Clemson University TigerPrints

All Dissertations Dissertations

12-2007 DEVELOPMENT OF NOVEL PROLACTIN AND GROWTH HORMONE RECEPTOR AGONISTS AND ANTAGONISTS John Langenheim Clemson University, [email protected]

Follow this and additional works at: https://tigerprints.clemson.edu/all_dissertations Part of the Molecular Biology Commons

Recommended Citation Langenheim, John, "DEVELOPMENT OF NOVEL PROLACTIN AND GROWTH HORMONE RECEPTOR AGONISTS AND ANTAGONISTS" (2007). All Dissertations. 157. https://tigerprints.clemson.edu/all_dissertations/157

This Dissertation is brought to you for free and open access by the Dissertations at TigerPrints. It has been accepted for inclusion in All Dissertations by an authorized administrator of TigerPrints. For more information, please contact [email protected].

DEVELOPMENT OF NOVEL PROLACTIN AND GROWTH HORMONE RECEPTOR AGONISTS AND ANTAGONISTS

A Dissertation Presented to the Graduate School of Clemson University

In Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy Microbiology

by John Fairbanks Langenheim December 2007

Accepted by: Dr. Wen Y. Chen, Committee Chair Dr. Thomas E. Wagner Dr. Steven E. Ellis Dr. Lyndon L. Larcom

ABSTRACT

Potential indications for human prolactin (hPRL) and human growth hormone

(hGH) are the support and sustainment of lactation and the enhancement or reconstitution

of the immune system when under dysregulation. A prolactin receptor (PRLR)

antagonist, hPRL-G129R (G129R), and a PRLR and growth hormone receptor (GHR)

antagonist, hGH-G120R (G120R), have been developed which may be useful for the treatment of breast cancer since PRLR is elevated in a majority of human breast tumors and makes cancer cells highly sensitive to the mitogenic and anti-apoptotic activity of hPRL and hGH.

As a single agent, G129R is more cytostatic than cytotoxic to human breast cancer cells xenografted into nude mice; when combined with common chemotherapies (e.g.

Tamoxifen or Herceptin) G129R has additive effects. In Chapter 2, we propose to enhance the cytotoxicity of G129R by fusing it to a toxin that potently inhibits protein synthesis. We demonstrate that the fusion protein specifically targets the PRLR and drastically enhances the cytotoxicity of G129R in vitro and that it can be administered safely to mice in vivo.

The low molecular weight of hPRL and G129R make it difficult to maintain

therapeutic levels in the plasma. To increase the plasma half-life we conjugated hPRL and

G129R with hydroxyethyl starch (HES); unfortunately, the ligands lost their intrinsic activities.

In Chapter 3, we propose to increase the plasma half-life of hPRL, hGH, and their respective

antagonists by fusing them to a small serum albumin binding peptide. We demonstrate that the

serum albumin binding peptide extends the plasma half-life of the ligands and enhances

the bioactivity of hPRL and hGH in vivo.

ii In Chapter 4, we propose to increase the plasma half-life and make hPRL and hGH more amendable to chemical modifications such as HESylation by fusing two molecules of hPRL and hGH in tandem (PRL-PRL and GH-GH). We demonstrate that their intrinsic activity is unaffected and that the plasma half-life of PRL-PRL is extended. We also fused two molecules of G129R or G120R in tandem (G129R-G129R and G120R-G120R) and unexpectedly ended up with agonists. This indicates that the PRLR is flexible and that the receptor binding sites on PRL-PRL and GH-GH can potentially interact with receptors in three different ways, reducing the probability that chemical modifications will affect their crucial receptor binding sites.

hGH can mediate its effects via a PRLR homodimer or a GHR homodimer and there is no reason to believe it might not also mediate its effects via a PRLR/GHR heterodimer.

Unfortunately, there is no easy way to discern which receptor hGH exploits. In Chapter 5, we extend the findings of Chapter 4 (i.e. that the fusion of two antagonists together results in an agonist) and propose to make two unique ligands that are only capable of binding to a

PRLR/GHR heterodimer. We confirm that PRLR/GHR heterodimers either pre-exist or can be induced, something which has previously only been shown to occur in rudiments. Also we made a GHR agonist, characterized it, and confirmed that it is a pure GHR agonist. This molecule can be used in place of hGH if one only wants to observe somatogenic effects in human cell lines.

iii DEDICATION

To my parents, Susan and John Langenheim, who made it all possible. To my grandfather, Lawson Langenheim, who made life a whole lot easier. To my grandmother, Louise Griffiths, and her brother, Brent Breedin, for their unconditional support. To my high school biology teacher, Larry Nemerow, who inspired me to become a biologist. And last but not least, Rick and Melanie Ulrich, for their friendship and encouragement over the years.

iv ACKNOWLEDGMENTS

I would like to thank all the people that have made this work possible. Dr. Wen

Y. Chen for his patience and guidance, he has helped me acquire knowledge in a

completely new field of study, cancer research, and taught me critical skills that will aid in my development as a scientist. I would also like to thank my committee members, Dr.

Thomas E. Wagner, Dr. Steven E. Ellis, and Dr. Lyndon L. Larcom for their helpful comments and guidance.

I would like to thank my colleagues and friends: Seth Tomblyn, Jang Pyo Park,

Mike Beck, Alison Springs, Karl Franek, Susan Peirce, Michele Scotti, Isabelle

Jacquemart, Lori and Eric Holle, Guillermo Deangulo, and Keneisha Burrell.

I would like to extend my sincerest appreciation to all of the contributors to the

Endowment Fund of the Greenville Hospital System, especially William and Frances

Byran, and the National Cancer Institute (CA105479) and The Susan G. Komen

Foundation (BCTR0402985 and BCTR091306) for their financial support.

v TABLE OF CONTENTS

Page

TITLE PAGE...... i

ABSTRACT...... ii

DEDICATION...... iv

ACKNOWLEDGMENTS ...... v

LIST OF TABLES...... viii

LIST OF FIGURES ...... ix

CHAPTER

I. LITERATURE REVIEW...... 1

Background on Prolactin, Growth Hormone, and Chorionic Somatomammotropin...... 1 Breast Cancer and Hormones...... 8 Current Therapies for Breast Cancer ...... 23 Targeting Prolactin and Growth Hormone Receptor Signaling ...... 27

II. DEVELOPMENT OF A PROLACTIN RECEPTOR TARGETING FUSION TOXIN USING A PROLACTIN RECEPTOR ANTAGONIST AND A RECOMBINANT FORM OF PSEUDOMONAS EXOTOXIN A...... 37

Abstract...... 37 Introduction...... 39 Materials and Methods...... 46 Results...... 56 Discussion...... 71

vi Table of Contents (Continued)

Page

III. IMPROVING THE PHARMACOKINETIC PROPERTIES OF PROLACTIN AND GROWTH HORMONE RECEPTOR AGONISTS AND ANTAGONISTS BY FUSION TO A SYNTHETIC ALBUMIN BINDING PEPTIDE...... 81

Abstract...... 81 Introduction...... 83 Materials and Methods...... 85 Results...... 95 Discussion...... 114

IV. IMPROVING THE PHARMACOKINETIC PROPERTIES OF PROLACTIN AND GROWTH HORMONE RECEPTOR AGONISTS BY FUSION IN TANDEM ...... 120

Abstract...... 120 Introduction...... 122 Materials and Methods...... 125 Results...... 134 Discussion...... 150

V. DEVELOPMENT OF NOVEL LIGANDS THAT ACTIVATE JAK2/STAT5 SIGNALING EXCLUSIVELY THROUGH THE HETERODIMER OF A PROLACTIN RECEPTOR AND GROWTH HORMONE RECEPTOR...... 161

Abstract...... 161 Introduction...... 163 Materials and Methods...... 166 Results...... 174 Discussion...... 187

VI. CONCLUSIONS...... 196

REFERENCES ...... 198

vii LIST OF TABLES

Table Page

2.1 Extrapolated concentrations of G129R-PE40-KDEL necessary for 50% inhibition of protein synthesis and 50% reduction in the relative number of viable cells in various breast cancer cell lines ...... 74

2.2 Toxicity of PE and G129R-PE40-KDEL to mice ...... 80

3.1 List of oligonucleotides used to construct DNA encoding albumin binding fusion proteins...... 86

3.2 Pharmacokinetic parameters after a single treatment of mice with hPRL, hPRL-SA20, or SA20-hPRL ...... 106

5.1 Effect of hGH and hPRL analogs on receptor dimers ...... 191

viii LIST OF FIGURES

Figure Page

1.1 Illustration of the signaling pathways activated as a result of the sequential binding of PRL to PRLRs...... 7

1.2 Mechanisms exploited by breast cancer cells to sustain PRLR activation...... 19

1.3 Representation of PRLR gene expression in 25 tumors and normal contiguous breast tissues...... 26

1.4 Model of hPRL and G129R complexes with the extracellular domain of the human PRLR ...... 33

1.5 Schematic representation of G129R competitively inhibiting the binding of hPRL to PRLRs...... 34

1.6 Apoptotic response of four PRLR-positive human breast cancer cell lines to G129R treatment ...... 35

1.7 Effects of hPRL and G129R on T-47D human breast cancer cell xenograft growth in nude mice...... 36

2.1 Structural domains of Pseudomonas exotoxin A...... 42

2.2 Diagram of intoxification by exotoxin A...... 43

2.3 Diagram of G129R-PE40KDEL...... 45

2.4 Schematic representation of plasmid used to produce G129R-PE40-KDEL in E. coli...... 55

2.5 Reducing SDS-PAGE and Western blot analysis of purified PE, G129R, and G129R-PE40-KDEL...... 57

2.6 Binding of hPRL, G129R, and G129R-PE40-KDEL to PRLRs on the surface of T-47D human breast cancer cells...... 58

ix List of Figures (Continued)

Figure Page

2.7 Stimulation of STAT5 phosphorylation by hPRL and PRL-PE40-KDEL using T-47D human breast cancer cells...... 61

2.8 Competitive inhibition of STAT5 phosphorylation by G129R and G129R-PE40-KDEL in T-47D human breast cancer cells...... 62

2.9 Examination of PARP cleavage and the effect of zVAD- fmk on the viability of MCF-7 human breast cancer cells treated with PE, hPRL, G129R, or G129R- PE40-KDEL ...... 63

2.10 The effect of G129R, PE, and G129R-PE40-KDEL on the subcellular location of Bax and Cytochrome C in MCF-7 human breast cancer cells ...... 65

2.11 The effect of G129R-PE40KDEL on protein synthesis and the relative viability of human breast cancer cell lines expressing different levels of PRLR...... 66

2.12 Comparison of the response of breast cancer cells to G129R, G129R-PE40-KDEL, and PE ...... 69

2.13 Prevention of G129R-PE40-KDEL-induced protein synthesis inhibition and cytotoxicity by hPRL, G129R, and anti-PE...... 70

2.14 Examination of non-specific cytotoxicity in 4T1 mouse breast cancer cells in response to G129R, PE40, and G129R-PE40-KDEL ...... 73

3.1 Dose-dependent activation of STAT5 and Erk1/2 signaling induced by hPRL, hGH, and their SA20 derivatives...... 97

3.2 Time course analysis of the activation of Erk1/2 in response to hPRL, hGH, and their SA20 derivatives...... 98

x List of Figures (Continued)

Figure Page

3.3 Comparison of the inhibitory effects of G129R, G120R, and their SA20 derivatives on hPRL and hGH- induced signaling...... 99

3.4 The effects of hPRL, G129R, and their SA20 derivatives on Nb2 cell proliferation...... 103

3.5 Mean plasma concentration-time profiles of hPRL, SA20-hPRL, and hPRL-SA20 after administration to mice...... 104

3.6 The rate of hPRL and SA20-hPRL clearance in mice and their effective molecular size ...... 107

3.7 Lobulo-alveolar response in the mammary glands of mature nulliparous mice treated with PBS, hPRL, or SA20-hPRL...... 110

3.8 Lobulo-alveolar response in the mammary glands of young nulliparous mice treated with PBS, hGH, or SA20-hGH...... 111

3.9 Tail length measurements after treatment with PBS, hGH, or SA20-hGH...... 112

3.10 Body weight measurement after treatment with PBS, hGH, or SA20-hGH...... 113

4.1 Non-reducing SDS-PAGE, reducing SDS-PAGE, and Western blot analysis of purified monomeric and dimeric hPRL derivatives...... 133

4.2 Competitive binding of monomeric and homodimeric hPRL derivatives to PRLRs on T-47D cells...... 136

4.3 Monomeric hPRL and homodimeric G129R-induced homodimerization of PRLRs...... 137

xi List of Figures (Continued)

Figure Page

4.4 Western blot analysis of the phosphorylation status of Akt, Erk1/2, and STAT5 in response to monomeric hPRL and hGH derivatives ...... 139

4.5 Western blot analysis of the phosphorylation status of Akt, Erk1/2, and STAT5 in response to dimeric hPRL and hGH derivatives ...... 140

4.6 Dose-dependent activation of STAT5 phosphorylation induced by monomeric and dimeric hPRL derivatives...... 141

4.7 Dose-dependent induction of STAT5, Erk1/2 and Akt phosphorylation induced by monomeric and dimeric hGH derivatives...... 145

4.8 Time course of Erk1/2 and STAT5 phosphorylation in T-47D cells in response to monomeric and dimeric hPRL derivatives...... 146

4.9 Time course of STAT5, Akt, and ERK1/2 phospho- rylation in T-47D cells in response to monomeric and dimeric hGH derivatives ...... 147

4.10 Rat Nb2 cell proliferation in response to monomeric and dimeric hPRL and hGH derivatives ...... 148

4.11 Examination of rat Nb2 cell proliferation in response to truncated forms of dimeric G129R ...... 149

4.12 Pharmacokinetics of monomeric and dimeric forms of hPRL in vivo ...... 159

4.13 Illustration of monomeric and homodimeric G129R and G120R binding to PRLRs ...... 160

5.1 Reducing SDS-PAGE analysis of purified agonists, antagonists, and fusions of the antagonists...... 173

xii List of Figures (Continued)

Figure Page

5.2 Fusions of a PRLR antagonist (G129R) and a PRLR/ GHR antagonist (G120R) or pure GHR antagonist (B2036) do not activate STAT5 signaling in cells expressing only GHRs...... 176

5.3 Fusions of a PRLR antagonist (G129R) and a PRLR/ GHR antagonist (G120R) or pure GHR antagonist (B2036) inhibit STAT5 signaling in cells expressing only GHRs...... 177

5.4 Proliferation of rat Nb2 cells in response to monomers or fusions of G129R and G120R or B2036 in comparison to hGH, hPRL, and B2036-REV...... 179

5.5 Fusions of B2036 and G129R retain the ability to bind PRLR as indicated by the activation of or competitive inhibition of STAT5 signaling in L-cells transfected with human PRLR...... 181

5.6 Western blot analysis of the phosphorylation status of STAT5 in T-47D cells in response to PRLR and GHR agonists and antagonists ...... 183

5.7 Western blot analysis of the phosphorylation status of STAT5 in T-47D cells in response to PRLR and GHR agonists and antagonist fusions ...... 184

5.8 Western blot analysis of the phosphorylation status of STAT5 in transfected HEK293 cells in response to hPRL, B2036-REV, and fusions of a PRLR antagonist (G129R) and GHR antagonist (B2036)...... 186

5.9 Illustration of the binding of G129R, B2036, or fusions of the two in tandem to homodimers or heterodimers of PRLR and GHR...... 190

xiii

CHAPTER ONE

LITERATURE REVIEW

Background on Prolactin, Growth Hormone, and Chorionic Somatomammotropin

Human prolactin (PRL), growth hormone (GH), chorionic somatomammotropin

(CS) and chorionic somatomammotropin-like hormone (CS-L) are members of the

cytokine protein superfamily. Although closely related, variations between these peptide

hormones are evident at the genetic level, transcriptional level, site of expression and in

their biological properties such as receptor specificity.

Gene Structure

The genes encoding these peptide hormones are composed of five exons which

are separated by four introns. The site of the introns has remained conserved while the

length of the introns has not. PRL is encoded by a single gene (PRL) located on

chromosome 6 (Owerbach et al., 1981) and the genes encoding GH (GH1, GH2), CS

(CSH1, CSH2) and CS-L (CSHL1) are closely linked on chromosome 17 (Chen et al.,

1989; Cooke and Liebhaber, 1995). GH1 encodes normal GH (GH-N), GH2 encodes a

placental GH variant (GH-V) which differs by 13 amino acids, and CSH1 and CSH2 give rise to identical CSs often referred to as placental lactogen (PL).

Gene Transcription

In the pituitary, transcription of the PRL and GH1 genes is mediated via the tissue-specific transcription factor Pit-1. In extrapituitary cells, transcription of the PRL gene occurs from a promoter located 5.8 kb upstream of the promoter used in the pituitary. Although the regulation of PRL transcription differs the translation and signal

1 sequence cleavage are identical, resulting in a single PRL isoform. In contrast,

differential splicing of the messenger ribonucleic acid (mRNA) precursor of GH1 gives

rise to five isoforms; GH2 and CSHL1, four isoforms; and CSH1 and CSH2, three

isoforms. Translation of these altered transcripts into protein may serve as a means of

further specialization (Lewis, 1984). For example, a 20 kD isoform of GH lacking amino

acid residues 32–46 of GH-N (Lewis et al., 1980; Cooke et al., 1988) is commonly expressed (Ishikawa et al., 1999) that exhibits identical agonism for growth hormone receptors (GHRs) (Wada et al., 1997) but weaker agonism for prolactin receptors

(PRLRs) (Tsunekawa et al., 1999).

Site of Expression

PRL and GH-N are synthesized in the anterior lobe of the pituitary gland, are stored in granules, and are released in a pulsatile manner in response to hypothalamic neuropeptides and growth factors. Secretion of PRL from lactotrophs is directly under the stimulatory and strong inhibitory control of thyrotropin releasing hormone (TRH) and dopamine, respectively; whereas, secretion of GH-N from somatotropes is directly under the stimulatory and inhibitory control of growth hormone releasing hormone (GHRH) and somatostatin, respectively. Alternative promoter usage for PRL and GH1 results in extrapituitary secretion of PRL (Gellersen et al., 1994; Sinha, 1995) and GH (Wu et al.,

1996; Liu et al., 1997). Extrapituitary production of PRL has been reported in the uterus

(decidualized endometrial stroma, myometrial layer), immune system cells (peripheral blood lymphocytes, thymocytes, hematopoetic system cells), and more recently in the mammary gland and breast cancers (Ben-Jonathan et al., 1996). This extrapituitary PRL

2 is not stored in granules and its secretion is not regulated by neuropeptides. GH-V, CS-L and the CSs are transcribed and secreted almost exclusively in the placenta during pregnancy (Barrera-Saldana et al., 1983; Baumann et al., 1991).

Physiology

Physiologically, PRL, GH and PL are intimately involved in modulating reproductive functions. Evidence suggests that these ligands induce signaling in different compartments (Gallego et al., 2001) and that the PRLR and GHR do not have an instructive role, but rather provide tissue specificity (Brisken et al., 2002). These ligands regulate ovarian, mammary gland and immune functions and adjust maternal metabolism during pregnancy and lactation. In addition, PRL, GH and PL activity is often observed in tissues exhibiting pathology which are normally targets during the establishment of pregnancy. PRL primarily controls mammary gland development (Horseman et al.,

1997; Ormandy et al., 1997a). PL mimics the action of PRL on mammary gland development, maintenance of the corpus luteum of pregnancy, and promotion of luteal progesterone synthesis during pregnancy (Handwerger and Freemark, 1987). In addition,

PL appears to substitute for GH-N in regulating IGF-I and stimulates growth of the fetus during pregnancy. GH-V has also been reported to replace GH-N during pregnancy

(Frankenne et al., 1988; Eriksson et al., 1989) and to modulate maternal IGF-I levels

(Hall et al., 1984).

Hormone Structure

Structurally, PRL (Keeler et al., 2003; Teilum et al., 2005), GH (de Vos et al.,

1992; Chantalat et al., 1995; Clackson and Wells, 1995) and PL (Walsh and Kossiakoff,

3 2006) are similar to one another. PL is more closely related to GH than PRL (Dietz et al.,

1992; Gootwine and Yossafi, 1998), sharing 93% and 42% sequence identity, respectively. Each hormone contains between 190-199 amino acids which are arranged into four α-helical segments that are connected via loops, allowing the helices to be bundled in an anti- parallel (up-up-down-down) manner. The structural integrity of the four-helix bundle is maintained by intra-chain disulfide bonds and non-covalent interactions between the helices. Two receptor binding sites are located on the surface of these peptide hormones that are structurally distinct.

Receptor Structure

The receptors to which PRL, PL, and GH bind belong to the class 1 cytokine receptor family. They are characterized as forming homodimers and as having no intrinsic kinase activity. The full length GHR and PRLR are composed of 620 and 598 amino acids, respectively. Topologically, each receptor has an N-terminal extracellular domain involved in hormone binding and receptor dimerization, a single-pass transmembrane domain involved in membrane anchorage, and a C-terminal intracellular domain involved in signal transduction. Although the PRLR and GHR exhibit little overall sequence identity, they have highly conserved sequences in their extracellular and intracellular domains.

The tertiary structure of the extracellular domain of the PRLR and GHR are similar due to four positionally conserved Cys residues and a Trp-Ser-X-Trp-Ser

(WSXWS) motif which stabilize the cytokine recognition site (Somers et al., 1994).

They each have fibronectin-like type III subdomains (S1 and S2) which are connected via

4 a hinge region. The N-terminal S1 domain is involved in ligand binding and the

membrane proximal S2 domain plays more of a structural role. The PRLR appears to be

more promiscuous, binding PRL, GH and PL with high affinity; whereas, the GHR only

binds GH with high affinity. The membrane proximal region of the intracellular domains

of these receptors have a conserved Box 1 motif which is the binding site for an

intracellular non-receptor tyrosine kinase, Janus-activated kinase 2 (JAK2).

Hormone-Induced Receptor Dimerization

The two structurally distinct binding sites of GH (Cunningham et al., 1991; de

Vos et al., 1992; Clackson and Wells, 1995; Sundstrom et al., 1996) each interact with a

GHR to form a trimeric 1:2 ligand-receptor complex (Cunningham et al., 1989;

Cunningham et al., 1991; de Vos et al., 1992). Binding of GH occurs sequentially with the higher affinity Site 1 interacting with the first receptor before lower affinity Site 2 interacts with a second receptor (Cunningham et al., 1991; Fuh et al., 1992; Ilondo et al.,

1994). In addition to binding to GHRs (Hughes and Friesen, 1985), GH also binds to

PRLRs (Kleinberg and Todd, 1980) in a sequential manner (Somers et al., 1994; Goffin

and Kelly, 1997) to form a trimeric 1:2 ligand-receptor complex (Wells et al., 1993). The

binding sites of GH for GHRs (Ultsch et al., 1994) and PRLRs (Somers et al., 1994)

overlap but are not identical (Cunningham and Wells, 1991; Kossiakoff et al., 1994).

Zinc influences the binding of GH to PRLRs but not to GHRs (Cunningham et al.,

1990b) and increases its bioactivity (Dattani et al., 1993; Fuh et al., 1993) by increasing

the affinity of Site 1 for the PRLR (Somers et al., 1994; Duda and Brooks, 2003).

Likewise, PL binds to PRLRs in a sequential manner (Fuh and Wells, 1995) and is

5 influenced by zinc (Lowman et al., 1991). PRL also binds to PRLRs in a sequential

manner as illustrated in Fig. 1.1A (Fuh et al., 1993; Goffin et al., 1994). This sequential binding process has been shown to be due to the functional coupling of the receptor binding sites. The interaction of Site 1 with the first receptor induces conformational changes in the ligand that create a functional Site 2 (Sivaprasad et al., 2004; Walsh et al.,

2004).

Receptor-Mediated Signal Transduction

Binding of GH to GHRs and binding of GH, PL or PRL to PRLRs induces the

formation of functional receptor homodimers, leading to the autophosphorylation of

receptor-associated JAK2 (Fig. 1.1A). Activation of JAK2 can lead to the stimulation of:

(1) signal transducer and activator of transcription (STAT) 1, 3, and 5 signaling; (2) phosphoinositide 3-kinase (PI3K)/3-phosphoinositide-dependent protein kinase 1

(PDK1)/Akt signaling; (3) Ras/Raf/MEK/mitogen-activated protein kinases (MAPKs) signaling by interacting with the Shc/Grb2/SOS complex; and (4) protein kinase C

(PKC)/MAPKs signaling by activating phospholipase C-gamma (PLC-γ). These signaling pathways ultimately affect cell motility, survival, proliferation, and adhesion

(Fig. 1.1B).

Using mutagenesis and by generating truncated forms of the PRLR, it has been

demonstrated that JAK2 is necessary for PRL-induced proliferation (Chang et al., 1998).

It has further been demonstrated that JAK2 is necessary for PRL-induced differentiation

of the mammary gland (Shillingford et al., 2002). PRL-induced JAK2 activation induces

both STAT5A signaling and Shc mediated MAPK signaling (Das and Vonderhaar, 1996).

6 A

1 2 4 3

2 2 1 3 1 4 4 3

Jak2 Jak2 Jak2 Jak2 P Jak2Jak2 P P P

P P P P [PRL:PRLR] 1:2 complex B 2 1 3 4

RAS RAF P Jak2Jak2 P SHC SOS STAT P FYNP Grb2 SHP-2 STAT P P PI3K P P MAPKK

Akt MAPK STAT P P

STAT Cell differentiation Cell survival Cell proliferation

Fig. 1.1 Illustration of the signaling pathways activated as a result of the sequential binding of PRL to PRLRs

(A) The higher affinity binding Site 1 of PRL (helices 1 & 4) interacts with the first PRLR before the lower affinity binding Site 2 (helix 3) can interact with a second receptor. Functional receptor dimerization leads to the autophosphorylation of a receptor-associated tyrosine kinase, JAK2. (B) Activated JAK2 phosphorylates the PRLR and associated proteins leading to the activation of JAK2/STAT signaling cascade, the PI3K/PDK1/Akt signaling pathway, and the Ras/Raf/MEK/MAPKs signaling pathway which are mainly associated with differentiation, survival, and proliferation, respectively.

7 PRL induced JAK2/STAT5A signaling has been shown induce the expression estrogen receptor alpha (ERα) (Frasor et al., 2001) and receptor activator of nuclear factor kappaB

ligand (RANKL) in the mammary gland (Srivastava et al., 2003) via STAT5 binding

elements within their promoters; whereas, JAK2-Shc-MAPK signaling has been shown to

induce bcl-2 expression (Das and Vonderhaar, 1996).

STAT5 activatioin in mammary epithelial cells usually only occurs during

pregnancy and lactation. Constitutively active STAT5A delays the onset of involution; in

contrast, deletion of STAT5 during pregnancy, after the mammary epithelium has entered

STAT5-mediated differentiation, results in premature cell death (Cui et al., 2004). This

suggests the activation of STAT5 signaling by PRL not only serves as a proliferation and

differentiation signal during lactation but also as a survival signal at the end of lactation.

Similar to the mammary gland, the capacity of PRL to increase survival and proliferation

of early T-cell precursors has been shown to be related with its ability to induce lymphoid

differentiation (Carreno et al., 2005). These findings suggest that the multiple biological

effects of PRL (proliferation, differentiation, survival) are not separate events.

Breast Cancer and Hormones

There are numerous hypotheses to explain how a cell is transformed and progresses to cancer. It is generally accepted that cancer is a multi-step process caused by genetic alterations to a cell and the clonal expansion of the altered cell.

Carcinogenesis can be divided into three stages. The first stage of cancer is initiation,

when the DNA of a cell is permanently damaged by carcinogens (hormone, virus,

radiation, trauma) and the growth of the cell is altered. These alterations result from

8 mutations in proto-oncogenes, tumor suppressor genes and stability (repair-mutator)

genes. The second stage of cancer is promotion, when non-mutagenic tumor promoters

expand the tumor cell population to the point where the benign tumor cell mass starts to

interfere with normal function. The third stage is progression, when the tumor becomes

malignant and invades surrounding tissue, blood vessels, lymphatics and metastasizes to

distant parts of the body.

Role of Proto-oncogenes, Tumor Suppressors, and Stability Genes

Proto-oncogenes are genes involved in the control of cell growth. Upon

activation by mutation or overexpression due to gene amplification an oncogene product

can drive unconstrained proliferation. An oncoprotein may induce a cell to release

growth factors (e.g. EGF) into the local environment which are utilized by the cell, alleviating the need for external stimuli to induce cell proliferation. It may be an overexpressed receptor (e.g. HER2) that hypersensitizes the cell to extremely low

concentrations of growth factor, convincing the cell to grow in response to sub-

physiological levels of growth factor. It may be an altered receptor (e.g. mutated EGFR) that is constitutively active and sends growth promoting signals without the need for growth factors. Or it may be a cytoplasmic protein (e.g. c-Src and Ras) or nuclear protein

(e.g. cylin D1) that is part of the signaling cascade downstream of the receptor that is altered in a manner that prevents its signal from being turned off by regulatory proteins.

Tumor suppressor genes are genes that negatively regulate cell growth and

proliferation. There function is lost as a result of mutations or deletions to both alleles

and are often found in familial malignancies. Many tumor suppressors turn off signal

9 transduction after it has been transmitted via a receptor (e.g. PTEN). Mutations in these tumor suppressors can lead to constitutive activation or prolonged signal transduction after stimulation, leading to uninhibited cell proliferation. Other tumor suppressors lie downstream and control cell cycle progression (e.g. p16-INK4A), normally blocking the activity of cyclin-dependent kinases (CDKs) and their interaction with cyclins.

Cells have to overcome many other hurdles in addition to altering growth factor signaling to be able to proliferate indefinitely. During progression to cancer a variety of phenotypic changes must occur, including: insensitivity to anti-growth factor signals; evasion of programmed cell death; immortalization; the induction of blood vessel

formation; and eventually invasion and metastasis (Hanahan and Weinberg, 2000). It is

hard to envision all of these changes occurring without considering the last class of genes

that are mutated in cancers, the stability genes. Stability genes are tumor suppressor

genes involved in DNA repair whose loss of function increases the frequency of mutation

or chromosomal rearrangements that result in loss of heterozygosity (e.g. TP53, BRCA1,

and BRCA2). When DNA is damaged, p53 accumulates and halts the cell cycle until it

has been repaired so as to prevent the copying of mutated DNA; if the DNA is damaged

beyond repair, p53 induces programmed cell death. Loss of p53 allows damaged DNA to

be incorporated and passed along to descendent cells and has been shown to increase the

chance for gene amplifications. Aneuploidy, which is a change in the number or balance

of chromosomes, is a “hallmark” of all cancer cells and can alter the expression of

thousands of genes, providing an appealing explanation for the complexity of the

abnormal phenotypes observed in cancers.

10 As mentioned above, a single mutation in anyone of these genes is unlikely to

turn a normal cell into a malignant cell as originally proposed in the one-molecule one-hit hypothesis. This is supported by epidemiological evidence which shows that the relationship between cancer incidence and age is not linear and that the rates increase sharply with age (American Cancer Society., 2007). This indicates that cancer is a long process and likely involves at least six steps (Beckman and Loeb, 2005). The current hypothesis is that cancer arises from the accumulation of multiple gene mutations within a cell (Knudson, 1971) and that these genetic alterations sequentially give rise to hyperplasia, dysplasia, carcinoma in situ, and eventually cancer (Wellings et al., 1975).

It has been hypothesized that the gene mutations must be in multiple oncogenes and at least one tumor suppressor gene. It has also been hypothesized that the first step in transformation and perhaps the most important genetic alteration is genetic instability (Li et al., 2000). Mutations to genes that preserve chromosomal stability, such as the breast cancer susceptibility gene (BRCA1), inevitably lead to initiation of cancer (Rosen et al.,

2005).

Role of Growth Promoting Agents

Carcinogens include not only agents that directly promote mutations (mutagens)

but also agents that facilitate the propagation of the affected cell. Growth promoting

agents further increase the chance of mutation. The more a cell replicates its DNA the greater the chance of a spontaneous mutation occurring. In addition, mutagens are more effective during DNA replication. This suggests that growth promoting agents may be

11 carcinogenic, even though they are not mutagenic, by shortening the time necessary to

acquire the multiple mutations necessary for cancer.

Growth Promoting Hormones Involved in Breast Development

The mammary gland is one of the few organs to undergo substantial changes

postnatally, including cycles of growth, differentiation, apoptosis, regression, and

remodeling (Topper and Freeman, 1980; Richert et al., 2000; Wiseman and Werb, 2002;

Bissell et al., 2003). The prenatal gland primarily consists of a stromal fat pad and

simple ducts that are lined with epithelial cells and a small number of stem cells critical

for development. Development of these ducts into a functional gland is controlled

primarily by steroid and peptide hormones produced by the ovary and pituitary, respectively. During early puberty, systemic increases in GH and estrogen stimulate the proliferation of the ductal epithelial cells and the formation of large buds at the termini of

the ducts. Proliferation of cap cells within these terminal end buds (TEBs) results in the

extension of the ducts throughout the fat pad. Later in puberty the TEBs regress as

progesterone and PRL induce ductal branching and the formation of lobular buds. By the end of puberty, elaborately branched ducts lined with resting lobules extend throughout the fat pad.

Proliferation within the lobules continues to occur in response to the release of steroid hormones from ovarian follicles during normal menstrual cycling (Ramakrishnan et al., 2002), with the highest proliferation occurring during diestrus. During pregnancy, increases in the systemic levels of PRL, PL, and progesterone induce the lobular epithelial cells to proliferate and form lobulo-alveolar-like structures. At parturition, PL

12 and progesterone levels fall and the lobulo-alveolar epithelial cells terminally

differentiate and form functional alveoli that produce and secrete milk proteins in

response to PRL. Upon weaning the systemic levels of PRL fall and the mammary gland

undergoes involution, at which point inflammatory cells infiltrate the lobules, the alveoli

regress, and lobules with acinar similar to those observed before pregnancy are reformed

from progenitor cells.

The changes within the lobules throughout mammary gland development have

been categorized based on their degree of differentiation (i.e. average number of ductules

per lobular unit). After puberty the TEBs differentiate into Type 1 lobules; at sexual

maturity into Type 2 lobules; during pregnancy into Type 3 lobules, and during lactation

into Type 4 lobules (Russo et al., 2001). Childless women have 65-85% Type 1 lobules,

10-35% Type 2 lobules, and 0-5% Type 3 lobules; whereas, childbearing women have

70-90% Type 3 lobules (Russo et al., 2000). Proliferation within the lobules has been correlated with estrogen receptor alpha (ERα) and progesterone receptor (PR) expression.

As the lobules become more differentiated there are fewer cells proliferating and expressing both ERα and PR (Russo et al., 2000). This indicates that parity reduces the

population of cells capable of proliferating in response to ovarian hormones (i.e. the cells

have differentiated). Parous women also have significantly shorter menstrual cycle

lengths, significantly lower estrogen levels (22%) (Bernstein et al., 1985; Hankinson et

al., 1995), and lower levels of PRL in the follicular phase than at mid-cycle or during the

luteal phase of the menstrual cycle.

13 Growth Promoting Hormones in Breast Cancer

Since hormones from the ovaries and pituitary have an intimate role in normal breast development, it’s not surprising that they may strongly influence the initiation and promotion of breast cancer. Hormones from ovaries and pituitary glands are absolutely essential for the proliferation and differentiation of mammary epithelial cells, which are the predominant carcinogen-susceptible cell type in the mammary gland (Imagawa et al.,

1990; Nandi et al., 1995). They play a role in the development of both hormone- independent and hormone-dependent mammary cancers (Nandi et al., 1995). For example, estrogen may stimulate ER positive cells to produce a growth factor (such as

PRL) which may stimulate theirself and neighboring ER negative cells to proliferate.

Hormone imbalances are one of the major risk factors for the development of breast cancer. When the biochemical balance of estrogens is upset they can be potent mitogens to mammary epithelial cells and can influence tumor development by stimulating the local production of growth factors and their cognate receptors. Estrogens, epidermal growth factor (EGF), and heregulin have all been implicated in mammary carcinomas and evidence suggests other factors involved in mammary gland development may have a role too.

Role of Ovarian Hormones in Breast Cancer

Several lines of evidence indicate that hormones associated with females are involved in precancerous ductal and lobular carcinomas in situ. First, breast cancer is rare in men, accounting for less than 1% of all cancer cases. Second, a majority of the risk factors associated with breast cancer are specific to women. The risk of breast

14 cancer in women increases with age and has been found to be higher in women who experience early menarche (< 12 years old) or late menopause (> 55 years of age), indicating that the greater number of menstrual cycles (i.e. exposure to estrogen) increases the risk factor for breast cancer; whereas, later menarche and early menopause have been found to have protective effects. Early pregnancy (< 30 years of age) reduces the risk of developing breast cancer; whereas, late pregnancy (>30 years of age) or nulliparity increases the risk of breast cancer. The protective effect of parity likely reflects the reduction in the number of responsive cells to ovarian hormones and the long- term reduction in exposure to both estrogens and PRL. Breastfeeding also provides a degree of protection against breast cancer development. Studies have correlated the protective effects of lactation with the permanent lowering of circulatory levels of PRL; however, it should also be noted that prolonged lactation reduces the number of menstrual cycles. Third, removal of the ovaries reduces the risk of breast cancer and can cause regression of some tumors as does the irradiation of the ovaries (2000a; Metcalfe et al., 2004; Prowell and Davidson, 2004). Fourth, prevention studies using anti-estrogens, tamoxifen or raloxifene, have shown that they reduce the risk of breast cancer by about half (Dunn and Ford, 2000; Cauley et al., 2001; Fisher et al., 2005; Vogel et al., 2006).

The Link Between Estrogen and Prolactin

Estrogens have been shown to mediate there effects in part by regulating PRL signaling. Studies have shown that the overall pattern of serum PRL during the menstrual cycle resembles that reported for circulating 17β-estradiol (Vekemans et al.,

1977). Neuropeptides modulated by ovarian steroids during menstrual cycling and

15 lactation reduce the inhibitory effects of the hypothalamus, permitting enhanced secretion

of PRL from the pituitary (Freeman et al., 2000; Takahashi, 2004). Estrogen-induced

increases in galanin cause the lactotrophs to proliferate and release PRL. Estrogen-

stimulated synthesis and release of vasoactive intestinal peptide (VIP) from the

hypothalamus regulates the release of hypothalamic oxytocin which in turn causes PRL

secretion. Antiserum to estradiol or oxytocin has been shown to block the proestrous

surge of PRL and antiserum to galanin has been shown to inhibit basal PRL secretion.

These findings suggest that estrogen’s effect on hypothalamic oxytocin or galanin may be

responsible for its regulation of pituitary PRL secretion (Arey and Freeman, 1989).

Estrogen can also promote PRL transcription by activating MAPK and by

interacting with multiple estrogen response elements within the PRL promoter. This

indicates that estrogen may induce the expression of PRL outside the pituitary in ER positive cells. It has been suggested that one of the most important actions of estrogen is to modulate growth factor secretion in human breast cancer and that this may be the critical event in progression of cancer (Dickson and Lippman, 1986). Consistent with this, PRL and estrogen have been shown to cross-talk with each other to synergistically cause mammary gland proliferation. Both the ERα and the PRLR are co-expressed in

many breast tumors (Murphy et al., 1984; Ormandy et al., 1997b). Tamoxifen has been

shown to down-regulate PRLRs in breast cancer cells (de Castillo et al., 2004), which

may indicate that estrogen increases PRLR levels. Likewise, endogenous PRL has been

shown to increase ERα levels and the responsiveness of cells to estrogens, suggesting

that PRL acting via the PRLR amplifies the effects of estrogen (Gutzman et al., 2004).

16 If PRL amplifies the effects of estrogen than it undoubtedly should be associated with

breast cancer and should increase the risk.

Prolactin and Breast Cancer

Several lines of evidence support the involvement of PRL in human breast cancer

(Vonderhaar, 1998, 1999; Llovera et al., 2000a). First, PRL interacts synergistically with

estrogen and progesterone to promote cancerous growth and the receptors of these

ligands are co-expressed in primary breast cancers and have been shown to be cross-

regulated in mammary tumor cell lines (Ormandy et al., 1997b). Second, approximately

80% of human breast cancer cell lines tested transcribe PRL and respond to its growth

stimulatory action in vitro (Ginsburg and Vonderhaar, 1995; Das and Vonderhaar, 1997;

Ginsburg et al., 1997). Addition of exogenous PRL has a proliferative effect upon human

breast cancer cell lines (Malarkey et al., 1983) and has been reported to increase DNA

synthesis, protein synthesis and serves as a chemoattractant (Maus et al., 1999). Similarly,

PRL mRNA and protein have been detected in both normal and cancerous human

mammary tissue (Purnell et al., 1982; Fields et al., 1993) and has been shown to originate

primarily from the mammary epithelium (Clevenger et al., 1995). Third, biopsies have

revealed that a majority of human breast tumors have higher PRLR mRNA and protein

levels than surrounding and normal mammary tissue (Reynolds et al., 1997; Touraine et

al., 1998). Both PRL and the PRLR are expressed in the epithelium of a majority (>90%)

of benign and malignant human breast tissue samples (Reynolds et al., 1997), indicating

that PRL may be acting in an autocrine/paracrine manner (Clevenger et al., 1995).

Together, these observations support the hypothesis that: 1) estrogen elevates PRLRs in

17 malignant breast tissue making these cells highly sensitive to stimulation by PRL; 2) that

PRL released from the epithelial cells acts locally via the PRLR in an autocrine/paracrine

loop; and 3) that the mitogenic effects of PRL play a role in the progression to

malignancy (Fig. 1.2).

Several lines of evidence support the proposed role of PRL in tumor progression.

Sufferers of Parkinson's disease, a dopamine deficient population, have been shown to have lower rates of breast cancer and other types of malignancies (Lalonde and

Myslobodsky, 2003). In contrast, dopamine antagonists, which increase the release of

PRL, slightly but significantly increase the risk of breast cancer (Wang et al., 2002). In a study of 424 women, PRL levels above the median value were associated with a 2.1-fold increase in the risk of breast cancer and a 1.7-fold increase in risk for women with benign epithelial hyperplasia (Ingram et al., 1990). A large study showed that postmenopausal women with high levels of PRL had an increased risk of breast cancer regardless of ER status

(Hankinson et al., 1999). A highly significant correlation was found between tumor size and plasma PRL levels in premenopausal patients with breast cancer (Olsson et al.,

1985). The least favorable prognosis based on nodal status, tumor size and histological grade was significantly associated with the highest PRL levels (Wang et al., 1986). High

PRL levels before surgery or high PRL levels post-mastectomy were also significantly related to shorter disease-specific survival in postmenopausal women (Wang et al., 1995).

These findings clearly show that PRL levels are associated with breast cancer risk and progression.

18 1 PRLR expression

PRL

P P Unknown Jak2 Jak2 Jak2 Jak2 P Jak2Jak2 P 5 kinasekinase Kinase P P P P PP PP 2 PRL autocrine PRL Jak2 loop Jak2 PP PP PRL PP PP Jak2 PRL se

a Jak2 PP PRL PP kin Mitogenic pS PP PRL Jak2 Ub Signaling PP Jak2 PP PP PRL PRL

pY pY pY pY P P pY pY P P P P pY pY P P pY pY P P K K P P Jak2Jak2 P P Jak2Jak2 P Jak2 Grb2 P P P

3 2

PRL PRL ErbBfamily PRL receptors 4 Changes In PRLR 333 Crosstalk Isoform Ratios

Fig. 1.2 Mechanisms exploited by breast cancer cells to sustain PRLR activation

(1) Overexpresion of PRLR may hypersensitize cells to low levels of PRL; (2) autocrine/paracrine action of PRL can lead to constitutive activation of the PRLR; (3) activated PRLR can exploit ErbB2 via Grb2 to sustain MAPK activation; (4) modulation the PRLR isoforms can expand and intensify signaling; and (5) reducing the level or activity of an unidentified kinase reduces the degradation of the PRLR after activation leading to sustained signaling.

19 Further supporting a role for PRL are rodent models. PRL has been shown to contribute to the formation of chemical carcinogen-induced mammary carcinomas and to accelerate tumor progression in rodents (Manni et al., 1977; Welsch et al., 1977; Welsch and Nagasawa, 1977; Mershon et al., 1995; Tomblyn et al., 2005). Transgenic mice expressing rat PRL or human GH were shown to have a higher incidence of breast cancer due to signaling mediated via the PRLR (Bartke et al., 1994; Cecim et al., 1994; Wennbo et al., 1997).

Growth Hormone and Breast Cancer

In humans, GH can mediate its effects via PRLRs in addition to GHRs. For this reason it is not possible to discriminate between GHR and PRLR mediated effects in the mammary gland. Like PRL, expression of GH has been detected in the mammary gland

(Mol et al., 1995) and breast cancer cell lines (Decouvelaere et al., 1995) and its transcription is independent of Pit-1 (Lantinga-van Leeuwen et al., 1999). In situ RT-

PCR and in situ hybridization revealed that GH mRNA is predominantly expressed in the epithelial and myoepithelial ductal cells of the normal human mammary gland (Raccurt et al., 2002). In comparison to normal surrounding tissue, the expression of GH was found to be elevated in intraductal and invasive ductal carcinomas and was even higher in metastatic mammary carcinoma cells, occurring not only in proliferative epithelial cells but also stromal cells (Raccurt et al., 2002). Similarly the levels of GH protein have been shown to be elevated in carcinomas in comparison to normal tissue of the mammary gland (Mol et al., 1995; Raccurt et al., 2002) and are elevated in breast cancer patients

(Emerman et al., 1985). Conditions which increase GH increase the risk of malignancy

20 including breast cancer (Nabarro, 1987). Intramuscular injection of GH prior to

chemotherapy in patients with advanced breast cancer induced a 2-fold increase in the

proliferative activity of the tumor cells (Baldini et al., 1994). The changes in the local

concentration of GH are thought to be pivotal to determine the effects on the behavior of

the mammary epithelial cells.

The receptors for GH are expressed in normal human mammary tissue (Sobrier et al., 1993; Lincoln et al., 1998) and in breast cancers (Decouvelaere et al., 1995; Reynolds et al., 1997; Lincoln et al., 1998; Mertani et al., 1998; Gebre-Medhin et al., 2001).

Contrary to some reports (Emerman et al., 1985; Mertani et al., 1998), the expression of the GHR appears to be up-regulated in breast cancers in comparison to adjacent normal mammary tissue (Gebre-Medhin et al., 2001). The GHR is also up-regulated in stroma cells of the breast carcinomas (Gebre-Medhin et al., 2001).

Transfection of MCF-7 human breast cancer cells with the hGH gene resulted in a hyperproliferative state (Kaulsay et al., 1999) due to increased mitogenesis and decreased apoptosis (Kaulsay et al., 2001). GH-induced mitogenesis was mediated in part by increased expression of HOXA1 (Zhang et al., 2003), which increases the expression of cyclin D1 which is necessary for cell cycle progression and the up-regulation of the anti- apoptotic factor bcl-2 (Graichen et al., 2002; Zhang et al., 2003). Autocrine hGH also inhibits apoptosis by suppressing the expression of a positive regulator of the TGF-β pathway (PTGF-β) (Graichen et al., 2002) and inducing a negative regulator, c-ski, which associates with Smads to repress TGF-β inducible genes (Mertani et al., 2001). CHOP also appears to be upregulated by autocrine hGH and appears to protect the cells from

21 apoptosis too (Mertani et al., 2001). Stimulation with hGH resulted in reorganization of both the microfilament (Goh et al., 1997) and microtubule networks (Goh et al., 1998) and dramatically increased the rate of human mammary carcinoma cell spreading on a collagen substrate (Kaulsay et al., 2000). Autocrine production of hGH by MCF-7 cells alters the cellular morphology (loss of cell-cell contact, formation of multiple cellular protrusions, flattening of the cells) and was found to involve the downregulation

(occludin, γ-catenin) or relocalization (E-cadherin) of epithelial markers and acquisition of mesenchymal markers (fibronectin, vimentin) which stimulated cell migration and invasion via enhanced matrix metalloproteinase (MMP-2 and MMP-9) activity (Mukhina et al., 2004).

Similarly, transfection of immortalized but otherwise normal MCF-10A human mammary epithelial cells with the hGH gene caused a change in the morphology of the cells from epithelial to mesenchmyal spindle-shaped phenotype. Autocrine hGH significantly increased mitogenesis and abrogated apoptotic cell death. When cultured in

Matrigel the organization of cells resembled that of aggressive mammary carcinoma cell lines and supported anchorage-independent proliferation in the soft agar colony formation assay reminiscent of oncogenic transformation, this was not observed for exogenous hGH

(Zhu et al., 2005). This oncogenic transformation of immortalized MCF-10A cells induced by autocrine hGH was mediated primarily via the up-regulation of HOXA1 and involves Bcl-2 (Zhang et al., 2003). These drastic changes provide solid evidence that the acquisition of hGH production can lead to transformation of mammary epithelial cells and and tumor cell progression.

22 Current Therapies for Breast Cancer

Breast cancer predominantly occurs in women between 45 to 50 years of age and is the most common cause of death among women 18-54 years of age, primarily due to their tendency to metastasize (Baselga and Norton, 2002). The mortality rate for this disease is highest in the U.S. and the American Cancer Society projects the incidences of breast cancer within the U.S. will continue to rise while deaths associated with the disease will continue to decline. These trends have primarily been attributed to earlier detection of breast disease by mammograms and treatment of the disease with tamoxifen and/or chemotherapy.

Standard post-operative therapy for post-menopausal women with early-stage, hormone-receptor-positive breast cancers is five years of tamoxifen therapy rather than cytotoxic chemotherapy. Tamoxifen is an anti-estrogen that blocks mammary epithelial proliferation induced by natural estrogen cycling. It has been shown to significantly decrease the recurrence of breast cancers and to reduce the incidence of ER positive tumors in women with extremely high risk of breast cancer. Like chemotherapy, tamoxifen has some potentially serious side effects that include an increased risk of uterine cancer, thromboembolism, hot flashes, sexual dysfunction and cataracts (Fabian

and Kimler, 2001). In addition, tamoxifen has proven ineffective for the prevention or

treatment of ER negative tumors, which represent approximately 30% of all breast

cancers and result in 12,000-15,000 deaths each year (Fabian and Kimler, 2001). The

short comings of tamoxifen, the failure of conventional chemotherapy against advanced

invasive disease, and the projection of over 178,480 new cases of breast cancer and

23 40,910 deaths this year alone (American Cancer Society., 2006), suggests new

approaches to control and prevent breast cancer are needed. The major challenges facing

investigators of breast cancer are to develop alternative therapies that have fewer side

effects, and to develop drugs that act independently of ER status (Arun and Hortobagyi,

2002).

To address the short comings of tamoxifen, numerous therapeutic agents are

under development to prevent the cyclic production of estrogens and progesterone by the ovary (aromatase inhibitors, gonadotrophin-releaseing hormone agonists) and to block

the receptor mediated effects of estrogen (selective estrogen receptor modulators) and

progesterone (von Minckwitz and Kaufmann, 1998). These endocrine agents may one

day replace tamoxifen as the primary therapy for prevention of cancer reoccurrence in postmenopausal women since they appear to be more potent, have few if any agonistic effects on other organs, and show considerably fewer side effects than tamoxifen.

Numerous targets other than estrogen production and signaling are being explored for the development of therapeutics with potentially broader application since they may prove more useful for premenopausal women and may be effective on ER negative cancers. Potential targets include proteases, tyrosine kinases, receptors, and ligands which are aberrantly expressed and / or activated within cancerous tissues. Utilizing a variety of recombinant techniques and strategies, therapies have been developed which interfere with angiogenesis, modulate an immune response, induce apoptosis or necrosis, interfere with signal-transduction pathways, downregulate proto-oncogenes, block processes involved in metastasis and invasion, or a combination of the above.

24 The most widely explored therapies to date are ligand-based. Ligand-based therapies utilize ligands (peptides, growth factors, or steroids) that bind to receptors or antibodies that bind to antigens on the surface of target cells. Ideally, the receptor or antigen to be targeted is uniformly expressed within the target cell population; present at high levels on the surface of the target cells; and is not released or down-regulated. The use of ligand-based therapies overcomes many of the limitations of traditional chemotherapeutics by reducing the exposure of the normal tissues to the drug while increasing the dose achieved within the targeted cancerous tissue.

Monoclonal antibodies have proven more effective than small molecule tyrosine kinase inhibitors at disrupting growth factor signaling via the human epidermal growth factor receptor (HER) family of receptors. Strategies to modulate an immune response against breast cancer are being explored using adjuvants (non-specific immunotherapy); tumor vaccines (specific immunotherapy); interleukins (cytokine immunotherapy); and humanized Abs (passive immunotherapy). To interfere with angiogenesis and invasion the use of inhibitors of the action of vascular endothelial growth factor (anti-VEGF antibodies, soluble VEGF receptors, dominant-negative VEGF receptors, tyrosine kinase inhibitors, ribozymes, antisense); endogenous inhibitors of endothelial cells (endostatin, angiostatin, anti-angiogenic antithrombin); and inhibitors of matrix-metalloproteases

(small molecules) are currently being explored. Strategies targeting the extracellular portion of the receptors have run into fewer issues.

25

Fig. 1.3 Representation of PRLR gene expression in 25 tumors and normal contiguous breast tissues

Results are presented from lowest to highest level of PRLR gene expression in tumors. Note that a majority of breast cancers express elevated levels of PRLR in comparison to adjacent normal tissue. Adapted from Touraine et al., (Touraine et al., 1998).

26 Targeting Prolactin and Growth Hormone Receptor Signaling

At first glance there may appear to be no need to develop drugs that prevent the activation of the GHR and PRLR because the secretion of GH and PRL from the pituitary is under the inhibitory control of somatostatin and dopamine, respectively. Octapeptide, which mimics the activity of somatostatin, has been used to prevent GH secretion and to shrink pituitary tumors (e.g. pituitary adenomas) and treat disorders associated with these tumors (e.g. acromegaly). Similarly, bromocriptine and cabergoline, which mimic the activity of dopamine, have been used to prevent PRL secretion and to treat pituitary tumors (e.g. prolactinomas) and disorders associated with hyperprolactinaemia (e.g. amenorrhoea and infertility).

In rodents, significant inhibition of tumor progression has been achieved by removing their pituitary gland or by inhibiting the release of PRL from the pituitary using bromocriptine. In contrast, these approaches have had little success in humans. Removal of the pituitary gland has been reported to reduce the circulating levels of PRL by only

50-70% while other pituitary hormones were undetectable (Lachelin et al., 1977), indicating that extrapituitary production of PRL occurs in humans. Drugs that inhibit the release of PRL from the pituitary are ineffective against PRL secreted from extrapituitary sources such as the mammary epithelium.

Since biologically active PRL may be produced locally by human breast cancer cells (Ginsburg and Vonderhaar, 1995) and acts in an autocrine/paracrine manner within the mammary gland (Clevenger et al., 1995), therapeutics should be designed that directly antagonize PRL/PRLR signaling rather than PRL production or release from the

27 pituitary. In addition, some disorders may not be due to hyperprolactinaemia but rather

to elevated levels of the PRLR which may increase the sensitivity of cells to normal

levels of PRL. Elevation of PRLR occurs in a majority of breast cancers (Fig. 1.3) providing a strong argument in support of the development of a PRLR targeting antagonist.

No therapeutic agents are on the market that disrupt PRLR mediated signaling. Since

both PRL and the PRLR are expressed widely in breast cancer, it was postulated that targeted

inhibition of signaling via the PRLR would have potential as a therapy for breast cancer

(Reynolds et al., 1997). Drugs that target the PRLR represent a new class of therapeutics

for disrupting the effects of autocrine/paracrine PRL or hypersensitivity to pituitary PRL

due to over-expression of the PRLR.

In vitro, anti-hPRL antibodies and anti-sense RNA directed against the hPRL gene

have been used to successfully inhibit T-47Dco cell proliferation (Ginsburg and Vonderhaar,

1995; Ginsburg et al., 1997) as have antibodies against PRLR in multiple breast cancer cell

lines (Fuh and Wells, 1995). An antibody raised against the PRLR has also been shown to

reduce the incidence of carcinogen induced breast cancers in vivo in mice (Sissom et al., 1988).

These findings validate PRL/PRLR signaling as a target for the development of breast cancer

therapeutics.

Development of a Growth Hormone Receptor Antagonist

PRL and GH share approximately 40% amino acid sequence identity and a similar

tertiary structure as do the extracellular regions of the receptors to which they bind

(Forsyth and Wallis, 2002). Two separate receptor binding sites have been identified in

these hormones, each of which interacts with one receptor to form a one ligand two

28 receptor complex that activates signal transduction. Because of differences in affinity between the two binding sites, binding occurs sequentially with the higher affinity site interacting with the first receptor before the lower affinity site can interact with the second receptor (Cunningham et al., 1990b; Goffin et al., 1992; Fuh and Wells, 1995).

Even with the limited identity among these hormones, hGH can bind to both GHR and

PRLR with high affinity and the introduction of as few as eight amino acid mutations to hPRL confers the ability to bind to GHR with high affinity (Cunningham et al., 1990a). These findings suggest the few amino acid residues that are conserved among these hormones are involved in receptor binding and/or the maintenance of the overall confirmation of their four- helix bundle, which made them a good starting point for the development of an antagonist.

While studying the effect of mutations on the growth promoting activity of bovine

GH, it was discovered that mutations in its third α-helix either enhanced or suppressed growth without affecting the binding affinity of GH towards GHR (Chen et al., 1990;

Chen et al., 1991a; Chen et al., 1991c; Chen et al., 1991b; Chen et al., 1994; Chen et al.,

1995). It was further elucidated that Gly 119 of bovine GH (Chen et al., 1991b) or Gly

120 of human GH (Chen et al., 1994; Xu et al., 1995) is critical for retaining this growth promoting activity. Substitution of this Gly residue with an Arg residue results in a GH analog, hGH-G120R, which behaves as an antagonist (Fuh et al., 1992; Chen et al.,

1994). In vitro, GH has been shown to induce the tyrosine phosphorylation of STAT5

(Silva et al., 1993), hGH-G120R does not (Chen et al., 1994), and hGH-G120R competitively inhibits this GH-induced tyrosine phosphorylation (Silva et al., 1993).

29 In vivo, hGH-G120R transgenic mice have a dwarf phenotype, indicating that hGH-

G120R inhibits the binding of endogenous GH to GHRs (Chen et al., 1994).

It was originally proposed that hGH-G120R inhibits signal transduction by

preventing the dimerization of GHRs (Fuh et al., 1992; Silva et al., 1993). However, in contrast to the sequential dimerization due to differential affinity model, hGH-G120R

does not appear to block dimerization or affect internalization of porcine GHRs in mouse

L-cells but rather leads to the formation of dimeric complexes which are non-functional

(Harding et al., 1996).

Since hGH binds to PRLR in addition to GHR, it is not surprising that the hGH-

G120R antagonist also inhibited GH and PRL responses mediated through the PRLR

(Fuh et al., 1993; Fuh and Wells, 1995). However, it did not appear to always behave as

an antagonist in species other than human. At about 25 to 50-fold higher concentrations

than GH, it promoted the growth in hypophysectomized rats possibly via PRLRs (Mode

et al., 1996) and stimulated the proliferation of rat Nb2 cells which are responsive to lactogens (Dattani et al., 1995). Evidence suggests that these effects were species specific, reflecting different affinities towards rat PRLRs (Goffin et al., 1996).

A pure GHR antagonist, B2036, was developed which combines an amino acid substitution at position 120 of hGH with an additional eight amino acid substitutions to improve the affinity of binding Site 1 towards the GHR (Thorner et al., 1999) and reduce its affinity towards the PRLR (Goffin et al., 1999). A poly(ethyene glycol) modified form of B2036 (PEG-B2036) known as Somavert™ has recently been approved by FDA for the treatment of acromegaly (Trainer et al., 2000; Muller et al., 2001; Goffin and

30 Touraine, 2002). Another potential use for this pure GHR antagonist is the treatment of breast cancer (Kaulsay et al., 2001). However, it remains to be determined if GH mediates its effects primarily via GHRs or PRLRs in breast cancer.

Development of a Prolactin Receptor Antagonist

The Gly residue mutated in hGH was found to be conserved in both hPL and hPRL. Using an approach similar to the one used to develop hGH-G120R, the equivalent mutation was introduced in to hPL. hPL-G120R was found to antagonize the PRLR; however, it had a lower affinity than hGH-G120R for the receptor (Fuh and Wells, 1995).

To develop a high affinity antagonist specific for the PRLR, an equivalent Gly at position

129 of hPRL was converted to an Arg (Fig. 1.4). This single mutation altered the structure of binding Site 2, greatly reducing the binding affinity of hPRL-G129R for a second receptor, resulting in an analog that behaves as a PRLR antagonist (Fig. 1.5). hPRL-G129R was found to selectively bind to PRLRs on cell surfaces with high affinity, where it blocked the binding of PRL, and thus interfered with PRL signal transduction

(Goffin et al., 1996; Ramamoorthy et al., 2001).

It has been demonstrated that hPRL-G129R inhibits cell proliferation in human breast cancer cell lines in a dose-dependent manner by inducing apoptosis (Fig. 1.6)

(Chen et al., 1999). It has also been shown that hPRL-G129R competitively blocks hPRL-induced JAK2/STAT signal transduction (Cataldo et al., 2000) and by doing so up- regulates tumor suppressors such as TGF-β and the down-regulates proto-oncogenes such

TGF-α (Chen et al., 1999) and bcl-2 (Beck et al., 2002; Peirce and Chen, 2004); leading to the arrest of cell cycle progression and in some circumstances the induction of

31 apoptosis (Chen et al., 1999). In vivo, hPRL-G129R appears to be more cytostatic than

cytotoxic in human breast cancer xenograft-bearing nude mice (Fig. 1.7) (Chen et al.,

2002).

Potential Indications for Prolactin Receptor Antagonists

The most apparent indication for a PRLR antagonist is the treatment of breast

cancer. It may also have utility for the treatment of immune diseases such as rheumatoid

arthritis (Figueroa et al., 1997), systemic lupus erythematosus (McMurray et al., 1995;

Peeva et al., 2004), and Reiter's syndrome (Bravo et al., 1992). PRL has been associated

with these diseases, has been shown to function as a cytokine in lymphocytes, and may be produced by immune system cells (peripheral blood lymphocytes, thymocytes,

hematopoetic system cells) where it can act in an autocrine manner.

Potential Indications for Prolactin Receptor Agonists

There are many circumstances where the growth promoting effects of PRL and

GH may be beneficial. They may have utility for restoring the immune system in

immunocompromised patients. PRLR agonists may have potential agricultural uses to

maintain milk yields in dairy animals or overcome lactation insufficiencies in humans.

Similar to how GH increases insulin-like growth factor (IGF)- I levels, PRL increases

IGF-II levels in the mammary gland. In addition, PRL represses the production of IGF

binding protein 5 (IGFBP-5), a protein associated with involution. The induction of IGF-

II and repression of IGFBP-5 by PRL suggests a coordinate regulation of IGF-II

bioavailability during pregnancy and lactation stages.

32 A Glycine Arginine

B

hPRL G129R

Fig. 1.4 Model of hPRL and G129R complexes with the extracellular domain of the human PRLR

The G129R substitution (yellow) in helix 3 of hPRL disrupts the integrity of binding Site 2 without destabilizing the 3rd α-helix (A). The G129R mutation (yellow) fills the crevice between the 1st and 3rd helices and prevents functional receptor dimerization from occurring (B). Generated using Swiss Modeler (http://www.expasy.ch/swissmod/SWISS-MODEL.html).

33 2 1 3 4 2 1 2 2 3 1 1 4 3 3 4 4

2 1 2 3 X 4 1 4

P P P Jak2 Jak2 P Jak2 Jak2

P P P P hPRL G129R PRLR/PRLR PRLR/PRLR

Fig. 1.5 Schematic representation of G129R competitively inhibiting the binding of hPRL to PRLRs

The high affinity binding Site 1 of G129R (helices 1 and 4) readily binds to a PRLR; however, its altered binding Site 2 (X) prevents it from interacting with a second receptor in a manner that will induce signal transduction. Bound G129R prevents hPRL from occupying the PRLR homodimers and therefore behaves as an antagonist of PRL, GH or PL signaling via PRLRs.

34 Control G129R (Untreated) (500 ng/ml/24hr)

MDA-MB-134

T-47D

BT-474

MCF-7

Fig. 1.6 Apoptotic response of four PRLR-positive human breast cancer cell lines to G129R treatment

Apoptosis was measured using the TUNEL assay. Cells treated with 500 ng/ml of hPRL showed results similar to the control cells (not shown). The green fluorescence staining represents cells undergoing apoptosis (DNA fragmentation). Adapted from Chen et al. (Chen et al., 1999), with modifications

35

Fig. 1.7 Effects of hPRL and G129R on T-47D human breast cancer cell xenograft growth in nude mice

Thirty 6-7 week old Nuj/Nude mice (Jackson Laboratory) pre-implanted s.c. with slow-releasing E2 pellets (0.72 mg/60 day, Innovative Research of America, Inc.) were injected with T-47D cells (5 x 106) pre-mixed with Matrigel (BD Biosciences; San Diego, CA) into the mammary fat pad. One week after tumor cell inoculation, the mice were randomized into three groups and treated 5 times per week with either Matrigel, 200μg hPRL in Matrigel or 200μg G129R in Matrigel for seven consecutive weeks. The hPRL treated mice exhibited enhanced tumor growth (mean tumor volume = 202 + 62 mm3 versus 124 + 31 mm3 in the control mice). Mice treated with the G129R showed inhibition of tumor growth (mean tumor volume was 79 + 32 mm3). The final tumor weight in the three groups was also significantly different (p<0.05), (control, 100 mg + 2 mg; PRL, 121 mg + 5 mg, G129R, 65 mg + 16 mg). Adapted from Chen et al. (Chen et al., 2002).

36 CHAPTER TWO

DEVELOPMENT OF A PROLACTIN RECEPTOR TARGETING

FUSION TOXIN USING A PROLACTIN RECEPTOR ANTAGONIST

AND A RECOMBINANT FORM OF PSEUDOMONAS

EXOTOXIN A

Abstract

Human (h) prolactin (PRL) promotes the proliferation and differentiation of

mammary epithelial cells during mammary gland development and has been linked to

breast tumor development. The receptor for PRL (PRLR) is elevated in a majority of

human breast tumors, making cancer cells highly sensitive to the mitogenic and anti- apoptotic activity of hPRL. These findings provide the rationale for the development of

PRLR targeted breast cancer therapeutics. Previously, a hPRL receptor antagonist,

G129R, was developed that competitively binds to the PRLR resulting in growth inhibition and the induction of apoptosis in certain types of breast cancer cells. To further increase the potency of G129R, we fused G129R to a truncated form of

Pseudomonas exotoxin A, PE40, which lacks the cell recognition domain of the toxin but

retains the domains necessary for it to translocate into the cytosol and inhibit protein

synthesis. We postulated that the fusion of G129R to PE40-KDEL would (1) deliver the

recombinant toxin to breast cancer cells where the PRLR is overexpressed; (2) block PRL

signaling via the G129R moiety; and (3) inhibit protein synthesis via the PE40-KDEL

moiety. We demonstrate that the fusion toxin competitively binds to PRLR on T-47D

human breast cancer cells and inhibits signal transducers and activators of transcription 5

37 (STAT5) phosphorylation induced by hPRL. In addition, we show that G129R-PE40-

KDEL is selectively cytotoxic to breast cancer cell lines expressing the PRLR and that cell death is associated with the inhibition of protein synthesis and does not involve caspase mediated apoptosis.

38 Introduction

Human (h) prolactin (PRL) promotes the proliferation and differentiation of mammary epithelial cells during mammary gland development and is necessary for the stimulation of lactogenesis. The receptor for PRL (PRLR) is up regulated in mammary tissue during development and studies have shown that a majority of human breast tumors have higher PRLR levels than surrounding normal mammary tissue (Clevenger et al., 1995; Reynolds et al., 1997; Touraine et al., 1998). It has also been shown that PRL mRNA and protein are present in both normal and cancerous human mammary tissue

(Purnell et al., 1982; Fields et al., 1993) and that it originates primarily from the mammary epithelium (Clevenger et al., 1995). Based on these observations, it has been hypothesized (1) that elevated PRLR levels in malignant breast tissue make these cells highly sensitive to the mitogenic and anti-apoptotic activity of PRL; (2) that PRL released from the epithelial cells acts locally via the PRLR in an autocrine/paracrine manner; and (3) that alterations in PRL signaling may play a role in malignant transformation and/or tumor development.

Since both PRL and PRLR are expressed widely in breast cancers, it was proposed that targeted inhibition of signaling via the PRLR would have potential as a therapy for breast cancers (Reynolds et al., 1997). In vitro, anti-hPRL antibodies and anti-sense RNA directed against the hPRL gene have been used to successfully inhibit T-

47Dco cell proliferation (Ginsburg and Vonderhaar, 1995; Ginsburg et al., 1997) as have antibodies against hPRLR in multiple breast cancer cell lines (Fuh and Wells, 1995). An antibody raised against the PRLR has also been shown to reduce the incidence of

39 carcinogen induced breast cancers in vivo in mice (Sissom et al., 1988). These findings

validate PRL/PRLR signaling as a target for the development of breast cancer

therapeutics.

A PRLR antagonist, G129R, was developed by substituting a single glycine residue at

position 129 of hPRL with a bulky arginine residue (Goffin et al., 1996; Chen et al., 1999).

This amino acid substitution disrupts the structural integrity of the lower affinity binding site of

hPRL so that it retains the ability to bind to the first receptor with high affinity but can no

longer interact with a second receptor to initiate signal transduction (Chen et al., 1999; Llovera

et al., 2000b). It has been demonstrated that G129R binds to the PRLR with similar affinity as

hPRL and is able to competitively antagonize the effects of hPRL on hormone responsive human breast cancer cell lines (Chen et al., 1999). It has further been demonstrated that the

antagonist competitively blocks hPRL-induced Janus-activated kinase 2/signal transducers

and activators of transcription (JAK2/STAT) signaling and by doing so results in the up-

regulation of a tumor suppressor (TGF-β) and the down-regulation of a proto-oncogene (bcl-2);

leading to the arrest of cell cycle progression and in some circumstances the induction of

apoptosis (Cataldo et al., 2000; Ramamoorthy et al., 2001; Beck et al., 2002; Peirce and Chen,

2004). Since the blocking of PRL signaling by G129R appears to be more cytostatic than

cytotoxic to breast cancer cells in vitro and in vivo using mouse xenografts (Chen et al., 2002), we proposed to enhance the cytotoxicity of G129R by fusing it to a toxin that potently inhibits protein synthesis.

Pseudomonas exotoxin A (PE) is a bacterial toxin with three structurally and functionally different domains (Allured et al., 1986; Hwang et al., 1987). Domain I is

40 responsible for cell recognition, domain II for translocation, and domain III for inhibition

of protein synthesis (Fig. 2.1). PE binds to the membrane-associated heavy chain of α2-

macroglobulin receptor and undergoes receptor-mediated endocytosis (Fig. 2.2) (Morris et al., 1983; Kounnas et al., 1992; Zdanovsky et al., 1996). Within the endosome PE is acidified and processed (Jinno et al., 1988; Ogata et al., 1990; Siegall et al., 1991; Fryling et al., 1992); the low pH induces conformational changes in PE necessary for translocation (Idziorek et al., 1990) and for processing by furan (Chiron et al., 1994; Gu

et al., 1996), which cleaves PE between arginine 279 and glycine 280 (Ogata et al.,

1992). The mechanism by which the 37 kDa C-terminal fragment of PE translocates to

the cytosol remains unclear; however, both the N-terminal (Theuer et al., 1993;

Zdanovsky et al., 1993) and C-terminal regions of PE37 are necessary for translocation.

The C-terminal five amino acids, REDLK, resemble an endoplasmic reticulum retention

signal (Chaudhary et al., 1990) and are necessary for cytotoxicity (Theuer et al., 1993).

After removal of the C-terminal lysine, this C-terminal fragment is thought to exploit the

KDEL receptor retrieval system by binding to KDEL receptors present in the Golgi,

which retrograde transport the toxin to the lumen of the endoplasmic reticulum (Hessler

and Kreitman, 1997; Jackson et al., 1999). In the endoplasmic reticulum, PE may be

reduced, unfolded and reverse translocated across the membrane into the cytosol where it

can inactivate elongation factor 2 (EF-2), thereby inhibiting protein synthesis and ultimately

causing cell death.

41

Fig. 2.1 Structural domains of Pseudomonas exotoxin A

Receptor binding domain Ia (red), cytoplasm translocation domain II (blue), linker domain Ib (yellow), and ADP-ribosyl transferase domain III (green). Crystallographic structure from Wedekind et al. (Wedekind et al., 2001).

42 LK D RE

III

REDLK II III II I I

n uli lob Coated pit rog ac or -m pt endocytosis 2 ece α r furan

I II

II

K III Acidification of L Release of active ED R R ED LK fragment III endosome II pH 5.1 ADPR Inhibition of EF2 EF2 protein synthesis & cell death (Active) NAD+ NADH (Inactive)

Fig. 2.2 Diagram of intoxification by exotoxin A

The toxin is internalized by receptor-mediated endocytosis. In the endosome the toxin is cleaved by furan and the cytoplasmic translocation domain of the toxin is activated. The activated toxin is retrograde transported through the Golgi to the endoplasmic reticulum and enters the cytosol where it binds to and irreversibly ADP- ribosylates elongation factor 2 (EF2). Inactivation of EF2 prevents protein synthesis from occuring and results in cell death.

43 Here we describe the development of a fusion toxin intended to specifically target

and kill PRLR-positive breast cancers using G129R and a recombinant form of

Pseudomonas exotoxin A that lacks the cell recognition domain (Fig. 2.3). Our data demonstrates that G129R-PE40-KDEL drastically increases the potency of G129R even though

it has a lower affinity for the PRLR when fused with PE40-KDEL. We show that G129R-PE40-

KDEL has the attributes of a dual function protein; it retains the ability to antagonize the PRLR by partially inhibiting hPRL-induced STAT5 phosphorylation and it drastically reduces the viability of PRLR-positive human breast cancer cells by inhibiting protein synthesis.

44

IN L U G N B IN IO G O D N T N L N IO A I T G I T L D S O B A Y IN I R R IN C S IN B N C O O IN O O T A L B A R G MA P M S A I M L A - E O N M R O R T 2 C D A O P D P N α E R D D A R T A 1 253 364 400 608

Exotoxin A (PE66) Ia II Ib III REDLK 1 199

G129R-PE40-KDEL G129R II Ib III KDEL

Fig. 2.3 Diagram of G129R-PE40KDEL

The receptor binding domain of exotoxin A was replaced with G129R to target the toxin to the PRLR. The C-terminal five amino acids (REDLK) of exotoxin A were replaced with a more acitive endoplasmic reticulum retention signal (KDEL) to enhance the retrograde transport of the toxin via the KDEL receptor retrieval system.

45 Materials and Methods

Cell Lines and Reagents

The human breast cancer cell lines T-47D, MCF-7, BT-474, BT-483, MDA-MB-

134, MDA-MB-453, MDA-MB-231 and the murine breast cancer cell line 4T1 were

purchased from the American Type Culture Collection (Manassas, VA) and were

maintained in medium containing 10 μg/ml gentamicin at 37°C with humidity and 5%

CO2 unless otherwise mentioned. All media and supplements were purchased from

Invitrogen (Carlsbad, CA) unless otherwise mentioned. T-47D and MCF-7 cells were maintained in RPMI Medium 1640 supplemented with 10% fetal bovine serum (FBS).

BT-474 cells were maintained in RPMI Medium 1640 supplemented with 10% FBS, 1 mM sodium pyruvate, and 10 μg/ml insulin (Sigma, St. Louis, MO). BT-483 cells were maintained in RPMI Medium 1640 supplemented with 20% FBS, 2.5 mg/ml glucose

(Sigma), 10 mM HEPES, 1 mM sodium pyruvate, and 10 μg/ml insulin. MDA-MB-134 cells were maintained in Leibovitz’s L-15 Medium supplemented with 20% FBS without

CO2. MDA-MB-453 and MDA-MB-231 cells were maintained in Leibovitz’s L-15

Medium supplemented with 10% FBS without CO2. 4T1 cells were maintained in RPMI

Medium 1640 supplemented with 10% FBS, 2.5 mg/ml glucose, 10 mM HEPES, and

36.4 ml of 7.5% (w/v) sodium bicarbonate per liter of medium.

Pseudomonas exotoxin A was purchased from List Biological Laboratories

(Campbell, CA). PE40, which retains the enzymatic activity of the toxin but has low

cytotoxicity due to the lack of the cell recognition domain Ia (Seetharam et al., 1991),

was prepared and purified as described previously from the periplasmic fraction (Kondo

46 et al., 1988). Recombinant hPRL and G129R were prepared in house as previously

described (Chen et al., 2002). Cbz-Val-Ala-Asp-(Ome)-fluoromethyl ketone (zVAD-

fmk) was purchased from Enzyme Systems Products (Livermore, CA) and dissolved in

DMSO (Fisher Scientific, Pittsburgh, PA) to a final concentration of 50 mM, aliquoted,

and stored at -20ºC. Working solutions were made in culture medium immediately before use.

Construction of PRL-PE40-KDEL and G129R-PE40-KDEL Expression Vectors

Previously, hPRL and G129R were cloned by inserting their cDNAs between the

Nde I and Xho I sites of the expression vector pET22b(+) obtained from Novagen

(Madison, WI). Plasmid JH8, which encodes a recombinant form of Pseudomonas

exotoxin A (PE40), was obtained from the American Type Culture Collection. All

oligonucleotides were synthesized by Integrated DNA Technologies (Coralville, IA), all

restriction and modifying enzymes were purchased from New England BioLabs (Beverly,

MA), and all PCR reactions were performed using the Advantage-HF™ Kit (BD

Biosciences, Palo Alto, CA). Internal BamH I and Xho I sites were removed from the

® DNA encoding PE40 using the QuickChange XL Site-Directed Mutagenesis Kit

(Stratagene, La Jolla, CA). The C-terminal codons of hPRL, G129R, and PE40 were modified and restriction sites were added by PCR using oligonucleotides; the stop codons of hPRL and G129R were replaced with codons encoding Gly-Gly-Gly-Gly-Ser (G4S) and the codons of PE40 encoding REDLK were replaced with codons encoding KDEL

since it has been shown to significantly increase the cytotoxicity of PE and its derivatives

(Seetharam et al., 1991). The PCR products were ligated with pCR2.1 T/A cloning

47 vector and transformed into E. coli TOP10 cells obtained from Invitrogen. Positive clones were identified by plating the cells on Luria-Bertani (LB) agar plates containing

100 µg/ml ampicillin and 80 µg/ml 5-bromo-4-chloro-3-indolyl-bD-galactoside. The plasmids were isolated using the QIAprep® Spin Miniprep Kit (Qiagen, Valencia, CA)

and the Nde I-PRL-G4S-BamH I, Nde I-G129R-G4S-BamH I, and BamH I-PE40-KDEL-

Xho I DNA fragments were isolated by restriction digestion, separated by electrophoresis,

purified from the agarose gel using the QIAquick® Gel Extraction Kit (Qiagen), and

ligated with Nde I and Xho I cut pET22b(+) to create pET22b-PRL-PE40-KDEL and

pET22b-G129R-PE40-KDEL. Various oligonucleotides linkers were inserted at the

BamH I site located between G129R and PE40-KDEL to increase the distance between

the two moieties.

Production and Purification of Fusion Toxins

Chemically competent E. coli BL21 (DE3) Star cells (Invitrogen) were transformed with pET22b-PRL-PE40-KDEL or pET22b–G129R-PE40-KDEL and

propagated overnight at 37°C with agitation in LB broth containing 100 µg/ml ampicillin.

The next morning 160 ml of each of the starter cultures was used to inoculate four liters

of LB broth and the cultures were grown to an O.D.600 of 0.9 before induction of protein

production with 1 mM isopropyl β-thiogalactoside (Alexis Biochemicals, San Diego,

CA). After 5 h of induction the cells were harvested by centrifugation at 5,000 x g for 5 min and resuspended in Solution 1 (0.2M NaPO4; 10 mM EDTA, pH 8.0; 0.5% Triton X-

100; pH adjusted to 8.0) to which 0.1 mg/ml lysozyme was added. The cells were

incubated at room temperature for 1 h to weaken the cell membranes before the cells

48 were sheared on ice with five 1-min ultrasonic pulses (30% duty at setting 7) using a 550

Sonic Dismembrator (Fisher Scientific). The insoluble fraction containing the inclusion

bodies was recovered by centrifugation at 12,000 x g for 30 min at 4ºC and resuspended

in Solution 2 (0.2M NaPO4; 10 mM EDTA, pH 8; 0.5% Triton X-100; 1 M urea; pH adjusted to 7.0). The inclusion bodies were collected by centrifugation at 12,000 x g for

30 min at 4ºC and were solubilized and denatured in Solution 3 (0.2M NaPO4; 8 M urea;

1% v/v β-mercaptoethanol; pH adjusted to 8.0). The proteins were refolded by dialysis against

baths of 20 mM Tris·HCl, pH 8.0 containing decreasing amounts of urea (4M urea, 2 M urea, 0

M urea, 0 M urea). The refolded fusion toxins were filtered through 0.45 μm filters and

purified by anion-exchange chromatography using a 5 ml HiTrap Q Sepharose XL column

(Amersham Biosciences, Piscataway, NJ) on an ÄKTA fast protein liquid chromatography

system (Amersham Biosciences). The purified fusion toxins were separated by reducing SDS-

PAGE along with BSA standards (Pierce, Rockford, IL) and stained with SYPRO® Orange

(Molecular Probes, Eugene, OR) to determine the concentration and purity of the fusion

toxins. The identity of the fusion toxins were confirmed by Western blotting.

Western Blot Analysis

PE, G129R, and G129R-PE40-KDEL were separated on a 12% polyacrylamide gel

by reducing SDS-PAGE. The proteins were transferred by Western blot to a Hybond™

enhanced chemiluminescence (ECL) nitrocellulose membrane (Amersham Biosciences)

at 16 W for 1 h. The nitrocellulose membrane was incubated in a blocking solution containing TBS with 0.05% Tween 20 (TBS-T) and 5% milk for 1 h at room temperature.

The membrane was incubated overnight at 4°C with agitation in blocking solution

49 containing a 1:2000 dilution of anti-exotoxin A (List Biological Laboratories). The blots were washed three times with TBS-T, and incubated for 2 h at room temperature with agitation in blocking solution containing a 1:2000 dilution of swine anti-goat horseradish peroxidase (HRP) conjugate (Roche, Indianapolis, IN). Blots were washed three times with TBS-T and incubated for 1 min with ECL™ Western Blotting Detection Reagents

(Amersham Biosciences). Blots were exposed to Biomax MR film (Kodak, Rochester,

NY) and the film was developed with a Konica SRX-101A processor (Konica Minolta

Medical Imagining, Wayne, NJ). The membrane was stripped and re-probed with a

1:4000 dilution of rabbit anti-hPRL antiserum (National Hormone and Pituitary Program,

NIH, Bethesda, MD) and 1:2000 dilution of goat anti-rabbit conjugated HRP (Bio-Rad,

Hercules, CA).

Radio-Receptor Binding Assay

One million cells per well of T-47D human breast cancer cells were seeded in six- well tissue culture plates and grown to confluency. Cells were starved in serum-free

RPMI Medium 1640 for 1 h, and incubated for 2 h at room temperature in serum-free medium containing 5 x 104 cpm of 125I-labeled hPRL (specific activity, 40 µCI/µg; NEN

Perkin-Elmer, Boston, MA) with or without various concentrations of hPRL, G129R, or

G129R-PE40-KDEL. Cells were washed three times with serum-free RPMI Medium

1640 and were lysed in 0.5 ml of 0.1 N NaOH/1% SDS. Bound radioactivity was determined using a Beckman LS6500 Liquid Scintillation Counter calibrated to detect

125I using a wide spectrometer window. The percentage of specific binding was calculated and compared among the samples.

50 STAT5 Phosphorylation Assay

T-47D cells were grown to 90% confluency in six-well plates in RPMI Medium

1640 supplemented with 10% (v/v) charcoal-stripped fetal bovine serum (CSS; HyClone,

Logan, UT) and 10 μg/ml gentamycin. The cells were depleted for 1 h in RPMI Medium

1640 supplemented with 0.5% CSS and then treated for 20 min with various

concentrations and combinations of hPRL, G129R, PRL-PE40-KDEL, and G129R-PE40-

KDEL. The cells were washed with ice cold phosphate buffered saline (PBS) and lysed

in 200 μl of Lysis Buffer (20 mM Tris⋅HCl, pH 7.4; 1% Igepal CA630; 6 mM

deoxycholic acid; 150 mM NaCl; 1 mM EDTA, pH 8.0) containing protease inhibitors (1

µg/ml aprotinin; 1 µg/ml leupeptin; 170 µg/ml PMSF; 180 µg/ml sodium orthovanadate).

The lysates were agitated for 10 min on an orbital shaker, transferred to 1.5 ml tubes,

passed six times through 21-gauge needles, and incubated on ice for 20 minutes. The

lysates were clarified by centrifugation for 10 min at 12,000 x g at 4ºC and 30 µl of the

supernatants were used for Western blot analysis as described previously using 4-15% polyacrylamide gels. Blots were performed with 1.5 µg/ml mouse anti-phosphate

STAT5A/B antibody (Upstate, Lake Placid, NY) and a 1:2000 dilution of goat anti- mouse-HRP conjugate (Bio-Rad). The membranes were re-probed with 3 µg/ml rabbit

anti-STAT5 polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) and a

1:2000 dilution of goat ant-rabbit HRP conjugate (Bio-Rad).

51 Protein Synthesis Inhibition Assay

Approximately 1 x 105 cells were seeded in each well of a 24-well plate in a final

volume of 0.5 ml of appropriate growth medium and the cells were incubated overnight.

The medium was replaced with 0.5 ml of appropriate treatments and incubated for 44 h

after which each well was supplemented with 1 µCi of 3H-leucine (Amersham

Biosciences) in 100 µl of medium and incubated an additional 4 h. The monolayers were washed with 1 ml PBS and the protein was precipitated with 0.5 ml of 26% trichloroacetic acid (TCA) overnight at -20 °C. The monolayers were solubilized with

0.5 ml of 0.2 nM NaOH and loaded onto GF/C glass fiber filters (Whatman, Clifton, NJ)

pre-rinsed with 1 ml of 10% TCA. The filters were washed twice with 1 ml of 10% TCA

and once with 3 ml of 95% ethanol and dried by vacuum. The dried filters were placed in

scintillation vials and aqueous scintillation cocktail was added before counting with a liquid scintillation counter. The amount of 3H-leucine incorporated into precipitable protein was determined and expressed as a percent of control.

Cell Proliferation Assay

The various cell lines to be assayed were seeded in 96-well culture plates at a

density of 5,000 cells per well in 200 µl of appropriate culture medium and grown as

recommended. After incubation for 24 h, the medium was discarded and replaced with

200 µl of medium containing various concentrations of PE, G129R, or G129R-PE40-

KDEL and the cells were incubated for an additional 48 h. The culture medium was

discarded and replaced with 100 µl of 3-(4,5-dimethylthiazol-2-yl)-5-(3-

52 carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium/phenazine methosulfate

(MTS/PMS) solution diluted in PBS at a 1:6 ratio (CellTiter 96 AQueous non-radioactive cell proliferation kit; Promega, Madison, WI). The relative viability of the cells was determined by colorimetric measurement of the reduction of MTS by the living cells using a Benchmark microplate reader (Bio-Rad) measuring absorbance at 490 nM. All assays were performed in quadruplicate and the relative cell viability was calculated as a percentage of control treatments.

Poly-(ADP-Ribose)-Polymerase (PARP) Cleavage Assay

Approximately 3 x 105 MCF-7 cells were resuspended in 5 ml of RPMI Medium

1640 supplemented with 10% CSS and seeded into T-25 flasks and allowed to grow 24 h.

For experiments that included zVAD-fmk, 50 μM of the peptide was added 1 h prior to treatment with 1 µg/ml of PE, hPRL, G129R, or G129R-PE40-KDEL. After 24 h treatment, the medium and cells were collected and transferred to 5 ml Falcon tubes and centrifuged at 5,000 x g. The pellets were washed once with ice-cold PBS and resuspended in 100 μl of Lysis Buffer with inhibitors and lysed for 30 min at 4ºC. The lysates were clarified by centrifugation for 10 min at 12,000 x g at 4ºC and the protein content of the lysates were assayed using the Coomassie Plus Protein Assay Reagent

(Pierce). Thirty micrograms of the lysates were resolved on a 10% gel by reducing SDS-

PAGE and were Western blotted as previously described using a blocking solution containing 10% milk. A 1:2000 dilution of rabbit anti-PARP (Roche) and 1:2000 dilution of goat anti-rabbit HRP conjugate (Bio-Rad) were used.

53 Bax Translocation and Cytochrome C Release Assay

MCF-7 cells were resuspended in RPMI Medium 1640 supplemented with 10%

CSS at a concentration of 1 x 106 cells/ml. Two milliliters of the suspension was used to

seed each well of a six-well culture plate. After 24 h incubation, the medium was

replaced with medium containing 1 µg/ml G129R, PE, G129R-PE40-KDEL, or TNF-α.

After 24 h, the cells were washed with PBS, collected in microcentrifuge tubes, and permeabilized in 200 µl of ice cold Digitonin Lysis Buffer (75 mM KCl, 1 mM

NaH2PO4, 8 mM Na2HPO4, 250 mM sucrose, 190 µg/ml digitonin). After 5 min on ice,

the cells were centrifuged for 5 min at 16,000 x g at 4°C and the supernatants were

transferred to new microcentrifuge tubes. Laemmli buffer was added to the supernatants

containing the cytosolic fraction and to the pellets containing the membrane fractions and

the samples were heated at 98°C for 5 min before being separated by reducing SDS-

PAGE using a 15% polyacrylamide gel. Western blot analysis was performed as

described previously. A 1:200 dilution of rabbit anti-Cytochrome C (Cell Signaling

Technology, Danvers, MA), a 1:200 dilution of rabbit anti-BAX (Santa Cruz) and a

1:2000 dilution of goat anti-rabbit HRP conjugate (Bio-Rad) were used.

54 G4S(TLRDLFDRAVVLSHYIHNLSSEMFSEGS)2 G4STLRDLFDRAVVLSHYIHNLSSEMFSEGS (G4S)3 G4S Nde I RPE 129 40 K G DE L Xho I

f I pET22b 1 c

a

L G129R-PE40-KDEL

A A

m m

p p

r r

Fig. 2.4 Schematic representation of plasmid used to produce G129R-PE40-KDEL in E. coli

The cDNA encoding G129R was fused to the DNA encoding an enzymatically active form of PE (PE40-KDEL) with a BamH I site separating the two. In an effort to increase the activity of the fusion toxin, linkers of increasing length were inserted at the BamH I site between G129R and PE40-KDEL to decrease steric hindrance caused by their close proximity. The diagram above shows the amino acids encoded by the linkers inserted between G129R and PE40-KDEL. The fusion toxin used for all studies is shown highlighted. The selection was based on its superior potency at inhibiting protein synthesis.

55 Results

Construction, Expression, and Purification of PRL-PE40-KDEL

and G129R-PE40-KDEL

To target the PE40 toxin to the PRLR, we fused cDNA encoding hPRL or G129R

to the N-terminus of DNA encoding PE40 and replaced the DNA sequence encoding the native endoplasmic reticulum retention sequence (REDLK) of the toxin with a sequence encoding KDEL to increase its cytotoxicity (Fig. 2.4). Linkers of increasing length were

inserted between hPRL or G129R and PE40-KDEL since the fusion toxins had little

cytotoxicity with the G4S linker (Fig. 2.4). For all studies we used the fusion toxins

containing the Gly-Gly-Gly-Gly-Ser-Thr-Leu-Arg-Asp-Leu-Phe-Asp-Arg-Ala-Val-Val-

Leu-Ser-His-Tyr-Ile-His-Asn-Leu-Ser-Ser-Glu-Met-Phe-Ser-Glu-Gly-Ser (helix1) linker

since they showed greater potency. PRL-PE40-KDEL and G129R-PE40-KDEL were

produced in E. coli and found to accumulate in the insoluble fraction as inclusion bodies.

The inclusion bodies were isolated and denatured and the fusion toxins were refolded and

purified by anion-exchange chromatography to yield preparations of greater than 90% purity.

Purified G129R-PE40-KDEL was analyzed by reducing SDS-PAGE and was found to have a

size corresponding to the predicted molecular mass of 66 kDa (Fig. 2.5A). Western blots with

antibodies that detect the G129R and PE40-KDEL moieties confirmed the identity of purified

protein to be G129R-PE40-KDEL (Fig. 2.5, B and C).

56 A BC 1 2 3 M 1 2 3 M 1 2 3 M 66 kDa

23 kDa

SDS-PAGE WB w/ anti-hPRL WB w/ anti-PE

Fig. 2.5 Reducing SDS-PAGE (A) and Western blot analysis (B and C) of purified PE (66 kD), G129R (23 kD), and G129R-PE40-KDEL (66 kD) in lanes 1 thru 3, respectively

The purity of G129R-PE40-KDEL was confirmed after reducing SDS-PAGE by staining with SYPRO® Orange (A). The identity of the protein was confirmed by Western blotting with antibodies that detect the G129R (B) and PE40-KDEL moieties (C) as indicated by arrows.

57 100

80 I-PRL I-PRL 125 60

40

PRL 20 % Specific Binding of G129R

G129R-PE40-KDEL

00.1110100

[Ligand] nM

Fig. 2.6 Binding of hPRL, G129R, and G129R-PE40-KDEL to PRLRs on the surface of T-47D human breast cancer cells

The ability of hPRL, G129R and G129R-PE40-KDEL to bind to PRLRs was determined by measuring their ability to compete with 125I-labeled hPRL for binding to the surface of the cells. Each data point represents the mean + SD of at least three independent experiments.

58 G129R-PE40-KDEL Binds to Prolactin Receptor

The ability of G129R and G129R-PE40-KDEL to specifically bind to PRLRs

expressed on the surface of T-47D human breast cancer cells was demonstrated using a

radio-receptor binding assay. Figure 2.6 shows that increasing doses of hPRL, G129R, or

125 G129R-PE40-KDEL were able to competitively reduce the binding of I-hPRL to

PRLRs. This confirmed that G129R-PE40-KDEL retains the ability to bind to PRLRs;

125 however, the higher percent specific binding of I-hPRL suggests that G129R-PE40-

KDEL has a reduced affinity for PRLRs in comparison to hPRL and G129R. The EC50s for hPRL, G129R and G129R-PE40-KDEL were determined to be 2.7 nM, 6.3 nM, and

76.5 nM, respectively.

PRL-PE40-KDEL Activates STAT5 and G129R-PE40-KDEL Inhibits STAT5

Phosphorylation

Human PRL induces PRLR dimerization and leads to the phosphorylation of

STAT5 via receptor associated JAK2. The PRL-PE40-KDEL fusion toxin retained the

ability of the hPRL moiety to bind to PRLRs on T-47D human breast cancer cells and

activate STAT5 phosphorylation (Fig. 2.7), albeit with only a fraction (1/10th) of the

activity of hPRL. The ability of this fusion toxin to activate STAT5 suggested that

G129R-PE40-KDEL should be able to bind to and antagonize the PRLR. Like G129R,

G129R-PE40-KDEL did not activate STAT5 (Fig. 2.8A) and was able to compete with

hPRL for occupancy of the PRLR as shown by its ability to partially block STAT5

phosphorylation (Fig. 2.8B). G129R-PE40-KDEL was unable to completely block the

action of hPRL at a ratio of 1:100 and had antagonistic activity less than 1/10th that of

59 G129R, which corresponds with the reduced activity observed for PRL-PE40-KDEL.

These findings are consistent with the fusion toxins retaining the ability to bind to PRLR

but with a reduced affinity.

The PE40-KDEL Moiety of G129R-PE40-KDEL Retains the Ability to Cleave PARP

Having confirmed that the G129R moiety retains its ability to bind and block

receptor activation, we wanted to determine if the PE40-KDEL moiety retained its

activity. Previously, it has been shown that treatment of MCF-7 cells with PE results in

the activation of caspases that cleave PARP into 89 kDa and 24 kDa fragments (Keppler-

Hafkemeyer et al., 1998). Figure 2.9A shows that PE cleaved this DNA repair enzyme and that this cleavage was inhibited by treating the cells with the broad spectrum caspase inhibitor, zVAD-fmk. Neither hPRL nor G129R caused cleavage of PARP, confirming that the ability of G129R-PE40-KDEL to cleave PARP is due the PE40-KDEL moiety of

the fusion toxin (Fig. 2.9A).

A cell proliferation assay was used to examine the cytotoxic effects of PE and

G129R-PE40-KDEL on MCF-7 cells in the presence or absence of zVAD-fmk. Both PE

and G129R-PE40-KDEL were found to have cytotoxic effects, confirming that at least a

portion of the fusion toxin was taken up and processed correctly by MCF-7 cells (Fig.

2.9B). Interestingly, pretreatment with zVAD-fmk had little effect on the viability of the

cells, indicating that caspase mediated apoptosis does not play a major role in cell death associated with protein synthesis inhibition (Fig. 2.9B).

60 PRL PRL-PE40-KDEL nM: - 4 4 8 20 40 80 120

p-STAT5

Fig. 2.7 Stimulation of STAT5 phosphorylation by hPRL and PRL-PE40-KDEL using T-47D human breast cancer cells

The ability of hPRL and PRL-PE40-KDEL to bind to PRLRs and activate STAT5 was determined by treatment of the cells with 4 nM of hPRL or increasing concentrations of PRL-PE40-KDEL. Phosphorylated STAT5 (p-STAT5) was detected by Western blot using an anti-phospho STAT5 antibody and is indicated by the arrow.

61 A

PRL G129R G129R-PE40-KDEL - 4nM 40 nM 40 nM p-STAT5

STAT5

B

PRL PRL:G129R PRL:G129R-PE40-KDEL - 4 nM 1:10 1:100 1:10 1:100 p-STAT5

STAT5

Fig. 2.8 Competitive inhibition of STAT5 phosphorylation by G129R and G129R- PE40-KDEL in T-47D human breast cancer cells

The inability of G129R and G129R-PE40-KDEL to activate STAT5 (A) and their ability to competitively block hPRL-induced STAT5 activation (B) was determined by Western blot with antibodies against phosphorylated STAT5 (p-STAT5) and STAT5.

62 A

(-) PE PRL G129R G129R-PE40KDEL 50 µM zVAD-fmk - + - + - + - + - +

24 kDa PARP fragment

B

no zVAD-fmk 50 µM zVAD-fmk

100 80 60 40

(% of Control) 20

Relative Cell Viability Viability Cell Relative 0 Control PE G129R-PE40-KDEL

Fig. 2.9 Examination of PARP cleavage and the effect of zVAD-fmk on the viability of MCF-7 breast cancer cells treated with PE, hPRL, G129R, or G129R-PE40- KDEL

Cells were treated for 24 h with 1 µg/ml of PE, hPRL, G129R, or G129R-PE40- KDEL in the absence or presence of a broad spectrum caspase inhibitor, zVAD-fmk (A and B). The cell lysate was separated by reducing SDS-PAGE, and immunblotted with anti-PARP antibody. The caspase mediated cleavage of PARP is indicated by the presence of a 24 kDa PARP fragment as shown by the arrow (A). The role of activated caspases in cytotoxicity caused by PE or G129R-PE40-KDEL was examined by measuring the reduction of MTS by living cells in the absence or presence of zVAD-fmk (B). Data presented is representative of the mean + SD from assays performed in quadruplicate on two separate occasions.

63 G129R-PE40-KDEL Triggers Bax Translocation to the Mitochondria and Induces

Cytochrome C Release

Cytochrome C release is an indicator of mitochondrial damage and is regulated in part by the translocation of Bax from the cytosol to the mitochondria. To assess Bax translocation to the mitochondria and cytochrome C release from the mitochondria we examined their levels in the cytosolic and membrane fractions of MCF-7 cells treated with G129R, PE, G129R-PE40-KDEL, or TNF-α (Fig. 2.10). Densitometric analysis of the Western blots revealed no difference in the total amount (cytosolic fraction + membrane fraction) of Cytochrome C and Bax between treatments, but a difference in their subcellular location. Cytochrome C was undetectable in the cytosolic fraction of untreated or G129R treated cells. Consistent with Cytochrome C release, treatment with

PE or G129R-PE40-KDEL increased cytosolic levels while reducing mitochondrial levels

of Cytochrome C. Consistent with Bax translocation, an increase in mitochondrial levels and a reduction in cytosolic levels of Bax were observed in PE and G129R-PE40-KDEL treated cells.

G129R-PE40-KDEL Inhibits Protein Synthesis and Decreases the Viability of Breast

Cancer Cell Lines Expressing PRLR

Human breast cancer cell lines expressing various levels of PRLR were treated for

48 h with increasing doses of G129R-PE40KDEL and the inhibition of protein synthesis

(Fig. 2.11A) and its effect on the relative viability of the cells (Fig. 2.11B) was

determined by measuring the incorporation of 3H-leucine into perceptible protein and by

colorimetrically measuring the reduction of MTS by living cells, respectively.

64 L DE K 40 E -P R R α 9 9 - ) 2 2 F (- 1 E 1 N G P G T

CYTOSOLIC Cyt C FRACTION Bax

Cyt C MEMBRANE FRACTION Bax

Fig. 2.10 The effect of G129R, PE, and G129R-PE40-KDEL on the subcellular location of Bax and Cytochrome C in MCF-7 human breast cancer cells

Cells were treated for 24 h with 1 µg/ml of G129R, PE, G129R-PE40-KDEL, or TNF-α. The membrane fractions containing the mitochondria and the cytosolic fractions were separated by reducing SDS-PAGE and immunblotted with antibodies against Cyt C and Bax. A reduction in the membrane fraction of Cyt C with an increase in the cytosolic fraction of Cyt C is in accordance with Cytochrome C release. A reduction in the cytosolic fraction of Bax with an increase in the membrane fraction of Bax is in accordance with Bax translocation.

65 A

PRLR-Negative 100 MDA-MB-231 MDA-MB-453

PRLR-Positive 50 T-47D MDA-MB-134 BT-474 MCF-7 Protein Synthesis of Control) (% 0 110100

[G129R-PE40-KDEL] (nM) B PRLR-Negative 100 MDA-MB-453 MDA-MB-231 PRLR-Positive 50 BT-483 MDA-MB-134 T-47D

Cell Viability (% of Control) (% Cell Viability BT-474 MCF-7 0110100

[G129R-PE40-KDEL] (nM)

Fig. 2.11 The effect of G129R-PE40KDEL on protein synthesis and the relative viability of human breast cancer cell lines expressing different levels of PRLR

Cells were treated for 48 h with increasing doses of G129R-PE40KDEL and the inhibition of protein synthesis was determined by measuring the incorporation of 3H- leucine into perceptible protein (A). The relative number of viable cells was determined colorimetrically by measuring the reduction of MTS by living cells (B). Data points presented are representative of the mean + SD from three assays performed in quadruplicate.

66 Cells expressing PRLRs were markedly more sensitive to G129R-PE40-KDEL than cell lines such as MDA-MB231 and MDA-MB453 which lack or have few PRLRs (Fig. 2.11,

A and B).

The Inhibitory Effect of G129R-PE40-KDEL is Mediated via the G129R Moiety and is Due to the Inhibition of Protein Synthesis by the PE40-KDEL Moiety

Previously, we have shown that G129R can inhibit the proliferation of breast cancer cells. At the concentrations used, G129R appeared to be cytostatic and did not decrease the viability of T-47D, MCF-7, or BT-474 breast cancer cells which have elevated levels of the PRLR (Fig. 2.12). Even with the greater than ten-fold reduced receptor binding affinity and reduced antagonistic activity as indicated by STAT5 phosphorylation, G129R-PE40-KDEL was extremely cytotoxic in comparison to G129R.

Fusion of G129R with PE40-KDEL drastically decreased the relative number of viable cells at concentrations well below those observed for G129R, suggesting that the increased cytotoxicity was due to the PE40-KDEL moiety of the fusion toxin.

To confirm that the killing observed with G129R-PE40-KDEL was not due to non- specific binding and to confirm that it was due to the inhibition of protein synthesis by

3 the PE40-KDEL moiety, we measured the incorporation of H-leucine into perceptible protein and the relative cell viability in response to G129R-PE40-KDEL in the absence or presence of competitors (Fig. 2.13, A and B). Treatment of T-47D cells with increasing doses of G129R-PE40-KDEL decreased protein synthesis (Fig. 2.13A); this inhibition was reduced by co-treatment of the cells with excessive amounts of G129R, suggesting that the cytotoxic effects of the toxin rely heavily on cell binding. Similarly, cells co-treated with anti-

67 PE had no reduction in protein synthesis, indicating that the PE40-KDEL moiety was

responsible for the inhibition of protein synthesis. Treatment of T-47D cells with

increasing doses of G129R-PE40-KDEL also decreased the relative number of viable cells

(Fig. 2.13B). Co-treatment of the cells with G129R-PE40-KDEL and G129R or hPRL revealed that the relative number of viable cells was increased by G129R and hPRL.

This confirmed that the inhibition of protein synthesis was mediated through the PRLR and that the decreased viability was due to the PE40-KDEL moiety. We also examined the

non-specific effects of G129R-PE40-KDEL in mouse 4T1 cells which are extremely

sensitive to PE and the non-specific effects of PE40; G129R-PE40-KDEL was found to

have lower cytotoxicity than PE40 by itself, consistent with the fusion toxin having lower

non-specific cytotoxicity than PE40 alone (Fig. 2.14).

68 T-47D MCF-7 BT-474

100 100 100

50 50 50 (% of Control) Relative Cell Viability Cell Viability Relative

0 1 10 100 0 1 10 100 0 1 10 100 [Ligand] nM

Fig. 2.12 Comparison of the response of human breast cancer cells to G129R (‘), G129R-PE40-KDEL ( ), and PE (U)

The relative viability of three cell lines was examined after 48 h treatment by measuring the reduction of MTS by living cells. Data points presented are representative of the mean + SD from assays performed in quadruplicate on three separate occasions.

69 A

100

No Competition 50 + G129R + Anti-PE Ab

0

Protein Synthesis (% of Control ) Control of (% Synthesis Protein 03.333

[G129R-PE40-KDEL] (nM)

B

100

50

No competition + G129R

Relative Cell Viability (% Control) Viability of Cell Relative + PRL

0 1 10 100 [G129R-PE40-KDEL] (nM)

Fig. 2.13 Prevention of G129R-PE40-KDEL-induced protein synthesis inhibition and cytotoxicity by hPRL, G129R, and anti-PE

The ability of G129R-PE40-KDEL to inhibit protein synthesis and decrease the relative number of viable T-47D cells in the absence or presence of competitors, G129R (4 μM), PRL (4 μM) or anti-PE antibody (5 μL) was determined by measuring the incorporation of 3H-leucine into precipitable protein (A) and by measuring the reduction of MTS by living cells (B). Data points presented in (B) are representative of the mean + SD from assays performed in quadruplicate on two separate occasions.

70 Discussion

Most growth factors and hormones are taken up by receptor-mediated

endocytosis, a broadly named process that can involve a variety of different

internalization pathways. For this reason, the ligand chosen for making a fusion toxin is

not only important for determining its host-range, but also the ability of the toxin to be

properly internalized and processed. Numerous fusion toxins have been constructed by gene

fusion or conjugation of PE40 with cytokines, growth factors, hormones, or antibodies. Most of

these fusion toxins have proven to be cytotoxic to the targeted cells in the picomolar range.

Based on the potency of PE based fusion toxins we wanted to examine the potential of a fusion

toxin that targets the PRLR. Studies have shown that a majority of human breast carcinomas

express higher levels of PRLR than normal mammary tissue (Clevenger et al., 1995;

Reynolds et al., 1997; Touraine et al., 1998) and that their expression patterns differ

markedly (Gill et al., 2001). These observations suggest a relationship between PRLR

overexpression and the pathology of breast carcinomas and were our justification for

making G129R-PE40-KDEL.

Previous studies have reported the mechanism by which hPRL and G129R bind to

the PRLR and activate or inhibit signal transduction, respectively (Goffin et al., 1996;

Chen et al., 1999; Llovera et al., 2000b). We have constructed G129R-interleukin 2 and

G129R-endostatin fusion proteins and showed that their G129R moieties retain the ability to bind to the PRLR and competitively block hPRL-induced signal transduction (Zhang et al., 2002; Beck et al., 2003). In these studies we also demonstrated that the interleukin 2 and endostatin portions of the fusion proteins retain the ability to bind to and activate

71 receptors on T-cells and endothelial cells, respectively. The intracellular fate of these

fusion proteins remains unknown and is inconsequential since their functions are believed

to be served solely on the surface of the cells. Unlike G129R-interleukin 2 and G129R-

endostatin, the fusion of G129R to PE40-KDEL would require PRLR mediated

internalization and processing of the toxin to enter the cytoplasm and inhibit protein synthesis.

Little is known of the intracellular fate of hPRL and even less is known of the fate of G129R. In rat liver cells it has been demonstrated that PRL undergoes receptor

mediated endocytosis and accumulates intact in the Golgi region (Josefsberg et al., 1979;

Posner et al., 1982) before degradation occurs (Bergeron et al., 1983). In lactating rat

mammary epithelial cells PRL has been shown to avoid degradation in lysosomes by undergoing transcytosis; possibly via the fusion of the vesicles in the Golgi region with

secretory vacuoles (Seddiki et al., 2002). In addition to remaining in endosomes or

undergoing degradation, PRL has also been shown to undergo nuclear retrotranslocation

in rat Nb2 lymphoma cells (Clevenger et al., 1990b; Clevenger et al., 1990a; Rao et al.,

1993). This nuclear PRL has been shown to have nuclear actions (Clevenger et al., 1991;

Rao et al., 1995) and can modulate transcriptional events in part by interacting with

chaperones such as cyclophilin B (Rycyzyn et al., 2000; Rycyzyn and Clevenger, 2002).

In rat pituitary mammotrophs PRL was observed in all of these locations except

lysosomes (Giss and Walker, 1985). The presence of PRL in the Golgi apparatus,

secretory vesicles, and the nucleus suggest multiple intracellular fates for PRL in addition

to endosome/lysosomes, the organelle PE exploits to gain access to the cytoplasm.

72 100

PE

PE40

50 G129R-PE40-KDEL G129R Relative Cell Viability (% Control) Viability Cell of Relative

0 0.1 1 10 100

[Ligand] nM

Fig. 2.14 Examination of non-specific cytotoxicity in 4T1 mouse breast cancer cells in response to G129R, PE40, and G129R-PE40-KDEL

The consequence of fusing G129R to PE40-KDEL was examined to determine the effect on non-specific cytotoxicity. Cells were treated with PE, PE40, G129R-PE40- KDEL, or G129R and the relative number of viable cells was examined after 48 h by measuring the reduction of MTS. Data points presented are representative of the mean + SD from assays performed in quadruplicate on two separate occasions.

73 Table 2.1 Extrapolated concentrations of G129R-PE40-KDEL necessary for 50% inhibition of protein synthesis and 50% reduction in the relative number of viable cells in various breast cancer cell lines Protein Cell PRL Synthesis Cytotoxicity Binding*

Cell Line IC50 (nM) IC50 (nM) MCF-7 0.9 4.75 ++ BT-474 2.1 4.8 ++ BT-134 6 22.7 ND T-47D 8.2 12.5 +++ MDA-MB-453 29.6 >100 + MDA-MB-231 >33 >100 +/-

*Ormandy et al. 1997. J Clin Endo & Metab. 82:3692-3699. Shiu et al., 1987. Recent Prog Horm Res. 43:277-303.

74 In this study we demonstrate that G129R-PE40-KDEL is able to inhibit protein

synthesis and cell proliferation in multiple human breast cancer cell lines that express the

PRLR (Fig. 2.11). We did not anticipate that the inhibitory concentrations for G129R-PE40-

KDEL would be in the nanomolar range rather than the picomolar range (Table 2.1). To

further examine this, we utilized T-47D breast cancer cells since they are hPRL responsive

and express relatively high levels of PRLR (Shiu et al., 1987). We demonstrate that the

inhibitory effects of G129R-PE40-KDEL can be reduced drastically by treatment with excess

amounts of G129R or antibody against PE (Fig. 2.13), suggesting that the cytotoxic effects of

the fusion toxin rely heavily on cell binding and are mediated via the toxin moiety. We also

125 examined the ability of G129R-PE40-KDEL to compete with I-labeled hPRL for binding to

T-47D cells and confirmed that G129R-PE40-KDEL is able to bind to the PRLR with an

approximately 28-fold lower affinity than hPRL and approximately eleven-fold lower

affinity than G129R (Fig. 2.6).

The lower receptor binding affinity of G129R-PE40-KDEL may be due its size

and a conformational change in the G129R moiety which make it less accessible to PRLR

than G129R. The lower binding affinity of G129R-PE40-KDEL provides a logical

explanation for its poor inhibitory effects on STAT5 activation by hPRL (Fig. 2.8B). The

reduced affinity of G129R-PE40-KDEL for the PRLR may also account for the necessity

to use nanomolar concentrations to inhibit protein synthesis and cause cell death. To

ensure that the cytotoxicity observed in the nanomolar range was not due to non-specific

binding, we examined the effects on mouse 4T1 cells which are extremely sensitive to the

specific and non-specific effects of PE and PE40, respectively. We observed that fusion

75 of G129R to PE40-KDEL actually reduced the non-specific side effects associated with

PE40 (Fig. 2.14), indicating that G129R-PE40-KDEL may actually decrease the non-

specific cytotoxicity associated with the PE40 moiety. To further ensure that the

cytotoxicity observed was due to specific binding, we examined the effects on cell lines

that express the PRLR. The concentrations of G129R-PE40-KDEL necessary to reduce

the relative number of viable cells were orders of magnitude lower than the concentrations of

G129R needed to show any affect (Fig. 2.12); similarly, treatment of the human breast cancer

cell lines with PE40 alone resulted in little or no killing of the cells (data not shown). This

suggests that the cytotoxic effects observed are mediated via the PRLR and only occur if

G129R is fused with PE40-KDEL.

We evaluated the relationship between sensitivity to G129R-PE40-KDEL and

PRLR mRNA expression level based on two studies that quantified the expression of the

long isoform of PRLR mRNA in human breast cancer cell lines using Northern blot

analysis (Ormandy et al., 1997b) and real-time PCR (Peirce et al., 2001). Both studies

identified T-47D, MDA-MB-134, MDA-MB-483, and BT-474 cells as having high

PRLR levels; MCF-7 cells as having moderate levels; and MDA-MB-453 and MDA-

MB-231 cells as having little or no expression. All of the cell lines reported to express high levels of PRLR were more sensitive than the cell lines reported to have few or no

PRLR (Fig. 2.11). We also noticed that MCF-7 cells, reported to have modest PRLR levels, were more sensitive than all of the cell lines with high PRLR expression levels

(Fig. 2.11). The most obvious reason for the discrepancy would be that the mRNA levels don’t necessarily correlate with actual PRLR levels; however, this would not explain why

76 T-47D cells are not the most sensitive to G129R-PE40-KDEL since studies have shown they have the highest PRLR binding activity (Shiu et al., 1987; Ormandy et al., 1997b).

Differences in the rate at which the PE40 moiety is processed and translocated may affect the sensitivity of the cells. As controls for all of the cell proliferation assays we examined the sensitivity of the cell lines to PE. All of the cell lines showed various degrees of sensitivity to PE, with MCF-7 cells being the most sensitive. This suggests that processing of PE is the rate limiting step in intoxification and that sensitivity may not necessarily reflect PRLR expression levels.

Another explanation for the lack of a clear correlation between long isoform of PRLR mRNA expression level and sensitivity to G129R-PE40-KDEL may be explained by the complexity of PRLR expression. Multiple promoters (Hu et al., 2002) and multiple PRLR isoforms have been identified that arise from alternative splicing in humans (Clevenger et al.,

2003). In addition to the long isoform of PRLR which activates all known PRL signaling pathways, isoforms displaying differences in the extracellular or cytoplasmic domains have been identified that may have roles in the development and function of normal human mammary tissue and breast carcinomas. A transcript that would result in a soluble form of

PRLR has been found to be lower in human breast carcinomas than in normal mammary tissue and may modulate the expression of the long isoform of PRLR (Laud et al., 2000).

An isoform of PRLR with an extracellular truncation has also been identified in normal and malignant human breast tissue that has reduced binding affinity for PRL (Kline et al.,

2002). Isoforms of PRLR with truncations in the cytoplasmic domain have also been identified that may affect intracellular signaling and the internalization rates. An intermediate

77 isoform was identified in human breast cancer biopsies (Clevenger et al., 1995) that lacks

STAT5 motifs and has diminished ability to activate Fyn (Kline et al., 1999) and two short isoforms have been identified that are incapable of activating JAK2/STAT5 signaling (Hu et al., 2001). These short isoforms are expressed widely in human tissues and may modulate PRL signaling through the long isoform by acting in a dominant negative manner (Hu et al., 2001;

Trott et al., 2003). Recently, it has been observed that the ratio of short forms to long forms appears to be lower in human breast carcinomas and breast cancer cell lines than in normal samples (Meng et al., 2004). The short forms of the rodent and bovine PRLR have been shown to have a slower internalization rate than the long isoform (Lu et al., 2002). The presence of multiple PRLR isoforms with variations in expression level; the potential of these PRLRs to form homodimers and heterodimers with differences in binding affinity and internalization rates; and the lack of data concerning the trafficking and degradation of these various PRLR isoforms make it difficult to predict the potential effectiveness of

G129R-PE40-KDEL based solely on PRLR mRNA expression levels.

Studies regarding the status of PRLR in human breast cancer patients are limited

in contrast to the estrogen receptor status. From the relatively small numbers of clinical

studies it appears that the PRLR levels are elevated in a majority of primary human breast

cancers (Clevenger et al., 1995; Reynolds et al., 1997; Touraine et al., 1998). The high

levels of PRLR in the malignant breast tissue provide an opportunity to design targeted

therapeutics such as ligand directed toxin in this case. We show that G129R-PE40-KDEL binds to the PRLR with 28-fold reduced affinity compared to that of hPRL and retains one-tenth of the ability of G129R to inhibit PRL-induced STAT5 phosphorylation yet

78 drastically increases the specific cytotoxicity. In addition, we show that it can be

administered safely to mice in vivo (Table 2.2). The clinical potential of this ligand targeted toxin will largely depend upon improving the binding affinity of the fusion toxin towards the PRLR.

79 Table 2.2 Toxicity of PE and G129R-PE40-KDEL to mice Protein µg/mouse Mortality Single Dose

PE66 10 4/4

G129R-PE40-KDEL 50 0/4

G129R-PE40-KDEL 100 3/4 Daily Dose

G129R-PE40-KDEL 15 0/4

Mice were injected i.p. with a single dose or five times a week for two weeks. The number of dead mice was determined 72 h after administration.

80 CHAPTER THREE

IMPROVING THE PHARMACOKINETIC PROPERTIES OF

PROLACTIN AND GROWTH HORMONE RECEPTOR

AGONISTS AND ANTAGONISTS BY FUSION TO A

SYNTHETIC ALBUMIN BINDING PEPTIDE

Abstract

Human growth hormone (hGH) must be administered daily in order to achieve therapeutic effects due to its short circulation half-life. Human prolactin (hPRL), with

similar molecular size as hGH, shares a similar pharmacokinetic profile. Prolactin

receptor (PRLR) and growth hormone receptor (GHR) antagonists, G129R and G120R,

developed by introducing amino acid substitutions into hPRL and hGH, respectively,

suffer a similar fate in vivo. To prolong the circulation half-life, and therefore reduce the

frequency of injection of these potential therapeutics, we examined the effects of fusing a

serum albumin binding peptide, SA20 (Dennis et al., 2002), to the N-terminus or C-

terminus of hPRL, hGH, and their respective antagonists. Fusion of the SA20 peptide to

the N-terminus of the ligands was less detrimental upon the intrinsic activities of the

ligands, i.e. the ability to induce or inhibit signal transduction or rat Nb2 cell

proliferation, than when fused to the C-terminus. For pharmacokinetic studies, SA20-

hPRL (4 mg/kg) and hPRL-SA20 (4 mg/kg) exhibited a 4-fold prolonged plasma half-life

compared to hPRL (3.59 mg/kg) and their clearance was reduced approximately 9.5-fold

and 7-fold, respectively, after intraperiotneal (i.p.) administration of a single dose to

mice. Based upon the signaling results in vitro and the pharmacokinetic profiles in vivo,

81 only the fusion proteins with SA20 fused to the N-terminus were further used for pharmacodynamic studies. SA20-hPRL (4 mg/kg), hPRL (3.59 mg/kg), or PBS was i.p. administered to 8-week-old female mice daily, every two days, or every third day over a

12-day period. The number of lobulo-alveolar-like structures in mammary glands was greater in the three groups of mice treated with SA20-hPRL in comparison to hPRL.

SA20-hGH (8 mg/kg), hGH (7.15 mg/kg), or PBS were i.p. administered to 23-day-old female mice daily over a 40-day period. The weight gain, body length, and number of lobulo-alveolar-like structures in the mammary glands were greater in the mice treated with SA20-hGH than hGH. These findings indicate that SA20-hPRL and SA20-hGH are biologically active, that the serum albumin binding peptide improves the pharmacokinetic properties of otherwise short lived hPRL and hGH, and that it enhances the efficacy of hPRL and hGH in vivo by reducing their clearance from plasma.

82 Introduction

Human prolactin (hPRL) and human growth hormone (hGH) are members of the

somatotropin/prolactin family of four-helical cytokines which regulate the survival,

proliferation and differentiation of cells in a variety of tissues and endocrine glands.

hPRL mediates its effects via prolactin receptors (PRLRs); whereas, hGH mediates its effects via PRLRs or growth hormone receptors (GHRs). Because hPRL and hGH have a low molecular weight (23 kDa and 22 kDa, respectively) and are protein in nature, they are rapidly removed from circulation via renal clearance and by proteolytic degradation

within responsive tissues. For these reasons recombinant (r) hPRL and hGH must be

administered subcutaneously on a daily basis to maintain physiologically relevant plasma

concentrations in humans (Kastrup et al., 1983; MacGillivray et al., 1996; 2000b).

To the best of our knowledge, no long-acting formulations of rhPRL are under

development that would reduce the frequency of administration by sustaining its release

or by extending its plasma half-life. Sustained release of rhPRL could be achieved by

encapsulating it in microspheres of biodegradable polymers such as poly(D,L-lactide-co-

glycolide) or by creating hydrogels using poly(ethylene glycol) (PEG) modified with

fluorocarbon end groups, these strategies have proven effective for rhGH (Johnson et al.,

1996; Brodbeck et al., 1999; Tae et al., 2005). The pharmacokinetic properties of rhPRL

and rhGH may also be enhanced by expanding their absorptional half-life and reducing

their elimination half-life. This has been achieved for rhGH by covalently attaching one

or more PEG chains to the N-terminus and/or amino groups of lysine residues within

rhGH (Clark et al., 1996), allowing the hydrophilic regions of PEG to interact with water

83 molecules and effectively increase the size of the conjugated protein above the limit of filtration by the kidney. Thus, PEGylation enhances the mean residence time by reducing the glomerulus filtration rate. The greater the degree of PEGylation, the higher the mean residence time in the bloodstream; unfortunately, the degree of PEGylation is also correlated with a reduction in potency (Clark et al., 1996). This random conjugation of

PEG to lysine residues decreases the accessibility of the ligands for receptor-binding.

Recently, site-specific PEGylation has been achieved by introducing an unpaired cysteine residue into a non-essential site of rhGH followed by modification with a cysteine- reactive PEG (Cox et al., 2007).

In this study, we examine another means to expand the absorptional half-life and reduce the elimination half-life of rhPRL, rhGH, and their antagonists. We report that the fusion of a small serum albumin binding peptide to the N-terminus of rhPRL or rhGH:

(1) results in minimal deleterious effects on their ability to induce signaling via the Janus- activated kinase 2 (JAK2)/signal transducer and activator of transcription 5 (STAT5) and

Ras/Raf/MEK/mitogen-activated protein kinases (MAPKs) signaling cascades; (2) induces proliferation of Nb2 cells in a manner similar to that of unmodified rhPRL and rhGH in vitro; (3) increases their mean residence time by reducing glomerulus filtration; and (4) enhances proliferation and differentiation of epithelial cells within the lobules of mammary glands in vivo.

84 Materials and Methods

Cell Lines and Reagents

T-47D human breast ductal carcinoma cells and IM-9 Epstein-Barr virus-

transformed human B-lymphoblastoid cells were obtained from the American Type

Culture Collection (Manassas, VA). Nb2-11 rat pre-T lymphoma cells (Nb2) were kindly

provided by Dr. Ameae Walker and are now commercially available from Sigma-Aldrich

(St. Louis, MI). All reagents were purchased from Invitrogen (Carlsbad, CA) unless

otherwise stated. T-47D cells were maintained in RPMI Medium 1640 supplemented

with 10% fetal bovine serum (FBS) and 10 µg/ml gentamicin. For IM-9 cells, the medium was further supplemented with 10 mM HEPES and 1.0 mM sodium pyruvate.

Nb2 cells were maintained in RPMI Medium 1640 supplemented with 10% FBS, 10% horse serum (HS), 10 µg/ml gentamicin, and 0.1 mM 2-mercaptoethanol. The cell lines were incubated at 37°C with humidity and 5% CO2.

Construction of Plasmids for Expression of Albumin Binding Fusion Protein

Plasmids encoding the cDNAs for hPRL, hGH, G129R, and G120R were used as

DNA templates for the polymerase chain reaction (PCR). PCR was performed using the

Advantage-HF™ Kit (BD Biosciences, Palo Alto, CA). All oligonucleotides were

synthesized by Integrated DNA Technologies (Coralville, IA) and are shown in Table

3.1. Oligonucleotide primers for PCR were designed to incorporate restriction

endonuclease cleavage sites upstream and downstream of the gene of interest. The PCR

products were ligated with pCR2.1 T/A cloning vector and transformed into Escherichia

coli (E. coli) TOP10 cells obtained from Invitrogen. Positive clones were identified by

85 Table 3.1 List of oligonucleotides used to construct DNA encoding albumin-binding fusion proteins Nde I-SA20-BamH I Linker (Cohesive Ends) Nde I-SA20 for 1 5’-TATGCAGCGTCTGATTGAAGAT-3’ SA20 for 2 5’-ATTTGCCTGCCGCGTTGGGGCTG-3’

SA20-Bam H I for 3 5’-CCTGTGGGAAGATGATTTTG-3’

SA20-Bam HI rev 1 5-GATCCAAAATCATCTTCCCACAGGCAGCCCCA-3’ Nde I-SA20 rev 2 5’-ACGCGGCAGGCAAATATCTTCAATCAGACGCTGCA-3’

BamH I-SA20-stop-Xho I Linker (Cohesive Ends) BamH I-SA20 for 1 5’-GATCCCAGCGTCTGATTGAAGAT-3’ SA20 for 2 5’-ATTTGCCTGCCGCGTTGGGGCTG-3’ SA20-Xho I for 3 5’-CCTGTGGGAAGATGATTTTTAAC-3’

SA20-Xho I rev 1 5’-TCGAGTTAAAAATCATCTTCCCACAGGCAGCCCCA-3

BamH I-SA20 rev 2 5’-ACGCGGCAGGCAAATATCTTCAATCAGACGCTGG-3’ Nde I-PRL-BamH I & Nde I-G129R-BamH I

Nde I-PRL for 5’-CATATGTTGCCCATCTGTCCCGGCGGGG-3’ PRL-BamH I rev 5’-CTGGGATCCGCCGCCGCAGTTGTTGTTGTG-3’

BamH I-PRL-stop-Xho I & BamH I-G129R-stop-Xho I BamH I-PRL for 5’-GGATCCTTGCCCATCTGTCCCGGCGG-3’ PRL-stop- Xho I rev 5’-CTCGAGTTAGCAGTTGTTGTTGTGGA-3’ Nde I-GH-BamH I & Nde I-G120R-BamH I

Nde I-GH for 5’-CATATGTTCCCAACCATTCCCTTATCCAG-3’

GH-BamH I rev 5’-GGATCCGAAGCCACAGCTGCCCTCCACA-3’ BamH I-GH-stop-Xho I & BamH I-G120R-stop-Xho I BamH I-GH for 5’-GGATCCTTCCCAACCATTCCCTTATCCAGG-3’ GH-stop-Xho I rev 5’-CTCGAGCTAGAAGCCACAGCTGCCCTCCAC-3’

86 plating the cells on Luria-Bertani (LB) agar plates containing 100 µg/ml ampicillin and

80 µg/ml 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside. Plasmids were isolated

using the QIAprep® Spin Miniprep Kit (Qiagen, Valencia, CA), restriction endonuclease

digested, and the DNA fragments were separated by electrophoresis. The Nde I-PRL-

BamH I, Nde I-G129R-BamH I, Nde I-GH-BamH I, Nde I-G120R-BamH I and BamH I-

PRL-stop-Xho I, BamH I-G129R-stop-Xho I, BamH I-GH-stop-Xho I, and BamH I-

G120R-stop-Xho I DNA fragments were purified from the agarose gel using the

QIAquick® Gel Extraction Kit (Qiagen). Double stranded DNA encoding the serum

albumin binding peptide, SA20, was prepared by annealing five oligonucleotides together

that had been treated with T4 polynucleotide kinase. The oligonucleotides were designed

to create cohesive ends ready for ligation with restriction digested DNA utilizing optimal

codons for expression in E. coli. The DNA fragments were ligated with Nde I and Xho I cut pET22b(+) (Novagen, Madison, WI) to create pET22b-SA20-hPRL, pET22b-hPRL-

SA20, pET22b-SA20-G129R, pET22b-G129R-SA20, pET22b-SA20-hGH, pET22b- hGH-SA20, pET22b-SA20-G120R, and pET22b-G120R-SA20. All restriction and modifying enzymes were purchased from New England BioLabs (Beverly, MA).

Production and Purification of Albumin Binding Fusion Proteins To understand the nomenclature, SA20-hPRL is hPRL fused at its N-terminus to the C-terminus of a serum albumin binding peptide and has a molecular mass of approximately 25.7 kDa. In contrast, hPRL-SA20 is hPRL fused at its C-terminus to the

N-terminus of a serum albumin binding peptide and has a molecular mass of approximately 25.8 kDa. In addition to hPRL, SA20 derivatives of G129R, hGH and

87 G120R were produced using a bacterial system. Chemically competent Rosetta™ (DE3)

pLysS E. coli cells (Novagen) were transformed with the constructs and propagated

overnight at 37°C with agitation in LB broth containing 100 µg/ml ampicillin. The next

morning 1 L baffled flasks containing 500 ml of Terrific Broth with 100 µg/ml ampicillin

were inoculated with 20 ml of each of the starter cultures. The cultures were grown to an

O.D.600 of 0.9 before induction of protein production with 1 mM isopropyl-β- thiogalactoside (Alexis Biochemicals, San Diego, CA). After 4 h of induction the cells were harvested by centrifugation at 5,000 x g for 5 min and resuspended in Solution 1 (50 mM Tris·HCl; 10 mM EDTA; 0.5% Triton X-100; pH adjusted to 8.0) to which 0.4 mg/ml lysozyme was added. The cells were incubated at room temperature for 1 h to weaken the cell membranes before the cells were sheared on ice with five 1-min ultrasonic pulses (30% duty at setting 7) using a 550 Sonic Dismembrator (Fisher

Scientific, Waltham, MA). The insoluble fraction containing the inclusion bodies was recovered by centrifugation at 12,000 x g for 30 min at 4°C and resuspended in Solution

2 (50 mM Tris·HCl; 10 mM EDTA; 1 M urea; pH adjusted to 8.0). The inclusion bodies were collected by centrifugation at 12,000 x g for 30 min at 4°C and were solubilized and

denatured in Solution 3 (50 mM Tris·HCl; 10 mM EDTA; 8 M urea; 1% v/v 2-

mercaptoethanol; pH adjusted to 12.0). The proteins were refolded by dialysis against baths of

50 mM Tris·HCl, pH 8.0 containing decreasing amounts of urea (4M urea, 2 M urea, 0 M urea,

0 M urea) at 4°C. The refolded proteins were filtered through 0.45 µm filters and purified by

anion-exchange chromatography using a 5 ml HiTrap Q Sepharose HP column (Amersham

Biosciences, Piscataway, NJ) on an ÄKTA fast protein liquid chromatography system

88 (Amersham Biosciences). The concentration of the purified proteins was determined using the

Coomassie Plus™ Protein Assay Reagent (Pierce, Rockford, IL). The purified proteins were

separated by reducing SDS-PAGE along with hPRL standards on a 12% polyacrylamide gel and stained with SYPRO® Orange (Molecular Probes, Eugene, OR) to confirm their size, concentration and purity.

STAT5 and Erk1/2 Phosphorylation Assays

T-47D cells were resuspended in RPMI Medium 1640 supplemented with 10% charcoal/dextran treated FBS (CSS; HyClone, Logan, UT) and 10 µg/ml gentamycin and

plated in twelve-well tissue culture plates. The cells were grown overnight to approximately 80% confluency and depleted for 1 h in RPMI Medium 1640 supplemented with 0.5% CSS. Cells were treated for the indicated amounts of times with the indicated concentrations of hPRL, G129R, hGH, G120R or their albumin binding counterparts. The cells were washed with ice cold phosphate buffered saline (PBS) and lysed in 100 µl of Lysis Buffer (20 mM Tris·HCl, pH 7.4; 1% Igepal CA630; 6 mM deoxycholic acid; 150 mM NaCl; 1 mM EDTA, pH 8.0) containing protease inhibitors (1

µg/ml aprotinin; 1 µg/ml leupeptin; 170 µg/ml PMSF; 180 µg/ml sodium orthovanadate).

The lysates were homogenized and incubated on ice for 20 minutes before clarification by centrifugation at 12,000 x g at 4°C for 30 min. Approximately 40 µg of the supernatants were used for Western blot.

IM-9 cells were harvested by centrifugation and resuspended in DMEM containing 0.5% CSS. The volume was adjusted with medium to yield a final

89 concentration of 1 x 106 cells per ml. The cells were aliquoted (1 mL) into Eppendorf

tubes and depleted for 1 h before addition of 111 μL of 10X treatments. Cells were

treated for 30 min at 37°C, pelleted, the medium aspirated and resuspend in 100 μL of

Lysis Buffer containing protease inhibitors. Approximately 80 µg of the supernatants were used for Western blot.

Western Blot Analysis

The cell lysates were separated on 10% polyacrylamide gels by reducing SDS-

PAGE at 150 V for 1 h and transferred by Western blot to Hybond™ Enhanced

Chemiluminescence (ECL) nitrocellulose membranes (Amersham Biosciences) at 16

Watts for 90 min. The nitrocellulose membranes were blocked for 30 min at room temperature in Tris buffered saline (TBS) containing 0.05% Tween 20 (TBS-T) and 5%

milk. The membranes were incubated at 4°C with agitation overnight in the blocking

solution containing the primary antibody. The blots were washed three times with TBS-

T, and incubated for 2 h at room temperature with agitation in blocking solution containing a secondary antibody conjugated to horseradish peroxidase. Blots were washed three times with TBS-T and incubated for 1 min with ECL™ Western Blotting

Detection Reagents (Amersham Biosciences). The blots were exposed to Biomax MR film (Kodak, Rochester, NY) and the film was developed with a Konica SRX-101A processor (Konica Minolta Medical Imagining, Wayne, NJ).

Blots were initially performed with a 1:1000 dilution of mouse anti-phospho

STAT5A/B antibody (Upstate Biotechnology, Inc. Lake Placid, NY) and 1:1333 dilution of mouse anti-phospho ERK (sc-7383; Santa Cruz Biotechnology, Santa Cruz, CA)

90 followed by a 1:2000 dilution of goat anti-mouse conjugated HRP (Bio-Rad, Hercules,

CA). Blots were reprobed with a 1:1333 dilution of rabbit anti-STAT5 polyclonal

antibody (sc-835), 1:1000 dilution of rabbit anti-ERK1 (sc-93) and 1:1000 dilution of

rabbit anti-ERK2 (sc-154) from Santa Cruz Biotechnology followed by a 1:2000 dilution of goat anti-rabbit conjugated HRP (Bio-Rad).

Nb2 Cell Proliferation Assay

Rat Nb2 cells were resuspended to a final concentration of 5.5 x 105 cells per ml

in RPMI 1640 supplemented with 10% low lactogen HS (Cat. 2921154; MP Biomedicals,

Irvine, CA), 10 µg/ml gentamicin and 0.1 mM 2-mercaptoethanol. The cells were

depleted overnight and reseeded in flat bottom 96-well plates at a density of 20,000 cells

per well in 150 µl of medium. Treatments were added to each well in 50 µl volumes and

the cells were incubated for 72 h. The relative number of cells was determined by

incubating the cells for 4 h with 40 µl of 3-(4,5-dimethylthiazol-2-yl)-5-(3-

carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium/phenazine methosulfate

(MTS/PMS) solution (CellTiter 96 AQueous non-radioactive cell proliferation kit;

Promega, Madison, WI) per well. The reduction of MTS by living cells was determined by measuring the absorbance at 490 nm using a Benchmark microplate reader (Bio-Rad).

All assays were performed at least three times in quadruplicate.

Pharmacokinetic Studies in Mice

Female FVB/N mice (Charles River Laboratories, Wilmington, MA) eight weeks of age and weighing 22-25 g were used for all pharmacokinetic studies. Mice were acclimated to a 12 h light/dark cycle for at least one week prior to the start of the studies

91 and provided LabDiet® Prolab® RMH 3000 feed (PMI Nutrition International; St. Louis,

MO) and water ad libitum. All animal studies were approved by the Animal Care and

Use Committee. For pharmacokinetic studies, mice were administered a single

intraperitoneal (i.p.) injection of SA20-hPRL or hPRL-SA20 at 4 mg/kg body weight, or

an equimolar amount of hPRL using 27 gauge needles with a 0.4 mm bore and 13 mm

length. Blood drippings (10 to 50 µl) were collected via tail-sectioning in hematocrit

tubes at time points determined empirically. The samples were centrifuged and the

plasma was harvested and stored at -80 C until assayed.

Concentrations of hPRL or the fusion proteins were measured using a commercial

IRMA kit (Coat-A-Count® Prolactin IRMA; DPC, Los Angeles, CA) or ELISA kit

(Active® Prolactin ELISA; DSL, Webster, TX) determined to be insensitive to mouse

PRL. hPRL, hPRL-SA20 and SA20-hPRL standards (2 ― 200 ng/ml) were prepared in

the zero calibrator buffer provided with the kits and used to prepare standard curves.

Dilutions of the serum samples were determined empirically and were made in the zero

calibrator buffers. Plasma concentrations were extrapolated from the standard curves

prepared for each protein.

The plasma concentration (Cp) verses time (t) profiles were plotted and non-

compartmental methods were used to estimate basic pharmacokinetic parameters. The

peak serum concentrations (Cmax) were measured. The terminal elimination rate constant

(k) was obtained from the terminal slope of the log-linear Cp versus t plots using the last

three data points within the linear portion of their elimination phases. The terminal half- life (t1/2) was calculated as ln2/k. The area under the serum concentration-time curve

92 (AUC) and area under the first moment of serum concentration-time curve (AUMC) were

calculated by numerical integration using the trapezoidal rule. AUC0 was extrapolated

using the linear trapezoidal rule; whereas AUC∝ was determined by dividing the last

2 measurable concentration level by k. AUMC∝ was determined as Cplast·tlast/k + Cplast/k .

The AUC and AUMC values were used to calculate the mean residence time (MRT), apparent elimination rate constant (kel’) and total body clearance (CL).

Pharmacodynamic Studies

Starting at 60 days of age, mice (n=6) were injected i.p. with hPRL, SA20-hPRL or PBS once daily for 12 consecutive days (days 1-12), every other day (days 1, 3, 5, 7, 9, and 11), or every third day (1, 4, 7 and 10). Individual body weights of the mice were determined using an analytical balance and the dose of hPRL administered was equimolar with the amount of hPRL in a 4 mg/kg dose of SA20-hPRL. Blood samples were not collected throughout the study because we did not want to stress the mice and potentially compromise the study. Mice were euthanized by cervical dislocation on day 13 and the right abdominal mammary glands were harvested to prepare whole-mounts to assess mammary gland morphology. The mammary glands were spread onto glass slides and fixed overnight in Carnoy’s fixative. The glands were then stained with alum carmine,

destained in water, dehydrated in xylene, and mounted with Permount (Fisher, Pittsburgh,

PA) and a glass cover slip. The whole mounts were scanned to obtain high resolution

images of the glands. The mammary glands were scored for lobulo-alveolar development

(LAD) on a scale of 1– 5. The scoring system used was: 1 = ducts with few side branches and no lobulo-alveolar buds, 2 = ducts with end buds and few lobulo-alveoli,

93 3 = ducts with limited lobulo-alveolar development, 4= ducts with moderate lobulo-

alveolar development, 5 = ducts with extensive lobulo-alveolar development.

Pharmacodynamic studies were also carried out for hGH and SA20-hGH.

Starting at 23 days of age, mice (n=6) were injected i.p. with hGH, SA20-hGH or PBS daily for 40 consecutive days. The dose of hGH was equimolar with the amount in a 8 mg/kg dose of SA20-hGH. Individual body weights of the mice were determined daily until 44 days of age and three times a week thereafter. Body weight gain was evaluated using repeated measures analysis of variance (ANOVA). Mice were euthanized by cervical dislocation on day 63. Tail lengths were measured from anus to tail-tip. The right posterior mammary glands were harvested, whole mounts were prepared, and lobulo-alveolar development was scored as mentioned above.

94 Results

Construction, Expression, and Purification of Serum Albumin Binding Derivatives

of hPRL, hGH, and Their Antagonists

SA20 derivatives of hPRL, G129R, hGH, and G120R were recombinantly

engineered by gene fusion. The cDNA encoding the mature form of hPRL, G129R,

hGH, or G120R were ligated with DNA encoding the SA20 peptide and cloned into an E.

coli expression vector. Expression of each of the fusion genes in E. coli resulted in a

single protein in which the SA20 moiety was connected to the N-terminus or C-terminus

of the hPRL, G129R, hGH, or G120R moiety via a Gly-Ser peptide linker. The proteins

were found to accumulate in the insoluble fraction as inclusion bodies. The inclusion bodies were isolated, denatured, refolded, and purified by anion-exchange chromatography to yield preparations of greater than 95% purity. Reducing SDS-PAGE revealed that the purified fusion proteins had sizes corresponding to their predicted molecular masses (not shown).

Serum Albumin Binding Derivatives of hPRL and hGH Induce Signal Transduction

Dose response and time course experiments were performed with hPRL, hGH and their SA20 derivates to determine their potency and efficiency at inducing signal transduction. Treatment of PRLR expressing T-47D cells with various concentrations of hPRL (Fig. 3.1A) or hGH (Fig. 3.1B) resulted in the phosphorylation of STAT5 in a

dose-dependent manner. The potency of SA20-hPRL appeared to be equivalent to that of

hPRL; in contrast, the potency of hPRL-SA20 was reduced, requiring approximately 3-

fold higher molar concentrations to achieve the same degree of STAT5 phosphorylation

95 observed for hPRL (Fig. 3.1A). The potency of SA20-hGH at activating STAT5 and

MAPKs, extracellular signal-regulated kinases 1 and 2 (Erk1/2), was reduced approximately 3.3-fold in comparison to hGH; whereas, the potency of hGH-SA20 was reduced approximately 10-fold (Fig. 3.1B). To determine if the SA20 derivatives of hGH also retain the ability to bind to GHRs and activate signaling, we examined the phosphorylation of STAT5 in IM-9 cells. Treatment with various concentrations of hGH resulted in the phosphorylation of STAT5 in a dose-dependent manner (Fig. 3.1C). The potency of SA20-hGH appeared to be approximately 10-fold less than that of hGH, and hGH-SA20 appeared to be far less potent (Fig. 3.1C). Time course experiments were also performed to examine the efficiency of signaling (Fig. 3.2). SA20-hPRL induced the phosphorylation of Erk1/2 as efficiently as hPRL (Fig. 3.2A); in contrast, SA20-hGH required a longer time course to achieve similar activation levels as hGH (Fig. 3.2B).

Serum Albumin Binding Derivatives of G129R and G120R Competitively Inhibit

hPRL and hGH-Induced Signal Transduction

To determine if the SA20 derivatives of G129R and G120R retain the ability to

bind and antagonize PRLR mediated signaling transduction, the phosphorylation of

signaling molecules associated with the JAK2/STAT5 and Ras/Raf/MEK/MAPK

signaling cascades were examined in T-47D cells. Treatment with various concentrations

of G129R (Fig. 3.3A) or G120R (Fig. 3.3B) resulted in the inhibition of hPRL and hGH-

induced STAT5 phosphorylation, respectively, in a dose-dependent manner.

96 A

hPRL SA20-hPRL hPRL-SA20 nM: - 1 3.3 10 33 1 3.3 10 33 1 3.3 10 33 p-STAT5

STAT5 B

hGH SA20-hGH hGH-SA20 nM: - 1 3.3 10 33 1 3.3 10 33 100 1 3.3 10 33 100 p-ERK1/2

p-STAT5

STAT5

C

hGH SA20-hGH hGH-SA20 nM: - 1 3.3 10 33 1 3.3 10 33 100 1 3.3 10 33 100 p-STAT5

STAT5

Fig. 3.1 Dose-dependent activation of STAT5 and Erk1/2 signaling induced by hPRL, hGH, and their SA20 derivatives

T-47D cells (A and B) or IM-9 cells (C) were plated and serum starved as described in the Materials and Methods. The cells were treated with increasing doses of hPRL, SA20-hPRL or hPRL-SA20 (A) or increasing doses of hGH, SA20-GH or hGH- SA20 (B and C) and response of the signaling molecules was examined after 30 min. Cell lysates were harvested and 80 μg of lysate from each treatment was separated by reducing SDS-PAGE using 10% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed with anti-phospho-STAT5 (p-STAT5) and anti-phospho-Erk1/2 (p-ERK1/2). Equal loading was confirmed by reprobing the membranes with an anti-STAT5 antibody (STAT5).

97 A 5 nM hPRL min: - 2 5 10 20 30 60 p-ERK1/2

ERK1/2 B 5 nM SA20-hPRL min: - 2 5 10 20 30 60 p-ERK1/2

ERK1/2

C 5 nM hGH min: - 2 5 10 20 30 60

p-ERK1/2

ERK1/2

D 5 nM SA20-hGH min: - 2 5 10 20 30 60 p-ERK1/2

ERK1/2

Fig. 3.2 Time course analysis of the activation of Erk1/2 in response to hPRL, hGH, and their SA20 derivatives

T-47D cells were plated and serum starved as described in the Materials and Methods. The cells were treated with hPRL (A), SA20-hPRL (B), hGH (C) or SA20- hGH (D) for various durations (min). Cell lysates were harvested, and 80 μg of lysate from each treatment was separated by reducing SDS-PAGE using 10% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed with a phospho-Erk1/2 antibody. Equal loading was confirmed by reprobing the membranes with anti-Erk1/2 antibodies.

98 A hPRL: hPRL: hPRL: hPRL G129R SA20-G129R G129R-SA20 - 5 nM 1:5 1:10 1:20 1:40 1:25 1:50 1:100 1:200 1:25 1:50 1:100 1:200 p-ERK1/2

pTyr-STAT5

pSer-STAT5

STAT5

B hGH: hGH: hGH: hGH G120R SA20-G120R G120R-SA20 5 nM 1:5 1:10 1:20 1:40 1:25 1:501:1001:200 1:251:50 1:100 1:200 p-ERK1/2

pTyr-STAT5

STAT5

Fig. 3.3 Comparison of the inhibitory effects of G129R, G120R, and their SA20 derivatives on hPRL and hGH-induced signaling

T-47D cells were plated and serum starved as described in the Materials and Methods. The cells were treated with 5 nM hPRL and increasing doses of G129R, SA20- G129R or G129R-SA20 (A) or 5 nM hGH and increasing doses of G120R, SA20-G120R or G120R-SA20 (B) and the response of signaling molecules was examined after 30 min. Cell lysates were harvested and 80 μg of lysate from each treatment was separated by reducing SDS-PAGE using 10% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed for activate Erk1/2 (p-ERK1/2), activate STAT5 (pTyr-STAT5), inactive STAT5 (pSer-STAT5) and total STAT5 (STAT5).

99 The potency of SA20-G129R was 5-fold less than that of G129R; in contrast, the potency of G129R-SA20 was reduced more than 10-fold (Fig. 3.3A). The potency of SA20-

G120R at inhibiting STAT5 and Erk1/2 phosphorylation was reduced approximately 10- fold in comparison to hGH; whereas, the potency of G120R-SA20 was reduced even more so (Fig. 3.3B).

Serum Albumin Binding Derivatives of hPRL and G129R Induce Proliferation of

Rat Nb2 Cells in a Dose-Dependent Manner

To examine the biological activity of SA20-hPRL and hPRL-SA20, we measured their ability to induce the proliferation of rat Nb2 cells, which proliferate in response to lactogenic hormones such as hPRL. The relative number of viable cells was determined after 72 h treatment by colorimetrically measuring the reduction of MTS by living cells.

Treatment of the cells with increasing doses of SA20-hPRL or hPRL-SA20 stimulated cell proliferation in a dose-dependent manner (Fig. 3.4). SA20-hPRL was less potent than hPRL but was more potent than hPRL-SA20 at stimulating cell proliferation. The proliferation induced by these ligands in rat Nb2 cells correlated well with their ability to activate signal transduction in T-47D cells, confirming that bioactivity of SA20-hPRL was superior to that of hPRL-SA20. Rat Nb2 cells also proliferate in response to G129R at concentrations more than 2-log higher than that of hPRL. This partial agonism has been attributed to the rat PRLR forming a more stable interaction with binding Site 2 of these ligands than the human PRLR. Interestingly, the residual agonism of SA20-G129R was significantly higher than that of G129R. This may indicate that the addition of the

100 SA20 moiety to the N-terminus enhances binding Site 2 affinity for the PRLR, an attribute that is undesirable for an antagonist.

Bioanalysis

Before performing pharmacokinetic studies it was necessary to confirm that

SA20-hPRL and hPRL-SA20 could be detected using commercial immunoassays. The

IRMA and ELISA kits tested showed specificity for hPRL and no cross-reactivity with mouse PRL. The IRMA kit was less sensitive at detecting hPRL when the SA20 peptide was fused to the N-terminus; whereas, the ELISA kit recognized SA20-hPRL and hPRL-

SA20 equally well with sufficient sensitivity. Examination of the stability of hPRL,

SA20-hPRL, or hPRL-SA20 in plasma revealed negligible metabolism indicating that plasma decay is not a major player in clearance. To ensure that the immunoassays were detecting biologically active hPRL, SA20-hPRL, or hPRL-SA20 we attempted to measure the proliferation of Nb2 cells in vitro in response to plasma samples. Plasma was harvested from the serum of mice and spiked with different concentrations of hPRL,

SA20-hPRL, or hPRL-SA20. Regardless of the volume of plasma used, the Nb2 bioassay had extremely low-sensitivity for measuring hPRL, SA20-hPRL and hPRL-

SA20 due to other components within the plasma.

Pharmacokinetics of hPRL, SA20-hPRL, and hPRL-SA20

To examine pharmacokinetic behavior, female FVB/N mice were i.p. administered a single dose (4 mg/kg) of hPRL-SA20, SA20-hPRL, or the molar equivalent amount of hPRL. The plasma concentrations (Cp) were monitored over time

(t) and a linear-plot of Cp versus t (Fig. 3.5A) was used to visually emphasize the initial

101 absorption, distribution and elimination phases; a semi-logarithmic plot of Cp versus t

(Fig. 3.5B) was used to emphasize the terminal elimination phase and to derive pharmacokinetic parameters. The pharmacokinetic parameters of hPRL, hPRL-SA20 or

SA20-hPRL were determined using non-compartmental methods and are summarized in

Table 3.2.

Absorption of hPRL and hPRL-SA20 into the bloodstream after i.p. administration did not differ significantly, reaching their maximal observed concentration after approximately 50 min and 1 h, respectively; however, the observed concentration achieved by hPRL-SA20 was 3.5-fold higher than that of hPRL (Fig. 3.5A). The maximum observed concentration achieved by SA20-hPRL was slightly lower than that of hPRL-SA20 but occurred after 3 h, indicating that SA20-hPRL was absorbed into the blood stream more slowly (Fig. 3.5A). Differences were also observed in the elimination of the SA20 derivatives of hPRL. The AUCs of hPRL-SA20 and SA20-hPRL were increased approximately 7- and 10-fold in comparison to hPRL, respectively; with the small difference between the two reflecting earlier distribution and elimination of hPRL-

SA20 (Fig. 3.5A). For the most part, the differences in AUC between hPRL and the

SA20 derivatives reflected alterations in their terminal elimination phase (Fig. 3.5B).

The elimination patterns for each protein had two distinct phases (α and β).

Elimination in the second phase occurred according to first-order kinetics, i.e. a linear relationship was observed between the semi-log plot of drug concentration versus time

(Fig. 3.5B). This β phase was used to determine the elimination half-life of the proteins.

102 3.0

2.5 hPRL SA20-hPRL 2.0 hPRL-SA20 SA20-G129R 1.5

O.D. 490 O.D. G129R-SA20 G129R 1.0

0.5

0.0 -4 -3 -2 -1 0 1 2 3

Log Concentration (nM)

Fig. 3.4 The effects of hPRL, G129R, and their SA20 derivatives on Nb2 cell proliferation

Rat Nb2 cells were treated with various doses of hPRL, SA20-hPRL, hPRL- SA20, G129R, SA20-G129R or G129R-SA20 for 72 h as described in the Materials and Methods. Cell proliferation was assessed colorimetrically by measuring the reduction of MTS. Each data point represents the mean ± SD of at least three independent experiments performed in quadruplicate.

103 A 16 hPRL 14 hPRL-SA20 12 SA20-hPRL g/mL)

μ 10 8 6 4

Concentration ( 2 0 04812162024 Time (hour) B 100,000 hPRL 10,000 hPRL-SA20 SA20-hPRL 1,000

100

10 Concentration (ng/mL)

1 0 1020304050 Time (hour)

Fig. 3.5 Mean plasma concentration-time profiles of hPRL, SA20-hPRL, and hPRL-SA20 after administration to mice

A single dose of hPRL (3.59 mg/kg), SA20-hPRL (4 mg/kg) or hPRL-SA20 (4 mg/kg) was i.p. administered to female FVB/N mice and their plasma concentrations were determined as described in Materials and Methods. (A) Linear-plot and (B) semi- log plot of plasma concentration versus time profiles. Points, mean of two mice per time point.

104 The terminal elimination half-life of hPRL-SA20 and SA20-hPRL were nearly identical and were increased approximately 4-fold in comparison to hPRL (Table 3.2). This

extension in terminal elimination half-life is reflected by a reduction in the clearance of

hPRL-SA20 and SA20-hPRL by approximately 7-fold and 9.5-fold, respectively (Table

3.2). The reduction in clearance from 535.72 ± 123.19 ml/h/kg to 56.70 ± 2.6 ml/h/kg by

addition of the serum albumin binding peptide to the N-terminus of hPRL is consistent

with an “effective” increase in molecular size as indicated in Figure 3.6. Distribution studies using iodinated hPRL, hPRL-SA20, and SA20-hPRL were not performed; however, the apparent volume of distribution of hPRL was observed to be much larger than the plasma volume, indicating that hPRL readily left the bloodstream and entered other tissues. In contrast, the apparent volume of distribution of hPRL-SA20 and SA20- hPRL were reduced 1.75- and 2.4-fold, respectively, consistent with the serum albumin binding peptide causing more hPRL to be retained in the plasma (Table 3.2). The steady

state volumes for hPRL-SA20 and SA20-hPRL were reduced 1.5- and 2.5-fold in

comparison to hPRL, respectively. Since steady state volume is inversely correlated with

molecular weight, the reduction in steady state volume likely reflects binding of SA20-

hPRL and hPRL-SA20 to serum albumin, which would effectively increase their

molecular size.

Pharmacodynamics of hPRL and SA20-hPRL

Since a reduction in volume of distribution generally results in a reduction in

biological half-life, we examined the biological activity of SA20-hPRL on mammary

gland lobulo-alveolar development in comparison to hPRL and PBS controls.

105 Table 3.2 Pharmacokinetic parameters after a single treatment of mice (n = 2) with hPRL, hPRL-SA20, or SA20-hPRL hPRL hPRL-SA20 SA20-hPRL (Mean) (Mean) (Mean) Administration Route IP IP IP Dose (mg/kg) 3.59 4 4 AUC (mg • hr /L) 7.67 58.8 78.7 CL ( mL/hr/kg) 535.7 77.4 56.7

Apparent Vd ( mL/kg) 1302 745.9 544.6

t 1/2 term (hr) 1.66 6.68 6.64

Observed C max (mg/L) 4.07 14.33 12.08

Observed Tmax (hr) 0.84 1.01 3.06 Abbreviations: AUC = area under the concentration-time curve; CL = clearance; Vd = volume of distribution; t1/2term = terminal elimination half-life; Cmax = maximum plasma concentration; Tmax = time to maximum plasma concentration.

106 hPRL

SA20-hPRL

Figure 3.6 The rate of hPRL and SA20-hPRL clearance in mice and their effective molecular size

The clearance of hPRL and SA20-hPRL were calculated as described in the Materials and Methods. Based on the calculated clearance values the “effective” molecular size was extrapolated. Clearance of hPRL (23 kD) was similar to that reported for hGH (22 kD) and had an “effective” molecular size deduced to be 23 kD (blue circle); in contrast, clearance of SA20-hPRL (26 kD) was reduced and had an “effective” molecular size deduced to be 54 kD (green circle). Adapted from Clark et al. (Clark et al., 1996), with modifications.

107 FVB/N female mice (n = 6) were i.p. administered 4 mg/kg SA20-hPRL, an equimolar

amount of hPRL or PBS everyday, every other day, or every third day over a period of 12

days. No adverse side effects were observed in the mice after administration of hPRL or its serum albumin binding derivative. Both hPRL and SA20-hPRL increased the number of lobulo-alveoli in the mammary glands in a dosing-dependent manner (Fig. 3.7), the more often they were treated, the greater the extent of lobulo-alveolar development. The group receiving SA20-hPRL consistently had more lobulo-alveoli than the group receiving hPRL. Mice receiving SA20-hPRL every other day exhibited more lobulo- alveoli than mice receiving hPRL every other day (Fig. 3.7).

Pharmacodynamics of hGH and SA20-hGH

Female FVB/N mice (n=6) were i.p. administered 8 mg/kg SA20-hGH, an equimolar amount hGH, or PBS in a volume of 200 μl for 40 consecutive days. No adverse side effects were observed in the mice after administration of hGH or its serum albumin binding derivative. Extremely significant alterations were observed in the number of lobulo-alveoli and extent of side-branching in the mammary glands of mice treated with hGH or SA20-hGH in comparison to mice treated with PBS (Fig. 3.8). Side- branching and the number of lobulo-aveoli were consistently and significantly elevated in the mammary glands of mice treated with SA20-hGH in comparison to hGH (Fig. 3.8).

Extremely significant changes were also observed in the length of the tails (Fig. 3.9) and overall body weight (Fig. 3.10) of mice treated with hGH or SA20-hGH in comparison to

PBS. After the 6th and 7th day of treatment with SA20-hGH and hGH, respectively, a

significant weight gain was observed in the mice in comparison to mice receiving PBS.

108 A significant difference in the body weights of mice treated with SA20-hGH in comparison to hGH was eventually observed after 20 days of treatment (Fig. 3.10).

109 hPRL PBS Every 3rd Day Every 2nd Day Daily

SA20-hPRL PBS Every 3rd Day Every 2nd Day Daily

Figure 3.7 Lobulo-alveolar response in the mammary glands of mature nulliparous mice treated with PBS, hPRL, or SA20-hPRL

Eight-week old nulliparous FVB/N female mice (n = 6) were treated (i.p.) with hPRL (3.59 mg/kg), SA20-hPRL (4 mg/kg) or PBS everyday, every other day, or every third day over a period of 12 days. The mammary glands were isolated and the whole mounts were prepared and scored for lobulo-alveolar development.

110 PBS

hGH

SA20-hGH

Fig. 3.8 Lobulo-alveolar response in the mammary glands of young nulliparous mice treated with PBS, hGH, or SA20-hGH

Nulliparous FVB/N female mice (n = 6) were treated (i.p.) with hGH (7.15 mg/kg), SA20-hGH ( 8 mg/kg) or PBS daily beginning at 24 days of age and ending at 63 days of age. Mammary gland whole mounts were prepared and scored for lobulo- alveolar development.

111 9.0

** 8.5 ** PBS hGH 8.0 SA20-hGH

Tail Length (cm) 7.5

7.0 Treatment

Fig. 3.9 Tail length measurements after treatment with PBS, hGH, or SA20-hGH

Nulliparous FVB/N female mice (n = 6) were treated (i.p.) with hGH (7.15 mg/kg), SA20-hGH (8 mg/kg) or PBS daily beginning at 24 days of age and ending at 63 days of age, at which time the the tail lengths were measured. Points, mean tail length of five mice; bars, SE; symbols, indicate a P ≤ 0.001 by t test after 5.5 wk of treatment compared with PBS controls (**).

112 30 28 26 24 22 20 18 16 Body Weight Body Weight (g) SA20-hGH* † 14 hGH* 12 PBS 10 20 30 40 50 60 70 Age (Days)

Fig. 3.10 Body weight measurement after treatment with PBS, hGH, or SA20-hGH

Nulliparous FVB/N female (n = 6) mice were treated (i.p.) with PBS, hGH (7.15 mg/kg), or SA20-hGH (8 mg/kg) beginning at 23 days of age and ending at 63 days of age. Points, mean weight of six mice per time point; bars, SE; symbols, indicate a P ≤ 0.0001 by t test after 5.5 wk of treatment compared with PBS controls (*) or SA20-hGH vs. hGH (†).

113 Discussion

A potential indication for rhPRL is the support and sustainment of lactation. PRL is critical for milk production in humans (Kauppila et al., 1987) and its secretion from lactotrophs in the anterior lobe of the pituitary gland is under the stimulatory and strong inhibitory control of thyrotropin releasing hormone (TRH) and dopamine, respectively.

Administration of dopamine receptor agonists, such as Cabergoline, prevent postpartum lactation (Melis et al., 1988); whereas, administration of TRH significantly increases milk production in postpartum women (Tyson et al., 1975; Peters et al., 1991). Approximately

15% of newly lactating women suffer from lactation insufficiencies (Powers, 1999).

Milk production can be compromised by low hPRL response to suckling in women whom are overweight or obese before conception (Rasmussen and Kjolhede, 2004) and difficulty maintaining milk production can arise due to premature birth (Ehrenkranz and

Ackerman, 1986) but is most often associated with inadequate breast-feeding rates

(Delvoye et al., 1977).

Breast-feeding is important and beneficial for both the infant and mother

(Williams, 1995). Breast milk contains maternal macrophages and antibodies that protect the infant against illness, benefits not afforded by formulations, and numerous studies have shown that breast-feeding reduces the risk of developing breast cancer (Beral et al.,

2002). Unfortunately, there are no drugs approved by the FDA for enhancing breast milk production in women suffering from lactation insufficiency. Dopamine antagonists, such as Metoclopramide, have been used to increase pituitary hPRL secretion (Brouwers et al.,

1980; Brown et al., 2000) and milk production (Kauppila et al., 1981; Ehrenkranz and

114 Ackerman, 1986); however, they can have severe side-effects, can be transferred by breastfeeding to infants where their effects are unknown, and can inadvertently increase the secretion of other hormones (Massara et al., 1985).

Rather than use drugs to modulate hPRL levels, rhPRL is a more logical choice to treat lactation insufficiencies. It can be administered with fewer side effects and is safe for infants since hPRL is naturally transferred via milk during breast feeding.

Preliminary studies have shown that rhPRL actually causes the secretion of milk in non- postpartum woman (Page-Wilson et al., 2007), indicating that it should support and sustain lactation in postpartum woman. Evidence also indicates that rhGH can improve breast milk volume in normal lactating women (Milsom et al., 1992) and in women suffering from lactational insufficiency due to premature birth (Gunn et al., 1996; Milsom et al., 1998); but also increases the level of IGF-I in the breast milk (Breier et al., 1993).

Another potential indication for rhPRL is the enhancement or reconstitution of the lymphopoietic portion of the immune system when under dysregulation (Richards and

Murphy, 2000). SCID mice engrafted with normal human peripheral blood lymphocytes

are perhaps the best model to study the effects of rhPRL on the human immune system.

Using this model, rhPRL was shown to increase the number of human natural killer (NK)

cells, T lymphocytes, and B lymphocytes (Sun et al., 2004b); to improve the anti-tumor

effects of normal human NK cells by enhancing their proliferation and cytotoxicity

(Zhang et al., 2005a); and to promote human B-cell secondary Ig responses following

rechallenge (Zhang et al., 2007). These findings provide further support that rhPRL may

be useful for recovering from myelosuppression caused by chemotherapy (Woody et al.,

115 1999b; Zhang et al., 2005b), bone marrow transplant (Woody et al., 1999a; Sun et al.,

2003), burn injury (Dugan et al., 2002; Dugan et al., 2004) or immunosuppressive disease

(Richards et al., 1998); may have adoptive immunotherapeutic value for cancer

(Oberholtzer et al., 1996; Matera et al., 1999; Matera et al., 2000; Majumder et al., 2002;

Sun et al., 2002; Sun et al., 2004a) or pathogen-induced infection (Di Carlo et al., 1993;

Meli et al., 1996); and may be of potential use as a B-cell adjuvant in vivo (Richards et

al., 1998). Evidence suggests that rhGH may have similar effects on multilineage human

hematopoiesis (Hanley et al., 2005) and may be useful for recovering from

myleosuppression (Chen et al., 2003; Carlo-Stella et al., 2004).

A more recently proposed indication for rhPRL is for the treatment of type 1

diabetes mellitus (Labriola et al., 2007b). Type 1 diabetes mellitus results from a

deficiency in insulin production in the pancreas caused by the autoimmune destruction of

beta cells in the islets of Langerhans. rhPRL increases the proliferation of beta cells, islet cell mass, and insulin synthesis and secretion in primary cultures of human pancreatic islets (Labriola et al., 2007a), much like is observed in during pregnancy. Consistent with this, hyperprolactinemia has been associated with hyperinsulinemia and insulin resistance (Bahceci et al., 2003). Like rhPRL, rhGH has been shown to increase insulin secretion in human islets in vitro (Brelje et al., 1993). These findings indicate that rhPRL

administration could prevent the destruction of beta cells or could be used to expand

primary islets for transplantation back into the pancreas.

Currently, administration of rhPRL is not practical due to its short half-life, the

necessity for continuous maintenance, and the high cost of production. This would

116 undoubtedly create problems with compliance and discomfort at the site of injection. In this study, we examined the potential of fusing a serum albumin binding peptide to hPRL and hGH to enhance their serum half-life. Serum albumin is the most common protein in serum; it has a long half-life and has the ability to reversibly bind proteins. As a transporter it prevents the renal clearance of small molecules, thereby increasing their serum half-life. Recently, a series of unique peptides were developed using phage display which bind to serum albumin in an apparently non-competitive manner (Dennis et al., 2002). In this study, we genetically fused one of these serum albumin binding peptides, SA20, to the N-terminus or C-terminus of hPRL, hGH, and their antagonists in an effort to increase their plasma half-life by reducing renal clearance.

SA20-hPRL was found to be as efficient as hPRL at inducing signal transduction

and was 3.3-fold more efficient than hPRL-SA20 (Fig. 3.1). Similarly, SA20-hPRL was superior to hPRL-SA20 at inducing the proliferation of Nb2 cells (Fig. 3.4). Three parameters were used to determine which version of the fusion protein would be used for in vivo bioactivity studies. The observed peak height concentration (Cmax) achieved; the

time to reach this peak concentration (Tmax) and the area under the blood concentration-

time curve which reflects absorption, distribution, metabolism and excretion. The rate of

absorption of SA20-hPRL was slower than hPRL-SA20 in mice as indicated by the larger

Tmax and smaller Cmax, the AUC was greater, and the duration during the first 24 h

appeared more favorable than for hPRL-SA20. In addition, SA20-hPRL had a lower

apparent volume of distribution and reduced clearance, consistent with binding to serum albumin and slower renal elimination.

117 Based on these observations we examined the pharmacodynamics of SA20-hPRL and SA20-hGH in comparison to hPRL and hGH, respectively, after being i.p. administered to mice. Drugs administered to the peritoneum are absorbed only four times slower than when administered intravenously (Morton et al., 2001) due to the large surface area and vasculature of the peritoneum. Once in the peritoneum drugs rapidly enter circulation via the colic, the intestinal, and the mesenteric veins and are delivered via the inferior and superior mesenteric veins to the portal vein and enter the liver (Lukas et al., 1971). In the liver, drugs may be subject to metabolism before being carried with

deoxygenated blood by the inferior vena cava into the right atrium of the heart. From the

right atrium the blood enters the right ventricle and is pumped into the pulmonary artery

to the lungs before returning via the pulmonary vein into the left atrium and left ventricle

into the circulatory system.

This route of administration undoubtedly subjects hPRL, hGH, and their

derivatives to first pass hepatic metabolism, meaning they pass through the liver before

they reach systemic circulation. Since hepatic blood flow is much higher in mice than

man, hepatic extraction is very high in mice and hPRL, hGH, and their derivatives are

subject to rapid metabolism in the liver, which expresses high levels of PRLR and GHR.

Previously, we have demonstrated that i.p. administered G129R is cleared rapidly from

the blood of mice primarily by the liver and kidneys with less than 1% of the injected

dose per gram of tissue remaining after one hour. Although the SA20 peptide should

reduce kidney clearance by allowing hPRL or hGH to interact with serum albumin; this

interaction probably has little effect on metabolism in the liver. Thus, although the route

118 of administration is convenient in mice it is less than ideal due to first pass hepatic

metabolism.

To partially overcome this problem we treated mice with relatively high doses of

hPRL, hGH, and their SA20 derivatives. Still we observed an increase in the number of

lobulo-alveoli-like structures in the mammary glands of mice treated with hPRL, hGH

and their SA20 derivatives. In contrast to hPRL and hGH, SA20-hPRL and SA20-hGH

consistently increased the number of lobulo-alveoli in all mammary glands to a greater

extent, indicating that serum albumin binding effectively increased the efficacy of hPRL

and hGH in vivo. The half-life of mouse serum albumin is only 1 day (Stevens et al.,

1992); in contrast the half-life of human serum albumin is 19 days (Sterling, 1951).

Since mice metabolize proteins much faster than humans the circulating half-lives of

SA20-hPRL and SA20-hGH should be increased substantially in humans.

In summary, genetic fusion of a serum albumin binding peptide to hPRL and hGH

caused it to associate with serum albumin and extended their plasma half-life in vivo.

This strategy didn’t require downstream conjugation and therefore provided an

economically feasible alternative to PEGylation or HESylation. In contrast to

PEGylation and HESylation, the fusion of the small serum albumin binding peptide to the

N-terminus of hPRL and hGH had minimal deleterious effects upon signal transduction

and bioactivity in vitro and in vivo. Most importantly, SA20-hPRL or SA20-hGH can be

combined with existing sustained release formulations to further enhance their

effectiveness.

119 CHAPTER FOUR

IMPROVING THE PHARMACOKINETIC PROPERTIES OF

PROLACTIN AND GROWTH HORMONE RECEPTOR

AGONISTS BY FUSION IN TANDEM

Abstract

Human (h) prolactin (PRL) and growth hormone (GH) have two distinct binding

sites (Site 1 with high affinity; Site 2 with low affinity) that each interact with a prolactin

receptor (PRLR) to form a functional receptor dimer that activates signal transduction.

The G129R mutation in hPRL and the G120R mutation in hGH disrupt the structural integrity

of Site 2 such that the ligands retain the ability to bind to the first receptor with high affinity,

but act as receptor antagonists. In this study, we examined the ability of monomeric and

dimeric forms of hPRL, hGH, hPRL-G129R (G129R), and hGH-G120R (G120R) to (1)

bind to PRLRs; (2) induce conformational changes in PRLRs; (3) activate signaling

pathways associated with the PRLR; and (4) mediate cell proliferation in vitro. In

contrast to monomeric G129R, homodimeric G129R induced PRLR dimerization;

activated Janus-activated kinase 2/signal transducer and activator of transcription 5,

Ras/Raf/MAPK kinase/Erks, and phosphatidylinositol 3-kinase/Akt signaling cascades;

and stimulated Nb2 cell proliferation. Similarly, homodimeric G120R was able to

mediate signaling via the PRLR and to stimulate Nb2 cell proliferation. These

experiments demonstrate that a ligand must have two functional binding sites, but that

these may be Site 1 plus Site 2 or two Site 1s, to elicit receptor mediated signal transduction. The size of the ligand plays less of a role in receptor activation, suggesting

120 that the extracellular portion of the PRLR (and possibly the growth hormone receptor) is rather flexible and can accommodate larger ligands. These findings may have implications for designing multi-functional therapeutics that target this class of cytokine receptors.

121 Introduction

Prolactin (PRL) and growth hormone (GH) are members of the hematopoietin

family of cytokines. Despite their low amino acid sequence identity they have similar

tertiary structures (de Vos et al., 1992; Keeler et al., 2003), as do the extracellular regions

of the receptors to which they bind (Forsyth and Wallis, 2002). These peptide hormones

have four α-helical segments that are connected via loops which provide flexibility and allow the helices to be bundled in an anti-parallel (up-up-down-down) manner. Two

separate asymmetric receptor binding regions have been identified in these hormones,

each of which interacts with the equivalent region of a receptor to form a one ligand two receptor complex. Due to differences in affinity between the two receptor binding sites, binding occurs sequentially with the higher affinity site (Site 1) interacting with the first receptor before the lower affinity site (Site 2) can interact with a second receptor

(Cunningham et al., 1991; Ultsch et al., 1991; Fuh et al., 1993; Goffin et al., 1994; Fuh and Wells, 1995). Recently, this sequential binding process has been shown to be due to the functional coupling of the receptor binding sites. The interaction of Site 1 with the first receptor induces conformational changes in the ligand that create a functional Site 2

(Sivaprasad et al., 2004; Walsh et al., 2004).

Studies of human (h) GH, PRL, and placental lactogen (PL) have shown that the third α-helix is important to their structure and the function of Site 2. Substitution of a

Gly residue in this region with an Arg residue results in antagonists (hGH-G120R, hPRL-

G129R, hPL-G120R) with little or no biological activity (Chen et al., 1991a; Fuh et al.,

1993; Chen et al., 1994; Fuh and Wells, 1995; Goffin et al., 1996; Chen et al., 1999).

122 Comparative studies of the interaction of hGH-G120R and hGH with the extracellular domain (ECD) of the hGH receptor (hGHR-ECD) have shown that the mutation has minimal effects upon initial hormone-receptor interactions at Site 1 but disrupts the critical hormone-receptor-interactions at Site 2 (de Vos et al., 1992; Sundstrom et al.,

1996; Clackson et al., 1998); thereby preventing proper receptor dimerization.

Proper ligand-induced receptor dimerization induces conformational changes in the receptors that bring about the activation of receptor associated Janus-activated kinase

2 (JAK2) (Carter-Su et al., 1989; Campbell et al., 1994; Dusanter-Fourt et al., 1994;

Lebrun et al., 1994; Rui et al., 1994a), which stimulates signal transducer and activators of transcription 5 (STAT5) signaling. Receptor activation also leads to the activation of

Ras/Raf/MAPK kinase/Erks (Das and Vonderhaar, 1996) and phosphatidylinositol 3- kinase/Akt signaling pathways. It is primarily via these pathways that the receptors for these ligands induce cell differentiation, proliferation, and/or survival (Socolovsky et al.,

2001; Shillingford et al., 2002).

In addition to the monomeric forms of these ligands, dimeric and oligomeric forms have been identified in plasma of mammals and within pituitary extracts

(Schneider et al., 1975a; Stolar et al., 1984; Smith and Norman, 1990; Sinha, 1995;

Fahie-Wilson et al., 2005). Dimeric forms may arise from interchain disulfide-linkages or non-covalent aggregation; oligomeric forms exist and larger forms may also arise from immunoglobulin G (IgG) covalent and non-covalent complexes with monomeric or dimeric forms of these ligands (Schneider et al., 1977; Brostedt et al., 1990; Baumann,

1991; Sinha, 1995). The significance of these larger forms with reduced affinities for the

123 PRLR remains largely unknown. In addition to these natural ligands of the PRLR,

bivalent monoclonal antibodies (mAbs) have been developed that exhibit various degrees

of agonistic activity (Djiane et al., 1985; Okamura et al., 1989; Elberg et al., 1990) which

correlates with their ability to induce the phosphorylation of JAK2 (Rui et al., 1994b).

To better understand the molecular interactions between ligands and the PRLR,

we prepared dimeric forms of hGH, hPRL, hGH-G120R (G120R), and hPRL-G129R

(G129R), including homodimers (PRL-PRL, GH-GH) with four potentially functional

binding sites, heterodimers (PRL-G129R, G129R-PRL) with three potential binding sites and homodimers (G129R-G129R, G120R-G120R) with two potential binding sites. We demonstrate that the fusion of two antagonists (G129R or G120R) together restores some biological activity, even though the two functional binding sites are located in separate moieties. This suggests that the PRL/GH receptor family exhibits a considerable degree of flexibility in accommodating ligands and transducing their signals. This finding may have implications for designing ligand-based therapeutics that target this family of receptors.

124 Materials and Methods

Cell Lines and Reagents

T-47D human breast cancer cells and transformed human embryonic kidney cells

(HEK293) were obtained from the American Type Culture Collection (Manassas, VA).

Nb2-11 rat pre-T lymphoma cells (Nb2) were kindly provided by Dr. Arthur Buckley.

All reagents were purchased from Invitrogen (Carlsbad, CA) unless otherwise stated. T-

47D cells were maintained in RPMI Medium 1640 supplemented with 10% fetal bovine serum (FBS) and 10 µg/ml gentamicin. HEK293 cells were maintained in Dulbecco's

Modified Eagle Medium (DMEM) containing high glucose, 1 mM sodium pyruvate, 1 mM pyridoxine hydrochloride, and 2 mM L-glutamine; the medium was further

supplemented with 10% FBS, 100 U/ml penicillin, and 0.7 µM streptomycin. Nb2 cells

were maintained in Minimum Essential Medium (MEM) supplemented with 10% FBS,

10% horse serum (HS), 0.1 mM non-essential amino acids, 50 µg/ml streptomycin, 50

U/ml penicillin, and 0.1 mM 2-mercaptoethanol. The cell lines were incubated at 37°C

with humidity and 5% CO2.

Construction of hPRL and G129R Homo- and Heterodimer Expression Vectors

Previously, hPRL and G129R were cloned by inserting their cDNAs between the

Nde I and Xho I sites of the expression vector pET22b(+) (Novagen; Madison, WI).

These plasmids were used as DNA templates for PCR amplification of each moiety of the

recombinant dimers. The first moiety of hPRL or G129R was amplified using the

oligonucleotide primers 5’-TCATATGTTGCCCATCTGTCCCGGCG-3’ and 5’-

TGGATCCGCAGTTGTTGTTGTGGATG-3’; which incorporate an Nde I site at the N-

125 terminus and replace the C-terminal stop codon with a BamH I site. The second moiety

of hPRL or G129R was amplified using the primers 5’-

TGGATCCTTGCCCATCTGTCCCGGCG-3’ and 5’-TCTCGAGTTAGCAGTTGTTGT

TGTGG-3’; which incorporate a BamH I site at the N-terminus and an Xho I site after the stop codon. The PCR products were ligated with pCR2.1 T/A cloning vector and transformed into E. coli TOP10 cells obtained from Invitrogen. Positive clones were identified by plating the cells on Luria-Bertani (LB) agar plates containing 100 µg/ml ampicillin and 80 µg/ml 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside. The

plasmids were isolated using the QIAprep® Spin Miniprep Kit (Qiagen, Valencia, CA)

and the Nde I-PRL-BamH I, Nde I-G129R-BamH I, BamH I-PRL-stop-Xho I and BamH

I-G129R-stop-Xho I DNA fragments were isolated by restriction digestion and separated

by electrophoresis. The DNA fragments were purified from the agarose gel using the

QIAquick® Gel Extraction Kit (Qiagen) and ligated with Nde I and Xho I cut pET22b(+)

to create pET22b-PRL-PRL, pET22b-G129R-G129R, pET22b-PRL-G129R and pET22b-

G129R-PRL. A similar strategy was used to create pET22b-GH-GH and pET22b-

G120R-G120R. All oligonucleotides were synthesized by Integrated DNA Technologies

(Coralville, IA), restriction and modifying enzymes were purchased from New England

BioLabs (Beverly, MA), and PCR reactions were performed using the Advantage-HF™

Kit (BD Biosciences, Palo Alto, CA).

Production and Purification of hPRL and G129R Homo- and Heterodimers

Chemically competent Rosetta™(DE3) pLysS E. coli cells (Novagen) were transformed with the constructs and propagated overnight at 37°C with agitation in LB

126 broth containing 100 µg/ml ampicillin. The next morning 40 ml of each of the starter

cultures was used to inoculate one liter of Terrific Broth in baffled flasks and the cultures were grown to an O.D.600 of 0.9 before induction of protein production with 1 mM isopropyl-β-thiogalactoside (Alexis Biochemicals, San Diego, CA). After 4 h of induction the cells were harvested by centrifugation at 5,000 x g for 5 min at 4°C and

resuspended in Solution 1 (20 mM Tris·HCl; 10 mM EDTA; 0.5% Triton X-100; pH

adjusted to 8.0) to which 0.4 mg/ml lysozyme was added. The cells were incubated at

room temperature for 1 h to weaken the cell membranes before the cells were sheared on

ice with five 1-min ultrasonic pulses (30% duty at setting 7) using a 550 Sonic

Dismembrator (Fisher Scientific, Waltham, MA). The insoluble fraction containing the inclusion bodies was recovered by centrifugation at 12,000 x g for 30 min at 4ºC and resuspended in Solution 2 (20 mM Tris·HCl; 10 mM EDTA; 1 M urea; pH adjusted to

8.0). The inclusion bodies were collected by centrifugation at 12,000 x g for 30 min at

4ºC and were solubilized and denatured in Solution 3 (20 mM Tris·HCl; 10 mM EDTA; 8

M urea; 1% v/v 2-mercaptoethanol; pH adjusted to 8.0). The proteins were refolded by dialysis against baths of 20 mM Tris·HCl; 10 mM EDTA, pH 8.0 containing decreasing amounts of urea (4M urea, 2 M urea, 0 M urea, 0 M urea). The refolded proteins were filtered through 0.45

µm filters and purified by anion-exchange chromatography using a 5 ml HiTrap Q Sepharose

HP column (Amersham Biosciences, Piscataway, NJ) on an ÄKTA fast protein liquid chromatography system (Amersham Biosciences). The concentration of the purified proteins was determined using the Coomassie Plus™ Protein Assay Reagent (Pierce, Rockford, IL).

The purified proteins were separated by non-reducing and reducing SDS-PAGE along with

127 BSA standards (Pierce, Rockford, IL) and stained with SYPRO® Orange (Molecular Probes,

Eugene, OR) to confirm their concentration and purity. The identity of the proteins was

confirmed by Western blotting.

Western Blot Analysis

Purified monomers, homodimers and heterodimers of hPRL and G129R were

separated on a 12% polyacrylamide gel by reducing SDS-PAGE. The proteins were

transferred by Western blot to a Hybond™ Enhanced Chemiluminescence (ECL)

nitrocellulose membrane (Amersham Biosciences) at 25 V overnight. The nitrocellulose membrane was blocked for 1 h at room temperature in Tris buffered saline (TBS) containing 0.05% Tween 20 (TBS-T) and 5% milk. The membrane was incubated at room temperature with agitation for 4 h in the blocking solution containing a 1:5000

dilution of rabbit anti-hPRL antiserum (National Hormone and Pituitary Program, NIH,

Bethesda, MD). The blots were washed three times with TBS-T, and incubated for 2 h at

room temperature with agitation in blocking solution containing a 1:2000 dilution of goat

anti-rabbit conjugated HRP (Bio-Rad, Hercules, CA). Blots were washed three times

with TBS-T and incubated for 1 min with ECL™ Western Blotting Detection Reagents

(Amersham Biosciences). The blots were exposed to Biomax MR film (Kodak,

Rochester, NY) and the film was developed with a Konica SRX-101A processor (Konica

Minolta Medical Imagining, Wayne, NJ).

Radio-Receptor Binding Assay

One million cells per well of T-47D human breast cancer cells were seeded in six-

well tissue culture plates and grown to confluency. Cells were starved in serum-free

128 RPMI Medium 1640 for 1 h, and incubated for 2 h at room temperature in serum-free

medium containing 5 x 104 cpm of 125I-labeled hPRL (specific activity, 40 µCI/µg; NEN

Perkin-Elmer, Boston, MA) with or without various concentrations of monomeric or

homodimeric hPRL and G129R. An excess of hPRL was used to determine non-specific

binding. Cells were washed three times with serum-free RPMI Medium 1640 and lysed

in 0.5 ml of 0.1 N NaOH/1% SDS. Bound radioactivity was determined using a

Beckman LS6500 Liquid Scintillation Counter calibrated to detect 125I using a wide

spectrometer window. The percentage of specific binding was calculated and compared

among the samples.

BRET2 Assay (Performed by Dunyong Tan at the University of California,

Riverside)

Bioluminescence resonance engegy transfer (BRET2) assays were performed

precisely as described by Tan et al. (Tan et al., 2005). Briefly, HEK293 cells were

seeded in six-well tissue culture plates at a density of 5 x 105 cells per well and grown

overnight to 90-95% confluency. Eukaryotic expression vectors encoding the long form

(LF) or short forms (SF1a and SF1b) of the human PRLR tagged with a variant of green

fluorescent protein (GFP2) [LF-GFP2, SF1a-GFP2, SF1b-GFP2] and Renilla reniformis

luciferase (Rluc) [LF-Rluc, SF1a-Rluc, SF1b-Rluc] at the C-terminus were transiently

transfected into the cells. After 48 h incubation the transfected cells were washed, detached, and resuspended in BRET2 buffer at a density of 2 x 106 cells per ml. After

incubation at 37°C for at least 1 h, the relative expression levels of the GFP2- and Rluc-

tagged PRLR constructs were estimated by comparing their fluorescence and

129 bioluminescence to that of cells transfected with a GFP2-Rluc fusion plasmid.

Bioluminescence scanning and BRET2 signal detection were carried out immediately

after the addition of the ligand (45.7 nM) and 5 µM of DeepBlueC™. The bioluminescence from the oxidation of the DeepBlueC™ by Rluc has a peak emission at

390–405 nm and the fluorescence from the GFP2 acceptor has a peak emission at 500–

520 nm. Energy transfer from Rluc to GFP2 was defined as the BRET ratio, which was determined by calculating the ratio of light emitted by receptor-GFP2 (500-520 nm) over

the light emitted by receptor-Rluc (385-420 nm). Values were corrected by subtracting

the BRET background signal detected in nontransfected cells (Emission500 nm-520 nm-

background500 nm-520 nm)/(Emission385 nm-420 nm-background385 nm-420 nm).

STAT5, Erk1/2, and Akt Phosphorylation Assays

T-47D cells were resuspended in RPMI Medium 1640 supplemented with 10%

charcoal/dextran treated FBS (CSS; HyClone, Logan, UT) and 10 µg/ml gentamycin and

plated in twelve-well tissue culture plates. The cells were grown overnight to approximately 80% confluency and depleted for 1-72 h in RPMI Medium 1640 supplemented with 0.5% CSS to determine which condition best achieved the lowest background activation of signaling molecules. Since prolonged depletion produced no improvement in this regard, 1 h depeletions were used. Cells were treated for the indicated amounts of time with the indicated concentrations of monomeric, homodimeric, or heterodimeric hPRL, G129R, hGH, or G120R. The cells were washed with ice cold

phosphate buffered saline (PBS) and lysed in 100 µl of Lysis Buffer (20 mM Tris·HCl, pH 7.4; 1% Igepal CA630; 6 mM deoxycholic acid; 150 mM NaCl; 1 mM EDTA, pH

130 8.0) containing protease inhibitors (1 µg/ml aprotinin; 1 µg/ml leupeptin; 170 µg/ml

PMSF; 180 µg/ml sodium orthovanadate). The lysates were homogenized and incubated on ice for 20 minutes before clarification by centrifugation at 12,000 x g at 4°C for 30 min. Approximately 40 µg of the supernatants were used for Western blot analysis as described above using 10% or 4-15% polyacrylamide gels.

Blots were initially performed with a 1:1000 dilution of mouse anti-phospo

STAT5A/B antibody (Upstate, Lake Placid, NY) and a 1:2000 dilution of goat anti- mouse-HRP conjugate (Bio-Rad). Blots were reprobed with a 1:1333 dilution of rabbit anti-STAT5 polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) and a

1:2000 dilution of goat anti-rabbit HRP conjugate (Bio-Rad); a 1:1333 dilution of mouse anti-phospho ERK (Santa Cruz Biotechnology), and a 1:2000 dilution of goat anti- mouse-HRP conjugate; and lastly with a 1:200 dilution of rabbit anti-phospo Akt antibody (Santa Cruz Biotechnology) and a 1:2000 dilution of goat anti-rabbit HRP conjugate.

Rat Nb2 Cell Proliferation Assay

Rat Nb2 cells were resuspended to a final concentration of 5.5 x 105 cells per ml in MEM supplemented with 0.1 mM non-essential amino acids, 10% low lactogen HS

(Cat. 2921154; MP Biomedicals, Irvine, CA), 50 µg/ml streptomycin, 50 U/ml penicillin, and 0.1 mM 2-mercaptoethanol. The cells were depleted overnight and reseeded in flat bottom 96-well plates at a density of 20,000 cells per well in 150 µl of medium.

Treatments were added to each well in 50 µl volumes and the cells were incubated for 72 h. The relative number of cells was determined by incubating the cells for 4 h with 40 µl

131 of 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H- tetrazolium/phenazine methosulfate (MTS/PMS) solution (CellTiter 96 AQueous non- radioactive cell proliferation kit; Promega, Madison, WI) per well. The reduction of

MTS by living cells was determined by measuring the absorbance at 490 nm using a

Benchmark microplate reader (Bio-Rad). Preliminary experiments demonstrated that cell plating densities up to 45,000 showed a linear relationship to OD 490. All assays were performed at least three times in quadruplicate.

In Vivo Pharmacokinetics

Two groups of four BALB/C mice were injected intraperitoneally with 200 µg of monomeric or dimeric hPRL. Blood was collected by tail-vein bleeding 1 h after injection and collected hourly for an additional 3 h. Plasma samples were diluted when necessary with the zero calibrator and subject to an immunoradiometric assay (IRMA) for

hPRL (Coat-A-Count® Prolactin IRMA; DPC, Los Angeles, CA). The concentrations of

monomeric and dimeric hPRL were extrapolated and plotted versus time and the areas

under the curve were calculated. Human PRL and dimeric hPRL standards of known

concentration were detected equally using the IRMA kit since the standard curve was in

nanograms per milliliter and not molar concentrations. Blood from control mice

indicated that the IRMA kit was insensitive to endogenous mouse PRL.

132 ABC Non-Reducing SDS-PAGE Reducing SDS-PAGE Western Blot M 1 2 3 4 5 6 M 1 2 3 4 5 6 1 2 3 4 5 6 100 80 50 40 30 25 20

Fig. 4.1 Non-reducing SDS-PAGE, reducing SDS-PAGE, and Western blot analysis of purified monomeric and dimeric hPRL derivatives

Monomers, homodimers, and heterodimers of hPRL and G129R were separated by non-reducing SDS-PAGE (A) or reducing SDS-PAGE (B) on 12% polyacrylamide gels and stained with SYPRO® Orange to confirm their size and purity. M, BenchMark Ladder (Invitrogen); Lane 1, PRL; Lane 2, G129R; Lane 3, PRL-PRL; Lane 4, G129R- G129R; Lane 5, PRL-G129R; and Lane 6, G129R-PRL. The identity of the proteins was confirmed by Western blotting with an antibody that detects the hPRL and G129R moieties (C).

133 Results

Construction, Expression, and Purification of Dimeric Derivatives of hPRL and

hGH

Homodimers and heterodimers of hPRL and G129R were recombinantly

engineered by gene fusion. The cDNAs encoding the mature form of the proteins were

ligated together via a small linker (GGATCC), and cloned into an E. coli expression

vector. Expression of the fusion gene in E. coli resulted in a single fusion protein in

which the C-terminus of the first moiety was connected to the N-terminus of the second

moiety by a Gly-Ser peptide linker. The proteins were found to accumulate in the

insoluble fraction as inclusion bodies. The inclusion bodies were isolated, denatured,

refolded, and purified by anion-exchange chromatography to yield preparations of greater than

95% purity. The dimeric fusion proteins were analyzed by non-reducing (Fig. 4.1A) and reducing SDS-PAGE (Fig. 4.1B) along with hPRL and G129R. Human PRL and G129R were predominantly monomeric (20 kDa) under non-reducing conditions with only a small portion remaining unfolded (25 kDa) or forming covalently linked dimers via interchain disulfide linkages (40 kDa) (Fig. 4.1A, lanes 1 and 2). No covalently linked multimeric forms were observed for the recombinantly engineered homodimers and heterodimers of hPRL and G129R after purification; however, multiple bands were observed which would suggest some improper intrachain disulfide linkages (Fig. 4.1A, lanes 3-6). This was anticipated since the C-terminal cysteines of the first moiety (Cys191, Cys199) are separated from the N-terminal cysteines

(Cys4, Cys11) of the second moiety by only a small linker (Gly-Ser). Reducing SDS-PAGE revealed that purified dimers have sizes corresponding to their predicted molecular mass of 46

134 kDa (Fig. 4.1B, lanes 3-6) and Western blotting with an antibody that detects hPRL and G129R

confirmed the identity of purified dimers (Fig. 4.1C, lanes 3-6).

Homodimers of hPRL and G129R Retain the Ability to Bind to PRLRs

To examine if the homodimers of hPRL and G129R retain the ability to bind to

PRLRs, we measured their ability to compete with 125I-labeled hPRL for binding to

PRLRs expressed on the surface of T-47D human breast cancer cells. Monomers and

homodimers of hPRL and G129R effectively displaced the binding of 125I-hPRL (Fig.

4.2), confirming that the dimers retain the ability to specifically bind to PRLRs. The

125 effective concentrations necessary to displace 50% of the I-labeled hPRL (EC50) were

calculated. There was no statistical difference between the EC50 of dimeric hPRL (0.97 ±

0.23 nM), dimeric G129R (1.17 ± 0.1 nM), and monomeric hPRL (0.96 ± 0.29 nM);

however, there were statistical differences between the EC50 of these ligands and monomeric G129R (1.96 ± 0.25 nM).

Homodimeric G129R Induces a BRET Signal Consistent with Receptor

Dimerization

To determine if dimeric G129R truly has a second functional binding site, we examined whether or not it could induce conformational changes in PRLRs consistent with dimerization. Figure 4.3A shows representative luminescence scans in the absence

or presence of monomeric hPRL, monomeric G129R, and homodimeric G129R. In the

absence of ligand or the presence of monomeric G129R there is minimal emission of a

BRET signal from HEK293 cells cotransfected with tagged human PRLRs.

135 100 PRL G129R**

I-hPRL I-hPRL PRL-PRL 125 G129R-G129R

50 %Specific Binding of %Specific Binding of

00.1110100

[Ligand] (nM)

Fig. 4.2 Competitive binding of monomeric and homodimeric hPRL derivatives to PRLRs on T-47D cells

The ability of monomeric and homodimeric hPRL and G129R to bind to PRLRs was determined by measuring their ability to compete with 125I-labeled hPRL for binding to the surface of T-47D cells as described in the Materials and Methods. Each data point represents the mean ± SD of at least three independent experiments. The EC50 for monomeric hPRL, homodimeric hPRL, homodimeric G129R, and monomeric G129R were determined to be 0.96 ± 0.29 nM, 0.97 ± 0.23 nM, 1.17 ± 0.1 nM, and 1.96 ± 0.25 nM, respectively. **, p<0.005, for difference in EC50 from that of monomeric hPRL.

136 A 1 DBC only 0.8 DBC + PRL 0.8 0.6 0.6 ) BRET 3 0.4 0.4

0.2 0.2

0.8 0.8 DBC + G129R DBC + G129R-G129R

Luminescence(10 0.6 0.6 BRET 0.4 0.4

0.2 0.2 300 400 500 600 300 400 500 600

Wavelength (nm) B

DBC only DBC + G129R-G129R

0.3

** 0.2 ** *

0.1

0 2 2 2 -GFP P GFP c/LF 1a-GF F1b- -Rlu c/SF uc/S LF a-Rlu b-Rl SF1 SF1

Fig. 4.3 Monomeric hPRL and homodimeric G129R-induced homodimerization of PRLRs

BRET2 assays were performed within 48 h after transfection of HEK 293 cells in the presence of 5 µM DeepBlueC™ (DBC) with or without monomeric hPRL, monomeric G129R, or homodimeric G129R as described in the Materials and Methods. Representative bioluminescence scans from the DBC reaction in the absence of ligand and after the addition of monomeric hPRL (45.7 nM), monomeric G129R (45.7 nM), or homodimeric G129R (45.7 nM) to cells transiently transfected with SF1a-GFP2/SF1a- Rluc (A). Note the absence of a BRET signal in the absence of monomeric hPRL or homodimeric G129R. Averaged BRET ratios from four or more separate determinations on different preparations of cells transfected with LF, SF1a, or SF1b (B). The BRET ratio was calculated as (Emission500–520 nm – Background500–520 nm)/(Emission385–420 nm – Background385–420 nm). *, p<0.05; **, p<0.01, for difference with and without monomeric hPRL or dimeric G129R.

137 In contrast, the addition of monomeric hPRL or homodimeric G129R induces a

significant BRET signal in HEK 293 cells cotransfected with tagged human PRLRs,

SF1b-Rluc/SF1b-GFP2 (Fig. 4.3A). The BRET ratios for HEK293 cells cotransfected with LF-Rluc/LF-GFP2, SF1a-Rluc/SF1a-GFP2, or SF1b-Rluc/SF1b-GFP2 increased

approximately 4, 3, and 3-fold, respectively, upon treatment with dimeric G129R (Fig.

4.3B).

Dimerization of G129R Restores Human PRLR Mediated Signal Transduction

To determine if the receptor dimers induced by homodimeric G129R are

functional we examined its ability to activate signaling molecules associated with these

pathways. We also produced homodimeric G120R, a GHR as well as PRLR antagonist,

to test whether dimerization of this antagonist would also produce an agonist. Treatment

of T-47D cells with various concentrations of hPRL or hGH resulted in the

phosphorylation of STAT5, Erk1/2, and Akt (Fig. 4.4, A and B). Monomeric G129R and

G120R, which block the conformational changes necessary for the formation of a

functional receptor dimer, did not elicit phosphorylation of any of these signaling

molecules (Fig. 4.4, A and B). In contrast to the monomeric PRLR antagonists,

homodimers of G129R and G120R activated STAT5, Erk1/2, and Akt in a dose-

dependent manner (Fig. 4.5, A and B).

Dose response and time course experiments were performed to examine the

potency (Fig. 4.6) and efficiency (Fig. 4.8) of dimeric G129R in the activation of STAT5

in comparison to hPRL. Dimeric G129R was able to induce STAT5 phosphorylation

(Fig. 4.6B), but required approximately 3- fold higher molar concentrations (Fig. 4.6A).

138 A PRL G129R nM : 0 5 50 500 5 50 500 0 p-Akt

p-Erk1 p-Erk2

p-STAT5

STAT5 B GH G120R nM : 0 5 50 5000 5 50 500 p-Akt

p-Erk1 p-Erk2

p-STAT5

STAT5

Fig. 4.4 Western blot analysis of the phosphorylation status of Akt, Erk1/2, and STAT5 in response to monomeric hPRL and hGH derivatives

T-47D cells were plated as described in the Materials and Methods and serum starved overnight. The cells were treated for 30 min with 5 nM, 50 nM, or 500 nM of monomeric hPRL or G129R (A) or treated with monomeric hGH or G120R (B). Cell lysates were harvested and 40 µg of lysate from each treatment was separated by reducing SDS-PAGE using 4-15% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed with antibodies recognizing phospho- Akt1/2/3, phospho-Erk1/2 and phospho-STAT5A/B. Equal loading was confirmed by reprobing for STAT5.

139 A PRL-PRL G129R-G129R PRL nM : 0 5 50 500 5 50 500 5500 p-Akt

p-Erk1 p-Erk2 p-STAT5

STAT5

B GH-GH G120R-G120R nM : 0 5 50 5000 5 50 500 p-Akt

p-Erk1 p-Erk2 p-STAT5

STAT5

C PRL-G129R G129R-PRL PRL nM : 0 5 50 500 5 50 500 5500 p-Akt

p-Erk1 p-Erk2 p-STAT5

STAT5

Fig. 4.5 Western blot analysis of the phosphorylation status of Akt, Erk1/2, and STAT5 in response to dimeric hPRL and hGH derivatives

T-47D cells were plated as described in the Materials and Methods and serum starved overnight. The cells were treated for 30 min with 5 nM, 50 nM, or 500 nM of: homodimeric hPRL or G129R (A); homodimeric hGH or G120R (B); or heterodimers of hPRL and G129R (C). Cell lysates were harvested and 40 µg of lysate from each treatment was separated by reducing SDS-PAGE using 4-15% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed with antibodies recognizing phospho-Akt1/2/3, phospho-Erk1/2 and phospho-STAT5A/B. Equal loading was confirmed by reprobing for STAT5.

140 A PRL nM : 0 1 3.3 10 33 100 333

p-STAT5

STAT5

B G129R-G129R nM : 0 1 3.3 10 33 100 333

p-STAT5

STAT5

C PRL-PRL nM : 0 1 3.3 10 33 100 333

p-STAT5

STAT5

D PRL-G129R nM : 0 1 3.3 10 33 100 333

p-STAT5

STAT5

E G129R-PRL nM : 0 1 3.3 10 33 100 333

p-STAT5

STAT5

Fig. 4.6 Dose-dependent activation of STAT5 phosphorylation induced by monomeric and dimeric hPRL derivatives

T-47D cells were plated as described in the Materials and Methods and serum starved overnight. The cells were treated with increasing doses of monomeric (A), homodimeric (B, C), or heterodimeric (D, E) hPRL or G129R and the efficiency of STAT5 phosphorylation was examined after 20 min. Cell lysates were harvested and 40 µg of lysate from each treatment was separated by reducing SDS-PAGE using 4-15% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed with anti-phospho-STAT5. Equal loading was confirmed by reprobing the membranes with an anti-STAT5 antibody.

141 The difference in potency between dimeric G129R and monomeric hPRL may be explained in part by the difference in their relative binding affinity to human PRLRs (Fig.

4.2), but may also be related to the adjustments in the receptor that are necessary to bind two Site 1s. We also compared the efficiency of dimeric G120R and monomeric hGH at activating STAT5, Erk1/2, and Akt (Fig. 4.7). Dimeric G120R induced the phosphorylation of these signaling molecules in a dose-dependent manner (Fig. 4.7B), but required approximately 30-fold higher concentrations to achieve similar levels of activation induced by monomeric hGH (Fig. 4.7A).

Consistent with having a lower affinity, the efficiency of signal transduction induced by 10 nM dimeric G129R (Fig. 4.8B) or dimeric G120R (Fig. 4.9B) were lower than that of 10 nM of hPRL (Fig. 4.8A) and hGH (Fig. 4.9A), requiring a longer time course to achieve similar activation levels. The phosphorylation of STAT5 induced by

10 nM dimeric G129R (Fig. 4.8A) was detectable after 20 min whereas it was detectable after 10 min for hPRL (Fig. 4.8B). The phosphorylation of STAT5 by dimeric G120R

(Fig. 4.9B) required 30 min to achieve the same level as hGH (Fig. 4.9A) after 2 min; similarly, it took much longer for dimeric G120R to stimulate phosphorylation of Erk1/2.

We further examined the signaling induced by homodimers of hPRL, homodimers of hGH, and heterodimers of hPRL and G129R. We show that these dimers activate the same signaling pathways as hPRL and hGH (Fig. 4.5). Heterodimers of hPRL and

G129R (Fig. 4.6, D and E, and Fig. 4.8, D and E) exhibited slightly lower signaling efficiencies than monomeric hPRL (Figs. 4.6A and 4.8A); whereas, homodimers of hPRL

142 (Figs. 4.6C and 4.8C) and hGH (Figs. 4.7C and 4.9C) activated STAT5 nearly as well as

monomeric hPRL (Figs. 4.6A and 4.8A) and hGH (Figs. 4.7A and 4.9A).

Dimeric G129R and G120R Induce Proliferation of Rat Nb2 Cells in a Dose-

Dependent Manner

To examine the biological activity of homodimeric G129R, homodimeric G120R, and the other dimeric derivatives we measured their ability to induce the proliferation of rat Nb2 cells which proliferate in response to lactogenic hormones. The relative number of viable cells was determined after 72 h treatment by colorimetrically measuring the reduction of MTS by living cells. Treatment of the cells with increasing doses of

homodimeric G129R (Fig. 4.10A) or homodimeric G120R (Fig. 4.10C) stimulated cell

proliferation in a dose-dependent manner. Homodimers of hPRL (Fig. 4.10B), homodimers of hGH (Fig. 4.10D), and heterodimers of hPRL and G129R (Fig. 4.10, E and F) were more potent agonists than homodimers of G129R (Fig. 4.10A) and

homodimers of G120R (Fig. 4.10C). The proliferation induced by these ligands in rat

Nb2 cells correlated well with the ability of the ligands to bind receptors (Fig. 4.2) and induce signaling in human T-47D cells (Figs. 4.5-4.7). Dimeric G129R appeared to be a more potent agonist than dimeric G120R.

To provide further evidence that the agonism observed for homodimeric G129R was due to the presence of two functional Site 1s, we made a series of deletions to the C- terminus of the second moiety of G129R (Fig. 4.11). Removal of a portion of helix IV by deleting the last 25 amino acid residues (G129R2ΔC25) reduced the agonism by

approximately 10-fold. Further removal of helix IV by deleting 40 amino acid residues

143 (G129R2ΔC40) resulted in further reduction in agonist activity to a level similar to monomeric G129R, strongly suggesting that helix IV in the second moiety of G129R plays an important role in forming the second functional binding site. Interestingly, truncations into and beyond helix III (G129R2ΔC80 thru G129R2ΔC160, with the exception of G129R2ΔC180) abolished all residual agonistic activity.

144 A GH nM : 0 1 3.3 10 33 100 333

p-Akt

p-Erk1 p-Erk2

p-STAT5

STAT5

B G120R-G120R GH nM : 0 1 3.3 10 33 100 333 1 10 0 p-Akt

p-Erk1 p-Erk2

p-STAT5

STAT5

C GH-GH GH nM : 0 1 3.3 10 33 100 333 1 10 0

p-Akt

p-Erk1 p-Erk2

p-STAT5

STAT5

Fig. 4.7 Dose-dependent induction of STAT5, Erk1/2 and Akt phosphorylation induced by monomeric and dimeric hGH derivatives

T-47D cells were plated as described in the Materials and Methods and serum starved overnight. The cells were treated with increasing doses of monomeric hGH (A), homodimeric G120R (B) or homodimeric hGH (C) and the efficiency of activation was examined after 30 min. Cell lysates were harvested and 40 µg of lysate from each treatment was separated by reducing SDS-PAGE using 10% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed with antibodies that recognize phospho-Akt1/2/3, phospho-Erk1/2, and phospho-STAT5A/B. Equal loading was confirmed by reprobing the membranes with an anti-STAT5 antibody.

145 A PRL (10 nM) min :0251020304050 60 0 p-Erk1 p-Erk2

p-STAT5

STAT5

B G129R-G129R (10 nM) min :0251020304050 60 0 p-Erk1 p-Erk2

p-STAT5

STAT5

C PRL-PRL (10 nM) min :0251020304050 60 0 p-Erk1 p-Erk2

p-STAT5

STAT5

D PRL-G129R (10 nM) min : 0251020304050 60 0 p-Erk1 p-Erk2

p-STAT5

STAT5

E G129R-PRL (10 nM) min :02 51020304050 60 0 p-Erk1 p-Erk2

p-STAT5

STAT5

Fig. 4.8 Time course of Erk1/2 and STAT5 phosphorylation in T-47D cells in response to monomeric and dimeric hPRL derivatives

T-47D cells were plated as described in the Materials and Methods and serum starved overnight. The cells were treated with 10 nM of monomeric hPRL (A), homodimeric G129R (B), homodimeric hPRL (C) or heterodimeric hPRL and G129R (D and E) for various durations (0, 2, 5, 10, 20, 30, 40, 50, and 60 min). Cell lysates were harvested and 40 µg of lysate from each treatment was separated by reducing SDS-PAGE using 4-15% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed with antibodies that recognize phospho-Erk1/2 and phospho- STAT5A/B. Equal loading was confirmed by reprobing the membranes with an anti- STAT5 antibody.

146 A 10 nM GH Min: 0 2 5 10 20 30 40 50 60 0 p-Akt

p-Erk1 p-Erk2 p-STAT5

STAT5

B 10 nM G120R-G120R Min: 0 2 5 10 20 30 40 50 60 0 p-Akt

p-Erk1 p-Erk2 p-STAT5

STAT5

C 10 nM GH-GH Min: 0 2 5 10 20 30 40 50 60 0 p-Akt

p-Erk1 p-Erk2 p-STAT5

STAT5

Fig. 4.9 Time course of STAT5, Akt, and ERK1/2 phosphorylation in T-47D cells in response to monomeric and dimeric hGH derivatives

T-47D cells were plated as described in the Materials and Methods and serum starved overnight. The cells were treated with 10 nM of monomeric hGH (A), homodimeric G120R (B) or homodimeric hGH (C) for various durations (0, 2, 5, 10, 20, 30, 40, 50, and 60 min). Cell lysates were harvested and 40 µg of lysate from each treatment was separated by reducing SDS-PAGE using 10% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed with antibodies that recognize phospho-Akt1/2/3, phospho-Erk1/2, and phospho-STAT5A/B. Equal loading was confirmed by reprobing the membranes with an anti-STAT5 antibody.

147 AB 1.6 1.6 1.4 1.4 1.2 1.2 1 1 490 490 0.8 0.8 O.D. O.D. 0.6 O.D. 0.6 0.4 PRL 0.4 PRL 0.2 G129R-G129R 0.2 PRL-PRL G129R G129R

0 -12 -11 -10 -9 -8 -7 0 -12 -11 -10 -9 -8 -7 Ligand (Log M) Ligand (Log M) CD 1.6 1.6 1.4 1.4 1.2 1.2

490 1 1 490 0.8 0.8 O.D. O.D. 0.6 O.D. 0.6 0.4 GH 0.4 GH 0.2 G120R-G120R 0.2 GH-GH G120R G120R

0 -12 -11 -10 -9 -8 -7 0 -12 -11 -10 -9 -8 -7 Ligand (Log M) Ligand (Log M) EF 1.6 1.6 1.4 1.4 1.2 1.2 1 1 490 490 0.8 0.8 O.D. 0.6 O.D. 0.6 0.4 PRL 0.4 PRL 0.2 PRL-G129R 0.2 G129R-PRL G129R G129R

0 -12 -11 -10 -9 -8 -7 0 -12 -11 -10 -9 -8 -7 Ligand (Log M) Ligand (Log M)

Fig. 4.10 Rat Nb2 cell proliferation in response to monomeric and dimeric hPRL and hGH derivatives

Rat Nb2 cells were treated with various doses of monomeric, homodimeric, or heterodimeric hPRL, G129R, hGH, or G120R for 72 h as described in the Materials and Methods. Cell proliferation was assessed by colorimetrically measuring the reduction of MTS. Each data point represents the mean ± SD of at least three independent experiments performed in quadruplicate. Rat Nb2 cells proliferate in response to picomolar concentrations of hPRL and hGH; G129R and G120R exhibit partial agonism at concentrations more than 2-log higher than that of hPRL and hGH. This partial agonism has been attributed to the rat PRLR forming a more stable interaction with binding site 2 of these ligands than the human PRLR.

148 1.6 PRL G129R2 1.4 G129R2ΔC25

G129R2Δ C40 1.2 G129R2Δ C60 490 G129R2Δ C180

O.D. G129R 1.0 G129R2Δ C80 G129R2Δ C100 0.8 G129R2Δ C120 G129R2Δ C142 G129R2Δ C160 0-11-9-7-12 -10 -8 Ligand (Log M)

Fig. 4.11 Examination of rat Nb2 cell proliferation in response to truncated forms of dimeric G129R

Rat Nb2 cells were treated with various doses of hPRL, G129R, or truncated forms of dimeric G129R (G129R2) for 72 h as described in the Materials and Methods. Cell proliferation was assessed by colorimetrically measuring the reduction of MTS. Each data point represents the mean ± SD of at least two independent experiments performed in quadruplicate.

149 Discussion

Human PRL and hGH share limited amino acid sequence identity, yet they both

can bind to PRLRs with high affinity to form functional receptor dimers. These ligands

have two separate and different binding sites that each interact with a PRLR in a

sequential manner to form a 1:2 ligand-receptor complex that activates signal

transduction (Cunningham and Wells, 1989, 1991; Ultsch et al., 1991; de Vos et al.,

1992; Fuh et al., 1993). Although the interactions of hPRL and hGH with the PRLR

overlap they are not identical (Rozakis-Adcock and Kelly, 1992), hGH requires the

presence of zinc to bind with high affinity (Cunningham et al., 1990b). The ability of

hGH to adapt to both the GHR and PRLR has largely been attributed to the local

conformational flexibility of its binding loops (Kossiakoff et al., 1994). Studies have

shown that the second half of the long loop between helices I and II is important to the

structure and/or function of both GH and PRL (Cunningham et al., 1989; Goffin et al.,

1992). Conformational changes in this region of the loop were observed for a high

affinity hGH variant (Ultsch et al., 1994) and amino acid changes to this region were

necessary to engineer PRL to bind to the GHR (Cunningham et al., 1990a). These studies

suggest that the natural ligands of the PRLR exhibit local conformational flexibility.

The ability of the PRLR and GHR to be activated by ligands other than the monomeric forms suggests that the ECD of the PRLR and GHR also exhibit flexibility.

In addition to monomers, dimeric and oligomeric forms of PRL, GH, and PL have been identified in the circulation of mammals and account for a small portion of the total plasma levels of these ligands (Stolar et al., 1984; Smith and Norman, 1990; Sinha, 1995;

150 Fahie-Wilson et al., 2005). Dimeric forms may arise from interchain disulfide linkages, linkages between glycosylated monomers, or be due to non-covalent interactions.

Purified disulfide-linked dimers of hPL have been shown to retain the ability to bind to mammary glands with high affinity and to be biologically active in vitro (Schneider et al.,

1975b; Schneider et al., 1977). A natural dimeric form of hGH with an extremely stable disulfide-linkage (MER-45-kDA hGH) has been characterized (Grigorian et al., 2005) which binds to both the GHR and PRLR with high affinity (Grigorian et al., 2002). It has been shown to stimulate Nb2 cell proliferation (Grigorian et al., 2003), but to inhibit the proliferation of breast cancer cell lines (Muñoz et al., 2004), suggesting either that it acts as an antagonist in the presence of autocrine PRL or that it antagonizes the GHR, or both

(Muñoz et al., 2005). In addition to these natural ligands, mAbs developed against the

ECD of the PRLR have been shown to be weak agonists (Djiane et al., 1985; Okamura et al., 1989; Elberg et al., 1990), with activities approximately 100-fold lower than natural monomeric ligands in the rat Nb2 assay (Rui et al., 1994b). These bivalent mAbs were inactive as monovalent Fab fragments, confirming that the dimerization and activation of

PRLRs occurs only when a ligand has two functional binding sites. Whether or not these mAbs recognize epitopes involved in PRL binding remains to be determined.

To specifically address whether or not the ligand-binding sites of the PRLR are flexible enough to accommodate larger ligands, we recombinantly engineered dimeric forms of the ligands (hPRL, hGH) and their antagonists (G129R, G120R). Previous studies have shown that mutations in helix III of GH and PRL disrupt the integrity of binding Site 2, resulting in mutants with little or no biological activity (Chen et al.,

151 1991a; Fuh et al., 1993; Chen et al., 1994; Fuh and Wells, 1995; Goffin et al., 1996; Chen

et al., 1999). This Gly to Arg mutation reportedly has minimal effects upon initial

hormone-receptor interactions at Site 1 but disrupts the critical hormone-receptor-

interactions at Site 2 (de Vos et al., 1992; Sundstrom et al., 1996; Clackson et al., 1998),

preventing proper receptor dimerization from occurring. We reasoned that the fusion of

two molecules of G129R or G120R together would create a bivalent ligand with only two potentially functional binding Site 1s located in different moieties. Dimerization and activation of PRLR by these dimers would provide further evidence that the PRLR exhibits considerable flexibility in accommodating ligands of larger size (Fig. 4.13).

Using a radio-receptor binding assay we examined the overall affinity of hPRL derivatives for human PRLRs expressed on the surface of T-47D cells. Since ligand binding occurs sequentially and because ligand binding is transient in nature, the radio-

receptor binding assay reflects 1:1 ligand-receptor interactions to a larger extent than 1:2 interactions. We observed that higher concentrations of G129R were needed to compete with 125I-labeled hPRL than were necessary for hPRL (Fig. 4.2) (Chen et al., 1999). This lower affinity may reflect the inability of G129R to interact with a second receptor due to the disruption of binding Site 2. Homodimers of G129R were found to have an affinity greater than that of monomeric G129R (Fig. 4.2), indicating that the addition of a second

G129R moiety was able to improve the interaction of the ligand with a second receptor.

Homodimeric hPRL was found to have an affinity for PRLRs similar to that of monomeric hPRL, suggesting that the fusion of two hPRL molecules together had minimal effect on 1:1 ligand-receptor interactions.

152 To determine if dimeric G129R truly has a second functional binding site, we

examined whether or not it could induce the dimerization of human PRLRs. Ligand-

induced dimerization causes conformational changes that bring the cytoplasmic domains

of the two PRLRs into close proximity to each other. To examine the conformational

changes induced by dimerization we fused Renilla luciferase (Rluc) or a mutant of green

fluorescent protein (GFP2) to the C-terminus of the long (LF) and short (SF1a and SF1b)

forms of the human PRLR and cotransfected the fusions of each isoform into HEK 293

cells. Upon addition of DeepBlueC™ (DBC), Rluc catalytically degrades the substrate and releases bioluminescence resonance energy (peak ~395 nm). If receptor dimerization occurs, Rluc is brought into close proximity to GFP2 and the bioluminescence resonance

energy is transferred to GFP2 which emits fluorescence at a higher wavelength (peak

~515 nm) referred to as a bioluminescence resonance energy transfer (BRET) signal.

Previously, it has been demonstrated that hPRL induces a BRET signal in this system and

that G129R does not; additionally, it has been shown that G129R can competitively

inhibit hPRL-induced BRET (Tan et al., 2005). In contrast to monomeric G129R, we

demonstrate that the addition of homodimeric G129R to the transfected cells results in

the emission of a BRET signal similar to that of monomeric hPRL (Fig. 4.3A). The

BRET ratios for HEK293 cells cotransfected with LF-Rluc/LF-GFP2, SF1a- Rluc/SF1a-

GFP2, or SF1b-Rluc/SF1b-GFP2 and treated with homodimeric G129R increased

approximately 4, 3, and 3-fold, respectively (Fig. 4.3B); which are comparably less than

that of monomeric hPRL which increased the BRET ratios approximately 9, 4, and 4- fold, respectively (Tan et al., 2005). The lower efficiency of BRET induction by

153 homodimeric G129R in comparison to monomeric hPRL may reflect alterations in the conformation of the receptor homodimers formed. These findings indicate that dimeric

G129R is able to induce conformational changes in the long and short isoforms of the

PRLR which are consistent with receptor dimerization. It also indicates that the second

G129R moiety provides the second functional receptor binding site necessary for receptor dimerization to occur (Fig. 4.13).

To determine whether the receptor dimers induced by dimeric G129R were functional we examined the ability of dimeric G129R to activate signaling molecules.

The conformational changes induced by the formation of a functional PRLR dimer activate receptor-associated JAK2 tyrosine kinase which stimulates the phosphorylation of STAT5. Activation of PRLRs by hPRL or hGH also leads to the activation of the

Ras/Raf/MAPK kinase/Erk and phosphatidylinositol 3-kinase/Akt signaling pathways.

We demonstrate that dimeric G129R and dimeric G120R are able to activate these three different intracellular signaling pathways in a manner qualitatively similar to that of hPRL and hGH (Fig. 4.5). The magnitude of response induced by the dimers was similar to that of monomeric hPRL and hGH (Figs. 4.6 and 4.7); however, activation of signaling by homodimeric G129R was slightly less efficient than hPRL, whereas homodimeric

G120R was significantly less efficient than hGH. These finding suggest that dimeric

G129R acts as a substantial agonist and that G120R acts as a weak agonist at the PRLR.

One explanation for the reduced efficiency of homodimeric G129R and homodimeric

G120R is that the receptor homodimers formed upon ligand binding have an altered conformation that reduces the efficiency of signal transduction. The even lower

154 efficiency of signaling induced by homodimeric G120R may be attributed to the signaling of hGH and its derivatives being mediated primarily via the PRLR and not the

GHR in T-47D cells (data not shown) (Tsunekawa et al., 1999). Studies have shown that

His18 and Glu 174 located within binding Site1 of hGH and Asp187 and His188 of the human PRLR interact with a single zinc cation to enhance the affinity of hGH for the

PRLR approximately 8000-fold (Cunningham et al., 1990b). Presumably, homodimeric hGH would require the coordination of two zinc cations to dimerize PRLRs (one for each binding Site 1) since the two binding Site 2s are not functional. This would suggest that the interactions of homodimeric G120R with PRLRs may be more stringent and less flexible than would be necessary for this ligand to interact with the GHR.

The quantitatively similar signaling activation observed for monomeric hPRL

(Figs. 4.4A, 4.6A, and 4.8A) and heterodimers of hPRL and G129R (Fig. 4.5A, Fig. 4.6,

D and E, and Fig. 4.8, D and E) suggests that fusion of two ligands together has little effect on their ability to interact with the first receptor, a finding in agreement with the binding data. The decreased agonism observed for homodimeric G129R (Figs. 4.5A,

4.6B, and 4.8B) as opposed to heterodimers likely reflects the fact that both moieties are required for activity, requiring some accommodation in the receptors. These data indicate that the receptor dimerization induced by dimeric derivatives of hPRL and hGH are functional and capable of activating the signaling pathways necessary for biological activity.

hPRL-mediated signaling can induce cell differentiation, proliferation, and/or survival (Socolovsky et al., 2001; Shillingford et al., 2002). We examined the bioactivity

155 of the dimers using rat Nb2 cells which proliferate in response to lactogens. The

bioactivity of the dimers on rat Nb2 cells (Fig. 4.10) correlated well with their ability to

transduce signals in human T-47D cells. This suggests that the restoration of the

biological activities of G129R and G120R was most likely due to the addition of a second

functional binding site (Site 1). This conclusion is further supported by our examination

of truncated forms of dimeric G129R in which a large portion of the second Site 1 (helix

IV) was deleted (Fig. 4.11). It is noteworthy that three types of responses were observed

in the Nb2 cell proliferation assay in response to these truncated mutants including: full

or near full agonism (PRL, G129R2 and G129R2∆C25), partial agonism (G129R,

G129R2∆C40, G129R2∆C60, and G129R2∆C180), and minimal agonism (G129R2∆C80,

G129R2∆C100, G129R2∆C120, G129R2∆C142, and G129R2∆C160). With the exception

of G129R2∆C180, truncations into and beyond helix III completely abolished the partial agonism, suggesting that helix III in the second G129R moiety is involved in the partial

agonism observed in G129R2ΔC25, G129R2ΔC40 and G129R2ΔC60. One possibility is

that the role of helix III is just to maintain the integrity of the second G129R moiety.

Loss of helix III may prevent the formation of a helical bundle and the remnants of the

second moiety may take on a structure that sterically hinders the ability of the first

G129R moiety to interact with the PRLR. Radio-receptor binding assays were not

performed for these truncated variants of dimeric G129R because of low production

yields.

In addition to studying homodimeric G129R, we demonstrate that homodimeric

hPRL binds to human PRLRs (Fig. 4.2), activates PRLR associated signaling pathways

156 (Fig. 4.8C), and exhibits bioactivity in vitro (Fig. 4.10B) similar to that of monomeric

hPRL (Figs. 4.2, 4.8A, and 4.10B). Similar results were observed for homodimeric hGH

(Figs. 4.9C and 4.10D) when compared to monomeric hGH (Figs. 4.9A and 4.10D). To

examine if the pharmacokinetic behavior of homodimers differs from that of monomers,

we injected mice with monomeric and homodimeric hPRL and measured the plasma

concentrations over time. We demonstrate that homodimeric hPRL has twice the area

under the serum concentration-time curve as monomeric hPRL (Fig. 4.12), indicating that

the larger size of the ligand enhances its pharmacokinetic behavior. This suggests that

homodimers of ligands in this cytokine family could be more clinically efficacious. The

presence of four potential binding sites, and the necessity for only two sites, suggests that

chemical modification to further improve their pharmacokinetics may be possible.

Chemical modifications such as polyethylene glycol, dextran chains, or fusion to albumin

or albumin binding peptides can affect the size and ability of a ligand to bind to a

receptor; use of homodimers of a ligand for derivatization may decrease the probability

of affecting crucial receptor binding sites. Conversely, it is clear that homodimerization

of antagonists would produce the opposite of what was intended. The findings in this

study also suggest the possibility that the different degree of partial agonism reported by

different laboratories for G129R and other antagonists may be reflective of the number of

dimeric or larger aggregates in the preparation.

Studies of the erythropoietin (EPO) have yielded similar results to our study. An

EPO receptor (EPOR) antagonist (EPO-R103A) was shown to regain its biological activity when homodimerized (Qiu et al., 1998). Similarly, dimeric forms of EPO have

157 been shown be fully active and to have beneficial pharmacokinetic behavior different

from that of monomeric EPO (Sytkowski et al., 1999; Dalle et al., 2001). These findings suggest that this class of hematopoietin receptors exhibits considerable plasticity and that modifications to their cognate ligands may have therapeutic value.

In summary, we demonstrate that the ligand binding sites of the PRLR are capable

of interacting with larger molecules to induce receptor dimerization. We confirm that

larger ligands with at least two functional binding sites can act as agonists and that a

second receptor binding Site 1 can replace the need for a receptor binding Site 2 (Fig.

4.13). The recent implications of autocrine hPRL and hGH in cancers and the

overexpression of their cognate receptors in certain cancers make the PRLR and GHR

targets of interest for the development of therapeutics. The results of this study suggest

the feasibility of designing novel multi-functional therapeutics that target the PRL/GH

receptor family as we have demonstrated using G129R fused with other proteins (G129R-

IL2, G129R-Endostatin, and G129R-PE40-KDEL) (Zhang et al., 2002; Beck et al., 2003;

Langenheim and Chen, 2005).

158 16

14 PRL 12 PRL-PRL 10

8 AUC of Dimer =2.01 AUC of Monomer 6

Concentration (µg/ml) 4

2

0 1234

Time (h)

Fig. 4.12 Pharmacokinetics of monomeric and dimeric forms of hPRL in vivo

Two groups of four BALB/C mice were injected intraperitoneally with 200 µg of monomeric (triangles) or dimeric (squares) hPRL. Blood was collected by tail-vein bleeding at 1 h intervals after injection and the serum was subjected to an immunoradiometric assay for hPRL. The concentrations of monomeric and dimeric hPRL were extrapolated and plotted vs. time and the difference in the area under the plasma concentration curves was calculated. Monomeric and dimeric hPRL standards of known concentration were detected equally well. Endogenous mouse PRL was not detected at all by the IRMA kit.

159 G129R G129R-G129R or or G120R G120R-G120R

2 2 X 1 X 1 2 4 4 1 X 4

Jak2 Jak2 P P P Jak2 Jak2 P

P P P P PRLR/PRLR PRLR/PRLR

Fig. 4.13 Illustration of monomeric and homodimeric G129R and G120R binding to PRLRs

Substitution of a Gly residue with an Arg residue in helix 3 (X) of hPRL or hGH prevents functional receptor dimerization and activation of signal transduction from occuring. Fusion of two antagonists in tandem restores receptor dimerization and signal transduction via the two functional binding Site 1s (composed of helix 1 and 4) each interacting with a receptor.

160 CHAPTER FIVE

DEVELOPMENT OF NOVEL LIGANDS THAT ACTIVATE JAK2/

STAT5 SIGNALING EXCLUSIVELY THROUGH THE

HETERODIMER OF A PROLACTIN RECEPTOR

AND GROWTH HORMONE RECEPTOR

Abstract

Prolactin (PRL), growth hormone (GH), and placental lactogen (PL) have two binding sites, Site 1 and Site 2, which each interact with a receptor to form a trimeric complex that stimulates signal transduction. In ruminants, PRL binds to prolactin receptors (PRLRs), GH to growth hormone receptors (GHRs) and PL to PRLRs or to heterodimerized GHR and PRLR. In humans, GH binds to PRLRs in addition to GHRs and PL has only been reported to bind PRLRs. To examine if a heterodimer of PRLR and GHR is formed in humans, we designed two novel ligands composed of G129R and

B2036 in tandem (G129R-B2036 and B2036-G129R). G129R and B2036 are antagonists of human (h) PRL and GH, respectively, with their binding Site 2 disrupted.

Furthermore, B2036 has its binding Site 1 altered to reduce its affinity towards PRLR.

Therefore, these novel fusion proteins should contain a single PRLR and single GHR binding site. The data presented in this study demonstrates that these ligands do not activate Janus-activated kinase 2/signal transducer and activator of transcription 5

(JAK2/STAT5) signaling in cells that express only GHR or PRLR, but rather behave as receptor antagonists. No biological activity was observed for these ligands in rat Nb2 cells, which predominately express PRLRs. When tested in human cell lines that co-

161 express PRLR and GHR, both B2036-G129R and G129R-B2036 activated JAK2/STAT5 signaling. These findings indicate that a functional heterodimer of PRLR and GHR may be formed in humans and that these novel ligands may be used to modulate heterodimer signaling.

162 Introduction

Prolactin (PRL) (Keeler et al., 2003; Teilum et al., 2005), growth hormone (GH)

(de Vos et al., 1992; Chantalat et al., 1995; Clackson and Wells, 1995) and placental lactogen (PL) (Walsh and Kossiakoff, 2006) are long-chain four α-helical cytokines.

They each have two distinct binding sites, Site 1 and Site 2, that bind to homodimeric type 1 cytokine receptors and stimulate Janus-activated kinase 2/signal transducer and activator of transcription 5 (JAK2/STAT5) signaling (Nicola and Hilton, 1998). PRL and

PL mediate their effects via prolactin receptor (PRLR) homodimers; whereas, GH mediates its effects via growth hormone receptor (GHR) homodimers. Binding to the receptor homodimers occurs sequentially with the higher affinity Site 1 interacting with the first receptor before the lower affinity Site 2 interacts with a second receptor to form a functional trimeric complex.

A recent finding in ruminants challenges the paradigm that that these peptide hormones only bind to receptor homodimers and indicates that ovine and bovine PL may also bind a heterodimer of PRLR and GHR. Like PRL, PL induces the formation of a functional PRLR homodimer in ovines and bovines (Byatt et al., 1994; Sakal et al.,

1997), but in contrast to GH, forms an inactive complex with GHR (Staten et al., 1993;

Herman et al., 2000) that competitively inhibits the action of GH (Herman et al., 1999;

Warren et al., 1999). Since receptor binding Site 2 of PL binds a PRLR, the possibility that PL might heterodimerize a GHR via Site 1 and PRLR via Site 2 was explored.

Using homologous systems, it was demonstrated that ovine (o) PL transiently heterodimerizes a oPRLR and oGHR (Biener et al., 2003; Gertler et al., 2005) and that

163 the cytoplasmic domains of heterodimerized receptors induce STAT5-responsive

transcriptional activity (Biener et al., 2003). The transcriptional response to oPL was

better in cells with both oPRLR and oGHR than in cells with only oPRLR (Gertler and

Djiane, 2002). Consistent with the existence of receptor heterodimers, binding of oPL in

the endometrial gland was displaced completely by oPL and oPRL but only partially by

oGH (Noel et al., 2003). Also, in contrast to bovine (b) PRL, bPL was shown to

stimulate DNA synthesis in the bovine mammary gland (Byatt et al., 1994). These

observations suggest that receptor heterodimers activated by PL may have a

physiological role in mammary, placental, and fetal tissues of ruminants.

There is no direct evidence that a PRLR can heterodimerize with a GHR in

humans. In contrast to ruminants, human (h) PL is more closely related to hGH than

hPRL (Dietz et al., 1992; Gootwine and Yossafi, 1998), sharing 93% and 42% sequence

identity, respectively. Also, the two binding sites of hGH (Cunningham et al., 1991; de

Vos et al., 1992; Clackson and Wells, 1995; Sundstrom et al., 1996) can interact with

PRLRs (Kleinberg and Todd, 1980) in addition to a GHRs (Cunningham et al., 1989;

Cunningham et al., 1991; de Vos et al., 1992). The binding sites of hGH for PRLRs

(Somers et al., 1994) and GHRs (Ultsch et al., 1994) overlap but are not identical

(Cunningham and Wells, 1991; Kossiakoff et al., 1994). Zinc influences the binding of hGH to PRLRs but not to GHRs (Cunningham et al., 1990b) and increases its bioactivity

(Dattani et al., 1993; Fuh et al., 1993) by increasing the affinity of Site 1 for the PRLR

(Somers et al., 1994; Duda and Brooks, 2003). Since hGH can readily homodimerize

PRLRs or GHRs, it may potentially heterodimerize a PRLR via Site 1 and GHR via Site

164 2 or vice versa.

Rather than speculate which ligands might heterodimerize a PRLR and GHR in humans and have to utilize elaborate systems in which the extracellular or intracellular portions of the receptors are modified, we chose to design novel ligands that retain a single functional PRLR binding site and a single functional GHR binding site. To do this we utilized a well characterized PRLR antagonist, hPRL-G129R (G129R); PRLR/GHR antagonist, hGH-G120R (G120R); and pure GHR antagonist, B2036. Previously, we have demonstrated that the fusion of two PRLR antagonist (G129R-G129R) or two

PRLR/GHR antagonist (G120R-G120R) in tandem results in ligands able to activate

STAT5 signaling via PRLRs (Langenheim et al., 2006) and that G120R-G120R also induces STAT5 signaling via GHRs (Yang et al., 2007). Based upon these findings we wanted to determine if G129R and G120R linked in tandem (G129R-G120R and G120R-

G129R), or G129R and B2036 linked in tandem (G129R-B2036 and B2036-G129R) could heterodimerize a PRLR and GHR. To do this we examined the ability of the fusion proteins to induce the tyrosine phosphorylation of STAT5 using cell lines determined to express human GHR, PRLR, or both receptors.

165 Materials and Methods

Cell Lines and Reagents

T-47D human breast cancer cells, IM-9 Epstein-Barr virus-transformed human B- lymphoblastoid cells and human embryonic kidney (HEK) 293 cells transformed by sheared human adenovirus type 5 DNA were obtained from the American Type Culture

Collection (Manassas, VA). PRL-dependent Nb2-11 rat pre-T lymphoma cells (Nb2)

were kindly provided by Dr. Ameae Walker and are now available from Sigma-Aldrich

(St. Louis, MO). All reagents were purchased from Invitrogen (Carlsbad, CA) unless otherwise stated. T-47D cells were maintained in RPMI Medium 1640 supplemented with 10% fetal bovine serum (FBS) and 10 µg/ml gentamicin. For IM-9 cells, the medium was further supplemented with 10 mM HEPES and 1.0 mM sodium pyruvate.

For Nb2 cells, the medium was further supplemented with 10% horse serum (HS) and 0.1 mM 2-mercaptoethanol. HEK293 cells were maintained in Dulbecco's Modified Eagle

Medium (DMEM) containing high glucose, 1 mM sodium pyruvate and 2 mM L- glutamine; the medium was supplemented with 10% FBS and 10 µg/ml gentamicin. The

cell lines were incubated at 37°C with humidity and 5% CO2.

Construction of Prokaryotic Expression Vectors with G129R and G120R or B2036

Linked in Tandem

Previously, a PRLR antagonist (G129R), a PRLR/GHR antagonist (G120R), and

a pure GHR antagonist (B2036) were cloned between the Nde I and Xho I sites of the

expression vector pET22b(+) (Novagen; Madison, WI). These plasmids were used as

templates for PCR to modify and amplify the DNA encoding G129R, G120R or B2036

166 so that the genes could be linked in tandem to create recombinant fusion proteins. The

first moiety of G120R or B2036 was amplified using the oligonucleotide primers 5’-

TCATATGTTCCCAACCATTCCCTTATC-3’ and 5’-TGGATCCGAAGCCACAGCTG

CCCTCC-3’; which incorporate an Nde I site at the N-terminus and replace the C-

terminal stop codon with a BamH I site. The second moiety of G120R or B2036 was

amplified using the primers 5’-TGGATCCTTCCCAACCATTCCCTTATC-3’ and 5’-

TCTCGAGCTAGAAGCCACAGCTGCCC-3’; which incorporate a BamH I site at the

N-terminus and an Xho I site after the stop codon. The primers used to create the first

and second moieties of G129R were reported in Chapter 4. The PCR products were

ligated with pCR2.1 T/A cloning vector and transformed into E. coli TOP10 cells

obtained from Invitrogen. Positive clones were identified by plating the cells on Luria-

Bertani (LB) agar plates containing 100 µg/ml ampicillin and 80 µg/ml 5-bromo-4-

chloro-3-indolyl-β-D-galactopyranoside. The plasmids were isolated using the QIAprep®

Spin Miniprep Kit (Qiagen, Valencia, CA) and the Nde I-G129R-BamH I, Nde I-G120R-

BamH I or Nde I-B2036-BamH I DNA fragments were isolated by restriction digestion and separated by electrophoresis as were the BamH I-G129R-stop-Xho I, BamH I-

G120R-stop-Xho I and BamH I-B2036-stop-Xho I DNA fragments. The DNA fragments were purified from the agarose gel using the QIAquick® Gel Extraction Kit (Qiagen) and

the appropriate combinations were ligated with Nde I and Xho I cut pET22b(+) to create

pET22b-G129R-G120R, pET22b-G129R-G120R, pET22b-G129R-B2036 and pET22b-

B2036-G129R. All oligonucleotides were synthesized by Integrated DNA Technologies

(Coralville, IA), restriction and modifying enzymes were purchased from New England

167 BioLabs (Beverly, MA) and PCR reactions were performed using the Advantage-HF™

Kit (BD Biosciences, Palo Alto, CA).

Production and Purification of the Fusion Proteins

Chemically competent Rosetta™(DE3) pLysS E. coli cells (Novagen) were

transformed with the constructs and propagated overnight at 37°C with agitation in LB

broth containing 100 µg/ml ampicillin. The next morning 40 ml of each of the starter

cultures was used to inoculate one liter of Terrific Broth in baffled flasks and the cultures were grown to an O.D.600 of 0.8 before induction of protein production with 1 mM isopropyl-β-thiogalactoside (Alexis Biochemicals, San Diego, CA). After 3 h of induction the cells were harvested by centrifugation at 5,000 x g for 5 min at 4°C and

resuspended in Solution 1 (50 mM Tris·HCl; 10 mM EDTA; 0.5% Triton X-100; pH

adjusted to 8.0) to which 0.4 mg/ml lysozyme was added. The cells were incubated at

room temperature for 1 h to weaken the cell membranes before the cells were sheared on

ice with five 1-min ultrasonic pulses (30% duty at setting 7) using a 550 Sonic

Dismembrator (Fisher Scientific, Waltham, MA). The insoluble fraction containing the inclusion bodies was recovered by centrifugation at 12,000 x g for 30 min at 4ºC and resuspended in Solution 2 (50 mM Tris·HCl; 10 mM EDTA; 1 M urea; pH adjusted to

8.0). The inclusion bodies were collected by centrifugation at 12,000 x g for 30 min at

4ºC and were solubilized and denatured in Solution 3 (50 mM Tris·HCl; 10 mM EDTA; 8

M urea; 1% v/v 2-mercaptoethanol; final pH after protein addition was adjusted to 11.0).

The proteins were refolded by dialysis against baths of 50 mM Tris·HCl, pH 8.0 containing decreasing amounts of urea (4 M urea, 2 M urea, 0 M urea, 0 M urea). The

168 refolded proteins were filtered through 0.45 µm filters and purified by anion-exchange

chromatography using a 5 ml HiTrap Q Sepharose HP column (Amersham Biosciences,

Piscataway, NJ) on an ÄKTA fast protein liquid chromatography system (Amersham

Biosciences). The concentration of the purified proteins was determined using the

Coomassie Plus™ Protein Assay Reagent with BSA standards (Pierce, Rockford, IL).

The purified proteins were separated by reducing SDS-PAGE along with BenchMark

Protein Ladder (Invitrogen) and stained with SYPRO® Orange (Molecular Probes,

Eugene, OR) to confirm their concentration and purity.

STAT5 Phosphorylation Assays

T-47D cells were resuspended in RPMI Medium 1640 supplemented with 10%

charcoal/dextran treated FBS (CSS; HyClone, Logan, UT) and 10 µg/ml gentamycin and

plated in twelve-well tissue culture plates. The cells were grown overnight to approximately 80% confluency and depleted for 3 h in RPMI Medium 1640 supplemented with 0.5% CSS. L-cells were resuspended in DMEM supplemented with

10% FBS and 10 µg/ml gentamycin and plated in twelve-well tissue culture plates. The cells were grown overnight to approximately 80% confluency and depleted for 3 h in

DMEM supplemented with 0.5% CSS. Cells were treated for the indicated amounts of time with the indicated concentrations of monomeric hPRL, hGH, B2036-REV, G129R,

G120R, or B2036; or with heterodimeric G129R-G120R, G120R-G129R, G129R-B2036, or B2036-G129R. The cells were washed with ice cold phosphate buffered saline containing 0.4 mM sodium vanadate and lysed in 100 µl of Lysis Buffer (20 mM

Tris·HCl, pH 7.4; 1% Igepal CA630; 6 mM deoxycholic acid; 150 mM NaCl; 1 mM

169 EDTA, pH 8.0) containing protease inhibitors (1 µg/ml aprotinin; 1 µg/ml leupeptin; 170

µg/ml PMSF; 180 µg/ml sodium orthovanadate). The lysates were homogenized and

incubated on ice for 20 minutes before clarification by centrifugation at 12,000 x g at 4°C

for 30 min. Approximately 80 µg of the supernatants were used for Western blot analysis

as described below.

IM-9 cells were harvested by centrifugation and resuspended in DMEM

containing 0.5% CSS. The volume was adjusted with medium to yield a final

concentration of 1 x 106 cells per ml. The cells were aliquoted (1 mL) into Eppendorf

tubes and depleted for 1 h before addition of 111 μL of 10X treatments. Cells were

treated for 30 min at 37°C, pelleted, the medium aspirated and resuspend in 100 μL of

Lysis Buffer. Approximately 80 µg of the supernatants were used for Western blot analysis as described below.

HEK293 cells stably transfected with human GHR or PRLR were trypsinized and resuspended in selection medium at a final concentration of 5 x 105 cells per ml. For

each treatment, 10 ml of the cell suspension was seeded into a 100 mm2 tissue culture

dish and incubated overnight at 37°C with 5% CO2 and humidity. The medium was

aspirated and the cells were depleted in 9 ml of DMEM containing 0.5% CSS for 3 h.

Treatments (1 ml) were added directly to the cells and incubated for 15 min. The

medium was aspirated and the cells were lysed in 1 ml of modified RIPA Buffer (50 mM

Tris-HCl, pH 7.4; 150 mM NaCl; 1% IGEPAL CA-6305; 0.25% sodium deoxycholate)

to which protease inhibitors (1 mM EGTA; 1 mM NaF, 1 mM Na3VO4; 1 mM

phenylmethylsulfonyl fluoride; 5 µg/ml aprotinin; 1 µg/ml pepstatin A; 2 µg/ml

170 leupeptin) were added. Cells lysate was collected into eppendorf tubes, centrifuged at

18,000 x g for 30 min at 4°C and the supernatant was subject to immunoprecipitation.

Immunoprecipitation

The concentrations of the clarified cell lysates were determined using the

Coomassie Plus™ Protein Assay Reagent with BSA standards (Pierce). Approximately

500 μg of cell lysate was incubated with 5 μg of anti-STAT5 (sc-835, Santa Cruz

Biotechnology, Santa Cruz, CA ) and 25 μl of protein A-sepharose (GE Healthcare

Sciences, Piscataway, NJ) overnight at 4°C on a slowly rotating wheel.

Immunoprecipitates were washed in 1 ml of washing buffer (50 mM HEPES, pH 7.5; 150 mM NaCl; 5 mM EDTA; 50 mM Tris-HCl, pH 7.5; 0.05% Triton X-100), centrifuged briefly and the liquid was aspirated. Washing was repeated twice before Sample Buffer

(1% SDS; 100 mM DTT; 50 mM Tris, pH 7.5; 0.05% bromophenol blue) was added to the beads and the samples were denatured at 98°C for 5 min. The immunoprecipitated products were then subject to Western blotting as described below.

Western Blot Analysis

Cell lysates or immunoprecipitate were separated on 10% polyacrylamide gels by reducing SDS-PAGE. The proteins were transferred by Western blot to Hybond™

Enhanced Chemiluminescence (ECL) nitrocellulose membranes (Amersham Biosciences) at 16 W for 1 h. The nitrocellulose membranes were blocked for 20 min at room temperature in Tris buffered saline (TBS) containing 0.05% Tween 20 (TBS-T) and 5% milk. The blocking solution was replaced with solution containing a 1:500 dilution of mouse anti-phospo STAT5A/B antibody (Upstate, Lake Placid, NY) and incubated

171 overnight at 4°C with agitation. The blots were washed three times with TBS-T, and

incubated for 2 h at room temperature with agitation in blocking solution containing a

1:2000 dilution of goat anti-mouse-HRP conjugate (Bio-Rad, Hercules, CA). Blots were

washed three times with TBS-T and incubated for 1 min with ECL™ Western Blotting

Detection Reagents (Amersham Biosciences). The blots were exposed to Biomax MR

film (Kodak, Rochester, NY) and developed with a Konica SRX-101A processor (Konica

Minolta Medical Imagining, Wayne, NJ). Blots were stripped and reprobed with a

1:1333 dilution of rabbit anti-STAT5 polyclonal antibody (Santa Cruz Biotechnology)

and a 1:2000 dilution of goat anti-rabbit HRP conjugate (Bio-Rad).

Rat Nb2 Cell Proliferation Assay

Rat Nb2 cells were resuspended to a final concentration of 5.5 x 105 cells per ml

in RPMI 1640 Medium supplemented with 10% low lactogen HS (Cat. 2921154; MP

Biomedicals, Irvine, CA), 10 µg/ml gentamicin and 0.1 mM 2-mercaptoethanol. The

cells were depleted overnight, counted and reseeded in flat bottom 96-well plates at a

density of 20,000 cells per well in 150 µl of medium. Treatments were added to each

well in 50 µl volumes and the cells were incubated for 72 h. The relative number of cells was determined by incubating the cells for 4 h with 40 µl of 3-(4,5-dimethylthiazol-2-yl)-

5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium/phenazine methosulfate solution (CellTiter 96 AQueous non-radioactive cell proliferation kit; Promega, Madison,

WI) per well. The reduction of the substrate by living cells was determined by measuring the absorbance at 490 nm using a GENios Pro microplate reader (Tecan, Durham, NC).

All assays were performed at least three times in quadruplicate.

172 R R 6 R 0 9 3 9 V 2 2 0 1 2 E 1 1 R G G B2 G - - - - - R R 6 6 R R R 6 L 9 0 3 3 9 0 9 3 R 2 H 2 0 0 2 2 2 0 P 1 G 1 2 1 1 1 h G h G B2 B G G G B2

50 kDa 40 kDa 30 kDa 25 kDa 20 kDa

Fig. 5.1 Reducing SDS-PAGE analysis of purified agonists, antagonists, and fusions of the antagonists

One microgram of the agonists (hPRL, hGH, and B2036-REV), their antagonists (G129R, G120R, and B2036) and the fusions of the antagonists (G129R-G120R, G120R- G129R, G129R-B2036, and B2036-G129R) were separated by reducing SDS-PAGE on 12% polyacrylamide gels and stained with SYPRO® Orange to confirm their size and purity.

173 Results

Construction, Expression, and Purification of Antagonist-Based Fusion Proteins

A PRLR antagonist (G129R) was covalently fused to the N-terminus or C-

terminus of a PRLR/GHR antagonist (G120R) or a pure GHR antagonist (B2036)

through gene fusion. Expression of the fusion genes in E. coli resulted in a protein in

which the C-terminus of the first moiety was connected the N-terminus of the second

moiety by a Gly-Ser peptide linker. The fusion proteins were found to accumulate in the insoluble fraction as inclusion bodies. The inclusion bodies were isolated, denatured and the fusion proteins were refolded and purified by anion-exchange chromatography to yield preparations greater than 95% purity (Fig. 5.1).

The Fusion Proteins Do Not Activate STAT5 Signaling Via GHRs

Before examining the ability of the fusion proteins to induce the STAT5 tyrosine

phosphorylation it had to be determined if the cell lines respond to stimulus via GHRs,

PRLRs or both receptors. IM-9 cells did not respond to hPRL but responded equally well

to hGH or an analog of hGH which doesn’t bind to PRLRs, B2036-REV (Fig. 5.2A). We also show that hGH-induced STAT5 activation was completely inhibited by GHR antagonists, G120R and B2036, but not by a PRLR antagonist, G129R (Fig. 5.3A). This indicates that hGH-induced STAT5 signaling in IM-9 cells was mediated entirely via

GHRs and not PRLRs. Since all four of the fusion proteins contain only a single GHR binding site, we suspected that the ligands would not induce signal transduction in IM-9 cells, and if their affinity was high enough, would behave as antagonists in the presence of hGH. We demonstrated that fusions of G129R and G120R or B2036 do not activate

174 STAT5 signaling in IM-9 cells (Fig. 5.2, C and D), and therefore do not induce the

formation of a functional GHR homodimer.

The Fusion Proteins Behave as GHR Antagonists

To confirm that the fusion proteins retain the ability to bind a single GHR, we

examined their ability to competitively inhibit hGH-induced STAT5 signaling. The

fusion proteins competitively inhibited hGH-induced STAT5 signaling to different

degrees (Fig. 5.3, B and C). Ligands in which the GHR binding site was located within

the C-terminal moiety (G129R-G120R, G129R-B2036) were more potent than ligands in

which the GHR binding site was located within the N-terminal moiety (G120R-G129R,

B2036-G129R) at inhibiting hGH-induced signaling (Fig. 5.3, B and C). Compared to

monomers of G120R and B2036, the potency of the G129R-G120R and G129R-B2036

fusion proteins was reduced only two-fold; in contrast, the potency of the G120R-G129R

and B2036-G129R fusion proteins was reduced over sixteen-fold (Fig. 5.3, A versus B

and C). These findings are consistent with our rudimentary 3D-models of the fusion

proteins, which indicate that the receptor binding site within the N-terminal moiety is

sterically hindered; whereas, the receptor binding site within the C-terminal moiety is not.

The Fusion Proteins Are Inactive in the Nb2 Cell Proliferation Assay

Having confirmed that the fusion proteins bind to a GHR but are unable to induce the formation of a functional GHR homodimer, we wanted to examine their behavior in

PRLR system. Rat Nb2 cells proliferate in response to extremely low concentrations of hPRL and hGH, but not readily in response to a PRLR antagonist (G129R), a

PRLR/GHR antagonist (G120R) or a pure GHR antagonist (B2306) (Fig. 5.4A).

175

A GH B2036-REV PRL nM: - 3.3 10 33 3.3 10 33 10 333.3 p-STAT5

STAT5

B G120R B2036 G129R nM: - 3.3 10 33 3.3 10 33 3.3 10 33 p-STAT5

STAT5

C GH (G120R-G129R) (G129R-G120R) nM: -103.3 10 33 100 3.3 1001033 p-STAT5

STAT5

D GH (B2036-G129R) (G129R-B2036) nM: -103.3 10 33 100 3.310 33 100 p-STAT5

STAT5

Fig. 5.2 Fusions of a PRLR antagonist (G129R) and a PRLR/GHR antagonist (G120R) or pure GHR antagonist (B2036) do not activate STAT5 signaling in cells expressing only GHRs

IM-9 cells were prepared as described in the Materials and Methods and serum starved for 1 h prior to assay. Cells were treated for 30 min with increasing doses of the agonists (hGH, B2036-REV, or hPRL) to confirm that the cells respond via GHR and not PRLR; increasing doses of the antagonists (G120R, B2036, or G129R) to confirm that they retain no residual agonistic activity (B); increasing doses of the heterodimers containing G129R and G120R (C) or the heterodimers containing G129R and B2036 (D) to confirm that they do not acquire the ability to activate GHR homodimers. Cell lysates were harvested and 80 μg of lysate from each treatment was separated by reducing SDS- PAGE using 10% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed with antibodies that recognize tyrosine phosphosphorylated STAT5A/B (p-STAT5). Equal loading was confirmed by reprobing the membranes with anti-STAT5 antibody.

176 A GH : GH : GH : G120R B2036 G129R GH 10 20 40 10 20 40 10 20 40 nM: -51: 1: 1: 1: 1: 1: 1: 1: 1: p-STAT5

STAT5

B GH : GH : (G120R-G129R) (G129R-G120R) GH 10 20 40 80 10 20 40 80 nM: -5 1: 1: 1: 1: 1: 1: 1: 1: p-STAT5

STAT5 C GH : GH : (B2036-G129R) (G129R-B2036) GH 10 20 40 80 10 20 40 80 nM: -5 1: 1: 1: 1: 1: 1: 1: 1: p-STAT5

STAT5

Fig. 5.3 Fusions of a PRLR antagonist (G129R) and a PRLR/GHR antagonist (G120R) or pure GHR antagonist (B2036) inhibit STAT5 signaling in cells expressing only GHRs

IM-9 cells were prepared as described in the Materials and Methods and serum starved for 1 h prior to assay. Cells were treated with 5 nM of hGH in the absence or presence of increasing doses of GHR antagonists or a PRLR antagonist (A), fusion proteins containing G129R and G120R in tandem (B), or G129R and B2036 in tandem (C). Cell lysates were harvested and 80 μg of lysate from each treatment was separated by reducing SDS-PAGE using 10% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed with an antibody that recognizes tyrosine phosphorylated STAT5A/B (p-STAT5). Equal loading was confirmed by reprobing the membranes with anti-STAT5 antibody.

177 When G129R and G120R were linked in tandem the ability to induce proliferation was

restored due to the presence of a PRLR binding site within each moiety of the fusion

proteins (Fig. 5.4B). However, when the PRLR/GHR antagonist (G120R) moiety was

replaced with a pure GHR antagonist (B2036) moiety, proliferation was once again lost

(Fig. 5.4B). This observation is consistent with the G129R-B2036 and B2036-G129R

fusion proteins having only one PRLR binding site and that they are unable to induce the

formation of a functional PRLR homodimer.

The Fusion Proteins Retain the Ability to Interact with PRLR

To confirm that G129R-B2036 and B2036-G129R still retain a functional PRLR

binding moiety we examined their effect on STAT5 phosphorylation using mouse L-

cells. L-cells do not respond to hPRL and respond poorly but equally to hGH and

B2036-REV (Fig. 5.5A), indicating that they endogenously express low levels of mouse

GHR, but not PRLR. Treatment of L-cells with G129R-B2036 or B2036-G129R did not result in STAT5 activation (not shown). However, transfection of the L-cells with human

PRLR enabled STAT5 activation in response to hPRL and enhanced STAT5 activation induced by hGH but not by B2036-REV (Fig. 5.5A). In addition, a small but significant activation of STAT5 was induced by B2036-G129R (Fig. 5.5B); however, it is unclear if

G129R-B2036 was able to activate STAT5 (Fig. 5.5B). Since G129R-B2036 is at best a weak agonist we examined if it could antagonize hPRL-induced signaling via PRLRs.

Consistent with G129R-B2036 being a very weak agonist, at best, it behaved as an antagonist at the concentrations examined (Fig. 5.5C); whereas, B2036-G129R, which has weak agonism (Fig. 5.5B), was a poor antagonist (Fig. 5.5C).

178 A 2 hGH hPRL 1.5 B2036-REV

1

O.D. 490 G120R 0.5 G129R B2036 0 0.001 0.01 0.1 1 10 [Ligand] nM B 2 (G129R-G120R) 1.5 (G120R-G129R)

1 O.D. 490 (G129R-B2036) 0.5 (B2036-G129R)

0 0.001 0.01 0.1 1 10 [Ligand] nM

Fig. 5.4 Proliferation of rat Nb2 cells in response to monomers or fusions of G129R and G120R or B2036 in comparison to hGH, hPRL, and B2036-REV

Rat Nb2 cells were treated for 72 h with various doses of the ligands indicated above as described in Materials and Methods. Cell proliferation was assessed by colorimetrically measuring the reduction of MTS. Each data point represents the mean ± SD of at least three independent experiments performed in quadruplicate.

179 These findings are consistent with our model of activation. The affinity of the

receptor binding sites within the N-terminal moiety and the C-terminal moiety of the fusion proteins is not equal due to steric hindrance caused by the linkage of the two ligands together. The potency of the fusion proteins at activating STAT5 appears to be determined by the receptor binding site in its C-terminal moiety and the abundance of its cognate receptor. In the L-cells stably transfected with the PRLR the levels of PRLR were higher than GHR; therefore, the fusion protein with its PRLR binding moiety on the

C-terminus, B2036-G129R, was more potent at activating STAT5 than G129R-B2036.

The Fusion Proteins Induce STAT5 Signaling in Human Cells Co-Expressing PRLR and GHR

Having confirmed that G129R-B2036 and B2036-G129R retain the ability to bind a GHR and PRLR but are unable to induce the formation of functional PRLR or GHR homodimers, we wanted to examine if they behave as agonists in human cells that naturally express both the GHR and PRLR. Treatment of T-47D cells with various concentrations of hPRL or hGH readily resulted in the phosphorylation of STAT5 (Fig.

5.5A). In contrast to hGH, a pure GHR agonist (B2036-REV) only slightly induced the activation of STAT5 (Fig. 5.5A), suggesting that hGH mediates its effects primarily via

PRLRs in T-47D cells. The effects of hPRL could be completely blocked by a PRLR antagonist (G129R) and a PRLR/GHR antagonist (G120R) but were not affected by a pure GHR antagonist (B2036) (Fig. 5.5C). The effects of a pure GHR agonist (B2036-

REV) could be completely blocked by a pure GHR antagonist (B2036) and a PRLR/GHR antagonist (G120R) but were not affected by a PRLR antagonist (G129R) (Fig. 5.6D).

180 A CCL1.4 CCL1.4 PRLR - PRL GH B2036-REV - PRL GH B2036-REV p-STAT5

STAT5

B PRL (B2036-G129R) (G129R-B2036) nM: 01005505000550500 p-STAT5

STAT5

C PRL : PRL : (B2036-G129R) (G129R-B2036) PRL 0 0 0 0 10 50 10 15 10 50 10 15 -5 nM 1: 1: 1: 1: 1: 1: 1: 1: p-STAT5

STAT5

Fig. 5.5 Fusions of B2036 and G129R retain the ability to bind PRLR as indicated by the activation of or competitive inhibition of STAT5 signaling in L-cells transfected with human PRLR

L-cells or L-cells transfected with human PRLR cells were plated as described in Materials and Methods and incubated overnight. The cells were depleted for 3 h before treatment for 30 min with 20 nM of the indicated ligands (A). L-cells transfected with the PRLR were treated with increasing concentrations of the B2036 and G129R fusion proteins in the absence of hPRL (B) or presence of hPRL (C). Cell lysates were harvested and 80 μg of lysate from each treatment was separated by reducing SDS-PAGE using 10% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed with an antibody that recognizes tyrosine phosphosphorylated STAT5A/B (p-STAT5). Equal loading was confirmed by reprobing the membranes with anti-STAT5 antibody.

181 These findings clearly indicate that T-47D cells express both PRLR and GHR.

In T-47D cells, monomers of the receptor antagonists (G129R, G120R and

B2036) did not induce STAT5 activation (Fig. 5.6B). When G129R and G120R were linked in tandem, regardless of orientation, they equally and robustly induced the activation of STAT5 (Fig. 5.7A). When G129R and B2036 were linked in tandem, they differentially induced the activation of STAT5 (Fig. 5.7B). G120R-G129R and G129R-

G120R had similar potency to hPRL since both moieties can bind to PRLRs (Fig. 5.7A).

B2036-G129R, which has a single PRLR binding site in the C-terminal moiety, had potency similar to that of hPRL (Fig. 5.7B); likewise, G129R-B2036 which has its GHR binding site in the C-terminal moiety had potency similar to that of a pure GHR agonist,

B2036-REV (Fig. 5.7B). These results suggest that the potency of each ligand corresponds with the receptor binding site located in the C-terminal moiety of the fusion protein and reflects the abundance of its cognate receptor. Since B2036-G129R has its

PRLR binding moiety on the C-terminus and PRLRs are more prevalent in T-47D cells, it was more potent at activating STAT5 signaling than G129R-B2036.

Since G129R-B2036 has its GHR binding site in the C-terminal moiety, it would be expected to be more potent at activating STAT5 when GHRs were more prevalent than PRLRs. To create this scenario we transfected HEK293 cells with either PRLR or

GHR and initially examined there response to hPRL and B2036-REV (Fig. 5.8, A and B).

HEK293 cells endogenously express low levels of PRLR and even lower levels of GHR.

182 A PRL GH B2036-REV nM: - 550500 - 550500 - 550500 p-STAT5

STAT5

B G129R G120R B2036 nM:- 5 50 500- 550 500 - 550500 p-STAT5

STAT5

C PRL: PRL: PRL: G129R G120R B2036 0 PRL 0 0 0 0 0 0 0 0 0 2 4 6 2 4 6 6 8 :1 nM: -5 1: 1: 1: 1: 1: 1: 1: 1: 1 p-STAT5

STAT5

D B2036-REV: B2036-REV: B2036-REV: G129R G120R B2036 B2036-REV 30 40 20 30 40 20 30 40 nM: -5 1: 1: 1: 1: 1: 1: 1: 1: p-STAT5 STAT5

Fig. 5.6 Western blot analysis of the phosphorylation status of STAT5 in T-47D cells in response to PRLR and GHR agonists and antagonists

T-47D cells were plated as described in Materials and Methods and incubated overnight. The cells were depleted for 3 h before treatment for 30 min with: (A) increasing doses of agonists (hPRL, hGH, B2036-REV); (B) increasing doses of antagonists (G129R, G120R, B2036-K120G); or 5 nM of (C) hPRL or (D) B2036-REV in the absence or presence of increasing doses of a PRLR antagonist (G129R), a PRLR/GHR antagonist (G120R), or a pure GHR antagonist (B2036). Cell lysates were harvested and 80 μg of lysate from each treatment was separated by reducing SDS-PAGE using 10% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed with antibodies that recognize tyrosine phosphosphorylated STAT5A/B (p-STAT5). Equal loading was confirmed by reprobing the membranes with anti-STAT5 antibody.

183 A GH PRL (G120R-G129R) (G129R-G120R) nM: -55 5 0 5 050005 50 500 p-STAT5 STAT5

B B2036-REV PRL (B2036-G129R) (G129R-B2036) nM: - 55 0 5 50 500 0 5 50 500 p-STAT5 STAT5

Fig. 5.7 Western blot analysis of the phosphorylation status of STAT5 in T-47D cells in response to PRLR and GHR agonists and antagonist fusions

T-47D cells were plated as described in Materials and Methods and incubated overnight. The cells were depleted for 3 h before treatment for 30 min with increasing doses of fusions of G129R and G120R (A) or fusions of G129R and B2036 (B). Cell lysates were harvested and 80 μg of lysate from each treatment was separated by reducing SDS-PAGE using 10% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were probed with an antibody that recognizes tyrosine phosphosphorylated STAT5A/B (p-STAT5). Equal loading was confirmed by reprobing the membranes with anti-STAT5 antibody.

184 In the HEK293 PRLR cell line, STAT5 was activated more readily by hPRL than B2036-

REV (Fig. 5.8A) as observed in T-47D cells (Fig. 5.6A). In the HEK293 GHR cell line,

STAT5 was activated more robustly by B2036-REV than hPRL (Fig. 5.8B). We next examined their responses to the B2036 and G129R fusion proteins (Fig. 5.8, C and D).

STAT5 was activated more robustly in the PRLR cell line by B2036-G129R than

G129R-B2036 (Fig. 5.8C) and was similar to the response seen in T-47D cells (Fig.

5.7B). In contrast to T-47D cells (Fig. 5.7B) and HEK293 PRLR cells (Fig. 5.8C),

STAT5 was activated as readily by G129R-B2036 as B2036-G129R in HEK293 GHR cells (Fig. 5.8D). By increasing the level of GHR, the potency of G129R-B2036 at activating STAT5 was increased.

185 A hPRL B2036-REV nM:- 5 50 500 - 550500 p-STAT5 STAT5 HEK293 PRLR

B hPRL B2036-REV nM:- 5 50 500 - 550500 p-STAT5

STAT5 HEK293 GHR

C (B2036-G129R) (G129R-B2036) nM: - 550500 - 550500 p-STAT5

STAT5 HEK293 PRLR

D (B2036-G129R) (G129R-B2036) nM: - 550500 - 550500 p-STAT5

STAT5 HEK293 GHR

Fig. 5.8 Western blot analysis of the phosphorylation status of STAT5 in transfected HEK293 cells in response to hPRL, B2036-REV, and fusions of a PRLR antagonist (G129R) and GHR antagonist (B2036)

HEK293 cells transfected with PRLR or GHR were plated as described in Materials and Methods and incubated overnight. The cells were depleted for 3 h before treatment for 15 min with increasing doses of a PRLR agonist (hPRL) or GHR agonist (B2036-REV (A and B); or fusions of B2036 and G129R (C and D). For each treatment the cell lysate was harvested and STAT5 was immunoprecipitated from 500 μg of lysate. Immunoprecipitate was separated by reducing SDS-PAGE using 10% polyacrylamide, transferred to nitrocellulose membranes, and probed with an antibody that recognizes tyrosine phosphosphorylated STAT5A/B (p-STAT5). Equal loading was confirmed by reprobing the membranes with anti-STAT5 antibody.

186 Discussion

hPRL (Keeler et al., 2003; Teilum et al., 2005), hGH (de Vos et al., 1992;

Chantalat et al., 1995; Clackson and Wells, 1995) and hPL (Walsh and Kossiakoff, 2006) are structurally similar to one another. They have 190 to 199 amino acids; a folded conformation of a four α-helix bundle; two disulfide bridges at homologous positions; and two asymmetric receptor binding sites that provide receptor specificity. The two

binding sites of hPRL, hGH, and hPL can each interact with a PRLR to form a one ligand

two receptor complex; in addition, hGH can also interact with GHRs to form a one ligand two receptor complex.

The PRLR and GHR are glycoproteins which belong to the type 1 cytokine

receptor family and are characterized as forming homodimers and as having no intrinsic

kinase activity. The long form of PRLR and GHR are similar in size and are composed

of 598 and 620 amino acids, respectively. Topologically, each receptor has an N-

terminal extracellular domain that binds ligand, or in the absence of ligand, prevents the

formation of active receptor homodimers; a single-pass transmembrane/juxtamembrane

domain involved in membrane anchorage and dimer formation; and a C-terminal

intracellular domain involved in signal transduction.

Although the PRLR and GHR exhibit little overall sequence identity, they have highly conserved sequences in both their extracellular and intracellular domains. The tertiary structure of the extracellular domain of the PRLR (Ultsch and de Vos, 1993;

Kossiakoff et al., 1994; Somers et al., 1994) and GHR (de Vos et al., 1992; Kossiakoff et al., 1994; Sundstrom et al., 1996) contain two fibronectin-like type III subdomains (S1

187 and S2) (Thoreau et al., 1991; de Vos et al., 1992) which are connected via a hinge region

(de Vos et al., 1992). A portion of the N-terminal S1 domain, which is stabilized by four

positionally conserved Cys residues, and the hinge region are involved in ligand binding; whereas, the membrane proximal S2 domain, which contains a Trp-Ser-X-Trp-Ser

(WSXWS) motif near its C-terminus, is thought to stabilize the ligand recognition site

(Somers et al., 1994). Much less is known about the structure of their cytoplasmic

domain. At the membrane proximal region of these receptors there is a conserved Box 1

motif which is the binding site for the intracellular non-receptor tyrosine kinase, JAK2,

which is proceeded by a Box 2 motif.

Evidence suggests that the PRLR (Gadd and Clevenger, 2006) and GHR (Gent et

al., 2002; Brown et al., 2005; Waters et al., 2006) exist as preformed homodimers and

that ligand binding occurs sequentially with the higher affinity binding Site 1 interacting

with the first receptor before the lower affinity binding Site 2 interacts with the second

receptor. Due the asymmetric nature of the receptor binding sites, minor alterations in

the conformation of the second receptor are necessary (Mellado et al., 1997; Clackson et

al., 1998; Kossiakoff, 2004). These ligand-induced alterations lead to the optimal

alignment and binding of the extracellular receptor dimerization domain residues

(Behncken et al., 2000) and lead to a rotation in the transmembrane domain that bring

cytoplasmic receptor-associated JAK2 molecules into close proximity, allowing their

kinase domains to transphosphorylate each other. Activated JAK2 catalyzes the

phosphorylation of the receptors and receptor-associated proteins and can lead to: STAT

1, 3, and 5 signaling; phosphoinositide 3-kinase/3-phosphoinositide-dependent protein

188 kinase 1/Akt signaling; Ras/Raf/mitogen-activated protein kinase (MAPK) kinase/MAPK signaling by interacting with the SHC/Grb2/SOS complex; and protein kinase C/MAPK signaling by activating phospholipase C-gamma.

In ruminants it has been demonstrated that PL can interact with a GHR via binding Site 1 and a PRLR via binding Site 2 leading to the stimulation of JAK2/STAT5 signaling (Biener et al., 2003; Gertler et al., 2005). There is no evidence that a PRLR can heterodimerize with a GHR in humans; however, the structural similarity between the

PRLR and GHR and the ability of hGH to homodimerize PRLRs or GHRs raises some interesting questions. Can a functional PRLR and GHR heterodimer be formed in humans? If so, what ligand(s) would induce the formation of a functional heterodimer?

In what tissue would this heterodimer occur? And would the receptor heterodimers have any physiological or pathological relevance in humans?

In this study we provide evidence that a PRLR and GHR can heterodimerize and transduce STAT5 signaling in human cells. To do this we designed novel fusion proteins using well characterized receptor antagonists (G129R, G120R and B2036) which retain a single functional PRLR and/or GHR binding Site 1 (Fig. 5.9). Previously, we demonstrated that fusions of two PRLR antagonists together (G129R-G129R) or

PRLR/GHR antagonists together (G120R-G120R) were able to activate STAT5 signaling via PRLRs (Langenheim et al., 2006) and that G120R-G120R also induced STAT5 signaling via GHRs (Yang et al., 2007). In this study we demonstrate that fusions of

G129R and G120R in either orientation (G129R-G120R and G120R-G129R) are able to activate STAT5 signaling (Fig. 5.7A) and induce cell proliferation (Fig. 5.4B).

189

A G129R B2036

2 2 X X 1 1 4 4

Jak2 Jak2 Jak2 Jak2

PRLR/PRLR GHR/GHR B G129R-B2036 G129R-B2036 G129R-B2036 or or or B2036-G129R B2036-G129R B2036-G129R

2 2 2 2 2 X 2 X 1 X 1 X 1 X 1 X 4 4 4 4 1 1 4 4

P P Jak2 Jak2 Jak2 Jak2 P Jak2Jak2 P

P P P P PRLR/PRLR GHR/GHR PRLR/GHR

Fig. 5.9 Illustration of the binding of G129R, B2036, or fusions of the two in tandem to homodimers or heterodimers of PRLR and GHR

(A) G129R can only bind to one PRLR and B2036 to one GHR due to the substitution of a Gly residue in helix 3 (X), preventing functional receptor dimerization and activation of signal transduction from occuring. Fusion of the two antagonists in tandem (G129R-B2036 or B2036-G129R) results in a ligand with one PRLR and one GHR binding site. (B) In a pure PRLR or GHR signaling system, G129R-B2036 and B2036-G129R behave as antagonists because they are unable to find a second receptor to partner with; in a heterologous signaling system, they can dimerize a PRLR and GHR via the two functional binding Site 1s (composed of helix 1 and 4) leading to signal transduction.

190 Table 5.1 Effect of hGH and hPRL analogs on receptor dimers

PRLR GHR PRLR/GHR GH and PRL Analogs Homodimer Homodimer Heterodimer Antagonists hGH-G120R (G120R) xxx xxx xxx B2036 (Somavert) - xxx xxx hPRL-G129R (G129R) xxx - xxx Agonists hGH +++ +++ nd hPRL +++ - nd B2036-REV(1) - +++ nd hGH-hGH(2,3) +++ +++ nd hPRL-hPRL(2) +++ - nd G129R-G129R(2) ++ - nd G120R-G120R(2,3) + ++ nd G129R-G120R(1) ++ xx nd G120R-G129R(1) ++ x nd B2036-G129R(1) x x ++ G129R-B2036(1) xx xx + + Behaves as receptor agonist. x Behaves as receptor antagonist. - Does not bind receptor. nd Behavior towards receptor not determined. + Unique profile of receptor agonistic activity. (1) Characterized in this study. (2) Characterized by Langenheim et al., 2006. (3) Characterized by Yang et al., 2007.

191 However, since G120R is able to bind a PRLR or GHR, it was impossible to differentiate

if G129R-G120R and G120R-G129R mediated their effects via PRLR homodimers or a

PRLR and GHR heterodimer.

To overcome this problem we used a pure GHR antagonist, B2036, which

combines an amino acid substitution at position 120 of hGH (Fuh et al., 1992) with an

additional eight amino acid substitutions to reduce the affinity of binding Site 1 towards

the PRLR (Goffin et al., 1999) without negatively affecting its affinity towards the GHR

(Maamra et al., 1999). Thus, fusion of B2036 and G129R resulted in two novel ligands

(G129R-B2036 and B2036-G129R) which retain a single functional PRLR binding site

and a single functional GHR binding site. These ligands provided the molecular tools

necessary to address the simple question: can a functional PRLR and GHR heterodimer

be formed in human tissues?

By examining the response to receptor agonists (hPRL, hGH, and B2036-REV) in

the absence or presence of receptor antagonists (G129R, G120R, B2036) we were able to

determine that IM-9 cells can be stimulated via GHRs (Fig. 5.2A and 5.3A), Nb2 cells via PRLRs (Fig. 5.4A), and T-47D cells via PRLRs or GHRs (Fig. 5.6, A, C and D). As monomers, the receptor antagonists did not exhibit any significant agonistic effects in

IM-9 cells (Fig. 5.2B), Nb2 cells (Fig. 5.4A) or T-47D cells (Fig. 5.6B). Linkage of a

PRLR antagonist (G129R) to a PRLR/GHR antagonist (G120R) resulted in a ligand capable of inducing proliferation in a PRLR system (Fig. 5.4B) but unable to induce

STAT5 signaling in a GHR system (Fig. 5.2C). Linkage of a PRLR antagonist (G129R) to a pure GHR antagonist (B2036) resulted in ligands (G129R-B2036 and B2036-

192 G129R) which no longer were capable of inducing proliferation in a PRLR system (Fig.

5.4B) and did not activate STAT5 in a GHR system (Fig. 5.2D), suggesting that these ligands are inactive in pure receptor systems.

Although the fusions of G129R and B2036 could not activate receptor homodimers, the ligands retained the ability to bind to GHRs on IM-9 cells (Fig. 5.3C) or

PRLRs on transfected L-cells (Fig. 5.5C) as indicated by their ability to competitively inhibit GHR and PRLR mediated signal transduction, respectively. More importantly, the fusions of G129R and B2036 behaved as agonists when examined in T-47D cells

(Fig. 5.7B) and transfected HEK293 cells expressing both PRLR and GHR (Fig. 5.8, C and D). Since these ligands behaved as antagonists in cells predominantly expressing either GHR or PRLR, it is concluded that these novel ligands elicit STAT5 activation via the formation of a functional PRLR and GHR heterodimer.

When the novel fusion proteins were designed it was assumed that the affinities of each moiety would differ depending upon their location in the fusion protein. Based upon rudimentary 3D models of the fusion proteins, it was predicted that the moiety on the C-terminus would have higher affinity than the moiety on the N-terminus due to steric hindrance, and that binding might occur sequentially with the higher affinity C-terminal moiety interacting with its cognate receptor before the lower affinity N-terminal moiety interacted with its cognate receptor. G129R-B2036 was expected to sequentially bind to a GHR followed by a PRLR and B2036-G129R was expected to sequentially bind to a

PRLR followed by a GHR. In agreement with the two receptor binding sites having different affinities, a bell-shaped pattern for STAT5 activity in response to increasing

193 doses of the B2036-G129R and G129R-B2036 was observed in the transfected HEK293

cells (Fig. 5.8, C and D). It was also observed that B2036-G129R was a better agonist in cell lines that expressed higher levels of PRLR than GHR (Fig. 5.7B and 5.8C). We also expected that the agonistic activity of G129R-B2036 could be improved by elevating the level of GHR (Fig. 5.8C versus 5.8D). It is interesting to point out that there was remarkable improvement in terms of the potency of G129R-B2036 in the STAT5 assay; however, a better activation by G129R-B2036 than B2036-G129R in GHR transfected cells was never achieved (Fig. 5.8D), even when GHR levels were in excess of PRLRs.

It was not taken into consideration that receptor heterodimers formed in the

absence of ligand might differ fundamentally from receptor homodimers. In the absence

of ligand, the interactions between the extracellular domains of PRLR homodimers and

GHR homodimers are likely to be symmetrical in nature; whereas, PRLR/GHR

heterodimers are more likely to be asymmetrical. Also, it was not taken into

consideration that the flexibility of the extracellular domains of the PRLR and GHR may

differ. In Chapter 4, we demonstrated that the PRLR in addition to being promiscuous is

quite flexible and can readily accommodate dimeric forms of ligand, such as G129R-

G129R. The asymmetry in preexisting PRLR/GHR heterodimers and the variability in

receptor flexibility may favor the initial binding to one receptor (likely the PRLR) over

that of another, which can explain the potency difference between G129R-B2036 and

B2036-G129R. B2036-G129R would be a better agonist of human PRLR/GHR

heterodimers because its PRLR binding moiety is fully accessible on the C-terminus. On

the other hand, G129R-B2036 would not be a better agonist than B2036-G129R, even

194 when GHR levels were elevated, because it’s PRLR binding moiety on the N-terminus is

sterically hindered, reducing its ability to initially interact with a PRLR.

In summary, we have created novel ligands and demonstrated that they only activate a PRLR and GHR heterodimer (Table 5.1). B2036-G129R was a superior

agonist of PRLR/GHR heterodimers (Figs. 5.5B and 5.7B, and Fig. 5.8, C and D); whereas, G129R-B2036 was a weaker agonist of PRLR/GHR heterodimers (Fig. 5.7B and Fig. 5.8, C and D) and exhibited strong antagonism towards PRLR (Fig. 5.5C) and

GHR homodimers (Fig. 5.3C). Using these ligands we have provided evidence that a functional PRLR and GHR heterodimer either preexists or can be formed in response to ligands in humans, resulting in STAT5 activation. In addition, B2036-REV was created which unlike hGH does not readily activate PRLR homodimers. These novel ligands provide the molecular tools to differentiate signaling events induced via PRLR and GHR homodimers or heterodimers in human tissue.

195 CHAPTER SIX

CONCLUSIONS

In this dissertation, the following conclusions can be drawn from the experiments.

From the experimental results presented in Chapter 2, we conclude that the G129R-

PE40KDEL fusion toxin drastically increased the cytotoxicity of G129R; however, it was

not nearly as potent as we anticipated. Nevertheless, the cytotoxicity was abolished by

anti-PE, could be blocked by hPRL or G129R, and was only observed in human breast

cancer cell lines expressing PRLR. We were unaware of it at the time, but hPL had been

fused to both diphtheria fragment A and the ricin A-beta subunit. The diphtheria based

fusion protein was non-functional and the ricin A based fusion protein only worked at

high concentrations. Since hPRL and hPL are similar in structure and both bind to the

PRLR, we believe that the reduced toxicity is likely a problem with endocytosis.

From the experimental results presented in Chapter 3, we conclude that the fusion

of the serum albumin binding peptide (SA20) to the ligands extended their serum half-life

as expected. Furthermore, fusion of SA20 to the N-terminus of hGH and hPRL was

superior to fusion to their C-terminus. SA20-hGH and SA20-hPRL were superior to

hGH and hPRL at inducing lobulo-alveolar development, respectively. It should also be

noted that these proteins are amendable to slow-releasing formulations since the serum

albumin binding peptide is only twenty amino acids in length.

From the experimental results presented in Chapter 4, we conclude that the

intrinsic activity of PRL-PRL and GH-GH are unaffected and that the plasma half-life of PRL-

PRL was extended. Surprisingly, G129R-G129R and G120R-G120R exhibited agonistic

196 activity. By making a series of trunctations in G129R-G129R we were able to show that binding Site 1 within each moiety was responsible for receptor homodimerization and activation. This indicates that PRL-PRL and GH-GH can activate PRLRs in three different manners (Site1 and Site2 of the 1st moiety; Site 1 and Site 2 of the 2nd moiety; and the Site 1s from each moiety). For this reason we believe that the PRL-PRL and GH-GH molecules will be more amendable to chemical modifcations such as HESylation and PEGylation.

From the experimental results presented in Chapter 5, we conclude that that there are functional PRLR/GHR heterodimers in humans. These results were obtained by extending the findings of Chapter 4 (i.e. that the fusion of two antagonists together results in an agonist). We believe that hGH is capable of binding to receptor heterodimers. Also we were able to make a pure GH agonist which will be quite useful for studying how hGH mediates its effects. We have examined different cell lines that respond to hGH and found that the effects are mediated almost completely through PRLRs in many breast cancer cell lines. This may indicate that hGHs effects are primarily mediated via PRLRs in breast cancer.

197 REFERENCES

(2000a) Ovarian ablation for early breast cancer. Early Breast Cancer Trialists' Collaborative Group. Cochrane Database Syst Rev:CD000485.

(2000b) Consensus guidelines for the diagnosis and treatment of growth hormone (GH) deficiency in childhood and adolescence: summary statement of the GH Research Society. GH Research Society. J Clin Endocrinol Metab 85:3990-3993.

Allured VS, Collier RJ, Carroll SF, McKay DB (1986) Structure of exotoxin A of Pseudomonas aeruginosa at 3.0-Angstrom resolution. Proc Natl Acad Sci U S A 83:1320-1324.

American Cancer Society. (2006) Cancer facts and figures 2006. Atlanta, GA: American Cancer Society.

American Cancer Society. (2007) Cancer facts and figures 2007. Atlanta, GA: American Cancer Society.

Arey BJ, Freeman ME (1989) Hypothalamic factors involved in the endogenous stimulatory rhythm regulating prolactin secretion. Endocrinology 124:878-883.

Arun B, Hortobagyi GN (2002) Progress in breast cancer chemoprevention. Endocr Relat Cancer 9:15-32.

Bahceci M, Tuzcu A, Bahceci S, Tuzcu S (2003) Is hyperprolactinemia associated with insulin resistance in non-obese patients with polycystic ovary syndrome? J Endocrinol Invest 26:655-659.

Baldini E, Giannessi PG, Gardin G, Alama A, Minuto F, Barreca A, Conte PF (1994) In vivo cytokinetic effects of recombinant human growth hormone (rhGH) in patients with advanced breast carcinoma. J Biol Regul Homeost Agents 8:113- 116.

Barrera-Saldana HA, Seeburg PH, Saunders GF (1983) Two structurally different genes produce the same secreted human placental lactogen hormone. J Biol Chem 258:3787-3793.

Bartke A, Cecim M, Tang K, Steger RW, Chandrashekar V, Turyn D (1994) Neuroendocrine and reproductive consequences of overexpression of growth hormone in transgenic mice. Proc Soc Exp Biol Med 206:345-359.

Baselga J, Norton L (2002) Focus on breast cancer. Cancer Cell 1:319-322.

198 Baumann G (1991) Growth hormone heterogeneity: genes, isohormones, variants, and binding proteins. Endocr Rev 12:424-449.

Baumann G, Davila N, Shaw MA, Ray J, Liebhaber SA, Cooke NE (1991) Binding of human growth hormone (GH)-variant (placental GH) to GH-binding proteins in human plasma. J Clin Endocrinol Metab 73:1175-1179.

Beck MT, Peirce SK, Chen WY (2002) Regulation of bcl-2 gene expression in human breast cancer cells by prolactin and its antagonist, hPRL-G129R. Oncogene 21:5047-5055.

Beck MT, Chen NY, Franek KJ, Chen WY (2003) Prolactin antagonist-endostatin fusion protein as a targeted dual-functional therapeutic agent for breast cancer. Cancer Res 63:3598-3604.

Beckman RA, Loeb LA (2005) Genetic instability in cancer: theory and experiment. Semin Cancer Biol 15:423-435.

Behncken SN, Billestrup N, Brown R, Amstrup J, Conway-Campbell B, Waters MJ (2000) Growth hormone (GH)-independent dimerization of GH receptor by a leucine zipper results in constitutive activation. J Biol Chem 275:17000-17007.

Ben-Jonathan N, Mershon JL, Allen DL, Steinmetz RW (1996) Extrapituitary prolactin: distribution, regulation, functions, and clinical aspects. Endocr Rev 17:639-669.

Beral V, Bull D, Doll R, Peto R, Reeves G (2002) Breast cancer and breastfeeding: collaborative reanalysis of individual data from 47 epidemiological studies in 30 countries, including 50302 women with breast cancer and 96973 women without the disease. Lancet 360:187-195.

Bergeron JJ, Resch L, Rachubinski R, Patel BA, Posner BI (1983) Effect of colchicine on internalization of prolactin in female rat liver: an in vivo radioautographic study. J Cell Biol 96:875-886.

Bernstein L, Pike MC, Ross RK, Judd HL, Brown JB, Henderson BE (1985) Estrogen and sex hormone-binding globulin levels in nulliparous and parous women. J Natl Cancer Inst 74:741-745.

Biener E, Martin C, Daniel N, Frank SJ, Centonze VE, Herman B, Djiane J, Gertler A (2003) Ovine placental lactogen-induced heterodimerization of ovine growth hormone and prolactin receptors in living cells is demonstrated by fluorescence resonance energy transfer microscopy and leads to prolonged phosphorylation of signal transducer and activator of transcription (STAT)1 and STAT3. Endocrinology 144:3532-3540.

199 Bissell MJ, Rizki A, Mian IS (2003) Tissue architecture: the ultimate regulator of breast epithelial function. Curr Opin Cell Biol 15:753-762.

Bravo G, Zazueta B, Lavalle C (1992) An acute remission of Reiter's syndrome in male patients treated with bromocriptine. J Rheumatol 19:747-750.

Breier BH, Milsom SR, Blum WF, Schwander J, Gallaher BW, Gluckman PD (1993) Insulin-like growth factors and their binding proteins in plasma and milk after growth hormone-stimulated galactopoiesis in normally lactating women. Acta Endocrinol (Copenh) 129:427-435.

Brelje TC, Scharp DW, Lacy PE, Ogren L, Talamantes F, Robertson M, Friesen HG, Sorenson RL (1993) Effect of homologous placental lactogens, prolactins, and growth hormones on islet B-cell division and insulin secretion in rat, mouse, and human islets: implication for placental lactogen regulation of islet function during pregnancy. Endocrinology 132:879-887.

Brisken C, Socolovsky M, Lodish HF, Weinberg R (2002) The signaling domain of the erythropoietin receptor rescues prolactin receptor-mutant mammary epithelium. Proc Natl Acad Sci U S A 99:14241-14245.

Brodbeck KJ, Pushpala S, McHugh AJ (1999) Sustained release of human growth hormone from PLGA solution depots. Pharm Res 16:1825-1829.

Brostedt P, Luthman M, Wide L, Werner S, Roos P (1990) Characterization of dimeric forms of human pituitary growth hormone by bioassay, radioreceptor assay, and radioimmunoassay. Acta Endocrinol (Copenh) 122:241-248.

Brouwers JR, Assies J, Wiersinga WM, Huizing G, Tytgat GN (1980) Plasma prolactin levels after acute and subchronic oral administration of domperidone and of metoclopramide: a cross-over study in healthy volunteers. Clin Endocrinol (Oxf) 12:435-440.

Brown RJ, Adams JJ, Pelekanos RA, Wan Y, McKinstry WJ, Palethorpe K, Seeber RM, Monks TA, Eidne KA, Parker MW, Waters MJ (2005) Model for growth hormone receptor activation based on subunit rotation within a receptor dimer. Nat Struct Mol Biol 12:814-821.

Brown TE, Fernandes PA, Grant LJ, Hutsul JA, McCoshen JA (2000) Effect of parity on pituitary prolactin response to metoclopramide and domperidone: implications for the enhancement of lactation. J Soc Gynecol Investig 7:65-69.

Byatt JC, Eppard PJ, Veenhuizen JJ, Curran TL, Curran DF, McGrath MF, Collier RJ (1994) Stimulation of mammogenesis and lactogenesis by recombinant bovine placental lactogen in steroid-primed dairy heifers. J Endocrinol 140:33-43.

200 Campbell GS, Argetsinger LS, Ihle JN, Kelly PA, Rillema JA, Carter-Su C (1994) Activation of JAK2 tyrosine kinase by prolactin receptors in Nb2 cells and mouse mammary gland explants. Proc Natl Acad Sci U S A 91:5232-5236.

Carlo-Stella C, Di Nicola M, Milani R, Longoni P, Milanesi M, Bifulco C, Stucchi C, Guidetti A, Cleris L, Formelli F, Garotta G, Gianni AM (2004) Age- and irradiation-associated loss of bone marrow hematopoietic function in mice is reversed by recombinant human growth hormone. Exp Hematol 32:171-178.

Carreno PC, Sacedon R, Jimenez E, Vicente A, Zapata AG (2005) Prolactin affects both survival and differentiation of T-cell progenitors. J Neuroimmunol 160:135-145.

Carter-Su C, Stubbart JR, Wang XY, Stred SE, Argetsinger LS, Shafer JA (1989) Phosphorylation of highly purified growth hormone receptors by a growth hormone receptor-associated tyrosine kinase. J Biol Chem 264:18654-18661.

Cataldo L, Chen NY, Yuan Q, Li W, Ramamoorthy P, Wagner TE, Sticca RP, Chen WY (2000) Inhibition of oncogene STAT3 phosphorylation by a prolactin antagonist, hPRL-G129R, in T-47D human breast cancer cells. Int J Oncol 17:1179-1185.

Cauley JA, Norton L, Lippman ME, Eckert S, Krueger KA, Purdie DW, Farrerons J, Karasik A, Mellstrom D, Ng KW, Stepan JJ, Powles TJ, Morrow M, Costa A, Silfen SL, Walls EL, Schmitt H, Muchmore DB, Jordan VC, Ste-Marie LG (2001) Continued breast cancer risk reduction in postmenopausal women treated with raloxifene: 4-year results from the MORE trial. Multiple outcomes of raloxifene evaluation. Breast Cancer Res Treat 65:125-134.

Cecim M, Bartke A, Yun JS, Wagner TE (1994) Expression of human, but not bovine, growth hormone genes promotes development of mammary tumors in transgenic mice. Transgenics 1:431-437.

Chang WP, Ye Y, Clevenger CV (1998) Stoichiometric structure-function analysis of the prolactin receptor signaling domain by receptor chimeras. Mol Cell Biol 18:896- 905.

Chantalat L, Jones N, Korber F, Navaza J, Pavlovsky AG (1995) The crystal-structure of wild-type growth-hormone at 2.5 angstrom resolution. Protein Pept Lett 2:333- 340

Chaudhary VK, Jinno Y, FitzGerald D, Pastan I (1990) Pseudomonas exotoxin contains a specific sequence at the carboxyl terminus that is required for cytotoxicity. Proc Natl Acad Sci U S A 87:308-312.

201 Chen BJ, Cui X, Sempowski GD, Chao NJ (2003) Growth hormone accelerates immune recovery following allogeneic T-cell-depleted bone marrow transplantation in mice. Exp Hematol 31:953-958.

Chen EY, Liao YC, Smith DH, Barrera-Saldana HA, Gelinas RE, Seeburg PH (1989) The human growth hormone locus: nucleotide sequence, biology, and evolution. Genomics 4:479-497.

Chen NY, Holle L, Li W, Peirce SK, Beck MT, Chen WY (2002) In vivo studies of the anti-tumor effects of a human prolactin antagonist, hPRL-G129R. Int J Oncol 20:813-818.

Chen WY, Wight DC, Wagner TE, Kopchick JJ (1990) Expression of a mutated bovine growth hormone gene suppresses growth of transgenic mice. Proc Natl Acad Sci U S A 87:5061-5065.

Chen WY, White ME, Wagner TE, Kopchick JJ (1991a) Functional antagonism between endogenous mouse growth hormone (GH) and a GH analog results in dwarf transgenic mice. Endocrinology 129:1402-1408.

Chen WY, Wight DC, Mehta BV, Wagner TE, Kopchick JJ (1991b) Glycine 119 of bovine growth hormone is critical for growth-promoting activity. Mol Endocrinol 5:1845-1852.

Chen WY, Chen NY, Yun J, Wagner TE, Kopchick JJ (1994) In vitro and in vivo studies of antagonistic effects of human growth hormone analogs. J Biol Chem 269:15892-15897.

Chen WY, Ramamoorthy P, Chen N, Sticca R, Wagner TE (1999) A human prolactin antagonist, hPRL-G129R, inhibits breast cancer cell proliferation through induction of apoptosis. Clin Cancer Res 5:3583-3593.

Chen WY, Wight DC, Chen NY, Coleman TA, Wagner TE, Kopchick JJ (1991c) Mutations in the third alpha-helix of bovine growth hormone dramatically affect its intracellular distribution in vitro and growth enhancement in transgenic mice. J Biol Chem 266:2252-2258.

Chen WY, Chen NY, Yun J, Wight DC, Wang XZ, Wagner TE, Kopchick JJ (1995) Amino acid residues in the third alpha-helix of growth hormone involved in growth promoting activity. Mol Endocrinol 9:292-302.

Chiron MF, Fryling CM, FitzGerald DJ (1994) Cleavage of pseudomonas exotoxin and diphtheria toxin by a furin-like enzyme prepared from beef liver. J Biol Chem 269:18167-18176.

202 Clackson T, Wells JA (1995) A hot spot of binding energy in a hormone-receptor interface. Science 267:383-386.

Clackson T, Ultsch MH, Wells JA, de Vos AM (1998) Structural and functional analysis of the 1:1 growth hormone:receptor complex reveals the molecular basis for receptor affinity. J Mol Biol 277:1111-1128.

Clark R, Olson K, Fuh G, Marian M, Mortensen D, Teshima G, Chang S, Chu H, Mukku V, Canova-Davis E, Somers T, Cronin M, Winkler M, Wells JA (1996) Long- acting growth hormones produced by conjugation with polyethylene glycol. J Biol Chem 271:21969-21977.

Clevenger CV, Sillman AL, Prystowsky MB (1990a) Interleukin-2 driven nuclear translocation of prolactin in cloned T-lymphocytes. Endocrinology 127:3151- 3159.

Clevenger CV, Altmann SW, Prystowsky MB (1991) Requirement of nuclear prolactin for interleukin-2--stimulated proliferation of T lymphocytes. Science 253:77-79.

Clevenger CV, Russell DH, Appasamy PM, Prystowsky MB (1990b) Regulation of interleukin 2-driven T-lymphocyte proliferation by prolactin. Proc Natl Acad Sci U S A 87:6460-6464.

Clevenger CV, Furth PA, Hankinson SE, Schuler LA (2003) The role of prolactin in mammary carcinoma. Endocr Rev 24:1-27.

Clevenger CV, Chang WP, Ngo W, Pasha TL, Montone KT, Tomaszewski JE (1995) Expression of prolactin and prolactin receptor in human breast carcinoma. Evidence for an autocrine/paracrine loop. Am J Pathol 146:695-705.

Cooke NE, Liebhaber SA (1995) Molecular biology of the growth hormone-prolactin gene system. Vitam Horm 50:385-459.

Cooke NE, Ray J, Watson MA, Estes PA, Kuo BA, Liebhaber SA (1988) Human growth hormone gene and the highly homologous growth hormone variant gene display different splicing patterns. J Clin Invest 82:270-275.

Cox GN, Rosendahl MS, Chlipala EA, Smith DJ, Carlson SJ, Doherty DH (2007) A long-acting, mono-PEGylated human growth hormone analog is a potent stimulator of weight gain and bone growth in hypophysectomized rats. Endocrinology 148:1590-1597.

203 Cui Y, Riedlinger G, Miyoshi K, Tang W, Li C, Deng CX, Robinson GW, Hennighausen L (2004) Inactivation of Stat5 in mouse mammary epithelium during pregnancy reveals distinct functions in cell proliferation, survival, and differentiation. Mol Cell Biol 24:8037-8047.

Cunningham BC, Wells JA (1989) High-resolution epitope mapping of hGH-receptor interactions by alanine-scanning mutagenesis. Science 244:1081-1085.

Cunningham BC, Wells JA (1991) Rational design of receptor-specific variants of human growth hormone. Proc Natl Acad Sci U S A 88:3407-3411.

Cunningham BC, Henner DJ, Wells JA (1990a) Engineering human prolactin to bind to the human growth hormone receptor. Science 247:1461-1465.

Cunningham BC, Jhurani P, Ng P, Wells JA (1989) Receptor and antibody epitopes in human growth hormone identified by homolog-scanning mutagenesis. Science 243:1330-1336.

Cunningham BC, Bass S, Fuh G, Wells JA (1990b) Zinc mediation of the binding of human growth hormone to the human prolactin receptor. Science 250:1709-1712.

Cunningham BC, Ultsch M, De Vos AM, Mulkerrin MG, Clauser KR, Wells JA (1991) Dimerization of the extracellular domain of the human growth hormone receptor by a single hormone molecule. Science 254:821-825.

Dalle B, Henri A, Rouyer-Fessard P, Bettan M, Scherman D, Beuzard Y, Payen E (2001) Dimeric erythropoietin fusion protein with enhanced erythropoietic activity in vitro and in vivo. Blood 97:3776-3782.

Das R, Vonderhaar BK (1996) Involvement of SHC, GRB2, SOS and RAS in prolactin signal transduction in mammary epithelial cells. Oncogene 13:1139-1145.

Das R, Vonderhaar BK (1997) Prolactin as a mitogen in mammary cells. J Mam Gland Biol Neoplasia 2:29-39.

Dattani MT, Hindmarsh PC, Brook CG, Robinson IC, Weir T, Marshall NJ (1993) Enhancement of growth hormone bioactivity by zinc in the eluted stain assay system. Endocrinology 133:2803-2808.

Dattani MT, Hindmarsh PC, Brook CG, Robinson IC, Kopchick JJ, Marshall NJ (1995) G120R, a human growth hormone antagonist, shows zinc-dependent agonist and antagonist activity on Nb2 cells. J Biol Chem 270:9222-9226.

204 de Castillo B, Cawthorn S, Moppett J, Shere M, Norman M (2004) Expression of prolactin receptor mRNA in oestrogen receptor positive breast cancers pre- and post-tamoxifen therapy. Eur J Surg Oncol 30:515-519.

de Vos AM, Ultsch M, Kossiakoff AA (1992) Human growth hormone and extracellular domain of its receptor: crystal structure of the complex. Science 255:306-312.

Decouvelaere C, Peyrat JP, Bonneterre J, Djiane J, Jammes H (1995) Presence of the two growth hormone receptor messenger RNA isoforms in human breast cancer. Cell Growth Differ 6:477-483.

Delvoye P, Demaegd M, Delogne-Desnoeck J (1977) The influence of the frequency of nursing and of previous lactation experience on serum prolactin in lactating mothers. J Biosoc Sci 9:447-451.

Dennis MS, Zhang M, Meng YG, Kadkhodayan M, Kirchhofer D, Combs D, Damico LA (2002) Albumin binding as a general strategy for improving the pharmacokinetics of proteins. J Biol Chem 277:35035-35043.

Di Carlo R, Meli R, Galdiero M, Nuzzo I, Bentivoglio C, Carratelli CR (1993) Prolactin protection against lethal effects of Salmonella typhimurium. Life Sci 53:981-989.

Dickson RB, Lippman ME (1986) Hormonal control of human breast cancer cell lines. Cancer Surv 5:617-624.

Dietz AB, Georges M, Threadgill DW, Womack JE, Schuler LA (1992) Somatic cell mapping, polymorphism, and linkage analysis of bovine prolactin-related proteins and placental lactogen. Genomics 14:137-143.

Djiane J, Dusanter-Fourt I, Katoh M, Kelly PA (1985) Biological activities of binding site specific monoclonal antibodies to prolactin receptors of rabbit mammary gland. J Biol Chem 260:11430-11435.

Duda KM, Brooks CL (2003) Differential effects of zinc on functionally distinct human growth hormone mutations. Protein Eng 16:531-534.

Dugan AL, Thellin O, Buckley DJ, Buckley AR, Ogle CK, Horseman ND (2002) Effects of prolactin deficiency on myelopoiesis and splenic T lymphocyte proliferation in thermally injured mice. Endocrinology 143:4147-4151.

Dugan AL, Malarkey WB, Schwemberger S, Jauch EC, Ogle CK, Horseman ND (2004) Serum levels of prolactin, growth hormone, and cortisol in burn patients: correlations with severity of burn, serum cytokine levels, and fatality. J Burn Care Rehabil 25:306-313.

205 Dunn BK, Ford LG (2000) Breast cancer prevention: results of the National Surgical Adjuvant Breast and Bowel Project (NSABP) breast cancer prevention trial (NSABP P-1: BCPT). Eur J Cancer 36 Suppl 4:S49-50.

Dusanter-Fourt I, Muller O, Ziemiecki A, Mayeux P, Drucker B, Djiane J, Wilks A, Harpur AG, Fischer S, Gisselbrecht S (1994) Identification of JAK protein tyrosine kinases as signaling molecules for prolactin. Functional analysis of prolactin receptor and prolactin-erythropoietin receptor chimera expressed in lymphoid cells. Embo J 13:2583-2591.

Ehrenkranz RA, Ackerman BA (1986) Metoclopramide effect on faltering milk production by mothers of premature infants. Pediatrics 78:614-620.

Elberg G, Kelly PA, Djiane J, Binder L, Gertler A (1990) Mitogenic and binding properties of monoclonal antibodies to the prolactin receptor in Nb2 rat lymphoma cells. Selective enhancement by anti-mouse IgG. J Biol Chem 265:14770-14776.

Emerman JT, Leahy M, Gout PW, Bruchovsky N (1985) Elevated growth hormone levels in sera from breast cancer patients. Horm Metab Res 17:421-424.

Eriksson L, Frankenne F, Eden S, Hennen G, Von Schoultz B (1989) Growth hormone 24-h serum profiles during pregnancy--lack of pulsatility for the secretion of the placental variant. Br J Obstet Gynaecol 96:949-953.

Fabian CJ, Kimler BF (2001) Chemoprevention for high-risk women: tamoxifen and beyond. Breast J 7:311-320.

Fahie-Wilson MN, John R, Ellis AR (2005) Macroprolactin; high molecular mass forms of circulating prolactin. Ann Clin Biochem 42:175-192.

Fields K, Kulig E, Lloyd RV (1993) Detection of prolactin messenger RNA in mammary and other normal and neoplastic tissues by polymerase chain reaction. Lab Invest 68:354-360.

Figueroa FE, Carrion F, Martinez ME, Rivero S, Mamani I (1997) Bromocriptine induces immunological changes related to disease parameters in rheumatoid arthritis. Br J Rheumatol 36:1022-1023.

Fisher B, Costantino JP, Wickerham DL, Cecchini RS, Cronin WM, Robidoux A, Bevers TB, Kavanah MT, Atkins JN, Margolese RG, Runowicz CD, James JM, Ford LG, Wolmark N (2005) Tamoxifen for the prevention of breast cancer: current status of the National Surgical Adjuvant Breast and Bowel Project P-1 study. J Natl Cancer Inst 97:1652-1662.

206 Forsyth IA, Wallis M (2002) Growth hormone and prolactin--molecular and functional evolution. J Mammary Gland Biol Neoplasia 7:291-312.

Frankenne F, Closset J, Gomez F, Scippo ML, Smal J, Hennen G (1988) The physiology of growth hormones (GHs) in pregnant women and partial characterization of the placental GH variant. J Clin Endocrinol Metab 66:1171-1180.

Frasor J, Park K, Byers M, Telleria C, Kitamura T, Yu-Lee LY, Djiane J, Park-Sarge OK, Gibori G (2001) Differential roles for signal transducers and activators of transcription 5a and 5b in PRL stimulation of ERalpha and ERbeta transcription. Mol Endocrinol 15:2172-2181.

Freeman ME, Kanyicska B, Lerant A, Nagy G (2000) Prolactin: structure, function, and regulation of secretion. Physiol Rev 80:1523-1631.

Fryling C, Ogata M, FitzGerald D (1992) Characterization of a cellular protease that cleaves Pseudomonas exotoxin. Infect Immun 60:497-502.

Fuh G, Wells JA (1995) Prolactin receptor antagonists that inhibit the growth of breast cancer cell lines. J Biol Chem 270:13133-13137.

Fuh G, Colosi P, Wood WI, Wells JA (1993) Mechanism-based design of prolactin receptor antagonists. J Biol Chem 268:5376-5381.

Fuh G, Cunningham BC, Fukunaga R, Nagata S, Goeddel DV, Wells JA (1992) Rational design of potent antagonists to the human growth hormone receptor. Science 256:1677-1680.

Gadd SL, Clevenger CV (2006) Ligand-independent dimerization of the human prolactin receptor isoforms: functional implications. Mol Endocrinol 20:2734-2746.

Gallego MI, Binart N, Robinson GW, Okagaki R, Coschigano KT, Perry J, Kopchick JJ, Oka T, Kelly PA, Hennighausen L (2001) Prolactin, growth hormone, and epidermal growth factor activate Stat5 in different compartments of mammary tissue and exert different and overlapping developmental effects. Dev Biol 229:163-175.

Gebre-Medhin M, Kindblom LG, Wennbo H, Tornell J, Meis-Kindblom JM (2001) Growth hormone receptor is expressed in human breast cancer. Am J Pathol 158:1217-1222.

Gellersen B, Kempf R, Telgmann R, DiMattia GE (1994) Nonpituitary human prolactin gene transcription is independent of Pit-1 and differentially controlled in lymphocytes and in endometrial stroma. Mol Endocrinol 8:356-373.

207 Gent J, van Kerkhof P, Roza M, Bu G, Strous GJ (2002) Ligand-independent growth hormone receptor dimerization occurs in the endoplasmic reticulum and is required for ubiquitin system-dependent endocytosis. Proc Natl Acad Sci U S A 99:9858-9863.

Gertler A, Djiane J (2002) Mechanism of ruminant placental lactogen action: molecular and in vivo studies. Mol Genet Metab 75:189-201.

Gertler A, Biener E, Ramanujan KV, Djiane J, Herman B (2005) Fluorescence resonance energy transfer (FRET) microscopy in living cells as a novel tool for the study of cytokine action. J Dairy Res 72 Spec No:14-19.

Gill S, Peston D, Vonderhaar BK, Shousha S (2001) Expression of prolactin receptors in normal, benign, and malignant breast tissue: an immunohistological study. J Clin Pathol 54:956-960.

Ginsburg E, Vonderhaar BK (1995) Prolactin synthesis and secretion by human breast cancer cells. Cancer Res 55:2591-2595.

Ginsburg E, Das R, Vonderhaar BK (1997) Prolactin: An autocrine growth factor in the mammary gland. In: Biological Signalling and the Mammary Gland. (Wilde CJ, Peaker M, Taylor E, eds), pp 47-57. Ayr, Scotland: Hannah Research Institute.

Giss BJ, Walker AM (1985) Mammotroph autoregulation: intracellular fate of internalized prolactin. Mol Cell Endocrinol 42:259-267.

Goffin V, Kelly PA (1997) The prolactin/growth hormone receptor family: structure/function relationships. J Mammary Gland Biol Neoplasia 2:7-17.

Goffin V, Touraine P (2002) Pegvisomant. Pharmacia. Curr Opin Investig Drugs 3:752- 757.

Goffin V, Norman M, Martial JA (1992) Alanine-scanning mutagenesis of human prolactin: importance of the 58-74 region for bioactivity. Mol Endocrinol 6:1381- 1392.

Goffin V, Struman I, Mainfroid V, Kinet S, Martial JA (1994) Evidence for a second receptor binding site on human prolactin. J Biol Chem 269:32598-32606.

Goffin V, Kinet S, Ferrag F, Binart N, Martial JA, Kelly PA (1996) Antagonistic properties of human prolactin analogs that show paradoxical agonistic activity in the Nb2 bioassay. J Biol Chem 271:16573-16579.

208 Goffin V, Bernichtein S, Carriere O, Bennett WF, Kopchick JJ, Kelly PA (1999) The human growth hormone antagonist B2036 does not interact with the prolactin receptor. Endocrinology 140:3853-3856.

Goh EL, Pircher TJ, Lobie PE (1998) Growth hormone promotion of tubulin polymerization stabilizes the microtubule network and protects against colchicine- induced apoptosis. Endocrinology 139:4364-4372.

Goh EL, Pircher TJ, Wood TJ, Norstedt G, Graichen R, Lobie PE (1997) Growth hormone-induced reorganization of the actin cytoskeleton is not required for STAT5 (signal transducer and activator of transcription-5)-mediated transcription. Endocrinology 138:3207-3215.

Gootwine E, Yossafi S (1998) Genetic mapping of the ovine placental lactogen gene to sheep chromosome 20. Anim Genet 29:69.

Graichen R, Liu D, Sun Y, Lee KO, Lobie PE (2002) Autocrine human growth hormone inhibits placental transforming growth factor-beta gene transcription to prevent apoptosis and allow cell cycle progression of human mammary carcinoma cells. J Biol Chem 277:26662-26672.

Grigorian AL, Bustamante JJ, Aguilar RM, Martinez AO, Haro LS (2002) 45 kDa Mercaptoethanol-resistant human growth hormone binds to somatogenic and lactogenic receptors. Program of the 84th Annual Meeting of The Endocrine Society, San Francisco, CA, 2002, p 555 (Abstract P3-273).

Grigorian AL, Bustamante JJ, Hernandez P, Martinez AO, Haro LS (2005) Extraordinarily stable disulfide-linked homodimer of human growth hormone. Protein Sci 14:902-913.

Grigorian AL, Bustamante JJ, Aguilar RM, Muñoz J, Martinez AO, Haro LS (2003) Characterization of 45 kDa mercaptoethanol-resistant human growth hormone by radioimmunoassay and bioassays. Program of the 85th Annual Meeting of The Endocrine Society, Philadelphia, PA, 2003, p 381 (Abstract P2-313).

Gu M, Gordon VM, Fitzgerald DJ, Leppla SH (1996) Furin regulates both the activation of Pseudomonas exotoxin A and the Quantity of the toxin receptor expressed on target cells. Infect Immun 64:524-527.

Gunn AJ, Gunn TR, Rabone DL, Breier BH, Blum WF, Gluckman PD (1996) Growth hormone increases breast milk volumes in mothers of preterm infants. Pediatrics 98:279-282.

209 Gutzman JH, Miller KK, Schuler LA (2004) Endogenous human prolactin and not exogenous human prolactin induces estrogen receptor alpha and prolactin receptor expression and increases estrogen responsiveness in breast cancer cells. J Steroid Biochem Mol Biol 88:69-77.

Hall K, Enberg G, Hellem E, Lundin G, Ottosson-Seeberger A, Sara V, Trygstad O, Ofverholm U (1984) Somatomedin levels in pregnancy: longitudinal study in healthy subjects and patients with growth hormone deficiency. J Clin Endocrinol Metab 59:587-594.

Hanahan D, Weinberg RA (2000) The hallmarks of cancer. Cell 100:57-70.

Handwerger S, Freemark M (1987) Role of placental lactogen and prolactin in human pregnancy. Adv Exp Med Biol 219:399-420.

Hankinson SE, Colditz GA, Hunter DJ, Manson JE, Willett WC, Stampfer MJ, Longcope C, Speizer FE (1995) Reproductive factors and family history of breast cancer in relation to plasma estrogen and prolactin levels in postmenopausal women in the Nurses' Health Study (United States). Cancer Causes Control 6:217-224.

Hankinson SE, Willett WC, Michaud DS, Manson JE, Colditz GA, Longcope C, Rosner B, Speizer FE (1999) Plasma prolactin levels and subsequent risk of breast cancer in postmenopausal women. J Natl Cancer Inst 91:629-634.

Hanley MB, Napolitano LA, McCune JM (2005) Growth hormone-induced stimulation of multilineage human hematopoiesis. Stem Cells 23:1170-1179.

Harding PA, Wang X, Okada S, Chen WY, Wan W, Kopchick JJ (1996) Growth hormone (GH) and a GH antagonist promote GH receptor dimerization and internalization. J Biol Chem 271:6708-6712.

Herman A, Helman D, Livnah O, Gertler A (1999) Ruminant placental lactogens act as antagonists to homologous growth hormone receptors and as agonists to human or rabbit growth hormone receptors. J Biol Chem 274:7631-7639.

Herman A, Bignon C, Daniel N, Grosclaude J, Gertler A, Djiane J (2000) Functional heterodimerization of prolactin and growth hormone receptors by ovine placental lactogen. J Biol Chem 275:6295-6301.

Hessler JL, Kreitman RJ (1997) An early step in Pseudomonas exotoxin action is removal of the terminal lysine residue, which allows binding to the KDEL receptor. Biochemistry 36:14577-14582.

210 Horseman ND, Zhao W, Montecino-Rodriguez E, Tanaka M, Nakashima K, Engle SJ, Smith F, Markoff E, Dorshkind K (1997) Defective mammopoiesis, but normal hematopoiesis, in mice with a targeted disruption of the prolactin gene. Embo J 16:6926-6935.

Hu ZZ, Meng J, Dufau ML (2001) Isolation and characterization of two novel forms of the human prolactin receptor generated by alternative splicing of a newly identified exon 11. J Biol Chem 276:41086-41094.

Hu ZZ, Zhuang L, Meng J, Tsai-Morris CH, Dufau ML (2002) Complex 5' genomic structure of the human prolactin receptor: multiple alternative exons 1 and promoter utilization. Endocrinology 143:2139-2142.

Hughes JP, Friesen HG (1985) The nature and regulation of the receptors for pituitary growth hormone. Annu Rev Physiol 47:469-482.

Hwang J, Fitzgerald DJ, Adhya S, Pastan I (1987) Functional domains of Pseudomonas exotoxin identified by deletion analysis of the gene expressed in E. coli. Cell 48:129-136.

Idziorek T, FitzGerald D, Pastan I (1990) Low pH-induced changes in Pseudomonas exotoxin and its domains: increased binding of Triton X-114. Infect Immun 58:1415-1420.

Ilondo MM, Damholt AB, Cunningham BA, Wells JA, De Meyts P, Shymko RM (1994) Receptor dimerization determines the effects of growth hormone in primary rat adipocytes and cultured human IM-9 lymphocytes. Endocrinology 134:2397- 2403.

Imagawa W, Bandyopadhyay GK, Nandi S (1990) Regulation of mammary epithelial cell growth in mice and rats. Endocr Rev 11:494-523.

Ingram DM, Nottage EM, Roberts AN (1990) Prolactin and breast cancer risk. Med J Aust 153:469-473.

Ishikawa M, Yokoya S, Tachibana K, Hasegawa Y, Yasuda T, Tokuhiro E, Hashimoto Y, Tanaka T (1999) Serum levels of 20-kilodalton human growth hormone (GH) are parallel those of 22-kilodalton human GH in normal and short children. J Clin Endocrinol Metab 84:98-104.

Jackson ME, Simpson JC, Girod A, Pepperkok R, Roberts LM, Lord JM (1999) The KDEL retrieval system is exploited by Pseudomonas exotoxin A, but not by Shiga-like toxin-1, during retrograde transport from the Golgi complex to the endoplasmic reticulum. J Cell Sci 112 ( Pt 4):467-475.

211 Jinno Y, Chaudhary VK, Kondo T, Adhya S, FitzGerald DJ, Pastan I (1988) Mutational analysis of domain I of Pseudomonas exotoxin. Mutations in domain I of Pseudomonas exotoxin which reduce cell binding and animal toxicity. J Biol Chem 263:13203-13207.

Johnson OL, Cleland JL, Lee HJ, Charnis M, Duenas E, Jaworowicz W, Shepard D, Shahzamani A, Jones AJ, Putney SD (1996) A month-long effect from a single injection of microencapsulated human growth hormone. Nat Med 2:795-799.

Josefsberg Z, Posner BI, Patel B, Bergeron JJ (1979) The uptake of prolactin into female rat liver. Concentration of intact hormone in the Golgi apparatus. J Biol Chem 254:209-214.

Kastrup KW, Christiansen JS, Andersen JK, Orskov H (1983) Increased growth rate following transfer to daily sc administration from three weekly im injections of hGH in growth hormone deficient children. Acta Endocrinol (Copenh) 104:148- 152.

Kaulsay KK, Mertani HC, Lee KO, Lobie PE (2000) Autocrine human growth hormone enhancement of human mammary carcinoma cell spreading is Jak2 dependent. Endocrinology 141:1571-1584.

Kaulsay KK, Zhu T, Bennett W, Lee KO, Lobie PE (2001) The effects of autocrine human growth hormone (hGH) on human mammary carcinoma cell behavior are mediated via the hGH receptor. Endocrinology 142:767-777.

Kaulsay KK, Mertani HC, Tornell J, Morel G, Lee KO, Lobie PE (1999) Autocrine stimulation of human mammary carcinoma cell proliferation by human growth hormone. Exp Cell Res 250:35-50.

Kauppila A, Kivinen S, Ylikorkala O (1981) A dose response relation between improved lactation and metoclopramide. Lancet 1:1175-1177.

Kauppila A, Chatelain P, Kirkinen P, Kivinen S, Ruokonen A (1987) Isolated prolactin deficiency in a woman with puerperal alactogenesis. J Clin Endocrinol Metab 64:309-312.

Keeler C, Dannies PS, Hodsdon ME (2003) The tertiary structure and backbone dynamics of human prolactin. J Mol Biol 328:1105-1121.

Keppler-Hafkemeyer A, Brinkmann U, Pastan I (1998) Role of caspases in immunotoxin- induced apoptosis of cancer cells. Biochemistry 37:16934-16942.

Kleinberg DL, Todd J (1980) Evidence that human growth hormone is a potent lactogen in primates. J Clin Endocrinol Metab 51:1009-1013.

212 Kline JB, Roehrs H, Clevenger CV (1999) Functional characterization of the intermediate isoform of the human prolactin receptor. J Biol Chem 274:35461-35468.

Kline JB, Rycyzyn MA, Clevenger CV (2002) Characterization of a novel and functional human prolactin receptor isoform (deltaS1PRLr) containing only one extracellular fibronectin-like domain. Mol Endocrinol 16:2310-2322.

Knudson AG, Jr. (1971) Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci U S A 68:820-823.

Kondo T, FitzGerald D, Chaudhary VK, Adhya S, Pastan I (1988) Activity of immunotoxins constructed with modified Pseudomonas exotoxin A lacking the cell recognition domain. J Biol Chem 263:9470-9475.

Kossiakoff AA (2004) The structural basis for biological signaling, regulation, and specificity in the growth hormone-prolactin system of hormones and receptors. Adv Protein Chem 68:147-169.

Kossiakoff AA, Somers W, Ultsch M, Andow K, Muller YA, De Vos AM (1994) Comparison of the intermediate complexes of human growth hormone bound to the human growth hormone and prolactin receptors. Protein Sci 3:1697-1705.

Kounnas MZ, Morris RE, Thompson MR, FitzGerald DJ, Strickland DK, Saelinger CB (1992) The alpha 2-macroglobulin receptor/low density lipoprotein receptor- related protein binds and internalizes Pseudomonas exotoxin A. J Biol Chem 267:12420-12423.

Labriola L, Montor WR, Krogh K, Lojudice FH, Genzini T, Goldberg AC, Eliaschewitz FG, Sogayar MC (2007a) Beneficial effects of prolactin and laminin on human pancreatic islet-cell cultures. Mol Cell Endocrinol 263:120-133.

Labriola L, Ferreira GB, Montor WR, Demasi MA, Pimenta DC, Lojudice FH, Genzini T, Goldberg AC, Eliaschewitz FG, Sogayar MC (2007b) Prolactin-induced changes in protein expression in human pancreatic islets. Mol Cell Endocrinol 264:16-27.

Lachelin GC, Yen SC, Alksne JF (1977) Hormonal changes following hypophysectomy in humans. Obstet Gynecol 50:333-339.

Lalonde FM, Myslobodsky M (2003) Are dopamine antagonists a risk factor for breast cancer? An answer from Parkinson's disease. Breast 12:280-282.

Langenheim JF, Chen WY (2005) Development of a prolactin receptor-targeting fusion toxin using a prolactin antagonist and a recombinant form of Pseudomonas exotoxin A. Breast Cancer Res Treat 90:281-293.

213 Langenheim JF, Tan D, Walker AM, Chen WY (2006) Two wrongs can make a right: dimers of prolactin and growth hormone receptor antagonists behave as agonists. Mol Endocrinol 20:661-674.

Lantinga-van Leeuwen IS, Oudshoorn M, Mol JA (1999) Canine mammary growth hormone gene transcription initiates at the pituitary-specific start site in the absence of Pit-1. Mol Cell Endocrinol 150:121-128.

Laud K, Gourdou I, Belair L, Peyrat JP, Djiane J (2000) Characterization and modulation of a prolactin receptor mRNA isoform in normal and tumoral human breast tissues. Int J Cancer 85:771-776.

Lebrun JJ, Ali S, Sofer L, Ullrich A, Kelly PA (1994) Prolactin-induced proliferation of Nb2 cells involves tyrosine phosphorylation of the prolactin receptor and its associated tyrosine kinase JAK2. J Biol Chem 269:14021-14026.

Lewis UJ (1984) Variants of growth hormone and prolactin and their posttranslational modifications. Annu Rev Physiol 46:33-42.

Lewis UJ, Bonewald LF, Lewis LJ (1980) The 20,000-dalton variant of human growth hormone: location of the amino acid deletions. Biochem Biophys Res Commun 92:511-516.

Li R, Sonik A, Stindl R, Rasnick D, Duesberg P (2000) Aneuploidy vs. gene mutation hypothesis of cancer: recent study claims mutation but is found to support aneuploidy. Proc Natl Acad Sci U S A 97:3236-3241.

Lincoln DT, Sinowatz F, Temmim-Baker L, Baker HI, Kolle S, Waters MJ (1998) Growth hormone receptor expression in the nucleus and cytoplasm of normal and neoplastic cells. Histochem Cell Biol 109:141-159.

Liu N, Mertani HC, Norstedt G, Tornell J, Lobie PE (1997) Mode of the autocrine/paracrine mechanism of growth hormone action. Exp Cell Res 237:196- 206.

Llovera M, Touraine P, Kelly PA, Goffin V (2000a) Involvement of prolactin in breast cancer: redefining the molecular targets. Exp Gerontol 35:41-51.

Llovera M, Pichard C, Bernichtein S, Jeay S, Touraine P, Kelly PA, Goffin V (2000b) Human prolactin (hPRL) antagonists inhibit hPRL-activated signaling pathways involved in breast cancer cell proliferation. Oncogene 19:4695-4705.

Lowman HB, Cunningham BC, Wells JA (1991) Mutational analysis and protein engineering of receptor-binding determinants in human placental lactogen. J Biol Chem 266:10982-10988.

214 Lu JC, Scott P, Strous GJ, Schuler LA (2002) Multiple internalization motifs differentially used by prolactin receptor isoforms mediate similar endocytic pathways. Mol Endocrinol 16:2515-2527.

Lukas G, Brindle SD, Greengard P (1971) The route of absorption of intraperitoneally administered compounds. J Pharmacol Exp Ther 178:562-564.

Maamra M, Finidori J, Von Laue S, Simon S, Justice S, Webster J, Dower S, Ross R (1999) Studies with a growth hormone antagonist and dual-fluorescent confocal microscopy demonstrate that the full-length human growth hormone receptor, but not the truncated isoform, is very rapidly internalized independent of Jak2-Stat5 signaling. J Biol Chem 274:14791-14798.

MacGillivray MH, Baptista J, Johanson A (1996) Outcome of a four-year randomized study of daily versus three times weekly somatropin treatment in prepubertal naive growth hormone-deficient children. Genentech Study Group. J Clin Endocrinol Metab 81:1806-1809.

Majumder B, Biswas R, Chattopadhyay U (2002) Prolactin regulates antitumor immune response through induction of tumoricidal macrophages and release of IL-12. Int J Cancer 97:493-500.

Malarkey WB, Kennedy M, Allred LE, Milo G (1983) Physiological concentrations of prolactin can promote the growth of human breast tumor cells in culture. J Clin Endocrinol Metab 56:673-677.

Manni A, Trujillo JE, Pearson OH (1977) Predominant role of prolactin in stimulating the growth of 7, 12-dimethylbenz(a)anthracene-induced rat mammary tumor. Cancer Res 37:1216-1219.

Massara F, Tangolo D, Godano A, Goffi S, Bertagna A, Molinatti GM (1985) Effect of metoclopramide, domperidone and apomorphine on GH secretion in children and adolescents. Acta Endocrinol (Copenh) 108:451-455.

Matera L, Contarini M, Bellone G, Forno B, Biglino A (1999) Up-modulation of interferon-gamma mediates the enhancement of spontanous cytotoxicity in prolactin-activated natural killer cells. Immunology 98:386-392.

Matera L, Geuna M, Pastore C, Buttiglieri S, Gaidano G, Savarino A, Marengo S, Vonderhaar BK (2000) Expression of prolactin and prolactin receptors by non- Hodgkin's lymphoma cells. Int J Cancer 85:124-130.

Maus MV, Reilly SC, Clevenger CV (1999) Prolactin as a chemoattractant for human breast carcinoma. Endocrinology 140:5447-5450.

215 McMurray RW, Weidensaul D, Allen SH, Walker SE (1995) Efficacy of bromocriptine in an open label therapeutic trial for systemic lupus erythematosus. J Rheumatol 22:2084-2091.

Meli R, Raso GM, Bentivoglio C, Nuzzo I, Galdiero M, Di Carlo R (1996) Recombinant human prolactin induces protection against Salmonella typhimurium infection in the mouse: role of nitric oxide. Immunopharmacology 34:1-7.

Melis GB, Mais V, Paoletti AM, Beneventi F, Gambacciani M, Fioretti P (1988) Prevention of puerperal lactation by a single oral administration of the new prolactin-inhibiting drug, cabergoline. Obstet Gynecol 71:311-314.

Mellado M, Rodriguez-Frade JM, Kremer L, von Kobbe C, de Ana AM, Merida I, Martinez AC (1997) Conformational changes required in the human growth hormone receptor for growth hormone signaling. J Biol Chem 272:9189-9196.

Meng J, Tsai-Morris CH, Dufau ML (2004) Human prolactin receptor variants in breast cancer: low ratio of short forms to the long-form human prolactin receptor associated with mammary carcinoma. Cancer Res 64:5677-5682.

Mershon J, Sall W, Mitchner N, Ben-Jonathan N (1995) Prolactin is a local growth factor in rat mammary tumors. Endocrinology 136:3619-3623.

Mertani HC, Zhu T, Goh EL, Lee KO, Morel G, Lobie PE (2001) Autocrine human growth hormone (hGH) regulation of human mammary carcinoma cell gene expression. Identification of CHOP as a mediator of hGH-stimulated human mammary carcinoma cell survival. J Biol Chem 276:21464-21475.

Mertani HC, Garcia-Caballero T, Lambert A, Gerard F, Palayer C, Boutin JM, Vonderhaar BK, Waters MJ, Lobie PE, Morel G (1998) Cellular expression of growth hormone and prolactin receptors in human breast disorders. Int J Cancer 79:202-211.

Metcalfe K, Lynch HT, Ghadirian P, Tung N, Olivotto I, Warner E, Olopade OI, Eisen A, Weber B, McLennan J, Sun P, Foulkes WD, Narod SA (2004) Contralateral breast cancer in BRCA1 and BRCA2 mutation carriers. J Clin Oncol 22:2328- 2335.

Milsom SR, Rabone DL, Gunn AJ, Gluckman PD (1998) Potential role for growth hormone in human lactation insufficiency. Horm Res 50:147-150.

Milsom SR, Breier BH, Gallaher BW, Cox VA, Gunn AJ, Gluckman PD (1992) Growth hormone stimulates galactopoiesis in healthy lactating women. Acta Endocrinol (Copenh) 127:337-343.

216 Mode A, Tollet P, Wells T, Carmignac DF, Clark RG, Chen WY, Kopchick JJ, Robinson IC (1996) The human growth hormone (hGH) antagonist G120RhGH does not antagonize GH in the rat, but has paradoxical agonist activity, probably via the prolactin receptor. Endocrinology 137:447-454.

Mol JA, Henzen-Logmans SC, Hageman P, Misdorp W, Blankenstein MA, Rijnberk A (1995) Expression of the gene encoding growth hormone in the human mammary gland. J Clin Endocrinol Metab 80:3094-3096.

Morris RE, Manhart MD, Saelinger CB (1983) Receptor-mediated entry of Pseudomonas toxin: methylamine blocks clustering step. Infect Immun 40:806-811.

Morton DB, Jennings M, Buckwell A, Ewbank R, Godfrey C, Holgate B, Inglis I, James R, Page C, Sharman I, Verschoyle R, Westall L, Wilson AB (2001) Refining procedures for the administration of substances. Report of the BVAAWF/FRAME/RSPCA/UFAW Joint Working Group on Refinement. British Veterinary Association Animal Welfare Foundation/Fund for the Replacement of Animals in Medical Experiments/Royal Society for the Prevention of Cruelty to Animals/Universities Federation for Animal Welfare. Lab Anim 35:1-41.

Mukhina S, Mertani HC, Guo K, Lee KO, Gluckman PD, Lobie PE (2004) Phenotypic conversion of human mammary carcinoma cells by autocrine human growth hormone. Proc Natl Acad Sci U S A 101:15166-15171.

Muller AF, Janssen JA, de Herder WW, van der Lely AJ (2001) [Growth hormone receptor antagonists: potential indications]. Ned Tijdschr Geneeskd 145:69-73.

Muñoz J, Esquivel J, Bustamante JJ, Martinez AO, Haro LS (2004) Anti-proliferative effects of a 2-mercaptoethanol resistant 45 kDa human growth hormone variant in MCF-7, MDA-MB-231 and T47D human breast cancer cell lines. Program of the 86th Annual Meeting of The Endocrine Society, New Orleans, LA, 2004, p 375 (Abstract P2-228).

Muñoz J, Aguilar RM, Vargas M, Martinez AO, Haro LS (2005) Beta-mercaptoethanol resistant 45-kDa human growth hormone inhibits STAT activation in T47-D, MCF-7 and MDA-MB-231 human breast cancer cell lines. Program of the 87th Annual Meeting of The Endocrine Society, San Diego, CA, 2005, p 563 (Abstract P114-5).

Murphy LJ, Murphy LC, Vrhovsek E, Sutherland RL, Lazarus L (1984) Correlation of lactogenic receptor concentration in human breast cancer with estrogen receptor concentration. Cancer Res 44:1963-1968.

Nabarro JD (1987) Acromegaly. Clin Endocrinol (Oxf) 26:481-512.

217 Nandi S, Guzman RC, Yang J (1995) Hormones and mammary carcinogenesis in mice, rats, and humans: a unifying hypothesis. Proc Natl Acad Sci U S A 92:3650-3657.

Nicola NA, Hilton DJ (1998) General classes and functions of four-helix bundle cytokines. Adv Protein Chem 52:1-65.

Noel S, Herman A, Johnson GA, Gray CA, Stewart MD, Bazer FW, Gertler A, Spencer TE (2003) Ovine placental lactogen specifically binds to endometrial glands of the ovine uterus. Biol Reprod 68:772-780.

Oberholtzer E, Contarini M, Veglia F, Cossarizza A, Franceschi C, Geuna M, Provinciali M, Di Stefano G, Sissom J, Brizzi MF, Pegoraro L, Matera L (1996) Prolactin increases the susceptibility of primary leukemia cells to NK and LAK effectors. Adv Neuroimmunol 6:233-247.

Ogata M, Chaudhary VK, Pastan I, FitzGerald DJ (1990) Processing of Pseudomonas exotoxin by a cellular protease results in the generation of a 37,000-Da toxin fragment that is translocated to the cytosol. J Biol Chem 265:20678-20685.

Ogata M, Fryling CM, Pastan I, FitzGerald DJ (1992) Cell-mediated cleavage of Pseudomonas exotoxin between Arg279 and Gly280 generates the enzymatically active fragment which translocates to the cytosol. J Biol Chem 267:25396-25401.

Okamura H, Zachwieja J, Raguet S, Kelly PA (1989) Characterization and applications of monoclonal antibodies to the prolactin receptor. Endocrinology 124:2499- 2508.

Olsson H, Ewers SB, Landin-Olsson M, Ranstam J (1985) Relation between tumour size and plasma prolactin levels in premenopausal patients with breast carcinoma. A preliminary report. Acta Radiol Oncol 24:57-59.

Ormandy CJ, Binart N, Kelly PA (1997a) Mammary gland development in prolactin receptor knockout mice. J Mammary Gland Biol Neoplasia 2:355-364.

Ormandy CJ, Hall RE, Manning DL, Robertson JF, Blamey RW, Kelly PA, Nicholson RI, Sutherland RL (1997b) Coexpression and cross-regulation of the prolactin receptor and sex steroid hormone receptors in breast cancer. J Clin Endocrinol Metab 82:3692-3699.

Owerbach D, Rutter WJ, Cooke NE, Martial JA, Shows TB (1981) The prolactin gene is located on chromosome 6 in humans. Science 212:815-816.

Page-Wilson G, Smith PC, Welt CK (2007) Short-term prolactin administration causes expressible galactorrhea but does not affect bone turnover: pilot data for a new lactation agent. Int Breastfeed J 2:10.

218 Peeva E, Venkatesh J, Michael D, Diamond B (2004) Prolactin as a modulator of B cell function: implications for SLE. Biomed Pharmacother 58:310-319.

Peirce SK, Chen WY (2004) Human prolactin and its antagonist, hPRL-G129R, regulate bax and bcl-2 gene expression in human breast cancer cells and transgenic mice. Oncogene 23:1248-1255.

Peirce SK, Chen WY, Chen WY (2001) Quantification of prolactin receptor mRNA in multiple human tissues and cancer cell lines by real time RT-PCR. J Endocrinol 171:R1-4.

Peters F, Schulze-Tollert J, Schuth W (1991) Thyrotrophin-releasing hormone--a lactation-promoting agent? Br J Obstet Gynaecol 98:880-885.

Posner BI, Verma AK, Patel BA, Bergeron JJ (1982) Effect of colchicine on the uptake of prolactin and insulin into Golgi fractions of rat liver. J Cell Biol 93:560-567.

Powers NG (1999) Slow weight gain and low milk supply in the breastfeeding dyad. Clin Perinatol 26:399-430.

Prowell TM, Davidson NE (2004) What is the role of ovarian ablation in the management of primary and metastatic breast cancer today? Oncologist 9:507-517.

Purnell DM, Hillman EA, Heatfield BM, Trump BF (1982) Immunoreactive prolactin in epithelial cells of normal and cancerous human breast and prostate detected by the unlabeled antibody peroxidase-antiperoxidase method. Cancer Res 42:2317-2324.

Qiu H, Belanger A, Yoon HW, Bunn HF (1998) Homodimerization restores biological activity to an inactive erythropoietin mutant. J Biol Chem 273:11173-11176.

Raccurt M, Lobie PE, Moudilou E, Garcia-Caballero T, Frappart L, Morel G, Mertani HC (2002) High stromal and epithelial human gh gene expression is associated with proliferative disorders of the mammary gland. J Endocrinol 175:307-318.

Ramakrishnan R, Khan SA, Badve S (2002) Morphological changes in breast tissue with menstrual cycle. Mod Pathol 15:1348-1356.

Ramamoorthy P, Sticca R, Wagner TE, Chen WY (2001) In vitro studies of a prolactin antagonist, hPRL-G129R in human breast cancer cells. Int J Oncol 18:25-32.

Rao YP, Buckley DJ, Buckley AR (1995) The nuclear prolactin receptor: a 62-kDa chromatin-associated protein in rat Nb2 lymphoma cells. Arch Biochem Biophys 322:506-515.

219 Rao YP, Olson MD, Buckley DJ, Buckley AR (1993) Nuclear co-localization of prolactin and the prolactin receptor in rat Nb2 node lymphoma cells. Endocrinology 133:3062-3065.

Rasmussen KM, Kjolhede CL (2004) Prepregnant overweight and obesity diminish the prolactin response to suckling in the first week postpartum. Pediatrics 113:e465- 471.

Reynolds C, Montone KT, Powell CM, Tomaszewski JE, Clevenger CV (1997) Expression of prolactin and its receptor in human breast carcinoma. Endocrinology 138:5555-5560.

Richards SM, Murphy WJ (2000) Use of human prolactin as a therapeutic protein to potentiate immunohematopoietic function. J Neuroimmunol 109:56-62.

Richards SM, Garman RD, Keyes L, Kavanagh B, McPherson JM (1998) Prolactin is an antagonist of TGF-beta activity and promotes proliferation of murine B cell hybridomas. Cell Immunol 184:85-91.

Richert MM, Schwertfeger KL, Ryder JW, Anderson SM (2000) An atlas of mouse mammary gland development. J Mammary Gland Biol Neoplasia 5:227-241.

Rosen EM, Fan S, Isaacs C (2005) BRCA1 in hormonal carcinogenesis: basic and clinical research. Endocr Relat Cancer 12:533-548.

Rozakis-Adcock M, Kelly PA (1992) Identification of ligand binding determinants of the prolactin receptor. J Biol Chem 267:7428-7433.

Rui H, Kirken RA, Farrar WL (1994a) Activation of receptor-associated tyrosine kinase JAK2 by prolactin. J Biol Chem 269:5364-5368.

Rui H, Lebrun JJ, Kirken RA, Kelly PA, Farrar WL (1994b) JAK2 activation and cell proliferation induced by antibody-mediated prolactin receptor dimerization. Endocrinology 135:1299-1306.

Russo J, Lynch H, Russo IH (2001) Mammary gland architecture as a determining factor in the susceptibility of the human breast to cancer. Breast J 7:278-291.

Russo J, Hu YF, Yang X, Russo IH (2000) Developmental, cellular, and molecular basis of human breast cancer. J Natl Cancer Inst Monogr:17-37.

Rycyzyn MA, Clevenger CV (2002) The intranuclear prolactin/cyclophilin B complex as a transcriptional inducer. Proc Natl Acad Sci U S A 99:6790-6795.

220 Rycyzyn MA, Reilly SC, O'Malley K, Clevenger CV (2000) Role of cyclophilin B in prolactin signal transduction and nuclear retrotranslocation. Mol Endocrinol 14:1175-1186.

Sakal E, Bignon C, Grosclaude J, Kantor A, Shapira R, Leibovitch H, Helman D, Nespoulous C, Shamay A, Rowlinson SW, Djiane J, Gertler A (1997) Large-scale preparation and characterization of recombinant ovine placental lactogen. J Endocrinol 152:317-327.

Schneider AB, Kowalski K, Sherwood LM (1975a) Identification of "big" human placental lactogen in placenta and serum. Endocrinology 97:1364-1372.

Schneider AB, Kowalski K, Sherwood LM (1975b) "Big" human placental lactogen: disulfide-linked peptide chains. Biochem Biophys Res Commun 64:717-724.

Schneider AB, Kowalski K, Buckman G, Sherwood LM (1977) Dimeric ("big") human placental lactogen. Immunological and biological activity. Biochim Biophys Acta 493:69-77.

Seddiki T, Delpal S, Aubourg A, Durand G, Ollivier-Bousquet M (2002) Endocytic prolactin routes to the secretory pathway in lactating mammary epithelial cells. Biol Cell 94:173-185.

Seetharam S, Chaudhary VK, FitzGerald D, Pastan I (1991) Increased cytotoxic activity of Pseudomonas exotoxin and two chimeric toxins ending in KDEL. J Biol Chem 266:17376-17381.

Shillingford JM, Miyoshi K, Robinson GW, Grimm SL, Rosen JM, Neubauer H, Pfeffer K, Hennighausen L (2002) Jak2 is an essential tyrosine kinase involved in pregnancy-mediated development of mammary secretory epithelium. Mol Endocrinol 16:563-570.

Shiu RP, Murphy LC, Tsuyuki D, Myal Y, Lee-Wing M, Iwasiow B (1987) Biological actions of prolactin in human breast cancer. Recent Prog Horm Res 43:277-303.

Siegall CB, Ogata M, Pastan I, FitzGerald DJ (1991) Analysis of sequences in domain II of Pseudomonas exotoxin A which mediate translocation. Biochemistry 30:7154- 7159.

Silva CM, Weber MJ, Thorner MO (1993) Stimulation of tyrosine phosphorylation in human cells by activation of the growth hormone receptor. Endocrinology 132:101-108.

Sinha YN (1995) Structural variants of prolactin: occurrence and physiological significance. Endocr Rev 16:354-369.

221 Sissom JF, Eigenbrodt ML, Porter JC (1988) Anti-growth action on mouse mammary and prostate glands of a monoclonal antibody to prolactin receptor. Am J Pathol 133:589-595.

Sivaprasad U, Canfield JM, Brooks CL (2004) Mechanism for ordered receptor binding by human prolactin. Biochemistry 43:13755-13765.

Smith CR, Norman MR (1990) Prolactin and growth hormone: molecular heterogeneity and measurement in serum. Ann Clin Biochem 27 ( Pt 6):542-550.

Sobrier ML, Duquesnoy P, Duriez B, Amselem S, Goossens M (1993) Expression and binding properties of two isoforms of the human growth hormone receptor. FEBS Lett 319:16-20.

Socolovsky M, Nam H, Fleming MD, Haase VH, Brugnara C, Lodish HF (2001) Ineffective erythropoiesis in Stat5a(-/-)5b(-/-) mice due to decreased survival of early erythroblasts. Blood 98:3261-3273.

Somers W, Ultsch M, De Vos AM, Kossiakoff AA (1994) The X-ray structure of a growth hormone-prolactin receptor complex. Nature 372:478-481.

Srivastava S, Matsuda M, Hou Z, Bailey JP, Kitazawa R, Herbst MP, Horseman ND (2003) Receptor activator of NF-kappaB ligand induction via Jak2 and Stat5a in mammary epithelial cells. J Biol Chem 278:46171-46178.

Staten NR, Byatt JC, Krivi GG (1993) Ligand-specific dimerization of the extracellular domain of the bovine growth hormone receptor. J Biol Chem 268:18467-18473.

Sterling K (1951) The turnover rate of serum albumin in man as measured by I131- tagged albumin. J Clin Invest 30:1228-1237.

Stevens DK, Eyre RJ, Bull RJ (1992) Adduction of hemoglobin and albumin in vivo by metabolites of trichloroethylene, trichloroacetate, and dichloroacetate in rats and mice. Fundam Appl Toxicol 19:336-342.

Stolar MW, Amburn K, Baumann G (1984) Plasma "big" and "big-big" growth hormone (GH) in man: an oligomeric series composed of structurally diverse GH monomers. J Clin Endocrinol Metab 59:212-218.

Sun R, Li AL, Wei HM, Tian ZG (2004a) Expression of prolactin receptor and response to prolactin stimulation of human NK cell lines. Cell Res 14:67-73.

Sun R, Wei H, Zhang J, Li A, Zhang W, Tian Z (2002) Recombinant human prolactin improves antitumor effects of murine natural killer cells in vitro and in vivo. Neuroimmunomodulation 10:169-176.

222 Sun R, Gault RA, Welniak LA, Tian ZG, Richards S, Murphy WJ (2003) Immunologic and hematopoietic effects of recombinant human prolactin after syngeneic bone marrow transplantation in mice. Biol Blood Marrow Transplant 9:426-434.

Sun R, Zhang J, Zhang C, Zhang J, Liang S, Sun A, Wang J, Tian Z (2004b) Human prolactin improves engraftment and reconstitution of human peripheral blood lymphocytes in SCID mice. Cell Mol Immunol 1:129-136.

Sundstrom M, Lundqvist T, Rodin J, Giebel LB, Milligan D, Norstedt G (1996) Crystal structure of an antagonist mutant of human growth hormone, G120R, in complex with its receptor at 2.9 A resolution. J Biol Chem 271:32197-32203.

Sytkowski AJ, Lunn ED, Risinger MA, Davis KL (1999) An erythropoietin fusion protein comprised of identical repeating domains exhibits enhanced biological properties. J Biol Chem 274:24773-24778.

Tae G, Kornfield JA, Hubbell JA (2005) Sustained release of human growth hormone from in situ forming hydrogels using self-assembly of fluoroalkyl-ended poly(ethylene glycol). Biomaterials 26:5259-5266.

Takahashi S (2004) Intrapituitary regulatory system of proliferation of mammotrophs in the pituitary gland. Zoolog Sci 21:601-611.

Tan D, Johnson DA, Wu W, Zeng L, Chen YH, Chen WY, Vonderhaar BK, Walker AM (2005) Unmodified prolactin (PRL) and S179D PRL-initiated bioluminescence resonance energy transfer between homo- and hetero-pairs of long and short human prolactin receptors in living human cells. Mol Endocrinol 19:1291-1303.

Teilum K, Hoch JC, Goffin V, Kinet S, Martial JA, Kragelund BB (2005) Solution structure of human prolactin. J Mol Biol 351:810-823.

Theuer CP, Buchner J, FitzGerald D, Pastan I (1993) The N-terminal region of the 37- kDa translocated fragment of Pseudomonas exotoxin A aborts translocation by promoting its own export after microsomal membrane insertion. Proc Natl Acad Sci U S A 90:7774-7778.

Thoreau E, Petridou B, Kelly PA, Djiane J, Mornon JP (1991) Structural symmetry of the extracellular domain of the cytokine/growth hormone/prolactin receptor family and interferon receptors revealed by hydrophobic cluster analysis. FEBS Lett 282:26-31.

Thorner MO, Strasburger CJ, Wu Z, Straume M, Bidlingmaier M, Pezzoli SS, Zib K, Scarlett JC, Bennett WF (1999) Growth hormone (GH) receptor blockade with a PEG-modified GH (B2036-PEG) lowers serum insulin-like growth factor-I but does not acutely stimulate serum GH. J Clin Endocrinol Metab 84:2098-2103.

223 Tomblyn S, Langenheim JF, Jacquemart IC, Holle E, Chen WY (2005) The role of human prolactin and its antagonist, G129R, in mammary gland development and DMBA-initiated tumorigenesis in transgenic mice. Int J Oncol 27:1381-1389.

Topper YJ, Freeman CS (1980) Multiple hormone interactions in the developmental biology of the mammary gland. Physiol Rev 60:1049-1106.

Touraine P, Martini JF, Zafrani B, Durand JC, Labaille F, Malet C, Nicolas A, Trivin C, Postel-Vinay MC, Kuttenn F, Kelly PA (1998) Increased expression of prolactin receptor gene assessed by quantitative polymerase chain reaction in human breast tumors versus normal breast tissues. J Clin Endocrinol Metab 83:667-674.

Trainer PJ, Drake WM, Katznelson L, Freda PU, Herman-Bonert V, van der Lely AJ, Dimaraki EV, Stewart PM, Friend KE, Vance ML, Besser GM, Scarlett JA, Thorner MO, Parkinson C, Klibanski A, Powell JS, Barkan AL, Sheppard MC, Malsonado M, Rose DR, Clemmons DR, Johannsson G, Bengtsson BA, Stavrou S, Kleinberg DL, Cook DM, Phillips LS, Bidlingmaier M, Strasburger CJ, Hackett S, Zib K, Bennett WF, Davis RJ (2000) Treatment of acromegaly with the growth hormone-receptor antagonist pegvisomant. N Engl J Med 342:1171- 1177.

Trott JF, Hovey RC, Koduri S, Vonderhaar BK (2003) Alternative splicing to exon 11 of human prolactin receptor gene results in multiple isoforms including a secreted prolactin-binding protein. J Mol Endocrinol 30:31-47.

Tsunekawa B, Wada M, Ikeda M, Uchida H, Naito N, Honjo M (1999) The 20-kilodalton (kDa) human growth hormone (hGH) differs from the 22-kDa hGH in the effect on the human prolactin receptor. Endocrinology 140:3909-3918.

Tyson JE, Khojandi M, Huth J, Andreassen B (1975) The influence of prolactin secretion on human lactation. J Clin Endocrinol Metab 40:764-773.

Ultsch M, de Vos AM (1993) Crystals of human growth hormone-receptor complexes. Extracellular domains of the growth hormone and prolactin receptors and a hormone mutant designed to prevent receptor dimerization. J Mol Biol 231:1133- 1136.

Ultsch M, de Vos AM, Kossiakoff AA (1991) Crystals of the complex between human growth hormone and the extracellular domain of its receptor. J Mol Biol 222:865- 868.

Ultsch MH, Somers W, Kossiakoff AA, de Vos AM (1994) The crystal structure of affinity-matured human growth hormone at 2 A resolution. J Mol Biol 236:286- 299.

224 Vekemans M, Delvoye P, L'Hermite M, Robyn C (1977) Serum prolactin levels during the menstrual cycle. J Clin Endocrinol Metab 44:989-993.

Vogel VG, Costantino JP, Wickerham DL, Cronin WM, Cecchini RS, Atkins JN, Bevers TB, Fehrenbacher L, Pajon ER, Jr., Wade JL, 3rd, Robidoux A, Margolese RG, James J, Lippman SM, Runowicz CD, Ganz PA, Reis SE, McCaskill-Stevens W, Ford LG, Jordan VC, Wolmark N (2006) Effects of tamoxifen vs raloxifene on the risk of developing invasive breast cancer and other disease outcomes: the NSABP Study of Tamoxifen and Raloxifene (STAR) P-2 trial. Jama 295:2727- 2741. von Minckwitz G, Kaufmann M (1998) New endocrine approaches in the treatment of breast cancer. Biomed Pharmacother 52:122-132.

Vonderhaar BK (1998) Prolactin: the forgotten hormone of human breast cancer. Pharmacol Ther 79:169-178.

Vonderhaar BK (1999) Prolactin involvement in breast cancer. Endocr Relat Cancer 6:389-404.

Wada M, Ikeda M, Takahashi Y, Asada N, Chang KT, Takahashi M, Honjo M (1997) The full agonistic effect of recombinant 20 kDa human growth hormone (hGH) on CHO cells stably transfected with hGH receptor cDNA. Mol Cell Endocrinol 133:99-107.

Walsh ST, Kossiakoff AA (2006) Crystal structure and site 1 binding energetics of human placental lactogen. J Mol Biol 358:773-784.

Walsh ST, Sylvester JE, Kossiakoff AA (2004) The high- and low-affinity receptor binding sites of growth hormone are allosterically coupled. Proc Natl Acad Sci U S A 101:17078-17083.

Wang DY, Stepniewska KA, Allen DS, Fentiman IS, Bulbrook RD, Kwa HG, De Stavola BL, Reed MJ (1995) Serum prolactin levels and their relationship to survival in women with operable breast cancer. J Clin Epidemiol 48:959-968.

Wang DY, Hampson S, Kwa HG, Moore JW, Bulbrook RD, Fentiman IS, Hayward JL, King RJ, Millis RR, Rubens RD, et al. (1986) Serum prolactin levels in women with breast cancer and their relationship to survival. Eur J Cancer Clin Oncol 22:487-492.

Wang PS, Walker AM, Tsuang MT, Orav EJ, Glynn RJ, Levin R, Avorn J (2002) Dopamine antagonists and the development of breast cancer. Arch Gen Psychiatry 59:1147-1154.

225 Warren WC, Byatt JC, Huynh M, Paik K, Pegg G, Staten NR (1999) Evaluation of the somatogenic activity of bovine placental lactogen with cell lines transfected with the bovine somatotropin receptor. Life Sci 65:2755-2767.

Waters MJ, Hoang HN, Fairlie DP, Pelekanos RA, Brown RJ (2006) New insights into growth hormone action. J Mol Endocrinol 36:1-7.

Wedekind JE, Trame CB, Dorywalska M, Koehl P, Raschke TM, McKee M, FitzGerald D, Collier RJ, McKay DB (2001) Refined crystallographic structure of Pseudomonas aeruginosa exotoxin A and its implications for the molecular mechanism of toxicity. J Mol Biol 314:823-837.

Wellings SR, Jensen HM, Marcum RG (1975) An atlas of subgross pathology of the human breast with special reference to possible precancerous lesions. J Natl Cancer Inst 55:231-273.

Wells JA, Cunningham BC, Fuh G, Lowman HB, Bass SH, Mulkerrin MG, Ultsch M, deVos AM (1993) The molecular basis for growth hormone-receptor interactions. Recent Prog Horm Res 48:253-275.

Welsch CW, Nagasawa H (1977) Prolactin and murine mammary tumorigenesis: a review. Cancer Res 37:951-963.

Welsch CW, Adams C, Lambrecht LK, Hassett CC, Brooks CL (1977) 17beta-oestradiol and Enovid mammary tumorigenesis in C3H/HeJ female mice: counteraction by concurrent 2-bromo-alpha-ergocryptine. Br J Cancer 35:322-328.

Wennbo H, Gebre-Medhin M, Gritli-Linde A, Ohlsson C, Isaksson OG, Tornell J (1997) Activation of the prolactin receptor but not the growth hormone receptor is important for induction of mammary tumors in transgenic mice. J Clin Invest 100:2744-2751.

Williams RD (1995) Breast-feeding best bet for babies. In: FDA consumer. [Rockville, MD]: U.S. Food and Drug Administration.

Wiseman BS, Werb Z (2002) Stromal effects on mammary gland development and breast cancer. Science 296:1046-1049.

Woody MA, Welniak LA, Richards S, Taub DD, Tian Z, Sun R, Longo DL, Murphy WJ (1999a) Use of neuroendocrine hormones to promote reconstitution after bone marrow transplantation. Neuroimmunomodulation 6:69-80.

226 Woody MA, Welniak LA, Sun R, Tian ZG, Henry M, Richards S, Raziuddin A, Longo DL, Murphy WJ (1999b) Prolactin exerts hematopoietic growth-promoting effects in vivo and partially counteracts myelosuppression by azidothymidine. Exp Hematol 27:811-816.

Wu H, Devi R, Malarkey WB (1996) Localization of growth hormone messenger ribonucleic acid in the human immune system--a Clinical Research Center study. J Clin Endocrinol Metab 81:1278-1282.

Xu BC, Chen WY, Gu T, Ridgway D, Wiehl P, Okada S, Kopchick JJ (1995) Effects of growth hormone antagonists on 3T3-F442A preadipocyte differentiation. J Endocrinol 146:131-139.

Yang N, Langenheim JF, Wang X, Jiang J, Chen WY, Frank SJ (2007) GH receptor activation by GH and GH antagonist dimers. Program of the 89th Annual Meeting of The Endocrine Society, Toronto, Canada, 2007, p 430 (Abstract P2-402).

Zdanovsky AG, Chiron M, Pastan I, FitzGerald DJ (1993) Mechanism of action of Pseudomonas exotoxin. Identification of a rate-limiting step. J Biol Chem 268:21791-21799.

Zdanovsky AG, Zdanovskaia MV, Strickland D, FitzGerald DJ (1996) Ligand-toxin hybrids directed to the alpha 2-macroglobulin receptor/low density lipoprotein receptor-related protein exhibit lower toxicity than native Pseudomonas exotoxin. J Biol Chem 271:6122-6128.

Zhang G, Li W, Holle L, Chen N, Chen WY (2002) A novel design of targeted endocrine and cytokine therapy for breast cancer. Clin Cancer Res 8:1196-1205.

Zhang J, Sun R, Tian Z (2007) Human prolactin promotes human secondary immunoglobulin response in human/SCID mouse chimeras. Clin Vaccine Immunol 14:60-64.

Zhang J, Sun R, Wei H, Tian Z (2005a) Antitumor effects of recombinant human prolactin in human adenocarcinoma-bearing SCID mice with human NK cell xenograft. Int Immunopharmacol 5:417-425.

Zhang WC, Sun R, Zhang J, Zhang J, Tian Z (2005b) Recombinant human prolactin protects against irradiation-induced myelosuppression. Cell Mol Immunol 2:379- 385.

Zhang X, Zhu T, Chen Y, Mertani HC, Lee KO, Lobie PE (2003) Human growth hormone-regulated HOXA1 is a human mammary epithelial oncogene. J Biol Chem 278:7580-7590.

227 Zhu T, Starling-Emerald B, Zhang X, Lee KO, Gluckman PD, Mertani HC, Lobie PE (2005) Oncogenic transformation of human mammary epithelial cells by autocrine human growth hormone. Cancer Res 65:317-324.

228