RESEARCH ARTICLE

Structure and function of a dual antagonist of the human

growth hormone and prolactin receptors with site-specific
PEG conjugates
Received for publication, November 22, 2022, and in revised form, June 6, 2023 Published, Papers in Press, July 11, 2023,
https://doi.org/10.1016/j.jbc.2023.105030
Reetobrata Basu1,‡, Rich Brody2,‡, Uday Sandbhor2,‡, Prateek Kulkarni1,3,4, Emily Davis1,3,4, Deborah Swegan1,4,
Lydia J. Caggiano1,5, Edward Brenya1,4 , Sebastian Neggers6, and John J. Kopchick1,3,7, *
From the 1Edison Biotechnology Institute, Ohio University, Athens, Ohio, USA; 2Infinix Bio LLC, Columbus, Ohio, USA; 3Molecular
and Cellular Biology Program, Ohio University, Athens, Ohio, USA; 4Department of Biological Sciences, Ohio University, Athens,
Ohio, USA; 5Honors Tutorial College, Ohio University, Athens, Ohio, USA; 6Department of Medicine, Endocrinology, Erasmus
Medical Centre, Rotterdam, Netherlands; 7Heritage College of Osteopathic Medicine, Ohio University, Athens, Ohio, USA
Reviewed by members of the JBC Editorial Board. Edited by Henrik Dohlman

Human growth hormone (hGH) is a pituitary-derived growth, organ development, directing the hepatic production
endocrine protein that regulates several critical postnatal of >75% of the circulating insulin-like growth factor 1 (IGF1),
physiologic processes including growth, organ development, body composition, and metabolic homeostasis (anabolic to
and metabolism. Following adulthood, GH is also a regulator of proteins and carbohydrates, catabolic to lipids) (2). Patholog-
multiple pathologies like fibrosis, cancer, and diabetes. ically, an excess of serum hGH due to a hypersecreting pitui-
Therefore, there is a significant pharmaceutical interest in tary adenoma causes acromegaly (3), while a subnormal
developing antagonists of hGH action. Currently, there is a production of pituitary hGH causes growth hormone (GH)
single FDA-approved antagonist of the hGH receptor (hGHR) deficiency and inactivating mutations of the growth hormone
prescribed for treating patients with acromegaly and discov- receptor (GHR) leads to the states of GHR resistance/insen-
ered in our laboratory almost 3 decades ago. Here, we present sitivity termed Laron syndrome (LS) (4).
the first data on the structure and function of a new set of In a normal state, there is a progressive decline in the pituitary
protein antagonists with the full range of hGH actions—dual output of hGH following adulthood, termed ‘somatopause’ (5)
antagonists of hGH binding to the GHR as well as that of hGH while, interestingly, the local, extra-pituitary production of GH
binding to the prolactin receptor. We describe the site-specific with autocrine/paracrine action in specific tissues continues
PEG conjugation, purification, and subsequent characteriza- throughout life (6) and often increases with age (7). Importantly,
tion using MALDI-TOF, size-exclusion chromatography, congenital suppression of GH action has been found to be
thermostability, and biochemical activity in terms of ELISA- protective against incidence of several age-associated comor-
based binding affinities with GHR and prolactin receptor. bidities such as cancer, diabetes, cardiomyopathy, glomerulo-
Moreover, these novel hGHR antagonists display distinct nephritis, and cognitive decline in several mouse models (Ames,
antagonism of GH-induced GHR intracellular signaling in vitro Snell, lit/lit, GHR−/−, GH−/−, and more) (8, 9). Remarkably, in
and marked reduction in hepatic insulin-like growth factor 1 Israel, a relatively large cohort of individuals with LS, studied for
output in vivo. Lastly, we observed potent anticancer biological over 50 years, presented zero cases of cancer (10, 11), while
efficacies of these novel hGHR antagonists against human another large Ecuadorian cohort of patients with LS studied for
cancer cell lines. In conclusion, we propose that these new over 30 years have also presented no cases of malignancy and are
GHR antagonists have potential for development towards resistant to diabetes (12). Moreover, adult-onset suppression of
multiple clinical applications related to GH-associated GH action in mouse models (1.5mGHRKO, 6mGHRKO) has
pathologies. been recently reported to recapitulate several of the aforemen-
tioned health benefits (13). Therefore, GH occupies a central
position in the research on lifespan and healthspan, resulting in
Human growth hormone (hGH) is an anterior pituitary- a significant interest in the development of pharmaceutical in-
derived endocrine protein that exerts a range of physiolog- hibitors of the GH–GHR interaction and subsequent intracel-
ical effects in the body, following binding to and activation of lular signaling (14).
its cognate pre-formed hGH receptor (hGHR) homodimer The GH–GHR binding is asymmetric (1:2 stoichiometry),
expressed differentially across a wide range of cellular types in wherein one GH molecule binds to a preformed GHR homo-
different tissues (1). The major processes regulated by the dimer. Accordingly, the GH molecule has two distinct sites of
hGH–hGHR interaction includes driving postnatal somatic interaction with the receptor—a high affinity–binding site 1
comprising amino acids in helices I and IV and a low affinity–
‡
These authors contributed equally to this work.
binding site 2 comprising amino acids in helix III (15).
* For correspondence: John J. Kopchick, [email protected]. Currently, the most successful way of interfering with the GH–

J. Biol. Chem. (2023) 299(8) 105030 1

© 2023 THE AUTHORS. Published by Elsevier Inc on behalf of American Society for Biochemistry and Molecular Biology. This is an open access article under the CC
BY license (http://creativecommons.org/licenses/by/4.0/).
New PEGylated antagonist of GH action
GHR interaction is an hGHR antagonist, Somavert (pegvisom- affinity of the hGHR antagonist to the hGHR to less than 5% of
ant for injection; Pfizer Inc), which was discovered in our lab- the original ligand hGH. Thereafter, a number of subsequent
oratory in the early 1990s and is currently an Food and Drug efforts at developing improved hGHR antagonists have
Administration (FDA)-approved treatment for patients with focused on improving the pegylation pattern in the original
acromegaly (16). The peptide core of the pegvisomant molecule, peptide core—B2036—of the pegvisomant molecule. In this
termed B2036, is the 191 amino acid hGH molecule with nine regard, Perry and Maynard have generated three successive
different amino acid substitutions of which one (G120K) makes new designs of the modified GH molecule. First, the group
it an hGHR antagonist by disrupting proper site 2 binding, while generated an hGH-G120R or B2036 with N-terminal thio-
the other eight substitutions (H18D, H21N, R167N, K168A, redoxin fusion and conjugated to multiple amine-reactive 5-
D171S, K172R, E174S, I179T) impart improved affinity for site 1 kDa methoxy PEG succinimidyl propionate, which had a
binding and increased specificity for the hGHR dimer (16–18). >15-h circulating half-life in mice and improved in vitro
Importantly, the human GH (but no other species GH) is known bioactivity than pegvisomant, albeit a significant end product
to bind to and activate the prolactin receptor (PRLR) where it heterogeneity (28). The two other antagonist designs from this
exerts a lactogenic effect (19–21). The G120K-modified GH group involved site-specific pegylations: (i) replacing tyrosine-
(hGH-G120K) also binds to the PRLR, unlike B2036 or pegvi- 35 with an unnatural amino acid propargyl tyrosine to which 5,
somant which binds exclusively to the hGHR (22). 10, or 20 kDa azide-containing PEGs were attached by Cu-
The versatile and pleiotropic role of hGH in normal growth catalyzed click reactions (29) and (ii) replacing serine-144 by
and development in the early stage of life and in promoting cysteine to which 20, 30, or 40 kDa methoxy-PEG maleimide
multiple pathophysiology in mature adult life makes it an were attached (30). Both modifications improved serum half-
attractive pharmaceutical candidate (1, 23). The bulk of the ef- life and bioactivity of the core B2036 molecule. Other at-
forts at synthesizing an active GH molecule and extending its tempts of an N-terminal site-specific pegylation of hGH-
short circulating half-life of
20-min involved various strategies G120R remained unsuccessful, while a dimeric hGH-G120R
of pegylation to treat congenital deficiencies in GH production generated by N-term to C-terminal fusion yielded an agonist
(24). In these cases, site-specific pegylations leading to increased in place of an antagonist (31). The only other study which
serum half-lives of hGH were performed by Cox et al generating performed successful N-terminal mono-pegylation of B2036
T3C and S144C variants of the hGH (https://pubchem.ncbi. with either a 20-kDa or a 40-kDa PEG reported complete loss
nlm.nih.gov/patent/NZ-513077-A) and by Cho et al. (25) of antagonist activity by the larger PEG, likely from steric
generating 20 different hGH variants pegylated site specifically hindrance in binding to the hGHR (32).
at six different conjugation sites (Y35, F92, Q131, R134, Y143, Although promising as hGHR antagonists, none of these
K143). As of now, there has been a major advancement in long- molecules are known to inhibit the binding of the GH to the
acting forms of hGH with several new approaches which have PRLR with an affinity similar to that of prolactin. Therefore, it
won regulatory board approvals recently. The notable candi- is apparent that there is a distinct viable space for the devel-
dates which have achieved an efficacious once/week adminis- opment of a dual antagonist of the hGH–GHR and hGH–
tration include but are not limited to lonapegsomatropin PRLR interactions by harnessing the ability of the G120K-
(unmodified hGH bound to a PEG carrier molecule with a pH hGH molecule to bind to and inhibit hGH binding to both
and temperature-dependent cleavable linker; Ascendis; of these receptors, unlike the GHR-specific B2036 molecule.
approved by FDA in 2021), somapacitan (modified GH non- Using site-specific pegylations (33, 34), there is the opportunity
covalently attached to albumin; Novo Nordisk; approved by to add unique large or charged PEGs to G120K-hGH to
FDA in 2020), and Jintrolong (hGH attached to a 40-kDa PEG; effectively increase half-life without reducing receptor affinity,
GeneScience; approved and marketed in China) (26). On the allowing a putatively reduced dosing frequency.
other hand, to extend the half-life of the hGHR antagonist We aimed to develop a ‘better’ antagonist of the full spectrum
B2036 in circulation for therapeutic viability, four to six mole- of hGH action, targeting both the hGH–GHR and hGH–PRLR
cules of linear 5-kDa amine-reactive PEGs are nonspecifically interactions, with a site-specific and reduced number of unique
covalently attached to internal lysine residues and the N-ter- pegylations towards maximally preserving the GH-binding af-
minus of the first amino acid, phenylalanine, resulting in the finity of the GHR antagonist hGH-G120K. Previously, we re-
final pegylated molecule—pegvisomant—with a mean half-life ported the design and preparation of multiple analogs of hGH-
of
70-h in humans (16–18). Pegvisomant is a competitive in- G120K where one or two specific amino acids were substituted
hibitor of the hGHR and in clinical trials was >90% effective in with cysteine at positions distal to the GH-GHR–binding
normalizing serum IGF1 in patients with acromegaly, despite a interface (31), as determined from the X-ray structure of hGH
>20-fold reduction in hGHR-binding affinity compared to hGH bound to a receptor dimer (25, 26) and the very similar struc-
or the hGH-G120K (27). ture of hGH-G120R bound to a receptor monomer (35). To
There remains considerable room for improvement in these substituted cysteines, maleimide PEGs were specifically
designing a new novel long-acting hGHR antagonist, given conjugated to build a library of multiple putative hGHR an-
that pegvisomant does present an end product heterogeneity tagonists. GHR-binding studies resulted in the selection of
given the variable number of PEGs attached per peptide, as amino acids T142 and H151 as the positions in hGH-G120K
well as the amino acid residues to which they are attached. whose substitution with cysteine and subsequent pegylation
Additionally, the pegylations drastically reduce the binding caused the least reduction in receptor-binding activity (31).

2 J. Biol. Chem. (2023) 299(8) 105030

 New PEGylated antagonist of GH action
In terms of characterizing the binding efficiencies of hGHR exclusion chromatography (SEC) prior to pegylation in all
antagonists to their target (GHR), the unmodified hGH-G120K cases. There was considerable variability in the oxidation state
was found by Ross et al. (36) to inhibit the binding of I125 GH of the native disulfides after glutathione treatment, as detected
to GHR on the surface of cells with a Ki of 4.7 nM, while hGH by small scale pegylation reactions. In some cases, SDS-PAGE
inhibited with a Ki of 3.5 nM. The Ki values for hGH and analysis showed that the number of added PEGs was equal to
hGH-G120K inhibition of I125 GH binding to the extracellular the number of substituted cysteines, indicating that the native
domain of GHR (GHBP) were 2.3 nM and 2.6 nM, respec- disulfides had formed. In other cases, the product was a
tively. In this paper, we used a competitive ELISA to determine mixture of multiple PEG conjugates with higher molecular
the relative abilities of hGH and the pegylated hGH-G120K weights, indicating that some of the native cysteines did not
mutants to inhibit the binding of biotin-hGH to the soluble form disulfides but instead reacted with the PEG-maleimides.
domain of the hGHR immobilized on an ELISA plate. To remedy the variability in the extent of thiol oxidation,
The biological assays described in this paper were performed small scale pegylations were performed after the glutathione
with two types of GHR antagonists that were pegylated on removal step. If SDS-PAGE analysis of the small-scale reaction
selected amino acid positions; hGH-G120K that contains a indicated that higher molecular weight species formed, then
single substituted cysteine (T142C) that is conjugated to a the native thiols were oxidized by the addition of copper (33,
40 kDa 2-branched PEG (GL2-400) and hGH-G120K that 41). After copper oxidation, the reaction was again monitored
contains two cysteine substitutions (T142C; H151C) which are by small scale pegylation and SDS-PAGE. Large scale pegyla-
each conjugated to a 4.5 kDa tribranched PEG (dPEGA) that has tion was performed only after higher molecular weight species
an anionic carboxyl group at the terminus of each branch (Ta- were no longer detected in the small-scale reactions, indicating
ble 1). Both types of pegylated hGHR antagonists showed robust that the native disulfides had formed.
inhibition of GH-induced downstream signaling and GH-driven
oncogenic processes in cultured human cancer cell lines. Characterization of pegylated hGHR antagonists
The 2-branched 40 kDa PEG was chosen because of its Sodium dodecyl sulfate-polyacrylamide gel electrophoresis
proven efficacy in increasing proteins’ half-lives. Substituting
The SDS-PAGE gel of the hGHR antagonists is shown in
this molecule on the N terminus (37) or C terminus (38) of GH
Figure 1C. There is a single major band for each antagonist at
or substituting a linear 40 kDa PEG on the N terminus on the
an apparent molecular weight higher than the calculated MW.
hGHR antagonist B2036 (39) resulted in greatly extended
However, pegylated proteins have been shown to have lower
in vivo half-lives although the losses in binding activities were
than expected mobilities during electrophoresis (42). There are
significant. The 4.5 kDa tribranched anionic PEG was chosen
also a number of fainter bands that can be seen in each lane.
for conjugation to two specific hGHR antagonist sites to
The gel was analyzed by ImageJ to determine purity of the
produce a smaller hGH antagonist with the potential to have
hGHR antagonists (43). The major product bands for com-
greater tumor penetration than the antagonist with the larger
pound G, compound D, and compound G0 comprised 92%,
40 kDa PEG (40). We previously showed that the half-life of a
89%, and 87% of the total protein respectively for each
50 kDa protein conjugated to one 4.5 kDa tribranched anionic
antagonist. No impurity band accounted for more than 6% of
PEG was
6 h compared with a half-life of
2 h for the same
the totals. Figure 1D shows the SDS-PAGE bands at each step
protein conjugated to a 4.5 kDa neutral analog (34). In this
of the compound G0 purification. The effect of the addition of
paper, we describe the structure and function of these three
the two molecules of dPEGA to hGH-G120K-T142C-H151C
novel hGHR antagonists.
can be seen by comparing the bands in lanes 3 and 4.

Results Matrix-assisted laser desorption ionization time-of-flight

Preparation of pegylated hGHR antagonists The MALDI-TOF mass spectra of Compounds D, G, and G’
In this study, Escherichia coli cells that expressed an hGHR (Fig. 1E) all show a single major [M + H]+ peak. The MALDI-
antagonist were disrupted by sonication and, after partial pu- TOF experimental [M + H]+/calculated [M + H]+ values are as
rification, any non-native disulfides were reduced by the follows: compound D, 65.6 kDa/62.5 kDa; compound G,
addition of low concentrations of reduced glutathione. In the 31.2 kDa/31.5 kDa; compound G0 , 30.9 kDa/31.2 kDa. This
case of the His-tagged compound G and compound D pre- data shows that the expected number of PEG molecules are
cursors, the glutathione reduction was done concomitantly conjugated to each antagonist. The large difference between
with the TEV cleavage. The glutathione was removed by size the experimental and calculated molecular weights for com-
pound D is due to the polydisperse nature of the 40 kDa PEG,
Table 1 resulting in a broad MALDI-TOF peak.
Abbreviations for GHRAs used in this study
Abbreviation GHR antagonist Analytical SEC
Compound G GGSSG-hGH-G120K-T142C-dPEGA- The SEC traces for the three pegylated hGHR antagonists
H151C-dPEGA
Compound G’ M-hGH-G120K-T142C- dPEGA- are shown in Figure 1F. Compound G shows a single major
H151C-dPEGA peak with a trailing shoulder and no significantly higher mo-
Compound D GGSSG-hGH-G120K-T142C-GL2-400
lecular weight peaks. Compound G0 shows a single large peak

J. Biol. Chem. (2023) 299(8) 105030 3

New PEGylated antagonist of GH action

Figure 1. Characterization of pegylated GHR antagonists. A, crystal structure of hGH (ribbon) bound to hGHR dimer (stick). B, white arrows indicating
Thr142 and His151 residues in hGH amino acids facing away from the GH-GHR–binding site chosen based on solvent accessibility and steric access, for
cysteine substitutions and subsequent site-specific PEG attachments. C, SDS-PAGE gel of purified GHR antagonists. D, SDS-PAGE gel after the different steps
of the compound G0 preparation. E, MALDI-TOF mass spectra of compound D, compound G, and compound G’. F, analytical SEC profile of compounds G, G0 ,
and D (top to bottom). hGH, human growth hormone; hGHR, hGH receptor; SEC, size-exclusion chromatography.

with a trailing shoulder and some small higher molecular Figure 2A shows the results of competitive inhibition assay,
weight peaks. Compound D shows a large peak that accounts and Table 2 shows the IC50 values from four independent
for 90% of the total peak area and a higher molecular weight competitive assays. Using the IC50 hGH/IC50 antagonist ratio,
peak that accounts for the remaining 10% of the peak area. compounds G, D, and G0 are calculated to retain 70%, 90%, and
40% of hGH’s affinity for the receptor, respectively. The result
Stability for compound D, which is conjugated to a 40 kDa 2-branched
PEG at position 142 and retains 90% of hGH’s binding activ-
The ability of the three hGHR antagonists and hGH to ity, can be compared to the 5% retention of binding activity
inhibit the binding of biotin-hGH to immobilized hGHR after when hGH is substituted with a 2-branched 40 kDa PEG on its N
three freeze-thaw cycles and after incubation overnight at 4  C terminus (37) or the 1% retention of binding activity when hGH
and 24  C is shown in Table S1. The three GHA antagonists is substituted with the same PEG on its C terminus (38). As per
retained over 80% of their activity after three freeze-thaw cy- the t test, the IC50s for compounds G and D are not significantly
cles and retained over 89% of this activity after overnight in- higher than the IC50 for hGH (p > 0.05; one tailed test). In
cubation at both 4  C and 24  C. These results indicate that contrast, the IC50 for compound G0 , which has 40% of the
the pegylated antagonists are generally stable and are expected hGH’s affinity for the receptor, is significantly higher than the
to perform well in the receptor-binding assays that are per- IC50 for hGH (p < 0.05; one tailed test). Considering the
formed with single-use frozen aliquots and where samples are sequenced similarity of compounds G and G0 and the fact that,
only subject to 4  C and room temperature (RT) incubations by SDS-PAGE (Fig. 1), both compounds are substantially pure, it
for short periods of time. hGH was similarly stable at 4  C and is likely that the lower binding activity of compound G0 is due to
24  C but retained only
70% of its activity after a single misfolded protein due to different preparation procedures.
freeze-thaw cycle.

Relative affinities of pegylated antagonists for the hPRLR

Relative affinities of the pegylated antagonists for the hGHR
A single competition ELSA was performed to determine
The hGHR antagonists were designed so that the conju- whether the three GHR antagonists, like hGH, bind to the
gated PEG molecules would not interfere with receptor hPRLR. Figure 2B clearly shows that these antagonists, unlike
binding (31) with the expectation that the binding constants pegvisomant (20), bind to the hPRLR with affinities similar to
for the pegylated antagonists and hGH would be similar. A that for hGH.
competitive ELISA assay was used to determine the IC50
values for the inhibition of biotin-hGH binding to the soluble
domain of the GHR immobilized on an ELISA plate. The ratio In vitro evaluation of hGHR antagonism by compound G and
of the IC50 value for hGH to the IC50 value for each antag- compound D
onist is equal to the relative binding affinities of that antagonist Tyrosine phosphorylation of STAT5, the hallmark surrogate
compared with hGH. assay for hGHR activation, was evaluated using hGHR-rich

4 J. Biol. Chem. (2023) 299(8) 105030

 New PEGylated antagonist of GH action

Figure 2. Receptor binding and inhibition of GHR antagonists. A, competitive ELISA assay for the ability of the GHR antagonists (GHRA) to inhibit the
binding of biotin-hGH to soluble GHR immobilized on an ELISA plate. B, competitive ELISA assay for the ability of the GHR antagonists to inhibit the binding
of biotin-hGH to soluble PRLR immobilized on an ELISA plate. C, successful binding of GH to GHR leads to downstream activation of STAT5 by phos-
phorylation at residues 694 and 699 by GHR-associated JAK2. This phosphorylation of STAT5 is a hallmark of GHR signaling and is used as a surrogate assay
for GHR activation/inhibition. C, Western blot and (D) competitive ELISA analysis of inhibitory effects of 200 nM of compounds (unpegylated hGH-G120K,
compound-D, or compound-G in STAT5 phosphorylation of MALME-3M cells treated with 2.5 nM (
50 ng/ml) human GH. Tukey’s multiple comparison test
was done for (C) and (D) using GraphPad Prism(v7) and * represents p < 0.001. E, Western blot analysis of STAT5 phosphorylation following treatment with
increasing doses (6.25–200 nM) of GHR antagonists versus 2.5 nM hGH. EC50 of compounds G120K and compound-G in inhibiting hGH induced STAT5
phosphorylation in GHR-rich human melanoma cells MALME-3M. hGH, human growth hormone; PRLR, prolactin receptor.

human melanoma cells, and employing both western blots and in inhibiting hGH binding (Fig. S3). Further evaluation of
ELISA showed that the efficacy of the MAL-dPEG-A conjugate in vivo efficacy in lowering circulating IGF1 levels in mice were
of double cysteine–substituted GGSSG-hGH-G120K (com- performed using purified compound G.
pound G) was comparable to the unpegylated GGSSG-hGH-
G120K and suppressed STAT5 phosphorylation by >90% In vivo efficacy of hGHR antagonist compound G0
(Fig. 2, C–E). The polydisperse GL2-400 conjugate of a single A principal physiologic action of hGH is the hepatic pro-
cysteine–substituted GGSSG-hGH-G120K, (compound D) duction of IGF1 contributing as much as 75% of the circulating
suppressed cellular STAT5 phosphorylation by 46%. IGF1. Therefore, the lowering of circulating IGF1 by the
Comparing increasing doses (6.25–200 nM) of compounds D hGHR antagonist pegvisomant is a robust surrogate biomarker
and G in the cell-based assay, compound G was found to be of its in vivo efficacy. However, pegvisomant, although is an
more efficacious (EC50 = 20.9 nM) than compound D (EC50 effective antagonist of the hGHR, does not act as a strong
>100 nM) and close to the unpegylated GGSSG-hGH-G120K antagonist of the mouse GHR, requiring high doses to cause
(compound A, EC50 = 7.9 nM) in inhibiting hGH-induced significant observable effects. Firstly, we observed that in
STAT5 phosphorylation in human melanoma cells (Fig. 2E). cultured mouse bladder cancer cells and mouse fibroblast cells,
Moreover, comparing increasing doses of hGH (0.25–200 nM) while compound D was comparable to pegvisomant in sup-
against 50 nM of compounds A or D or G, compound G was pressing bGH-induced STAT5 phosphorylation downstream
found to be as efficacious as unpegylated GGSSG-hGH-G120K of GHR activation, compound G exhibited a clear and

Table 2
Inhibition of the binding of Biotin-hGH to immobilized GHRa
IC50 nM Relative affinity for
immobilized GHR
Inhibitor Run 1 Run 2 Run 3 Run 4 Average IC50, nM (IC50 hGH/IC50 GHA)
hGH 9.1 5.4 8.1 5.8 7 ± 2 1 (defined)
Cpd G 15.9 4.5 10 11 10 ± 5 0.7
Cpd D 9.4 3.6 9.5 8.6 8 ± 3 0.9
Cpd G’ 15.9 14.8 21.2 14.1 17 ± 3 0.4
a
The IC50 entries in this table are average values ± 1 STDEV from four independent assays (Runs 1–4) of the type shown in Figure 2.

J. Biol. Chem. (2023) 299(8) 105030 5

New PEGylated antagonist of GH action

Figure 3. Suppression of GH-target genes IGF1 and IGFBP3 levels by compound-G0 in mice. 3-months-old male Nude mice (n = 4) were intraperi-
toneally injected with (A and B) either a single intraperitoneally (i.p.) injection of compound-G0 at either 100 or 200 mg/kg body weight or PBS (control), and
serum was collected 48 and 96 h of the injection or (C and D) i.p. injected with 100 mg/kg body weight of compound-G0 once daily (intraperitoneal in-
jection) for 7 days and serum was collected on day 4 and day 8 1 h after of the injections. Commercial kits for circulating levels of IGF1 (A and C) and IGFBP3
(B and D) were used as per manufacturers protocol. (* p < 0.01, Bonferroni’s multiple comparisons test, n = 6). IGF1, insulin-like growth factor 1; IGFBP3,
IGF1-binding protein 3.

profoundly superior GHR inhibition than either pegvisomant addition significantly increases cell viability over a 72-h period
or compound D, by almost completely inhibiting the down- including a rescue from the cytotoxic effects of the chemo-
stream STAT5 phosphorylation (Fig. S4). Next, we treated therapy doxorubicin in three different melanoma cell lines
Nude male mice with compound G0 in two different regimens MALME-3M, SK-Mel-28, and SK-Mel-30 (Fig. 4, A–C).
—either a single i.p. injection of either 100 or 200 mg/kg body Compounds D and G markedly suppressed this effect of hGH
weight or seven consecutive once daily i.p. injections of across all three cell lines and significantly improved the ther-
100 mg/kg body weight. We observed that unlike a single dose apeutic efficacy of doxorubicin as observed by resazurin assay
of 100 mg/kg compound G0 which did not decrease circulating (Fig. 4, A–C, Supplemental Excel file) and posttreatment
IGF1, a single 200 mg/kg dose suppressed circulating IGF1 by crystal violet staining of viable colonies (Fig. 4D). In most
24% at 48-h and by 19% at 96-h postinjection (Fig. 3A). Cor- cases, the compounds D and G had better antagonist activity
responding decreases in circulating IGF1-binding protein 3 over the 72-h period compared to the unpegylated counterpart
(IGFBP3), another hepatic biomarker of GH action, were (hGH-G120K), while there was no consistent significant dif-
suppressed by 38% at 48-h and by 31% at 96-h postinjection in ference in effects between the two pegylated compounds.
the same mice (Fig. 3B). On the other hand, seven once daily We and others have previously shown that hGH action in
i.p. injections of 100 mg/kg compound G0 suppressed IGF1 melanoma upregulates the epithelial-to-mesenchymal transi-
levels by 26% at day 4 and 27% at day 8 of treatment (Fig. 3C). tion program (via upregulation of expression of epithelial-to-
Compound G0 also suppressed circulating IGFBP3 levels at day mesenchymal transition–related transcription factors) in
4 and day 8 in the same mice significantly by 43% and 50%, these cancer cells, thereby promoting their migration and in-
respectively (Fig. 3D). vasion potential (45–50). Accordingly, the cell migration rate
in the absence and presence of hGH was evaluated over a 48-h
Anticancer bioactivity of compound G in human cancer cells period including treatments with doxorubicin alone or in
Human melanoma cells express a consistently high level of conjunction with the candidate GHR antagonists. Both pegy-
hGHR (44) and are highly responsive to hGH action, which lated compounds D and G as well as the unpegylated G120K
regulates multiple oncogenic pathways and processes in this significantly suppressed melanoma cell migration (increased %
cancer (45–50). Therefore, a GHR antagonist is a candidate area free from migrating cancer cells), more pronouncedly in
negative regulator of melanoma cell growth and therapeutic combination with doxorubicin (Figs. 4E and S5, Supplemental
response. Herein, we compared our pegylated compounds D Excel file). A similar effect was seen in the basement mem-
and G against the unpegylated GGSSG-hGH-G120K (referred brane invasion assays over a 48-h period, wherein both pegy-
to here as ‘G120K’) in human melanoma cells. Exogenous hGH lated compounds D and G as well as unpegylated G120K

6 J. Biol. Chem. (2023) 299(8) 105030

 New PEGylated antagonist of GH action

Figure 4. Anticancer efficacy of GHR antagonists. The unpegylated GGSSG-hGH-G120K, compound-D, and compound-G were tested for their anticancer
properties by virtue of their GH inhibitory effects. A–C, cell viability of human melanoma cells MALME-3M (A), SK-Mel-28 (B), SK-Mel-30 (C) following 72-h
treatments with either saline (control) or 5 nM hGH or 200 nM doxorubicin or doxorubicin and hGH in either presence or absence of saline or G120K or
compound-D or compound-G at 500 nM was assessed using resazurin cell viability assay. D, similar 72-h treatments followed by crystal violet staining shows
number of viable colonies on the plate. E–H, the efficacy of G120K or compound-D or compound-G in inhibiting MALME-3M migration rate (E), basement
membrane invasion rate (F), DiOC2 (dye, ABC transporter substrate, and surrogate for chemotherapy) retention (G), and postchemotherapy colony for-
mation (H). All experiments were done at least three times (n = 3, *p < 0.01, Tukey’s multiple comparisons test). hGH, human growth hormone.

inhibited the invasive effects of GH and improved the cyto- particularly relevant for therapeutic approaches against breast
toxicity of doxorubicin treatment (Fig. 4F, Supplemental Excel and prostate cancers. This fact is highlighted by a markedly
file). These inhibitory effects of compounds D and G on higher hazard ratio and decreased survival in breast cancer
melanoma cell invasion was further corroborated by evaluating patients (the cancer Genome Atlas data base) with higher
viable colony-forming abilities of detached cells in the super- expression signature of hGH, hGHR, and hPRLR together,
natant, following doxorubicin treatment in the presence or compared to hGHR or hPRLR alone (Fig. S7).
absence of hGH. Over a 14-day period, compounds D and G
but not the unpegylated G120K effected a 3- to 4-fold decrease
in the number of viable colonies, as observed by crystal violet Discussion
staining (Fig. 4H). Moreover, hGH action is known to drive In the current study, we describe the synthesis, purification,
drug resistance in human melanoma cells by upregulating the and characterization of biologically active dual antagonists of
expression of multidrug efflux pumps—ABC transporters the hGH-induced activation of hGHR and hPRLR, through
which allows lower chemotherapeutic retention in the tumor introduction of cysteine residues at specific sites that are distal
cells, thus enabling chemoresistance (48). One week pre- to the ligand-receptor–binding interfaces and subsequent
treatment of melanoma cells with GH and/or GHR antago- conjugation with activated PEG reagents. The purification and
nists, followed by drug-retention assay loading DiOC2 (3) as a pegylation procedures were designed to facilitate the forma-
drug surrogate, showed a 1.5- to 2-fold increase in the amount tion of the two native hGH disulfide linkages while keeping the
of dye retained within the cells over a 2-h period by com- two cysteines introduced at positions 142 and 151 of com-
pounds D and G compared to the hGH-induced efflux pounds G and G0 or the single cysteine at position 142 of
(Fig. 4G). Altogether, both compound D and compound G compound D in the reduced state so they could be pegylated.
exhibit potent anticancer properties, as a function of inhibiting After pegylation, purification by either SEC or SEC followed by
hGH action, in human melanoma cell lines. Additionally, ion exchange chromatography resulted in highly purified
compound G also strongly inhibited the growth-promoting preparation of compounds G, D, and G’. The pegylated GHR
effects of hGH and improved doxorubicin cytotoxic effects antagonists described in this study are significantly superior in
in human breast cancer cell lines T47D and MDA-MB-231. maintaining their binding affinities than pegvisomant (<5%),
Both these cell lines have a much higher hPRLR:hGHR ratio, with compound G, compound D, and compound G0 exhibit-
compared to the human melanoma cells (Fig. S6). The abro- ing, respectively, binding affinities of about 70%, 90%, and 40%
gation of hGH-regulated hGHR and hPRLR activation is of the original hGH molecule. In the biological activity assays,

J. Biol. Chem. (2023) 299(8) 105030 7

New PEGylated antagonist of GH action
compound G was particularly effective and is the current focus binding to the hGHR and normalizes serum IGF1 in >90% of
of our anticancer studies. the patients in clinical trials. Additionally, it has an impressive
Attachment of PEG moieties onto therapeutic compounds, safety profile (18). However, the pegvisomant molecule dis-
first discovered in the 1970s by Davis et al. (51), increases the plays (i) end product heterogeneity due to the nonspecific and
net size, hydrodynamic radius, and the molecular weight of variable (four to six) number of linear pegylations per mole-
proteins and also alters their physicochemical properties like cule, (ii) reduced affinity to the target hGHR compared to
spatial hindrance, conformation, and electrostatic binding. hGH, requiring an elevated dose (10–30 mg/day), and (iii) does
Successful pegylation results in decreased systemic clearance, not inhibit the hGH–PRLR binding which can be physiologi-
macrophage uptake, and proteolysis of the proteins, as well as cally crucial especially in the context of increased circulating
decreased immunogenicity (52). For example, pegylation of the hGH levels following pegvisomant treatment. In addition, the
hGH antagonist B2036 with 4 to 6 5-kDa PEG moieties on high economic burden of treatment with pegvisomant em-
random lysines and the N-terminal phenylalanine (pegvisom- phasizes the requirement for developing a comparably cost-
ant) significantly decreased B2036’s binding affinity by about effective next-in-class hGHR antagonist (39).
20-fold but greatly increased its circulating half-life in humans hGH itself can bind to and potently (as much as by pro-
to
70-h from 20-min (14). hGH substituted with a 2- lactin itself) activate the hPRLR (20) as well as hGHR/hPRLR
branched 40 kDa PEG on its N terminus (37) or its C termi- hybrid receptors (53). Importantly, the net amino acid mod-
nus (38) has been shown to also have a greatly extended in vivo ifications on pegvisomant prevents it from antagonizing the
half-life. However, the affinity for the hGHR decreased by
20 hGH-hPRLR binding (22)—a critical factor for multiple rea-
times after conjugation to the N terminus of hGH and by over sons as follows. Anywhere from 15 to 40% of patients with
100 times after C-terminus conjugation. Likewise, when the acromegaly (a condition of GH excess due to a hypersecreting
hGHR antagonist B2036 (unpegylated pegvisomant) was con- pituitary adenoma) also suffer from hyperprolactinemia due
jugated to a linear 40 kDa PEG on its N terminus (39), the half- to elevated secretion of hPRL from the pituitary cells (54–56).
life was greatly extended but the binding to the hGHR was These patients with acromegaly and hyperprolactinemia
decreased
20 fold. In contrast, we have found that conju- (adenomas secreting both hGH and hPRL) have significantly
gating a 40 kDa 2-branched PEG to amino acid position 142 of more severe menstrual disorders, galactorrhea, poorer sur-
hGH-G120K (compound D) results in almost no decrease in gical control, and higher adenoma relapse rates than GH-only
receptor binding. From the literature precedents described pituitary adenoma patients (55). Moreover, there are
above, it is expected that compound D will have an extended numerous reports that in specific cancers, especially breast,
in vivo half-life. endometrial, liver, and prostate cancers, the hPRLR is often
The 4.5 kDa tribranched discrete PEG (dPEG-A), which is a upregulated along with a concomitant increase in the hGHR
component of compounds G and G0 , was chosen for three rea- expression (22, 55, 57–60). In these cancers, either the hGHR
sons. First, it was previously shown that a 50 kDa antibody or the hPRLR or both have been targeted by either receptor-
fragment conjugated to a single molecule of this tri-anionic specific antagonists or antibodies as well as by bispecific
4.5 kDa PEG had an approximate half-life of 6 h in mice (hGHR + hPRLR) antibodies (61, 62). Therefore, the
compared to an approximate 2 h half-life for the same antibody concomitant inhibition of hGHR and hPRLR activation is a
fragment conjugated to the analogous neutral PEG (34). It was promising approach in cancer therapy, especially in target
postulated that attaching two 4.5 kDa tribranched PEGs to the scarce situations like triple-negative breast cancer. Further-
22 kDa hGH-G120K protein would result in a similarly extended more, different hPRLR antagonists are under development for
half-life. Second, large PEG molecules, such as the 40 kDa– the treatment of different kinds of ‘unresolved systemic and
branched PEG used to make compound D, are made by poly- local hyperprolactinemia’ (63). Lastly, therefore, the hPRLR-
merization and contain a range of molecular weight molecules. activating effect of hGH is crucial in all the above cases of
The 4.5 kDa PEG, in contrast, is a single molecule “discrete PEG” acromegaly, cancer, and systemic/local hyperprolactinemia.
made by step-by-step chemical synthesis (https://pubchem.ncbi. The current study can potentially address these factors as we
nlm.nih.gov/patent/NZ-513077-A). Finally, the smaller antago- present novel antagonists of the complete hGH action, with
nist with two tri-anionic 4.5 kDa PEGs has the potential to have distinct advantages. As we showed here, compared to the
greater tumor penetration than the antagonist with the larger currently approved pegylated hGHR antagonist pegvisomant,
40 kDa PEG (40). compounds G, D, and G0 have the advantages of site speci-
Following the discoveries of the covert actions of hGH in ficity of pegylations, number of pegylations and hence no end
driving multiple human pathologies, a major pharmaceutical product heterogeneity, and markedly higher binding affinity
interest in targeting the hGH–hGHR binding towards antag- to the hGHR (Table S2). Additionally, unlike pegvisomant,
onizing hGH action ensued. The targetability of the hGHR was our compounds also exhibit strong hPRLR-binding affinity
established by the landmark discovery of the first protein which allows them to be novel candidates for a wider spec-
antagonist, pegvisomant, in our laboratory in the early 1990s, trum inhibition of hGH action. Lastly, they can be an excel-
which was a modified hGH molecule with multiple 5-kDa lent investigational tool as they appear to be superior to
nonspecific pegylations. Pegvisomant is a successful drug pegvisomant in antagonizing mouse GHR activation.
(marketed by Pfizer and prescribed for acromegaly) which acts In our earlier report, we described a short peptide (16 amino
as a competitive inhibitor/antagonist of endogenous hGH acids)–S1H–as a novel antagonist of the GHR (64). By virtue

8 J. Biol. Chem. (2023) 299(8) 105030

 New PEGylated antagonist of GH action
of its design mimicking the residues 36 to 51 comprising the detrimental prognoses are acromegaly and cancer. The possi-
mini-helix of hGH molecule, which is also important for bility of inhibiting both hGHR and hPRLR allows for a puta-
hGH–hPRLR binding, S1H successfully inhibited STAT5 tively improved treatment outcome in acromegaly, especially
activation following hGH as well as hPRL treatment on cells in female patients with pronounced effects of hyper-
harboring both hGHR and hPRLR receptors. Therefore, our prolactinemia. On the other hand, in breast cancer, as we show
previous hGHR antagonist candidate S1H is also a dual here (Fig. S7), patients overexpressing both hGHR and hPRLR
antagonist of hGH action but markedly differs in size, have a significantly higher probability of mortality than pa-
composition, mechanism of action and lacks any pegylations tients with upregulation in either one of the receptors.
compared to compounds G, G0 , and D. Therefore, S1H Therefore, a dual antagonist of the full spectrum of hGH ac-
compared to the compounds in our current report is also tion, like compound G, is expected to be highly efficacious in
expected to have a distinct pharmacokinetic and pharmaco- multiple therapeutic areas (20, 72, 73). In fact, targeted hPRLR
dynamic profile and awaits empirical validation. Direct inhibition in breast cancer treatment by small molecules and
comparative studies of antagonistic properties of S1H against monoclonal antibodies are under active development (58, 61,
compound G or pegvisomant are forthcoming and not re- 74). Here, our in vitro assessments with human melanoma and
ported in this manuscript. breast cancer cells exhibit a distinct anticancer property of
The marker for in vivo bioactivity of hGHR antagonists is compound G, especially in sensitizing cells to chemotherapy.
the lowering of serum IGF1 and IGFBP3. Studies in rodents Further in vivo studies using xenograft or syngeneic mouse
with pegvisomant at a variety of doses and regimen have re- models will highlight the in vivo efficacy of such a treatment
ported differential levels of IGF1 reduction and do not regimen.
perform well overall. However, pegvisomant administered In summary, a successful preparation of pegylated hGHR
(intra-peritoneal (IP)) in Nude mouse model of colorectal antagonists with markedly higher binding affinities than the
cancer over 30-days at 60 mg/kg lowered IGF1 by 57 to 64% currently approved hGHR antagonist, pegvisomant, is reported
(65), while a 14-days treatment (IP) at 250 mg/kg reduced (Table S2). In light of the pleiotropic action of hGH, especially
IGF1 in Nude mouse model of breast cancer by 70 to 80% in later life, in driving several age-associated comorbidities
(66). From earlier successful pegylations of the hGHR including cancer, insulin resistance, cardiac hypertrophy, renal
antagonist, a 20-kDa N-terminal pegylated B2036 adminis- insufficiency, and tissue fibrosis, differential antagonism of
tered in rats at 2 mg/kg suppressed IGF1 by 30 to 43% (32), hGH action is highly promising (23, 75). The congenital
while Perry and Maynard’s 40-kDa S144C pegylated B2036 GHRKO mouse is the longest-lived laboratory mouse in the
administered in mice at 10 mg/kg suppressed IGF1 by almost world but has a short stature (9). Recent reports of adult-onset
51% (30). In our current study, we performed two sets of ablation of GH action in fully grown mice have resulted in
treatment regimens. In a single bolus of compound G at marked increase of lifespan, improved insulin sensitivity,
100 mg/kg, we did not observe a significant drop in serum decreased glomerulosclerosis, and cancer incidence in a sex-
IGF1 or IGFBP3 at 48 or 96-h post-treatment. However, a specific manner (13). Therefore, an adult life pharmaceutical
200 mg/kg single IP injection caused a 24% reduction in intervention to reduce hGH action has been suggested (1).
serum IGF1 and a 39% drop in serum IGFBP3 at 48-h post- Herein, we provide an opportunity of such an intervention in
treatment. A 7-days once daily injection at a dose of future with a repertoire of strong preclinical candidates.
100 mg/kg caused a 26% drop in 4 days and 40% drop at the Extensive pharmacologic developmental studies will identify
end of study (day 8) in serum IGF1, with a more pronounced the best molecule for future use.
effect observed for serum IGFBP3. Therefore, it is apparent
that compound G is a potent antagonist of GH-induced he-
patic IGF1 in mice, while an empirical comparison with Experimental procedures
pegvisomant is pending. Importantly, due to a much faster The pegylated hGHR antagonists used in the in vitro cell
rate of metabolism in mice than humans (67), the half-life of studies (compound G and compound D) and the nonpegylated
pegvisomant is significantly lower in mice (
20-h) than in hGHR antagonist hGH-G120K were prepared by first
humans (
70-h). In the current study, the pharmacokinetic expressing a His-tagged protein with a TEV protease cleavage
studies to evaluate the in vivo half-life of compound G have site. After affinity purification, removal of the His-tag, and
not been performed. Future studies incorporating multiple pegylation, the final products contained five additional amino
types of pegylated hGH-G120K constructs compared against acids on their N termini (GGSSG). The hGHR antagonist that
pegvisomant in mice will allow a distinct comparison, was prepared in larger quantities for the subsequent in vivo
allowing room for additional chemical modifications to select studies (compound G0 ) was expressed without a His-tag and
the best candidate compound balanced in serum half-life, contains only a single additional methionine (M) on its N
GHR binding, and in vivo bioactivity profiles. Additional terminus.
variabilities are expected across strain and sex of the animals. The structures of the two maleimide-activated PEG mole-
These studies are currently underway. cules used in this study (dPEGA and GL2-400 MA) are shown
Compound G binds strongly to both hGHR and hPRLR, as in Fig. S1. These molecules are conjugated to the thiol groups
does hGH itself (68–71). Principal human pathologies where of cysteine molecules in the hGHR antagonist proteins by well-
hyperactivation of hGH-regulated signaling cascades leads to characterized chemical reactions (76). Table 1 shows the three

J. Biol. Chem. (2023) 299(8) 105030 9

New PEGylated antagonist of GH action
pegylated hGHR antagonists prepared and tested in this study MAHHHHHHGSSGENLYFQ-GGSSG-hGH-G120K-T142C-
along with their abbreviations. The X-ray structure of hGH H151C. The inoculated broth was incubated at 37  C for 4 h
bound to the hGHR dimer (25, 26) is shown in Figure 1A. This with shaking at 200 RPM. This culture was then used to
structure was used as a model for the binding of hGH-G120K inoculate 30 ml of LB broth, containing 100 μg/ml
to the receptor (31), and the amino acid positions chosen for carbenicillin, and incubated overnight at 37  C with shaking
substitution with cysteine are shown in Figure 1B. The primary at 200 RPM. Four 2.5 L flasks, each containing 750 ml of TB
amino acid structures of pegvisomant, hGH-G120K, and the broth and carbenicillin (100 μg/ml), were inoculated with
hGHR antagonists (compound D and compound G) described 7.5 ml of the overnight culture and incubated at 37  C until
in this study, along with the positions of the native disulfide the A(600) of the culture was between 0.6 and 1.0 OD units
bonds and the positions of the PEG substitutions, are shown in (
2 h 45 min). The culture was then cooled to 16  C, made
cartoon form in Fig. S2. 1 mM IPTG, and incubated overnight with shaking at 200
RPM to induce expression. The culture was harvested by
Preparation of pegylated hGHR antagonists from precursors centrifugation at 6080 rcf for 15 min at 4  C. Cell pellets
containing N-terminal His-tags and TEV protease cleavage from 250 ml aliquots of the cell culture were stored at −20

sites (compounds G and D) C until purification.
Cell disruption—Cell pellets obtained from centrifugation of
The pegylated hGHR antagonists compound G and com-
250 ml of growth medium containing the expressed mutant
pound D were prepared by substituting either two or one
were suspended in 10 ml PBS and combined with 0.05 ml of a
amino acid, respectively, of the GHR antagonist GGSSG-hGH-
protease inhibitor cocktail that did not contain EDTA (Sigma
G120K with cysteine and conjugating the added cysteines to
P8849). The solutions were cooled in an ice water mixture and
different PEG molecules. The precursor molecules were
sonicated (Fisher 100 sonic dismembrator, 5 mm probe tip)
expressed with an N-terminal His-tag to facilitate purification
five times in 30 s bursts. After each sonication, the samples
and a TEV cleavage site to enable removal of the His-tag.
were cooled in the ice-water mixture until the temperature was
below 4  C. The sonicated suspensions were then centrifuged
Preparation of compound G at 4  C and 25,000g for 30 min and the supernatants collected
Gene construction for His-tagged hGH mutants—The WT and kept on ice.
hGH gene (somatotropin) (UniProt: UniProtKB - P01241) with IMAC purification of expressed protein—The sonicate su-
an N-terminal hexa-histidine tag and a TEV cleavage site (His- pernatants were adjusted to 0.3 M sodium chloride and made
TEV-hGH) was synthesized and cloned into pET-21 (a vector 5 mM imidazole by addition of a pH 7 solution of 150 mM
by Genscript). The His-TEV-hGH mutant was made by site- imidazole. The samples were then applied to a stoppered gravity
specific mutagenesis (QuickChange II site-directed flow column packed with 5 ml of Talon (Clontech) immobilized
mutagenesis kit; Agilent #200522) to substitute a lysine for the metal affinity resin (IMAC) equilibrated in 0.05 M sodium phos-
glycine at position 120 and to substitute a cysteine for amino phate buffer, pH 7.0 containing 5 mM imidazole and 0.3 M so-
acids codons at positions T142 and H151. The construct was dium chloride. The tops of the columns were then stoppered, and
transformed into Lemos (DE3) (New England Biolabs) for the columns mixed end-over-end at RT for 30 min. The columns
protein expression. The complete amino acid sequence for the were allowed to drain and washed with at least five 5 ml aliquots of
His-tagged construct of the compound G precursor is shown the equilibration buffer. The washing was continued until the
below. The cysteines that are conjugated to dPEGA and the A(280) nm of the eluents no longer decreased. The columns
amino acid position responsible for converting hGH into a were then eluted with the pH 7 equilibration buffer containing
hGH antagonist (G120K) are highlighted and shown in bold: 150 mM imidazole, and the product containing fractions were
MAHHHHHHGSSGENLYFQGGSSGFPTIPLSRLFDNAM made 5 mM EDTA by addition of a 100 mM solution of
LRAHRLHQLAFDTYQEFEEAYIPKEQKYSFLQNPQTSLCF disodium EDTA adjusted to pH 7.
SESIPTPSNREETQQKSNLELLRISLLLIQSWLEPVQFLRSV- TEV protease cleavage of His-Tag and reduction of non-
FANSLVYGASDSNVYDLLKDLEEKIQTLMGRLEDGSPRTG native disulfides—The IMAC-purified mutants were
QIFKQCYSKFDTNSCNDDALLKNYGLLYCFRKDMDKVET concentrated by molecular filtration (Amicon Ultra-4
FLRIVQCRSVEGSCGF centrifugal 10 filter, 10 kDa; Millipore) to 2 mg/ml, and
The amino acid sequence after TEV cleavage is shown aliquots of the solutions were made 1.5 mM reduced
below. glutathione + 0.15 mM oxidized glutathione. TEV Protease
GGSSGFPTIPLSRLFDNAMLRAHRLHQLAFDTYQEFEEA (TurboTev, Accelagen) was then added (0.04 mg protease/mg
YIPKEQKYSFLQNPQTSLCFSESIPTPSNREETQQKSNLELL protein), and the solutions incubated for 2 h at RT followed
RISLLLIQSWLEPVQFLRSVFANSLVYGASDSNVYDLLKDL by overnight incubation at 4  C. The imidazole- and
EEKIQTLMGRLEDGSPRTGQIFKQCYSKFDTNSCNDDALL glutathione-containing buffers were then exchanged on SEC
KNYGLLYCFRKDMDKVETFLRIVQCRSVEGSCGF. spin columns (Zeba 7 kDa, Millipore) for a buffer containing
Protein expression—Primary cultures were prepared by inoc- 0.05 M sodium phosphate, pH 7.0 and 0.3 M sodium chloride.
ulating 2 ml of LB broth containing 100 μg/ml of carbenicillin Pegylation—After removal of the glutathione and imidazole,
with a frozen glycerol stock of E. coli containing the plasmid for the product was analyzed to determine whether the native

10 J. Biol. Chem. (2023) 299(8) 105030

 New PEGylated antagonist of GH action
disulfide bonds had formed so that only the amino acids that with a salt gradient from 0 to 100 mM NaCl in the same Tris
were replaced with cysteine could be pegylated. Pegylation was buffer.
performed in 0.05 M sodium phosphate, pH 7.0 and 0.3 M
sodium chloride. Small aliquots of the solution were made Preparation of a pegylated GHR antagonist expressed without
1 mM Mal-dPEG-A and incubated for 1 h at RT before being a His-tag (compound G0 )
analyzed by SDS-PAGE. In cases when the native cysteines This section describes the preparation of compound-G’ (M-
were not fully oxidized, the gels showed bands with lower hGH-G120K-T142C- dPEG-A-H151C-dPEG-A), which was
mobilities (higher MW) than the mobility of compound G, expressed and purified without a His-Tag.
caused by pegylating both the native and the inserted
cysteines. If the gel showed the product and no higher
Gene construction for an hGH mutant without a His-tag
molecular weight species, then large scale pegylation was
performed. If higher molecular weight bands appeared, then The gene for M-hGH-G120K-T142C-H151C was synthe-
the formation of the native disulfides was catalyzed by the sized and cloned into the pET-21a vector by Genscript. The
addition of divalent copper. Copper oxidation was performed complete amino acid sequence of this molecule is shown
by adjusting the product solution to 2.5 μM copper sulfate below. The construct was transformed into Lemos (DE3) (New
and incubating for 10 min. This sample was then analyzed England Biolabs) for protein expression. The three site-specific
again by SDS-PAGE to ensure that the native disulfide bonds substitutions (G120K, T142C, and H151C) are highlighted and
had formed. The non-native cysteines in the compound G shown in bold:
precursor were pegylated in the large-scale reaction by MFPTIPLSRLFDNAMLRAHRLHQLAFDTYQEFEEAYIPK
exchanging the buffer to the pH 7 phosphate buffer solution EQKYSFLQNPQTSLCFSESIPTPSNREETQQKSNLELLRISLLL
with 0.5 mM maleimide-dPEGA and incubating the reactions IQSWLEPVQFLRSVFANSLVYGASDSNVYDLLKDLEEKIQTL
for 2 h at RT followed by overnight incubation at 4  C. MGRLEDGSPRTGQIFKQCYSKFDTNSCNDDALLKNYGLLY
Purification—The pegylation reaction mixture, which con- CFRKDMDKVETFLRIVQCRSVEGSCGF
tained compound G, was applied to gravity flow IMAC columns
containing 5 ml Talon resin equilibrated in a 0.05 M sodium Protein expression
phosphate buffer, pH 7.0, 0.3 M sodium chloride, and the col- A primary culture was prepared by inoculating 2 ml of LB
umns were washed with 5 column volumes (CVs) of the same broth containing 100 μg/ml of carbenicillin with a frozen
buffer. Compound G, which did not bind to the column, was glycerol stock of E. coli containing the M-hGH-G120K-
found in the Talon flow throughs and washes, which were then T142C-H151C plasmid. The inoculated broth was incubated
concentrated on a centrifugal concentrator to 0.3 ml and pu- at 37  C for 4 h with shaking at 200 RPM. This culture was
rified by SEC on a Superdex 200 Increase 10/300 Gl column then used to inoculate 30 ml of LB broth, containing 100 μg/
(GE Healthcare) equilibrated in 0.05 M Tris Buffer, pH 8, ml carbenicillin, and incubated overnight at 30  C with
containing 0.15 M sodium chloride and 10% glycerol. The shaking at 150 RPM. Four 2.5 L flasks, each containing 750 ml
product fractions were combined and analyzed for protein of TB broth and carbenicillin (100 μg/ml), were inoculated
concentration by absorption at A(280) nm and for purity by with 7.5 ml of the overnight culture and incubated for 2 h
SDS-PAGE. 45 min, at which time the A(600) nm of the cultures reached
0.6 to 1.0 OD. The cultures were then cooled to 16  C, made
1 mM IPTG, and incubated overnight with shaking at 200
Preparation of compound D
RPM to induce expression. The cultures were then harvested
The gene construction of the compound D precursor was by centrifugation at 6080 rcf for 15 min at 4  C, resulting in
the same as that shown for compound G, except that for cell pellets that were stored at −20  C until purification.
compound D, only one amino acid (T142) was changed to
cysteine. Protein expression, cell disruption, IMAC purifica- Cell disruption
tion, and TEV cleavage were performed as described for
compound G. Pegylation was performed with GL2-400 MA Cell pellets from 1.5 L batches of culture were suspended by
using the procedure described for compound G. vortexing in 150 ml of cold PBS containing 5 mM cysteine and
Purification of compound D—Compound D was purified, 0.1 ml of a protease inhibitor cocktail (Sigma P8849). The
after pegylation and passage through the second IMAC col- suspended cells were cooled in an ice water bath and sonicated
umn, by anion exchange chromatography (HiTrap Q FF, 5 ml, (Fisher 100 sonic dismembrator, 5 mm probe tip) for 30 s
Cytiva). To lower the salt concentrations prior to application followed by a 5-min incubation to cool the solution. The
to the column, samples were diluted 10× with 5 mM Tris sequence of sonication followed by cooling was repeated five
buffer, pH 8, 10% glycerol. The samples were concentrated to more times for a total of six sonication cycles. The samples
their original volumes on molecular concentrators and then were then centrifuged for 30 min at 30,600 rcf at 4  C.
diluted 10× a second time with the same buffer. The samples
were then applied to the HiTrap Q columns equilibrated in Ammonium sulfate precipitation
the 5 mM Tris, pH 8, 10% glycerol buffer, and the columns The supernatant from the centrifuged sonicate was stirred
were washed with 5 mM Tris, pH 8, 10% glycerol and eluted at 4  C and made 1.3 M ammonium sulfate by the addition of

J. Biol. Chem. (2023) 299(8) 105030 11

New PEGylated antagonist of GH action
cold 3 M ammonium sulfate in PBS, pH 7. After stirring for 20 CV 100 mM sodium chloride, 10 CV 300 mM sodium
15 min at 4  C, the precipitate was pelleted by centrifugation at chloride, and 10 CV of 500 mM sodium chloride. Fractions
4  C for 30 min at 30,600 rcf. (2 ml) were collected in tubes that contained 0.2 ml 1 M po-
tassium phosphate, pH 8.5 to adjust the pH of the eluted
Glutathione reduction fractions to neutrality. The majority of the product eluted in
The ammonium sulfate precipitate was redissolved in the buffer containing 300 mM NaCl. Fractions that contained
100 ml of cold PBS by gentle stirring at 4  C for 30 min. The pure product were identified by SDS-PAGE. These fractions
solution was then made 1.5 mM reduced glutathione, incu- were collected, and the buffer was exchanged for PBS using the
bated for 2 h at RT, and concentrated to 50 ml using a 10 kDa concentration/dilution procedure described above. The final
concentrator. The concentrated sample was applied at RT to a product was found to contain <1 EU endotoxin/mg protein
1 L Sephadex G-25 column that was equilibrated with aerated using the Pierce Chromogenic Endotoxin Quantitation Kit
PBS to remove the glutathione and allow oxidation of the (S39552).
native cysteines of the M-hGH-G120K mutant to form the
native disulfide bonds. Characterization of the pegylated GHR antagonists
Pegylation—After removal of the glutathione, the oxidation SDS-PAGE, analytical SEC, and MALDI-TOF
state of the native hGH disulfides was determined as described
SDS-PAGE was performed with 4%-20% gradient gels (Bio-
for compound G and copper sulfate oxidation was performed
Rad 4561093) using a 1X Tris/Glycine/SDS buffer (Bio-Rad
if needed. When the native disulfide bonds had formed, the
1610732). Samples were heated at 95% C for 5 min in 1X
samples, which were in pH 7.4 PBS, were adjusted to 1 mg/ml
Laemmli buffer (Bio-Rad 1610737) under nonreducing con-
and then made 0.125 mM Mal-dPEG-A. The reaction was
dition. SEC was performed on a Superdex 200 Increase col-
allowed to proceed at RT for 1 h and then stored at 4  C
umn, 10 × 300 mM (Cytiva 28990944) on an Agilent 1200
prior to purification.
HPLC system. The flow rate was 0.75 ml/min and the buffer
was PBS. Mass spectrometry was performed on a Bruker
Size-exclusion chromatography Ultraflextreme MALDI TOF-TOF.
Compound G0 was concentrated on a 10 kDa centrifugal
concentrator to
10 mg/ml and then aliquots of 6 to 9 ml GHR antagonist stability
were applied to a HiLoad Superdex 200 pg size exclusion
The hGHR antagonists and hGH in pH 7.4 PBS were tested
column at RT (Cytiva; 26 mm × 600 mm) that was equilibrated
for freeze-thaw stability by rapidly thawing small aliquots
in filter-sterilized Tris-buffered saline, pH 8 containing 10%
(10 μl) of the GHR antagonists in a RT water bath, incubating
glycerol. The SEC buffer was made just prior to use with
the aliquots on ice for 2 h, and then refreezing the samples in
endotoxin-free water. Fractions that contained compound G
a −80  C freezer overnight. One, two, and three freeze-thaw
were identified by SDS-PAGE and combined.
cycles were performed. “Room Temperature” stability was
determined by incubating samples for 2 h, 4 h, and overnight
Cation exchange chromatography
(25 h) in a 24  C incubator. The stability of the samples at 4  C
All buffers for cation exchange chromatography were made was determined by incubating the samples for 2 h, 4 h, and
using endotoxin-free water. The combined SEC fractions were overnight in a refrigerator. At both temperatures, the 2 h and
concentrated to
10 mg/ml on a 10 kDa centrifugal concen- 4-h samples were frozen, and the three time points assayed
trator (Amicon Ultra 15) and then diluted to 1 mg/ml with together the next day. Competitive binding assays were per-
10 mM sodium acetate, pH 4. The sequence of concentration formed at 17 nM GHR antagonist, and the percent activity
and dilution was repeated two additional times, and the con- retained after freeze-thaw cycles or incubation at 4  C or 24  C
ductivity was measured. If the conductivity was <1 mSiemens, were determined from the assay response of the starting
then the sample was applied to the cation exchange column. If sample divided by the assay response of the treated sample.
not, then the sequence of concentration and dilution was
repeated. The product was purified and endotoxin removed
Receptor-binding assays
using cation exchange chromatography that was performed at
4  C. Five 5-mL HiTrap SP Sepharose FF cation exchange Preparation of biotin-hGH
columns (Cytiva) were connected and equilibrated in 10 mM hGH (Thermo RP-10928) was dissolved at 1 mg/ml in a
sodium acetate, pH 4. The product, which was in a low salt pH 0.025 M sodium phosphate buffer at pH 7 and purified from
4 buffer, was applied to the column that was then washed with any small molecule contaminants on a 2 mL Zeba SEC spin
the following sequence of buffers at 4  C: 5 CV 10 mM sodium column (Thermo Fisher Scientific, 89890) equilibrated in the
acetate, pH 4; 30 CV 0.1% Triton X-114 in 10 mM sodium same buffer. The solution was then made 0.25 mM Biotin-
acetate, pH 4; 10 CV 1% CHAPS in 10 mM sodium acetate, pH dPEG4-TFP (Quanta BioDesign 10009) and incubated for 1 h
4; and 5 CV 10 mM sodium acetate, pH 4. The column was at RT. The reaction product was then purified on a second
then eluted stepwise with a 10 mM sodium acetate buffer (pH 2 mL Zeba spin column equilibrated in PBS and frozen in
4) containing different concentrations of sodium chloride; single use aliquots at −80  C.

12 J. Biol. Chem. (2023) 299(8) 105030

 New PEGylated antagonist of GH action
Affinity for the hGH receptor incubation. No hGH was present in or added to the growth
The relative affinities of hGH and the different pegylated media unless specifically indicated. After seeding in 6-well
GHR antagonists for the hGHR were compared in competi- plates at a concentration of 300,000 cells/ml overnight, cells
tive ELISA assays. To develop this assay, it was necessary to were switched to serum-free growth media (no GH) and
determine the concentration of biotin-hGH that would be starved for 4 h before the addition of hGH (to human cells) or
included in every well. Soluble, recombinant GHR (R&D bovine (b)GH (to mouse cells) at the concentrations noted.
Systems 1210-GR) was coated on an ELISA plate at 0.125 μg/ Cells were then allowed to incubate for an additional 15 min
ml in PBS for 1 h at 37  C, and after washing with PBS/0.05% before being harvested.
Tween 20 and blocking with 2% bovine serum albumin in
PBS, the wells were incubated with different concentrations Protein extraction
of biotin-hGH. After washing, treatment with 0.5 μg/ml Following treatment, cells were harvested at the indicated
streptavidin-horse radish peroxidase (Thermo Fisher Scien- time points and total protein was extracted. To promote
tific, 21130), washing, incubation with 3,30 ,5,50 -tetrame- protein extraction, the treatment media was removed by
thylbenzidine (SeraCare 5120-0083), and addition of 1 M aspiration and the cells were washed twice with ice-cold PBS.
HCL, the concentration of biotin-hGH that gave an Following washing, total protein was extracted from the cells
A(450 nm) assay response of approximately 2 was selected. using RIPA buffer (100 μl/million cells) supplemented with
The selected biotin-hGH concentration (0.1 μg/ml) was then 1.5× Halt protease and phosphatase inhibitor cocktail. Briefly,
incubated with different concentrations of hGH or the chilled RIPA buffer was added to the cells and was allowed to
pegylated GHR antagonists, and the abilities of these com- incubate for 5 min at 4  C. Adherent cells were harvested for
pounds to inhibit the binding of biotin-hGH to the plate were lysis using a sterile cell scraper. The resultant cell suspensions
determined in the assay described above. The IC50 values were sonicated on ice for 5 min with ON (2-s)/OFF (1-s)
from these assays, determined graphically (Table 2), reflect pulses at 2% power. The cell lysates were then cleared by
the relative affinities of hGH and the pegylated GHR antag- centrifugation at 8000g for 10 min at 4  C. Following
onists for the GHR. centrifugation, the supernatants were collected and stored
at −80  C until further use. Sample protein concentrations
Affinity for the hPRL receptor were determined using a Bradford assay with bovine serum
The relative affinities of hGH and the hGHR antagonists for albumin as the protein standard. Absorbances at 595 nm were
the hPRLR was done using the same procedure described measured using a Spectramax 250 multi-mode plate reader
above for the hGHR except that the plate was coated with (Molecular Devices) and processed using Softmax Pro v4.7.1
0.3 μg/ml of the recombinant hPRLR (R&D Systems 1167-PR), software (https://support.moleculardevices.com/s/article/
and the inhibition assays were performed in the presence of SoftMax-Pro-7-1-software-Download-page).
0.5 μg/ml biotin-hGH.
Western blot
Biological assays for GHR inhibition Cytosolic fractions from treated cells were thawed on ice,
and 100 μg protein was loaded onto polyacrylamide gels for
The abilities of the purified and pegylated compounds to
separation by SDS-PAGE. All SDS-PAGE gels were resolved at
inhibit hGHR activation was assessed using a surrogate assay,
120 V for 1.5 h at 20  C. Following separation, the proteins
tyrosine phosphorylation of STAT5, as hGH binding to
were transferred to activated PVDF membranes using a wet-
hGHR is followed by immediate activation of hGHR-
transfer method at 20 V for 16 h at 4  C. The membranes
associated JAK2 kinase which in turn tyrosine phosphory-
were then blocked with 5% nonfat dry milk in 1× Tris buffered
lates STAT5 (77).
saline, 0.1% Triton-X 100, pH 7.2 for 1 h at 25  C and incu-
bated with the primary antibody (rabbit anti-STAT5A/B;
Cell culture and hGH treatments Biotechne) at specified dilutions for 16 h at 4  C. Following
The human melanoma cell line MALME-3M was used for addition of the primary antibody, the membranes were washed
most of the assays, in addition to SK-Mel-28 (human mela- and subsequently incubated with appropriate secondary anti-
noma), SK-Mel-30 (human melanoma), MDA-MB-231 (hu- bodies (Goat anti-Rabbit horse radish peroxidase; Cell
man breast cancer), T47D (human breast cancer), MB49 Signaling Technology) at the indicated dilutions for 2 h at 25

(mouse bladder cancer), and L cells (mouse fibroblast). C. Membranes were then washed with Tris buffered saline,
MALME-3M has relatively high levels of GHR (Fig. S6). Cells 0.1% Triton-X100 and treated with West Femto Chem-
were cultured in either Eagle’s Minimum Essential Medium iluminescence detection reagents (Thermo Fisher Scientific).
(EMEM), Roswell Park Memorial Institute 1640, IMDM, or Chemiluminescence signals were detected using an Odyssey
Dulbecco’s Modified Eagle Medium (ATCC) supplemented XF (LI-COR) imaging system. Densitometry analyses of the
with 10% fetal bovine serum (FBS) and antibiotics (penicillin- blots were performed using ImageJ Software (https://imagej.
streptomycin) as per manufacturer’s instructions. Cells were nih.gov/ij/) by measuring the integrated band density and
grown at 37  C under a 5% CO2 atmosphere in a humidified normalizing against band intensity of the loading control (b-
incubator, and growth media was replaced every 48 h during actin) in the same lane. The integrated densities of pSTAT5

J. Biol. Chem. (2023) 299(8) 105030 13

New PEGylated antagonist of GH action
bands from cells after different treatments were then Cell migration assay
normalized to integrated densities of pSTAT5 bands from cells This assay was performed to assess the effect of hGHR
treated with hGH alone to obtain an induction ratio. Percent antagonist on hGH-induced migration property of the cancer
inhibition was then measured by subtracting the pSTAT5 cells. The Radius Cell Migration Assay kit from Cell Biolabs
band intensity in each lane from the pSTAT5 band intensity of (Cell Biolabs #CBA-125) was used for migration assay, and
hGH-treated samples. experiments were performed as per manufacturer’s protocol.
Briefly, the assay was performed using 24-well plates con-
Assays for anticancer properties of hGHR antagonist taining a nontoxic, 0.68 mm biocompatible hydrogel spot that
compound G is present at the center of the well, where cells cannot attach.
Crystal violet cell viability assay Prior to plating, cells were treated for 48 h, trypsinized,
This assay was performed to assess the number of cells with counted, and seeded at 5000 cells/well in the pretreated
colony-forming capacity, following hGHR antagonist and/or hydrogel spot containing 24-well plate. The hydrogel spot was
chemotherapy treatment. Human melanoma cells cultured in gently removed after 24 h incubation at 37  C/5% CO2, and
10% FBS and 1X Pen-Strep containing EMEM media were cells were allowed to migrate for up to 48 h while images were
plated onto 12-well plates at a concentration of 200 cells/cm2 captured every 24 h using a 4× objective (total magnification
and, following cell-attachment (overnight), they were treated 40×) employing an inverted Olympus IX70 microscope fitted
with either saline or recombinant hGH (@2.5 nM) or the with a Retiga 1300 camera (QImaging). The total uncovered
chemotherapy doxorubicin (@200 nM) or both GH and area at the beginning and end of assay were quantitated using
doxorubicin for 72 h. After 72 h, cells were stained with crystal ImageJ software. Experiments were done in triplicates.
violet staining solution (Abcam, cat# ab232855) for 15 to 20-
min, washed, air dried, photographed, and then solubilized, Cell invasion assay
and the absorbance was measured at 570 nm using Spectramax This assay was performed to assess the effect of hGHR
250 (Molecular Devices) spectrophotometer with a Softmax antagonist on hGH-induced invasion property of the cancer
Pro v4.7.1 software. cells. The cells were pretreated with GH in 2% FBS-containing
Dulbecco’s Modified Eagle Medium or EMEM for 48 h. The
Resazurin cell viability assay CytoSelect 96-well Cell Invasion Assay kit (CBA-112, Cell
This assay was performed to quantify the effect of hGHR Biolabs, INC) was used as per the manufacturer’s instructions.
antagonist and/or chemotherapy treatments on cell viability. Briefly, on the day of the assay, the cells were trypsinized,
Human melanoma cells cultured in 10% FBS and 1X Pen-Strep counted, and 100,000 thousand cells were then seeded per well
containing EMEM media were plated in 96-well plates at a in the upper chamber of the 96-well invasion assay well coated
concentration of 10,000 cells/well, and following cell- with basement membrane and incubated with respective
attachment (overnight), they were treated with 2% FBS and treatments in serum-free media for 24 h. After 24 h, the cells
1X Pen-Strep containing EMEM media with either saline or from under the membranes were dislodged, lysed, and stained
recombinant hGH (@5 nM) or the GHR antagonist with CyQuant GR dye solution. The fluorescence intensity
(@500 nM) or both hGH and hGHR antagonist for 72 h. After correlating with invasive cell number was measured at ex-
72 h, the media was removed and Resazurin reagent (Abcam, 480 nm/em-520 nm using a fluorescence plate reader.
cat# 129732) was added, incubated for 4-h, and the absorbance
was measured at 570 nm (reference wavelength 600 nm) using Drug retention assay
a Spectramax 250 (Molecular Devices) spectrophotometer This assay was performed to assess the effect of hGHR
with Softmax Pro v4.7.1 software. antagonist on hGH-induced drug efflux property of the cancer
cells. MALME-3M cells were pretreated with 2.5 nM hGH or
Colony formation assay 2.5 nM hGH + 500 nM hGHR antagonist in 2% FBS-
This assay was performed to assess the number of cells in containing IMDM media for a week with the media being
the supernatant, following treatments (hGHR antagonist or refreshed every 48 h. On the assay day, cells were trypsinized,
chemotherapy or both) with colony-forming capacity. counted, and suspended at 1 million cells per ml in cold
MALME-3M cultured in 10% FBS and 1X Pen-Strep con- DiOC2 (3) dye on ice for 15 min (Chemicon International,
taining EMEM media were plated onto 6-well plates at a #ECM910). The cells were then centrifuged, and the super-
concentration of 200,000 cells/well and treated with 200 nM natant removed. The cell pellets were subsequently resus-
doxorubicin for 72 h. After 72 h, the supernatant (2 mL) was pended in cold efflux buffer and distributed into different
collected (only to assess detached but viable cells) and 200 μL Eppendorf tubes under the following conditions: one set of
of it was plated on to 24-well plates in 500-μL 10% FBS and 1X tubes was kept on ice to deactivate the drug-efflux pumps as
Pen-Strep containing EMEM media and incubated for 14 days. controls, while the other two sets were kept in a 37  C water
After 14 days, the media was removed and cells were washed bath for 30 min and 120 min, respectively, allowing the active
and stained with crystal violet staining solution (Abcam, cat# drug-efflux pumps to drive out the DiOC2 (3) dye. Afterward,
ab232855) for 15 to 20 min and then washed, air dried, the cells were centrifuged and washed. Cell suspensions were
photographed. then dispensed into the wells of a black-walled 96-well plate,

14 J. Biol. Chem. (2023) 299(8) 105030

 New PEGylated antagonist of GH action
and the fluorescence was measured in a fluorescence plate Acknowledgments—We thank Dr Silvana Duran-Ortiz for her
reader at an excitation wavelength of 485 nm and an emission assistance and training in statistical analyses and graphing using
wavelength of 530 nm. GraphPad Prism. We also thank Arshad Ahmad (Translational
Biomedical Sciences Graduate program, Ohio University, Athens,
OH) for their assistance with cell culture and western-blot
In vivo efficacy of hGHR antagonist compound G’
experiments.
Nude mice (Jackson Laboratories) animals were purchased
at the age of 6 to 8 weeks and maintained in sterile cages with Author contributions—Reetobrata Basu, Rich Brody, and U. S.
ad libitum access to autoclaved water and feed (ProLab RMH methodology; Reetobrata Basu, Rich Brody, U. S., P. K., E. D., D. S.,
3000; 14% of energy from fat, 60% from carbohydrates, and L. J. C., E. B., and S. N. investigation; Reetobrata Basu formal
analysis; Reetobrata Basu, Rich Brody, and U. S. writing–original
26% from protein; PMI Nutrition International). The cages
draft; Reetobrata Basu, Rich Brody, U. S., and J. J. K. writing–
were maintained at 22  C in a humidity-controlled room and
review and editing; S. N. and J. J. K. supervision.
exposed to a 14-h light, 10-h dark cycle. Twelve-week-old male
Nude mice were injected intra-peritoneally (i.p.) with 100 or Funding and additional information—This work was supported in
200 μg/g body weight of GHR antagonists as outlined in the part by InfinixBio LLC, Columbus, Ohio, and the State of Ohio’s
Results section. Body weight was measured using a Mettler Eminent Scholar Program to J. J. K. that includes a gift from Milton
Toledo PL 202-S balance. Body composition was measured and Lawrence Goll, and by the Edison Biotechnology Institute at
using the Minispec mq Benchtop NMR analyzer (Bruker In- Ohio University, Athens, Ohio.
struments, Minispec ND2506). There were six mice in each
Conflict of interest—R. Ba., R. Br., U. S., and J. J. K. hold patent rights
treatment group. to the novel pegylated GHR antagonists designed and discussed in
this manuscript. All other authors declare that they have no con-
Assays for circulating IGF1 and IGFBP3 flicts of interest with the contents of this article.
Blood collection at specified time points was performed by
Abbreviations—The abbreviations used are: CV, column volume;
placing the mice in Plexiglass restrainers, snipping <1 mm
EMEM, Eagle’s Minimum Essential Medium; FBS, fetal bovine
from the tip of the tail with sharp surgical scissors wiped with serum; FDA, Food and Drug Administration; GH, growth hormone;
70% alcohol and kneading the tail gently to obtain 200 μl of GHR, growth hormone receptor; hGH, human growth hormone;
blood collected in a heparinized capillary tube. The end-of-life hGHR, hGH receptor; IGF1, insulin-like growth factor 1; IGFBP3,
blood collection was done just before euthanasia by retro- IGF1-binding protein 3; IMAC, immobilized metal affinity resin; IP,
orbital bleeding. All procedures performed were approved by intra-peritoneal; LS, Laron syndrome; PRLR, prolactin receptor; RT,
the Ohio University IACUC. IGF1 levels were measured using room temperature; SEC, size-exclusion chromatography.
an IGF1 mouse/rat ELISA kit (22-IG1MS-E01; ALPCO), and
IGFBP3 levels were measured using an IGFBP3 mouse ELISA
kit (EMIGFBP3; Invitrogen). References
1. Vd, L., A, A., A, B., N, B., Hm, B.-B., C, C., et al. (2015) Interventions to
Slow aging in humans: are We Ready? Aging cell 14, 497–510
Statistics
2. Tidblad, A. (2022) The history, physiology and treatment safety of growth
All values for the compound characterization assays are hormone. Acta Paediatr. 111, 215–224
given as mean ± SD. Statistics were performed using GraphPad 3. Colao, A., Grasso, L. F. S., Giustina, A., Melmed, S., Chanson, P., Pereira,
Prism software (v7.04; https://www.graphpad.com/features). A. M., et al. (2019) Acromegaly. Nat. Rev. Dis. primers 5, 20
4. Aguiar-Oliveira, M. H., and Bartke, A. (2019) Growth hormone defi-
For all in vitro biological assays, Tukey’s Multiple Compari- ciency: health and Longevity. Endocr. Rev. 40, 575–601
sons test was performed while for all in vivo experimental 5. van den Beld, A. W., Kaufman, J. M., Zillikens, M. C., Lamberts, S. W. J.,
comparisons, Bonferroni’s Multiple Comparison’s test was Egan, J. M., and van der Lely, A. J. (2018) The physiology of endocrine
performed and corresponding scatter-plots with bars are pre- systems with ageing. Lancet. Diabetes Endocrinol. 6, 647–658
sented. Differences were statistically significant when p < 0.01. 6. Harvey, S., Martínez-Moreno, C. G., Luna, M., and Arámburo, C. (2015)
Autocrine/paracrine roles of extrapituitary growth hormone and prolactin
in health and disease: an overview. Gen. Comp. Endocrinol. 220, 103–111
Data availability 7. Chesnokova, V., Zonis, S., Apostolou, A., Estrada, H. Q., Knott, S.,
All data described in the manuscript are contained in the Wawrowsky, K., et al. (2021) Local non-pituitary growth hormone is
induced with aging and facilitates epithelial damage. Cell Rep. 37, 110068
main text or supporting information. The proprietary materials 8. Basu, R., Qian, Y., and Kopchick, J. J. (2018) Lessons from growth hor-
used in this manuscript can be made available to other in- mone receptor gene-disrupted mice: are there benefits of endocrine de-
vestigators at payments to cover the costs of production and fects? Eur. J. Endocrinol. 178, R155–R181
reagents. All requests for protocols and proprietary material 9. Qian, Y., Berryman, D. E., Basu, R., List, E. O., Okada, S., Young, J. A.,
distribution should be addressed to both Nicholas Henderson et al. (2022) Mice with gene alterations in the GH and IGF family. Pi-
tuitary 25, 1–51
(nhenderson@infinixbio.com) and John J Kopchick (kopchick@ 10. Laron, Z., and Kauli, R. (2016) Fifty seven years of follow-up of the Israeli
ohio.edu). cohort of Laron Syndrome patients-From discovery to treatment. Growth
Horm. IGF Res. 28, 53–56
11. Shevah, O., and Laron, Z. (2007) Patients with congenital deficiency of
Supporting information—This article contains supporting IGF-I seem protected from the development of malignancies: a pre-
information. liminary report. Growth Horm. IGF Res. 17, 54–57

J. Biol. Chem. (2023) 299(8) 105030 15

New PEGylated antagonist of GH action
12. Guevara-Aguirre, J., Balasubramanian, P., Guevara-Aguirre, M., Wei, M., growth hormone analog is a potent stimulator of weight gain and bone
Madia, F., Cheng, C. W., et al. (2011) Growth hormone receptor defi- growth in hypophysectomized rats. Endocrinology 148, 1590–1597
ciency is associated with a major reduction in pro-aging signaling, cancer, 34. Dozier, J. K., and Distefano, M. D. (2015) Site-specific PEGylation of
and diabetes in humans. Sci. Transl. Med. 3, 70ra13 therapeutic proteins. Int. J. Mol. Sci. 16, 25831–25864
13. Duran-Ortiz, S., List, E. O., Ikeno, Y., Young, J., Basu, R., Bell, S., et al. 35. Clackson, T., Ultsch, M. H., Wells, J. A., and de Vos, A. M. (1998)
(2021) Growth hormone receptor gene disruption in mature-adult mice Structural and functional analysis of the 1:1 growth hormone:receptor
improves male insulin sensitivity and extends female lifespan. Aging cell complex reveals the molecular basis for receptor affinity. J. Mol. Biol. 277,
20, e13506 1111–1128
14. Lu, M., Flanagan, J. U., Langley, R. J., Hay, M. P., and Perry, J. K. (2019) 36. Ross, R. J., Leung, K. C., Maamra, M., Bennett, W., Doyle, N., Waters, M.
Targeting growth hormone function: strategies and therapeutic applica- J., et al. (2001) Binding and functional studies with the growth hormone
tions. Signal Transduct. Target. Ther. 4, 3 receptor antagonist, B2036-PEG (pegvisomant), reveal effects of pegyla-
15. Dehkhoda, F., Lee, C. M. M., Medina, J., and Brooks, A. J. (2018) The tion and evidence that it binds to a receptor dimer. J. Clin. Endocrinol.
growth hormone receptor: mechanism of receptor activation, cell Metab. 86, 1716–1723
signaling, and physiological Aspects. Front. Endocrinol. 9, 35 37. Finn, R. F. (2009) PEGylation of human growth hormone: strategies and
16. Kopchick, J. J., Parkinson, C., Stevens, E. C., and Trainer, P. J. (2002) properties. In: Veronese, F. M., ed. PEGylated Protein Drugs: Basic Science
Growth hormone receptor antagonists: discovery, development, and Use and Clinical Applications, Milestones in Drug Therapy, Birkhäuser, Basel:
in patients with acromegaly. Endocr. Rev. 23, 623–646 187–203
17. Muller, A. F., Kopchick, J. J., Flyvbjerg, A., and van der Lely, A. J. (2004) 38. Peschke, B., Zundel, M., Bak, S., Clausen, T. R., Blume, N., Pedersen, A.,
Clinical review 166: growth hormone receptor antagonists. J. Clin. et al. (2007) C-Terminally PEGylated hGH-derivatives. Bioorg. Med.
Endocrinol. Metab. 89, 1503–1511 Chem. 15, 4382–4395
18. Pradhananga, S., Wilkinson, I., and Ross, R. J. M. (2002) Pegvisomant: 39. Brue, T., Chanson, P., Rodien, P., Delemer, B., Drui, D., Marié, L., et al.
structure and function. J. Mol. Endocrinol. 29, 11–14 (2021) Cost-utility of acromegaly pharmacological treatments in a French
19. Bartke, A., and Kopchick, J. J. (2015) The forgotten lactogenic activity of context. Front Endocrinol. (Lausanne) 12, 745843
growth hormone: important implications for rodent studies. Endocri- 40. Thurber, G. M., Schmidt, M. M., and Wittrup, K. D. (2008) Antibody
nology 156, 1620–1622 tumor penetration: transport opposed by systemic and antigen-mediated
20. Xu, J., Sun, D., Jiang, J., Deng, L., Zhang, Y., Yu, H., et al. (2013) The role clearance. Adv. Drug Deliv. Rev. 60, 1421–1434
of prolactin receptor in GH signaling in breast cancer cells. Mol. Endo- 41. Fass, D., and Thorpe, C. (2018) Chemistry and Enzymology of disulfide
crinol. 27, 266–279 cross-linking in proteins. Chem. Rev. 118, 1169–1198
21. Wennbo, H., Gebre-Medhin, M., Gritli-Linde, A., Ohlsson, C., Isaksson, 42. Kurfürst, M. M. (1992) Detection and molecular weight determination
O. G. P., and Törnell, J. (1997) Activation of the prolactin receptor but of polyethylene glycol-modified hirudin by staining after sodium
not the growth hormone receptor is important for induction of mammary dodecyl sulfate-polyacrylamide gel electrophoresis. Anal. Biochem. 200,
tumors in transgenic mice. J. Clin. Invest. 100, 2744–2751 244–248
22. Goffin, V., Bernichtein, S., Carrière, O., Bennett, W. F., Kopchick, J. J., and 43. Rasband, W. S. (1997) ImageJ, National Institutes of Health, Bethesda,
Kelly, P. A. (1999) The human growth hormone antagonist B2036 does MD
not interact with the prolactin receptor. Endocrinology 140, 3853–3856 44. Sustarsic, E. G., Junnila, R. K., and Kopchick, J. J. (2013) Human meta-
23. Bartke, A. (2011) Pleiotropic effects of growth hormone signaling in ag- static melanoma cell lines express high levels of growth hormone receptor
ing. Trends Endocrinol. Metab. 22, 437–442 and respond to GH treatment. Biochem. Biophys. Res. Commun. 441,
24. Miller, B. S., Velazquez, E., and Yuen, K. C. J. (2020) Long-acting growth 144–150
hormone preparations - current Status and future considerations. J. Clin. 45. Basu, R., Qian, Y., Mathes, S., Terry, J., Arnett, N., Riddell, T., et al. (2022)
Endocrinol. Metab. 105, dgz149 Growth hormone receptor antagonism downregulates ATP-binding
25. Cho, H., Daniel, T., Buechler, Y. J., Litzinger, D. C., Maio, Z., Putnam, A.-M. cassette transporters contributing to improved drug efficacy against
H., et al. (2011) Optimized clinical performance of growth hormone with an melanoma and hepatocarcinoma in vivo. Front. Oncol. 0, 3369
expanded genetic code. Proc. Natl. Acad. Sci. U. S. A. 108, 9060–9065 46. Qian, Y., Basu, R., Mathes, S. C., Arnett, N. A., Duran-Ortiz, S., Funk, K.
26. Steiner, M., Frank, J., and Saenger, P. (2023) Long-acting growth hor- R., et al. (2020) Growth hormone upregulates Mediators of melanoma
mone in 2022. Pediatr. Investig. 7, 36–42 drug efflux and epithelial-to-mesenchymal transition in vitro and in vivo.
27. Stewart, P. M. (2003) Pegvisomant: an advance in clinical efficacy in Cancers (Basel) 12, 3640
acromegaly. Eur. J. Endocrinol. 148, S27–S32 47. Basu, R., Kulkarni, P., Qian, Y., Walsh, C., Arora, P., Davis, E., et al. (2019)
28. Wang, Y., Langley, R. J., Tamshen, K., Jamieson, S. M., Lu, M., Maynard, Growth hormone upregulates Melanocyte-Inducing transcription factor
H. D., et al. (2020) Long-acting human growth hormone receptor an- expression and activity via JAK2-STAT5 and SRC signaling in GH
tagonists produced in E. coli and conjugated with polyethylene glycol. receptor-Positive human melanoma. Cancers (Basel) 11, 1352
Bioconjug. Chem. 31, 1651–1660 48. Basu, R., Baumgaertel, N., Wu, S., and Kopchick, J. J. (2017) Growth
29. Tamshen, K., Wang, Y., Jamieson, S. M. F., Perry, J. K., and Maynard, H. hormone receptor knockdown sensitizes human melanoma cells to
D. (2020) Genetic code expansion enables site-specific PEGylation of a chemotherapy by attenuating expression of ABC drug efflux pumps.
human growth hormone receptor antagonist through click Chemistry. Horm. Cancer 8, 143–156
Bioconjug. Chem. 31, 2179–2190 49. Basu, R., Wu, S., and Kopchick, J. J. (2017) Targeting growth hormone
30. Wang, Y., Langley, R. J., Tamshen, K., Harms, J., Middleditch, M. J., receptor in human melanoma cells attenuates tumor progression and
Maynard, H. D., et al. (2021) Enhanced bioactivity of a human GHR epithelial mesenchymal transition via suppression of multiple oncogenic
antagonist generated by Solid-Phase site-specific PEGylation. Bio- pathways. Oncotarget 8, 21579–21598
macromolecules 22, 299–308 50. Buckels, A., Zhang, Y., Jiang, J., Athar, M., Afaq, F., Shevde-Samant, L.,
31. Yang, N., Langenheim, J. F., Wang, X., Jiang, J., Chen, W. Y., and Frank, et al. (2020) Autocrine/paracrine actions of growth hormone in human
S. J. (2008) Activation of growth hormone receptors by growth hormone melanoma cell lines. Biochem. Biophys. Rep. 21, 100716
and growth hormone antagonist dimers: insights into receptor triggering. 51. Abuchowski, A., McCoy, J. R., Palczuk, N. C., van Es, T., and Davis, F. F.
Mol. Endocrinol. 22, 978–988 (1977) Effect of covalent attachment of polyethylene glycol on immu-
32. Wu, L., Ho, S. V., Wang, W., Gao, J., Zhang, G., Su, Z., et al. (2013) N- nogenicity and circulating life of bovine liver catalase. J. Biol. Chem. 252,
terminal mono-PEGylation of growth hormone antagonist: correlation of 3582–3586
PEG size and pharmacodynamic behavior. Int. J. Pharm. 453, 533–540 52. Zuma, L. K., Gasa, N. L., Makhoba, X. H., and Pooe, O. J. (2022) Protein
33. Cox, G. N., Rosendahl, M. S., Chlipala, E. A., Smith, D. J., Carlson, S. J., PEGylation: navigating recombinant protein stability, aggregation, and
and Doherty, D. H. (2007) A long-acting, mono-PEGylated human bioactivity. Biomed. Res. Int. 2022, 8929715

16 J. Biol. Chem. (2023) 299(8) 105030

 New PEGylated antagonist of GH action
53. Xu, J., Zhang, Y., Berry, P. A., Jiang, J., Lobie, P. E., Langenheim, J. F., et al. 65. Dagnaes-Hansen, F., Duan, H., Rasmussen, L. M., Friend, K. E., and
(2011) Growth hormone signaling in human T47D breast cancer cells: Flyvbjerg, A. (2004) Growth hormone receptor antagonist administration
potential role for a growth hormone receptor-prolactin receptor complex. inhibits growth of human colorectal carcinoma in nude mice. Anticancer
Mol. Endocrinol. 25, 597–610 Res. 24, 3735–3742
54. PABLO, F. D., EASTMAN, R. C., ROTH, J., and GORDEN, P. (1981) 66. Divisova, J., Kuiatse, I., Lazard, Z. W., Weiss, H., Vreeland, F., Hadsell, D.
Plasma prolactin in acromegaly before and after treatment. J. Clin. L., et al. (2006) The growth hormone receptor antagonist pegvisomant
Endocrinol. Metab. 53, 344–352 blocks both mammary gland development and MCF-7 breast cancer
55. Wang, M., Mou, C., Jiang, M., Han, L., Fan, S., Huan, C., et al. (2012) The xenograft growth. Breast Cancer Res. Treat. 98, 315–327
characteristics of acromegalic patients with hyperprolactinemia and the 67. McIntyre, T. A., Han, C., Xiang, H., Bambal, R., and Davis, C. B. (2008)
differences in patients with merely GH-secreting adenomas: clinical Differences in the total body clearance of lead compounds in the rat and
analysis of 279 cases. Eur. J. Endocrinol. 166, 797–802 mouse: impact on pharmacokinetic screening strategy. Xenobiotica 38,
56. Burkart, F., Baitsch, G., Braun, S., Bertel, O., and Cocco, G. (2008) 605–619
Antiarrhythmic effects of Tiapamil on exercise-induced arrhythmias in 68. Fuh, G., Cunningham, B. C., Fukunaga, R., Nagata, S., Goeddel, D. V., and
patients with coronary artery disease. Cardiology 69, 165–171 Wells, J. A. (1992) Rational design of potent antagonists to the human
57. Trott, J. F., Vonderhaar, B. K., and Hovey, R. C. (2008) Historical Per- growth hormone receptor. Science 256, 1677–1680
spectives of prolactin and growth hormone as Mammogens, Lactogens 69. Fuh, G., Colosi, P., Wood, W. I., and Wells, J. A. (1993) Mechanism-
and Galactagogues—Agog for the future. J. Mammary Gland Biol. based design of prolactin receptor antagonists. J. Biol. Chem. 268,
Neoplasia 13, 3–11 5376–5381
58. Schuler, L. A., and O’Leary, K. A. (2022) Prolactin: the third hormone in 70. Somers, W., Ultsch, M., De Vos, A. M., and Kossiakoff, A. A. (1994) The
breast cancer. Front Endocrinol. (Lausanne) 13, 910978 X-ray structure of a growth hormone-prolactin receptor complex. Nature
59. Wu, Z.-S., Yang, K., Wan, Y., Qian, P.-X., Perry, J. K., Chiesa, J., et al. 372, 478–481
(2011) Tumor expression of human growth hormone and human pro- 71. Wells, J. A., Cunningham, B. C., Fuh, G., Lowman, H. B., Bass, S. H.,
lactin predict a worse survival outcome in patients with mammary or Mulkerrin, M. G., et al. (1993) The molecular basis for growth hormone-
endometrial carcinoma. J. Clin. Endocrinol. Metab. 96, E1619–1629 receptor interactions. Recent Prog. Horm. Res. 48, 253–275
60. Kong, X., Wu, W., Yuan, Y., Pandey, V., Wu, Z., Lu, X., et al. (2016) 72. Goffin, V. (2017) Prolactin receptor targeting in breast and prostate
Human growth hormone and human prolactin function as autocrine/ cancers: new insights into an old challenge. Pharmacol. Ther. 179,
paracrine promoters of progression of hepatocellular carcinoma. Onco- 111–126
target 7, 29465–29479 73. Wennbo, H., and Törnell, J. (2000) The role of prolactin and growth
61. Chen, X., Wu, D., Zheng, Y., Liu, X., and Wang, J. (2020) Preparation of a hormone in breast cancer. Oncogene 19, 1072–1076
growth hormone receptor/prolactin receptor bispecific antibody antag- 74. Borcherding, D. C., Hugo, E. R., Fox, S. R., Jacobson, E. M., Hunt, B. G.,
onist which exhibited anti-cancer activity. Front. Pharmacol. 11, 598423 Merino, E. J., et al. (2021) Suppression of breast cancer by small mole-
62. Tallet, E., Rouet, V., Jomain, J.-B., Kelly, P. A., Bernichtein, S., and Goffin, cules that block the prolactin receptor. Cancers (Basel) 13, 2662
V. (2008) Rational design of competitive prolactin/growth hormone re- 75. Kopchick, J. J., Basu, R., Berryman, D. E., Jorgensen, J. O. L., Johannsson,
ceptor antagonists. J. Mammary Gland Biol. Neoplasia 13, 105–117 G., and Puri, V. (2022) Covert actions of growth hormone: fibrosis, car-
63. Goffin, V., Touraine, P., Culler, M. D., and Kelly, P. A. (2006) Drug diovascular diseases and cancer. Nat. Rev. Endocrinol. 18, 1–16
insight: prolactin-receptor antagonists, a novel approach to treatment of 76. Hermanson, G. T. (2013) Bioconjugate Techniques. 3. Academic Press,
unresolved systemic and local hyperprolactinemia? Nat. Clin. Pract. Boston: 241
Endocrinol. Metab. 2, 571–581 77. Wang, X., Darus, C. J., Xu, B. C., and Kopchick, J. J. (1996) Identification
64. Basu, R., Nahar, K., Kulkarni, P., Kerekes, O., Sattler, M., Hall, Z., et al. of growth hormone receptor (GHR) tyrosine residues required for GHR
(2021) A novel peptide antagonist of the human growth hormone re- phosphorylation and JAK2 and STAT5 activation. Mol. Endocrinol. 10,
ceptor. J. Biol. Chem. 296, 100588 1249–1260

J. Biol. Chem. (2023) 299(8) 105030 17