Developing Synthetic Peptide-Based Inhibitors of Human Growth Receptor

A Thesis

Presented to

The Honors Tutorial College

Ohio University

In Partial Fulfillment of the Requirements for Graduation from the Honors Tutorial College with the degree of

Bachelor of Science in Chemistry

by

Maya R. Sattler

May 2018

2

Acknowledgements

This thesis represents the work a large number of people, to whom I am incredibly grateful. Thank you first and foremost to my adviser, Dr. Justin Holub, whose patience and teaching has been outstanding for the entire three years I have worked with him and been his student, all the way to the final deadline. Thank you to Dr. John

Kopchick for allowing me the opportunity to explore a different area of research and a different laboratory environment. Both Dr. Lauren McMills and Dean Cary Frith have offered wonderful support and guidance through my entire HTC career and I will forever be appreciative.

There are also many fellow students that have helped me with work and encouragement. Thank you to my fellow students in the Holub laboratory, especially

Najah Alqaeisoom who was my first teacher and a delight to work with, and Danushka

Arachchige who was always willing to help, no matter the problem. I would also like to thank my colleagues at Edison Biotechnology Institute for their support in all things trivial and otherwise. Dr. Reetobrata Basu in particular made all of the cell work possible through his teaching, assistance, and reassurances.

Lastly, I am grateful to a larger community of people in Athens and worldwide that helped me through my four years of undergraduate studies and will continue to support me as I continue to progress.

3

Table of Contents

Abstract 4

1. Introduction

1.1 Biology of hGH and GHR 5

1.2 Clinical Relevance 8

1.3 Inhibition of hGH Signaling 11

1.4 Peptide design 13

1.5 Scanning Alanine Library 17

1.6 Project Rationale 19

2. Materials and Methods

2.1 Reagents and Chemicals 21

2.2 Peptide Production and Characterization 22

2.3 Biological Assays 27

3. Results and Discussion

3.1 Peptide Production and Characterization 30

3. 2 Biological Assays 32

4. Conclusions 41

References 43

Appendices

A. MS Spectra 52

B. Analytical Spectra 57

C. CD Spectra 63

4

Abstract

The human (hGH) is an important endocrine mediator throughout life with myriad effects and receptors in every tissue of the body and therefore an excess of hGH signaling can negatively impact health. This thesis covers the development of peptide-based mimetics of hGH site 1 that are designed to antagonize the (GHR) and inhibit hGH signaling via the Stat5 pathway. In order to investigate the contributions of individual amino acid residues on hGH•GHR interaction, a control peptide and ten peptides that differed by one substitution from the control were synthesized, purified, tested for helical propensity, and screened for hGH signaling inhibition. A comparison of the residues in contact with GHR, the helical propensity, and the biological activity indicated that the ability to form a helix was more important to the inhibitory function of the peptides than was the presence of residues that make direct contact with the GHR. Further development of these peptide-based chemical genetic agents could offer new insights into the fundamental nature of hGH•GHR interactions and improve upon existing therapeutics.

5

1. Introduction

1.1 Biology of hGH and GHR

Human growth hormone (hGH) is a 22 kDa single-chain peptide hormone with

191 amino acid residues produced in the anterior .1 Its structure consists of a four-helical bundle with an unusual up-up-down-down connectivity and two disulfide linkages.2 Growth hormone signaling is mediated through the binding of hGH with the growth hormone receptor (GHR), a 28 kDa transmembrane protein in the class I cytokine receptor family. One molecule of hGH binds to two molecules of GHR, which exist as homodimers on the cellular surface even prior to ligand binding.3 The structure of hGH bound to two GHRs can be seen in Figure 1.1 A. There are two distinct binding sites on hGH; site 1 is mainly formed by the A and D helices, whereas site 2 is mainly formed by the A and C helices (Figure 1.1 B).4 The hGH binds sequentially to two overlapping sites on GHR. Site 1 of the hormone binds to the first receptor, followed by subsequent binding of site 2 to the second receptor. This binding provides the driving energy for the change in the relation between the helices of the transmembrane domains of the receptors.5 Upon binding, the helices rotate from a parallel state to a left-handed crossover state, which in turn changes the arrangement between the cytoplasmic domains of the receptor. 6

Figure 1.1. A. hGH (pink) bound to two GHR (purple). Adapted from PDB 1HWG. B. hGH structure with labeled helices. Adapted from PDB 1HGU.

The conformational changes in GHR associated with hGH binding initiate an intracellular signaling cascade that begins with the tyrosine kinase, Janus kinase 2

(JAK2).3,5 The GHR lacks intrinsic kinase activity but it binds JAK2 at a proline-rich motif known as Box 1 in its cytoplasmic intracellular domain, close to the cell membrane.6 When the cytoplasmic domains move apart from each other, the pseudokinase inhibitory domain of each JAK2 slides away from the kinase domain of the other JAK2, allowing them to activate each other. This initiates cross-phosphorylation of key tyrosine residues on the GHRs, allowing other molecules to dock and be phosphorylated by JAK2. JAK2 has the potential to activate a number of pathways including the mitogen-activated protein kinase (MAPK) pathway and the

Akt/phosphoinositide 3-kinase (PI3K) pathway. However, the primary route of action is through family of gene transcription factors called Signal Transducers and Activators of

Transcription (STATs), which has seven members. STAT1, STAT3, and STAT5 are 7 activated by JAK2, but STAT5 in particular mediates the majority of the genomic effects of hGH, such as cell proliferation. Two isoforms, STAT5a and STAT5b, exist and form a dimer when phosphorylated, which translocates to the nucleus to modulate target gene expression. The presence of phosphorylated STAT5 can be used as an indicator of hGH signaling.7

The most notable target of STAT5 activation is the gene encoding -like growth factor 1 (IGF-I), a protein so integral to hGH action that its effects are credited not solely to hGH, but to the hGH/IGF-I axis.8 IGF-I is produced in many tissues in the body where it exerts local effects as an auto- and paracrine hormone. However, liver production of IGF-I is particularly important because the hormone produced there acts as an endocrine signal, having broad-spectrum effects throughout the body.9 One effect is to inhibit the production of hGH from the anterior pituitary, acting as part of a negative feedback cycle that stops its own production. Though hGH and IGF-I production are interdependent, they act in synergistic but independent ways to produce their effect.

Receptors for both hGH and IGF-I are present in a wide variety of tissues and thus they have myriad important effects.3 The growth effect for which this hormone axis is most well known is achieved through bone growth and mediated through paracrine

IGF-I in the bone plate. The effect of the hGH/IGF-I axis in muscle is to promote protein synthesis and positive nitrogen balance, which leads to an increase in muscle and thus, lean body weight.9 Further contributing to a change in body composition, the hormone axis promotes lipolysis, or the breakdown of fatty acids in adipose tissue, and inhibits lipogenesis, the formation of fatty acids.6 Other effects on adipose tissue include control over preadipocyte proliferation, differentiation, and senescence, and altering levels of 8 adipokines, the released only adipose fat tissue. This axis also affects the critical metabolic hormone, insulin. While hGH promotes insulin resistance10, IGF-I has some insulin-like activity, which could be explained in part by the structural and functional homology between IGF-I and insulin and the IGF-I receptor and insulin receptor.11

1.2 Clinical Relevance

Because growth hormone is implicated in not only growth, but also metabolism and endocrine signaling, abnormal levels of hGH in the body can have serious physiological and pathological ramifications. and are two related conditions caused by an excess of hGH activity, usually due to excess hGH secretion from a , a type of tumor in the gland that produces hGH.12 Gigantism results from hypersecretion of hGH in childhood, before the fusion of the epiphyseal growth plates, and is most prominently characterized by excessive growth.13 Acromegaly is a similar condition in adulthood, characterized by excessive growth of organs and tissues, including skeletal tissue. The long-term exposure of hGH and IGF-I resulting from untreated acromegaly is associated with numerous detrimental effects, including arthritis, cardiomyopathy, hypertension, arrhythmias, insulin resistance, diabetes, sleep apnea, and renal failure.12 Furthermore, metabolic, respiratory, and cardio- or cerebrovascular comorbidities contribute to an increase in standardized mortality rates by a factor of two and a decrease in life expectancy by ten years.

Acromegaly provides a clear illustration of the adverse health effects of excess hGH, but such severe deviations in hGH levels are not needed for the manifestation of 9 these effects. For example, growth hormone has well-documented diabetogenic activity.10,11,14 The IGF-I produced via hGH signaling can mimic insulin action due to its homology with insulin. This action improves glucose homeostasis and decreases insulin resistance, but the diabetogenic affects of hGH ultimately dominate. Growth hormone promotes diabetes by antagonizing insulin action in peripheral tissues, increasing glucose production, decreasing glucose uptake in adipose tissue, increasing insulin production, stimulating lipolysis in visceral adipose tissue, and increasing the expression of some gluconeogenic genes that increase available levels of glucose.14 The diabetogenic properties are also demonstrated by a greatly increased insulin sensitivity with the removal of growth hormone signaling in the GHR knockout mouse model.15 A better understanding of and ability to manipulate hGH may contribute to the understanding and management of diabetes.

The hGH-IGF-I axis has also been shown to affect the incidence and progression of various cancers.16,17 Both hGH and IGF-I have mitogenic and anti-apoptotic effects, thus promoting tumor formation and progression. As such, antagonism of the hGH-IGF-I axis has been proposed as a way to control and treat cancer. This axis has also been implicated in the progression of chronic kidney disease (CKD), a problem that affects as much as one tenth of the general human population.18 While the effects of IGF-I may be more modest, hGH negatively affects the severity and rate of development of glomerulosclerosis, a major component of CKD.19 Growth hormone deficiency may have a protective effect on this condition, which suggests hGH antagonism could be a potential route to slow CKD progress. 10

A large population-based cohort study found that elevated fasting hGH levels were associated with stroke, coronary artery disease, congestive heart failure, cardiovascular mortality, and all-cause mortality after controlling for traditional cardiovascular risk factors, with the strongest association found with cardiovascular mortaility.20 These negative consequences of elevated hGH are supported by a higher prevalence of cardiovascular disease and hypertension in acromegalic patients and a mouse model with excess growth hormone action through the expression of the bovine growth hormone (bGH).21 Because growth hormone levels affect the structure and function of the heart, hormone excess and deficiency each may have negative health consequences.

Considering the number and severity of the health issues associated with excess hGH, it is not surprising that decreased growth hormone signaling is associated with an increased life span. To date, the growth hormone receptor knock out mouse holds the

“Methuselah Mouse Prize” that recognizes the longest lived mouse worldwide.15 One mouse of this type lived 1819 days and in general their life span is increased 38% for females and 55% for males. The inhibition of the hGH/IGF-I axis may directly mediate this increase in lifespan, based on a study of mice with no GH signaling and differing body sizes.22 This differs from a previous hypothesis suggesting that the benefit was derived from a difference in body composition. Thus, antagonism of GH signaling may play an important role in aging research.

11

1.3 Inhibition of hGH Signaling

The numerous ways in which hGH affects human health suggests that agents that suppress hGH signaling have many potential therapeutic uses. Three distinct classes are currently available to treat acromegaly and gigantism: dopamine agonists, analogues, and growth hormone receptor antagonists.23 Dopamine analogues such as the ergot-derived bromocriptine and cabergoline are used to treat conditions unrelated to hGH and normally stimulate hGH secretion; however, in acromegalic patients they inhibit hGH secretion.24 Somatostatin is a short peptide that inhibits the release of hGH and other hormones such as insulin, glucagon, and gastrin.25 Somatostatin analogues such as and lanreotide replicate this inhibition and are used to treat conditions related to the pancreas, pituitary, and neuroendocrine and gastrointestinal systems.23,26

Somatostatin analogues are generally more effective and better tolerated than dopamine agonists, but they inhibit the secretion of a number of other hormones, including thyroid stimulating hormone, insulin, and glucagon.23 Furthermore, they are only effective at controlling serum IGF-I levels in fewer than 2/3 of acromegalic patients, due in part to a difference in the levels of expression of different subtypes of somatostatin receptors and dopamine receptors,27 necessitating alternative therapies.

Pegvisomant is the first example of an hGH receptor antagonist and, to the best of the author’s knowledge, the only hGH receptor antagonist currently available as a therapeutic.28 It was developed by Dr. John J. Kopchick at Ohio University and received

FDA approval in 2003.29 Development began with an altered bovine growth hormone

(bGH) gene and its protein product, which was first reported in 1990.30 Three amino acid substitutions in the C α-helix of bGH were made: E117L, G119R, and A122D. While the 12 protein was found to have the same binding affinity to GHR-rich mouse liver preparations, transgenic mice that expressed the gene had a growth-suppressed phenotype. Further investigation of the contribution of individual residues identified

G119 as critical to successful GH signal transduction. In humans, the critical residue was found to be G120, and mutation at this position was used as a starting point for the antagonist.31,32

In order to increase half-life and decrease immunogenicity, the antagonist was pegylated, a process by which poly(ethylene glycol) (PEG) is conjugated to specific locations on the protein.33,34 Pegylation at the primary amine groups of lysines and the N- termini of peptides is common. Accordingly, is pegylated at lysines 39, 120,

140, and 158, and the N-terminal phenylalanine. The addition of 5 kDa PEG moieties has the potential to cause problematic steric hindrance, and two lysines within site 1 were mutated (K168A and K172R) to preserve the molecule’s ability to bind to the receptor.35

Though pegylation decreases affinity for the GH receptor by a factor of 20, it increases the serum half-life from 15 minutes to 100 hours and decreases the binding affinity to GH binding protein.34,35 Lastly, to overcome the potential decrease in binding affinity of the modified hGH molecule to the GHR, six more mutations were introduced that were previously shown to increase the site 1 binding affinity (H18D, H21N, R167N, D171S,

E174S, and I179T).23,35,36 Whether these mutations had their intended affect was later disputed, but they did eliminate the affinity of the antagonist for the prolactin receptor, reducing potential side effects.23,34,37

The final product of these modifications is pegvisomant, which has been marketed as a drug for acromegaly under the name Somavert®, pegvisomant for injection.23 Several 13 short- and long-term studies have established the efficacy of pegvisomant in treating acromegaly.23,35,38,39 Dose-dependent decreases in serum IGF-I levels are observed and when dose levels are titrated properly, normal levels can be achieved. Nearly one-third of acromegalic patients develop diabetes, but markers of insulin resistance including serum insulin and glucose concentrations decrease with pegvisomant use, which is not observed with somatostatin anologues.35,40

However, despite the successes of pegviosmant, several drawbacks exist. Growth of the pituitary tumors that cause acromegaly is reported in some cases. Another serious adverse effect is an elevation of liver enzymes, although the long term ACROSTUDY did not find any case of sustained liver damage.39 Beyond the health effects, pegvisomant treatment is a considerable inconvenience, requiring daily injection to be effective.

Additionally, because pegvisomant is produced recombinantly in E. coli, a bacterium that produces no endogenous growth hormone, and is then further modified chemically, production is expensive and can result in product heterogeneity, such as misfolded isoforms and different extents of pegylation.41 Combination therapy of weekly pegvisomant injections and monthly somatostatin analogue injections has been reported as a rational treatment approach,42 but other possibilities may exist to improve the treatment of excess hGH.

1.4 Peptide Design

Synthetic peptides offer an alternative to large proteins like pegvisomant to target large-scale protein-protein interactions like that between hGH and GHR. The Holub laboratory at Ohio University routinely synthesizes such peptides using solid-phase 14 peptide synthesis (SPPS) to study the biological function of proteins and protein-protein interactions.43,44 A synthetic peptide that mimics the sequence of hGH has the potential to improve upon pegvisomant as an inhibitor and to serve as a more easily manipulated chemical genetic agent to investigate the interaction between the hormone and its receptor. This thesis focuses on developing a library of synthetic peptides designed to mimic the sequence of a small helical portion of hGH involved in site 1 binding.

Synthetic peptides offer an array of advantages over traditional small molecule or biologic (full protein or antibody) approaches.45,46 The greater size of peptides compared to small molecules may allow for higher affinities because of increased numbers of interactions with a target, especially at shallow or hydrophobic binding surfaces.47 This results in molecules that have greater specificity, higher potency, increased efficacy, and decreased toxicity, side effects and drug-drug interactions. Peptides with fewer than fifty amino acids are highly amenable to SPPS, which reduces manufacturing costs compared to biologics. This method of synthesis also greatly increases the types of modifications that can be made on the molecule, expands control over primary sequence, and offers superior routes of purification over recombinant or cell-free synthesis systems. The immunogenicity and toxicity of peptide drugs is generally lower regardless of synthesis method because they degrade into naturally occurring amino acids, decreasing bioaccumulation.48 Additionally, their smaller size relative to protein-based therapeutics allows them to better penetrate into tissues, which could improve their , and they can even be engineered to gain cell entry.49 In addition to their potential therapeutic benefits, researchers can more readily manipulate their structure during synthesis to create specialized tools to better understand biological processes. Even short 15 peptides can form protein-like structures, such as helices, sheets, and turns, allowing them to better mimic and interact with native biological molecules.50 In theory, peptide- based drugs could be designed to target virtually any biomolecular interaction.

Using peptides to target protein interactions does present unique challenges, such as susceptibility to degradation by proteases, low oral bioavailability, high clearance rates from the body, and low membrane permeability, but such problems are quickly being overcome.45,46 Importantly, the interaction of hGH and GHR takes place on the cell surface, obviating the need to design the peptide to penetrate the cell membrane. Several methods can be employed to reduce degradation of the peptide by native proteases,51 including cyclization,52,53 the use of non-natural amino acids,54,55 capping of the N- terminal end of the peptide with an acetyl group,56 or pegylation57. Pegylation can also counteract a high rate of clearance from the body.33 High conformational flexibility has the potential to increase off-target effects, but this issue can be reduced by cyclization or the incorporation of a motif into a more stable scaffold. Peptides designed to fold into stable secondary and tertiary structures are often both more functional and more resistant to proteolytic degradation.50 Lastly, the potentially low oral bioavailability could be overcome by a combination of protease inhibitors and membrane penetration enhancers or alternative delivery methods such as nasal, pulmonary, or sublingual routes, transdermal patches, or controlled release subcutaneous or intramuscular routes.46 Such improvements in drug delivery would represent a substantial improvement over the daily injections currently required for pegvisomant treatment.

Because peptides violate the traditional principles for good drug lead compounds,58 their development has only accelerated more recently.46 Despite this, as of 16

March 2017, there were 60 peptide-based drugs approved for use in the United States,

Europe, and/or Japan, with 155 more in active clinical development.59 The Holub laboratory seeks to take advantage of this largely unexplored chemical space to design better chemical genetic agents to probe large-scale protein-protein interactions. An ongoing project in the Holub laboratory utilizes the scorpion toxin scyllatoxin (ScTx) as a scaffold to display residues of the helical BH3 domain to target Bcl-2 proteins.43,60 ScTx is a peptide found in the venom of the scorpion Leirus quinquestriatus hebraeus that is 32 amino acids in length and contains one alpha-helix and two beta-strands that are stabilized by three disulfide linkages (Figure 1.2).61 Several therapeutics based on scorpion toxins with similar structural motifs and disulfide linkages have already been successfully developed.50 Work outlined in this thesis utilizes similar ScTx backbones and other peptide designs to mimic helical portions of hGH site 1 to target GHR. These constructs were developed in the Holub laboratory and tested in collaboration with the lab of Dr. John Kopchick at the Edison Biotechnology Institute at Ohio University.

Figure 1.2. A ribbon representation of ScTx rendered in PyMol from the PDB file 1SCY.

The three disulfide linkages are shown as ball-and-stick in yellow.

17

A representative example of ScTx and the process by which ScTx-hGH mimetics were designed can be seen in Figure 1.3. For both peptides design strategies, residues 36-

51 of hGH were chosen. These residues correspond to a small helical region in site 1 of hGH that does not belong to any of the four main helices of hGH. It is hypothesized that either of these peptide-based hGH mimetics will interact with the GHR at the same location that site 1 of hGH does and that this interaction may be able to inhibit the binding of hGH to the receptor. However, unlike hGH, the small size of these mimetics means that they will bind in a 1:1 ratio with the GHR instead of a 1:2 ratio.

Figure 1.3. A. Schematic representation of how hGH residues were designed to be displayed using a ScTx scaffold to target GHR. B. Sequence alignment of wild type ScTx (wtScTx) and residues 36-51 of hGH compared to the ScTx-hGH construct with marked areas of alpha or beta sheet structure. Structural cysteines are shown in orange and the residues of hGH to be displayed on the ScTx-hGH construct are shown in blue.

1.5 Scanning Alanine Library

To better understand the contribution of each individual amino acid in the binding of hGH and GHR, a scanning alanine library based on the hGH (36-51) construct was 18 also designed. The library consisted of ten peptides in which one residue in the sequence is systematically replaced with an alanine residue. Alanine is used for this technique because its small, nonionic, nonpolar methyl side chain is relatively unreactive and maintains proper spacing of the peptide sequence.62 While the side chain of glycine is even smaller (a single hydrogen), its unusual backbone dihedral angle can disrupt helicity in the context of peptide secondary structure.62 In theory, the contribution of the residue that was replaced should be apparent in the difference between the peptide with the replaced residue and the hGH (36-51) peptide various physicochemical properties, such as helical propensity, ability to inhibit hGH signaling, or the binding affinity with the

GHR. Based on which amino acid side chains face towards the GHR in the crystal structure of hGH bound to GHR, it is predicted that residues K39, K42, Y43, L46, and

Q47 will have the greatest affect on the activity of the peptides (Figure 1.4).

Figure 1.4. Representation of hGH (pink, yellow, green) bound to dimeric GHR (purple) from the PDB file 1HWG, rendered in PyMol. Residues 36-51 are shown in yellow or green, with side chains included, and residues expected to affect binding are shown in green.

19

1.6 Project Rationale

The peptides used in this project were generated by established Fmoc-based solid- phase peptide synthesis (SPPS) procedures.63,64 After confirmation of identity and purity, the peptides were tested for structural features and biological activity. Circular dichroism

(CD) spectropolarimetry was used to determine what features of peptide secondary structure were present in each peptide. These results were then correlated to each peptide’s ability to inhibit the hGH signaling pathway as determined by the presence of phosphorylated Stat5 (pStat5) in western blot or enzyme-linked immunosorbent assay

(ELISA). A simplified representation of several of the steps in the hGH signaling pathway, including the phosphorylation of Stat5 and where the peptides are anticipated to inhibit the pathway, as shown in Figure 1.5. Collecting and interpreting these data has facilitated a better understanding of the hGH•GHR binding interaction and may lead to the development of a new peptide-based therapeutic to treat acromegaly and other conditions of excess hGH action.

20

Figure 1.5. Schematic representation of the hGH signaling pathway including steps relevant to this project.8 The extracellular binding of hGH to its dimeric, membrane- bound receptor leads to the activation of intracellular pathways. One pathway leads to the phosphorylation of Stat5, which translocates to the nucleus where it facilitates the transcription of genes such as IGF-I that are critical to hGH action. The peptides designed for this project are proposed to inhibit hGH signaling by blocking the binding of hGH to its receptor.

21

2. Materials and Methods

2.1 Reagents and Chemicals

Fmoc-PAL-AM resin, Fmoc-protected amino acids, and PyClock were purchased from Novabiochem (Billerica, MA). Piperidine, N,N-diisopropylethylamine (DIEA), and

N-methyl-2-pyrrolidone (NMP) were obtained from Sigma-Aldrich (St. Louis, MO).

Acetonitrile (ACN) was purchased from Alfa Aesar (Ward Hill, MA).

Human malignant melanoma cell line, SK-MEL-28 (HTB-72); human B lymphoblast cell line, IM9 (CCL-159); EMEM, RPMI 1640 medium, and fetal bovine serum (FBS) were purchased from American Type Culture Collection (ATCC; Manassas,

VA). The antibiotic-antimycotic (containing penicillin, streptomycin, and amphotericin

B), Halt protease and phosphatase inhibitor cocktail, ElisaOne kit (catalogue # 85-86112-

11), and West Femto Chemiluminescence detection reagents were purchased from

Thermo Fisher Scientific (Waltham, MA). RIPA lysis buffer (10 mM Tris-Cl (pH 8.0), 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, 140 mM NaCl, and 1 mM PMSF) and Bradford reagent were purchased from Sigma-Aldrich.

Recombinant hGH (catalogue # 1067-GH) was purchased from R&D Systems

(Minneapolis, MN). The monoclonal, rabbit, anti-β-actin (#4970) primary antibody, the rabbit anti-pStat5 (Y694) (#9351) primary antibody, and the donkey anti-rabbit IgG,

HRP-linked Antibody (#7074) were purchased from Technology

(Danvers, MA). All other materials were purchased from commercial sources and used without modification except where otherwise noted.

22

2.2 Peptide Production and Characterization

Solid-phase peptide synthesis: All peptides described in this thesis were synthesized by solid-phase peptide synthesis (SPPS) using standard Fmoc-based procedures.65,66 Iterative cycles of amino acid deprotection and coupling were repeated until all amino acids of the desired sequence were added. A schematic representation of this process can be found in Figure 2.1. All deprotection and coupling reactions were performed in a microwave-accelerated reaction system (CEM, Matthews, NC) using software programs written in-house.44 All peptides were synthesized in reaction vessels on a 25 or 50 µmol scale on Fmoc-PAL-AM resin. To remove terminal Fmoc protection groups, the resin was treated with 25% (v/v) piperidine in NMP containing 0.1 M hydroxybenzotriazole (HOBt) to minimize aspartamide formation67. To couple the next amino acid via an amide bond, the deprotected resin was treated with 5 equivalents (eq) of amino acid, 5 eq of PyClock and 10 eq of DIEA in NMP. All equivalents were based on resin loading level. The resin was washed three times with NMP after each deprotection and each coupling step to remove protecting groups and unreacted starting reagents.

23

Figure 2.1. SPPS process.64

Acetylation: All peptides were capped with an N-terminal acetyl group to enhance metabolic stability68. Following synthesis, resin-bound peptides were deprotected as described above and subsequently treated with 6% (v/v) acetic anhydride and 6% (v/v) 4- methylmorpholine in NMP for 20 minutes at room temperature. This reaction was repeated and then resin-bound peptides were washed with NMP to remove any unreacted starting reagents.

Cleavage of peptides from resin: Peptides were washed with NMP and dichloromethane (DCM) and then dried under vacuum to remove residual solvent. A cleavage mixture of 88% (v/v) trifluoroacetic acid (TFA), 5% (v/v) water, 5% (v/v) phenol and 2% (v/v) triisopropylsilane was added to the resin and the reaction proceeded for 30 minutes at 38 ˚C in a CEM microwave reactor. The peptide was then precipitated in cold diethyl ether, pelleted by centrifugation, and resuspended in an appropriate 24 volume of aqueous acetonitrile. This solution was then frozen and lyophilized to remove all solvent. Crude peptides in powder form were stored at -20 ˚C until further use.

Purification by HPLC: Crude peptide powders were resuspended in an appropriate volume of 15% (v/v) ACN in water for purification by reversed-phase high performance liquid chromatography (RP-HPLC). The solubilized peptides were purified across a semi-preparative scale reversed-phase C18 column (Grace, 10 µm, 250 x 10 mm) using an Agilent ProStar HPLC system. Peptides were eluted with a linear gradient of 15-

45% solvent B (0.1% TFA in ACN) over solvent A (0.1% TFA in water) over 23 minutes. Absorbance was monitored at 214 and 280 nm to detect the peptide product peaks as they were eluted from the column. All crude peptides were purified over multiple runs and then product peaks were combined, frozen, and lyophilized twice.

Peptides in powder form were stored at -20 ˚C until further use.

Characterization by MS: The identity of each peptide was confirmed by comparing a mass obtained by mass spectrometry to the expected mass, as calculated with an online peptide mass calculator.69 A small quantity of each purified peptide was dissolved in water and the solution was injected directly onto a Thermo Scientific Q

Exactive Plus Hybrid Quadrupole-Orbitrap mass spectrometer set to positive ion mode.

Mass data were processed using MassLynx Software version 4.1 (Waters) and MagTran version 1.03 deconvolution software (Amgen, Thousand Oaks, CA). Representative sample spectra from hGH 36-51 before and after deconvolution can be seen in Figure 2.2.

25

Figure 2.2. Mass spectra of hGH 36-51, raw (top) and deconvoluted (bottom). Expected mass is 2025.26 Da and observed mass is 2026.04 Da.

Characterization by Analytical HPLC: The purities of the peptides were quantified by analytical HPLC. For each peptide, 100 µL of 4 µM peptide in water were injected onto an analytical scale reversed-phase C18 column (Grace, 5 µm, 50 x 2.1 mm) using an Agilent ProStar HPLC system. Peptides were eluted with a linear gradient of 5- 26

95% solvent B over solvent A in 20 minutes. All peptides purities were determined by product peak integration of the chromatograms at 214 nm using OpenLab CDS

ChemStation Software version 1.06 (Agilent) and KaleidaGraph version 4.5 (Synergy

Software). All peptides were purified to >95%. A representative sample HPLC trace from hGH-ScTx and the table used to calculate the percent purity can be seen in Figure 2.3.

Figure 2.3. Analytical HPLC trace from hGH-ScTx. The area under the curve is determined for all peaks and then the purity is expressed as a percent of the area contributed by the peak of interest.

Characterization by CD: The helical propensity of each peptide was determined by wavelength-dependent CD spectropolarimetry. Peptides were diluted to a concentration of 20 µM in phosphate buffered saline (PBS) with or without 30% (v/v)

2,2,2-trifluoroethanol (TFE) and incubated at 25 ˚C for 10 minutes. A Jasco J-715 spectropolarimeter was used to perform wavelength scans from 260 to 190 nm at 25 ˚C. 27

Four scans of each peptide were averaged and background subtracted (buffer only) and the data were processed using J-700 Software version 1.5 (Jasco) and KaleidaGraph version 4.5 (Synergy Software). The percent α-helical nature was calculated from the same data source using online K2D2 software, which compares observed CD data to the spectra of samples with known dominant secondary structures.70

2.3 Biological Assays

Cell culture: SK-MEL-28 and IM9 were grown and maintained in EMEM medium and RPMI-1640 medium, respectively, supplemented with 10% FBS and 1Χ

71 antibiotic-antimycotic. Cells were kept in a humidified incubator at 37 ˚C and 5% CO2.

Media was replaced every 48 hours.

Plating cells: Before treatment with hGH, GHA, or peptides, cells were serum- starved to reduce background hGH signaling. Cells were counted using an automated cell counter (Countess; Thermo Fisher). For SK-MEL-28, cell concentration was adjusted to

105 cells/mL and 2 mL (200,000 cells) were plated per well of a 6-well plate, and the cells were allowed to attach for 16 hours prior to treatment. Media for SK-MEL-28 cells were replaced with serum-free growth media for 2 hours prior to treatment. The IM9 cell cultures were centrifuged and re-suspended in 1% serum growth medium. Cells were counted, distributed to each well of a 6-well plate such that each had 2 mL at 106 cells/mL in 1% serum, and were starved for 16 hours before treatment.

Cell treatment: To the serum-starved cells, a volume of GHA or peptide was added to achieve a concentration of 50 µg/mL or a desired test concentration, 28 respectively. A volume of hGH was then added to achieve a concentration of 50 ng/mL.

Cells were then incubated for 30 minutes.

Protein extraction: At the end of treatment, 6-well plates were placed on ice to slow cell metabolism. For SK-MEL-28 cells, the growth media was removed and cells were washed twice with 1.5 mL chilled, sterile, 1 × PBS, and the cells were lysed with

200 µL lysis buffer (RIPA lysis buffer, 1.5 × Halt protease, and 1 × phosphatase inhibitor cocktail). Cells were incubated for 5 minutes at 4 ˚C and then were harvested by scraping. The lysate was sonicated for 5 minutes and clarified by centrifuging at 8,000 × g for 10 minutes at 4 ˚C and the supernatant was collected and stored at -80 ˚C until further use. For IM9 cells, cells were pelleted at 150 × g for 4 minutes at 4 ˚C and the medium was removed. Cells were washed with chilled 1 × PBS and pelleted again using the same conditions. All PBS was removed and the cells were resuspended in 120 µL lysis buffer. The lysate was sonicated in a water bath for 5 minutes at 4 ˚C and was then clarified by centrifuging at 10,000 × g for 10 minutes at 4 ˚C. The supernatant was collected and stored at -80 ˚C until further use.

Bradford Assay: The protein concentration of the lysate was determined using the absorbance of Bradford reagent at 595 nm, measured using a Spectramax250 spectrometer (Molecular Devices, Sunnyvale, CA) and processed with SoftmaxPro v4.7.1 software. Bovine serum albumin (BSA) at serial dilutions from 1 mg/mL to 0.063 mg/mL was used as a standard.

Western blot: Western blot was performed to visualize the presence of pStat5.

Proteins from the cell lysates were first separated by SDS-PAGE. To accomplish this, lysates were boiled for five minutes in SDS and 4 × loading dye to denature the proteins. 29

They were then loaded onto a 10- or 15-well, polyacrylamide (8%) gel and a voltage of

120 V was applied for 90 minutes or until adequate separation was achieved. Proteins were next transferred to a polyvinylidene fluoride (PVDF) membrane using a current of

300 mA for 2 hours at 4 ˚C. The membrane was blocked with 5% BSA in 1 × Tris buffered saline (pH 7.2) with 0.1% Triton-X100 (TBS-T) at room temperature for 2 hours. The membrane was then incubated with primary antibody overnight (14-16 hour) and washed three times with TBS-T. Lastly, it was incubated with secondary antibody for

1.5 hours at room temperature and washed four times with TBS-T. The membrane was then treated with West Femto Chemiluminescence detection reagents and the signal produced was imaged using a GelDoc (Bio-Rad) reader. Densitometry analysis from measured band intensity from the developed gels was performed using ImageJ software version 1.51s.72

ELISA: The InstantOne ELISA kit was used to quantify pStat5 levels following manufacturer’s instructions. First, 50 µL of sample at a concentration of 1 mg/mL was added to ELISA wells with 50 µL antibody cocktail, and the mixture was incubated at room temperature with 300 rpm shaking for 2 hour. Next, wells were thoroughly washed and stop and detection reagents were added and incubated (room temperature, 300 rpm shaking, 1 hour). Finally absorbance was recorded at 450 nm using a spectrophotometer.

30

3. Results and Discussion

3.1 Peptide Production and Characterization

The full peptide library (Table 3.1) was synthesized, purified, and characterized by MS and analytical HPLC by the author, Zak Hall, and Danushka Arachchige in the

Holub laboratory. All mass spectra can be found in Appendix A and all analytical HPLC traces can be found in Appendix B. The helical propensity of the peptides was determined by CD spectropolarimetry (Table 3.1) by Olivia Kerekes in the Holub laboratory. All CD spectra can be found in Appendix C. The observed masses deviated within a small, acceptable range (1%) of the calculated masses and all purities exceeded

95%. As expected, the peptides had a pronounced α-helical structure (>60%), which increased in the structure-inducing cosolvent, TFE,73 for each peptide.

The control peptide that lacked any substitutions, 36-51, lacks α-helical structure in the absence of TFE (22.51% helical). Several substitutions allowed for greatly increased helical propensity (~60%): K39A, Q41A, Y43A, S44A, F45A, L46A, and

N48A. This implies that the amino acids replaced in those peptides may disrupt helix formation in the control. Additionally, alanine is the residue that confers the highest helical propensity and it is therefore reasonable to expect that the addition of an alanine would contribute to the higher helicity.74 These relationships change in the presence of

TFE. The control 36-51 has a much higher helical propensity, 84.27%, while two of the substitutions that previously increased helicity, Q41A and F45A, decrease helicity from the control.

The contribution of the residue to the helical nature does not seem to depend upon the nature of the amino acid that was substituted. Nonpolar F45A and L46A substitutions 31 did not follow the same trend and neither the pair of glutamines nor the pair of lysines followed the same pattern. Some substitutions followed expected trends in helical propensity. For example, tyrosine has a much lower helical propensity than alanine74 and the removal of tyrosine in Y43A increased helicity without TFE and maintained similar levels of helicity with TFE. However, others had the opposite trend. Because leucine is close to alanine in helical propensity, it would be expected that they would be somewhat interchangeable in that aspect, but the L46A substitution increased helicity in the absence of TFE. This would suggest some other interactions are affecting the helicity of these peptides.

Table 3.1. Sequences, masses, and helical propensities of the hGH mimetic scanning alanine library. Mass (Da) Percent α-Helix Peptide Sequence Calculated Observed - TFE + TFE 36-51 YIPKEQKYSFLQNPQT 2025.26 2026.04 22.51 84.27 K39A YIPAEQKYSFLQNPQT 1968.99 1968.99 61.33 84.27 E40A YIPKAQKYSFLQNPQT 1967.23 1967.03 32.63 67.45 Q41A YIPKEAKYSFLQNPQT 1968.22 1968.02 59.98 61.33 K42A YIPKEQAYSFLQNPQT 1967.18 1967.98 32.77 60.86 Y43A YIPKEQKASFLQNPQT 1933.17 1933.02 60.44 84.27 S44A YIPKEQKYAFLQNPQT 2010.50 2010.10 60.86 84.27 F45A YIPKEQKYSALQNPQT 1950.15 1950.00 60.44 60.86 L46A YIPKEQKYSFAQNPQT 1983.19 1983.00 60.44 84.27 Q47A YIPKEQKYSFLANPQT 1968.22 1968.02 59.98 84.27 N48A YIPKEQKYSFLQAPQT 1983.10 1983.00 32.63 58.24

32

3.2 Biological Assays

The biological activity of the peptides was tested by treating IM9 cells with specific amounts of hGH, pegvisomant, and the prepared peptides. The ability of hGH

36-51 to inhibit hGH function was measured first and the results (Figure 3.1) supported those from previous SK-MEL-28 cell treatments that showed that the peptide had a dose- dependent, antagonistic effect. Next, the alanine scanning library was tested to determine how the alanine substitutions affected antagonist ability of the peptides (Figure 3.2).

While the image may suggest certain patterns, the lack of antagonist activity from pegviosmant (GHA) and hGH 36-51 and the substantial reduction in signal in the last lane indicate that further data collection was needed.

Problems in image quality may have arisen from a number of different sources.

Inconsistent addition of hGH or peptides during treatment could account for some of the variability. Incorrect addition of peptide for the 36-51 treatment could account for the lack of inhibition seen in Figure 3.2 as compared to Figure 3.1 and previous results.

Additionally, problems may have been introduced during the western blot stage of the process. Incorrect loading of the lysates could account for the much lighter last lane in

Figure 3.2. Subsequent to this image, 10-well gels were used instead of 15-well gels in order to facilitate lane loading.

33

600

500

400

300

200

100 % change from control

0 Control +GH + GHA + GH +GH + 25 + GH + 250 nM 36-51 nM 36-51

Figure 3.1 Western blot (top) to test antagonism of hGH 36-51 in IM9 cells. Graph (bottom) shows the percent change from baseline in Stat5 phosphorylation, normalized to protein loading levels as determined by actin band intensity.

34

500 450 400 350 300 250 200 150

% change from control 100 50 0

Figure 3.2. Western blot (top) to test antagonism of the peptide library in IM9 cells. Graph (bottom) shows the percent change from baseline in Stat5 phosphorylation, normalized to protein loading levels as determined by actin band intensity.

35

After these images were obtained, attempts were made to repeat the experiments in order to make more firm conclusions. However, the protein concentrations of the three subsequent treatments were too low to properly load on the SDS-PAGE gels to run the western blots. Concentrations were generally under 1 mg/mL, which meant that the volume required to load an appropriate amount of protein far exceeded the volume of the loading wells. In order to address this problem, different concentrations of cells were tested with and without the presence of hGH to determine how many cells were needed per 2 mL treatment well in order to collect enough protein (Figure 3.3). A concentration of 2×106 cells/mL was chosen to continue.

Figure 3.3 Higher concentrations of cells during treatment yielded higher concentrations of protein after cell lysis.

36

Treatment of the cells at this concentration with the peptides did yield a high enough protein concentration to continue with the western blot. However, the image obtained was not of sufficient quality to draw meaningful conclusions (Figure 3.4). It was hypothesized that the problem may have arisen because of problems with the polymerization of the gel so the same cell lysates were run again on a pre-made 4-20% gradient polyacrylamide gel, but it yielded a similar image (data not shown). Having eliminated the gel as the potential source of the problem, it was next hypothesized that the amount of lysis buffer used was insufficient for the large number of cells, causing incomplete isolation of the proteins from other cellular material and thus the unclear bands. Stat5 is a transcription factor, which means it could be associating with DNA and contaminating the protein, causing the streaking seen in Figure 3.4. To compensate for this, the concentration of cells was reduced to 5×105 cells/mL while the amount of lysis buffer remained constant. The treatment run with these conditions did yield enough protein to continue with the western blot. However, while the actin band separated as expected, the pStat5 band was faint where expected and much more intense at the top of the gel (Figure 3.5). Stat5 can form higher order oligomers such as tetramers,75 and this oligomerization may account for why pStat5 did not migrate through the gel as expected.

In order to obtain results while working to solve these problems, ELISA was applied as an alternate method of visualizing the phosphorylation of Stat5.

37

Figure 3.4 Western blot image obtained after treatment at 2×106 cells/mL.

Figure 3.5 Western blot image obtained after treatment at 5×105 cells/mL.

38

ELISA was performed by Dr. Reetobrata Basu using SK-MEL-28 cell treatments.

Two tests were run in duplicate for a total of four treatments to assess antagonist action of hGH 36-51 (Figure 3.6) and the full peptide library (Figure 3.7). These results support the hypothesis that hGH 36-51 antagonizes hGH action and provide a clearer picture of which alanine substitutions affected the function of the hGH 36-51 peptide. With this information, comparisons can be made between the residues hypothesized to affect function because of their contact with the GHR, how alanine substitutions at each residue affected the biological activity, and how biological activity is affected by the helical propensity (Table 3.2).

Figure 3.6. Change in Stat5 phosphorylation from baseline based on ELISA in SK-MEL- 28 cells for hGH 36-51.

39

Figure 3.7. Change in Stat5 phosphorylation from baseline based on ELISA in SK-MEL- 28 cells for full peptide library. Peptides that failed to inhibit Stat5 are indicated with arrows.

Table 3.4. The interaction between which residues contacted GHR, biological activity, and helical propensity. Peptide Contact with GHR pStat5 Inhibition Percent α-Helix 36-51 N/A Yes 84.27 K39A Yes Yes 84.27 E40A No No 67.45 Q41A No No 61.33 K42A Yes No 60.86 Y43A Yes Yes 84.27 S44A No Yes 84.27 F45A No No 60.86 L46A Yes Yes 84.27 Q47A Yes No 84.27 N48A No Yes 58.24

40

While the biological activity matched the hypothesis for only half the residues tested, the helical nature of the peptides correlated to inhibition of Stat5 phosphorylation in 9 of the 11 peptides. Residues E40, Q41, K42, F45, and Q47 were found to be critical for Stat5 phosphorylation inhibition. Of these residues, all but Q47 were also less able to form a helix. In contrast, out of the five, only K42A and Q47A were predicted to contact the GHR and thus be important in binding and inhibition. These results may suggest that the ability to form an α-helix is more critical to the biological function of these peptide mimetics than the presence of individual residues that interact with the GHR in the native hGH.

41

4. Conclusions

The purpose of this thesis project was to develop a library of synthetic peptides mimetics of hGH and to test their function as antagonists of the GHR, with a focus on how individual residues affect inhibitory activity. This library was successfully synthesized and purified, and progress was made on assessing biological activity. Despite these advancements, not all experiments were successful, but there are many more questions that could potentially be explored using these materials and methods.

To build on these preliminary results, further repetitions of the western blots and

ELISAs discussed here will be needed. Additionally, repeating these tests in a number of different cell lines will help to establish the consistency and validity. While we hypothesize that the inhibition of Stat5 phosphorylation exhibited by these peptides is mediated by antagonism of the GHR, it is possible that they could be exerting their effect through other pathways76 and testing their ability to reduce non-hGH mediated phosphorylation of Stat5 will be important. Further evidence of their mechanism of action can be provided by completing direct binding studies in vitro. Fluorescently- labeled versions of the peptides will be tested against the recombinantly expressed extracellular domain of GHR and direct binding can be measured by fluorescence polarization.77

In addition to continuing studies with the peptides outlined here, further modifications could be made to investigate other hypotheses. Because helical propensity seems to have such an impact on activity, different methods for stabilizing that secondary structure could be explored. Peptide stapling could lock the peptide in the helical structure by replacing some of the residues with non-natural amino acids and performing 42 a chemical reaction to cyclize the peptide.52,53 Further experimentation with the ScTx- based mimetics also provides different opportunities. The scaffold structure could help to stabilize the helix60 so that the effect of individual residues can be better explored without the confounding variable of structure and varying the number of disulfide bonds in the

ScTx backbone could be used to explore the effect of intrinsic peptide disorder on binding.43

In addition to better understanding the binding dynamics between hGH and GHR, these peptides have potential applications as improved GHR antagonists. Drug development is an involved process and many routes are possible for further refinement of these peptides to be the most therapeutically viable.48 Peptides with the modifications discussed above have the potential to improve upon the efficacy or the bioavailability of the peptides already synthesized. It also may be possible to utilize smaller helices by testing how many of the 16 residues used in these peptides are needed for inhibition to occur. These site 1 mimics could also be paired with or linked to a site 2 mimic designed in a similar way to further increase antagonism. Therapeutic potential can eventually be tested in mouse models, either through transgenic expression of the peptides or another more traditional delivery method.78 An improved drug to treat disorders of excess hGH generated from the peptides tested here has the potential to increase quality of life for people with a wide variety of conditions, though much more development and experimentation is needed.

43

References

1. Cunningham, B. C. et al. Dimerization of the Extracellular Domain of the Human

Growth Hormone Receptor by a Single Hormone Molecule. Science. 254, 821–825

(1991).

2. De Vos, A. M., Ultsch, M., & Kossiakoff, A. A. Human growth hormone and

extracellular domain of its receptor: crystal structure of the complex. Science. 255,

306–312 (1992).

3. Waters, M. J. The growth hormone receptor. Growth Horm. IGF Res. 28, 6–10

(2016).

4. Demchuk, E., Mueller, T., Oschkinat, H., Sebald, W. & Wade, R. C. Receptor

binding properties of four-helix-bundle growth factors deduced from electrostatic

analysis. Protein Sci. 3, 920–935 (1994).

5. Waters, M. J. & Brooks, A. J. JAK2 activation by growth hormone and other

cytokines. Biochem. J. 466, 1–11 (2015).

6. Dehkhoda, F., Lee, C. M. M., Medina, J., Brooks, A. J. & Brooks, A. J. The

Growth Hormone Receptor: Mechanism of Receptor Activation , Cell Signaling ,

and Physiological Aspects. Front. Endocrinol. (Lausanne). 9, 1–23 (2018).

7. Hayashi, M. et al. GH directly stimulates UCP3 expression. Growth Hormone and

IGF Research (2018). doi:10.1016/j.ghir.2018.01.002

8. Carter-su, C., Schwartz, J. & Argetsinger, L. S. Growth hormone signaling

pathways. Growth Horm. IGF Res. 28, 11–15 (2016).

9. Le Roith, D., Bondy, C., Yakar, S., Liu, J.-L. & Butler, A. The Somatomedin

Hypothesis: 2001. Endocr. Rev. 22, 53–74 (2001). 44

10. Cotes, P. M., Red, E. & Young, F. G. Diabetogenic action of pure anterior

pituitary growth hormone. Nature 164, 209–211 (1949).

11. Holly, J. M. P., Amiel, S. A., Sandhu, R. R., Rees, L. H. & Wass, J. A. H. The role

of growth hormone in diabetes mellitus. J. Endocrinol. 118, 353–364 (1988).

12. Melmed, S. Acromegaly pathogenesis and treatment. J Clin Invest 119, 3189–3202

(2009).

13. Hannah-Shmouni, F., Trivellin, G. & Stratakis, C. A. Genetics of gigantism and

acromegaly. Growth Horm. IGF Res. 30–31, 37–41 (2016).

14. Kim, S.-H. & Park, M.-J. Effects of Growth Hormone on Glucose Metabolism and

Insulin Resistance in Human. Ann. Pediatr. Endocrinol. Metab. 22, 145–152

(2017).

15. List, E. O. et al. Endocrine parameters and phenotypes of the growth hormone

receptor gene disrupted (GHR-/-) mouse. Endocr. Rev. 32, 356–386 (2011).

16. Banerjee, I. & Clayton, P. E. Growth Hormone Treatment and Cancer Risk.

Endocrinol. Metab. Clin. North Am. 36, 247–263 (2007).

17. Yakar, S., LeRoith, D. & Brodt, P. The role of the growth hormone/insulin-like

growth factor axis in tumor growth and progression: Lessons from animal models.

Cytokine Growth Factor Rev. 16, 407–420 (2005).

18. Mak, R. H., Cheung, W. W. & Roberts, C. T. The growth hormone-insulin-like

growth factor-I axis in chronic kidney disease. Growth Horm. IGF Res. 18, 17–25

(2008).

19. Mak, R. H. & Cheung, W. W. Transforming growth factors and insulin-like

growth factors in chronic kidney disease. J. Organ Dysfunct. 5, 59–64 (2009). 45

20. Hallengren, E. et al. Fasting levels of high-sensitivity growth hormone predict

cardiovascular morbidity and mortality: The malmö diet and cancer study. J. Am.

Coll. Cardiol. 64, 1452–1460 (2014).

21. Isgaard, J., Arcopinto, M., Karason, K. & Cittadini, A. GH and the cardiovascular

system: an update on a topic at heart. Endocrine 48, 25–35 (2015).

22. Gesing, A. et al. A Long-lived Mouse Lacking Both Growth Hormone and Growth

Hormone Receptor: A New Animal Model for Aging Studies. Journals Gerontol.

Ser. A Biol. Sci. Med. Sci. 72, glw193 (2016).

23. Van Der Lely, A. J. & Kopchick, J. J. Growth hormone receptor antagonists. in

Neuroendocrinology 83, 264–268 (2006).

24. Sandret, L., Maison, P. & Chanson, P. Place of cabergoline in acromegaly: A

meta-analysis. J. Clin. Endocrinol. Metab. 96, 1327–1335 (2011).

25. Shen, L. P., Pictet, R. L. & Rutter, W. J. Human somatostatin I: sequence of the

cDNA. Proc. Natl. Acad. Sci. U. S. A. 79, 4575–4579 (1982).

26. Murray, R. D. & Melmed, S. A critical analysis of clinically available somatostatin

analog formulations for therapy of acromegaly. J. Clin. Endocrinol. Metab. 93,

2957–2968 (2008).

27. Venegas-Moreno, E. et al. Association between dopamine and somatostatin

receptor expression and pharmacological response to somatostatin analogues in

acromegaly. J. Cell. Mol. Med. 22, 1640–1649 (2018).

28. Kopchick, J. J. Lessons Learned from Studies with the Growth Hormone Receptor.

Growth Horm. IGF Res. (2015). doi:10.1016/j.ghir.2015.06.003

29. FDA. Drug Approval Package: Somavert. U.S. Department of Health and Human 46

Services Available at:

https://www.accessdata.fda.gov/drugsatfda_docs/nda/2003/021106_somavert.cfm.

30. Chen, W. Y., Wight, D. C., Wagner, T. E. & Kopchick, J. J. Expression of a

mutated bovine growth hormone gene suppresses growth of transgenic mice. Proc.

Natl. Acad. Sci. U. S. A. 87, 5061–5 (1990).

31. Chen, W. Y., Wight, D. C., Mehta, B. V, Wagner, T. E. & Kopchick, J. J. Glycine

119 of bovine growth hormone is critical for growth-promoting activity. Mol.

Endocrinol. 5, 1845–1852 (1991).

32. Chen, W. Y., Chen, N., Yun, J., Wagner, T. E. & Kopchick, J. J. In vitro and in

vivo studies of the antagonistic effects of human growth hormone analogs. J. Biol.

Chem. 269, 15892–15897 (1994).

33. Pfister, D. & Morbidelli, M. Process for protein PEGylation. J. Control. Release

180, 134–149 (2014).

34. Ross, R. J. M. et al. Binding and functional studies with the growth hormone

receptor antagonist, B2036-PEG (pegvisomant), reveal effects of pegylation and

evidence that it binds to a receptor dimer. J. Clin. Endocrinol. Metab. 86, 1716–

1723 (2001).

35. Kopchick, J. J., List, E. O., Kelder, B., Gosney, E. S. & Berryman, D. E.

Evaluation of growth hormone (GH) action in mice: Discovery of GH receptor

antagonists and clinical indications. Mol. Cell. Endocrinol. 386, 34–45 (2014).

36. Cunningham, B. C. & Wells, J. a. High-resolution epitope mapping of hGH-

receptor interactions by alanine-scanning mutagenesis. Science 244, 1081–1085

(1989). 47

37. Goffin, V. et al. The human growth hormone antagonist B2036 does not interact

with the prolactin receptor. Endocrinology 140, 3853–3856 (1999).

38. van der Lely, A. J. et al. Long-term treatment of acromegaly with pegvisomant, a

growth hormone receptor antagonist. Lancet (London, England) 358, 1754–9

(2001).

39. Trainer, P. J. ACROSTUDY: The first 5 years. Eur. J. Endocrinol. 161, (2009).

40. Barkan, A. L. et al. Glucose homeostasis and safety in patients with acromegaly

converted from long-acting octreotide to pegvisomant. J. Clin. Endocrinol. Metab.

90, 5684–5691 (2005).

41. European Medicines Agency. Somavert: Scientific Discussion. EMEA 1–26

(2004). Available at:

http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-

_Scientific_Discussion/human/000409/WC500054625.pdf.

42. Feenstra, J. et al. Combined therapy with somatostatin analogues and weekly

pegvisomant in active acromegaly. Lancet 365, 1644–1646 (2005).

43. Arachchige, D., Margaret Harris, M., Coon, Z., Carlsen, J. & Holub, J. M. Role of

single disulfide linkages in the folding and activity of scyllatoxin-based BH3

domain mimetics. J. Pept. Sci. 23, 367–373 (2017).

44. Harris, M. M., Coon, Z., Alqaeisoom, N., Swords, B. & Holub, J. M. Targeting

anti-apoptotic Bcl2 proteins with scyllatoxin-based BH3 domain mimetics. Org.

Biomol. Chem. 14, 440–446 (2016).

45. Craik, D. J., Fairlie, D. P., Liras, S. & Price, D. The Future of Peptide-based

Drugs. Chem. Biol. Drug Des. 81, 136–147 (2013). 48

46. Vlieghe, P., Lisowski, V., Martinez, J. & Khrestchatisky, M. Synthetic therapeutic

peptides: science and market. Drug Discov. Today 15, 40–56 (2010).

47. Tsou, L. K., Cheng, Y. & Cheng, Y. C. Therapeutic development in targeting

protein-protein interactions with synthetic topological mimetics. Current Opinion

in Pharmacology 12, 403–407 (2012).

48. McGregor, D. P. Discovering and improving novel peptide therapeutics. Current

Opinion in Pharmacology 8, 616–619 (2008).

49. Stewart, K. M., Horton, K. L. & Kelley, S. O. Cell-penetrating peptides as delivery

vehicles for biology and medicine. Org. Biomol. Chem. 6, 2242 (2008).

50. Holub, J. M. Small Scaffolds, Big Potential: Developing Miniature Proteins as

Therapeutic Agents. Drug Dev. Res. 78, 268–282 (2017).

51. Werle, M. & Bernkop-Schnürch, A. Strategies to improve plasma half life time of

peptide and protein drugs. Amino Acids 30, 351–367 (2006).

52. Marastoni, M. et al. Synthesis and activity of new linear and cyclic peptide T

derivatives. Arzneimittelforschung. 44, 1073–1076 (1994).

53. Ovadia, O. et al. The effect of backbone cyclization on PK/PD properties of

bioactive peptide-peptoid hybrids: The melanocortin agonist paradigm. Bioorganic

Med. Chem. 18, 580–589 (2010).

54. Agersø, H. et al. and renal excretion of desmopressin after

intravenous administration to healthy subjects and renally impaired patients. Br. J.

Clin. Pharmacol. 58, 352–358 (2004).

55. Harris, a G. Somatostatin and somatostatin analogues: pharmacokinetics and

pharmacodynamic effects. Gut 35, S1–S4 (1994). 49

56. Brinckerhoff, L. H. et al. Terminal modifications inhibit proteolytic degradation of

an immunogenic MART-127-35 peptide: Implications for peptide vaccines. Int. J.

Cancer 83, 326–334 (1999).

57. Lee, S. H. et al. Synthesis, characterization, and pharmacokinetic studies of

PEGylated glucagon-like peptide-1. Bioconjug. Chem. 16, 377–382 (2005).

58. Lipinski, C. A. Rule of five in 2015 and beyond: Target and ligand structural

limitations, ligand chemistry structure and drug discovery project decisions.

Advanced Drug Delivery Reviews 101, 34–41 (2016).

59. Lau, J. L. & Dunn, M. K. Therapeutic peptides: Historical perspectives, current

development trends, and future directions. Bioorganic Med. Chem. (2017).

doi:10.1016/j.bmc.2017.06.052

60. Harris, M. M., Coon, Z., Alqaeisoom, N., Swords, B. & Holub, J. M. Targeting

anti-apoptotic Bcl2 proteins with scyllatoxin-based BH3 domain mimetics. Org.

Biomol. Chem. 14, 440–446 (2016).

61. Martins, J. C., Van De Ven, F. J. M. & Borremans, F. A. M. Determination of the

three-dimensional solution structure of scyllatoxin by 1H nuclear magnetic

resonance. J. Mol. Biol. 253, 590–603 (1995).

62. Morrison, K. L. & Weiss, G. A. Combinatorial alanine-scanning. Curr. Opin.

Chem. Biol. 5, 302–307 (2001).

63. Pires, D. A. T., Bemquerer, M. P. & Do Nascimento, C. J. Some mechanistic

aspects on Fmoc solid phase peptide synthesis. Int. J. Pept. Res. Ther. 20, 53–69

(2014).

64. Wang, W. et al. Integrated SPPS on continuous-flow radial microfluidic chip. Lab 50

Chip 11, 929 (2011).

65. Fields, G. B. & Noble, R. L. Solid phase peptide synthesis utilizing 9-

fluorenylmethoxycarbonyl amino acids. Int. J. Pept. Protein Res. 35, 161–214

(1990).

66. Chang, C. -D & Meienhofer, J. Solid-phase peptide synthesis using mild base

cleavage of Nα-fluorenylmethyloxycarbonylamino acids, exemplified by a

synthesis of dihydrosomatostatin. Int. J. Pept. Protein Res. 11, 246–249 (1978).

67. Palasek, S. A., Cox, Z. J. & Collins, J. M. Limiting racemization and aspartimide

formation in microwave-enhanced Fmoc solid phase peptide synthesis. J. Pept.

Sci. 13, 143–148 (2007).

68. Jenssen, H. & Aspmo, S. I. Serum Stability of Peptides. in Peptide-Based Drug

Design (ed. Otvos, L.) 177–186 (Humana Press, 2008). doi:10.1007/978-1-59745-

419-3

69. Rozenski, J. Peptide Mass Calculator v3.2. (1999).

70. Perez-Iratxeta, C. & Andrade-Navarro, M. A. K2D2: Estimation of protein

secondary structure from circular dichroism spectra. BMC Struct. Biol. 8, (2008).

71. Basu, R., Wu, S. & Kopchick, J. J. Targeting growth hormone receptor in human

melanoma cells attenuates tumor progression and epithelial mesenchymal

transition via suppression of multiple oncogenic pathways. Oncotarget 8, 21579–

21598 (2017).

72. Schindelin, J., Rueden, C. T., Hiner, M. C. & Eliceiri, K. W. The ImageJ

ecosystem: An open platform for biomedical image analysis. Molecular

Reproduction and Development 82, 518–529 (2015). 51

73. Roccatano, D., Colombo, G., Fioroni, M. & Mark, A. E. Mechanism by which

2,2,2-trifluoroethanol/water mixtures stabilize secondary-structure formation in

peptides: A molecular dynamics study. Proc. Natl. Acad. Sci. 99, 12179–12184

(2002).

74. Pace, C. N. & Scholtz, J. M. A helix propensity scale based on experimental

studies of peptides and proteins. Biophys. J. 75, 422–427 (1998).

75. Lin, J. X. et al. Critical Role of STAT5 Transcription Factor Tetramerization for

Cytokine Responses and Normal Immune Function. Immunity 36, 586–599 (2012).

76. Bartalucci, N. et al. Inhibitors of the PI3K/mTOR pathway prevent STAT5

phosphorylation in <i>JAK2V617F</i> mutated cells through PP2A/

CIP2A axis. Oncotarget (2017). doi:10.18632/oncotarget.18073

77. Stoddart, L. A., White, C. W., Nguyen, K., Hill, S. J. & Pfleger, K. D. G.

Fluorescence- and bioluminescence-based approaches to study GPCR ligand

binding. Br. J. Pharmacol. 173, 3028–3037 (2016).

78. Chen, W. Y., Wight, D. C., Wagner, T. E. & Kopchick, J. J. Expression of a

mutated bovine growth hormone gene suppresses growth of transgenic mice.

Proc.Natl.Acad.Sci.USA 87, 5061–5065 (1990).

52

Appendices

Appendix A: Mass Spectra

Figure A1. Raw mass spectrum of hGH 36-51.

Figure A2. Raw mass spectrum of K39A. 53

Figure A3. Raw mass spectrum of E40A.

Figure A4. Raw mass spectrum of Q41A. 54

Figure A5. Raw mass spectrum of K42A.

Figure A6. Raw mass spectrum of Y43A. 55

Figure A7. Raw mass spectrum of S44A.

Figure A8. Raw mass spectrum of F45A. 56

Figure A9. Raw mass spectrum of L46A.

Figure A10. Raw mass spectrum of Q47A. 57

Figure A11. Raw mass spectrum of N48A.

Appendix B: Analytical HPLC Spectra

Figure B1. Analytical HPLC trace of hGH 36-51 (top) with accompanying table of the area-under-the-curve analysis (bottom).

58

Figure B2. Analytical HPLC trace of K39A (top) with accompanying table of the area- under-the-curve analysis (bottom).

Figure B3. Analytical HPLC trace of E40A (top) with accompanying table of the area- under-the-curve analysis (bottom).

59

Figure B4. Analytical HPLC trace of Q41A (top) with accompanying table of the area- under-the-curve analysis (bottom).

Figure B5. Analytical HPLC trace of K42A (top) with accompanying table of the area- under-the-curve analysis (bottom). 60

Figure B6. Analytical HPLC trace of Y43A (top) with accompanying table of the area- under-the-curve analysis (bottom).

Figure B7. Analytical HPLC trace of S44A (top) with accompanying table of the area- under-the-curve analysis (bottom).

61

Figure B8. Analytical HPLC trace of F45A (top) with accompanying table of the area- under-the-curve analysis (bottom).

Figure B9. Analytical HPLC trace of L46A (top) with accompanying table of the area- under-the-curve analysis (bottom).

62

Figure B10. Analytical HPLC trace of Q47A (top) with accompanying table of the area- under-the-curve analysis (bottom).

Figure B11. Analytical HPLC trace of N48A (top) with accompanying table of the area- under-the-curve analysis (bottom).

63

Appendix C: Circular Dichroism Spectra

For all CD spectra, red data was taken without the structure-inducing cosolvent, TFE, and all blue data was taken in the presence of TFE.

Figure C.1 CD spectrum of hGH 36-51. Figure C2. CD spectrum of K39A.

Figure C3. CD spectrum of E40A. Figure C4. CD spectrum of Q41A. 64

Figure C5. CD spectrum of K42A. Figure C6. CD spectrum of Y43A.

Figure C7. CD spectrum of S44A. Figure C8. CD spectrum of F45A. 65

Figure C9. CD spectrum of L46A. Figure C10. CD spectrum of Q47A.

Figure C11. CD spectrum of N48A.