University of Southern California Dissertations and Theses

Proinsulin-transferrin fusion protein as a liver targeting insulin therapeutic in Type 1 diabetes treatment

(USC Thesis Other)

Proinsulin-transferrin fusion protein as a liver targeting insulin therapeutic in Type 1 diabetes treatment

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

PROINSULIN-TRANSFERRIN FUSION PROTEIN AS A LIVER TARGETING INSULIN
THERAPEUTIC IN TYPE 1 DIABETES TREATMENT
By

Juntang Shao

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PHARMACEUTICAL SCIENCES)
August 2016

Copyright 2016 Juntang Shao

ii
Dedication

To my parents, Youhua Zhao and Zhongde Shao for their love and support.

iii
Acknowledgements

First of all, words cannot express my gratitude to my mentor, Dr. Wei-Chiang Shen.
One of the highlights in my graduate study was during the summer of 2012, the moment Dr.
Shen was willing to take me to join his lab. I was the lucky one among many other talented
graduate students who were competing for the same spot. Ever since then, I have been
benefited with tremendous help, insightful guidance and generous support from Dr. Shen.
During my work on the dissertation project, Dr. Shen has given specific instructions about
research directions and trustfully provided great freedom for me to expand ideas and explore
uncertainty. During those ups and downs in research, Dr. Shen's scientific instincts and
persisted passion to science guided me to sail on the sea of the unknown and unexpected. For
those eureka moments, Dr. Shen encouraged me to think about other possibilities and trained
me to develop critical thinking. Besides of growing into an independent scientific researcher,
I have been motivated and deeply influenced by Dr. Shen’s enthusiasm to science, his
persistence to work, his kindness to everyone and his humility. By day and day, I wish I
could learn a little bit more to be a more whole person. Also, I want to give my deep
gratitude to Mrs. Daisy Shen for her loving care and warm thoughtfulness, which always
reminded me of my mother in China. Throughout the past years, I have enjoyed the
Thanksgiving dinner cooked by Daisy and received those beautiful Christmas gifts every
year. Dr. Shen and Daisy have given countless care and support for my daily life, and for my
career path beyond the Ph.D study. I will always miss the lunch table in the hallway of PSC
404, where we can find the delicious pastries and sweets brought by Daisy as well as the
recent issues of Science magazines with Dr. Shen's reading suggestion tags.

iv
I can’t give enough thanks to Dr. Jennica Zaro. She is a smart and great scientist with
the warmest heart. She has continuously helped me in assay design and data analysis. I have
deeply learned that many details that I had overlooked could be the key in research. She has
also very patiently revised every piece of my conference abstracts and manuscripts, which
helped me to improve my skills in scientific writing. Whenever I have encountered
confusions or been trapped in "bottle neck" phase, she is very kind to offer me great
suggestions.
I also want give my sincere appreciation to all my committee members, Dr. Alicia A
McDonough, Dr. Curtis Okamoto, Dr. Kathleen Rodgers and Dr. Bangyan Stiles. Their
suggestions really broaden my horizon of my research interest. I am truly honored to have the
opportunity to speak with experts in the field and absorb nutrients from top faculty members.
Many thanks also goes to every member in Dr. Shen lab and the other fellow graduate
students. Without their help and suggestions, my graduate study life would be less colorful. I
want to sincerely thank Dr. Yan Wang and Hsin-Fang Lee for teaching me lab techniques
when I started to join the lab. Zoe Folchman-Wagner who has organized so many interesting
Pharmacy Graduate Alliance events. And I also like to thank her for her generous help in
proof-reading this dissertation. I thank Lina He in Dr. Stiles lab and Shuhua Chen in Dr.
Brinton Lab for their technique support about mouse handling. I also want to give special
thanks to Yu-Sheng Chen. We share many unforgettable and beautiful memories about those
overnight animal studies and long time experiments. Without his help and encouragement, I
might not finish this dissertation in time.
Deep in my heart, the love I felt from my parents is constantly supporting me to
improve myself and keep positive. When I look into their eyes, I will have the strength to be

v
courageous and confident. The desire that “I want to make you proud” drives me to conquer
any obstacles and fuel my dream that one day I could make a little contribution to the world.
Lastly, I will always remember the nurture from School of Pharmacy, University of
Southern California. Once a Trojan, for life time a Trojan. I will carry the spirit “fight on” in
my future life, and fearlessly embrace the day of tomorrow with smile!

vi
Table of Contents

Dedication ................................................................................................................................ ii
Acknowledgements ................................................................................................................. iii
Table of Contents .................................................................................................................... vi
List of Figures ......................................................................................................................... ix
Abbreviations .......................................................................................................................... xi
Abstract .................................................................................................................................... 1
Chapter 1 Introduction ............................................................................................................. 5
1.1 Comparison of endogenous versus exogenously administered insulin ....................................... 5
1.2 Current strategies in developing hepatopreferential insulin therapeutics .................................... 7
1.2.1 PEGylated INS lispro .......................................................................................................... 8
1.2.2 Thyroxyl-INS analog ........................................................................................................... 9
1.2.3 Hepatic-directed vesicle INS ............................................................................................. 11
1.2.4 IR-B selective INS analog ................................................................................................. 12
1.2.5 Proinsulin-transferrin ......................................................................................................... 13
1.3 Potential unanswered questions ................................................................................................ 14
1.3.1 Relative dose independent BG regulation .......................................................................... 14
1.3.2 Ambiguous biodistribution ................................................................................................ 15
Chapter 2 Characterization of hepatic conversion of Proinsulin-Transferrin ......................... 22
2.1 Background ............................................................................................................................... 22
2.2 Materials and methods .............................................................................................................. 24
2.2.1 Protein preparation ............................................................................................................ 24
2.2.2 Cell culture ........................................................................................................................ 25
2.2.3 ProINS-Tf conversion assay .............................................................................................. 26
2.2.4 TfR binding assay .............................................................................................................. 26
2.2.5 ProINS-Tf conversion on Hep G 2 cells with INS degradation inhibitor Bacitracin ......... 27
2.2.6 ProINS-Tf conversion on precision-cut liver slices (PCLS) .............................................. 27
2.2.7 C-peptide detection along with ProINS-Tf conversion ...................................................... 28
2.3 Results ....................................................................................................................................... 29
2.3.1 ProINS-Tf conversion studies on cell lines and primary mouse hepatocytes .................... 29
2.3.2 ProINS-Tf conversion on Hep G2 cells with insulin degradation inhibitor ....................... 30
2.3.3 Conversion studies on precision-cut liver slices ................................................................ 31
2.3.4 C-peptide detection during ProINS-Tf conversion in various in vitro liver models .......... 31
2.3.5 C-peptide detection after Trypsin and Carboxypeptidase B enzyme treatment ................. 32
2.4 Discussion ................................................................................................................................. 40
Chapter 3 Evaluation of liver selective action of Proinsulin-Transferrin ............................... 45
3.1 Background ............................................................................................................................... 45
3.2 Materials and Methods .............................................................................................................. 48
3.2.1 Diabetic mouse model ....................................................................................................... 48
3.2.2 Comparison of IR phosphorylation in liver versus skeletal muscles ................................. 48
3.2.3 Pyruvate tolerance test (PTT) ............................................................................................ 49

vii
3.2.4 Measurement of tissue glycogen ........................................................................................ 50
3.2.5 Measurement of glycogen synthase (GSase) enzyme activity ........................................... 50
3.2.6 Time courses of Akt dephosphorylation ............................................................................ 51
3.3 Results ....................................................................................................................................... 52
3.3.1 Preferential hepatic IR activation ...................................................................................... 52
3.3.2 Inhibition of hepatic gluconeogenesis ............................................................................... 52
3.3.3 Promotion of hepatic glycogen accumulation during feeding ........................................... 53
3.3.4. Glycogen synthase activation by ProINS-Tf in H-4-II-E .................................................. 53
3.3.5 Time-lapse INS signaling after ProINS-Tf treatment ........................................................ 54
3.4 Discussion ................................................................................................................................. 61
Chapter 4 Exploration of therapeutic potentials of Proinsulin-Transferrin as basal insulin
replacement in diabetes treatment .......................................................................................... 64
4.1 Background ............................................................................................................................... 64
4.2 Materials and Methods .............................................................................................................. 68
4.2.1 Hypoglycemic efficacy study under prolonged fasting or free-feeding ............................. 68
4.2.2 Dose response study .......................................................................................................... 69
4.2.3 Hypoglycemic efficacy study under fast/feed cycles ......................................................... 69
4.2.4 L-Alanine tolerance test and L-Glutamine tolerance test ................................................... 69
4.2.5 Glycodynamic control of premixed ProINS-Tf and INS ................................................... 70
4.2.6 Glycodynamic control of repeated dose of ProINS-Tf /INS mixture ................................ 70
4.2.7 IGF-1R activation assay on MCF-7 cells ........................................................................... 71
4.2.8 MCF-7 cells proliferation assay ......................................................................................... 71
4.2.9 PAI-1 secretion assay on Hep G2 cells .............................................................................. 72
4.2.10 Data analysis .................................................................................................................... 72
4.3 Results ....................................................................................................................................... 73
4.3.1 Hypoglycemic duration of ProINS-Tf during prolonged fasting ....................................... 73
4.3.2 Hypoglycemic effect of ProINS-Tf during free feeding .................................................... 73
4.3.3 Hypoglycemic efficacy of ProINS-Tf during fast/fed cycles ............................................. 74
4.3.4 Dose response study of ProINS-Tf .................................................................................... 74
4.3.5 Influence of systematic gluconeogenesis of ProINS-Tf ..................................................... 75
4.3.6 Glycodynamic control of premixed ProINS-Tf and INS ................................................... 76
4.3.7 Glycodynamic control of repeated dose of ProINS-Tf /INS mixture ................................ 77
4.3.8 Mitogenic potential evaluation of ProINS-Tf .................................................................... 77
4.3.9 Effect of ProINS-Tf to PAI-1 synthesis in Hep G2 cells ................................................... 78
4.4 Discussion ................................................................................................................................. 89
Chapter 5 Summary ................................................................................................................ 93
5.1 Conclusions ............................................................................................................................... 93
5.1.1 Characterization of hepatic conversion of ProINS-Tf ........................................................ 93
5.1.2 Evaluation of liver selective action of ProINS-Tf .............................................................. 94
5.1.3 Exploration of the therapeutic potentials of ProINS-Tf as basal INS replacement in
diabetes treatment ....................................................................................................................... 96
5.2 Future perspective ..................................................................................................................... 98
5.2.1 Exploitation of Tf fusion technology as a platform for hepatic activated prodrug ............ 98
5.2.2 Evaluation of the long-term glycemic management and weight neutrality advantage in
Type 2 diabetes model ................................................................................................................ 99
Reference ............................................................................................................................. 101

viii

ix
List of Figures

Chapter 1: Introduction

Figure 1.1 Comparison of physiological INS distribution and exogenous INS distribution. . 17
Figure 1.2 Two targeting approaches are exploited in the development of liver-targeted INS
therapeutics. ............................................................................................................................ 18
Figure 1.3 Hepatic sinusoid structure and mechanisms of liver-targeted INS analogs. ......... 20
Figure 1.4 Peripheral blood capillary structure and transit of liver targeted INS analogs. ..... 21

Chapter 2: Characterization of hepatic conversion of Proinsulin-Transferrin

Figure 2.1 Correlation of the TfR binding and the appearance of irINS-Tf in the medium. . 34
Figure 2.2 Conversion of ProINS-Tf on Hep G2 cells with or without INS degradation
inhibitor Bacitracin after 24 h incubation. .............................................................................. 35
Figure 2.3 Conversion of ProINS-Tf and ProINS on precision-cut liver slices. .................... 36
Figure 2.4 C-peptide stability test and detection during ProINS-Tf conversion. ................... 37
Figure 2.5 Trypsin and Carboxypeptidase B treatment of ProINS-Tf and ProINS. ............... 39
Figure 2.6 Schematic graph of hypothesized conversion steps, conversion intermediates, and
irINS-Tf formation. ................................................................................................................. 44

Chapter 3: Evaluation of liver selective action of Proinsulin-Transferrin

Figure 3.1 IR phosphorylation in the liver versus muscle. ..................................................... 55
Figure 3.2 BG increase after pyruvate bolus. ......................................................................... 56
Figure 3.3 Measurement of glycogen levels in the liver and skeletal muscle. ....................... 57
Figure 3.4 GSase activation in H-4-II-E cells. ........................................................................ 58
Figure 3.5 Time course of Akt dephosphorylation. ................................................................ 59
Figure 3.6 Time course of Akt dephosphorylation with Tf chase. ......................................... 60

Chapter 4: Exploration of therapeutic potentials of Proinsulin-Transferrin as basal insulin
replacement in diabetes treatment

Figure 4.1 Hypoglycemic effect of ProINS-Tf during fasting. ............................................... 79
Figure 4.2 Comparison of the glycodynamic effect of ProINS-Tf and INS during free-
feeding. .................................................................................................................................... 80
Figure 4.3 Hypoglycemic efficacy under fast/feed cycles. ..................................................... 81
Figure 4.4 Dose response study. ............................................................................................. 82
Figure 4.5 BG increase during (A) L-Ala tolerance test and (B) L-Gln tolerance test. .......... 83
Figure 4.6 Comparison of the glycodynamic effect between premixed ProINS-Tf/INS and
INS alone during feeding and fasting conditions. ................................................................... 84
Figure 4.7 Hypoglycemic effect of premixed ProINS-Tf/INS after repeated dose during
feeding and fasting conditions. ............................................................................................... 85
Figure 4.8 Activation of IGF-1R on MCF-7 cells after short-term protein treatment. ........... 86
Figure 4.9 Effects of ProINS-Tf and H-4-II-E pretreated ProINS-Tf on cell proliferation of
MCF-7 cells. ........................................................................................................................... 87

x

Figure 4.10 Effects of ProINS-Tf on plasminogen activator inhibitor type-1(PAI-1) secretion
in Hep G2. ............................................................................................................................... 88

xi
Abbreviations

AA Amino acid
ACTH Adrenocorticotropic hormone
ALT Alanine aminotransferase
AST Aspartate aminotransferase
AUC Area under the curve
BA Bioavailability
BCA Bicinchoninic acid
BG Blood glucose
CPDB Carboxypeptidase B
CPE Carboxypeptidase E
DMEM Dulbecco’s Modified Eagle Medium
FBS Fetal bovine serum
G6P Glucose-6-phosphate
GAPDH Glyceraldehyde-3-phosphate dehydrogenase
GI Glucose integration
GSA Glycogen synthase activity
GSase Glycogen synthase
HbA1c Hemoglobin A1c
HDV-I Hepatic-directed vesicle insulin
HGO Hepatic glucose output
HGP Hepatic glucose production
HSA Human serum albumin
HTM Hepatocyte-targeting molecule
i.p Intraperitoneal
IDE Insulin degradation enzyme
IGF-1R Insulin-like growth factor-1 receptor
INS Insulin
INS-B Insulin receptor-B selective insulin analog
INS-Tf Insulin-transferrin
IR Insulin receptor
irINS-Tf Immuno-reactive insulin-transferrin
MTT 3-(4,5-dimeth- ylthiazol-2-yl)-2,5-diphenyltetrazolium bromide
MW Molecular weight
NHGU Net hepatic glucose uptake
NIDDM Non-insulin-dependent diabetes mellitus
NMR Nuclear magnetic resonance
NPH Neutral protamine Hagedorn
NTA Nitrilotriacetic acid

xii
O-GlcNAc O-linked N-acetylglucosamine
PAI-1 Plasminogen activator inhibitor-1
PC Proprotein convertases
PCLS Precision-cut liver slices
PEG Polyethylene glycol
PGD Peripheral glucose disposal
PK/PD Pharmacokinetic and pharmacodynamic
ProINS-Tf Proinsulin-transferrin
PTT Pyruvate tolerance test
Ra Glucose hepatic production rate
Rd Glucose peripheral disposal rate
s.c Subcutaneous
SD Standard deviation
STZ Streptozotocin
TBG Thyroxine-binding globulin
Tf Transferrin
TfR Transferrin receptor
TGN Trans-Golgi network
TIA Thyroxyl-insulin analog
TRH Thyrotropin-releasing hormone
TSH Thyroid-stimulating hormone
TTR Transthyretin
UKPDS United Kingdom Prospective Diabetes Study

1
Abstract
Long-acting insulin (INS) analogs have been used in the treatment of INS-deficient
diabetes patients to meet INS needs between meal times and during night-time hours. Current
products result in overexposure of INS to muscle and adipose tissues, with insufficient
amount of INS delivered to the liver following subcutaneous (s.c) injection. This non-
physiological distribution raises risks of severe hypoglycemia, and causes weight gain after
long-term treatment. Therefore, development of better long-acting INS analogs are urgently
required to enable prolonged and stringent glycemic control, while eliminating severe
hypoglycemia. Since INS is self-administered by diabetic patients at home, a wider safe
dosage range with lower risk of hypoglycemia also remains an unmet demand.
Among a myriad of efforts in optimizing pharmacokinetic and pharmacodynamic
(PK/PD) profiles of basal INS therapeutics, novel delivery devices applied to various
administration routes are reported in an extensive volume of studies, e.g., micro-needle
technology based INS patch in transdermal delivery, and INS loaded lipid-based particles in
oral or pulmonary delivery. However, within the scope of this dissertation, the content
focuses on s.c-injected INS therapeutics, including structure modified INS analogs with
related formulation refinement. The importance of hepatopreferential action of exogenous
INS as an attempt to mimic INS's physiological distribution is also emphasized.
In contrast to current s.c-injected INS products, an ideal INS replacement requires the
distribution or action of exogenous INS to more closely mimic physiological INS by
demonstrating preferential hepatic action. However, specific to s.c-injected INS, factors that
have traditionally barred products from restoring the liver:peripheral gradient in INS

2
deficient diabetes patients include: (1) the physiological structures of subcutaneous and
peripheral tissues; (2) peripheral INS receptor (IR) mediated degradation of INS.
In this introduction, the barriers inherent to the s.c injection site and efficient liver targeting
of INS are discussed and current advances in developing hepatoselective INS therapeutics are
reviewed and compared.
Our laboratory has recently prepared a proinsulin-transferrin fusion protein (ProINS-
Tf) that is an inactive human ProINS analog with a long plasma half-life. Once it reaches the
liver, ProINS-Tf is slowly converted to an active insulin-transferrin (INS-Tf) form via the
transferrin receptor (TfR)-mediated recycling pathway in hepatocytes. ProINS-Tf
significantly inhibits glucose production and promotes glycogen accumulation in liver with
little effect on glucose deposition in muscle and adipose tissues. In chapter 2, ProINS-Tf
conversion was verified on two different liver models including in vitro mouse primary
hepatocytes and ex vivo mouse precision-cut liver slices (PCLS) culture. Conditions or
factors required for conversion were further investigated with various TfR expressing cell
lines. Generation of the active form immuno-reactive (ir) INS-Tf in rat hepatoma cells was
compared with the natural maturation process of ProINS in pancreas, in which C-peptide was
simultaneously released at equal amount as INS.
Other than the physiological hepatic: peripheral INS gradient, the liver is also a
primary therapeutic target in diabetes treatment. Aberrantly regulated glucose metabolism in
liver predisposes blood glucose (BG) excursion during both postprandial and fasting phases
in INS deficient diabetes. In addition, pathological changes including hepatic INS resistance
and hepatic steatosis are linked with the progression of type 2 diabetes and other
cardiovascular complications. Therefore, liver-targeting provides an efficient approach to

3
achieve satisfactory glycemic control and other therapeutic responses. Reported small
molecule and peptide-based drugs achieved anti-diabetic effect by regulation of glucose
metabolism related enzymes, e.g., liver-selective glucokinase activator or by targeting
specific receptors, e.g., liver-selective glucocorticoid receptor antagonist in liver. In chapter
3, ProINS-Tf's pharmacological action in liver was investigated from two aspects: (1)
activation status of hepatic IR signaling network; and (2) intervention of glucose metabolism
in liver. Desired therapeutic outcomes include restrictions on hepatic glucose output (HGO)
and improvement of the defective glycogen storage in liver. Methods to assess ProINS-Tf's
modulation in HGO and glycogenesis include gluconeogenesis substrate challenge, tissue
glycogen quantification, and enzyme activity test. ProINS-Tf preferentially activated IR in
liver, significantly increased the active form of glycogen synthase, and enhanced hepatic
glycogen storage. Moreover, ProINS-Tf displayed an extended duration of INS action in
liver, which was proved from both in vitro (Akt phosphorylation) and in vivo (pyruvate
tolerance) studies in comparison with INS.
Next, the therapeutic potential of ProINS-Tf on streptozotocin (STZ) induced Type 1
diabetic models was evaluated under various experimental settings. The condition of
fast/feed cycles was designed to simulate the basal condition / postprandial phase intervals in
human. Dose response studies were intended to examine the tolerable dose range of fusion
protein. In biphasic INS therapeutics studies, the hypoglycemic effect of ProINS-Tf was
assessed when supplemented with peripheral INS action. Whether or not ProINS-Tf led to
the potential side effects previously reported among other INS analogs or ProINS was
examined. The assessment was based on molecular characterization on in vitro assays. The
glycodynamic profiles of ProINS-Tf displayed a delayed but sustained hypoglycemic effect

4
under fasting condition following s.c injection. Compared with current INS glargine, ProINS-
Tf exhibited a longer BG lowering efficacy (40 h) and improved safety. ProINS-Tf also led
to relative dose-independent BG reduction immediately after s.c bolus. In vitro studies
suggested that ProINS-Tf has significantly reduced mitogenic effect compared with INS
itself.
Chapter 2-4 were descriptions of three independent research directions, which were
an attempt to unmask the ProINS-Tf conversion in liver, to explore ProINS-Tf's regulation of
hepatic glucose metabolism, and to assess its application in Type 1 diabetes treatment,
respectively. Meanwhile, results from those separated projects were integrated as one larger
picture for ProINS-Tf. Specifically, mechanism studies in chapter 2 provided instructive
information to design those in vivo studies in chapter 4. Similarly, key findings discussed in
chapter 2 and 3 rationalized the observation from animal experiments in chapter 4.
In summary, the Tf fusion technology ProINS-Tf demonstrated an in situ liver
activation, which rendered hepatopreferential INS action as well as a prolonged effect in
lowering BG levels. Therefore, these studies have suggested that ProINS-Tf is a promising
long-acting INS analog for Type 1 diabetes treatment with improved safety. Tf fusion
technology also provides a unique strategy for hepatic targeted protein prodrug delivery.

5
Chapter 1 Introduction

1.1 Comparison of endogenous versus exogenously administered insulin
The liver is the primary site of action for physiologically secreted INS. INS secretion
bursts are synchronous with BG fluctuations and oscillations of intracellular calcium levels
[1]. Following exocytosis of secretory granules from pancreatic beta cells, released INS is
first delivered to the hepatic portal vein, and then reaches the hepatic sinusoids. Hepatic
sinusoids are unique capillary cells characterized by their open endothelial pores (50-300 nm
diameter in humans) and discontinuous basal lamina lacking a diaphragm [2]. These
sinusoids function as a loose molecular sieve allowing substance exchange in the
perisinusoidal space between the hepatic artery and hepatocytes. In the lumen of hepatic
sinusoids, INS freely diffuses through the highly permeable endothelium and is exposed to
IR on hepatocytes. A substantial amount of INS is degraded in the liver after IR mediated
endocytosis [3]. In the hepatic portal vein, the concentration of INS fluctuates from 100 to
1000 pM. After a extensive hepatic clearance (~ 80%), the remaining INS in systemic
circulation ranges from 10 to 30 pM at basal conditions [4]. In contrast to the microvessels
aligning hepatocytes, peripheral blood capillaries are considered continuous and tight
(fenestration 6-12 nm) [5]. The transendothelial movement of INS is non-receptor mediated
and non-saturable. According to previous studies, the ratio of INS concentration in the
plasma versus the interstitial fluid of skeletal muscles significantly decreased during INS
infusion at pharmacological levels (1.37 ± 0.25) compared to physiological levels (1.98 ±
0.21) [6]. In other words, when plasma INS is at ultra-physiological level, INS's transport
efficiency and capacity to peripheral tissue increase. This suggests that a higher portion of
INS is delivered to skeletal muscles when therapeutic dose of INS is administered.

6
Current INS therapeutics administered via the s.c route are unable to recapitulate
physiological INS distribution (Fig 1.1). After s.c bolus, INS appears in the interstitial space
of the hypodermis, which is located between the skin and deep fascia covering skeletal
muscle. In the hypodermis, extracellular matrix and plasma derived proteases in the
connective tissue septa comprise the first physical and catabolism barriers against INS
absorption. INS, which is below the upper limit of capillary uptake (16 kDa), primarily
accesses systemic circulation via passive diffusion through small pores (<30 Å) or via the
paracellular pathway on capillary endothelium [7]. The overall influence of "first-pass
catabolism" on protein biologics disposition is not conclusive. However, studies have shown
that the bioavailability (BA) of s.c-injected INS in humans was as high as 84%, indicating
minimal catabolic clearance [8]. Even so, BA may vary on a case-dependent manner for
individual INS analogs. When evaluating the influence of the s.c catabolic barrier for each
INS analog, factors including injection site (abdominal vs appendicular side), retention time
at injection site (fast-acting INS vs depot effect of long-acting INS), as well as amino acid
(AA) sequence (native INS vs INS analogs with AA substitution) should be considered [9-
11]. The plasma concentration of pharmacological INS is approximately 2.5-3.5 times of
circulated INS in normal subjects [12]. This higher pharmacological INS concentration
predisposes peripheral tissues to hyperinsulinemia and leaves the liver under-insulinized.
Studies on rats showed that when arterial INS levels were clamped at twofold above basal
levels during peripheral INS infusion, the portal vein INS concentration was still marked as
"deficient", or below the basal level [13]. Peripheral IR mediated degradation and the short
plasma T
1/2
(5 min) of INS presumably resulted in observations in those early studies, in
which only 1% of exogenous INS was delivered to the liver [14]. Therefore, in order to

7
obtain sufficient INS concentrations in the liver, peripheral hyperinsulinemia is necessary.
Moreover, s.c INS administration shifts the main burden of INS metabolism from the liver to
the kidney compared with endogenous INS. Subsequently, it also raises the hypoglycemia
risk for renally impaired patients [3]. Further, the higher amount of INS exposed to the
peripheral tissues is responsible for many adverse effects and increased risk of mitogenicity
associated with exogenous INS therapy [15].
Direct delivery of INS to the hepatic portal circulation allows efficient liver-targeting.
According to empirical calculations, continuous intraportal INS infusion (0.35–0.56
U/kg/day) together with peripheral INS supplementation (0.08–0.11 U/kg/day) can reproduce
the physiological distribution of INS in Type 1 diabetic patients [1]. However, such a
complicated delivery system is unfeasible and is not suitable for self-administration by
diabetic patients. A s.c bolus, with its convenience and good patients’ compliance, remains
the main route for long-term INS treatment [16]. However, the physiological barriers of
peripheral tissues and liver when using a remote administration route (e.g., s.c) hamper the
ability of exogenous INS to mimic portal vein INS secretion. Recent advance in s.c-injected
INS analogs and formulation refinement to achieve liver-targeting were summarized in Fig.
2. Some examples are emerging novel concepts and some have been approved for following
clinical studies. By comparison and discussion of diverse targeting machineries, the goal is to
shed light on potential solutions to develop effective liver-targeted INS therapy.

1.2 Current strategies in developing hepatopreferential insulin therapeutics
INS analogs with increased size achieve hepatoselectivity by exploiting permeability
differences between hepatic sinusoids and peripheral blood capillaries. Large molecules,

8
which have restricted transport across peripheral capillaries, can easily diffuse through the
open-pored hepatic sinusoids. Moreover, molecules with a molecular size above the renal
filtration threshold (>50-60 kDa) attain longer plasma T
1/2
and protracted absorption as an s.c
depot.

1.2.1 PEGylated INS lispro
Peglispro (LY2605541, Eli Lilly and Company) is composed of a 20 kDa
polyethylene glycol (PEG) attached to lysine B28 on INS lispro (Fig 1.2). Each PEG
monomer binds three molecules of water, resulting in a hydrodynamic radius as large as that
of a 71-98 kDa protein for Peglispro [17]. Phase 2 clinical studies demonstrated that
Peglispro notably reduced intraday glycemic variation and nocturnal hypoglycemia
incidences in diabetes treatment [18,19]. After 12-week crossover studies on type 2 diabetic
patients, Peglispro displayed weight neutrality advantages (-0.6 kg) compared to INS
glargine (+0.3 kg).
The preferential hepatic action of Peglispro has been substantiated by several studies.
In euglycemic clamp studies on conscious dogs, the change from baseline in glucose
peripheral disposal rate (Rd)-to-hepatic production rate (Ra) was 0.5-0.6 in the Peglispro
infusion group compared with 1.4 ± 0.3 in the INS group. Moreover, net hepatic glucose
uptake (NHGU) was suppressed faster with a greater extent of net glycogenesis. In contrast,
non-hepatic glucose uptake in the Peglispro group was delayed and less than that in the INS
group. Molecular markers of INS signaling (phospho-Akt, phospho-GSK3β) in the liver were
also more significantly elevated in the Peglispro compared to the INS group [20]. It is
believed that the large hydrodynamic radius of Peglispro impedes its transport across

9
peripheral blood capillaries to the interstitial fluid of skeletal muscles and adipose tissue,
which leaves more Peglispro available for filtration through hepatic sinusoids [17]. In
addition, the reduced binding affinity of Peglispro to IR (less than 6% of INS lispro) possibly
decreases peripheral IR-mediated clearance and contributes to a longer residence time in liver
microcirculation [21]. However, hepato-preference of Peglispro was prominent only below a
certain dose. When the amount of infused Peglispro increased, Rd also increased and
subsequently lowered the Ra-to-Rd ratio, indicating decreased hepatoselectivity [20]. In
clinical studies, it was reported that lipid metabolism profiles and liver enzyme levels of
Peglispro-treated patients were different compared to the INS glargine treatment group. The
Peglispro group showed higher low-density lipoprotein and lower high-density lipoprotein
levels, as well as higher alanine aminotransferase (ALT) and aspartate aminotransferase
(AST) levels. These changes may imply hepatic toxicity in response to the loaded PEG [17].
Those changes, especially for the rising level of ALT, which is an indicator for drug-related
liver injury, were further evaluated for the risk and benefits in Phase 3 clinical trials [22].
Recently, the company announced that development of Peglispro was discontinued due to
hepatic lipid accumulation observed in Phase 3 clinical studies [23].

1.2.2 Thyroxyl-INS analog
Thyroxyl-INS analog (TIA) (MW 6.5 kDa) is synthesized by conjugating thyroxy
(T4) to the N-terminal of INS B-chain (Fig 1.2). It is designed to bind thyroid hormones
associated plasma proteins, i.e. thyroxine-binding globulin (TBG; MW 54 kDa), transthyretin
(TTR; MW 54 kDa) and human serum albumin (HSA; MW 65 kDa). Specifically, the
determining factors for TIA's transit from systemic circulation to IR on action sites involve

10
(a) association/dissociation rate of TIA/TBG complex, (b) mean transit times in capillaries
and (c) diffusion rate of free TIA through endothelial barriers [24]. The enlarged complex
(MW 120-150 kDa) not only obscures TIA access to peripheral IR but also overpasses the
last rate-limiting factor to expedite TIA delivery to hepatic IR.
Intravenous infused TIA in hyperinsulinaemic-euglycaemic clamped beagles
displayed a lower Rd and a similar hepatic glucose production (HGP) inhibition compared
with native INS, which suggested TIA's action was hepatopreferential. Its peripheral action
was further reduced (lower Rd) after pre-incubation with human TBG, which has a stronger
binding affinity to TIA than dog TBG, whereas the Ra was not altered. The study supported
that TBG associated TIA hampered its binding to peripheral IR to enhance its action in liver
[25]. The thyroxyl modification also markedly reduced the metabolic clearance rate of INS
by 70-80%. In vitro competitive IR binding assays also demonstrated that binding of TIA to
TBG shifted its ED50 from 9.6 ±1.4 nM to 50.7 ± 6.8 nM, likely due to steric hindrance.
Studies on a small size group (n=5) of normal human subjects have shown that s.c-injected
TIA has a similar hypoglycemic duration as the intermediate-acting neutral protamine
Hagedorn (NPH) INS, and that the circulated form of TIA was mostly protein bound (~86%).
Compared to NPH INS, it displayed similar Ra with lower Rd, suggesting its
hepatoselectivity.
The potential influences from TIA's binding to TBG on the homeostasis of
thyrotropin-releasing hormone (TRH) / thyroid-stimulating hormone (TSH) / thyroxine were
not thoroughly elucidated. However, a dose-dependent suppression of TSH following
intraperitoneal (i.p) injection of TIA has been reported [26]. Physiologically, binding to TBG
enables a site/temporal selective delivery of tyroxine, e.g., free tyroxine is released by

11
elastase cleavage at inflammation site [27]. Whether or not enriched elastase at inflammation
sites could interfere with free TIA's distribution and its glycodynamic regulation also remains
unclear. Moreover, the ratio of TBG bound TIA in circulation may change with increased
dose of TIA due to saturation of binding to endogenous protein, leaving the unbound fraction
more susceptible to renal removal or peripheral IR action and clearance.

1.2.3 Hepatic-directed vesicle INS
Hepatic-directed vesicle INS (HDV-I) entraps INS in hepatocyte-targeting molecule
(HTM) modified vesicles (Fig 1.2). INS is encapsulated inside of vesicles, along with HTM
attached to the amphipathic lipid layers. Following absorption and vesicle release, free INS
transits to systemic circulation and triggers peripheral IR activation. Meanwhile, lipid layer-
associated INS is driven to hepatocytes by virtue of the HTM/hepatic receptor mediation.
INS is then dissociated from the lipid complex over time and exerts INS action in the liver
[28]. HTM has undergone a battery of refinements in search of the optimal hepatic receptor
as a delivery target, including digalactosyl diglyceride, asialoglycoprotein, disofenin
chromium, and hepatobiliary receptor. The most recent HTM is biotin, a supplemented
vitamin, which obliterates many safety concerns. During oral glucose tolerance tests on Type
1 diabetic patients, equal molar human INS and HDV-I after s.c bolus rendered similar
peripheral levels of INS, glucagon and glycerol. A lower plasma ß-hydroxybutyrate
concentration in the HDV-I treatment group suggested an acute hepatic inhibition of ketone
production [29]. In diabetic rats, HDV-I more markedly blunted BG excursion during i.p
[14C] glucose tolerance tests than INS. Machinery studies elucidated that the difference was

12
attributed by HDV-I's increased stimulation of hepatic glucose storage [30]. The FDA has
approved HDV-I plus pre-meal INS for Phase 3 clinical trials.
The amount of INS delivered to the liver has been enhanced in the HDV-I
formulation. Still, the vast majority of HDV-I was the free form encapsulated inside of
vesicles, with 99% free INS and 1% bound to HDV [30]. In addition, vesicles with a
diameter over 90 nm displayed enhanced uptake by nonparenchymal cells and decreased
specificity towards hepatocytes [31]. In consideration of BA, molecules larger than 100 nm
in diameter were trapped at injection sites for longer duration and resulted in a reduced extent
of uptake [32].

1.2.4 IR-B selective INS analog
IR has two isoforms, IR-A and IR-B, due to alternative splicing on exon 11. IR-A is
considered as a proliferative receptor playing important roles in fetus development.
Neoplastic cell lines also have overexpressed IR-A [33]. In comparison, IR-B presents an
additional 12 AA at the C-terminal of its alpha subunit. This terminal AA string functions as
a protein motif in sorting IR isoforms to different plasma membrane microdomains, and may
possibly lead to discrete downstream signaling [34]. IR-B is more involved in metabolic
events in glucose and lipid metabolism with canonical Akt activation. The expression pattern
of IR-B is tissue-specific. The liver predominantly expresses IR-B (more than 75%), while in
peripheral tissues the ratio between two isoforms varies with age and among different species
[35]. The IR-B selective INS analog (INS-B) has AA substitutions at B25 and B27 (Fig 1.2),
which engage in IR binding [36]. The corresponding binding site on IR was adjacent to the
region that differentiates the two IR isoforms. Compared to native INS, INS-B achieved

13
preferential binding towards IR-B over IR-A (2 to 4-fold difference) [37] with a discrete
level of decreased binding affinity to both isoforms. In ex vivo rodent studies, INS-B
exhibited higher potency in hepatic glycogen accumulation (75% potency of human INS)
than glycogen synthesis in skeletal muscles (45% potency of human INS) [38].
The hepatoselectivity of INS-B largely depends on tissue-specific expression of IR
isoforms. Although human liver predominantly express IR-B, IR-B in skeletal muscles and
adipose tissues also counts substantially (30-60%) in the total amount of IR [35]. In addition,
the ratio of IR-B:IR-A mRNA in human liver was reportedly reduced during diabetes
progression [39]. This pathological change may challenge INS-B's liver-preferential
efficiency after chronic treatment.

1.2.5 Proinsulin-transferrin
ProINS-Tf fusion protein consists of a human INS precursor, ProINS and Tf (Fig
1.2). In cultured adipocytes and after short-term incubation in hepatocytes, ProINS-Tf itself
displayed weak INS potency. However, following incubation with hepatocytes, ProINS-Tf
was converted to immuno-reactive INS-Tf (irINS-Tf). The resultant irINS-Tf acquired
enhanced INS potency and was recognized by human INS specific radioimmunoassay [40].
Conversion screens on various TfR expressing cell lines and on ex vivo cultured mouse liver
slices suggested that the conversion is liver-specific. Although the specific enzymes
responsible for the conversion are not elucidated, the conversion process occurred via the
TfR mediated slow recycling pathway only in hepatocytes, indicating a liver-specific enzyme
or compartment. When evaluated in a Type 1 diabetes mouse model (STZ-treated mice),
ProINS-Tf exerted a delayed but prolonged BG lowering effect in vivo [41]. The delayed

14
effect was subject to the required conversion from ProINS-Tf to irINS-Tf to exert the
bioactivity. The prolonged effect was also partly attributed to Tf, an enzymatically inert
serum protein established for its ability to prolong the plasma T
1/2
of fused proteins [42].
Compared with native INS, ProINS-Tf after s.c bolus led to a higher ratio of
hepatic/peripheral IR activation and enhanced hepatic glycogen accumulation.
The unique feature of ProINS-Tf is that it is the first described liver targeting INS
prodrug with low stimulation of peripheral glucose disposal (PGD) [40]. This fusion protein
meets the criteria for an optimal prodrug [43,44] with selective absorption in the liver due to
its larger size and the high expression of TfR in this organ, selective activation via a liver-
specific enzyme or intracellular compartment, and high retention in the liver. In contrast, the
aforementioned INS analogs and formulation refinements still bring circulated active ligands
that can act on peripheral tissues. Consequently, the efficiency of their hepatopreference may
be attenuated during dosage escalation (smaller Ra to Rd ratio), which is a potential
drawback. However, compared with the other INS analogs listed in Fig 1.2, ProINS-Tf's anti-
diabetic effect was only tested on a single rodent species. Therefore, its hepatic action on
human and liver toxicity needs further exploration.

1.3 Potential unanswered questions
1.3.1 Relative dose independent BG regulation
Interestingly, among the aforementioned INS analogs, many displayed their anti-
hyperglycemia effect in a dose-independent fashion. When HDV-I was infused via the
external jugular vein to diabetic dogs together with a constant portal vein glucose infusion,
HDV-I (0.025 mU/kg/min) was able to convert hepatic glucose output (HGO) to hepatic

15
glucose uptake (HGU) at less than 1% of the dose of regular INS (6.25 mU/kg/min).
However, no further increment of HGU was observed when the HDV-I dose was escalated
[30]. Similarly, on conscious dogs with peripherally infused [3-
3
H] glucose, three doses of
Peglispro via peripheral vein infusion (0.25, 0.375 and 0.5 nmol/kg/h) maintained a similar
level on net HGO inhibition and net glycogenesis stimulation. In dose-response studies of
ProINS-Tf on diabetic mice, ProINS-Tf (67.5, 135, 202.5 nmol/kg) rendered similar BG
reduction at 8 h post s.c bolus.
Although the liver is exposed to a 2 to 4-fold higher INS level compared to peripheral
tissues, HGO is more sensitive to INS compared with PGD. The half-maximal portal INS
concentration for HGO is approximately 300 pM, whereas the half-maximal INS
concentration for PGD is about 860 pM [1,45]. This result indicated that a lower dose of INS
is required to achieve complete suppression of HGO than full PGD stimulation. Therefore,
the doses applied in those studies may have already reached maximum HGO inhibition and
resulted in a less dose-dependent glycemic response.

1.3.2 Ambiguous biodistribution
Hepatoselectivity in current studies is mostly validated by pharmacodynamic features,
such as comparing the activated IR level on liver and peripheral tissues or the Ra to Rd ratio.
Given the fact that the liver extensively metabolizes INS, direct comparison of INS
distribution between liver and systematic circulation is scarce. Even with external jugular
vein infusion (0.4-mU/kg/min) of HDV-I in diabetic dogs, no portal vein INS was
measurable by standard radioimmunoassay [30]. An indirect biodistribution study of i.v
bolus HDV-1 (labeled with [
14
C] cholesterol incorporated into the lipid layers) showed that

16
the vesicle, with and without INS, disappeared rapidly from the blood and then quickly
appeared in the liver, with approximately 80% of the radioactivity appearing in the liver [29].
In silico modeling may provide alternatives other than direct liver sampling. Studies on
Metabolism Physiolab® predicted that INS glargine was distributed in the plasma, muscles
and liver at similar level, whereas the Peglispro was likely enriched in the liver than the
muscle [46,47].
INS therapy with improved glycemic control without compromising the safety issues
remains a critical challenge for chronic treatment in diabetes. Current applied s.c INS analogs
or formulations mostly achieve an anti-hyperglycemic effect via promoting PGD, reversing
the physiological portal: peripheral INS gradient. Liver-targeted INS mimicking endogenous
INS distribution is able to restore glucose metabolism in liver. Additional hepatic INS action
also reduces the required dose of peripheral-acted INS, thereby reducing the risk of INS
overdose and subsequent hypoglycemia. Liver-targeted INS analogs may serve as the next
generation of INS therapy in diabetes treatment with improved safety.

17

Figure 1.1 Comparison of physiological INS distribution and exogenous INS distribution.
Under physiological conditions, INS was secreted by pancreatic beta cells and delivered to
hepatic portal vein. 50-80% of INS was cleared out in liver before its exposure upon
peripheral tissues [3]. Exogenous INS after s.c injection appeared in blood capillaries through
diffusion. It imposed hepatic hypoinsulinemia and peripheral hyperinsulinmemia compared
to the physiological hepatic: peripheral insulin gradient.

18
Passive targeting Active targeting

Hepatic-directed vesicle INS (HDV-I)

PEGylated INS lispro

Analog INS-B [A8H,B25N,B27E,desB30
human INS]

Thyroxyl-INS analog (TIA)

ProINS-transferrin (ProINS-Tf)

Figure 1.2 Two targeting approaches are exploited in the development of liver-targeted INS
therapeutics.
Passive targeting by size increment of INS molecule (5.8 kDa) takes advantage of the
different endothelial vascular sieves in peripheral tissues versus the liver. Large molecules,
which minimally diffuse through peripheral capillary walls, are freely filtered through
fenestrated hepatic sinusoids. PEGylated INS lispro (Peglispro, LY2605541) is covalently

19
coupled with a 20 kDa polyethylene glycol (PEG) and attains a hydrodynamic radius as large
as a 71-98 kDa globular protein. Thyroxyl INS analog (TIA) is semi-synthesized by
covalently linking thyronine to the ε-amino group of PheB1. In the active liver targeting area,
hepatic-directed vesicle INS (HDV-I) encapsulates human INS in lipid vesicles (<150 nm in
diameter) which have hepatocyte-targeting molecules (HTM) incorporated on the surface.
HTMs include molecules targeting the asialoglycoprotein (galactose) receptor, hepatobiliary
receptor or biotin receptor on hepatocytes. INS-B [A8H, B25N, B27E, desB30 human INS]
is produced as an INS precursor in yeast with site directed mutagenesis, and the precursor is
then enzymatically converted into two-chain desB30 analogs [36]. ProINS-Tf is expressed in
mammalian HEK-293 cells by recombinant DNA techniques. MW increments of HDV-I and
ProINS-Tf also presumably contribute to their hepatoselectivity by passive targeting effect.

20

Figure 1.3 Hepatic sinusoid structure and mechanisms of liver-targeted INS analogs.
The fenestrated hepatic sinusoids are highly permeable with discontinuous basal lamina.
Native INS (1) and size increased INS analogs easily diffuse into space of Disse through the
open pore (150-175 nm diameter in humans [2]). (2) After diffusion, Peglispro directly acts
on IR on hepatocytes. (3) Thyroxyl INS analogs form larger complex via binding to plasma
proteins: thyroid hormone binding proteins (THBPs). TIA may undergo disassociation
process from THBPs before binding to IR. (4) At the s.c injection site, HDV-1 releases free
INS and drives the residual or amphipathic lipid layer associated INS to hepatocytes by
virtue of HTM/receptor mediation, e.g., biotin/ biotin receptor. INS is dissociated from the
lipid complex over time and then binds to hepatic IR. (5) INS-B displays a 2 to 4-fold higher
binding affinity towards the dominant isoform expressed on human liver IR-B than IR-A. (6)
ProINS-Tf is specifically activated to irINS-Tf in TfR mediated slow recycling pathway in
hepatocytes.

21

Figure 1.4 Peripheral blood capillary structure and transit of liver targeted INS analogs.
Peripheral blood capillary is tight with intact basal lamina. (1) The transendothelial
movement of native INS is non-receptor mediated and non-saturable. (2) Diffusion of size
increased INS analogs across the endothelium are more restricted compared with native INS.
(3) The association/disassociation rate of TIA/THBPs complex in circulation determines the
amount of free TIAs. Free TIAs, which have a similar MW as native INS, diffuse through
endothelium much easier than the THBPs bound form. (4) Both IR-A and IR-B are expressed
on skeletal muscles. Since INS-B preferentially binds to IR-B, its stimulation to peripheral
glucose disposal (PGD) is weaker than native INS.

22
Chapter 2 Characterization of hepatic conversion of Proinsulin-
Transferrin

2.1 Background
The natural processing of human ProINS into INS and C-peptide involves the initial
cleavage and the subsequent removal of basic amino acid residues at the C-terminal of both
INS B-chain (Arg-31, Arg-32) and C-peptide (Lys-64, Arg-65). Enzymes responsible for the
cleavages are two neuroendocrine convertases PC1/3 and PC2 [48], which trigger cleavage at
B-chain/C-peptide junction and C-peptide/A-chain junction respectively. The trimming
enzyme is Carboxypeptidase E (CPE) [49]. Based on in vitro screening studies on various
cells including human fibroblasts, myoblasts and stabilized cell lines (Hep G2 and NIH3T3),
all the cells transfected with wild-type ProINS cDNA released intact ProINS. In contrast,
considerable amount of mature INS (23-59% of the secreted (pro)INS related protein) was
released after cell transfection with modified ProINS cDNA, in which Furin sensitive motifs
replaced the original sequences at those two cleavage sites [50]. Interestingly, early studies
suggested that ProINS cDNA stably transfected AtT-20 cells - which were derived from
mouse anterior pituitary and exploited the same enzymatic mechanism in converting
precursor proopiomelanocortin into adrenocorticotropic hormone (ACTH) - were able to
store and secrete INS-like protein, not ProINS, into the medium upon secretagogue
stimulation [51]. Results from in vivo studies suggested that native human ProINS sequence
did not go through any conversion in vascular compartment after i.v administration. Even
after s.c bolus, the majority remained unchanged and 4-11% was converted into the
intermediate forms, i.e., des-Arg31, Arg32-ProINS and Arg65/Gly66-split-ProINS, but not

23
INS itself [52]. Comparison study of IR binding affinity revealed that the potency relative to
INS for des (31,32) HPI, des (64,65) HPI and human ProINS were 2-8 %, 10-20 %, and 0.56-
1.4% respectively [53].
Secreted human Tf in circulation is mainly produced in liver [54]. TfR /Tf mediated
endocytosis and recycling pathway plays the main role in transporting iron into hepatocytes
[55]. In physiological conditions, iron chelation by Tf improves the iron solubility as well as
prevents iron-mediated free radical toxicity [56]. It involves both a fast recycling process
with T
1/2
of 2 minutes and a slow recycling pathway with T
1/2
about 14 minutes [57]. In the
fast recycling process, Tf/TfR follow the itinerary from clathrin-coated pits to sorting
endosome and then return to plasma membrane. On the other hand, in the route of slow
recycling, Tf/TfR go through an additional intracellular venue, the endocytic recycling
compartment. The mildly acidic endocytic recycling compartment (pH 6.4-6.5) serves as an
accumulation reservoir for most recycled proteins and enzymes between the cell interior and
cell surface since it is the slowest stop in their return to plasma membrane [58]. It also
involves connection with trans-Golgi network (TGN), which serves as a traffic hub for either
forward or retrograde transport, and is also where Furin is concentrated [59].
The mammalian secretory proprotein convertases (PC) are part of the subtilisin-type
serine proteinases family and they are responsible for the production of many circulated
proteins which require proprotein digestion [60]. Among the PC family, 7 members belong to
the classic basic amino acid specific PC groups and share similar cutting cite at the C-
terminal of the last K/R in the consensus motif (K/R)-(X)n-(K/R), where n = 0, 2, 4 or 6 and
X is any AA [61]. These PC members are distributed at multiple intracellular locations (cell
surface: PACE4, PCSK9, PC5; plasma membrane bound: Furin, PC5; ER: PC7; Golgi: SKI-

24
1, PC5B; TGN: Furin, PC7, PC5A, PACE4; Endosome: Furin, PCSK9) [62]. Protein
secretion from hepatocytes is different compared with that from neuroendocrine cells. The
former has only consistent secreted pathway whilst the latter manages additional regulated
secretion pathway. Several differences are reported between the regulated secretory pathway
and consistent secretory pathway in context of protein secretion. The discrepancies result
from tissue specific PC members, the presence or absence of secretory storage vesicles, and
different sorting machineries [63]. Hepatocytes were reported to express redundant PC
family members, and their substrate spectrums largely overlapped [60]. For example,
multiple PC members including PACE4, PC8 and Furin were able to process proalbumin in
Hep G2 cells [64]. Our assumption is that TfR mediated recycling pathway facilitates the
access of endocytosed ProINS-Tf to the redundant PC members in hepatocytes, and therefore
generates the active form irINS-Tf. In this chapter, the ability of the liver to convert ProINS-
Tf is validated, which serves as a direct proof for ProINS-Tf's hepatic activation in addition
to previous findings on rat hepatoma cells [40]. Additionally, the difference in processing of
the ProINS portion of ProINS-Tf between hepatocytes and endocrine cells, by detecting C-
peptide during the conversion, is examined.

2.2 Materials and methods
2.2.1 Protein preparation
Vector construction and protein expression of ProINS-Tf were performed as
previously described [40]. ProINS-Tf was harvested from serum-free CD 293 medium
(Invitrogen) after transfection in HEK-293 cells (American Type Culture Collection, ATCC).
A His-tag incorporated on the C-terminal of ProINS-Tf enabled further purification by

25
nickel-nitrilotriacetic acid (NTA) agarose (Qiagen). The purity was estimated by Coomassie
blue staining after SDS-PAGE. The purified product was recognized by both anti-ProINS
antibody (ab8304, Abcam) and anti-Tf antibody (T2027, Sigma) in Western blot. ProINS-Tf
was quantified according to human serum Tf standards in anti-Tf antibody Western blot
based on band density using Image Lab™ Software (Bio-Rad). Recombinant human INS was
purchased from Sigma expressed in yeast.

2.2.2 Cell culture
H-4-II-E (rat hepatoma cells), Huh-7.5 (a human hepatocyte cell line), HeLa
(human carcinoma cervical cells), MCF7 (a human breast cancer cell line), Hep G2 (human
hepatoma cells), 5637 (a human bladder carcinoma cell line), CHO-K1 (Chinese hamster
ovary cell line), HEK-293 (human embryonic kidney cells) and A549 (a human pulmonary
alveolar epithelial cell line) cells were all obtained from ATCC. Cells were seeded in 12-well
culture plates and cultured in Dulbecco’s Modified Eagle Medium (DMEM) with high
glucose (GIBCO) containing 10% fetal bovine serum (FBS), non-essential amino acids
supplements (Life Technologies), L-glutamine and 1% penicillin/streptomycin (Invitrogen).
Confluent cells were processed with either ProINS-Tf conversion or TfR binding assays.
Primary hepatocytes were isolated from livers of wild type male C57BL/J6 mice (Jackson
Laboratory) using two-step collagenase perfusion techniques [65]. Primary hepatocytes were
cultured in DMEM/F12 medium supplemented with 1 mM L-Methionine, 14 mM NaHCO3,
15 mM HEPES, 150 nM INS, 50 nM hydrocortisone and 10% FBS at 37 °C in a humidified
atmosphere containing 5% CO2. Assays were conducted after hepatocytes were fully
attached.

26

2.2.3 ProINS-Tf conversion assay
Immediately before experiments, confluent cells were washed with DMEM
and cultured in supplemented DMEM with 0.1% bovine serum albumin (BSA) for Tf
starvation. After 30 min, culture medium was replaced with 10 nM ProINS-Tf dosing
solution followed by incubation 12 h at 37 °C in a 5% CO2 balanced cell incubator. irINS-Tf
was detected from aliquots of the medium in each well using a human INS specific
radioimmunoassay (RIA) kit (Millipore) with ProINS cross activity of less than 0.2%. At the
end of the 12 h incubation, cells were extensively washed in ice-cold PBS and lysed in 1N
NaOH. Cellular proteins were quantified by MicroBCA kit (Pierce) from lysate supernatant
after centrifugation (13,000 x g, 10 min).

2.2.4 TfR binding assay
Holo-Tf was prepared by iron saturation of human apo-Tf (Sigma) following
protocols in reference [66] with slight modification. Apo-Tf (25 mg) was dissolved in 1 mL
of 100 mM ammonium bicarbonate. Ferrous ammonium sulfate was mixed with NTA in an
equal molar ratio. Fe-NTA complex was then added into apo-Tf in the molar ratio of 2:1 iron
to Tf. The reaction was incubated at 37 °C for 2 h. Excess iron was removed by step-wise
dialysis (molecular weight cutoff 12 kDa) in 100 mM ammonium bicarbonate (pH 7.4).
Holo-Tf was then iodinated using the chloramine-T method [67]. To test TfR cell surface
binding, 10 nM 125I-holo-Tf was added to confluent cell monolayers and primary
hepatocytes. Binding assays were conducted at 4° C for 2 h, followed by aspiration of the
culture medium, extensive washing of cells with ice-cold PBS and cell lysis in 1 N NaOH.

27
The total amount of cell surface bound 125I-holo-Tf in cell lysates was measured using
Packard Cobra II gamma counter (GMI). Non-specific surface binding of 125I-holo-Tf on
the cell surface was determined in wells supplemented with a 100-fold excess of cold holo-
Tf. TfR-bound 125I-holo-Tf was calculated by subtracting the quantity of non-specific
surface bound from the total amount of 125I-holo Tf in the lysates. Cellular proteins were
quantified by MicroBCA kit (Pierce) from lysate supernatant after centrifugation (13,000 x g,
10 min).

2.2.5 ProINS-Tf conversion on Hep G 2 cells with INS degradation inhibitor Bacitracin
Hep G2 cells were seeded on 6-well culture plates until confluent. Before protein
treatment Hep G2 cells were washed with pre-warmed DMEM, and cells were incubated in
0.1% BSA DMEM with or without 0.125 mM Bacitracin (Sigma) for 0.5 h. Afterwards, the
dosing solution containing 10 nM ProINS-Tf was added to each well. For the Bacitracin
group, the dosing medium also contained 0.125 mM Bacitracin. After 24 h incubation, 300
uL of the medium supernatant were collected and the irINS-Tf level was analyzed using the
same method as described in section 2.B.3.

2.2.6 ProINS-Tf conversion on precision-cut liver slices (PCLS)
Liver samples collected from wild type male C57BL/6J mice (6–7 weeks old) were
precisely sliced at a defined thickness (250 µm) using a McIlwainTM Tissue Chopper (The
Mickle Laboratory Engineering Co. Ltd). Freshly prepared liver slices were cultured in
William’s E medium (Life Technologies) supplemented with 25 mM D-glucose and 0.1%
BSA. For the conversion assay, liver slices were treated with oxygenated (95% air / 5%

28
CO2) dosing medium containing ProINS-Tf (10 nM) with or without holo-Tf blockage (10
µM). Then, PCLS were placed on an orbital shaker (50 RPM) and cultured at 37° C. At
indicated time points, the conversion product irINS-Tf was detected from aliquots of
incubation medium using the same method as described in section 2.B.3. The conversion
assay was also performed using 10 nM human recombinant ProINS (R&D Systems). To
ensure the viability of liver slices, all the conversion assays were performed within 12 h after
liver collection. At the endpoint, liver slices were homogenized for protein quantification
using a bicinchoninic acid (BCA) kit (Pierce).

2.2.7 C-peptide detection along with ProINS-Tf conversion
2.2.7.1 C-peptide detection during conversion on H-4-II-E cells
Rat hepatoma cells H-4-II-E were seeded on 6-well culture plates and the conversion
experiment was carried out on confluent cells. Cells were treated with 10 nM ProINS-Tf for
24 hrs. Following incubation, the supernatant medium was collected and applied to both
human C-peptide RIA (HCP-20K, Millipore) and human INS RIA to check C-peptide and
irINS-Tf level, respectively. To test the stability of C-peptide in the cell culture conditions,
cells were treated with 400 pM human C-peptide (Phoenix Pharmaceutics) and the remaining
C-peptide level was measured by the same method. All the proteins were dissolved in
DMEM with 0.1% BSA.

2.2.7.2 C-peptide detection during ProINS-Tf conversion on PCLS
The PCLS culture was the same as described in section 2.B.6. IrINS-Tf and C-peptide
appearance was measured after 8 h incubation using 10 nM ProINS-Tf. The stability of C-

29
peptide in the same culture condition on PCLS was characterized by spiking the dosing
medium with 100 pM C-peptide.

2.2.7.3 C-peptide detection after Trypsin and Carboxypeptidase B enzyme treatment
Freshly mixed enzyme solutions containing Trypsin (T1426, Sigma) and
Carboxypeptidase B (CPDB, C9584, Sigma) were kept on ice. The enzyme mixture of
Trypsin and CPDB was added to 50 uL ProINS-Tf protein solution. The enzyme and
substrate ratio was 1:1 (w:w) and the final concentration of each enzyme was 0.05 µg/µL.
After enzyme addition, reaction tubes were immediately transferred to 37° C water bath. At
indicated time points, aliquots from the reaction solution were collected and diluted into ice-
cold stopping buffer. Stopping buffer was composed of protease inhibitor cocktails (Sigma)
in 0.1 M Tris/HCl buffer (pH 7.55). Samples were kept on ice until RIA analysis for both
INS and C-peptide appearance.

2.3 Results
2.3.1 ProINS-Tf conversion studies on cell lines and primary mouse hepatocytes
Previous studies showed that the ProINS-Tf conversion in H-4-II-E was through TfR
mediated endocytosis and the slow recycling pathway [6]. To explore whether ProINS-Tf
conversion was specific to hepatoma cell lines or it was a prevalent phenomenon on any TfR
expressing cells, cell lines derived from different tissues and mouse primary hepatocytes
were treated with ProINS-Tf and the appearance of irINS-Tf in the incubation medium was
measured. From the scatter plot (Fig 2.1), the amount of TfR specific binding did not
correlate with the release of irINS-Tf. HeLa cells, for example, had the highest cell surface

30
TfR level but showed negligible conversion capacity of ProINS-Tf. H-4-II-E and Huh-7.5,
two liver carcinoma cell lines, exhibited greater conversion capacity with much lower TfR
amount compared to HeLa. Interestingly, mouse primary hepatocytes, given a relative low
level of TfR, were able to efficiently convert ProINS-Tf to irINS-Tf (12.3 pmol/mg cell
protein/12 h). These data suggested that conversion of ProINS-Tf was not only modulated by
TfR but also determined by factors in hepatocyte-specific intracellular environment.

2.3.2 ProINS-Tf conversion on Hep G2 cells with insulin degradation inhibitor
A significant discrepancy in the conversion ability between rat hepatoma cell H-4-II-
E and human hepatoblastoma cell Hep G2 was noticed in Fig 2.1. As summarized in
reference [68], Hep G2, different from cultured rodent liver cell lines, maintained the
property of extensive INS degradation. Therefore, INS degradation was possibly an
explanation for the minimal conversion in Hep G2. To explore whether or not the intrinsic IR
mediated irINS-Tf degradation could mask the detection of conversion product, we compared
the irINS-Tf increase during ProINS-Tf incubation on Hep G2 cells with or without the INS
degradation inhibitor Bacitracin. Bacitracin inhibits INS degradation enzyme (IDE), which
is responsible for the initial cleavage of INS B-chain during endosomal degradation [3]. In
Fig 2.2, Bacitracin treatment slightly increased the level of detected irINS-Tf. However, the
net irINS-Tf increase was less than 20% of that on H-4-II-E cells, indicating INS degradation
is not a significant contributor to the observed minimal conversion capacity of Hep G2.

31
2.3.3 Conversion studies on precision-cut liver slices
PCLS cultures compared to primary hepatocytes have preserved liver tissue
architecture, cell heterogeneity, and hepatocyte bipolarity [69]. It closely mimics in vivo
hepatic processing, and is therefore a widely applied model to study glucose metabolism and
INS action in the liver [70]. The defined thickness (250 µm) is optimal to maintain liver slice
function and viability in PCLS cultures at least up to a 24 h incubation [71,72,73]. As shown
in Fig 2.3A, liver slices were able to convert ProINS-Tf, and released irINS-Tf (156.7±31.2
pmol/g liver protein) to the medium after 8 h incubation. Complete blockage of conversion
by holo-Tf indicated that the hepatic conversion was mediated by TfR. The lack of
conversion in the ProINS-alone treatment group also confirmed the importance of TfR. (Fig
2.3B).

2.3.4 C-peptide detection during ProINS-Tf conversion in various in vitro liver models
In physiological conditions, C-peptide and INS are exocytosed from the secreted
vesicles of pancreatic beta cells as 1:1 ratio [74]. C-peptide, which is mainly metabolized by
renal removal with minimal hepatic extraction [75], has a longer plasma T
1/2
(20-30 min)
compared with INS (3-5 min). Therefore, the plasma level of C-peptide is used as a
biomarker to evaluate the pancreas production of INS [76]. To characterize ProINS-Tf
conversion C-peptide, in addition to irINS-Tf, was included for detection after prolonged
incubation with H-4-II-E. If C-peptide was also released with irINS-Tf, it could be a
relatively stable surrogate to measure and to decipher the conversion kinetic in vivo.
C-peptide itself remained stable under both experimental conditions in H-4-II-E (Fig
2.4B) and PCLS culture (Fig 2.4E). There were no significant changes based on the RIA

32
readings from C-peptide spiked samples before and after incubation, suggesting neither H-4-
II-E nor liver slices had notable degradation of C-peptide. In Fig. 4A, at the end of 24 h
incubation with H-4-II-E, 10% of the total 10 nM ProINS-Tf in the dosing solution was
converted to irINS-Tf. However, a corresponding increase of C-peptide was not observed.
The ability of PCLS to convert ProINS-Tf to irINS-Tf was substantiated in Fig 2.3A and the
result of the simultaneous detection of C-peptide was plotted in Fig 2.4C. Due to the cross
activity (4%) of the anti-C-peptide antibody to ProINS-Tf, the RIA reading of C-peptide was
around 400 pM for the initial dosing solution containing 10 nM ProINS-Tf. During the
incubation, we did not observe a clear trend of C-peptide increase other than the baseline
fluctuation. Equal molar ProINS (Fig 2.4D) showed a lower cross activity (2%) compared to
the fusion protein (Fig 2.4C). No C-peptide increase was seen during ProINS treatment,
which was as predicted since no irINS was detectable during the PCLS incubation. (Fig
2.3B).

2.3.5 C-peptide detection after Trypsin and Carboxypeptidase B enzyme treatment
Early studies have used in vitro enzymatic digestion by Trypsin and CPB to treat
ProINS, attempting to mimic the pancreatic maturation process of INS [77]. Additionally, the
cleaving kinetics and enzyme susceptibility of those two connecting regions were thoroughly
studied [78]. Tryptic cleavage was also applied in INS manufacture by recombinant DNA
technology from E.coli. [79].
In the stability test of C-peptide under enzymatic treatment (Fig 2.5C), when the
reaction was processed to 10 min, the detected C-peptide in the solution was reduced to
24.4±1.2% (n=3) of the starting dose. This dramatic decrease suggested that the time frame

33
of the assay should be restrained within 2 min to facilitate data interpretation. The instability
of C-peptide in the experiment was unexpected, and this observation might due to a much
higher enzyme to substrate ratio (w:w=1:1) compared with those ratios (1:48-2000) in
previous studies [80, 78]. After 2 min of Trypsin and CPB treatment, the irINS signal
appeared in both the ProINS and ProINS-Tf group (Fig 2.5A). In comparison, only the
ProINS group displayed a significant C-peptide peak after 2 min enzymatic reaction (Fig
2.5B). The diminished C-peptide signal at 10 min in ProINS reaction could possibly due to
the instability of C-peptide.

34

Figure 2.1 Correlation of the TfR binding and the appearance of irINS-Tf in the medium.
The correlations of the TfR binding (pmol/mg cell protein) and the appearance of irINS-Tf in
the medium (pmol/mg cell protein) was tested after 12 h incubation with various cells. Dash
lines labeled the mean value of the test parameters. Numerical notation indicated the
following cells: 1, H-4-II-E. 2, Huh-7.5. 3, HeLa. 4, MCF7. 5, Hep G2. 6, 5637. 7, CHO-K1.
8, HEK-293. 9, A549. 10, Mouse primary hepatocytes.

35

Figure 2.2 Conversion of ProINS-Tf on Hep G2 cells with or without INS degradation
inhibitor Bacitracin after 24 h incubation.

0 24
0
100
200
300
400
Time (h)
irINS-Tf(pM)
Control
Bacitracin

36

Figure 2.3 Conversion of ProINS-Tf and ProINS on precision-cut liver slices.
Liver slices were treated with (A) 10 nM of ProINS-Tf in the absence or presence of 1000-
fold holo-Tf blockage (ProINS-Tf and ProINS-Tf + Holo-Tf respectively). In the medium
only group, liver slices were incubated in supplemented WE medium without fusion protein.
At indicated time points, aliquots of the incubation medium were applied to human INS
specific RIA kit to detect irINS-Tf. (B) 10 nM human recombinant ProINS with or without
10 uM holo-Tf. Data were expressed as average value and error bars indicated SD (n=3).

37

Figure 2.4 C-peptide stability test and detection during ProINS-Tf conversion.
Whether or not C-peptide was released during ProINS-Tf conversion was tested from
incubation medium with (A) H-4-II-E cells and (C) precision-cut liver slices. (A) After 24 h
incubation with H-4-II-E cells or with cell free medium, irINS-Tf and C-peptide were
measured from the dosing medium containing 10 nM ProINS-Tf. (B) The stability of C-
peptide during incubation with H-4-II-E cells was tested over 24 h incubation. Liver slices
C-peptide irINS-Tf
0
500
1000
1500
pM
10 nM ProINS-Tf in medium (24 hr)
H-4-II-E treated with 10 nM
ProINS-Tf (24 hr)
(A)
0 2 4 6 8
0
200
400
600
Time (h)
C- peptide (pM)
10 nM ProINS-Tf
(C)
10 nM ProINS-Tf +1 uM Holo-Tf
0 4 8
0
20
40
60
80
100
Time (h)
C-peptide (pM)
100 pM C-peptide
(E)
0
500
1000
1500
C-peptide (pM)
0 hr 400 pM C-peptide
24 hr medium only
24 hr H-4-II-E cells
(B)
0 2 4 6 8
0
200
400
600
Time (h)
C-peptide (pM)
10 nM ProINS
(D)
10 nM ProINS +1 uM Holo-Tf

38
were treated with 10 nM (C) ProINS-Tf or (D) ProINS and no obvious C-peptide increment
was observed regardless of TfR blockage. (E) The stability of C-peptide during incubation
with liver slices was tested. The residual C-peptide level was evaluated from the spiked
dosing medium containing 100 pM C-peptide over various time length of incubation.

39

Figure 2.5 Trypsin and Carboxypeptidase B treatment of ProINS-Tf and ProINS.
irINS (A) and C-peptide (B) were detected from reaction aliquots at indicated time points.
(C) The stability of C-peptide during enzymes treatment was also tested. The weight ratio
between substrate and enzyme was 1:1 and the final concentration (0.05 µg/µL) of both
enzymes were kept the same among all the experiment.

0 2 4 6 8 10
0
100
200
300
400
Time (min)
irINS (nM)
ProINS-Tf
ProINS
0 2 4 6 8 10
0
200
400
600
800
1000
Time (min)
C-peptide (nM)
ProINS-Tf
ProINS
0 2 10
0
20
40
60
80
100
Time (min)
%
C-peptide
(A)
(B)
(C)

40
2.4 Discussion
ProINS-Tf conversion via TfR mediated endocytosis has been thoroughly studied in
rat hepatoma cell line H-4-II-E. In this chapter, ProINS-Tf conversion was further
investigated on various TfR expressing cells and on in vitro and ex vivo liver models. A
noted discrepancy (Fig 2.1) in the conversion ability of ProINS-Tf among various TfR
expressing cells implied that, in addition to TfR expression, other necessary factors were
required for irINS-Tf generation. Conversion assays on PCLS culture further supported that
the liver was able to convert ProINS-Tf to irINS-Tf. Our previous studies on H-4-II-E cells
have pinpointed the intracellular venue for conversion was specifically to an acidic
compartment along the TfR-mediated slow recycling pathway [6]. When the TfR slow
recycling pathway was disrupted using hypothermia (16° C), lysosomotropic agents or
microtubule-disrupting reagents, ProINS-Tf conversion was drastically inhibited. Moreover,
in vitro Trypsin digestion assay indicated that irINS-Tf generation involved Trypsin-like
enzymes with restricted cutting sites after basic amino acids, rather than non-specific
protease degradation [7].
Minimal conversion of ProINS-Tf was observed during incubation with Hep G2, a
human hepatoblastoma-derived cell line [81] (Fig 2.1). Conversion assays along with
Bacitracin treatment suggested that INS degradation was not the main cause of the
conversion difference compared with H-4-II-E cells (Fig 2.2). Hep G2 was reported to
contain extremely low amounts of hepatic metabolism enzymes [82] and altered TfR
recycling kinetics [83]. These varied levels possibly provided an explanation and supported
the requirement of a specific endoprotease for conversion, such as one in the PC family.
PCLS is a practical tool to simulate the in vivo hepatic metabolism of xenobiotics

41
[84]. PCLS brings greater data reproducibility than liver homogenate and enables drug
metabolism studies on disease related liver models, which is more advantageous than
primary hepatocyte culture [69]. According to literature, the optimal thickness for each liver
slice is around 300 µm, which ensures sufficient oxygen and nutrients exchange between
tissue and medium as well as avoids excessive damage due to mechanical slicing [69,85]. In
our study, PCLS delivered a better estimation of the conversion rate and capacity in
exploring the liver's ability to process ProINS-Tf. After 8 h incubation, 1% of ProINS-Tf was
converted to irINS-Tf and released to the medium. To exclude possible interference from the
viability of liver slices, experiments were not tested beyond 8 h. In order to test irINS-Tf
production after longer incubation, relayed PCLS treatment could be applied by
consecutively transferring the incubation medium to another batch of freshly prepared PCLS
at the end of every 8 h.
The fact that C-peptide was not detected during ProINS-Tf conversion on H-4-II-E
cells and PCLS culture suggested that the processing of ProINS in hepatocytes was different
from that in the regulated secretory pathway of pancreatic beta cells. One of the possibilities
is, after cleavage, the byproduct C-peptide or C-peptide like fragment is released in the
recycling compartments and then delivered to lysosomes for degradation. However, the
irINS-Tf still binds to TfR in the recycling compartment and is then recycled back to the cell
surface. The C-peptide signal is possibly unable to be captured if one end of the peptide is
still attached to the fusion protein. The possibility that C-peptide is generated and then
deposited in the degradation pathway could be further investigated by utilization of specific
inhibitors against lysosome degradation or TfR recycling. For example, the V-ATPase
inhibitor Bafilomycin selectively inhibited the transport of marker destined for lysosomal

42
degradation in early endosomes, whereas the transport of Tf to the perinuclear recycling
compartment still occurred [86]. If C-peptide is identified in medium after Bafilomycin
treatment in experiment 2.B.7.a, this could validate the hypothesis. Another possibility is
conversion of ProINS-Tf solely involves one single cutting step at either B-chain /C-peptide
junction or C-peptide/A-chain junction, which doesn’t yield free C-peptide in either scenario.
The resulted conversion product irINS-Tf could be either des 31,32 ProINS-Tf (cut at B/C
junction) or des 64,65 ProINS-Tf (cut at C/A junction). Data from in vitro enzymatic
reactions (Fig 2.5B) also suggested that ProINS-Tf, compared with ProINS, might acquire
increased enzyme resistance possibly due to the steric hindrance by the Tf moiety. Based on
the information provided in INS RIA kit, by which we used to monitor and define the
conversion product, the applied human INS antiserum's recognition ability for des 64,65
ProINS is as high as 76% and for des 31,32 ProINS is less than 0.2%. In addition, the human
ProINS RIA kit can recognize des 31,32 ProINS (95%), not des 64,65 form (<0.1%). If the
converted form was Des 64,65 ProINS-Tf (which means the N-terminal of C-peptide is
attached to INS B-chain, leaving the C-terminal of C-peptide exposed), the readings from
both INS and ProINS RIA are also in accordance with observed phenomenon. Therefore, Des
64,65 ProINS-Tf and INS-Tf are less discernable in determining the actual form of irINS-Tf
simply based on RIA assays (Fig 2.6). Moreover, cross-activity information of C-peptide
from different species provided in the RIA manual suggested that both ends of C-peptide are
indispensable in the antibody recognition. So the C-peptide signal is possibly unable to be
captured if one end of the peptide is still attached to the fusion protein.
For future studies, considering the stability of C-peptide during Trypsin and
Carboxypeptidease B treatment, more informative results could be generated after ratio

43
optimization between substrate and enzyme. In addition, in order to identify the absolute
form of irINS-Tf, Des 64,65 ProINS-Tf is a highly likely candidate. However, other studies
on ProINS concluded that the B/C cleavage was more favorable (around 65%) than the
alternative C/A cleavage (around 35%) [78]. All the results of C-peptide detection assays
(section 2.B.7) indicated that C-peptide could not be used as a marker to study the
conversion kinetics. In addition to the reported data in this chapter, other C-peptide detection
methods, which have been tried but failed to detect C-peptide included: quantitative MS/MS
of C-peptide and MW characterization of Trypsin/CPB treated ProINS-Tf by MALDI-TOF,
and CPB treatment of H-4-II-E pretreated ProINS-Tf solution, which did not render different
readings on C-peptide RIA. The attempt to identify the INS-B chain before and after ProINS-
Tf conversion was ceased due to the fact that the tested antibody (M093-3, Anti-human
Insulin B chain mAb) was not able to recognized reduced INS-B chain and had significant
cross activity toward ProINS-Tf itself.

44

Figure 2.6 Schematic graph of hypothesized conversion steps, conversion intermediates, and
irINS-Tf formation.
Conversion of ProINS-Tf in the TfR mediated slow recycling pathway possibly requires
Trypsin like enzyme as well as one trimming enzyme functioning as Carboxypeptidase E.
Combined the current data from conversion and C-peptide studies as well as the antibody
recognition information provided in the RIA kits, two most possible candidates of irINS-Tf
are INS like-Tf and Des 64,65 ProINS like-Tf. Modified from the processing of human
ProINS to humanINS[50].

45
Chapter 3 Evaluation of liver selective action of Proinsulin-
Transferrin

3.1 Background
INS plays a major role in hepatic glucose metabolism by suppression of glucose
production (gluconeogenesis and glycogenolysis) as well as increase of glucose utilization
and storage (glycolysis and glycogenesis), thus restricting the net hepatic glucose output
(NHGO) [87]. INS regulates NHGO either through altering key enzyme expression levels via
transcription factors (e.g., INS down-regulates glucose-6-phosphatase mRNA via
IRS/PI3K/Akt/FoxO1 pathway [88]) or by changing the enzymes activity through
phosphorylation status (e.g., INS activates glycogen synthase by dephosphorylates activation
[87].
In the pathology of diabetes mellitus, the main cause for fasting hyperglycemia is the
aberrant hepatic glucose production [89]. Two contributing processes of HGO include
gluconeogenesis, i.e., de novo glucose synthesis from non-carbohydrate precursors, and
glycogen breakdown (glycogenolysis). In the non-INS-dependent diabetes mellitus
(NIDDM) patients, the former pathway appeared to be more dramatically increased
(p<0.001) than the latter (p=0.057) when in comparison with normal subjects [90]. Further
more, it was reported in hepatic UDP-glucose pool of glycogen synthesis that the indirect
pathway (the glycogenesis substrate UDP-glucose is converted from carbon 3 compounds via
gluconeogenesis) had a larger contribution than the direct pathway in which UDP-glucose
was derived from the uptake glucose [91]. The increased flux via the indirect pathway was
likely caused by an augmentation of the availability of carbon 3 compounds, which is also
indicated by an increased gluconeogenesis [92]. The authors further showed that exogenous

46
INS which was able to normalize the synthesis rate of hepatic glycogen, but failed to reduce
the ratio of newly synthesized glycogen derived from gluconeogenesis substrates [92].
Therefore, inhibiting the excessive hepatic glucose production, especially gluconeogenesis, is
of indispensable value to restoring normal BG.
INS stimulates glycogen synthesis and approximately 20% and 30% of the
carbohydrate intake is stored in the form of glycogen in liver and muscle, respectively, after a
mixed meal [93]. Defects in meal-time glycogen synthesis increase postprandial
hyperglycemia. 13C nuclear magnetic resonance (NMR) spectroscopy revealed that Type 1
diabetic patients exhibited a defective net hepatic glycogen synthesis and the defect was most
pronounced after meal ingestions [91]. Glucagon is a representative hormone inducing BG
elevation and counteracts INS's action. Glucagon increases HGP mainly through stimulating
glycogen breakdown, rather than gluconeogenesis [93]. Since INS's intrapancreatic inhibition
against glucagon secretion is destructed during Type 1 diabetes progression, a rise of the
glucagon-to-INS ratio in hepatic portal vein attenuates glycogen storage in liver [94].
Impaired enzyme activitiy of glycogen synthase (GSase) is one of the intrinsic factors
resulting in less glycogen accumulation in the liver. GSase conducts the final step of
glycogen chain elongation, during which a glucose unit is transferred from UDP-glucose to
glycogen [88, 95]. GSase is activated by dephosphorylation at multiple serine sites. The
active GSase (I form) and inactive form (D form) are converted by
phosphorylation/dephosphorylation. Also, GSase activity is tuned by the metabolite inputs.
Glucose-6-phosphate (G6P) is an allosteric activator, and in vitro even the most highly
phosphorylated GSase (D form) is activated by 10 mM or more G6P. However, only the I
form is considered to be the active form under physiological conditions [96]. In addition,

47
GSase activity is affected by its glycosylation degree. Studies on STZ-induced diabetic mice
have shown that hyperglycemia led to increased O-linked N-acetylglucosamine (O-GlcNAc)
modification on GSase, which increased its resistance to INS and decreased I-form
percentage compared with normal controls (p < 0.01) [97]. INS activates GSase via
inhibiting GSK-3, which inactivates GSase by phosphorylation.
In terms of peripheral INS sensitive tissues, skeletal muscle and adipose tissues are
the main sites for INS dependent peripheral glucose disposal (PGD). Specifically, INS
stimulates the translocation of glucose transporter GLUT4 to the plasma membrane and
therefore, facilitates the glucose uptake. In comparison, glucose distribution to the brain is
mediated by insulin-independent GLUT3 with a low Km (1.8 mM), which is saturated under
physiological conditions (4-7 mM in normal subjects) [98,88].
In this chapter, the hepatic INS effect of ProINS-Tf was investigated from several
aspects: receptor activation (IR phosphorylation), canonic INS signaling (Akt, GSase) and
INS's regulation in hepatic glucose metabolism (pyruvate tolerance test and tissue glycogen
content). Specifically, pyruvate tolerance assay was used to assess ProINS-Tf's inhibitory
potency against gluconeogenesis. Glycogen content from the liver and peripheral tissue
samples from STZ induced diabetic mice were compared among PBS, INS or ProINS-Tf
treated groups. Glycogen changes reflected ProINS-Tf's regulation of glycogen
accumulation.

48
3.2 Materials and Methods
3.2.1 Diabetic mouse model
Male C57BL/6J mice (6 weeks old) were purchased from Jackson Laboratory. The
protocols of the animal studies were in accordance with NIH guidance from “Guide for the
Care and Use of Laboratory Animals” and approved by the Institutional Animal Care and
Use Committee (IACUC) at USC. Animals were housed at 12 h light and 12 h dark cycle
under standard conditions (room temperature at 22 ± 3° C and relative humidity at 50 ± 20%)
with access to regular rodent chow (Labdiet) and water. For studies on diabetic mice, STZ
induction procedures were modified from protocols in the Animal Models of Diabetic
Complications Consortium (AMDCC). In brief, after a 6 h morning fast, mice were
administrated with freshly prepared STZ (150 mg/kg) in sodium citrate buffer (100 mM, pH
4.5) via i.p injection. After injection, mice were maintained under standard care with free
access to food for 4-5 days to develop diabetic symptoms. Mice with BG over 300 mg/dL
after 2 h fasting were used for later investigations.

3.2.2 Comparison of IR phosphorylation in liver versus skeletal muscles
Diabetic mice, after 2 h fasting, were randomly grouped to receive an s.c injection of
either PBS, INS (22.5 nmol/kg) or ProINS-Tf (22.5 nmol/kg) respectively. Mice were then
kept restrained from rodent food except for water and BG was checked by tail vein with
OneTouch® glucose meter (BG assay range of 20 – 600 mg/dL, Lifescan Technologies) at 1
h and 8 h post-injection. At each time point, after BG measurement, half of the mice in each
treatment group were euthanized for tissues collection. Perfused liver and skeletal muscles
were dissected immediately, snap frozen in liquid nitrogen and were stored at -80° C for later
analysis.

49
Samples from the main lobes of liver and gastrocnemius muscle were grinded by
mechanical homogenization in dissolving buffer containing protease inhibitors (P2714,
Sigma) and phosphatase inhibitors (NaF and Na3VO4). Supernatants of homogenized
samples, after agitation and centrifugation (12,000 x g for 25 min, 4° C ) were used for
protein quantification by BCA protein assay (Thermo) according to bovine serum albumin
standards. IR in tissue homogenization samples was immunoprecipitated after incubation
with polyclonal antibody against IR beta unit (sc-711, Santa Cruz Biotechnology) and then
protein A/G-coupled agarose beads (SC-2003, Santa Cruz). The immunocomplex was
released from beads after boiling and subjected to 10% SDS-PAGE followed by Western blot
against anti-phosphotyrosine (Clone 4G10 05-321, Millipore) or anti-IR antibodies. Band
densities were measured using Quantity One (Bio-Rad, Hercules, CA) software.

3.2.3 Pyruvate tolerance test (PTT)
PTT was performed on normal 6-7 weeks old male C57BL/6J mice. After 6 h fasting,
mice were randomly grouped to receive s.c injection of respective treatment (saline, 22.5
nmol/kg INS or 22.5 nmol/kg ProINS-Tf). Mice were then kept on an overnight fast (12 h)
for hepatic glycogen depletion. Afterwards, the main precursor of hepatic gluconeogenesis,
sodium pyruvate (Sigma) was i.p injected to mice (2 g/kg body weight). BG levels were
recorded at indicated time points following pyruvate bolus using the same method described
above. Mice were recovered with rodent chow after PTT.

50
3.2.4 Measurement of tissue glycogen
The experiment was carried out on a diabetic rodent model as described in section
(3.B.1). Free-fed mice were s.c-injected with either saline, INS (22.5 nmol/kg), or ProINS-Tf
(22.5 nmol/kg), and the liver tissues were isolated at 5 or 17 h post-injection. Perfused liver
was homogenized and the glycogen content was measured as previously described [99].
Briefly, tissues were hydrolyzed with hot alkaline solution (33% KOH), and the glycogen
was precipitated with ethanol. The glycogen was then dissolved in distilled water and reacted
with 0.2% anthrone in 96% sulfuric acid to yield a blue-colored compound. The absorbance
of the solutions was measured at 620 nm. The glycogen concentration was calculated against
a D-glucose standard curve.

3.2.5 Measurement of glycogen synthase (GSase) enzyme activity
H-4-II-E cells were seeded on 6-well culture plates 2 days before experimentation.
Confluent cells were washed with pre-warmed PBS and cultured in low glucose (5 mM)
supplemented DMEM with reduced FBS (0.5%) for 24 h to deplete glycogen. Then cells
were either treated with PBS, INS (100 nM) or ProINS-Tf (100 nM) for 4 h at 37° C.
Following the treatment, cells were rapidly washed with ice old PBS and collected using cell
scrapers into cell lysis buffer containing 50mM HEPES (pH 7.8) 10mM EDTA, 100 mM
NaF, 0.5% Triton X-100, 1 mM PMSF and protease inhibitors cocktail (P2714, Sigma).
Supernatant of cell lysates (5,000 rpm, 10 min at 4° C, Eppendorf 5415R microcentrifuge)
was added into GSase working buffer to test the enzyme activity. GSase working buffer was
composed of Krebs Ringer buffer with 30mM HEPES (pH 7.4), 12 mM UDP-glucose, 0.2
µCi UDP-[6-3H] glucose (60 µCi/mmol, American Radiolabeled Chemicals, Inc) and 16

51
mg/mL rabbit liver glycogen, type III (Sigma). For total GSase activity, additional 25 mM
Glucose-6-Phosphate was added. Reactions were performed in a 37° C water bath for 2 h.
After that, 0.6 N perchloric acid was added to stop reactions. Glycogen was eluted and
separated from unincorporated tracer using Sephadex G-50 mini quick spin column (Roche;
1,000 x g, 4 min). After mixing with OptiPhase Supermix Cocktail scintillation fluid (Perkin
Elmer), the amount of incorporated radiolabeled UDP-glucose in the isolated glycogen was
measured using 1450 MicroBeta Trilux microplate scintillation and luminescence counter
(Perkin Elmer). Fractional GSase activity (active/total) was determined using equations
modified from reference [100]. Glucose integration (GI) into glycogen was calculated as
GI=(DPM
sample
−DPM
basal
)/DPM
total
. Glycogen synthase activity (GSA) was then calculated as
GSA=GI
lowG6P
/GI
highG6P
. Relative GSA was then calculated as Ratio=GSA
sample
/ GSA
PBS
.

3.2.6 Time courses of Akt dephosphorylation
H-4-II-E cells were seeded on 12-well plates and cultured in DMEM with 10% FBS.
Prior to experiments, the culture medium was replaced with DMEM supplemented with 0.1%
BSA for overnight serum starvation. Cells were then treated with equal molar (1 nM) of
ProINS-Tf, ProINS or INS in serum free DMEM for 4 h at 37° C. After treatment and
extensive wash, the dosing medium was replaced with fresh DMEM. At the indicated time
points, cells were solubilized in a cell extraction buffer (Life Technologies) containing
phosphatase inhibitors NaF and Na3VO4, and additional protease inhibitors cocktail (P2714,
Sigma) and PMSF. Cell lysates were extracted by centrifugation (13,000 x g, 10 min) and
subjected to 10% SDS-PAGE. The residual phosphorylated Akt was detected by anti-
Phospho Ser473-Akt antibody (Cell Signaling Technology). Band densities were quantified

52
by Image Lab™ Software (Bio-Rad). Akt phosphorylation was normalized with
glyceraldehyde-3-phosphate dehydrogenase (GAPDH) detected by anti-GAPDH (D16H11)
XP® antibody (Cell Signaling Technology) based on band densities.

3.3 Results
3.3.1 Preferential hepatic IR activation
After s.c bolus of protein treatment on STZ induced diabetic mice, IR activation (i.e.,
phosphorylation) was compared in the liver versus skeletal muscles, which is a representative
of peripheral INS sensitive tissues. Consistent with the short time frame of INS's
hypoglycemic efficacy, IR phosphorylation was detectable in liver and muscle tissue only at
1 h, but not 8 h post-injection. ProINS-Tf treated mice, on the other hand, showed a
preferential IR activation in liver and this activation was prolonged to at least 8 h (Fig 3.1A–
C) [41].

3.3.2 Inhibition of hepatic gluconeogenesis
Our previous study showed that ProINS-Tf displayed a longer BG lowering effect
than INS after s.c bolus on STZ induced diabetic mice [41]. To evaluate the inhibitory effect
of ProINS-Tf against HGO, a PTT was conducted 12 h after protein injection. INS with short
action duration did not show apparent restrictions of HGO (Fig 3.2). In both saline and INS
groups, pyruvate induced peak BG elevation around 120 mg/dL. In contrast, ProINS-Tf was
able to inhibit 50 % of BG increase, therefore showing a prolonged effect on the inhibition of
HGO.

53
3.3.3 Promotion of hepatic glycogen accumulation during feeding
The effect of ProINS-Tf on stimulating glycogen synthesis was also measured under
free-feeding conditions. Following s.c injection of either saline, INS, or ProINS-Tf, liver
glycogen levels were measured at time points with similar BG levels (i.e., at 5 and 17 h post-
injection). As shown in Fig 3.3A and Fig 3.3B, there was no significant difference in
glycogen levels for either saline or INS-treated groups. However, in the ProINS-Tf group,
hepatic glycogen levels significantly increased at 17 h compared with 5 h post-injection (Fig
3.3C). In addition, glycogen levels in skeletal muscles were not notably altered among all the
treatments (Fig 3.3D).

3.3.4. Glycogen synthase activation by ProINS-Tf in H-4-II-E
To test GSase activity, measurement of enzyme substrate incorporation rate using
tritium labeled UDP-glucose was chosen over the photometrical method, i.e., measurement
of the released UDP. The former method has improved sensitivity in samples with small
volumes [101]. A convenient gel filtration method other than the conventional filter paper
method was adopted to separate newly synthesized glycogen form free UDP-glucose [100].
The incubation time was chosen to be 2 h as the reference [100] showed that the
incorporation rate of glucose into glycogen was linear for up to 240 min of incubation and
the glycogen recovery rate was over 95%. Formation of UDPH during the reaction progress
was shown not to interfere with enzyme activity given a broad pH profile of GSase (EC
2.4.1.11) [102]. The reaction was carried out in HEPES buffer to prevent potential enzyme
inhibition caused by inorganic phosphate [96]. Active GSase (I form) after INS and ProINS-
Tf 's 4 h stimulation both increased significantly compared to the control group (Fig 3.4). The

54
increase was comparable between INS treated group (148.8 ± 0.3%) and ProINS-Tf treated
group (139.1± 18.2%). It was also noticed that the comparison of in vitro GSase activation
efficacy between INS and ProINS-Tf was also accordingly consistent with the phospho-Akt
level after 4 h treatment [40]. To avoid any potential inhibition of GSase by newly
synthesized glycogen in H-4-II-E cell during treatment [103], INS or ProINS-Tf was
dissolved in glucose deprived solution. Therefore, a longer treatment over 4 h was not
adopted in order to maintain cell viability.

3.3.5 Time-lapse INS signaling after ProINS-Tf treatment
After the 4 h stimulation (the initial time point 0 h in Fig 3.5), treatment with ProINS-
Tf and with INS induced significant phosphorylation of Akt. When the dosing solution was
removed and cells were starved in serum free medium, the residual Akt phosphorylation in
INS group was rapidly decreased and returned to basal levels within 1 h. On the other hand,
the phosphorylated Akt in ProINS-Tf-treated cells was maintained at a similar level and
lingered for at least 4 h after removing the fusion protein (Fig 3.5). ProINS did not lead to
notable phosphorylation of Akt at the experimental dose (1 nM) due to its low potency to
activate IR. When 1 nM Tf was added after removal of the extracellular protein treatment
(Fig 3.6), the time course of residual Akt dephosphorylation was not affected in either INS or
ProINS-Tf group.

55

Figure 3.1 IR phosphorylation in the liver versus muscle.
Mice were fasted for 2 h prior to s.c injection with PBS (vehicle), ProINS-Tf, or INS. Liver
and muscle tissues were collected and homogenized at 1 and 8 h post-injection[41]. A: IR
was isolated by immunoprecipitation using anti-IR antibody and subjected to 10% SDS-
PAGE followed by Western blot against antiphosphotyrosine or anti-IR antibodies. The bar
graphs represent the quantified band densities from the Western blot analysis of (B) liver and
(C) muscle samples. pIR, IR phosphorylation.

56

Figure 3.2 BG increase after pyruvate bolus.
After respective treatment (Saline, 22.5 nmol/kg INS or 22.5 nmol/kg ProINS-Tf), mice were
restrained from food for 12 h. A single dose of sodium pyruvate (2g/kg) was then i.p injected
and BG levels were recorded at indicated time points. Data were presented as the average
value of BG changes in each group (n=4). (BG increment = BG at X time point -initial BG
level before pyruvate injection).

57

Figure 3.3 Measurement of glycogen levels in the liver and skeletal muscle.
Free-fed mice were injected subcutaneously with (A) Saline, (B) INS, or (C) ProINS-Tf (22.5
nmol/kg), and the liver tissue was isolated at 5 or 17 h post-injection. The liver tissue was
perfused and homogenized, and the glycogen content was measured. Individual data points (n
= 3) are represented as closed symbols, along with the mean ± standard deviation (SD)
(horizontal lines). Data marked with an asterisk indicate statistically significant differences
comparing different time points using Student's t test. **p < 0.01. (D) Glycogen content in
skeletal muscle was compared among three treatment groups.

Liver glycogen in saline group
5
17
0
50
100
150
Time (h)
ug/mg
Liver glycogen in ProINS-Tf group
5
17
0
50
100
150
Time (h)
ug/mg
**
Liver glycogen in INS group
5
17
0
50
100
150
Time (h)
ug/mg
5
17
5
17
5
17
0
5
10
15
Time (h)
muscle glycogen (ug/mg)
Saline
ProINS-Tf
Insulin
(B) (A)
(C) (D)

58

Figure 3.4 GSase activation in H-4-II-E cells.
Cells were treated with 100 nM INS, 100 nM ProINS-Tf or PBS in glucose deprived
condition for 4 h. GSase activation was assessed from cell lysates based on the ratio of I form
in the total GSase activity. The level of I form in PBS treated cells were set as 100%. Data
was taken the average values of two independent experiments, each of which has duplicate
samples.

PBS
INS
ProINS-Tf
100
120
140
160
GSase activation I/(I+D)ratio
(%)
PBS
INS
ProINS-Tf

59

Figure 3.5 Time course of Akt dephosphorylation.
H-4-II-E cells were initially exposed for 4 h with control (DMEM), 1 nM ProINS-Tf, 1 nM
INS or 1 nM ProINS. Phosphorylated Akt was determined by Western blot at 0, 1 and 4 h
after the removal of the treatment medium. Phosphorylated-Akt bands were normalized with
GAPDH band densities. Data were expressed as the average values and error bars indicated
SD (n=3).

60

Figure 3.6 Time course of Akt dephosphorylation with Tf chase.
H-4-II-E cells were initially exposed for 4 h with control (DMEM), 1 nM ProINS-Tf, 1 nM
INS or 1 nM ProINS. After that, cells were extensively washed and replenished with 0.1%
BSA supplemented DMEM with or without holo-Tf (1 nM) at 37° C. Phosphorylated-Akt
was determined by Western blot at 0, 1 and 4 h after the removal of the treatment medium.
Phosphorylated-Akt bands were normalized with GAPDH band density. Data were expressed
as the average values and error bars indicated standard error. (n=2).

Medium only ProINS-Tf Insulin
0.00
0.05
0.10
0.15
0.20
P-Akt/GAPDH
0 h
1 h
4 h
1 h+ Tf
4 h+Tf

61
3.4 Discussion
In INS deficient diabetes, aberrant regulation of glucose metabolism in the liver leads
to undesired glycemic levels as well as deleterious morbidity in macrovascular/microvascular
systems. Previous therapeutics have focused on PGD stimulation, however INS therapeutics
targeting HGP and hepatic glycogenesis not only resemble the physiological distribution
pattern of INS, but, more importantly, enable a stringent glycemic control without raising the
risk of severe hypoglycemia. In this chapter, ProINS-Tf displayed a preferential IR activation
in the liver rather than on peripheral tissues and a preferential stimulation of hepatic glycogen
accumulation. In addition, the prolonged INS action of the fusion protein was manifested in
multiple events both in vivo (PTT, phosphorylated IR) and in vitro (phosphorylated Akt).
Hepatoselectivity, in current studies, is mostly validated by pharmacodynamic
features rather than by direct comparison of INS distribution between the liver and peripheral
tissues. Direct measurement of IR activation level from the liver and skeletal muscle (Fig
3.1) provided an alternative approach to testing the Ra to Rd ratio. PTT assays (Fig 3.2)
supported that ProINS-Tf displayed efficient gluconeogenesis inhibition during prolonged
fasting. Another area significance was that the inhibition was still significant even after 12 h
post-injection. The mRNA level of two enzyme markers in HGP, phosphoenolpyruvate
carboxykinase and glucose-6-phosphatase were also significantly decreased [41]. During
free-feeding condition, hepatic glycogen in ProINS-Tf's was increased with little intervention
on glycogen deposition in skeletal muscles. As Geho pointed out [104], in order to achieve
effective control of postprandial BG, an INS therapeutic must enhance hepatic glycogen
storage, and so would alleviate the burden of PGD stimulation. As a result, less peripheral
acted INS could be required to normalize feeding BG. Secondly, during BG drops over

62
fasting, hepatic glycogen is hydrolyzed to supplement glucose in circulation. Therefore,
increased hepatic glycogen storage can cushion BG fluctuation to prevent fasting
hypoglycemia. In vitro studies (Fig 3.4) demonstrated that ProINS-Tf displayed a similar
potency in activating the key enzyme GSase in glycogen synthesis after 4 h stimulation. In
the in vivo settings, hepatic IR activation for ProINS-Tf was extended and reinforced more
than INS, therefore the potency for ProINS-Tf's GSase stimulatory is presumably more
pronounced in vivo, which was suggested by Fig 3.3.
Tf has long been used to improve the serum stability of fused protein drugs through
several pathways including re-routing through the TfR-recycling pathway, blocking
protease/peptidase degradation, and reducing renal elimination by increasing the molecular
weight [105,106]. Previous studies demonstrated that after Tf fusion ProINS's elimination
T
1/2
following s.c was extended from 0.5 h to 7.3 h. ProINS- Tf also showed higher area
under the curve (AUC) value than ProINS, suggesting an overall enhanced in vivo exposure.
These PK features partially contributed ProINS-Tf's prolonged hypoglycemic effect. In
addition, it was also subject to an increase in its intra-hepatocyte retention. As shown in Fig
3.5, both ProINS-Tf and irINS-Tf exploited TfR recycling pathway as intracellular
reservoirs, and led to time-lapse INS signaling. Together with result in Fig 3.6, extracellular
Tf did not abolish the lingering effect of Akt phosphorylation, implying that that endosomes
could also be a legitimate signaling platform for IR [107, 108]. After binding to IR, the
majority of internalized INS is degraded. However, a significant portion is released intact
from the cell via the retroendocytotic pathway [109]. This outcome, to some extent, is similar
to TfR mediated recycling. The fusion protein may able to take advantage of TfR recycling
pathways and therefore enhance its survival by the retroendocytotic route. Another

63
conceivable explanation for the prolonged effect of ProINS-Tf was the intrinsic features of
its action site. Studies on healthy human subjects demonstrated that deactivation of INS
suppression on HGO took two times longer than the disappearance of INS-induced PGD
after ceasing INS infusion [110]. Therefore, the prolonged efficacy of ProINS-Tf may be
related to its preferential action on HGO in the liver.
Results discussed in this chapter describe an overall evaluation of ProINS-Tf's effect
in de novo gluconeogenesis and glycogenesis. For a more in depth assessment about ProINS-
Tf's hepatoselectivity or the biological efficacy relative to INS, more sophisticated and
complex techniques are required to generate quantitative information. For example, the gold
standard approach to assess gluconeogenesis is by measuring the incorporation of deuterium
from the body water pool into newly formed glucose during a euglycemic hyperinsulinemic
clamp [111].

64
Chapter 4 Exploration of therapeutic potentials of Proinsulin-
Transferrin as basal insulin replacement in diabetes treatment

4.1 Background
Basal INS replacement in Type 1 diabetes patients is to mimic the continuous INS
secretion between meals and during night-time hours. The physiological secretion profile of
INS is the outcome superimposed by both high-frequency INS pulses (every 5-15 min) and
ultradian oscillations release (every 80-120 min) [112]. In the basal state, INS concentration
in the circulation is controlled within 10-30 pM due to extensive hepatic extraction [4].
Therefore, it raises strict bounds upon the PK profiles of basal INS therapeutics in the context
of delivery rate and extent. Advancements in the identification of amino acid regions in INS
hexamer formation and protein engineering techniques have allowed new approaches in
optimizing the PK profile of basal INS. Ideally, treatment renders a sustained and uniform
INS plateau at low concentration in plasma after administration. The first success of long-
acting INS analog, INS glargine (Latus®, Sanofi-Aventis) came to the market in the 2000s,
which was 10 years later than the first fast-acting INS analog INS lispro (Humalog®, Eli
Lilly and Company) [113]. At present, the main strategy to protract INS's action relies on a
delayed absorption phase after s.c bolus. The s.c depot effect has been achieved by either
isoelectric point shift (INS glargine), albumin binding (INS detemir) or increased tendency of
INS monomer self association (INS degludec) [114]. However, the sole effort in extending
the half-life of INS, without considering its distribution pattern in physiological release, is
not enough for a successful translation into a better INS therapeutics. For example, ultralente
INS, which has a longer duration (20 h) than NPH INS (14 h), was withdrawn from the
market in 2005 due to higher incidence of hypoglycemia [115]. Exogenous INS exerts a BG-

65
independent exposure to peripheral tissues. In other words, when BG drops below the normal
range, the impaired counteraction due to hyperglucagon and attenuated sympathoadrenal
response leads to BG falling and causes iatrogenic hypoglycemia [116]. Severe
hypoglycemia may cause lethal cardiac arrhythmias including premature ventricular
contractions, tachycardia, and high-degree heart block [117].
A basal INS therapeutic of the next generation should be able to maintain a flatter and
peakless PK/PD profile with 24 h coverage. One dilemma of marketed INS glargine and
detemir is dose titration. On one hand, the high dose would yield a peak on PK/PD profiles
and on the other hand lower dose inevitably shorten the duration of glycemic control [23].
The second requirement for basal INS delivery is mimicking physiological INS distribution.
In a recent comparison study about hypoglycemia incidence between portal (Po) vein
delivered INS and peripheral (Pe) vein delivered INS on glucagon response diminished
conscious dogs, result suggested that Pe delivered INS exacerbated hypoglycemia. Therefore,
basal INS delivery that mimics Po INS delivery (liver-preferential INS analogs) may lessen
the adverse effects [118].
Aside from achieving a desirable glycemic control, a myriad of INS therapeutic
candidates displaying prominent BG lowering efficacy were dropped from further
development due to suspiciously increased tumorigenesis tendency [119]. Both IR and
insulin-like growth factor-1 receptor (IGF-1R) belong to the receptor tyrosine kinase family.
The sequences and overall structures of the two receptors remain very similar. However,
their downstream signaling pathways after receptor activation diverge considerably [120]. IR
signaling mostly enables metabolic regulation in BG and lipid homeostasis. On the other
hand, IGF-1R activation largely triggers mitogenic related events and promotes cell survival,

66
proliferation, and migration [121]. Human IGF-1 is a single polypeptide composed of 70
amino acids and has three intramolecular disulfide bonds similar to INS. The 𝛼- helix in IGF-
1 is in the same location as that of INS B-chain. The binding pattern of IGF-1 to IGF-1R is
comparable to that of INS/IR, which involves two surfaces on the ligands, and many
equivalent amino acids were found on those regions between IGF-1 and INS [121,122].
Native INS is able to bind IGF-1R with a much weaker potency, which is less than 1% of
IGF-1 [123]. Unfortunately changes on the molecular structure of INS analogs as an attempt
to optimize their binding affinity to IR or INS monomer association may accidently alter the
binding properties toward IGF-1R and results in an increased mitogenic potential. One
example is Aspartic B10 (AspB10) INS, in which aspartic acid replaced the original 10th
histidine residue (Zinc binding site) at INS B-chain to reduce the hexamer formation. As a
result, the monomeric molecule has faster absorption than native INS and was designed as
fast-acting INS analog to control prandial BG [124]. However, it was withdrawn from
clinical development due to a higher incidence of breast cancer in rats [125]. In vitro
molecular characterization proved that AspB10 INS compared with native INS exhibited (1)
increased binding affinity to IGF1-R and (2) increased residence on IR [126]. Although the
specific mechanism by which AspB10 INS exerted its mitogenic effect is not clear, its
greater affinity to the IGF-1R is at least partly responsible for the increased tumorigenecity
[127].
Analogs derived from modified INS sequences may have some uncertainty in their
IGF-1R binding properties. In comparison, ProINS-Tf contains the natural sequence of
human ProINS. In contrast to INS, ProINS was proven not to bind IGF-1R [128]. In addition,
Tf fused at the C-terminal of A-chain may provide steric hindrances against the interaction

67
between the fusion protein and IGF-1R. Therefore, ProINS-Tf may offer an advantage over
other INS analogs by demonstrating a decreased mitogenic effect. For safety evaluation, use
of higher doses of INS analogs in traditional animal toxicological studies are hampered due
to premature animal death from hypoglycemia. Therefore, the approach has shifted to
molecular characterization of INS analogs in preclinical studies [119].
In addition, the risk for myocardial infarction will also be analyzed. Investigations
about the therapeutic potential of ProINS, which displayed minor hepatopreferential effect,
were halted due to concerns that long-term treatment of ProINS posed a risk for myocardial
infarction [129]. Although the causal relationship between ProINS and acute myocardial
infarction has not been definitive, increased ProINS concentration over physiological levels
in circulation has been suggested to be associated with atherosclerosis and cardiovascular
disease, as the level was elevated along with progression of T2DM and aging, both of which
were involved in higher risk of cardiovascular diseases [130,131]. In addition, ProINS levels
are also associated with increased activity of the plasminogen activator inhibitor-1(PAI-1),
the main regulator of fibrinolysis, and plasma PAI-1 level over 45 pM is considered a marker
of high cardiovascular risk [132]. As summarized in the review [133], PAI-1 predisposes
cardiovascular risks by promoting the nascent thrombi to form macroscopic thrombosis in
blood and by precipitating vulnerable plaques development at neointintima in tissues. Factors
that can upregulate PAI-1 expression include INS, ProINS, split ProINS, triglycerides, free
fatty acids, glucose and inflammatory cytokines [133]. INS increases PAI-1 expression by
reducing the degradation rate of PAI-1 mRNA [134] as well as by stimulating its
transcription factor FoxO3a [135]. Studies on Hep G2 cells suggested that ProINS split
products des (31,32) and des (64,65) ProINS processed higher potential in inducing PAI-1

68
protein secretion. The threshold concentration of ProINS’s agonist effect was 10 nM,
whereas the split forms were at 1 nM [136]. PAI-1 mRNA was identified in heart, lung,
vessel endothelium, vascular smooth muscles, adipose tissues, liver and kidney. Some studies
suggested that PAI-1 expression was more enriched in vascularized organs (heart, lung) and
adipose [137,138]. Conversely, some studies concluded that hepatic PAI-1 synthesis was a
main source of plasma PAI-1[139,140]. In vivo studies on mice shown that i.p injected INS
increased plasma PAI-1 as well as tissue PAI-1 mRNA. Among tissues with observed
increase of PAI-1 mRNA levels, heart and adipose tissues were most susceptible to INS
stimulus [137]. It is not known whether or not the Tf-fused ProINS will elicit the same PAI-
1-stimulatory activity compared with INS and ProINS.
In this chapter, the main focus of the studies included two subsections: (1) evaluation
of ProINS-Tf's glycodynamic features by comparison with current long-acting INS analog
INS glargine and (2) in vitro evaluation of related safety concerns (mitogenic potential and
PAI-1 secretion).

4.2 Materials and Methods
4.2.1 Hypoglycemic efficacy study under prolonged fasting or free-feeding
Diabetic mice were induced using the same method described in section 3.B.1.
Diabetic mice were randomly grouped to receive either saline, INS, or ProINS-Tf via s.c
injection. For INS glargine treatment, the injection solution at low dose (22.5 nmol/kg) was
prepared by diluting Lantus® (600 nmol/mL, Sanofi-Aventis) in 0.85% saline buffer. The
original formulation was applied when higher dose (135 nmol/kg) was s.c injected. To test
ProINS-Tf's control of fasting glycemic level, mice were fasted after protein treatment. To

69
test ProINS-Tf's glycodynamic effect during free-feeding, mice were provided free access
with regular rodent chow during the entire experiment. BG levels were checked through tail
veil blood samples by OneTouch® glucose meter (BG assay range of 20 – 600 mg/dL,
Lifescan Technologies) at indicated time points. BG levels lower than 30 mg/dL were
considered as severe hypoglycemia [141].

4.2.2 Dose response study
ProINS-Tf (67.5 nmol/kg, 135 nmol/kg and 202.5 nmol/kg), INS glargine (Lantus®,
Sanofi-Aventis) (135 nmol/kg) or saline (control) were s.c-injected to diabetic mice in
respective groups. INS glargine was injected without dilution of the original formulation of
Lantus®. BG levels were recorded at indicated time points following treatment bolus using
the same method described above.

4.2.3 Hypoglycemic efficacy study under fast/feed cycles
ProINS-Tf (135 nmol/kg), INS glargine (135 nmol/kg) or saline solution was s.c-
injected in diabetic mice in each group. Mice after injection were subjected to 8 h fasting/o8
h free-feeding cycles. BG levels were monitored at the end of each fasting or feeding phase
using OneTouch® glucose meter.

4.2.4 L-Alanine tolerance test and L-Glutamine tolerance test
Tissue specific gluconeogenesis substrate tolerance test was performed on normal 6-7
week old male C57BL/6J mice. Before the gluconeogenesis substrate challenge, mice were
kept under 24 h fast for hepatic glycogen depletion. Afterwards, the main precursor of

70
hepatic gluconeogenesis, L-Ala (Sigma) or intestinal/renal gluconeogenesis substrate L-Gln
(Sigma) was administered (1.5 g/kg) by i.p injection. During the 24 h fasting, mice were
randomly grouped to receive s.c injection of respective treatment (saline, 22.5 nmol/kg INS
or 22.5 nmol/kg ProINS-Tf) at indicated time prior to substrate injection (6 h or 12 h).
Gluconeogenesis substrate solutions were prepared by dissolving L-Ala or L-Gln in 0.85%
isotonic saline solution and the solutions were sterilized by filtering through 0.2 µm syringe
filters (VWR). At indicated time points following substrate bolus, BG levels were recorded
using the same method described above.

4.2.5 Glycodynamic control of premixed ProINS-Tf and INS
Diabetic mice after a 6 h fast were randomly grouped to receive a s.c bolus of
respective treatment (saline, 12 nmol/kg INS or premixed ProINS-Tf and INS (22.5/ 12
nmol/kg)). The mice were then provided rodent chow for 4 h. After the short feeding periods,
mice were transferred to new cages and kept under fasting conditions for 8 h. BG levels were
monitored at indicated time points to evaluate the hypoglycemic efficacy during meals and
under extended fasting phase.

4.2.6 Glycodynamic control of repeated dose of ProINS-Tf /INS mixture
Diabetic mice after a 2 h fast were randomly grouped to receive a s.c bolus of
respective treatment (saline, 45 nmol/kg INS, premixed ProINS-Tf and INS (22.5/22.5
nmol/kg) or 45 nmol/kg ProINS-Tf alone). The mice were then provided rodent chow for 4 h.
After the short feeding periods, mice were transferred to new cages and kept under fasting
conditions for 8 h. At the end of the fasting phase, mice in each group were received the

71
second s.c bolus of previous treatment. Afterwards, mice were put under the same feed/ fast
cycle. BG levels were monitored at indicated time points to evaluate the hypoglycemic
efficacy during meals and under extended fasting phase.

4.2.7 IGF-1R activation assay on MCF-7 cells
MCF-7 cells, after overnight serum starvation, were treated with various
concentration of either INS or ProINS-Tf in 0.1% BSA DMEM for 30 min. After treatment
and extensive washing with ice-cold PBS, cells were solubilized in cell extraction buffer
(Life Technologies) containing phosphatase inhibitors sodium fluoride and sodium
orthovanadate, plus additional protease inhibitor cocktail (Sigma) and 1 mM PMSF. Cell
lysates were extracted by centrifugation (13,000 x g, 10 min) and subjected to human IGF-1R
[pY1135/pY1136] Phospho-ELISA Kit (KHO050, Thermo) to determine the level of IGF-1R
phosphorylation. Cell protein was quantified by BCA kit for normalization.

4.2.8 MCF-7 cells proliferation assay
MCF-7 Cells were seeded in triplicates in 96-well plates (5,000 cells per well). Cells
after 24 h serum starvation were treated with 10 nM of human INS, holo-Tf or ProINS-Tf in
phenol red free RPMI medium supplemented with 0.1% BSA (day 1). This treatment was
repeated in two consecutive days (day 2 and day 3). On day 4, cells were incubated for 4 h at
37° C with 0.5 mg/ml 3-(4,5-dimeth- ylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)
prior to cell lysis by acidified isopropanol (4 mM HCl, 0.1% Nondet P-40). Formazan dye
production was measured using Infinite® F200 (TECAN) at 570 nM. Absorbance readings
were normalized with control group cultured in 0.1% BSA RPMI and the OD value was

72
referred as 100%. For the H-4-II-E cell pretreated set of protein treatment, 10 nM of the
aforementioned protein was incubated with confluent H-4-II-E cells for 24 h. Afterwards, the
supernatant of the dosing medium was collected and applied to MCF-7 cells for proliferation
assays. For the ProINS-Tf group, aliquots of the medium after H-4-II-E incubation were
applied to human INS RIA kit to measure the irINS-Tf level.

4.2.9 PAI-1 secretion assay on Hep G2 cells
Monolayers of confluent Hep G2 cells were serum-starved in DMEM supplemented
with 0.1% BSA for at least 16 h. The 16 h serum starvation was able to lower the PAI-1
synthesis to basal level [136]. After serum starvation, cells were exposed to respective
protein treatment prepared in 0.5% BSA supplemented DMEM. Protein treatment groups
included 10 nM human INS, 10 nM human ProINS and 10 nM ProINS-Tf. Supernatant
medium was collected at indicated time points and was applied to human PAI-1 ELISA kit
(KHC3071, Thermo) for PAI-1 measurement. Cell protein was quantified by BCA kit for
normalization.

4.2.10 Data analysis
BG levels were expressed as the group mean value (n ≥ 3) with SD. Data were
analyzed by Student's t-test, where P< 0.05 was considered as statistically significant.

73
4.3 Results
4.3.1 Hypoglycemic duration of ProINS-Tf during prolonged fasting
The hypoglycemic effect of ProINS-Tf was compared with the long-acting INS
analog, INS glargine (Fig 4.1)[41]. At the lower dosage of 22.5 nmol/kg for the three
treatments (human INS, INS glargine and ProINS-Tf), INS glargine showed a fast onset with
a short-term hypoglycemic response. The resulted BG curves were mostly superimposable
with equal molar INS treatment. At a five-fold higher dose (135 nmol/kg), INS glargine
exhibited a more prolonged hypoglycemic effect at least 12 h post-injection. However, the
mice suffered from severe hypoglycemia with BG levels < 30 mg/dL, especially within the
initial 2 h phase. Meanwhile, ProINS-Tf exhibited a late onset and sustained hypoglycemic
effect at both 22.5 and 135 nmol/kg doses and no signs of severe hypoglycemia were
observed.

4.3.2 Hypoglycemic effect of ProINS-Tf during free feeding
From Fig 4.2, ProINS-Tf fusion protein alone (22.5 nmol/kg) displayed minor BG
lowering effect during free-feeding condition. The 20% BG reduction in ProINS-Tf group
compared with saline group was possibly contributed by improved hepatic glycogen
synthesis as well as inhibition against HGP. The hypoglycemic effect of INS (22.5 nmol/kg)
was most pronounced at 1 h post-injection and brought down BG near euglycemia (110
mg/dL). However, INS's glycemic control was rapidly diminished and the BG level at 3 h
was no statistically different from that in saline group.

74
4.3.3 Hypoglycemic efficacy of ProINS-Tf during fast/fed cycles
To explore the duration of the BG lowering effect of ProINS-Tf and its glycemic
control between prandial and basal conditions, BG levels were monitored during the fast/feed
cycles on diabetic mice after s.c injection of INS analogs. As shown in Fig 4.3, INS glargine
at 135 nmol/kg induced severe hypoglycemia at 2 h post-injection, and lost efficacy after the
first fast/feed cycle at 18 h post-injection. ProINS-Tf exhibited a 20-40 % BG lowering
activity during feeding phases and was unable to normalize feeding BG. However, an
effective and sustained blunting effect of fasting hyperglycemia was observed at 40 h post-
injection. In comparison with saline control, BG levels in ProINS-Tf group at the end of
three subsequent fasting phases, i.e., 8, 24 and 40 h, were 126±52, 99±30, 162±41 mg/dL
respectively, while BG in the control group were all above 300 mg/dL.

4.3.4 Dose response study of ProINS-Tf
In the dose response study (Fig 4.4), after ProINS-Tf treatment, BG gradually
decreased at similar rates among three doses in the first fasting phase. Even the highest dose
of ProINS-Tf (202.5 nmol/kg) did not cause severe hypoglycemia in mice. In contrast, INS
glargine (135 nmol/kg) induced an instant BG reduction, which was reflected as a steep slope
on the BG curve. All mice (n=5) in the INS glargine group suffered from severe
hypoglycemia and were not recovered until after 2 h feeding. Additionally, INS glargine
exerted a robust inhibition of hyperglycemia during both fasting and feeding conditions. BG
during the initial 4 h feeding was controlled around euglycemia. However, its hypoglycemic
effect diminished after 12 h post-injection, and BG bounced back to hyperglycemia with no
significant difference from saline group. In contrast, the BG lowering effect of ProINS-Tf

75
was more prominent during fasting than feeding. At 4 h post-injection, three doses of
ProINS-Tf exerted similar efficacy and maintained fasting BG around normal level at
subsequent fasting time points. During the feeding phase, only the highest dose of ProINS-Tf
was able to cause a 40% BG reduction compared with saline group. The dose dependent
effects of ProINS-Tf manifested on glycemic control at later time points. Compared with the
saline group, BG reduction observed after the second fasting (24 h) was more pronounced at
202.5 nmol/kg ProINS-Tf (81%) than those lower doses (40-45% BG decrease). Therefore,
ProINS-Tf exhibited a dose-dependent efficacy in lowering BG levels.

4.3.5 Influence of systematic gluconeogenesis of ProINS-Tf
Alanine is the specific gluconeogenesis (GNG) substrate for liver while glutamine is
the specific substrate for GNG in intestine and kidney [142,143]. By comparing BG changes
after organ specific GNG substrate challenges, ProINS-Tf's influence on hepatic versus
extrahepatic GNG was evaluated. The relationship between ProINS-Tf’s hypoglycemic
potency and duration post-injection (6 h vs 12 h) was also explored. In the experimental
setting, due to the limited solubility of L-Gln (36 mg/mL in water, 25°C), the i.p injection
volume for L-Gln tolerance test was as high as 1 mL. To avoid potential interference on BG
due to the large volume of injected solution, one group of mice, instead of receiving GNG
substrate, was injected with equal volume of saline solution and their BG fluctuation served
as the negative control to substrates-induced BG changes. Result of comparing BG curves
between Saline-Gln and Saline-Saline group (Fig 4.5B) suggested that i.p injected saline
induced BG change summited at 30 min post-injection and substrate induced BG peaks
appeared at 1 h post-injection. Similar phenomenon was also observed in Fig 4.5A. Results

76
in Fig 4.5 supported that ProINS-Tf effectively inhibited BG increase induced by both liver
specific GNG substrate L-Ala and intestine /kidney specific GNG substrate L-Gln. For the L-
Ala tolerance test, the potency of ProINS-Tf’s restriction against BG elevation was more
prominent at 6 h compared to 12 h. ProINS-Tf (6 h) was able to inhibit 30% of peak BG
increase in L-Ala challenge. Interestingly, ProINS-Tf post 6 h group was almost completely
superimposed with saline-saline group and suggested a robust abolishment of L-Gln induced
GNG. Based on PD profiles, a BG plateau was maintained from 4 h to 12 h after s.c-injected
ProINS-Tf (Fig 4.1). Results from Fig.5A provided a more specific evaluation of s.c-
injection interval versus its potency in liver. This was also consistent with ProINS-Tf's T
max

(5.5 h) from previous PK studies [6].

4.3.6 Glycodynamic control of premixed ProINS-Tf and INS
From Fig 4.6, re-feeding after prolonged fasting (6-7 h) triggered a quick BG burst as
much as 150 mg/dL within the first hour in saline group. INS at 12 nmol/kg was able to
restrict 45% BG excursion, however, failed to bring BG back to normal level. Also, in
consistency with the result in Fig.2, INS’s hypoglycemic effect was rapidly diminished
within 3 h after injection. BG regulation after premixed ProINS-Tf/INS solution was
statistically similar as the INS group during the initial 2 h. Conversely, at 4 h post feeding,
the premixed group exhibited an efficient hypoglycemic effect and maintained feeding BG
below 200 mg/dL. Moreover, a BG plateau at normal level was observed in the following 8 h
fasting phase.

77
4.3.7 Glycodynamic control of repeated dose of ProINS-Tf /INS mixture
In Fig 4.7, premixed ProINS-Tf/INS (22.5/22.5 nmol/kg) displayed effective
hypoglycemic effect during both fasting and feeding phases. During the first hour in feeding
phase, premixed treatment exerted a rapid BG reduction and the BG curve was superimposed
with INS treatment (45 nmol/kg). In addition, during prolonged fasting, BG level was
maintained around euglycemia in ProINS/INS group (110 mg/dL). ProINS-Tf/INS displayed
a glycodynamic feature of biphasic INS therapeutic, which included both a fast-onset INS
action to control prandial hyperglycemia and a long-acting INS analog to cover basal INS
needs.

4.3.8 Mitogenic potential evaluation of ProINS-Tf
In Fig 4.8, after 30 min treatment, INS led to significantly increased IGF-1R
activation in MCF-7 cells (1.4 and 2.0-fold compared to non-treated cells at 100 nM and
1000 nM INS, respectively, p<0.05). ProINS-Tf did not trigger IGF-1R activation even at
1000 nM and the phospho-IGF-1R level was similar to the non-treated level in control group.
In Fig 4.9, compared with non-treated cells, MCF-7 cell proliferation were increased to
239±5% in 10 nM INS treated group, 138±7% in 10 nM ProINS-Tf treated group, and no
significant increase was observed after 10 nM holo-Tf treatment. In addition, ProINS-Tf
elicited similar cell proliferation level as ProINS (154±15%). Since ProINS was proved not
to bind IGF-1R [128], the increased cell proliferation level in both ProINS-Tf and ProINS
groups than the control could be resulted from intrinsic cell growth stimulatory effects of IR
signaling. It was also noticed that H-4-II-E pretreatment of protein did not induce apparent
changes in MCF-7 proliferation. The treatment group, 800 pM INS was an equivalent dose of

78
the released irINS-Tf level of 10 nM ProINS-Tf after 24 h incubation with H-4-II-4. Clearly,
the pretreated ProINS-Tf demonstrated weaker cell proliferation activity than 800 pM INS.

4.3.9 Effect of ProINS-Tf to PAI-1 synthesis in Hep G2 cells
From Fig 4.9, 10 nM INS treatment rendered the most significant stimulation of PAI-
1 secretion from Hep G2 cells. PAI-1 release increased over treatment duration as the
detected PAI-1 doubled at 48 h versus 24 h. In comparison, equal molar dose ProINS and
ProINS-Tf displayed similar activity in stimulating PAI-1. The released PAI-1 in both
treatments was only 20-35% of that from INS treatment, which suggested a much weaker
PAI-1 stimulation than INS.

79

Figure 4.1 Hypoglycemic effect of ProINS-Tf during fasting.
A comparison with INS glargine was made following s.c injection of a low (22.5 nmol/kg) or
high (135 nmol/kg) dose. Data obtained from four mice per group were shown as average
with error bars indicating the SD [6].

80

Figure 4.2 Comparison of the glycodynamic effect of ProINS-Tf and INS during free-
feeding.
The glycodynamic effects of ProINS-Tf (22.5 nmol/kg) and INS (22.5 nmol/kg) on mice
were compared under free-feeding condition(n=4). After s.c-injected protein treatment, mice
were kept under feeding condition during the entire experiment.

0 1 2 3 4 5
200
400
600
Time (h)
BG (mg/dL)
Insulin
Saline
ProINS-Tf
*
* indicates P< 0.05

81

Figure 4.3 Hypoglycemic efficacy under fast/feed cycles.
Equal dose (135 nmol/kg) of ProINS-Tf and INS glargine were s.c-injected to STZ induced
diabetic mice. After injection, mice were kept under fast/feed cycles and BG levels were
monitored at the end of each phase (8 h). White zones represented fasting phases and grey
zones indicated feeding phases. The dashed line indicated the severe hypoglycemia threshold
(BG=30 mg/dL). BG curves reflected average values and error bars indicated SD (n=3).

82

Figure 4.4 Dose response study.
STZ induced diabetic mice after s.c injection of respective treatment were kept under 8 h
fasting (white zone)/feeding (grey zone) phase intervals. Data represented average value with
SD. The dashed line indicated BG threshold (30 mg/dL) of severe hypoglycemia in mice.

83

Figure 4.5 BG increase during (A) L-Ala tolerance test and (B) L-Gln tolerance test.
Normal 6-7 weeks old male C57BL/6J mice after 24 h fasting were i.p injected with hepatic
gluconeogenesis substrate L-Ala (1.5 g/kg Sigma) or intestinal/renal gluconeogenesis
substrate L-Gln (1.5 g/kg, Sigma). During the 24 h fasting, mice were randomly grouped to
receive s.c injection of respective treatment (saline, 22.5 nmol/kg INS or 22.5 nmol/kg
ProINS-Tf) at indicated time prior to substrate injection (6 h or 12 h).

(A)
(B)

84

Figure 4.6 Comparison of the glycodynamic effect between premixed ProINS-Tf/INS and
INS alone during feeding and fasting conditions.
The glycodynamic effects between premixed ProINS-Tf/INS and INS alone were compared
during feeding (grey area) and fasting (white area) conditions. STZ induced diabetic mice
after 6 h fasting was randomly grouped to receive respective treatment by s.c bolus.
Treatment groups included premixed ProINS-Tf/INS (22.5/12 nmol/kg), INS (12 nmol/kg)
and saline control.

0 2 4 6 8 10 12
0
100
200
300
400
500
600
Time (h)
BG (mg/dL)
Premixed ProINS-Tf +INS
(22.5/12 nmol/kg)
INS(12 nmol/kg)
Saline

85

Figure 4.7 Hypoglycemic effect of premixed ProINS-Tf/INS after repeated dose during
feeding and fasting conditions.
The hypoglycemic effects of premixed ProINS-Tf/INS after repeated dose during feeding
and fasting conditions were compared. STZ induced diabetic mice after 2 h fasting was
randomly grouped to receive respective treatment by s.c bolus. Treatment groups included
premixed ProINS-Tf/INS (22.5/22.5 nmol/kg), INS (45 nmol/kg), ProINS-Tf (45 nmol/kg)
and saline control. After the first 4 h feed/ 8 h fast cycle, mice in each groups were received
another s.c bolus of previous treatment.

2 4 6 8 10 14 16 18 20 22 24
Inj.1 Inj.2
0
100
200
300
400
500
600
Time (h)
BG (mg/dL)
ProINS-Tf/INS (22.5/22.5 nmol/kg)
ProINS-Tf (45 nmol/kg)
INS(45 nmol/kg)
Saline

86

Figure 4.8 Activation of IGF-1R on MCF-7 cells after short-term protein treatment.
MCF-7 cells after overnight serum starvation were treated with various concentration of
either INS or ProINS-Tf in 0.1% BSA DMEM for 30 min, lysed, and subjected to ELISA
analysis to determine level of IGF-1R phosphorylation. Cell protein was quantified by BCA
kit for normalization (n=3). IGF-1R activation level was compared with untreated MCF-7
cells and * refers to p ≤0.05, *** refers to p≤0.001

ProINS-Tf Insulin Untreated
0
200
400
600
800
1000
IGF-1R

Phosphorylation
(units/ mg cell protein)
1000 nM
100 nM
Untreated
*
***

87

Figure 4.9 Effects of ProINS-Tf and H-4-II-E pretreated ProINS-Tf on cell proliferation of
MCF-7 cells.
MCF-7 cells were treated with 10 nM ProINS-Tf, human INS or holo-Tf for 72 h. Cell
viability was determined by MTT assay (n=3). And for each protein, cell proliferation level
was also compared between intact protein (filled bar) and H-4-II-E pretreated protein (bar
with strips). * refers to p ≤0.05.

88

Figure 4.10 Effects of ProINS-Tf on plasminogen activator inhibitor type-1(PAI-1) secretion
in Hep G2.
Concentration of PAl-1 protein in culture media of Hep G2 after expos to 10 nM of human
INS, ProINS or ProINS-Tf at various duration of exposure. PAI-1 level in plain culture
medium group was referred as basal level. (n=3)

24 h 48 h
2
4
6
Basal level
INS
ProINS
ProINS-Tf
PAI-1/cell protein
(pg/ug)

89
4.4 Discussion
In this chapter, from a battery of comparison studies on the hypoglycemic effect
among ProINS-Tf, INS and INS glargine, the glycodynamic profiles of ProINS-Tf after s.c
bolus was further investigated and its main features was concluded as: late-onset of
hypoglycemic effect, prolonged INS action, preferential inhibition against fasting
hyperglycemia and dose-independent BG reduction immediately after s.c injection.
Additionally, the clinical application of ProINS-Tf has been assessed as both basal INS
replacement and biphasic INS therapeutics. Also, gleaning the information based on in vivo
and in vitro studies, we summarized that ProINS-Tf displayed improved safety over INS
glargine and less mitogenic potential than INS.
ProINS-Tf exhibited a delayed onset of BG lowering efficacy, as BG levels declined
gradually within the first 4 h post-injection (Fig 4.1). This delayed onset has been attributed
to both a protracted absorption and the rate/activity-controlled conversion [5]. The
rate/activity-limited conversion can explain the relative dose-independent BG reduction
immediately after ProINS-Tf treatment (Fig 4.4), and demonstrated a major difference in
tailoring BG lowering efficacy between INS glargine and ProINS-Tf. While INS glargine
was tightly related with dose titration and dose formulation (Fig 4.1), ProINS-Tf exhibited
less reliance on dose possibly due to a stringently regulated conversion in vivo. This type of
dose-response regiment prevents INS overdose-induced hypoglycemia, since the conversion
of ProINS-Tf is conceivably regulated in vivo by the competition with excess endogenous Tf
in Tf/TfR pathway.
ProINS-Tf exhibited a prolonged BG lowering efficacy, which was still active at 40 h
(Fig 4.3) post-injection compared to 12 h for INS glargine (Fig 4.4). The prolonged efficacy

90
is presumably related to the prolonged half-life of ProINS-Tf, but may also be related to an
increase in intrahepatic retention. As discussed in chapter 2, either ProINS-Tf or irINS-Tf
exploited TfR recycling pathway as intracellular reservoirs, and led to time-lapse IR/Akt
signaling. Another conceivable reason contributing to the prolonged effect of ProINS-Tf was
the intrinsic feature of its action site. Studies on healthy human subjects demonstrated that
deactivation of INS suppression on HGO took two times longer than the disappearance of
INS-induced PGD after ceasing INS infusion [110]. Therefore, the prolonged efficacy of
ProINS-Tf may be related to its preferential action on HGO in the liver.
The hepatopreferential INS action of ProINS-Tf was thoroughly discussed in chapter
3. It was also supported by ProINS-Tf's selective BG control during fasting rather than
feeding (Fig 4.2 and Fig 4.3). Hepatic contribution in glycemic homeostasis varies between
basal and prandial condition [93]. A hepatopreferential INS therapeutic presumably is more
effective in control fasting hyperglycemia, which was mainly induced by an excessive HGO
[144]. Conversely, peripheral acted INS is necessary to regulate feeding BG since the main
part of absorbed dietary glucose is disposed to brain and the peripheral (skeletal muscle,
adipose tissues) via INS sensitive glucose transporters [145].
L-Ala and L-Gln tolerance tests were used to evaluate ProINS-Tf's influences on
hepatic GNG and extrahepatic GNG respectively. Currently, the causative relationship
between renal and intestinal GNG and diabetic hyperglycemia is still ambiguous [146,
147,148]. However, abnormally elevated renal GNG was observed in both Type 1 and Type
2 diabetic patients [149]. ProINS-Tf effectively inhibited both L-Ala and L-Gln induced BG
elevation (Fig.5) and fusion protein at 6 h post-injection was more potent in restricting BG
spike after GNG substrate challenge. Alanine stimulates HGO via dual action mechanisms,

91
increasing both GNG and glycogenosis [150]. Therefore, a 24 h fasting was used to rule out
the influence of glycogen breakdown in interpreting BG increase in Fig 4.5A. The various
ability of using glutamine in GNG among those organs is subject to a combination of
differences in glutamine supply, transport, and enzymatic activity (glutaminase) of the rate-
limiting steps [151]. Glucagon doubles alanine GNG rate in liver [152]. In comparison, renal
GNG from glutamine is hardly influenced by glucagon. Although kidney has glucagon
receptors expressed on distal nephron, the renal GNG site is in the proximal tubules [153].
Therefore, on a cautionary note, a quantitative comparison regarding the potency of ProINS-
Tf 's inhibition against hepatic versus extrahepatic GNG was not available.
Another therapeutic advantage displayed by ProINS-Tf was that it synergistically
prolonged the duration of INS's control against feeding hyperglycemia (Fig 4.6). More
importantly, the injected dose of INS was reduced to about half amount as in Fig.2. This
synergistic control of feeding glycemic level by ProINS-Tf and INS could be translated into
many clinical merits, e.g., the reduced dose of required prandial INS therapeutics can
alleviate the economic burden of patients and payers as well as reduce the risk of
overinsulinization. After dose optimization of each component in the ProINS-Tf/INS
mixture, result in Fig 4.7 suggested that ProINS-Tf is a promising candidate to prepare
biphasic INS therapeutics.
MCF-7 is a human breast cancer cell line already in neoplastic state and
predominantly expresses IGF-1R over IR. Therefore, in the IGF-1R rich MCF-7 cells, IR
activation mediated cell growth is minimal [119] and is a suitable in vitro cell platform to
evaluate ProINS-Tf's mitogenic potential as well as other INS analogs [154]. Molecular
characterization of ProINS-Tf’s mitogenic potential was assessed after both short-term

92
stimulus (30 min) and long-term treatment (72 h). Phosphorylation status of IGF-1R, together
with the mitogenic functional readout of cell proliferation, suggested that ProINS-Tf
displayed significantly reduced mitogenic potential compared with INS.
As in a one year ProINS clinical trials, the cardiovascular events were observed with
17-fold increased frequency and the plasma ProINS level in patients was as high as 8 nM.
Therefore, in the in vitro studies, we choose 10 nM as the dosing concentration [129,136].
The result demonstrated that ProINS-Tf had less PAI-1 stimulatory effect than native INS
and the discrepancy increased over time of incubation. It implied that ProINS-Tf, even with a
demonstrated enhanced and prolonged in vivo exposure compared to INS, is unlikely to result
in abnormally increased PAI-1 if it is given equal therapeutic dose to INS. Also, it should be
noted that in vitro studies on both endothelial cells as well as vascular muscle cells (VMC)
suggested that PAI-1 expression was increased when cells were cultured in a high
concentration of glucose (25 mM) compared to a low glucose medium (5 mM) [155]. In our
studies, the assay was carried out in high glucose medium mimicking diabetic hyperglycemia
scenario. Another possibility for the observed increased mitogenic activity of ProINS was
that it was subject to a preferential binding and activation ability towards IR-A isoform (7-
fold higher affinity than IR-B) in the mitogenic signaling ERK/p70S6K pathway more than
the Akt response [128]. In future, the same concern of ProINS-Tf could be tested if MCF-7
cells exclusively expressing IR-A or IR-B are available.

93
Chapter 5 Summary

5.1 Conclusions
5.1.1 Characterization of hepatic conversion of ProINS-Tf
Tf has long been used to improve the serum stability of fused protein drugs through
several pathways including the re-routing through the TfR-recycling pathway, blocking
protease/peptidase degradation and reducing renal elimination by increasing the molecular
weight [30,31]. In the case of ProINS-Tf, the fusion to Tf not only increased circulation time
in plasma [6], but also enabled ProINS to irINS conversion, presumably in the slow TfR
recycling pathway [5].
Conversion of ProINS-Tf into irINS-Tf was identified during prolonged incubation
with rat hepatoma cells. A conversion screen on a panel of cell lines expressing TfR
suggested that conversion requires additional factors to complete the intracellular processing.
The INS degradation inhibitor Bacitracin didn't significantly enhance the generation of
irINS-Tf, which suggested that INS degradation was not the causal factor for the minimal
conversion observed on Hep G2. Instead, the abundance of hepatic specific enzymes as well
as the Tf/TfR endocytosis pathway or the kinetics of TfR recycling could be the reason. More
importantly, irINS-Tf was detected in the medium after incubation with mouse primary
hepatocytes and precision-cut liver slices (PCLS), which further validated liver's ability to
convert ProINS-Tf. PCLS preserves the normal tissue architecture, the cell heterogeneity and
cell–cell interactions and hepatocyte bipolarity. Specific for our study it is most close to the
in vivo conditions for a conversion rate and capacity quantitative study [69]. The incubation
conditions have been modulated to test the importance of TfR in PCLS conversion assays.

94
The absence of conversion phenomenon in either the presence of TfR blockage by holo-Tf or
ProINS alone consolidated the hypothesis that conversion in vivo was also mediated by TfR,
which was consistent with that on H-4-II-E cells.
To summarize the results from C-peptide detection during conversion assays on both
rat hepatoma cells and mouse PCLS, no C-peptide increase from the incubation medium was
observed among all the experiments, suggesting no C-peptide was released beyond the
detecting threshold or C-peptide wasn't released in proportion to the generated irINS-Tf.
These findings provide at least two fold significance. Firstly, it suggested the processing of
ProINS-Tf in the slow recycling intracellular compartments of TfR in hepatocyte was
different from that in the secretary granules in pancreatic beta cells. In the latter pathway, C-
peptide was yielded in a 1:1 ratio with INS. Secondly, it is informative in identifying the
absolute form of irINS-Tf in considering the possibility of Des 64,65 ProINS-Tf, in which C-
peptide was attached to INS B-chain. The comparison study between ProINS and ProINS-Tf
after Trypsin and Carboxypeptidase B enzyme treatment indicated that ProINS-Tf may
acquire additional enzymatic resistance due to the steric hindrance of the Tf part. Instead of
being cut at both connection regions B-chain/C-peptide and C-peptide/A-chain as in ProINS,
the generated product irINS-Tf possibly underwent one single cleavage of ProINS-Tf.

5.1.2 Evaluation of liver selective action of ProINS-Tf
Due to the extensive hepatic extraction of INS, the hepatopreference of INS analogs
are evaluated based on the INS biologic responses in liver versus peripheral tissues. By
comparing the IR activation status and duration from liver and skeletal muscles after s.c
bolus, we concluded that ProINS-Tf displayed preferential and prolonged INS action in liver

95
after s.c bolus. A hepatic GNG substrate pyruvate tolerance study provided direct proof that
the fusion protein was able to inhibit HGO. Consistent with long-lasting hypoglycemic
effect observed in vivo, ProINS-Tf decreased 50% of pyruvate induced BG elevation even at
12 h post s.c injection.
During free-feeding, ProINS-Tf was also able to significantly increase the hepatic
glycogen storage, by which it exerted a minor effect in alleviating feeding hyperglycemia.
Additionally, the activity of GSase after short-term fusion protein treatment was increased
and the stimulatory extent was comparable to equal molar INS based on in vitro
observations.
The delayed time course of Akt dephosphorylation provided another rational to
explain the long-acting INS action of the fusion protein. Either ProINS-Tf or irINS-Tf may
exploit the endosomes (dynamic/static early endosomes pH 6.0 or recycling endosomes pH
5.6) along the TfR mediated slow recycling pathway as intracellular reservoirs to
continuously elicit IR signaling, even when the extracellular IR stimulus were abolished.
ProINS itself has a marginal effect in lowering BG level and it shows weak
hepatospecific characteristics. Studies suggested that ProINS’s hepatospecificity resulted
from pharmacodynamics owing to its prolonged intracellular retention [156] or enhanced
liver access due to the passive targeting machinery by increased MW [157]. In terms of
ProINS-Tf, which is an INS prodrug and undergoes an intracellular activation, it stands a
good chance that its mechanism for hepatoselectivity is not identical to ProINS. Several
possible mechanisms were proposed based on current studies. One is the Tf moiety in the
fusion protein increased its intracellular retention after TfR mediated endocytosis in
hepatocytes, which provided a liver-localized ProINS-Tf depot. In addition, TfR slow cycling

96
pathway enabled ProINS-Tf to convert into more active INS analogues. In other words, liver
specificity benefited from TfR mediated improved liver exposure, prolonged intra-hepatocyte
retention, as well as a liver selective conversion. The Tf fusion protein was able to improve
the bioavailability of orally delivered protein therapeutics via enhanced intracellular retention
[158]. The extended IR signaling also set a new example of Tf fusion technology to extend
the protein drug's biological response by the endosomal platform. Tf fusion may serve as
another approach to enhance the efficacy for intracellular delivered therapeutic protein.

5.1.3 Exploration of the therapeutic potentials of ProINS-Tf as basal INS replacement
in diabetes treatment
In a series of hypoglycemic effect comparison studies, equal molar ProINS-Tf was
compared with short acting product native INS and long-acting INS glargine, both of which
were proved with significant peripheral INS action. Additional, in the in vitro molecular
characterization related with mitogenic potential, human ProINS was added as another
control group.
On STZ induced diabetic mice, INS (22.5 nmol/kg) immediately reduced BG to near
normal level within 1 h after s.c bolus regardless of diet conditions. In comparison, ProINS-
Tf treated group led to a gradual BG reduction during fasting condition among all the tested
doses ranging from 22.5 nmol- 202.5 nmol/kg. This late-onset of its hypoglycemic feature
suggested an internal factor contributed to the 3-4 h delay except for absorption. Previous
studies [5] also supported that when higher concentration of ProINS-Tf was incubated with
H-4-II-E cells for 24 h, the ratio of irINS-Tf to the total ProINS-Tf turned out to be lower,
e.g., the conversion ratio for 1 nM ProINS-Tf was 20% and for 10 nM ProINS-Tf was 8%. In

97
other words, when more abundant ProINS-Tf are available in the extracellular environment,
the conversion capacity is limited by some intrinsic factors, such as the kinetics of TfR
recycling or the access to cleavage enzymes. When this in vitro finding is extrapolated into in
vivo situation, the conversion rate is further regulated given the competition with endogenous
Tf.
In comparison with INS glargine and INS, ProINS-Tf preferentially inhibited fasting
hyperglycemia whilst the other two blunted BG increase in both feeding and fasting phases.
During the fast/feed cycles, INS glargine action was uninfluenced by food intake situation.
At the dose of 135 nmol/kg, the duration of its BG lowering effect was much shorter than
ProINS-Tf and triggered severe hypoglycemia within 2 h post-injection. ProINS-Tf's
selective restriction against fasting hyperglycemic allows it to be a candidate as biphasic INS
preparation. The PK profile of biphasic INS is a combination of both a rapid onset and peak
INS action to cover postprandial period and followed by a rapid return to a long flat profile to
cover basal needs [159]. Its clinical advantages including lower incidence of nocturnal
hypoglycemia and more flexibility of intensified treatment [160]. Eliminating the strict
formulation requirement for INS glargine, ProINS-Tf is readily mixed with native INS
without jeopardizing its extended BG lowering effect. The synergistic effect of premixed
ProINS-Tf/INS in controlling both feeding and fasting BG was demonstrated in chapter 5.
Lastly, results from in vitro IGF-1R activation and cell proliferation assays
demonstrated that ProINS-Tf significantly lowered the mitogenic potential of INS. ProINS-
Tf also displayed lower stimulation of PAI-1 secretion compared with INS. In vitro assays
did not directly address the question about irINS-Tf's mitogenic potential. However, our
studies concluded that 24 h of H-4-II-E pretreatment did not induce significant difference on

98
MCF-7 cell proliferation. Taken account into the possibility that ProINS-Tf conversion might
be lower in vivo due to competition with endogenous Tf, it is unlikely that irINS-Tf would
cause increased tumorigenesity risk. With that being said, the tumorigeneicity concern would
have more definitive answers from long-term animal toxicity studies.

5.2 Future perspective
5.2.1 Exploitation of Tf fusion technology as a platform for hepatic activated prodrug
ProINS-Tf was converted into the active form irINS-Tf and the phenomenon was
previously identified in rat hepatoma cells, and later has been substantiated on both mice
primary hepatocytes and PCLS culture. The conversion process attributed ProINS-Tf as a
hepatopreferential INS action and demonstrated a late-onset and prolonged hypoglycemic
effect in vivo. The conversion process was likely inducted by one enzyme member belong to
the proprotein convertase family, which is inherited with Trypsin-like properties and is
enriched in the slow recycling pathway of TfR in hepatocytes. Inspired by the example of
ProINS-Tf, the slow recycling pathway of TfR provide a potential platform for selective
targeting given the following advantages: (1) prolonged therapeutic signaling or biologic
responses by acquired intracellular retention (2) activation of prodrug with enzyme-cleavable
modification at constant rate in the intracellular endosomes (3) liver targeting. For a small
molecule drug, the enzyme sensitive sequence can be integrated as a peptide linker between
the drug molecule and carrier protein Tf. Once it is endocytosis, the free drug can be
liberated after enzyme cleavage. In comparison with the complex prodrug strategy such as
antibody-directed enzyme prodrug therapy [161], TfR slow recycling pathway can meet the
requirement for site specific release using required enzymes provided in situ.

99

5.2.2 Evaluation of the long-term glycemic management and weight neutrality
advantage in Type 2 diabetes model
Since single time point BG monitoring has limits in evaluating overall glycemic
control and cannot present a full picture of BG fluctuation, a comprehensive evaluation of
ProINS-Tf's glycemic regulation can be assessed by checking the level of glycosylated
hemoglobin. Hemoglobin A1c (HbA1c) is the major form of glycosylated hemoglobin in
human, and HbA1c concentration in diabetic patients is often two to three-fold higher than
that of healthy people [162]. A 1% decrease in HbA1c, according to United Kingdom
Prospective Diabetes Study (UKPDS), leads to 17- 34% reduction of complication incidence
on microvasculature and microvasculature, respectively [163]. However, the hypoglycemic
risk associated with current INS therapeutics prevents more aggressive dose schemes to
reduce HbA1c level under the expected 6.5% [164]. HbA1c is an indicator of BG in the
previous 2-3 months in human [165]. Glucose is blood samples derived from different
vessels (artery, venous and capillary) under all food intake situation (fasting and
postprandial) are in good correspondence with HbA1c [166]. Therefore, HbA1c monitoring
after long-term animal studies gives an overall evaluation on ProINS-Tf’s BG management
and may prove its superiority over same dose of INS in reducing HbA1c index without
hypoglycemic risk.
Long-term s.c INS treatment inevitably increases weight gain. For Type 2 diabetes
patients, weight gain has proved to be tightly correlated with increased cardiovascular risk
[167] and INS absorption was slowed down by increased thickness of fat in s.c tissue [168].
The accompanied weight gain for current INS therapeutics is subject to a defensively

100
increased calorie intake coping with hypoglycemia and the non-physiologycial INS
distribution between liver and peripheral tissue (fat and skeleton muscle) [169]. Since
ProINS-Tf displayed a hepatopreferential INS action with improved safety, it potentially
averts weight gain and is beneficial for long-term treatment in Type 2 diabetes. Type 2
diabetic mouse model resulted from gene knock-out could be applied to test ProINS-Tf's
therapeutic outcome. B6.Cg-Lepob/J is genetically mutated model carrying a singe gene
mutation in the leptin (Lepob) gene with C57BL/6J background. After long-term ProINS-Tf
treatment, indicators including body weight and body composition (lean tissue and body fat)
can be compared before and after treatment. ProINS-Tf's influences on lipid metabolism can
be analyzed by measuring total cholesterol amount, plasma triglycerides and high-density
lipoprotein cholesterol level using CardioCheck blood analyzer.

101
Reference
1. Ferrannini, E. Physiology of glucose homeostasis and insulin therapy in type 1 and
type 2 diabetes. Endocrinol. Metab. Clin. North Am. 41, 25–39 (2012).
2. Braet, F. & Wisse, E. Structural and functional aspects of liver sinusoidal endothelial
cell fenestrae: a review. Comp Hepatol 1, 1 (2002).
3. Duckworth, W. C., Bennett, R. G. & Hamel, F. G. Insulin degradation: progress and
potential. Endocr. Rev. 19, 608–624 (1998).
4. Matteucci, E. et al. Insulin administration: present strategies and future directions for
a noninvasive (possibly more physiological) delivery. Drug Des Devel Ther 9, 3109–
3118 (2015).
5. Sarin, H. Physiologic upper limits of pore size of different blood capillary types and
another perspective on the dual pore theory of microvascular permeability. J
Angiogenes Res 2, 14 (2010).
6. Steil, G. M., Ader, M., Moore, D. M., Rebrin, K. & Bergman, R. N. Transendothelial
insulin transport is not saturable in vivo. No evidence for a receptor-mediated
process. J. Clin. Invest. 97, 1497–1503 (1996).
7. McDonald, T. A., Zepeda, M. L., Tomlinson, M. J., Bee, W. H. & Ivens, I. A.
Subcutaneous administration of biotherapeutics: current experience in animal models.
Curr. Opin. Mol. Ther. 12, 461–470 (2010).
8. Hoffman, A. & Ziv, E. Pharmacokinetic considerations of new insulin formulations
and routes of administration. Clin Pharmacokinet 33, 285–301 (1997).
9. Braak, ter, E. W. et al. Injection site effects on the pharmacokinetics and
glucodynamics of insulin lispro and regular insulin. Diabetes Care 19, 1437–1440
(1996).
10. Jockel, J. P. L., Roebrock, P. & Shergold, O. A. Insulin depot formation in
subcutaneoue tissue. J Diabetes Sci Technol 7, 227–237 (2013).
11. Kuerzel, G. U. et al. Biotransformation of insulin glargine after subcutaneous
injection in healthy subjects. Curr Med Res Opin 19, 34–40 (2003).
12. Hayford, J. T. & Thompson, R. G. Free and total insulin integrated concentrations in
insulin dependent diabetes. Metab. Clin. Exp. 31, 387–397 (1982).
13. Farmer, T. D. et al. Comparison of the physiological relevance of systemic vs. portal
insulin delivery to evaluate whole body glucose flux during an insulin clamp. Am. J.
Physiol. Endocrinol. Metab. 308, E206–22 (2015).
14. Canfield, R. E., Kaye, G. I. & West, S. B. The preparation and evaluation of tritiated
polyalanyl insulin derivatives. Endocrinology 90, 112–122 (1972).
15. Stout, R. W. Insulin as a mitogenic factor: role in the pathogenesis of cardiovascular
disease. Am. J. Med. 90, 62S–65S (1991).
16. Richter, W. F. & Jacobsen, B. Subcutaneous absorption of biotherapeutics: knowns
and unknowns. Drug Metab. Dispos. 42, 1881–1889 (2014).
17. Madsbad, S. LY2605541--a preferential hepato-specific insulin analogue. Diabetes
63, 390–392 (2014).
18. Rosenstock, J. et al. Better glycemic control and weight loss with the novel long-
acting basal insulin LY2605541 compared with insulin glargine in type 1 diabetes: a
randomized, crossover study. Diabetes Care 36, 522–528 (2013).
19. Bergenstal, R. M. et al. A randomized, controlled study of once-daily LY2605541, a

102
novel long-acting basal insulin, versus insulin glargine in basal insulin-treated
patients with type 2 diabetes. Diabetes Care 35, 2140–2147 (2012).
20. Moore, M. C. et al. Novel PEGylated basal insulin LY2605541 has a preferential
hepatic effect on glucose metabolism. Diabetes 63, 494–504 (2014).
21. Pandyarajan, V. & Weiss, M. A. Design of non-standard insulin analogs for the
treatment of diabetes mellitus. Curr. Diab. Rep. 12, 697–704 (2012).
22. Navarro, V. J. & Senior, J. R. Drug-related hepatotoxicity. N. Engl. J. Med. 354,
731–739 (2006).
23. Eliaschewitz, F. G. & Barreto, T. Concepts and clinical use of ultra-long basal
insulin. Diabetol Metab Syndr 8, 2 (2016).
24. Shojaee-Moradie, F. et al. Effect of thyroid hormone binding proteins on insulin
receptor binding of B1-thyronine-insulin analogues. Biochem. J. 381, 51–57 (2004).
25. Shojaee-Moradie, F. et al. Novel hepatoselective insulin analogues: studies with
covalently linked thyroxyl-insulin complexes. Diabet. Med. 15, 928–936 (1998).
26. Shojaee-Moradie, F. et al. Novel hepatoselective insulin analog: studies with a
covalently linked thyroxyl-insulin complex in humans. Diabetes Care 23, 1124–
1129 (2000).
27. Jirasakuldech, B. et al. A characteristic serpin cleavage product of thyroxine-binding
globulin appears in sepsis sera. J. Clin. Endocrinol. Metab. 85, 3996–3999 (2000).
28. Geho, W. B. & Lau, J. R. Hepatocyte Directed Vesicle delivery system. (1986).
29. Davis, S. N. et al. The effects of HDV-insulin on carbohydrate metabolism in Type 1
diabetic patients. J. Diabetes Complicat. 15, 227–233 (2001).
30. Geho, W. B., Geho, H. C., Lau, J. R. & Gana, T. J. Hepatic-directed vesicle insulin:
a review of formulation development and preclinical evaluation. J Diabetes Sci
Technol 3, 1451–1459 (2009).
31. Harashima, H. & Kiwada, H. Liposomal targeting and drug delivery: kinetic
consideration. Adv. Drug Deliv. Rev. 19, 425–444 (1996).
32. Shojaee-Moradie, F. Hepatic and peripheral effects of four insulin analogues. (1995).
33. Denley, A., Wallace, J. C., Cosgrove, L. J. & Forbes, B. E. The insulin receptor
isoform exon 11- (IR-A) in cancer and other diseases: a review. Horm. Metab. Res.
35, 778–785 (2003).
34. Uhles, S., Moede, T., Leibiger, B., Berggren, P.-O. & Leibiger, I. B. Isoform-
specific insulin receptor signaling involves different plasma membrane domains. J.
Cell Biol. 163, 1327–1337 (2003).
35. Benecke, H., Flier, J. S. & Moller, D. E. Alternatively spliced variants of the insulin
receptor protein. Expression in normal and diabetic human tissues. J. Clin. Invest. 89,
2066–2070 (1992).
36. Glendorf, T., Sørensen, A. R., Nishimura, E., Pettersson, I. & Kjeldsen, T.
Importance of the solvent-exposed residues of the insulin B chain alpha-helix for
receptor binding. Biochemistry 47, 4743–4751 (2008).
37. Glendorf, T. et al. Engineering of insulin receptor isoform-selective insulin
analogues. PLoS ONE 6, e20288 (2011).
38. Vienberg, S. G. et al. Receptor-isoform-selective insulin analogues give tissue-
preferential effects. Biochem. J. 440, 301–308 (2011).
39. Besic, V., Shi, H., Stubbs, R. S. & Hayes, M. T. Aberrant liver insulin receptor
isoform a expression normalises with remission of type 2 diabetes after gastric

103
bypass surgery. PLoS ONE 10, e0119270 (2015).
40. Wang, Y., Chen, Y.-S., Zaro, J. L. & Shen, W.-C. Receptor-mediated activation of a
proinsulin-transferrin fusion protein in hepatoma cells. J Control Release 155, 386–
392 (2011).
41. Wang, Y., Shao, J., Zaro, J. L. & Shen, W.-C. Proinsulin-transferrin fusion protein as
a novel long-acting insulin analog for the inhibition of hepatic glucose production.
Diabetes 63, 1779–1788 (2014).
42. Chen, X., Lee, H.-F., Zaro, J. L. & Shen, W.-C. Effects of receptor binding on
plasma half-life of bifunctional transferrin fusion proteins. Mol. Pharm. 8, 457–465
(2011).
43. Han, H. K. & Amidon, G. L. Targeted prodrug design to optimize drug delivery.
AAPS PharmSci 2, E6–58 (2000).
44. Stella, V. J. & Himmelstein, K. J. Prodrugs and site-specific drug delivery. J. Med.
Chem. 23, 1275–1282 (1980).
45. Rizza, R. A., Mandarino, L. J. & Gerich, J. E. Dose-response characteristics for
effects of insulin on production and utilization of glucose in man. Am. J. Physiol.
240, E630–9 (1981).
46. Caparrotta, T. M. & Evans, M. PEGylated insulin Lispro, (LY2605541)--a new basal
insulin analogue. Diabetes Obes Metab 16, 388–395 (2014).
47. Topp, B. G. et al. Effects of a novel PEGylated basal insulin, LY2605541, on hepatic
glucose output and muscle glucose uptake: simulations based on data from
euglycaemic clamp studies. 55, S376–S376 (2012).
48. Bailyes, E. M. et al. A member of the eukaryotic subtilisin family (PC3) has the
enzymic properties of the type 1 proinsulin-converting endopeptidase. Biochem. J.
285 ( Pt 2), 391–394 (1992).
49. Bailyes, E. M., Bennett, D. L. & Hutton, J. C. Proprotein-processing endopeptidases
of the insulin secretory granule. Enzyme 45, 301–313 (1991).
50. Arcelloni, C. et al. Processing and release of human proinsulin-cleavage products
into culture media by different engineered non-endocrine cells: a specific assessment
by capillary electrophoresis. J. Endocrinol. 166, 437–445 (2000).
51. Moore, H. P., Walker, M. D., Lee, F. & Kelly, R. B. Expressing a human proinsulin
cDNA in a mouse ACTH-secreting cell. Intracellular storage, proteolytic processing,
and secretion on stimulation. Cell 35, 531–538 (1983).
52. Given, B. D. et al. Biochemical and clinical implications of proinsulin conversion
intermediates. J. Clin. Invest. 76, 1398–1405 (1985).
53. Peavy, D. E., Brunner, M. R., Duckworth, W. C., Hooker, C. S. & Frank, B. H.
Receptor binding and biological potency of several split forms (conversion
intermediates) of human proinsulin. Studies in cultured IM-9 lymphocytes and in
vivo and in vitro in rats. J. Biol. Chem. 260, 13989–13994 (1985).
54. Huebers, H. A. & Finch, C. A. The physiology of transferrin and transferrin
receptors. Physiol. Rev. 67, 520–582 (1987).
55. Morgan, E. H., Smith, G. D. & Peters, T. J. Uptake and subcellular processing of
59Fe-125I-labelled transferrin by rat liver. Biochem. J. 237, 163–173 (1986).
56. Conrad, M. E., Umbreit, J. N. & Moore, E. G. Iron absorption and transport. Am. J.
Med. Sci. 318, 213–229 (1999).
57. Grant, B. D. & Donaldson, J. G. Pathways and mechanisms of endocytic recycling.

104
Nat. Rev. Mol. Cell Biol. 10, 597–608 (2009).
58. Maxfield, F. R. & McGraw, T. E. Endocytic recycling. Nat. Rev. Mol. Cell Biol. 5,
121–132 (2004).
59. Bonifacino, J. S. & Rojas, R. Retrograde transport from endosomes to the trans-
Golgi network. Nat. Rev. Mol. Cell Biol. 7, 568–579 (2006).
60. Fries, E., Gustafsson, L. & Peterson, P. A. Four secretory proteins synthesized by
hepatocytes are transported from endoplasmic reticulum to Golgi complex at
different rates. EMBO J. 3, 147–152 (1984).
61. Seidah, N. G. et al. The activation and physiological functions of the proprotein
convertases. Int. J. Biochem. Cell Biol. 40, 1111–1125 (2008).
62. Seidah, N. G. & Prat, A. The biology and therapeutic targeting of the proprotein
convertases. Nat Rev Drug Discov 11, 367–383 (2012).
63. Kelly, R. B. Pathways of protein secretion in eukaryotes. Science 230, 25–32 (1985).
64. Mori, K. et al. Subtilisin-like proprotein convertases, PACE4 and PC8, as well as
furin, are endogenous proalbumin convertases in HepG2 cells. J. Biochem. 125, 627–
633 (1999).
65. Li, W.-C., Ralphs, K. L. & Tosh, D. Isolation and culture of adult mouse hepatocytes.
Methods Mol. Biol. 633, 185–196 (2010).
66. Luck, A. N., Bobst, C. E., Kaltashov, I. A. & Mason, A. B. Human serum transferrin:
is there a link among autism, high oxalate levels, and iron deficiency anemia?
Biochemistry 52, 8333–8341 (2013).
67. Sonoda, S. & Schlamowitz, M. Studies of 125I trace labeling of immunoglobulin G
by chloramine-T. Immunochemistry 7, 885–898 (1970).
68. Syed, N. A. Regulation of glycogen synthase and glycogen phosphorylase by insulin
in HepG2 cells. (2001).
69. Guo, Y., Wang, H. & Zhang, C. Establishment of rat precision-cut fibrotic liver slice
technique and its application in verapamil metabolism. Clin. Exp. Pharmacol.
Physiol. 34, 406–413 (2007).
70. Buettner, R. et al. Efficient analysis of hepatic glucose output and insulin action
using a liver slice culture system. Horm. Metab. Res. 37, 127–132 (2005).
71. Smith, P. F. et al. Dynamic organ culture of precision liver slices for in vitro
toxicology. Life Sci. 36, 1367–1375 (1985).
72. Olinga, P. et al. Comparison of five incubation systems for rat liver slices using
functional and viability parameters. J Pharmacol Toxicol Methods 38, 59–69 (1997).
73. Clayton, D. F. & Darnell, J. E. Changes in liver-specific compared to common gene
transcription during primary culture of mouse hepatocytes. Mol. Cell. Biol. 3, 1552–
1561 (1983).
74. Horwitz, D. L., Starr, J. I., Mako, M. E., Blackard, W. G. & Rubenstein, A. H.
Proinsulin, insulin, and C-peptide concentrations in human portal and peripheral
blood. J. Clin. Invest. 55, 1278–1283 (1975).
75. Katz, A. I. & Rubenstein, A. H. Metabolism of proinsulin, insulin, and C-peptide in
the rat. J. Clin. Invest. 52, 1113–1121 (1973).
76. Jones, A. G. & Hattersley, A. T. The clinical utility of C-peptide measurement in the
care of patients with diabetes. Diabet. Med. 30, 803–817 (2013).
77. Kemmler, W., Peterson, J. D. & Steiner, D. F. Studies on the conversion of
proinsulin to insulin. I. Conversion in vitro with trypsin and carboxypeptidase B. J.

105
Biol. Chem. 246, 6786–6791 (1971).
78. Gardner, Q.-T.-A. A., Younas, H. & Akhtar, M. Studies on the regioselectivity and
kinetics of the action of trypsin on proinsulin and its derivatives using mass
spectrometry. Biochim. Biophys. Acta 1834, 182–190 (2013).
79. Miller, W. L. & Baxter, J. D. Recombinant DNA--a new source of insulin.
Diabetologia 18, 431–436 (1980).
80. Jonasson, P. et al. Single-step trypsin cleavage of a fusion protein to obtain human
insulin and its C peptide. Eur. J. Biochem. 236, 656–661 (1996).
81. López-Terrada, D., Cheung, S. W., Finegold, M. J. & Knowles, B. B. Hep G2 is a
hepatoblastoma-derived cell line. Hum. Pathol. 40, 1512–1515 (2009).
82. Wilkening, S., Stahl, F. & Bader, A. Comparison of primary human hepatocytes and
hepatoma cell line Hepg2 with regard to their biotransformation properties. Drug
Metab. Dispos. 31, 1035–1042 (2003).
83. Tanaka, T., Fujishima, Y. & Kaneo, Y. Receptor mediated endocytosis and
cytotoxicity of transferrin-mitomycin C conjugate in the HepG2 cell and primary
cultured rat hepatocyte. Biol. Pharm. Bull. 24, 268–273 (2001).
84. Thohan, S. & Rosen, G. M. Liver slice technology as an in vitro model for metabolic
and toxicity studies. Methods Mol. Biol. 196, 291–303 (2002).
85. de Graaf, I. A. M. et al. Preparation and incubation of precision-cut liver and
intestinal slices for application in drug metabolism and toxicity studies. Nat Protoc 5,
1540–1551 (2010).
86. Baravalle, G. et al. Transferrin recycling and dextran transport to lysosomes is
differentially affected by bafilomycin, nocodazole, and low temperature. Cell Tissue
Res. 320, 99–113 (2005).
87. Saltiel, A. R. & Kahn, C. R. Insulin signalling and the regulation of glucose and lipid
metabolism. Nature 414, 799–806 (2001).
88. Sharabi, K., Tavares, C. D. J., Rines, A. K. & Puigserver, P. Molecular
pathophysiology of hepatic glucose production. Mol. Aspects Med. 46, 21–33 (2015).
89. Consoli, A. Role of liver in pathophysiology of NIDDM. Diabetes Care 15, 430–441
(1992).
90. Consoli, A., Nurjhan, N., Capani, F. & Gerich, J. Predominant role of
gluconeogenesis in increased hepatic glucose production in NIDDM. Diabetes 38,
550–557 (1989).
91. Hwang, J. H. et al. Impaired net hepatic glycogen synthesis in insulin-dependent
diabetic subjects during mixed meal ingestion. A 13C nuclear magnetic resonance
spectroscopy study. J. Clin. Invest. 95, 783–787 (1995).
92. Bischof, M. G. et al. Hepatic glycogen metabolism in type 1 diabetes after long-term
near normoglycemia. Diabetes 51, 49–54 (2002).
93. Cherrington, A. D. Banting Lecture 1997. Control of glucose uptake and release by
the liver in vivo. Diabetes 48, 1198–1214 (1999).
94. Roden, M. et al. The roles of insulin and glucagon in the regulation of hepatic
glycogen synthesis and turnover in humans. J. Clin. Invest. 97, 642–648 (1996).
95. Radziuk, J. & Pye, S. Hepatic glucose uptake, gluconeogenesis and the regulation of
glycogen synthesis. Diabetes Metab. Res. Rev. 17, 250–272 (2001).
96. Villar-Palasí, C. & Guinovart, J. J. The role of glucose 6-phosphate in the control of
glycogen synthase. FASEB J. 11, 544–558 (1997).

106
97. Parker, G., Taylor, R., Jones, D. & McClain, D. Hyperglycemia and inhibition of
glycogen synthase in streptozotocin-treated mice: role of O-linked N-
acetylglucosamine. J. Biol. Chem. 279, 20636–20642 (2004).
98. Burant, C. F. & Bell, G. I. Mammalian facilitative glucose transporters: evidence for
similar substrate recognition sites in functionally monomeric proteins. Biochemistry
31, 10414–10420 (1992).
99. SEIFTER, S. & DAYTON, S. The estimation of glycogen with the anthrone reagent.
Arch Biochem 25, 191–200 (1950).
100. Niederwanger, A., Kranebitter, M., Ritsch, A., Patsch, J. R. & Pedrini, M. T. A gel
filtration assay to determine glycogen synthase activity. J. Chromatogr. B Analyt.
Technol. Biomed. Life Sci. 820, 143–145 (2005).
101. Oron, Y. & Larner, J. A modified rapid filtration assay of glycogen synthase. Anal.
Biochem. 94, 409–410 (1979).
102. Trón, L., Matkó, J., Szücs, K. A. & Enyedi, A. A continuous assay method for
glycogen synthesis from UDPglucose. J. Biochem. Biophys. Methods 4, 167–175
(1981).
103. Nielsen, J. N. et al. Glycogen synthase localization and activity in rat skeletal muscle
is strongly dependent on glycogen content. J. Physiol. (Lond.) 531, 757–769 (2001).
104. Geho, W. B. The importance of the liver in insulin replacement therapy in insulin-
deficient diabetes. Diabetes 63, 1445–1447 (2014).
105. Kim, B.-J. et al. Transferrin fusion technology: a novel approach to prolonging
biological half-life of insulinotropic peptides. J. Pharmacol. Exp. Ther. 334, 682–
692 (2010).
106. Chen, X., Zaro, J. L. & Shen, W.-C. Pharmacokinetics of recombinant bifunctional
fusion proteins. Expert Opin Drug Metab Toxicol 8, 581–595 (2012).
107. Bevan, A. P. et al. Selective activation of the rat hepatic endosomal insulin receptor
kinase. Role for the endosome in insulin signaling. J. Biol. Chem. 270, 10784–10791
(1995).
108. Murphy, J. E., Padilla, B. E., Hasdemir, B., Cottrell, G. S. & Bunnett, N. W.
Endosomes: a legitimate platform for the signaling train. Proc. Natl. Acad. Sci.
U.S.A. 106, 17615–17622 (2009).
109. Marshall, S. Dual pathways for the intracellular processing of insulin. Relationship
between retroendocytosis of intact hormone and the recycling of insulin receptors. J.
Biol. Chem. 260, 13524–13531 (1985).
110. Glauber, H. S. et al. In vivo deactivation of proinsulin action on glucose disposal and
hepatic glucose production in normal man. Diabetes 35, 311–317 (1986).
111. Chung, S. T., Chacko, S. K., Sunehag, A. L. & Haymond, M. W. Measurements of
Gluconeogenesis and Glycogenolysis: A Methodological Review. Diabetes 64,
3996–4010 (2015).
112. Meier, J. J., Veldhuis, J. D. & Butler, P. C. Pulsatile insulin secretion dictates
systemic insulin delivery by regulating hepatic insulin extraction in humans.
Diabetes 54, 1649–1656 (2005).
113. Tibaldi, J. M. Evolution of insulin: from human to analog. Am. J. Med. 127, S25–38
(2014).
114. Shah, V. N., Moser, E. G., Blau, A., Dhingra, M. & Garg, S. K. The future of basal
insulin. Diabetes Technol. Ther. 15, 727–732 (2013).

107
115. Lepore, M. et al. Pharmacokinetics and pharmacodynamics of subcutaneous
injection of long-acting human insulin analog glargine, NPH insulin, and ultralente
human insulin and continuous subcutaneous infusion of insulin lispro. Diabetes 49,
2142–2148 (2000).
116. Cryer, P. E. Glycemic goals in diabetes: trade-off between glycemic control and
iatrogenic hypoglycemia. Diabetes 63, 2188–2195 (2014).
117. Reno, C. M. et al. Severe hypoglycemia-induced lethal cardiac arrhythmias are
mediated by sympathoadrenal activation. Diabetes 62, 3570–3581 (2013).
118. Gregory, J. M. et al. Insulin Delivery Into the Peripheral Circulation: A Key
Contributor to Hypoglycemia in Type 1 Diabetes. Diabetes 64, 3439–3451 (2015).
119. Baricevic, I. et al. A framework for the in vitro evaluation of cancer-relevant
molecular characteristics and mitogenic potency of insulin analogues.
Carcinogenesis 36, 1040–1050 (2015).
120. Lammers, R., Gray, A., Schlessinger, J. & Ullrich, A. Differential signalling
potential of insulin- and IGF-1-receptor cytoplasmic domains. EMBO J. 8, 1369–
1375 (1989).
121. Westley, R. L. & May, F. E. B. A twenty-first century cancer epidemic caused by
obesity: the involvement of insulin, diabetes, and insulin-like growth factors. Int J
Endocrinol 2013, 632461–37 (2013).
122. De Meyts, P. Insulin/receptor binding: the last piece of the puzzle? What recent
progress on the structure of the insulin/receptor complex tells us (or not) about
negative cooperativity and activation. Bioessays 37, 389–397 (2015).
123. Peyrat, J. P. et al. Characterization of insulin-like growth factor 1 receptors (IGF1-R)
in human breast cancer cell lines. Bull Cancer 76, 311–319 (1989).
124. Milazzo, G., Sciacca, L., Papa, V., Goldfine, I. D. & Vigneri, R. ASPB10 insulin
induction of increased mitogenic responses and phenotypic changes in human breast
epithelial cells: evidence for enhanced interactions with the insulin-like growth
factor-I receptor. Mol. Carcinog. 18, 19–25 (1997).
125. Hansen, B. F., Kurtzhals, P., Jensen, A. B., Dejgaard, A. & Russell-Jones, D. Insulin
X10 revisited: a super-mitogenic insulin analogue. Diabetologia 54, 2226–2231
(2011).
126. Hansen, B. F. et al. Molecular characterisation of long-acting insulin analogues in
comparison with human insulin, IGF-1 and insulin X10. PLoS ONE 7, e34274
(2012).
127. Hernández Mijares, A., Solá Izquierdo, E., García Malpartida, K. & Verge, D.
[Safety of insulin analogues: what to evaluate, how to do it, and how to interpret the
results]. Endocrinol Nutr 57, 376–380 (2010).
128. Malaguarnera, R. et al. Proinsulin binds with high affinity the insulin receptor
isoform A and predominantly activates the mitogenic pathway. Endocrinology 153,
2152–2163 (2012).
129. Galloway, J. A. et al. Biosynthetic human proinsulin. Review of chemistry, in vitro
and in vivo receptor binding, animal and human pharmacology studies, and clinical
trial experience. Diabetes Care 15, 666–692 (1992).
130. Bryhni, B., Arnesen, E. & Jenssen, T. G. Associations of age with serum insulin,
proinsulin and the proinsulin-to-insulin ratio: a cross-sectional study. BMC Endocr
Disord 10, 21 (2010).

108
131. Wiberg, B., Sundström, J., Zethelius, B. & Lind, L. Insulin sensitivity measured by
the euglycaemic insulin clamp and proinsulin levels as predictors of stroke in elderly
men. Diabetologia 52, 90–96 (2009).
132. Pfützner, A. et al. Intact and total proinsulin: new aspects for diagnosis and treatment
of type 2 diabetes mellitus and insulin resistance. Clin. Lab. 50, 567–573 (2004).
133. Schneider, D. J. & Sobel, B. E. PAI-1 and diabetes: a journey from the bench to the
bedside. Diabetes Care 35, 1961–1967 (2012).
134. Fattal, P. G., Schneider, D. J., Sobel, B. E. & Billadello, J. J. Post-transcriptional
regulation of expression of plasminogen activator inhibitor type 1 mRNA by insulin
and insulin-like growth factor 1. J. Biol. Chem. 267, 12412–12415 (1992).
135. Jag, U. R., Zavadil, J. & Stanley, F. M. Insulin acts through FOXO3a to activate
transcription of plasminogen activator inhibitor type 1. Mol. Endocrinol. 23, 1587–
1602 (2009).
136. Nordt, T. K., Schneider, D. J. & Sobel, B. E. Augmentation of the synthesis of
plasminogen activator inhibitor type-1 by precursors of insulin. A potential risk
factor for vascular disease. Circulation 89, 321–330 (1994).
137. Samad, F. & Loskutoff, D. J. Tissue distribution and regulation of plasminogen
activator inhibitor-1 in obese mice. Mol. Med. 2, 568–582 (1996).
138. Oishi, K. et al. Tissue-specific augmentation of circadian PAI-1 expression in mice
with streptozotocin-induced diabetes. Thromb. Res. 114, 129–135 (2004).
139. Simpson, A. J., Booth, N. A., Moore, N. R. & Bennett, B. Distribution of
plasminogen activator inhibitor (PAI-1) in tissues. J. Clin. Pathol. 44, 139–143
(1991).
140. Alessi, M.-C. et al. Plasma PAI-1 levels are more strongly related to liver steatosis
than to adipose tissue accumulation. Arterioscler. Thromb. Vasc. Biol. 23, 1262–
1268 (2003).
141. Diggs-Andrews, K. A. et al. Brain insulin action regulates hypothalamic glucose
sensing and the counterregulatory response to hypoglycemia. Diabetes 59, 2271–
2280 (2010).
142. Mithieux, G., Rajas, F. & Gautier-Stein, A. A novel role for glucose 6-phosphatase
in the small intestine in the control of glucose homeostasis. J. Biol. Chem. 279,
44231–44234 (2004).
143. Mutel, E. et al. Control of blood glucose in the absence of hepatic glucose
production during prolonged fasting in mice: induction of renal and intestinal
gluconeogenesis by glucagon. Diabetes 60, 3121–3131 (2011).
144. Wu, C., Okar, D. A., Kang, J. & Lange, A. J. Reduction of hepatic glucose
production as a therapeutic target in the treatment of diabetes. Curr. Drug Targets
Immune Endocr. Metabol. Disord. 5, 51–59 (2005).
145. Bryant, N. J., Govers, R. & James, D. E. Regulated transport of the glucose
transporter GLUT4. Nat. Rev. Mol. Cell Biol. 3, 267–277 (2002).
146. Cano, N. Bench-to-bedside review: glucose production from the kidney. Crit Care 6,
317–321 (2002).
147. Mithieux, G., Gautier-Stein, A., Rajas, F. & Zitoun, C. Contribution of intestine and
kidney to glucose fluxes in different nutritional states in rat. Comp. Biochem. Physiol.
B, Biochem. Mol. Biol. 143, 195–200 (2006).
148. Watford, M. Is the small intestine a gluconeogenic organ. Nutr. Rev. 63, 356–360

109
(2005).
149. Gerich, J. E., Meyer, C., Woerle, H. J. & Stumvoll, M. Renal gluconeogenesis: its
importance in human glucose homeostasis. Diabetes Care 24, 382–391 (2001).
150. Fernandes, J. & Blom, W. The intravenous L-alanine tolerance test as a means for
investigating gluconeogenesis. Metab. Clin. Exp. 23, 1149–1156 (1974).
151. Stumvoll, M. et al. Human kidney and liver gluconeogenesis: evidence for organ
substrate selectivity. Am. J. Physiol. 274, E817–26 (1998).
152. Chiasson, J. L., Liljenquist, J. E., Sinclair-Smith, B. C. & Lacy, W. W.
Gluconeogenesis from alanine in normal postabsorptive man. Intrahepatic
stimulatory effect of glucagon. Diabetes 24, 574–584 (1975).
153. Stumvoll, M., Meyer, C., Kreider, M., Perriello, G. & Gerich, J. Effects of glucagon
on renal and hepatic glutamine gluconeogenesis in normal postabsorptive humans.
Metab. Clin. Exp. 47, 1227–1232 (1998).
154. Bronsveld, H. K. et al. Treatment with insulin (analogues) and breast cancer risk in
diabetics; a systematic review and meta-analysis of in vitro, animal and human
evidence. Breast Cancer Res. 17, 100 (2015).
155. Chen, Y. Q. et al. Sp1 sites mediate activation of the plasminogen activator
inhibitor-1 promoter by glucose in vascular smooth muscle cells. J. Biol. Chem. 273,
8225–8231 (1998).
156. Levy, J. R., Ullrich, A. & Olefsky, J. M. Endocytotic uptake, processing, and
retroendocytosis of human biosynthetic proinsulin by rat fibroblasts transfected with
the human insulin receptor gene. J. Clin. Invest. 81, 1370–1377 (1988).
157. Herring, R., Jones, R. H. & Russell-Jones, D. L. Hepatoselectivity and the evolution
of insulin. Diabetes Obes Metab 16, 1–8 (2014).
158. Widera, A., Norouziyan, F. & Shen, W.-C. Mechanisms of TfR-mediated
transcytosis and sorting in epithelial cells and applications toward drug delivery. Adv.
Drug Deliv. Rev. 55, 1439–1466 (2003).
159. Evans, M., Schumm-Draeger, P. M., Vora, J. & King, A. B. A review of modern
insulin analogue pharmacokinetic and pharmacodynamic profiles in type 2 diabetes:
improvements and limitations. Diabetes Obes Metab 13, 677–684 (2011).
160. Liebl, A. et al. Ten years of experience with biphasic insulin aspart 30: from drug
development to the latest clinical findings. Drugs 72, 1495–1520 (2012).
161. Kratz, F., Müller, I. A., Ryppa, C. & Warnecke, A. Prodrug strategies in anticancer
chemotherapy. ChemMedChem 3, 20–53 (2008).
162. Bunn, H. F., Gabbay, K. H. & Gallop, P. M. The glycosylation of hemoglobin:
relevance to diabetes mellitus. Science 200, 21–27 (1978).
163. Intensive blood-glucose control with sulphonylureas or insulin compared with
conventional treatment and risk of complications in patients with type 2 diabetes
(UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group. Lancet 352, 837–
853 (1998).
164. Valla, V. Therapeutics of diabetes mellitus: focus on insulin analogues and insulin
pumps. Exp Diabetes Res 2010, 178372–14 (2010).
165. Herold, K. C., Vignali, D. A. A., Cooke, A. & Bluestone, J. A. Type 1 diabetes:
translating mechanistic observations into effective clinical outcomes. Nat. Rev.
Immunol. 13, 243–256 (2013).
166. Sikaris, K. The correlation of hemoglobin A1c to blood glucose. J Diabetes Sci

110
Technol 3, 429–438 (2009).
167. Lebovitz, H. E. Insulin: potential negative consequences of early routine use in
patients with type 2 diabetes. Diabetes Care 34 Suppl 2, S225–30 (2011).
168. Sindelka, G., Heinemann, L., Berger, M., Frenck, W. & Chantelau, E. Effect of
insulin concentration, subcutaneous fat thickness and skin temperature on
subcutaneous insulin absorption in healthy subjects. Diabetologia 37, 377–380
(1994).
169. Russell-Jones, D. & Khan, R. Insulin-associated weight gain in diabetes--causes,
effects and coping strategies. Diabetes Obes Metab 9, 799–812 (2007).

Abstract (if available)

Abstract Long-acting insulin (INS) analogs have been used in the treatment of INS-deficient diabetes patients to meet INS needs between meal times and during night-time hours. Current products result in overexposure of INS to muscle and adipose tissues, with insufficient amount of INS delivered to the liver following subcutaneous (s.c) injection. This non-physiological distribution raises risks of severe hypoglycemia, and causes weight gain after long-term treatment. Therefore, development of better long-acting INS analogs are urgently required to enable prolonged and stringent glycemic control, while eliminating severe hypoglycemia. Since INS is self-administered by diabetic patients at home, a wider safe dosage range with lower risk of hypoglycemia also remains an unmet demand. ❧ Among a myriad of efforts in optimizing pharmacokinetic and pharmacodynamic (PK/PD) profiles of basal INS therapeutics, novel delivery devices applied to various administration routes are reported in an extensive volume of studies, e.g., micro-needle technology based INS patch in transdermal delivery, and INS loaded lipid-based particles in oral or pulmonary delivery. However, within the scope of this dissertation, the content focuses on s.c-injected INS therapeutics, including structure modified INS analogs with related formulation refinement. The importance of hepatopreferential action of exogenous INS as an attempt to mimic INS's physiological distribution is also emphasized. ❧ In contrast to current s.c-injected INS products, an ideal INS replacement requires the distribution or action of exogenous INS to more closely mimic physiological INS by demonstrating preferential hepatic action. However, specific to s.c-injected INS, factors that have traditionally barred products from restoring the liver:peripheral gradient in INS deficient diabetes patients include: (1) the physiological structures of subcutaneous and peripheral tissues

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Mechanism study of proinsulin-transferrin fusion protein as a long-acting and liver-targeting insulin prodrug

PDF

Characterization of rice-derived proinsulin-transferrin fusion protein and its oral bioactivity as a basal insulin analogue

PDF

Proinsulin-transferrin recombinant fusion protein: mechanism of activation and potential application in diabetes treatment

PDF

Reconstruction of a proinsulin-transferrin fusion protein

PDF

Characterization of the transferrin oligomer and its potential application in drug delivery

PDF

The peptide angiotensin-(1-7) as a novel treatment for complications induced by type 2 diabetes mellitus

PDF

Downregulation of osteopontin by PTEN in liver-specific Pten-null mouse model

Asset Metadata

Creator Shao, Juntang (author)

Core Title Proinsulin-transferrin fusion protein as a liver targeting insulin therapeutic in Type 1 diabetes treatment

School School of Pharmacy

Degree Doctor of Philosophy

Degree Program Pharmaceutical Sciences

Publication Date 06/09/2017

Defense Date 04/29/2016

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Diabetes,fusion protein,liver targeting,long-acting insulin therapeutic,OAI-PMH Harvest,proinsulin-transferrin

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Shen, Wei-Chiang (committee chair), McDonough, Alicia (committee member), Okamoto, Curtis T. (committee member), Rodgers, Kathleen (committee member), Stiles, Bangyan L. (committee member)

Creator Email juntangs@usc.edu,shaojuntang2011@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-250950

Unique identifier UC11281138

Identifier etd-ShaoJuntan-4427.pdf (filename),usctheses-c40-250950 (legacy record id)

Legacy Identifier etd-ShaoJuntan-4427.pdf

Dmrecord 250950

Document Type Dissertation

Format application/pdf (imt)

Rights Shao, Juntang

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA