasmid. Qiagen Maxi Prep kits were used to isolate the MMTV-Ha-RhoA and MMTV-Ha-RhoB plas mid DNA by following the protocol provided with the kit. The MMTV-Ha-RhoA and MMTV-Ha-RhoB vectors were then sequenced at the Molecular Biology Core Facility at Moffitt Cancer Center using the following primers (GIBCO BR L Life Technologies, Carlsbad, CA): #1 RhoB 5: ATG GCG GCC ATC CGC AAG AAG C Used to sequence RhoB from begi nning of insert downstream. #2 RhoA 5: GCT GCC ATC CGG AAG AAA CT Used to sequence RhoA from begi nning of insert downstream. #3 RhoB 3: CGC AGG CGG TCG TAG TCC TCC Used to sequence RhoB from middle of insert upstream to Beta Globin portion of vector. #4 RhoA 3: GCC TCA GGC GAT CAT AAT CTT CCT G Used to sequence RhoA from middle of the insert upstream to Beta Globin portion of vector. #5 MMTV 5: GGC GTA TCA CGA GGC CCT TTC G Used with RhoA and RhoB to sequence MMTV promoter from beginning of the promoter downstream. #6 MMTV 3: GGG TCC CCA AAC TCA CCC TGA AG Used with RhoA and RhoB to sequence MMTV promoter from end of the promoter upstream.
145 These sequencing results verified that the RhoA and the RhoB genes were successfully inserted downstream of the MMTV promoter, without mutation. MMTV-RhoB and MMTV-RhoA transgenes were then isolated from the remaining portions of the MMTV vector s via restriction enzyme digestion and agarose gel electrophoresis for the transge ne constructs to be inserted into the genome of mouse models through oocyte inject ions. This step was originally intended to be accomplished using KpnI and SgrAI restriction endonucleases for the MMTV-RhoB transgene, while the MMT V-RhoA transgene co uld be isolated using XhoI. The result of this XhoI digest, however, was to cut the MMTV-RhoA plasmid exactly in half, resulting in tw o 3.3 kb bands that ov erlapped in agarose gels. This problem was solved by adding PvuI restriction endonuclease, which cuts the undesired portion of the vector into 2.6 kb and 700 base pair bands (Figure 33). After reaching this step in the transgene generation process, it was found that the SgrAI enzym e that had been chosen to digest and isolate the MMTV-Ha-RhoB transgene was not functioning correctly (Figure 33).
8.4kb4.8/4.3kb2.3kb1.3kb1.2kb700bp1 2 3 4 5 6 7Figure 33. RhoA, but not RhoB DNA insertion is confirmed in MMTVvector after restriction enzyme digestionFigure 33. MMTV-RhoB and MMTV-RhoAtransgeneisolation. Restriction enzyme digestion to isolate the MMTV-RhoBtransgenewithKpnIandSgrAIenzymes was done to confirm insertion into the MMTV vector. Lane1 shows MMTV-RhoB plasmid uncut. Lanes 2 and 3 show MMTV-RhoB plasmid digested with onlySgrAIandKpnI respectively. Lane 4 shows the result of combining these two enzymes. Lane 5 shows the Lambda DNA digest ladder used as a standard. Likewise RhoA insertion was verified by restriction enzyme digestion. Lane 6 shows MMTV-RhoA plasmid uncut and lane 7 shows the MMTV-RhoA plasmid digested withXhoIalone. 146
147 Because there was no restriction enzyme that would cut the vector within 200 base pairs downstream of the poly-A sequence, it was necessary to again subclone the MMTV-Ha-RhoB transgene into a new vector in order to acquire a unique restriction enzyme site immediat ely downstream of the transgene. This was accomplished by digesting the MMTVRhoB transgene with K pnI and XhoI in NEB buffer #2, which yielded two bands of interest, one 2.7 kb band and one 600 base pair band and a third undes ired 3.2 kb band (Figure 34 B.). The new vector, pcDNA3 (Figure 34 A.) (approximately 5. 4 kb), was digested and also linearized with these enzymes in NEB buffer #2 and the resulting sticky ends were dephosphorylated by incubation with CIP. These three bands were then isolated via agarose gel (1.5%) electrophoresis and purified as described above. The pcDNA3 vector was chosen for this model because it contains a KpnI restriction site upstream of the XhoI si te with a unique XbaI site immediately downstream of the cloning site. The pcDNA3 vector and t he 2.7 kb band, both containing KpnI and XhoI digested ends we re then ligated using T4 ligase and the protocol mentioned earlier, along with a c ontrol lacking the 2.7 kb band.
Figure 34. pcDNA3 vector is linearized by KpnIand XhoIdigestion and MMTV-RhoB is inserted into the plasmid 1 2 3 48.4kb4.8/4.3kb2.3kb1.9kb1.3kb1.2kb700bp A.B. 148
Figure 34. A. The pcDNA3 vector map and insertion of MMTV-RhoB. This map shows the restriction sites that were important in the selection of this vector for its use in creating the new MMTV-RhoB construct. B. Isolation of pcDNA3 vector and MMTV-RhoBtransgene2.7 and 600bpbands byXhoIplusKpnIdigestion. Lane 1 shows MMTV-RhoB plasmid digested withKpnIalone (6.5 kb). Lane 2 shows MMTV-RhoB plasmid digested withXhoIalone (3.3 kb, 2.7 kb and 600bpbands). Lane 3 shows MMTV-RhoB plasmid digested with bothKpnIandXhoI(3.3 kb, 2.7 kb and 600bpbands). Lane 4 shows the pcDNA3 vectorlinearized throughKpnIplusXhoIdigestion. 149
150 DH10B competent bacteria cells were transformed using the ligation product and plated onto penicillin LB pl ate, following the protocol mentioned above. The control plate grew no colonies, while t he ligation plate contained 12 colonies. Clones were then picked, cultured and mini-prepped using a Qiagen kit. The resulting plasmid DNA was digested with XhoI, incubated with CIP and run on agarose gel (1.0%) in order to locate clones that had the 2. 7 kb band integrated (Figure 35). Clones containing this insert would result in an 8.1 kb band, such as the clones in lanes 4,6,10 and 11, while those lacking the insert would have a 5.4 kb band. The DNA from the pcDNA3 clones that had integrated the 2.7 kb insert of MMTV-HA-RhoB transgene were then isol ated and used for ligation of the 600 base pairs band. Both the pcDNA3 plus 2.7 kb linearized plasmid and the 600 base pairs insert include two XhoI sticky ends.
Figure 35. MMTV-RhoB insertion into the pcDNA3 vector is verified by the presence of the 2.7 kb MMTV-RhoB ligationband 1 2 3 4 5 6 7 8 9 10 118.4kb4.8/4.3kb2.3kb1.2kb700bpFigure 35. Verification of pcDNA3 plus MMTV-RhoB 2.7 kb bandligation. This gel demonstrates uncut pcDNA3 (Lane1),XhoIdigested pcDNA3 (Lane 2) and uncut clone #5 in lane 3. Clones 5 through 12 all digested withXhoIare located respectively in lanes 4-11. This gel was run in 1.5%agaroseand clones 5, 7, 11 and 12, which shift up with respect to lane 2, are positive for the 2.7 kb MMTV-RhoB insert. 151
152 DH10B cells were then transformed with this ligation pr oduct and plated. Following the over night incubation, 10 clones formed on the control plate while 39 grew on the ligatio n plate with the 60 0 base pair insert. 12 of the 39 clones that resulted on the ligation plate were picked, cultured and purified by mini-prep. The DNA was then digested with EcoRI re striction enzyme in order to locate clones that had incorporated this 600 base pairs band in the correct orientation. Correctly orientated clones resulted in 8 kb and 600 base pairs bands, while incorrectly orientated clones resulted in 7.5kb and 1.2 kb bands (Figure 36). The two positive plasmids that we re found were then sequenced using the MMTV and RhoB primers listed above so that the desired orient ation and lack of mutations could be verified. Sequencing results confirmed that the MMTV-HaRhoB transgene had been successfully subcloned into the pcDNA3 vector without mutation. The final step in this pr ocedure was to digest 50 g pcDNA3MMTV-Ha-RhoB plasmid with KpnI and XbaI along with digesting 50 g MMTVHa-RhoA plasmid with XhoI and PvuI, both in 400 l reactions to isolate the transgenes (Figure 37).
3.0kb2.0kb1.0kb500bp10.0kb1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16Figure 36. Verification of the integration of the 600 bp MMTV-RhoB insertFigure 36. Verification of the integration of the 600bpMMTV-RhoB insert. After the pcDNA3 plus 2.7 kb band wasligatedwith the 600bpband of the MMTV-RhoBtransgene, clones were picked and digested withEcoRIin order to verify integration and orientation of the insert.EcoRIdigested pcDNA3 (Lane 1) and MMTV-RhoB plasmid (Lane 2) are shown as controls in this 1.5 % gel. The remaining lanes containEcoRIdigested clones 1-14, respectively. Clones resulting in a 600bpband are oriented correctly while those resulting in 1.2 kb band are not and those resulting in neither 500bpband nor 1.2 kb band did not integrate the insert. 153
3.0kb2.0kb1.0kb500bp10.0kb1 2 3 4Figure 37. Final isolation of MMTV-RhoA and MMTV-RhoBtransgenesFigure 37. Final isolation of MMTV-RhoA and MMTV-RhoBtransgenes. This 1.5% gel displays the final isolation of the two desiredtransgenes. Both MMTV-RhoA and MMTV-RhoBtransgenesare 3.3 kb in length. The pcDNA3-MMTV-RhoB plasmid is shown uncut (Lane 1), then digested withKpnIandXbaIresulting in 5.4 kb and 3.3 kb bands (Lane 2). The MMTV-RhoA plasmid is shown uncut (Lane3), then digested withXhoIandPvuIresulting in 3.3 kb, 2.6 kb and 700bpbands (Lane 4). 154
155 A small aliquot of these sa mples were then run on agarose gel. The transgene restriction endonuclease reactions were then submitted to the Moffitt Cancer Center Mouse Models Core Facility. The first step in the transgenic mouse model generation is the isolation of the MMTV-Ha-RhoB and MMTV-Ha-R hoA transgenes via agarose gel electrophoresis. Once this was accompli shed, FvB/N strain female mice were bred with FvB/N males. 12 hours follo wing copulation, the female was euthanized, ovaries and fallopian tubes were harvested. Fertilized eggs (i.e. oocytes) were isolated from the reproduc tive organs. For each of the RhoA and RhoB transgenes about 180 eggs were harvested and mi croinjected with the DNA constructs. Only one of the two pre-nuclei of each oocyte was microinjected with about .003 l transgene DNA solution (9 femtograms DNA). The injected oocytes were t hen incubated for 24-48 hours at 39 C, the normal body temperature of mice. Once oocytes had divided into two or more cells, 2030 of these microinjected eggs were implanted into the fallopian tube of each of 6 foster mothers. Of these 180 inject ed eggs, only about 20% were expected to survive. A projected 20% of these expected 36 survivors (approximately 7-8 mice) should have the transgene incorpor ated into the genome, assuming that the transgenes are not lethal.
156 Southern Blot and Genotyping One RhoA founder and 5 RhoB founde rs were identified based on genotyping from tail DNA. The insertio n of the transgene was confirmed by Southern blot, which confirmed the int egration of the MMTV-HA-RhoB or MMTVHA-RhoA in B1, B2, B5, B6, and B9 founders, as well as A5 founder (Figure 38). Founder mice were tail snipped and ~1 cm of the tail snip was used for each mouse for genomic DNA extraction with the DNA-easy kit (Qiagen) following the provided protocol. Southern blot analysis of founder mi ce was done by isolating approximately 5 ug tail DNA, and samples were digested overnight with EcoRI restriction enzyme prior to electrophoresis on 0.8 % agarose in TAE gels. Gels were then transferred to Zetaprobe GT (BioRad) membranes. Southern blots were hybridized overnight to 32-PdCTP-labelled Rho A or Rho B cDNAs. Additionally, F2 generation mice were gen otyped by slot blot hybridization of genomic DNA isolated from ta il biopsies to 32P-dCTP-labe lled probes specific for non-murine transgene sequences ; SV40 poly A signal for MMTV-ErbB2 mice and rabbit beta-globin intron for MMTV-Rho A a nd -Rho B mice. Slot blot genotyping analysis was performed by the Mouse Model s Core facility at Moffitt Cancer Center as described previously(194).
Figure 38. Southern blot of DNA from transgenic founder mice confirms integration of RhoA or RhoB into the host genomeFigure 38. MMTV-HA-RhoA and MMTV-HA-RhoB contructswere given to the mouse models core facility at Moffitt Cancer Center to generate transgenic mice over expressing the HA-RhoA or HA-RhoB in the mammary fat pads. DNA from the tails of the founder mice were analyzed by southern blot to confirm integration into the host DNA. 157
158 DNA Preparation and PCR The mice were tail snipped and ~1 cm of the tail snip was used for each mouse for genomic DNA extraction with t he DNA-easy kit purchased from Qiagen following the provided protocol. The pur ified DNA samples were dissoloved in 0.1 x TE and stored in 4 degree until PCR reaction wa s set up. The standard PCR cycles were utilized to examine t he presence of transgenic RhoB or RhoA genes in the DNA samples. In brief, for each PCR reaction, 50 ng (`0.5-1 l) of DNA was mixed with 1.5 l dNTPs mixture, 5.0 l of 10x buffer, 2 unit Taq enzyme, and 100 ng of 5 and 3 primer each, the final volume of the reaction was supplemented to 50 l with autoclaved DD water. The PCR reactions were started at 94 degree for 3 minutes, then 50 degree for annealing for 1min, then 72 degree for 2 min for elongation. The PCRs were cycled for 35 cycles and the PCR products were run in 1.3% agar ose gels and the picture taken. Protein Preparation and Analysis Tissue from tumor resection dissect ion was homogenized in a lysis buffer containing 20 mM Tris-HCl (pH 7. 5), 150 mM NaCl, 1% NP-40, 1 mM phenylmethylsulfonyl fluoride, 1.5 g each of aprotinin and leupeptin per ml, 10 mM NaF, and 10 mM NaPPi. T he crude lysates were placed on ice for 25 min and vortexed every 5 min and finally spun at 13000 rpm for 15 min. The cleared
159 lysates were stored at -80 degree until Western Blo tting analysis. 50 g of the lysates was loaded into SDS-PAGE gel and analyzed for each sample. Antigenbound antibody was detected by enhanced chemiluminescence Western blotting kit (Amersham Pharmacia Biotech, Piscatawa y, New Jersey). Anti-HA antiboby (12AC5) was purchased from Roche. Monoclonal antibody to -actin was obtained from Sigma. Tumor Onset, Growth Rate, and Multiplicity Calculation F1 generations of ErbB2, ErbB2/RhoA and ErbB2/RhoB mice were backcrossed to generate F2 litters. The F2 generation mice were used to study the tumor onset of each strain. Mice were physically checked for new tumors by palpitation of the mammary fat pads 3-4 days per week by vivarium technicians as well as laboratory staff. Once mice formed tumors the onset dat e was noted and tumor measurements were taken every other day to determine the growth rates. Tumors were measured by digital Vern ier caliper, and volume was calculated using the formula V = W 2 x L, where width is the la rgest diameter measurements and length is the smaller diameter meas urement perpendicular to the width. Tumors were allowed to grow until the largest diameter reached the predetermined size of 2 centim eters. At this time t he mice were sacrificed and tumor tissue was harvested. Tumor growth rates were determi ned by subtracting the volume at time of sacrifice from t he first measurable volume and dividing by the number of days between the measurement s. Multiplicity was determined by
160 counting the number of distinct tumors at time of sacrifice. Graph Pad software was used for graphical representation and statistical analysis. Tumor onset data was analyzed using a log-ranked (Mant el-Cox) test comparing ErbB2 and ErbB2/RhoA or ErbB2 and ErbB2/RhoB mice. Students t-test was used to determine statistical significance compari ng growth rates and multiplicity for the three groups.
161 Results Generation of DNA constructs for the creation of MMTV-HA-RhoB and MMTV-HA-RhoA transgenic mice DNA sequences for the MMTV prom oter, HA epitope tagged human RhoB and HA tagged RhoA were sourced fr om vectors containing MMTV-TGF and pcDNA3 vectors containing HA-RhoB and HA-RhoA as de scribed in det ail in the Materials and Methods. Figure 39 demonstr ates a schematic representation for generation of the constructs where RhoA and RhoB in serts were removed from pcDNA vectors and subclon ed into the linearized MMTV vector in which the TGF gene was removed. The insertion of RhoA into the MMTV vector was verified by subsequent removal of the HA-RhoA insert and identification by agarose gel electrophoresis, as described in Material s and Methods. However, the insertion of RhoB could not be verifi ed by this method. Ultimately, the HARhoB insert along with the MMTV promot er was removed from the MMTV vector and subcloned back into the pcDNA3 vector The insertion of MMTV-RhoB was then confirmed by excision and identification by agar ose gel electrophoresis. The MMTV-HA-RhoA and MMTV-HA-RhoB tr ansgenes were removed from their plasmid vectors and submitted to the Mouse Models Core facility at Moffitt
162 Cancer Center for injection into FVB/N mouse 0.5 days post coitum zygotes, and implanted into the pseudopregnant CDI foster recipients for development.
RhoBFigure 39. Schematic for generation of MMTV-RhoB and MMTV-RhoA constructs for generation of transgenic mice that over express human HA-RhoB and HA-RhoA under the MMTV promoter pcDNA3 pcDNA3 RhoB EcoRI RhoB RhoB MMTV Promoter EcoRI EcoRI RhoARhoA MMTV-HA-RhoB RhoB MMTV-HA-RhoB 621 bp612 bp EcoRIEcoRI RhoB RhoB Linearized pcDNA3Linearized MMTVKpnI XhoI EcoRIEcoRIEcoRI MMTV-HA-RhoB in pcDNA3 RhoB Mice over expressinghuman MMTV-HA-RhoBMice over expressinghuman MMTV-HA-RhoA RhoA RhoA 163
164 Microinjection of MMTV-HA-RhoB and MMTV-HA-RhoA inserts into mouse zygotes Fertilized eggs (oocytes) were harvest ed from the reproductive organs of superovulated female FVB/N mice as de scribed in Materials and Methods. The DNA inserts containing MMTV-HA -RhoB and MMTV-HA-RhoA were microinjected into approximately 180 zy gots each. Which we re then incubated for 24-48 hours and, once cell divisi on occurred, 20-30 eggs each were subsequently implanted into the oviduc ts of 6 pseudopregnant CDI mice per transgene (12 mice total). Females impl anted with the RhoB-injected embryos gave birth to 20 pups, while the females implanted with the RhoA-injected embryos give birth to 8 pups. PCR and Southern blot genotyping r eveals 6 MMTV-HA-RhoB founders and 2 MMTV-HA-RhoA founders PCR and Southern blots on DNA from tail snips were used to confirm the integration of the transgene into the host mouse geno me. Six MMTV-HA-RhoB founders and 2 MMTV-HA-RhoA founders we re confirmed, and the founders were then backcrossed with wild type mice to generate F1 colonies of RhoB and RhoA heterozygous mice. Littermates fr om the F1 generation were intercrossed to increase the portion of transgenic mice and potentially generate homozygous RhoB and RhoA transgenic F2 colonies. The genotyping of the pups from the F1
165 and F2 generations was done by PCR as well as slot blot analysis as detailed in Materials and Methods. Furthermore, R hoB and RhoA transgenic mice from the F2 generations were crossed with mice th at over express the receptor tyrosine kinase ErbB2 under the MMTV promoter to study the a ffects of RhoB and RhoA overexpression in the mammary fa t pads on ErbB2-mediated breast tumorigenesis. Overexpression of RhoB, but not RhoA delays ErbB2-mediated mammary tumor onset ErbB2, the Human Epidermal Growth Factor Receptor-2, is the protein product of the ErbB2 gene. Er bB2 is found over expre ssed in 25-30% of breast and ovarian cancers, and c onfers poor prognosis clinic ally(195). Transgenic mouse models of breast cancer have bee n engineered using overexpression of both mutated and wild type ErbB2 genes in mammary epithelial cells driven by the mouse mammary tumor virus (MMT V) promoter(196, 197). The exact signaling events leading to transformati on downstream of ErbB2 overexpression are still under invest igation, however, signaling through Ras and PI3K are likely mediators(89). The prot ooncogenic transgenic model of breast tumor formation driven by overexpression of ErbB2 has been well c haracterized and studied since the early 1990s. With this model, female mice produce spontaneous breast tumors around the age of 250 days, which is far less aggressive than the mutated Neu transgenic model which de velops multiple tumors around 80
166 days(196, 198, 199). In this Chapter of the thesis we describe work where we generated transgenic mice that over express HA-Rho A and HA-RhoB under the control of the mouse mammary tumor vi rus promoter, and crossed these mice with transgenic mice that over ex press the protooncogenic ErbB2 gene under MMTV promoter control. These bitrans genic mice were used to determine the tumor effects of the small GTPases in-vivo on ErbB2 breast oncogenesis. Figure 40 shows the percent tumor free animals for the ErbB2 (EE) monotransgenic, ErbB2/RhoB (EB) and ErbB 2/RhoA (EA) bitransgenic mice. Likewise, Table 7 reports tumor onset, growth rates and multiplicity for EE, EA and EB mice with statistical analysis. In accordance with previously reported data from the liter ature, the EE females averaged 246.5 4.6 days to tumor onset. Overexpression of RhoA along wi th ErbB2 neither significantly enhanced nor suppressed tumor onset, with EA females developing tumors at approximately 261 9.1 days. Log-rank statistical analysis was used to determine significance of the entire curv e of EE compared to EA mice, which gave a non-significant p valu e of 0.514. Hazard rati o analysis determined that EA mice were only 1.14 times more likely to have a delayed tumor onset event compared to EE mice. However, there wa s a statistically significant delay in tumor onset when RhoB was over expressed in concert with ErbB2. The average tumor onset for EB females was 306 10.4 days compared to 246.5 4.6 for the EE females. The Log-ranked statistical output gave a significant difference between the two groups with a p value of 0.0003. Hazard ratio analysis showed EB females to be over tw ice as likely to have a delayed tumor
167 onset over the EE females. The tumor gr owth rates for all 3 groups showed no statistically significant difference, wi th EE mice growing at 89 10.7 cubic millimeters per day, and EA and EB mice gr owing at 75.1 9.7 (p = 0.16) and 74.4 7.7 (p = 0.13) cubic millimeters per day, respectively. Finally, tumor multiplicity was calculated to determine if overexpression of RhoA or RhoB affected the number of tumors that developed per mouse. Mice over expressing ErbB2 and ErbB2/RhoA dev eloped 2.1 0.22 and 2.1 0.3 tumors per mouse, respectively, while mice over expressing ErbB2/RhoB developed a significantly lower 1.5 0.15 tumors per mouse (p = 0.018). Students T-test was used to determine significance between growth rates as well as multiplicity.
0 100 200 300 400 500 0 50 100 Pe r c e n t su r v i v a l % Tumor Free Mice(ErbB2 alone)(ErbB2/RhoA)(ErbB2/RhoB)(ErbB2 alone)(ErbB2/RhoA)(ErbB2/RhoB)EE EA EB Figure 40. RhoB, but not RhoA over expression results in a statistically significant delay in ErbB2-mediated tumor onset Figure 40. Tumor onset dates for bi-transgenic mice over expressing ErbB2/RhoB (EB) or ErbB2/RhoA (EA) were compared to tumor onsetsfor mono-transgenic ErbB2 mice as described in Materials and Methods. Mouse Age (Days) 168
Table 7. RhoB, but not RhoA over expression results in a significant delay in ErbB2-mediated tumor onset and multiplicity, but not tumor growth rateTable 7. Tumor onset, growth rate and multiplicity for bi-transgenic mice over expressing ErbB2/RhoB (EB) or ErbB2/RhoA (EA) were compared to tumor onsets for mono-transgenic ErbB2 (EE) mice. Kaplan-Meyer survival curves were generated using graph pad software. Log-ranked and student T-tests were used for statistical analysis.2.1 0.301.5 0.152.1 0.22Tumor Multiplicity0.450.018-P Value0.130.16-P Value1.142.022-Hazard Ratio74.4 7.775.1 9.789.0 10.7Tumor Growth Rate0.5141261.0 9.1ErbB2/RhoA(n=28)0.0003306.0 10.4ErbB2/RhoB(n=17)-246.5 4.6ErbB2(n=127)P ValueTumor Onset(Days)Mouse Strain 2.1 0.301.5 0.152.1 0.22Tumor Multiplicity0.450.018-P Value0.130.16-P Value1.142.022-Hazard Ratio74.4 7.775.1 9.789.0 10.7Tumor Growth Rate0.5141261.0 9.1ErbB2/RhoA(n=28)0.0003306.0 10.4ErbB2/RhoB(n=17)-246.5 4.6ErbB2(n=127)P ValueTumor Onset(Days)Mouse Strain 169
170 Taken together, these data demonstrate a tumor suppre ssive role for RhoB, in that RhoB overexpression resulted in a significant delay in tumor onset and decreased the overall tumor multiplicity. In contrast, RhoA had no significant effect on ErbB2 oncogenesis. However, it is not known to what extent RhoB is over expressed in the tumor tissue, or which signaling events RhoB affects to implement these tumor suppressive effect s. Tumor tissue was harvested after development to validate the expression of RhoB and RhoA. Figure 41 shows the western blot analysis of lysates from EB, EA and EE tumors. Exogenous RhoA and RhoB are tagged with an HA epitope for identification, wher eas the EE mice do not express the HA epitope tag. All tu mors aside from the one from mouse, number EB24, demonstrated HA-RhoB expr ession to some degree. Figure 41 also shows that all tumors from HA -RhoA transgenic mice expressed HA-RhoA, and that the HA antibody wa s specific to exogenous RhoB and RhoA as none of the tumors from EE mice demonstrated HA expression.
2 16 22 68 84 112 116HA-actin2 14 16 20 24 30 32 34 36 38 48 56 58 84 7276 80 Figure 41. Tumors from EB and EA transgenic mice express RhoB and RhoA as determined by detection of HA by western blot ErbB2/RhoB Tumor LysateFigure 41. Tumors were resectedfrom ErbB2/RhoB, ErbB2/RhoA and ErbB2 transgenic mice once the maximum tumor size was reached according to IACUC guidelines. Tissue was snap frozen upon collection, homogenized in lysisbuffer and the protein was separated by 10 % PAGE, as describedin Materials and Methods. 4 6 16 24 26 32 38 116 42 48 62 64 66 68 72 74 80 98HA-actin ErbB2/RhoA Tumor Lysate 104 108 110 112 116 120 128 152 160 162HA-actin ErbB2/RhoA Tumor Lysate ErbB2 Tumor Lysate MDA-MB-468GGTI-2417 -+RhoB 171
172 Discussion Rho proteins in the RhoA-like fam ily are highly identical and were once thought to be evolutionar ily redundant because they share many cellular functions(155, 174). However, recent ly RhoA and RhoB have been shown to also have divergent roles in the cont ext of oncogenesis(16, 21, 22, 168). The role of RhoA in cell cycle progressi on, malignant transformation and tumor invasiveness is well established(171, 174, 186). These traits demonstrate physiological necessities for cellular tr ansformation. RhoA has also been associated with STAT3 and STAT5, which leads to increased tumor cell survival, cell proliferation, cell migration and epithelial to mesenchymal transition (EMT)(184, 185). RhoB, on the other hand, has recently been shown to suppress oncogenic events th at lead to tumorigenesi s. Supporting evidence comes from data that shows RhoB ex pression is decreased by upstream oncogenic signaling through EGFR, ErbB2 and Ras, which is mediated through the PI3K/Akt pathway(15, 16, 200). Furthermore, ec topic expression of RhoB has been shown to inhibit transformati on induced by H-Ras and PI3K, overcome resistance to chemotherapeutic induced apoptosis and bl ock cell migration and invasion(15, 175, 181). Likewise, ectopi c expression of RhoB significantly inhibited metastasis in the B16/ F10 mouse melanoma animal model, and
173 transgenic mice lacking RhoB showed a hi gher propensity for chemically induced skin carcinogenesis(16). On the contrary RhoA did not have tumor suppressive activity in any of these events compared to RhoB. Correlations in the clinic have been made in head and neck as well as lung cancers that demonstrate loss of RhoB expression as the tumors became mo re aggressive, from carcinoma in-situ to highly infiltrating carcinoma(22, 23). To further elucidate the anti-cancer properties of RhoB we set out to determine the effects of overexpre ssion of both RhoA and RhoB on ErbB2 mediated breast neoplasia. By utilizing tr ansgenic mouse models we bridge the gap of data between ectopic overexpression in-vitro, and the correlative evidence from the studies done in human cancer patients. This strategy allows for examination of overexpression of these Rho family members in a physiologically relevant setting for breast cancer by pr oducing tissue-specific expression in the mammary fat pads. By crossing these mice with commercially available ErbB2 transgenic mice we determined that over expression of RhoB, but not RhoA, led to a significant delay in the tumor onset mediated by ErbB2. RhoA, on the other hand, did not affect the onset of ErbB2 mediated mammary tumors. These results were not surprising, however, because RhoA is downstream of ErbB2, constitutively expressed and most likely is already activated by ErbB2 in these transgenic mice. Therefore, it is likely t hat the increased RhoA expression in this system is redundant to ErB2 signaling and cannot add to transforming effects. RhoB on the other hand, is not constitu tively expressed and its expression is downregulated by ErbB2 in cu ltured cells. Therefore, forced expression of RhoB
174 could antagonize ErbB2 oncoge nesis in this model. This is consistent with our work in cell culture that showed that ectopic expressi on of RhoB blocked ErbB2 transformation. RhoB expression in the mammary fat pads also reduced the number of tumors that developed per mouse. This suggests that RhoB not only delayed the tumorigenic events, but also reduced the oncogenic potency of ErbB2 signaling creat ing a higher threshold for trans formation leading to fewer tumor sites in the mammary fat pads. However, RhoB did not affect tumor growth, suggesting that once the ErbB2 oncogenic signaling overcame the tumor suppressive threshold of the over expre ssed RhoB the tumors grew at the same rate. Similarly, overexpr ession of RhoA did not add to the ability of ErbB2 to induce tumor growth in multip le sites, or increase the gr owth rates of the tumors once established. Once again, this is most likely due to redundancy in the signaling between ErbB2 and RhoA. The mechanism by which RhoB delays tumor onset and multiplicity is not known. We hypothesize that this is most likely due to RhoB antagonizing ErbB2 signali ng such as those mediated by Ras, PI3K, Akt, Erk, and STAT3. These studi es should be done in the mammary fat pads prior to tumor onset as well as early on while the tumors are fairly small. This would allow for the comparison of the effects of trans gene expression on signaling in the fat pads ju st prior to, or during the transition from normal ductal epithelial cells to hyperplasia or neo-pl asia. Furthermore, future work should investigate the role of RhoB and RhoA on normal breast morphogenesis. Future analysis may be done on the fat pads of RhoA and RhoB transgenic mice by comparing timed matings during the proliferative, differentiation, and involution
175 time points. Here, P-Akt, P-Erk1/2, and P-STAT3 signaling along with HA, RhoB and RhoA expression can be analyz ed by both western blot and immunohistochemistry. These data both support the recent findi ngs of the divergent roles of RhoA and RhoB in cancer, as well as bring up additional ques tions regarding signaling events in which RhoB antagonizes, while RhoA potentiates tumorigenesis. Future work with this model and the tumors that arise shoul d focus on signaling pathways that are responsible for the anti-neoplastic role of RhoB. Furthermore, the physiological consequences of RhoA and RhoB ex pression in the mammary fat pads should be analyzed during the natural progression of fat pads throughout pregnancy. To this end, timed matings should be done to analyze fat pads harvested from R hoA and RhoB mice during the proliferative per iod of early pregnancy. Likewise, fat pads should be harvested from mice at day 18 of pregnancy to determine if RhoA or RhoB expression alte rs the differentiation of the ductal epithelial cells that develop gla ndular structures fo r milk production. Finally, fat pads should be harvested from mice post pregnancy, after weaning of the pups, to determine the effects of R hoA and RhoB expression on involutioninduced apoptosis of the ductal epithelial cells. Taken together, this data clearly demonstrate that RhoB suppresses Er bB2-mediated mammary oncogenesis by delaying the tumor onset and by decreasin g tumor multiplicity. These studies further validate RhoB tumor suppr essive activity and enhanced our understanding of RhoBs involvement in the regulation of tumorigenesis.
176 Conclusions and Future Directions In an attempt to discover novel inhi bitors of the Raf/Mek/Erk1/2 kinase cascade we identified pirbuterol, a known 2 selective adrenergic receptor agonist(150), as an activator of a tumor suppressive pathway that was poorly characterized in cancer. In Chapter 1 of the thesis we dem onstrated that in MDA-MB-231 (breast), ACHN (renal) and SF-539 (CNS) cancer cells, stimulation of the 2 AR with ARA-211 (pirbuterol) l ed to the production of cAMP and activation of PKA, which resulted in in hibition of C-Raf, but not B-Raf, kinase activity. By blocking C-Raf, and subsequ ently Mek kinase activity, the activation of Erk1/2 proteins were inhibited, correlating with inhibition of anchoragedependent and independent cell growth, i nduction of apoptosis, and complete inhibition of the growth of MDA-MB -231 and ACHN xenografts in nude mice. Through the use of pharmacological inhi bitors, and genetic manipulation, we found that the inhibition of C-Raf/Mek /Erk1/2 signaling wa s dependent on cAMPmediated activation of PKA, but not cAMP activation of the guanine nucleotide exchange factors EPAC 1 and 2. Furt hermore, we demonstrated that the inhibition of cell proliferation in MDA-MB-231 cells was dependent upon the ability of ARA-211 to bind 2 AR and to inhibit Mek1/2 activity.
177 Prior to our studies there was ev idence that cAMP can control the proliferation of nor mal cells through regu lation of the Raf/Mek/Erk1/2 pathway. However, this regulation has been shown to be cell type specific, and can result in either activation or inhibition of Ra f/Mek/Erk signaling(201). For example, in adipocytes, endothelial cells, NIH 3T3 cell s, rat fibroblasts, smooth muscle cells, hepatocytes and pancreatic acinar cells the stimulation of cAMP results in inhibition of Erk1/2 activation and dec reased cell proliferation(40, 48-56). However, in rat thyroid cells, bone cells, polycystic kidney epithelium, Sertoli cells, cardiac myocytes, granulosa cells, pre-adiposites, pituitary cells and PC12 cells the stimulation of cAMP results in the activation of Erk1/2 resulting in differentiation or stimulati on of proliferation( 39-46). As discussed previously, the physiological effects of cAMP stimulation have been well characterized in normal cells. The effects of cAMP stimulation has only been studied in a few cancer cells, and most importantly unt il this thesis work, ther e has never been a succinct report detailing the signaling mechanisms by which cAMP affects cancer cells. For instance, two reports demonstrated that in MDA-MB-231 (breast), HL-60 (leukemia) and SH-CY5Y (neuroblastoma) cancer cells stimulation of cAMP results in inhibition of DNA synthesis, but no mechanistic sig naling studies were done(66, 67). Inconsistent with this, other reports demonstrate that activation of AR results in stimulation of growth in cell lines derived from pancreatic ductal carcinoma, as well as pulmonary ade nocarcinoma, through activation of arachidonic acid(202-204). Fu rthermore, the effects or 2 AR stimulation of
178 cAMP production on anchorage-dependent and independent growth, apoptosis and growth of human tumors in nude mice are not known. In this thesis we demonstrated that stimulation of 2 AR inhibits anchorage-dependent and independent growth, induces apoptosis and induces regression of human tumors in nude mice. In normal cells, PKA-mediated inhibition of C-Raf kinase activi ty has been demonstrated, although the mechanism by which this occurs is still controversial in the literature for normal cells, and lacking for cancer cells. As mentioned earlier in the discussion of Chapter 1, we have dem onstrated the dependency of the crosstalk between cAMP and Raf/Mek/Erk1/2 si gnaling on PKA-medi ated inhibition of C-Raf kinase activity, but were unable to pinpoint the mechanism by which this occurs. It has been suggested in the literature that PK A can directly phosp horylate C-Raf on inhibitory serine residues 43 and 259, ho wever, we were unable to detect any changes in phosphorylati on of these residues by west ern blot analysis. Further studies are warranted to this end, and should include phosphopeptide mapping of C-Raf in the presence or absence of ARA-211 to determine the major amino acid residues phosphorylated by PKA. Li kewise, the binding partners of C-Raf in the presence or absence of ARA-211 should be analyzed by a proteomic approach to determine if PKA is activating an inhibitory protein that binds C-Raf and sequesters it in an inactive form. These results would be important in the determination of the mechanism by which PKA inhibits C-Raf activity, and could lead to the design of better therapeutic ap proaches to inhibit C-Raf for cancer therapy.
179 The studies herein provide significant insight into the tumor suppressive activity of the 2 AR in cancer cells, resultin g in inhibition of anchoragedependent and independe nt growth, induction of apoptosis and complete inhibition of tumor xenogra ft growth. However, from this work we have not demonstrated how the cAMP/PKA-dependent suppression of P-Erk1/2 levels results in apoptosis. There are studies in the literature that show that inhibition of P-Erk1/2 can lead to apoptosis in cancer cells through modulation of mitochondrial-associated proapo ptotic proteins such as Bim and Bad(136, 139). However, further studies are warranted to determine if t he afore mentioned Erk1/2 effectors mediate the apoptosis indu ced via stimulation of cAMP, or if the apoptosis events are independent of Erk1/2 activity. To this end, constitutively activated Mek should be overexpressed followed by ARA-2 11 treatment to determine if apoptosis is rescued when Er k1/2 cannot be inhibi ted by ARA-211. Similarly, PKA could be inhibited by H89 or siRNA prior to ARA-211 treatment to determine if apoptosis induction is mediat ed by cAMP/PKA dependent signaling. Also, effectors downstream of Erk1/2 su ch as Bim and Bad should be knocked down, and ARA-211-mediated apoptosis should be exami ned to determine if the ablation results in a re scue from apoptosis. In Chapter 2 of the thesis we dem onstrate from analysis of both tumor tissue array and fresh human tumor biopsies that 2 AR and P-Erk1/2 are expressed in approximately 25% of the tumors tested. This data, combined with the preclinical data in Chapt er 1, led us to believe t hat ARA-211 treatment was a realistic goal for cancer therapy. Furt hermore, the fact t hat pirbuterol was
180 previously tested clinically in patient s with asthma and congestive heart failure will allow for an easier transition to anticancer clinical testing since the safe doses in humans is already known. Ho wever, since in normal cells cAMP has been shown to either stimulate of inhibi t proliferation, we investigated the response of fresh tumor biops ies to ARA-211 treatment ex-vivo. This method of testing ARA-211 in fresh biopsies, unfort unately, yielded highly variable results, which in some cases was masked by necrosis. The detection of 2 AR and PErk1/2 proteins proved to be problematic as well, as the cellular localization and staining percentages were not as expected. Much of the 2 AR staining was found in the cytoplasm or nucleus, but was expected to be in the plasma membrane. Similarly, P-Erk1/2 staining was localized to the nucleus and cytoplasm, which was expected, but t he % positive staining was lower than expected. After consultati on with a molecular pathologi st it was determined that the antibody concentrations for 2 AR and P-Erk1/2 were not correctly optimized. The 2 AR antibody concentration was mo st likely too high exhibiting nonspecific staining, and the P-Erk1/2 antibody was most likely too dilute resulting in false negatives for tissues expected to expr ess high levels of P-Erk1/2, such as late stage breast and pancreatic tumors. As ide from optimizing the IHC staining procedures, western blot analysis of protein levels could be used to determine expression of 2 AR and P-Erk1/2. However, th is method would not account for the tumor stroma that was captured with the tumor duri ng the biopsy. This could result in 2 AR and P-Erk1/2 expression detection from blood vessels or fibroblasts that are not actually indicative of the expre ssion levels in the tumor.
181 Ultimately, the best method for the detection of the 2 AR would be laser capture microdissection (LCM) of tumor tissue, followed by RT-PCR. Furthermore, microarray analysis of LCM from fresh human tumor biopsy treated ex-vivo, with and without ARA-211, could be done to dist inguish the effects of the treatment on global gene expression. Another approach to determini ng the efficacy of ARA211 on human tumors would be to implant sm all pieces of the fresh tumor tissues in athymic, nude mice, and treat the mice with ARA-211, as discussed earlier. This method would not be practical for all patient biopsies, but would serve as a good method to determine proof of concept for a human clinical trial. Ultimately, these studies would provide validation of the tumor suppressive effects of 2 AR signaling in human cancer. Chapter 3 of the thesis focuses on validating the tumor suppressive effects of the small GTPase RhoB We set out to show that, in-vivo, RhoB can suppress the tumorigenic effects of upstream signaling through the ErbB2 receptor. As discussed earlier, the ErbB family of receptors is found overexpressed or mutated in many human cancers, such as lung, breast, head and neck, and bladder(205). In-vitro data from the S ebti lab and others demonstrated the ability of RhoB to abr ogate transformation induced by H-Ras and PI3K. Studies also demonstrate th at RhoB expression is decreased by oncogenic signaling through EGFR, ErbB2 and H-Ras. Likewise, IHC studies in paraffin-embedded tumor specimens from patients with lung and head and neck cancer shows that RhoB expression is lo st as cancers progress from hyperplastic to deeply invasive carcinoma. However, data supporting the tumor suppressive
182 effects of RhoB in-vivo is lacking. Therefore, we used a physiologically relevant transgenic animal model to test the tumor suppressive effects of RhoB compared to its close family member RhoA. Mi ce were generated that express RhoB or RhoA in the mammary fat pads under the control of the MMTV promoter. These mice were then crossed with mice that overexpress the ErbB2 receptor under the same promoter control to create Erb2/RhoB and ErbB2/RhoA bitransgenic mice. ErbB2 overexpressing mice are known to develop spontaneous breast tumors at approximately 250 days(199). By introduc ing human RhoB in the mammary fat pads, the tumor onset mediated by ErbB2 was delayed by approximately 60 days. This data was statistically signifi cant with a p value of 0.0003 as measured by a log-ranked statistical test. We hypot hesized that this delay was due to the ability of RhoB to antagonize ErbB2 signa ling through Ras, PI3K Akt, Erk, and other well-known downstream ErbB2 si gnals. However, gathering data to support this hypothesis was difficult as the tumors were allowed to grow to maximum capacity to determine the effects of RhoB on tumor gr owth rate. It is possible that once the tumors were har vested they had already overcome the suppressed signaling from RhoB. To test this hypothesis, fat pads would need to be tested in a time course study to det ermine if RhoB expression resulted in inhibition of P-Akt, P-Erk1/2 or P-STAT 3 during the transition from normal to transformed tissue. Although the tumor onset was delayed breast tumors eventually developed, and the tumor growth rate was not significantly different. This suggests that RhoB acts as a gatekeeper for transformation early on. However, as the tumors accumulate mo re oncogenic alterations the RhoB tumor
183 suppressive effects can be overcome. This may occur through further genetic alterations that surpass the regulation of RhoB, or signaling events that could eventually lead to expression of negative regulators of RhoB that re sult in faster degradation, or protein tu rnover. Time course studies are needed for RhoB expression in normal, hyperplasic and transformed cells isolated from the fat pads of ErbB2/RhoB bitransgenic mice. These studies could determine if RhoB is suppressed at the transcriptional level by methylation or acetylation during the transformation of the mammary ductal epi thelial cells, although silencing at the promoter level is not likely since RhoB is under the co ntrol of the MMTV promoter in this case. However, a proteomic approa ch could allow for t he identification of novel proteins that bind and inactivate RhoB. Here, RhoB would be immunoprecipitated from fat pad lysate at different time points and isolated by native gel electrophoresis. Once separ ated, RhoB binding partners could be identified, cloned and tested in tissue culture for their abi lity to reverse the tumor suppressive effects of RhoB. Although RhoB overexpression did not change the growth rate of the tumors once formed, the multiplicity, or tumor number per mouse, was significantly decreased by 25% (p value = 0.018) by RhoB, but not RhoA. Once again, the means by which R hoB inhibited tumor multiplicity in this setting was not determined, but it suggests that overall the oncogenic potency of ErbB2 signaling was diminis hed to the point that trans formation was blocked in some of the mammary fat pad tissue. To gain insight to the mechanism by which RhoB decreases the oncogenic potential of ErbB2, microarray studies could be done comparing ErbB2 to ErbB2/RhoB and Er bB2/RhoA fat pads. In contrast to
184 RhoB, RhoA, which has been shown to promote tumorigenesis by inducing proliferation and contribut ing to invasion and metast asis, was found to have no effects on the tumor onset, growth rate or multiplicity of ErbB2-mediated oncogenesis in our transgenic mouse model. It is likely that t he increase in RhoA expression is redundant to ErbB2 ov erexpression and could not add to the transforming effects of ErbB 2. For example, ErbB2 is known to activate Ras, which in turn, can activate RhoA. Ther efore, overexpression of RhoA in this setting would not be expected to add to t he transforming activity of ErbB2. While the generation of RhoA and RhoB transgenic mice allowed for the investigation of the effects of these small G-pr oteins in ErbB2 breast oncogenesis, they will also be instrument al in determining their role in normal mammary morphogenesis. As described prev iously, future work will be done with RhoA and RhoB monotransgenic mice for timed mating experiments to analyze their effects on ductal lobular format ion, normal cell proliferation at day 7 pregnancy, differentiation at day 18 pregnancy and apoptosis at day 4 weaning. This would allow for the prec ise timed measuring of the effect of RhoA and RhoB on signaling events linked to the induction of apoptos is or inhibition of proliferation. In conclusion, this work adds signif icant value to our understanding of the tumor suppressive roles of the 2 AR and small GTPase RhoB. We now know that in cancer cells, PKA can mediate crosstalk between the 2 AR and CRaf/Mek/Erk1/2 kinase cascade resulti ng in inhibition of anchorage-dependent and independent tumor cell growth, i nduction of apoptosis and complete
185 inhibition of human tumor growth in nude mice. Here we have shown that the inhibition of proliferat ion is dependent on stimulat ing cAMP production and inhibiting Mek/Erk1/2 signa ling. However, translating this work to the clinic proved to be problematic. We determi ned that approximately 25% of tumors tested for 2 AR and P-Erk1/2 actually co -express these markers from IHC examining a tissue array and fresh human tumor biopsies from Moffitt Cancer Center, but the expression patterns alone do not predict response to ARA-211 therapy on tissue ex-vivo. Further testing of ex-vivo tumor treatment techniques is necessary in order to more accurately predict a patients tumor response to ARA-211. Finally, we provide data that supports the tumor suppressive effects of the small GTPase RhoB, but not RhoA, in a transgenic mouse model highly relevant to human breast cancer where ErbB2 drives oncogenesis. Future work should include studies designed to in terrogate the signaling mechanisms by which RhoB delays tumor onset and reduces tumor multiplicity.
186 List of References 1. Stern, D. F. Tyrosi ne kinase signalling in breast cancer: ErbB family receptor tyrosine kinase s. Breast Cancer Res, 2: 176-183, 2000. 2. Yarden, Y. Biology of HER2 and its importance in breast cancer. Oncology, 61 Suppl 2: 1-13, 2001. 3. Yarden, Y. and Sliwko wski, M. X. Untangling the ErbB signalling network. Nat Rev Mol Cell Biol, 2: 127-137, 2001. 4. Graus-Porta, D., Beerli, R. R., Daly J. M., and Hynes, N. E. ErbB-2, the preferred heterodimerization par tner of all ErbB recept ors, is a mediator of lateral signaling. Embo J, 16: 1647-1655, 1997. 5. Tzahar, E., Waterman, H., Chen, X., Levkowitz, G., Karunagaran, D., Lavi, S., Ratzkin, B. J., and Yarden, Y. A hierarchical network of interreceptor interactions determines signal tr ansduction by Neu differentiation factor/neuregulin and epidermal gr owth factor. Mo l Cell Biol, 16: 52765287, 1996. 6. Li, Q. and Loeb, J. A. N euregulin-heparan-sulf ate proteoglycan interactions produce sustained erbB receptor activation required for the induction of acetylcholine rec eptors in muscle. J Biol Chem, 276: 3806838075, 2001. 7. Brennan, P. J., Kumagai, T., Berezo v, A., Murali, R., and Greene, M. I. HER2/neu: mechanisms of dimeriza tion/oligomerizat ion. Oncogene, 19: 6093-6101, 2000. 8. Ross, J. S., Fletcher, J. A., Bloom, K. J., Linette, G. P., Stec, J., Clark, E., Ayers, M., Symmans, W. F., Pusztai, L., and Hortobagyi, G. N. HER-2/neu testing in breast cancer. Am J Clin Pathol, 120 Suppl: S53-71, 2003. 9. Bos, J. L. ras oncogenes in hu man cancer: a review. Cancer Res, 49: 4682-4689, 1989. 10. Varghese, H. J., Davidson, M. T., MacDonald, I. C., Wilson, S. M., Nadkarni, K. V., Groom, A. C., and Chambers, A. F. Activated ras regulates the prolifer ation/apoptosis balance and early survival of developing micrometastases. Cancer Res, 62: 887-891, 2002. 11. Sahai, E., Olson, M. F., and Marshall, C. J. Cross-talk between Ras and Rho signalling pathways in transform ation favours proliferation and increased motility. Embo J, 20: 755-766, 2001. 12. Zondag, G. C., Evers, E. E., ten Klooster, J. P., Janssen, L., van der Kammen, R. A., and Collard, J. G. Oncogenic Ras downregulates Rac activity, which leads to increased R ho activity and epithelial-mesenchymal transition. J Cell Biol, 149: 775-782, 2000.
187 13. Vial, E., Sahai, E., and Marshall, C. J. ERK-MAPK signaling coordinately regulates activity of Rac1 and RhoA for tumor cell motility. Cancer Cell, 4: 67-79, 2003. 14. Berenjeno, I. M., Nunez, F., and Bustel o, X. R. Transcrip tomal profiling of the cellular transformation induced by Rho subfamily GTPases. Oncogene, 26: 4295-4305, 2007. 15. Jiang, K., Delarue, F. L., and Sebti, S. M. EG FR, ErbB2 and Ras but not Src suppress RhoB expression while ectopic expression of RhoB antagonizes oncogene-mediated transformation. Oncogene, 23: 11361145, 2004. 16. Jiang, K., Sun, J., Cheng, J., Djeu, J. Y., Wei, S., and Sebti, S. Akt mediates Ras downregulat ion of RhoB, a suppressor of transformation, invasion, and metastasis. Mol Cell Biol, 24: 5565-5576, 2004. 17. Huang, M., Kamasani, U., and Prendergas t, G. C. RhoB facilitates c-Myc turnover by supporting efficient nuclear accumulation of GSK-3. Oncogene, 25: 1281-1289, 2006. 18. Huang, M. and Pr endergast, G. C. RhoB in cancer suppression. Histol Histopathol, 21: 213-218, 2006. 19. Kamasani, U., Liu, A. X., and Prendergast, G. C. Genetic response to farnesyltransferase inhibitors: proapoptot ic targets of RhoB. Cancer Biol Ther, 2: 273-280, 2003. 20. Liu, A. X., Rane, N., Liu, J. P., and Prendergast, G. C. RhoB is dispensable for mouse development, but it modifies susceptibility to tumor formation as well as cell adhesi on and growth factor signaling in transformed cells. Mol Cell Biol, 21: 6906-6912, 2001. 21. Liu, A., Cerniglia, G. J., Bernhard, E. J., and Prendergast, G. C. RhoB is required to mediate apoptos is in neoplastically transformed cells after DNA damage. Proc Natl Acad Sci U S A, 98: 6192-6197, 2001. 22. Adnane, J., Muro-Cac ho, C., Mathews, L., S ebti, S. M., and MunozAntonia, T. Suppression of rho B expr ession in invasive carcinoma from head and neck cancer patients. Clin Cancer Res, 8: 2225-2232, 2002. 23. Mazieres, J., Antonia, T., Daste, G., Muro-Cacho, C., Berchery, D., Tillement, V., Pradines, A., Sebti, S., and Favre, G. Loss of RhoB expression in human lung cancer progression. Clin Cancer Res, 10: 27422750, 2004. 24. Revillion, F., Bonneterre, J., and Pe yrat, J. P. ERBB2 oncogene in human breast cancer and its clinical significance. Eur J Cancer, 34: 791-808, 1998. 25. Benovic, J. L., Shorr, R. G., Ca ron, M. G., and Lefkowitz, R. J. The mammalian beta 2-adrenergic receptor: purification and characterization. Biochemistry, 23: 4510-4518, 1984. 26. Bilezikian, J. P., Dornfeld, A. M. and Gammon, D. E. Structure-bindingactivity analysis of beta-adrenergic am ines--II. Binding to the beta receptor and inhibition of adenylate cycl ase. Biochem Pharmacol, 27: 1455-1461, 1978.
188 27. Bilezikian, J. P., Dornfeld, A. M. and Gammon, D. E. Structure-bindingactivity analysis of beta-adrenergic ami nes--I. Binding to the beta receptor and activation of adenylate cyclase. Biochem Pharmacol, 27: 1445-1454, 1978. 28. Lefkowitz, R. J. and Hamp, M. Com parison of specificity of agonist and antagonist radioligand binding to beta adrenergi c receptors. Nature, 268: 453-454, 1977. 29. Lefkowitz, R. J. and W illiams, L. T. Catecholamine binding to the betaadrenergic receptor. Proc Natl Acad Sci U S A, 74: 515-519, 1977. 30. Harden, T. K., Wolfe, B. B., and Molinoff, P. B. Binding of iodinated beta adrenergic antagonists to prot eins derived from rat heart. Mol Pharmacol, 12: 1-15, 1976. 31. Leichtling, B. H., Su, Y. F., Wimalasena, J., Har den, T. K., Wolfe, B. B., and Wicks, W. D. Studies of cAMP metabolism in cultured hepatoma cells: presence of functional adenylate cyclas e despite low cAMP content and lack of hormonal responsiveness. J Cell Physiol, 96: 215-223, 1978. 32. Cho, Y. J., Kim, J. Y., Jeong, S. W ., Lee, S. B., and Kim, O. N. Cyclic AMP induces activation of extracellular si gnal-regulated kinase s in HL-60 cells: role in cAMP-induced differentiation. Leuk Res, 27: 51-56, 2003. 33. Vossler, M. R., Yao, H., York, R. D., Pan, M. G., Rim, C. S., and Stork, P. J. cAMP activates MAP kinase and Elk-1 through a B-Rafand Rap1dependent pathway. Cell, 89: 73-82, 1997. 34. Stork, P. J. and Schm itt, J. M. Crosstalk between cAMP and MAP kinase signaling in the regulation of cell proliferation. Trends Cell Biol, 12: 258266, 2002. 35. Maudsley, S., Pierce, K. L., Zamah, A. M., Miller, W. E., Ahn, S., Daaka, Y., Lefkowitz, R. J., and Luttrell, L. M. The beta(2)adrenergic receptor mediates extracellular signal-regulated ki nase activation via assembly of a multi-receptor complex with the epider mal growth factor receptor. J Biol Chem, 275: 9572-9580, 2000. 36. Enserink, J. M., Christensen, A. E ., de Rooij, J., van Triest, M., Schwede, F., Genieser, H. G., Doskeland, S. O., Blank, J. L., and Bos, J. L. A novel Epac-specific cAMP analogue demonstr ates independent regulation of Rap1 and ERK. Nat Cell Biol, 4: 901-906, 2002. 37. Rangarajan, S., Enserink, J. M., Kuiper ij, H. B., de Rooij, J., Price, L. S., Schwede, F., and Bos, J. L. Cyclic AMP induces int egrin-mediated cell adhesion through Epac and Rap1 upon stimulation of the beta 2adrenergic receptor. J Cell Biol, 160: 487-493, 2003. 38. Christensen, A. E., Selheim, F., de Rooij, J., Dremier, S., Schwede, F., Dao, K. K., Martinez, A., Maenhaut, C ., Bos, J. L., Genieser, H. G., and Doskeland, S. O. cAMP analog mapping of Epac1 and cAMP kinase. Discriminating analogs demonstrate that Epac and cAMP kinase act synergistically to promote PC-12 cell neurite extension. J Biol Chem, 278: 35394-35402, 2003.
189 39. Iacovelli, L., Capobianco, L., Salvator e, L., Sallese, M. D'Ancona, G. M., and De Blasi, A. Thyrotropin activates mitogenactivated protein kinase pathway in FRTL-5 by a cAMP-dependent protein ki nase A-independent mechanism. Mol Pharmacol, 60: 924-933, 2001. 40. Fujita, T., Meguro, T. Fukuyama, R., Nakamuta, H., and Koida, M. New signaling pathway for parathyroid hormone and cyclic AMP action on extracellular-regulated kinase and cell proliferation in bone cells. Checkpoint of modulation by cyclic AMP. J Biol Chem, 277: 22191-22200, 2002. 41. Yamaguchi, T., Pelling, J. C., Ramasw amy, N. T., Eppler J. W., Wallace, D. P., Nagao, S., Rome, L. A., Sulliv an, L. P., and Grantham, J. J. cAMP stimulates the in vitro proliferation of renal cyst epithelial cells by activating the extracellular signal-regulat ed kinase pathway. Kidney Int, 57: 14601471, 2000. 42. Deeble, P. D., Murphy, D. J., Parsons S. J., and Cox, M. E. Interleukin-6and cyclic AMP-mediated signali ng potentiates neuroendocrine differentiation of LNCaP prostate tumor cells. Mol Cell Biol, 21: 84718482, 2001. 43. Crepieux, P., Marion, S., Martinat, N., Fafeur, V., Vern, Y. L., Kerboeuf, D., Guillou, F., and Reiter, E. T he ERK-dependent signalling is stagespecifically modulated by FSH, duri ng primary Sertoli cell maturation. Oncogene, 20: 4696-4709, 2001. 44. Zou, Y., Komuro, I., Yamazaki, T ., Kudoh, S., Uozumi, H., Kadowaki, T., and Yazaki, Y. Both Gs and Gi prot eins are critica lly involved in isoproterenol-induced cardiomyo cyte hypertrophy. J Biol Chem, 274: 9760-9770, 1999. 45. Seger, R., Hanoch, T., Rosenberg, R., Dantes, A., Merz, W. E., Strauss, J. F., 3rd, and Amsterdam, A. The ERK signaling cascade inhibits gonadotropin-stimulated ster oidogenesis. J Biol Chem, 276: 13957-13964, 2001. 46. Greenberg, A. S., Shen, W. J., Muliro, K., Patel, S., Souza, S. C., Roth, R. A., and Kraemer, F. B. Stimulation of lipolysis and hormone-sensitive lipase via the extracellular signal-regulated kinase pathway. J Biol Chem, 276: 45456-45461, 2001. 47. Hattori, M. and Minato, N. Rap1 GTPase: functions, regulation, and malignancy. J Biochem, 134: 479-484, 2003. 48. Sevetson, B. R., Kong, X., and Lawrence, J. C., Jr. Increasing cAMP attenuates activation of mitogen-activat ed protein kinase. Proc Natl Acad Sci U S A, 90: 10305-10309, 1993. 49. D'Angelo, G., Lee, H., and Weiner, R. I. cAMP-dependent protein kinase inhibits the mitogenic action of va scular endothelial growth factor and fibroblast growth factor in capillary endothelial cells by blocking Raf activation. J Cell Biochem, 67: 353-366, 1997.
190 50. Schmitt, J. M. and Stork, P. J. C yclic AMP-mediated inhibition of cell growth requires the small G protein Rap1. Mol Cell Biol, 21: 3671-3683, 2001. 51. Chen, J. and Iyengar, R. Suppression of Ras-induced transformation of NIH 3T3 cells by activated G alpha s. Science, 263: 1278-1281, 1994. 52. Cook, S. J. and McCormick, F. Inhibition by cAMP of Ras-dependent activation of Raf. Science, 262: 1069-1072, 1993. 53. Wu, J., Dent, P., Jelinek, T., Wolfman, A., Weber, M. J., and Sturgill, T. W. Inhibition of the EGF-activated MAP kinase signaling pathway by adenosine 3',5'-monophosphate. Science, 262: 1065-1069, 1993. 54. Scott, P. H., Belham, C. M., al-Hafidh, J., Chilvers, E. R., Peacock, A. J., Gould, G. W., and Plevin, R. A regulatory role for cAMP in phosphatidylinositol 3-kinase/p70 ri bosomal S6 kinase-mediated DNA synthesis in plateletderived-growth-factor-stimulated bovine airway smooth-muscle cells. Biochem J, 318 ( Pt 3): 965-971, 1996. 55. Osinski, M. T. and Schror, K. Inhibi tion of platelet-der ived growth factorinduced mitogenesis by phosphodiesterase 3 inhibitors: role of protein kinase A in vascular smooth muscle cell mitogenesis. Biochem Pharmacol, 60: 381-387, 2000. 56. Thoresen, G. H., Johasen, E. J., and Ch ristoffersen, T. Effects of cAMP on ERK mitogen-activated protein kinas e activity in hepatocytes do not parallel the bidirectional regulation of DNA synthesis. Cell Biol Int, 23: 1320, 1999. 57. Tortora, G. and Ciardiello, F. Pr otein kinase A as target for novel integrated strategies of cancer therapy. Ann N Y Acad Sci, 968: 139-147, 2002. 58. Cho-Chung, Y. S., Nesterova, M., Be cker, K. G., Srivas tava, R., Park, Y. G., Lee, Y. N., Cho, Y. S., Kim, M. K., Neary, C., and Cheadle, C. Dissecting the circuitry of protein kinase A and cA MP signaling in cancer genesis: antisense, microarray, gene overexpression, and transcription factor decoy. Ann N Y Acad Sci, 968: 22-36, 2002. 59. Schmitt, J. M. and Stork, P. J. PKA phosphorylation of Src mediates cAMP's inhibition of cell growth via Rap1. Mol Cell, 9: 85-94, 2002. 60. Hafner, S., Adler, H. S., Mischak, H., Janosch, P., Heidecker, G., Wolfman, A., Pippig, S., Lohse, M., Ueffing, M., and Kolch, W. Mechanism of inhibition of Raf-1 by pr otein kinase A. Mol Cell Biol, 14: 6696-6703, 1994. 61. Graves, L. M., Bornfeldt, K. E., Rai nes, E. W., Potts, B. C., Macdonald, S. G., Ross, R., and Krebs, E. G. Prot ein kinase A ant agonizes plateletderived growth factor-induced signali ng by mitogen-activated protein kinase in human arterial smooth muscle cells. Proc Natl Acad Sci U S A, 90: 10300-10304, 1993.
191 62. Dumaz, N. and Marais, R. Integr ating signals between cAMP and the RAS/RAF/MEK/ERK signalling pathways. Based on the anniversary prize of the Gesellschaft fur Biochemie und Molekularbiologie Lecture delivered on 5 July 2003 at the Sp ecial FEBS Meeting in Brussels. Febs J, 272: 3491-3504, 2005. 63. English, J. M. and Cobb, M. H. Pharmacological inhibitors of MAPK pathways. Trends Pharmacol Sci, 23: 40-45, 2002. 64. Khosravi-Far, R., White, M. A ., Westwick, J. K., Solski, P. A., Chrzanowska-Wodnicka, M., Van Aelst, L. Wigler, M. H., and Der, C. J. Oncogenic Ras activation of Raf/mitogen-activated protein kinaseindependent pathways is sufficient to cause tumorigenic transformation. Mol Cell Biol, 16: 3923-3933, 1996. 65. Weinstein-Oppenheimer, C. R., Blalock, W. L., Steelman, L. S., Chang, F., and McCubrey, J. A. The Raf signal tr ansduction cascade as a target for chemotherapeutic intervention in gr owth factor-responsive tumors. Pharmacol Ther, 88: 229-279, 2000. 66. Kim, S. N., Ahn, Y. H., Kim, S. G., Park, S. D., Cho-Chung, Y. S., and Hong, S. H. 8-Cl-cAMP induces ce ll cycle-specific apoptosis in human cancer cells. Int J Cancer, 93: 33-41, 2001. 67. Slotkin, T. A., Zhang, J., Dancel, R., Garcia, S. J., Willis, C., and Seidler, F. J. Beta-adrenoceptor si gnaling and its control of cell replication in MDAMB-231 human breast cancer cells. Breast Cancer Res Treat, 60: 153166, 2000. 68. Hanahan, D. and Wei nberg, R. A. The hallm arks of cancer. Cell, 100: 5770, 2000. 69. Borg, A., Tandon, A. K., Sigurdsson, H., Clark, G. M., Ferno, M., Fuqua, S. A., Killander, D., and McGuire, W. L. HER-2/neu amplif ication predicts poor survival in node-positive breast cancer. Cancer Res, 50: 4332-4337, 1990. 70. Slamon, D. J., Clark, G. M., Wong, S. G., Levin, W. J., Ullrich, A., and McGuire, W. L. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science, 235: 177-182, 1987. 71. Johnson, G. L. and Lapadat, R. Mitogen-activated protein kinase pathways mediated by ERK, JNK, and p38 protein kinases. Science, 298: 1911-1912, 2002. 72. Ballif, B. A. and Blenis, J. Mo lecular mechanisms mediating mammalian mitogen-activated protein kinase (MAPK) kinase (MEK)-MAPK cell survival signals. Cell Growth Differ, 12: 397-408, 2001. 73. Hagemann, C. and Blank J. L. The ups and downs of MEK kinase interactions. Cell Signal, 13: 863-875, 2001. 74. Fresno Vara, J. A., Casado, E., de Cast ro, J., Cejas, P., Belda-Iniesta, C., and Gonzalez-Baron, M. PI3K/Akt si gnalling pathway and cancer. Cancer Treat Rev, 30: 193-204, 2004.
192 75. Alvarez, J. V. and Frank, D. A. Genome-wide analysis of STAT target genes: elucidating the mechanis m of STAT-mediated oncogenesis. Cancer Biol Ther, 3: 1045-1050, 2004. 76. Sebolt-Leopold, J. S. Development of anticancer drugs targeting the MAP kinase pathway. Oncogene, 19: 6594-6599, 2000. 77. Sebolt-Leopold, J. S., Dudley, D. T., Herrera, R., Van Becelaere, K., Wiland, A., Gowan, R. C., Tecle, H., Barrett, S. D., Bridges, A., Przybranowski, S., Leopold, W. R., and Salt iel, A. R. Blockade of the MAP kinase pathway suppresses growth of colon tumors in vivo. Nat Med, 5: 810-816, 1999. 78. Blaskovich, M. A., Sun, J., Cantor, A., Turkson, J., Jove, R., and Sebti, S. M. Discovery of JSI-124 (cucurbitaci n I), a selective J anus kinase/signal transducer and activator of transcripti on 3 signaling pathway inhibitor with potent antitumor activity against human and murine cancer cells in mice. Cancer Res, 63: 1270-1279, 2003. 79. Redell, M. S. and Tweardy, D. J. Targeting tr anscription factors for cancer therapy. Curr Pharm Des, 11: 2873-2887, 2005. 80. Hao, D. and Rowinsky, E. K. Inhibiting signa l transduction: recent advances in the development of receptor tyrosine kinase and Ras inhibitors. Cancer Invest, 20: 387-404, 2002. 81. Johnson, J. R., Cohen, M., Sridhar a, R., Chen, Y. F., Williams, G. M., Duan, J., Gobburu, J., Booth, B., Benson, K., Leighton, J., Hsieh, L. S., Chidambaram, N., Zimmerman, P., and Pa zdur, R. Approval summary for erlotinib for treatment of patients with locally adv anced or metastatic nonsmall cell lung cancer after failure of at least one prior chemotherapy regimen. Clin Cancer Res, 11: 6414-6421, 2005. 82. Sebti, S. M. and Hamilt on, A. D. Inhibition of Ras prenylation: a novel approach to cancer chemotherapy. Pharmacol Ther, 74: 103-114, 1997. 83. Shields, J. M., Pruitt, K., McFa ll, A., Shaub, A., and Der, C. J. Understanding Ras: 'it ain't over 'til it's over'. Trends Cell Biol, 10: 147-154, 2000. 84. Lehman, T. A., Reddel, R., Peiifer, A. M., Spillare, E., Kaighn, M. E., Weston, A., Gerwin, B. I., and Harris, C. C. Oncogenes and tumorsuppressor genes. Environ Health Perspect, 93: 133-134, 1991. 85. Lerner, E. C., Zhang, T. T., Knowles, D. B., Qian, Y., Ha milton, A. D., and Sebti, S. M. Inhibition of the prenylation of K-Ras, but not Hor N-Ras, is highly resistant to CAAX pept idomimetics and requires both a farnesyltransferase and a geranylgeranyltransferase I inhibitor in human tumor cell lines. Oncogene, 15: 1283-1288, 1997. 86. Barbacid, M. ras genes. Annu Rev Biochem, 56: 779-827, 1987. 87. Khosravi-Far, R. and Der, C. J. The Ras signal transduction pathway. Cancer Metastasis Rev, 13: 67-89, 1994.
193 88. Shukla, S., Maclennan, G. T., Marengo, S. R., Resnick, M. I., and Gupta, S. Constitutive activation of PI3K -Akt and NF-kappaB during prostate cancer progression in autochthonous transgenic mouse model. Prostate, 2005. 89. Lackey, K. E. Lessons from the dr ug discovery of lapatinib, a dual ErbB1/2 tyrosine kinase inhibitor. Curr Top Med Chem, 6: 435-460, 2006. 90. Barber, M. A. and Welc h, H. C. PI3K and RAC si gnalling in leukocyte and cancer cell migration. Bull Cancer, 93: E44-52, 2006. 91. Kallergi, G., Agelaki, S., Ma rkomanolaki, H., Georgoulias, V., and Stournaras, C. Activation of FAK/PI 3K/Rac1 signaling controls actin reorganization and inhibits cell motility in human cancer cells. Cell Physiol Biochem, 20: 977-986, 2007. 92. Kolsch, V., Charest, P. G., and Firtel, R. A. The regulation of cell motility and chemotaxis by phospholip id signaling. J Cell Sci, 121: 551-559, 2008. 93. Qian, Y., Corum, L., M eng, Q., Blenis, J., Zheng, J. Z., Shi, X., Flynn, D. C., and Jiang, B. H. PI 3K induced actin filament remodeling through Akt and p70S6K1: implication of essential ro le in cell migration. Am J Physiol Cell Physiol, 286: C153-163, 2004. 94. Falsetti, S. C., Wang, D. A., Peng, H ., Carrico, D., Cox, A. D., Der, C. J., Hamilton, A. D., and Sebt i, S. M. Geranylgeranylt ransferase I inhibitors target RalB to inhibit anchorag e-dependent growth and induce apoptosis and RalA to inhibit anchorage-independent growth. Mol Cell Biol, 27: 8003-8014, 2007. 95. Hamad, N. M., Elconi n, J. H., Karnoub, A. E., Bai, W., Rich, J. N., Abraham, R. T., Der, C. J., and Counter, C. M. Distinct requirements for Ras oncogenesis in human versus mouse cells. Genes Dev, 16: 20452057, 2002. 96. Lim, K. H., Baines, A. T ., Fiordalisi, J. J., Shipitsin, M., Feig, L. A., Cox, A. D., Der, C. J., and Counter, C. M. Acti vation of RalA is critical for Rasinduced tumorigenesis of human cells. Cancer Cell, 7: 533-545, 2005. 97. Lim, K. H., O'Hayer, K., Adam, S. J ., Kendall, S. D., Cam pbell, P. M., Der, C. J., and Counter, C. M. Divergent roles for RalA and RalB in malignant growth of human pancreatic carcinoma cells. Curr Biol, 16: 2385-2394, 2006. 98. Peterson, S. N., Trabalzini, L., Brtva, T. R., Fischer, T., Altschuler, D. L., Martelli, P., Lapetina, E. G., Der, C. J., and White, G. C., 2nd Identification of a novel RalGDS-related protein as a candidate effector for Ras and Rap1. J Biol Chem, 271: 29903-29908, 1996. 99. Gonzalez-Garcia, A., Pr itchard, C. A., Paterson, H. F., Mavria, G., Stamp, G., and Marshall, C. J. RalGDS is required for tumor formation in a model of skin carcinogenesis. Cancer Cell, 7: 219-226, 2005. 100. Roberts, P. J. and Der, C. J. Targeting the Raf-MEK-ERK mitogenactivated protein kinase cascade for the treatment of cancer. Oncogene, 26: 3291-3310, 2007.
194 101. Wellbrock, C., Karasarides, M., and Marais, R. The RAF proteins take centre stage. Nat Rev Mol Cell Biol, 5: 875-885, 2004. 102. Weston, C. R., Lambright, D. G., and Davis, R. J. Signal transduction. MAP kinase signaling spec ificity. Science, 296: 2345-2347, 2002. 103. Yoon, S. and Seger, R. The extrac ellular signal-regulated kinase: multiple substrates regulate diverse cellu lar functions. Growth Factors, 24: 21-44, 2006. 104. Zuber, J., Tchernitsa, O. I., Hinzm ann, B., Schmitz, A. C., Grips, M., Hellriegel, M., Sers, C., Rosenthal A., and Schafer, R. A genome-wide survey of RAS transformation targets. Nat Genet, 24: 144-152, 2000. 105. Brose, M. S., Volpe, P., Feldman, M., Kumar, M., Rishi, I., Gerrero, R., Einhorn, E., Herlyn, M., Minna, J., Nicholson, A., Roth, J. A., Albelda, S. M., Davies, H., Cox, C., Brignell, G., Stephens, P., Futreal, P. A., Wooster, R., Stratton, M. R., and Weber, B. L. BRAF and RAS mutations in human lung cancer and melanoma. Cancer Res, 62: 6997-7000, 2002. 106. Davies, H., Bignell, G. R., Cox, C., Stephens, P., Edkins, S., Clegg, S., Teague, J., Woffendin, H., Garnett, M. J., Bottomley, W., Davis, N., Dicks, E., Ewing, R., Floyd, Y., Gray, K., Hall, S., Hawes, R., Hughes, J., Kosmidou, V., Menzies, A ., Mould, C., Parker, A. Stevens, C., Watt, S., Hooper, S., Wilson, R., Jayatilake, H., Gusterson, B. A., Cooper, C., Shipley, J., Hargrave, D., Pritchar d-Jones, K., Maitland, N., ChenevixTrench, G., Riggins, G. J., Bigner, D. D., Palmieri, G., Cossu, A., Flanagan, A., Nicholson, A., Ho, J. W., Leung, S. Y., Yuen, S. T., Weber, B. L., Seigler, H. F., Darro w, T. L., Paterson, H., Marais, R., Marshall, C. J., Wooster, R., Stratton, M. R., and Fu treal, P. A. Mutations of the BRAF gene in human cancer. Nature, 417: 949-954, 2002. 107. Yuen, S. T., Davies, H., Chan, T. L ., Ho, J. W., Bignell, G. R., Cox, C., Stephens, P., Edkins, S., Tsui W. W., Chan, A. S., Fu treal, P. A., Stratton, M. R., Wooster, R., and Leung, S. Y. Similarity of the phenotypic patterns associated with BRAF and KRAS mutations in colorectal neoplasia. Cancer Res, 62: 6451-6455, 2002. 108. Li, N., Batt, D., and Warmuth, M. B-Raf kinase inhibitors for cancer treatment. Curr Opin Investig Drugs, 8: 452-456, 2007. 109. Tsai, J., Lee, J. T., Wang, W., Zhang, J., Cho, H., Mamo S., Bremer, R., Gillette, S., Kong, J., Haass, N. K., Sproesser, K., Li, L., Smalley, K. S., Fong, D., Zhu, Y. L., Marimuthu, A., Nguyen, H., Lam, B., Liu, J., Cheung, I., Rice, J., Suzuki, Y., Luu, C., Settachatgul, C., Shel looe, R., Cantwell, J., Kim, S. H., Schlessinger, J., Zhang, K. Y., West, B. L., Powell, B., Habets, G., Zhang, C., Ibrahim, P. N., Hirth, P ., Artis, D. R., Herl yn, M., and Bollag, G. Discovery of a selective inhibito r of oncogenic B-Raf kinase with potent antimelanoma activity. Proc Natl Acad Sci U S A, 105: 3041-3046, 2008. 110. McClean, M. N., Mody, A., Broach, J. R., and Ramanathan, S. Cross-talk and decision making in MAP kinase pathways. Nat Genet, 39: 409-414, 2007.
195 111. Raman, M., Chen, W., and Cobb, M. H. Differential regulation and properties of MAPKs. Oncogene, 26: 3100-3112, 2007. 112. Dhillon, A. S. and Kolc h, W. Untying the regulat ion of the Raf-1 kinase. Arch Biochem Biophys, 404: 3-9, 2002. 113. Dritschilo, A., Huang, C. H., Rudin, C. M., Marshall, J., Collins, B., Dul, J. L., Zhang, C., Kumar, D., Gokhale, P. C., Ahmad, A., Ahmad, I., Sherman, J. W., and Kasid, U. N. Phase I study of liposome-encapsulated c-raf antisense oligodeoxyribonucleot ide infusion in combination with radiation therapy in patients with advanced malignancies. Clin Cancer Res, 12: 1251-1259, 2006. 114. Wilhelm, S. M., Carter, C., Tang, L., Wilkie, D., McNabola, A., Rong, H., Chen, C., Zhang, X., Vincent, P., Mc Hugh, M., Cao, Y., Shujath, J., Gawlak, S., Eveleigh, D., Rowley, B., Liu, L., Adnane, L., Lynch, M., Auclair, D., Taylor, I., Gedrich, R., Voznesensky, A., Riedl, B., Post, L. E., Bollag, G., and Trail, P. A. BAY 43-9006 exhibits broad spectrum oral antitumor activity and targets the RAF/MEK/ERK pathway and receptor tyrosine kinases involved in tumor progression and angiogenesis. Cancer Res, 64: 7099-7109, 2004. 115. Hahn, O. and Stadler, W. Sorafenib. Curr Opin Oncol, 18: 615-621, 2006. 116. Zheng, C. F. and Guan, K. L. Cloning and characterization of two distinct human extracellular signal-regulated kinase activator kinases, MEK1 and MEK2. J Biol Chem, 268: 11435-11439, 1993. 117. Catling, A. D., Schaeffer, H. J., Reuter, C. W., Reddy G. R., and Weber, M. J. A proline-rich sequence unique to MEK1 and MEK2 is required for raf binding and regulates MEK function. Mol Cell Biol, 15: 5214-5225, 1995. 118. Chen, Z., Gibson, T. B., Robinson, F. Silvestro, L., Pearson, G., Xu, B., Wright, A., Vanderbilt, C., and Cobb, M. H. MAP kinases. Chem Rev, 101: 2449-2476, 2001. 119. Lewis, T. S., Shapiro, P. S., and A hn, N. G. Signal tr ansduction through MAP kinase cascades. Adv Cancer Res, 74: 49-139, 1998. 120. Crews, C. M., Alessandrini, A., and Er ikson, R. L. The pr imary structure of MEK, a protein kinase that p hosphorylates the ERK gene product. Science, 258: 478-480, 1992. 121. Haycock, J. W., Ahn, N. G., Cobb, M. H., and Krebs, E. G. ERK1 and ERK2, two microtubule-associated protein 2 kinases, mediate the phosphorylation of tyrosine hydroxylase at serine-31 in situ. Proc Natl Acad Sci U S A, 89: 2365-2369, 1992. 122. Seger, R. and Krebs, E. G. The MAPK signaling ca scade. Faseb J, 9: 726-735, 1995. 123. Duff, J. L., Monia, B. P., and Berk, B. C. Mitogen-activa ted protein (MAP) kinase is regulated by the MAP ki nase phosphatase (MKP-1) in vascular smooth muscle cells. Effect of actinomycin D and antisense oligonucleotides. J Biol Chem, 270: 7161-7166, 1995.
196 124. Loda, M., Capodieci, P., Mishra, R., Yao, H., Corless, C., Grigioni, W., Wang, Y., Magi-Galluzzi, C., and Stor k, P. J. Expression of mitogenactivated protein kinase phosphatas e-1 in the early phases of human epithelial carcinogenesis. Am J Pathol, 149: 1553-1564, 1996. 125. Magi-Galluzzi, C. and Loda, M. Molecular events in the early phases of prostate carcinogenesis. Eur Urol, 30: 167-176, 1996. 126. Cox, M. E., Deeble, P. D., Bissonette E. A., and Parsons, S. J. Activated 3',5'-cyclic AMP-dependent protein ki nase is sufficient to induce neuroendocrine-like differentia tion of the LNCaP prostate tumor cell line. J Biol Chem, 275: 13812-13818, 2000. 127. Carman, C. V. and Benov ic, J. L. G-protein-coupl ed receptors: turn-ons and turn-offs. Curr Opin Neurobiol, 8: 335-344, 1998. 128. Troadec, J. D., Marien, M., Mour levat, S., Debeir, T., Ruberg, M., Colpaert, F., and Michel, P. P. Activation of the mitogen-activated protein kinase (ERK(1/2)) signaling pathway by cyclic AMP potentiates the neuroprotective effect of the neurotransmitter noradrenaline on dopaminergic neurons. Mol Pharmacol, 62: 1043-1052, 2002. 129. Steinberg, S. F. The cellular actions of beta-adrenergic receptor agonists: looking beyond cAMP. Circ Res, 87: 1079-1082, 2000. 130. Sun, J., Wang, D. A., Jain, R. K ., Carie, A., Paquette, S., Ennis, E., Blaskovich, M. A., Baldini, L., Coppola, D., Hamilton, A. D., and Sebti, S. M. Inhibiting angiogenesis and tumorigenes is by a synthetic molecule that blocks binding of both VEGF and PDGF to thei r receptors. Oncogene, 24: 4701-4709, 2005. 131. Sun, J., Ohkanda, J., Coppola, D., Yin, H., Kothare, M., Busciglio, B., Hamilton, A. D., and Sebt i, S. M. Geranylgeranylt ransferase I inhibitor GGTI-2154 induces breast carcinoma apo ptosis and tumor regression in H-Ras transgenic mice. Cancer Res, 63: 8922-8929, 2003. 132. Jiang, Z., Wu, C. L., Woda, B. A., I czkowski, K. A., Chu, P. G., Tretiakova, M. S., Young, R. H., Weiss, L. M., Blute, R. D., Jr., Brendler, C. B., Krausz, T., Xu, J. C., Rock, K. L., Amin, M. B., and Yang, X. J. Alphamethylacyl-CoA racemase: a multi-inst itutional study of a new prostate cancer marker. Histopathology, 45: 218-225, 2004. 133. Sidovar, M. F., Kozlowski, P., Lee, J. W., Collins, M. A., He, Y., and Graves, L. M. Phosphorylation of seri ne 43 is not required for inhibition of c-Raf kinase by the cAMP-dependent protein kinase. J Biol Chem, 275: 28688-28694, 2000. 134. Veber, N., Prevost, G., Planchon, P ., and Starzec, A. Evidence for a growth effect of epidermal growth factor on MDA-MB-231 breast cancer cells. Eur J Cancer, 30A: 1352-1359, 1994. 135. Yan, L., Herrmann, V., Hofer, J. K., and Insel, P. A. beta-adrenergic receptor/cAMP-mediated signaling and apoptosis of S49 lymphoma cells. Am J Physiol Cell Physiol, 279: C1665-1674, 2000.
197 136. Eisenmann, K. M., V anBrocklin, M. W., Staffend, N. A., Kitchen, S. M., and Koo, H. M. Mitogen-activat ed protein kinase pathway-dependent tumor-specific survival signaling in melanoma cells through inactivation of the proapoptotic protei n bad. Cancer Res, 63: 8330-8337, 2003. 137. Zha, J., Harada, H., Yang, E., Jocke l, J., and Korsmeyer, S. J. Serine phosphorylation of death agoni st BAD in response to survival factor results in binding to 14-3-3 not BCL-X(L). Cell, 87: 619-628, 1996. 138. Harada, H. and Grant, S. Apoptosis regulators. Rev Clin Exp Hematol, 7: 117-138, 2003. 139. Harada, H., Quearry, B., Ruiz-Vela, A., and Korsmeyer, S. J. Survival factor-induced extracellular signal-r egulated kinase phosphorylates BIM, inhibiting its association with BAX and proapoptotic activity. Proc Natl Acad Sci U S A, 101: 15313-15317, 2004. 140. O'Neill, E. and Kolch, W. Taming t he Hippo: Raf-1 controls apoptosis by suppressing MST2/Hippo. Cell Cycle, 4: 365-367, 2005. 141. Di Marco, E., Pierce, J. H., Flemi ng, T. P., Kraus, M. H., Molloy, C. J., Aaronson, S. A., and Di Fi ore, P. P. Autocrine interaction between TGF alpha and the EGF-receptor : quantitative requirement s for induction of the malignant phenotype. Oncogene, 4: 831-838, 1989. 142. Fleming, T. P., Matsui T., and Aaronson, S. A. Platelet-derived growth factor (PDGF) receptor activation in cell transformation and human malignancy. Exp Gerontol, 27: 523-532, 1992. 143. Watanabe, M., Nobuta, A., Tanaka, J. and Asaka, M. An effect of K-ras gene mutation on epidermal growth fact or receptor signal transduction in PANC-1 pancreatic carcinoma cells. Int J Cancer, 67: 264-268, 1996. 144. Graells, J., Vinyals, A., Figueras, A ., Llorens, A., Moreno, A., Marcoval, J., Gonzalez, F. J., and Fabra, A. Overproduction of VEGF concomitantly expressed with its recept ors promotes growth and survival of melanoma cells through MAPK and PI3K si gnaling. J Invest Dermatol, 123: 11511161, 2004. 145. Weldon, C. B., Scandurro, A. B., Rolf e, K. W., Clayton, J. L., Elliott, S., Butler, N. N., Melnik, L. I., Alam, J., McLachlan, J. A., Jaffe, B. M., Beckman, B. S., and Burow, M. E. Identification of mitogen-activated protein kinase kinase as a chemores istant pathway in MCF-7 cells by using gene expression microarray. Surgery, 132: 293-301, 2002. 146. Jin, W., Wu, L., Liang, K., Liu, B., Lu, Y., and Fan, Z. Roles of the PI-3K and MEK pathways in Ras-mediated ch emoresistance in breast cancer cells. Br J Cancer, 89: 185-191, 2003. 147. Strobl, J. S., Wonderlin, W. F ., and Flynn, D. C. Mitogenic signal transduction in human breast cancer cells. Gen Pharmacol, 26: 16431649, 1995. 148. Erhardt, P., Troppmair, J., Rapp, U. R., and Cooper, G. M. Differential regulation of Raf-1 and B-Raf and Ras-dependent activation of mitogenactivated protein kinase by cyclic AMP in PC12 cells. Mol Cell Biol, 15: 5524-5530, 1995.
198 149. Grassi, V., Daniotti, S., Schiassi, M., Dottorini, M., and Tantucci, C. Oral beta 2-selective adrenergi c bronchodilators. Int J Clin Pharmacol Res, 6: 93-103, 1986. 150. Moore, P. F., Constantine, J. W., and Barth, W. E. Pirbuterol, a selecttve beta2 adrenergic bronchodilato r. J Pharmacol Exp Ther, 207: 410-418, 1978. 151. Steen, S. N., Ziment, I., and Th omas, J. S. Pyrbuterol: a new bronchodilator. Phase I-single dose study. Curr Ther Res Clin Exp, 16: 1077-1081, 1974. 152. Boucher, M. J., Duchesne, C., Laine, J., Morisset, J., and Rivard, N. cAMP protection of pancreatic cancer ce lls against apoptosis induced by ERK inhibition. Biochem Biophys Res Commun, 285: 207-216, 2001. 153. Budillon, A., Di Gennaro, E., Caraglia, M., Barbar ulo, D., Abbruzzese, A., and Tagliaferri, P. 8-Cl-cAMP ant agonizes mitogen-activated protein kinase activation and cell growth stimulation induced by epidermal growth factor. Br J Cancer, 81: 1134-1141, 1999. 154. Marangoni, E., Vincent-Salomon, A ., Auger, N., Degeorges, A., Assayag, F., de Cremoux, P., de Plat er, L., Guyader, C., De Pinieux, G., Judde, J. G., Rebucci, M., Tran-Perennou, C., Sa stre-Garau, X., Sigal-Zafrani, B., Delattre, O., Dieras, V., and Poupon, M. F. A new model of patient tumorderived breast cancer xen ografts for preclinical assays. Clin Cancer Res, 13: 3989-3998, 2007. 155. Bishop, A. L. and Hall, A. Rho GTPases and their effector proteins. Biochem J, 348 Pt 2: 241-255, 2000. 156. Van Aelst, L. and D'Souza-Scho rey, C. Rho GTPases and signaling networks. Genes Dev, 11: 2295-2322, 1997. 157. Bourne, H. R., Sanders, D. A ., and McCormick, F. The GTPase superfamily: conserved structure and molecular mechanism. Nature, 349: 117-127, 1991. 158. Kjoller, L. and Hall, A. Signaling to Rho GTPases. Exp Cell Res, 253: 166179, 1999. 159. Gosser, Y. Q., Nomanbhoy, T. K., Aghazadeh, B., Manor, D., Combs, C., Cerione, R. A., and Rosen, M. K. C-te rminal binding domain of Rho GDPdissociation inhibitor directs N-termi nal inhibitory peptide to GTPases. Nature, 387: 814-819, 1997. 160. Hoffman, G. R., Nassar, N., and Cerion e, R. A. Structure of the Rho family GTP-binding protein Cdc42 in comple x with the multifunctional regulator RhoGDI. Cell, 100: 345-356, 2000. 161. Jaffe, A. B. and Hall, A. Rho GTPa ses in transformation and metastasis. Adv Cancer Res, 84: 57-80, 2002. 162. Olofsson, B. Rho guani ne dissociation inhibitors : pivotal molecules in cellular signalling. Cell Signal, 11: 545-554, 1999. 163. Adamson, P., Marshall, C. J., Hall, A., and Tilbrook, P. A. Posttranslational modifications of p21rho proteins. J Biol Chem, 267: 2003320038, 1992.
199 164. Fritz, G. and Kai na, B. Rho GTPases: promisin g cellular targets for novel anticancer drugs. Curr Cancer Drug Targets, 6: 1-14, 2006. 165. Michaelson, D., Sille tti, J., Murphy, G., D'Eust achio, P., Rush, M., and Philips, M. R. Differential localiza tion of Rho GTPases in live cells: regulation by hypervariable regions and RhoGDI binding. J Cell Biol, 152: 111-126, 2001. 166. Lebowitz, P. F., Sakamuro, D ., and Prendergast, G. C. Farnesyl transferase inhibitors induce apoptosis of Ras-transformed cells denied substratum attachment. Cancer Res, 57: 708-713, 1997. 167. Wherlock, M., Gampel, A., Futter, C., and Mellor, H. Farnesyltransferase inhibitors disrupt EGF receptor traffic through modulation of the RhoB GTPase. J Cell Sci, 117: 3221-3231, 2004. 168. Chen, Z., Sun, J., Pradines, A., Fa vre, G., Adnane, J., and Sebti, S. M. Both farnesylated and geranylgerany lated RhoB inhibit malignant transformation and suppress human tu mor growth in nude mice. J Biol Chem, 275: 17974-17978, 2000. 169. Hall, A. Rho GTPases and the actin cytoskeleton. Science, 279: 509-514, 1998. 170. Pruitt, K. and Der, C. J. Ras a nd Rho regulation of the cell cycle and oncogenesis. Cancer Lett, 171: 1-10, 2001. 171. Khosravi-Far, R., Solski, P. A., Clark, G. J., Kinc h, M. S., and Der, C. J. Activation of Rac1, RhoA, and mit ogen-activated protein kinases is required for Ras transform ation. Mol Cell Biol, 15: 6443-6453, 1995. 172. Kobune, M., Chiba, H., Kato, J., Kato, K., Nakamura, K., Kawano, Y., Takada, K., Takimoto, R., Takayama T., Hamada, H., and Niitsu, Y. Wnt3/RhoA/ROCK signaling pathway is involved in adhesion-mediated drug resistance of multiple myelom a in an autocrine mechanism. Mol Cancer Ther, 6: 1774-1784, 2007. 173. Adamson, P., Paterson, H. F., and Hall, A. Intracellular lo calization of the P21rho proteins. J Cell Biol, 119: 617-627, 1992. 174. Wheeler, A. P. and Ridley, A. J. Why three Rho proteins? RhoA, RhoB, RhoC, and cell motility. Exp Cell Res, 301: 43-49, 2004. 175. Zalcman, G., Closson, V., Linare s-Cruz, G., Lerebours, F., Honore, N., Tavitian, A., and Olofsson, B. Regula tion of Ras-related RhoB protein expression during the cell cycle. Oncogene, 10: 1935-1945, 1995. 176. Allal, C., Pradines, A ., Hamilton, A. D., Sebti, S. M., and Favre, G. Farnesylated RhoB prevents cell cyc le arrest and actin cytoskeleton disruption caused by the geranylgeranyltransferase I inhibitor GGTI-298. Cell Cycle, 1: 430-437, 2002. 177. Du, W., Lebowitz, P. F., and Prendergas t, G. C. Cell growth inhibition by farnesyltransferase inhibitors is mediated by gain of geranylgeranylated RhoB. Mol Cell Biol, 19: 1831-1840, 1999. 178. Du, W. and Prendergast, G. C. Ge ranylgeranylated RhoB mediates suppression of human tumor cell growth by farnesyltransferase inhibitors. Cancer Res, 59: 5492-5496, 1999.
200 179. Fritz, G. and Kaina, B. Ras-related GTPase Rhob represses NF-kappaB signaling. J Biol Chem, 276: 3115-3122, 2001. 180. Clark, E. A., Golub, T. R., Lander E. S., and Hynes, R. O. Genomic analysis of metastasis reveals an essential role for RhoC. Nature, 406: 532-535, 2000. 181. Yoshioka, K., Matsumur a, F., Akedo, H., and Itoh, K. Small GTP-binding protein Rho stimulates the actomyosin system, leading to invasion of tumor cells. J Biol Chem, 273: 5146-5154, 1998. 182. Ridley, A. J. Rho proteins and cancer. Breast Cancer Res Treat, 84: 1319, 2004. 183. Wu, M., Wu, Z. F., Kumar-Sinha, C ., Chinnaiyan, A., and Merajver, S. D. RhoC induces differentia l expression of genes in volved in invasion and metastasis in MCF10A breast cells. Breast Cancer Res Treat, 84: 3-12, 2004. 184. Benitah, S. A., Valeron, P. F., Ru i, H., and Lacal, J. C. STAT5a activation mediates the epithelial to mesenchymal transition induced by oncogenic RhoA. Mol Biol Cell, 14: 40-53, 2003. 185. Debidda, M., Wang, L., Zang, H., Poli, V., and Zheng, Y. A role of STAT3 in Rho GTPase-regulated cell migration and prolifer ation. J Biol Chem, 280: 17275-17285, 2005. 186. Sun, H. W., Tong, S. L., He, J., Wang, Q., Zou, L., Ma, S. J., Tan, H. Y., Luo, J. F., and Wu, H. X. RhoA and RhoC -siRNA inhibit the proliferation and invasiveness activity of human gas tric carcinoma by Rho/PI3K/Akt pathway. World J Gastroenterol, 13: 3517-3522, 2007. 187. Teramoto, H., Malek, R. L., Behbahani, B., Castellone, M. D., Lee, N. H., and Gutkind, J. S. Ident ification of H-Ras, RhoA, Rac1 and Cdc42 responsive genes. Oncogene, 22: 2689-2697, 2003. 188. Richert, M. M., Schwertfeger, K. L., Ryder, J. W., and Anderson, S. M. An atlas of mouse mammary gland dev elopment. J Mammary Gland Biol Neoplasia, 5: 227-241, 2000. 189. Nandi, S. Endocrine control of mammarygland development and function in the C3H/ He Crgl mouse. J Natl Cancer Inst, 21: 1039-1063, 1958. 190. Howlett, A. R. and Bissell, M. J. The influence of tissue microenvironment (stroma and extracellular matrix) on the development and function of mammary epithelium. Epithelial Cell Biol, 2: 79-89, 1993. 191. Quarrie, L. H., Addey, C. V., an d Wilde, C. J. Programmed cell death during mammary tissue involution induc ed by weaning, litter removal, and milk stasis. J Cell Physiol, 168: 559-569, 1996. 192. Furth, P. A. Introduction: mamma ry gland involution and apoptosis of mammary epithelial cells. J Ma mmary Gland Biol Neoplasia, 4: 123-127, 1999. 193. Strange, R., Li, F., Saurer, S., Burk hardt, A., and Friis, R. R. Apoptotic cell death and tissue remodelling during mouse mammary gland involution. Development, 115: 49-58, 1992.
201 194. Sheikh, S. N. and Lazarus, P. Re-usable DNA template for the polymerase chain reaction. Nucleic Acids Res, 25: 3537-3542, 1997. 195. Slamon, D. J., Godolphin, W., Jones, L. A., Holt, J. A., Wong, S. G., Keith, D. E., Levin, W. J., Stuart, S. G., Udove, J., Ullrich, A., and et al. Studies of the HER-2/neu prot o-oncogene in human breas t and ovarian cancer. Science, 244: 707-712, 1989. 196. Siegel, P. M., Dankort, D. L., Hard y, W. R., and Muller, W. J. Novel activating mutations in the neu prot o-oncogene involved in induction of mammary tumors. Mol Cell Biol, 14: 7068-7077, 1994. 197. Muller, W. J., Sinn, E., Pattengal e, P. K., Wallace, R., and Leder, P. Single-step induction of mammary adenocarcinoma in transgenic mice bearing the activated c-neu oncogene. Cell, 54: 105-115, 1988. 198. Guy, C. T., Webster, M. A., Schaller, M., Parsons, T. J., Cardiff, R. D., and Muller, W. J. Expression of th e neu protooncogene in the mammary epithelium of transgenic mice induces me tastatic disease. Proc Natl Acad Sci U S A, 89: 10578-10582, 1992. 199. Muller, W. J., Arteaga, C. L., Muthus wamy, S. K., Siegel, P. M., Webster, M. A., Cardiff, R. D., Meis e, K. S., Li, F., Halter, S. A., and Coffey, R. J. Synergistic interaction of t he Neu proto-oncogene product and transforming growth factor alpha in the mammary epithelium of transgenic mice. Mol Cell Biol, 16: 5726-5736, 1996. 200. Adnane, J., Seijo, E., Chen, Z., Bi zouarn, F., Leal, M., Sebti, S. M., and Munoz-Antonia, T. RhoB, not RhoA, represses the transcription of the transforming growth factor beta type II receptor by a mechanism involving activator protein 1. J Biol Chem, 277: 8500-8507, 2002. 201. Houslay, M. D. and Kolch, W. Cell-type specific integration of cross-talk between extracellular signal-regulat ed kinase and cAMP signaling. Mol Pharmacol, 58: 659-668, 2000. 202. Weddle, D. L., Tithoff, P., Willia ms, M., and Schuller, H. M. Betaadrenergic growth regulat ion of human cancer cell lines derived from pancreatic ductal carcinomas. Carcinogenesis, 22: 473-479, 2001. 203. Schuller, H. M., Tithof, P. K., Williams, M., and Plummer, H., 3rd The tobacco-specific carcinogen 4-(met hylnitrosamino)-1-(3-pyridyl)-1butanone is a beta-adrenergic agonist and stimulates DNA synthesis in lung adenocarcinoma via beta-adrenergic receptor -mediated release of arachidonic acid. Cancer Res, 59: 4510-4515, 1999. 204. Schuller, H. M., Porter, B., and Riec hert, A. Beta-adrener gic modulation of NNK-induced lung carcinogenesis in hamsters. J Cancer Res Clin Oncol, 126: 624-630, 2000. 205. Mendelsohn, J. and Baselga, J. Status of epidermal growth factor receptor antagonists in the biology and treatm ent of cancer. J Clin Oncol, 21: 27872799, 2003.
About the Author Adam Carie received his bachelors degree from the University of South Florida in the fall of 2002, and received the Bioche mistry Student of the Year award. He joined the Sebti lab during his senior ye ar for undergraduate research credit. After graduation he joined the Cancer Biology Ph.D. program, a joint collaboration between Moffitt Cancer Center and USF. He stayed in the Sebti lab for the remainder of his graduate work, where he published a first author paper in the journal Oncogene, and co-authored several other papers from interand intra-laboratory collaborations. Adam also presented his work in poster form at American Association of Cancer Research international meet ings in 2006 and 2007. Likewise, Adam presented posters at a Gordon Research Conference in 2005, Moffitt Research Days in 2006 and 2007 and USF Health Science Research Days in 2006 and 200 7. Adam was also invited to talk at the St. Jude National Graduate Student Symposia in 2007 and the CTEP Early Drug Discovery Meeting in 2007.