Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Mechanism study of proinsulin-transferrin fusion protein as a long-acting and liver-targeting insulin prodrug
(USC Thesis Other)
Mechanism study of proinsulin-transferrin fusion protein as a long-acting and liver-targeting insulin prodrug
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
MECHANISM STUDY OF PROINSULIN-TRANSFERRIN FUSION PROTEIN AS A
LONG-ACTING AND LIVER-TARGETING INSULIN PRODRUG
By
Yuqian Liu
______________________________________________________________________________
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 2018
Copyright 2018 Yuqian Liu
ii
DEDICATION
To my parents, Shuren Liu and Shuxia Du, for their unconditional love and supports.
iii
ACKNOWLEDGEMENT
First of all, I would like to express my sincere appreciation to my advisor Dr. Wei-
Chiang Shen, the best mentor I could ever wish to have. This dissertation would not be possible
without his genuine guidance and intelligent advice. His passion and devotion to science,
wisdom and experience in life, kindness to people, have constantly educated and motivated me
during my life of being a graduate student. He is always available to help when I encountered
any difficulties either in research or in daily life. The tremendous respect and freedom I received
from Dr. Shen are just as much as the supports. The same gratefulness also goes to Dr. Shen’s
wife, Mrs. Daisy Shen. Daisy is always so supportive and so caring to all the students. As an
international student being in the States without family, the lavish hospitality given by Dr. Shen
and Daisy at every holiday really made me feel warm and less lonely.
I want to acknowledge all my committee members including Dr. Curtis Okamoto, Dr.
Bangyan Stiles, Dr. Kathleen Rogers, and Dr. Jang-Hyun Youn, for their insights and
suggestions regarding my dissertation work since the qualifying exam. I also want to thank Dr.
Jennica Zaro, who has given me valuable advice and feedback on my research work.
Furthermore, I really appreciate all the friendship and assistance from my labmates,
without whom my graduate life would be much more struggling. I want to thank Hsin-Fang Lee
for training me all the lab techniques since the first day I joined the lab. I am grateful to have
these peer labmates, Dr. Zoe Folchman-Wagner, Dr. Li Zhou and Hsuan-Yao (Sean) Wang. We
entered the program at the same time and had been through everything together. The supports
they provided and the faith we shared pulled me through all those tough times. I also would like
to thank all the current and previous members in our lab, Dr. Chunmeng Sun, Dr. Tzyy-Harn
(Mindy) Yeh, Yang Su, as well as my other fellow friends in the school, Dr. Yu Jeong Kim and
iv
Zhe Li. I am so blessed to have the chance to work and study with these wonderful people during
the past six years.
I deeply appreciated the endless love and unconditional supports from my parents. They
made me who I am today and will always love me no matter what I am. They always believe I
am doing something great and always cheer for me for even the tiniest bit of achievement. Their
expectations and encouragements drive me towards being a better person.
At the end, I want to acknowledge all the people and all the things I have encountered
during my graduate school life. Good or bad, up or down, high or low, beautiful or ugly, together
they made up an unneglectable part of my life, which I will always be missing.
Studies in this dissertation were partially supported by a JDRF Grand, 2-SRA-2016-239-
SB.
v
TABLE OF CONTENTS
DEDICATION .............................................................................................................................. ii
ACKNOWLEDGEMENT ........................................................................................................... iii
LIST OF FIGURES AND TABLES ......................................................................................... viii
ABBREVIATIONS ....................................................................................................................... x
ABSTRACT ................................................................................................................................. xii
CHAPTER 1. INTRODUCTION ................................................................................................ 1
1.1 Strategies for Liver-Targeted Drug Delivery ..................................................................................1
1.1.1 Passive Targeting .........................................................................................................................1
1.1.2 Receptors ......................................................................................................................................1
1.1.3 Transporters ..................................................................................................................................2
1.1.4 Enzymes .......................................................................................................................................3
1.2 Current Therapies for Insulin Resistance/Type 2 Diabetes ..........................................................3
Insulin resistance and Type 2 Diabetes .................................................................................................3
1.2.1 Lifestyle Changes .........................................................................................................................4
1.2.2 Metformin ....................................................................................................................................5
1.2.3 Sulfonylureas ................................................................................................................................5
1.2.4 Thiazolidinediones .......................................................................................................................6
1.2.5 Insulin ...........................................................................................................................................6
1.3 Proinsulin-Transferrin Fusion Protein ............................................................................................7
CHAPTER 2. BINDING MECHANISM CHARACTERIZATION OF ACTIVATED
PROINSULIN-TRANSFERRIN FUSION PROTEIN .............................................................. 9
2.1 Background ........................................................................................................................................9
2.2 Materials and methods ....................................................................................................................12
2.2.1 Cell Cultures ...............................................................................................................................12
2.2.2 Protein Preparation .....................................................................................................................12
2.2.3 Akt Phosphorylation Assays ......................................................................................................13
2.2.4 Insulin Resistance Induction by Palmitate .................................................................................13
2.2.5 Insulin Receptor Binding Assays ...............................................................................................14
2.2.6 Statistical Analysis .....................................................................................................................15
2.3 Results ...............................................................................................................................................15
vi
2.3.1 Akt Phosphorylation Induced by Active and Inactive ProINS-Tf .............................................15
2.3.2 Time Course Akt Phosphorylation Assay ..................................................................................16
2.3.3 Pulse-Chase Akt phosphorylation Assay ...................................................................................16
2.3.4 Akt Phosphorylation Assay with Competition of Excess Tf .....................................................17
2.3.5 INS Receptor Binding of INS and ProINS-Tf ...........................................................................17
2.3.6 INS Receptor Binding of INS and Activated ProINS-Tf with Competition of Excess Tf ........17
2.3.7 Effect of Palmitate on Inducing INS Resistance in HepG2 Cells ..............................................18
2.3.8 Akt Phosphorylation Assay in INS Resistant Cells ...................................................................18
2.4 Discussion .........................................................................................................................................28
2.5 Summary ..........................................................................................................................................31
CHAPTER 3. BIODISTRIBUTION, ACTIVATION, AND RETENTION OF
PROINSULIN-TRANSFERRIN FUSION PROTEIN IN THE LIVER ............................... 32
3.1 Background ......................................................................................................................................32
3.2 Materials and Methods ...................................................................................................................33
3.2.1 Animals ......................................................................................................................................33
3.2.2 Protein Iodination .......................................................................................................................33
3.2.3 ProINS-Tf Conversion in Precision-Cut Tissue Slices ..............................................................34
3.2.4 ProINS-Tf Retention in Liver and Muscle Precision-Cut Tissue Slices ....................................35
3.2.5 Biodistribution Study .................................................................................................................35
3.2.6 Statistical Analysis .....................................................................................................................36
3.3 Results ...............................................................................................................................................36
3.3.1 Ex Vivo Conversion of ProINS-Tf in Precision-Cut Tissue Slices ............................................36
3.3.2 ProINS-Tf Retention in Liver and Muscle Precision-Cut Tissue Slices ....................................37
3.3.3 Biodistribution of
125
I-Tf,
125
I-INS and
125
I-ProINS-Tf in CF-1 Mice .......................................37
3.3.4 Retention of
125
I-ProINS-Tf in Various Organs in CF-1 Mice ..................................................38
3.3.5 Intact
125
I-ProINS-Tf Levels in Liver and Other Organs ...........................................................38
3.4 Discussion .........................................................................................................................................48
3.5 Summary ..........................................................................................................................................50
CHAPTER 4. EFFICACY STUDY OF PROINSULIN-TRANSFERRIN FUSION
PROTEIN IN NON-OBESE DIABETES MICE ..................................................................... 51
4.1 Background ......................................................................................................................................51
4.2 Materials and Methods ...................................................................................................................53
4.2.1 Animals ......................................................................................................................................53
vii
4.2.2 Hypoglycemic Efficacy Study in NOD Mice ............................................................................53
4.2.3 PAI-1 Secretion in CF-1 Mice ...................................................................................................54
4.2.4 IGF-1 and IGF-1R Phosphorylation Measurements in CF-1 Mice ............................................54
4.2.5 IGF-1R Phosphorylation in HepG2 Cells ..................................................................................54
4.2.6 Long-Term Protein Treatment in NOD Mice ............................................................................55
4.2.7 Detection of Anti-Drug Antibodies Developed in NOD Mice ..................................................55
4.3 Results ...............................................................................................................................................56
4.3.1 Diabetes Development in Female NOD Mice ...........................................................................56
4.3.2 Hypoglycemic Efficacy of ProINS-Tf in Fasted NOD Mice .....................................................56
4.3.3 Hypoglycemic Efficacy of ProINS-Tf in NOD Mice under a Single Fasting/Feeding Cycle ...57
4.3.4 Hypoglycemic Efficacy of ProINS-Tf in NOD Mice under Multiple Fasting/Feeding Cycles .58
4.3.5 24-Hour Meal Challenge Test in NOD Mice .............................................................................58
4.3.6 PAI-1 Secretion in CF-1 Mice ...................................................................................................59
4.3.7 IGF-1 and IGF-1 Receptor Phosphorylation in CF-1 Mice .......................................................59
4.3.8 IGF-1 Receptor Phosphorylation in HepG2 Cells .....................................................................59
4.3.9 Long-Term ProINS-Tf Treatment in NOD Mice .......................................................................60
4.3.10 Anti-Drug Antibody Detection in NOD Mice .........................................................................60
4.4 Discussion .........................................................................................................................................73
4.5 Summary ..........................................................................................................................................76
CHAPTER 5. SUMMARY ......................................................................................................... 78
5.1 Conclusions ......................................................................................................................................78
5.2 Future Perspectives .........................................................................................................................82
5.2.1 Identification of irINS-Tf Structure and In Vitro Activation of ProINS-Tf ...............................82
5.2.2 Exploration of Therapeutic Potentials of ProINS-Tf in T2D Animal Model ............................83
REFERENCES ............................................................................................................................ 85
viii
LIST OF FIGURES AND TABLES
Figure 1.1. ProINS-Tf conversion to irINS-Tf in H4IIE hepatoma cells ...................................... 8
Figure 2.1.Bivalent binding mechanism proposed for activated ProINS-Tf. .............................. 11
Figure 2.2. Akt phosphorylation in HepG2 cells. ........................................................................ 19
Figure 2.3. Time-course Akt phosphorylation in HepG2 cells. ................................................... 20
Figure 2.4. Pulse-chase Akt phosphorylation assay in HepG2 cells. ........................................... 21
Figure 2.5. Akt phosphorylation assay with Tf competition in HepG2 cells. .............................. 22
Figure 2.6. IR binding profile of INS on HepG2 cells. ................................................................ 23
Figure 2.7. IR binding profiles of ProINS-Tf on HepG2 cells before and after its activation ..... 24
Figure 2.8. IR binding profiles of INS and activated ProINS-Tf on HepG2 cells with or without
Tf competition. ...................................................................................................................... 25
Figure 2.9. Palmitate induced INS resistance in HepG2 cells. .................................................... 26
Figure 2.10. Time-course Akt phosphorylation assay in INS resistant HepG2 cells. .................. 27
Figure 3.1. Conversion of ProINS-Tf to irINS-Tf in precision-cut tissue slices. ........................ 40
Figure 3.2. I
125
-ProINS-Tf retention in liver and muscle slices. .................................................. 41
Figure 3.3. Biodistribution of
125
I-Tf,
125
I-INS and
125
I-ProINS-Tf in CF-1 mice ...................... 42
Table 3.1. Dunnett’s multiple comparison test for biodistribution of
125
I-Tf,
125
I-INS and
125
I-
ProINS-Tf. ............................................................................................................................ 44
Figure 3.4. Retention of
125
I-ProINS-Tf in various organs in CF-1 mice. ................................... 45
Table 3.2. Estimated retention half-life of
125
I-ProINS-Tf in different tissues. ........................... 46
Figure 3.5. Percentage of intact
125
I-ProINS-Tf in liver over time. ............................................. 47
Table 4.1. Diabetes development process in female NOD mice. ................................................. 61
Figure 4.1. Hypoglycemic efficacy of ProINS-Tf in fasted NOD mice. ..................................... 62
Figure 4.2. Hypoglycemic efficacy of ProINS-Tf under one fasting/feeding cycle. ................... 63
Figure 4.3. Hypoglycemic efficacy of ProINS-Tf in NOD mice under multiple fasting/feeding
cycles. .................................................................................................................................... 64
Figure 4.4. 24-H Meal challenge test in NOD mice. ................................................................... 65
Figure 4.5. PAI-1 secretion in CF-1 Mice. .................................................................................. 66
Figure 4.6. Serum concentrations of IGF-1 in CF-1 mice. .......................................................... 67
ix
Figure 4.7. IGF-1R phosphorylation in liver tissue of CF-1 mice. .............................................. 68
Figure 4.8. IGF-1R phosphorylation in HepG2 cells. .................................................................. 69
Figure 4.9. 28-Day Fasting BG in NOD mice with different treatment. ..................................... 70
Table 4.2. Long-term treatment effect in NOD mice. .................................................................. 71
Figure 4.10. Anti-Tf and Anti-ProINS antibody detection in NOD mice serum. ........................ 72
Figure 5.1. Mechanism of ProINS-Tf as a long-acting and liver-targeted INS prodrug. ............ 81
x
ABBREVIATIONS
BCA Bicinchoninic acid
BG Blood glucose
BSA Bovine serum albumin
DMEM Dulbecco’s Modified Eagle Medium
FBS Fetal bovine serum
GAPDH Glyceraldehyde-3-phosphate dehydrogenase
HbA1c Hemoglobin A1c (glycated hemoglobin)
HDL High density lipoprotein
HEK293 Human embryonic kidney 293
IC
50
Half maximal inhibitory concentration
IGF-1R Insuin-like growth factor-1 receptor
INS Insulin
INS-Tf Insulin-transferrin
IR Insulin receptor
irINS-Tf Immunoreactive insulin-transferrin
i.v. Intravenous
LDL Low density lipoprotein
Ni-NTA Nickel-nitrilotriacetic acid
PAI-1 Plasminogen activator inhibitor-1
PBS Phosphate buffered saline
PD Pharmacodynamics
PK Pharmacokinetics
ProINS Proinsulin
ProINS-Tf Proinsulin-transferrin
RIA Radioimmunoassay
s.c. Subcutaneous
STZ Streptozotocin
T1D Type 1 diabetes
T2D Type 2 diabetes
TZD Thiazolidinedione
xi
Tf Transferrin
TFF Tangential flow filtration
TfR Transferrin receptor
xii
ABSTRACT
A recombinant proinsulin-transferrin fusion protein (ProINS-Tf) has been previously
reported to be a novel long-lasting insulin analog, acting specifically on the inhibition of hepatic
glucose output. In this dissertation, ProINS-Tf was investigated to elucidate the mechanism
responsible for its sustained and liver targeted anti-diabetic effects. Studies were carried out to
determine its binding properties, in vivo biodistribution, as well as the in vivo hypoglycemic
effects in Type 1 Diabetes mice. The biodistribution study revealed that ProINS-Tf exhibited
liver specific accumulation after a single intravenous injection, whereas transferrin (Tf) or
insulin (INS) showed relatively even distribution among different organs. The conversion of
inactive ProINS-Tf into an active immune-reactive INS-Tf form (irINS-Tf) via a Tf receptor
(TfR) mediated process only occurred in the liver, but not in other organs. In addition, ProINS-Tf
demonstrated a prolonged retention in the liver after an intravenous injection, suggesting the
enhanced association of the bifunctional active form, irINS-Tf, within liver cells. ProINS-Tf
after being activated possessed enhanced binding affinity to INS receptor (IR), as suggested by
the high Akt phosphorylation level and low half maximal inhibitory concentration (IC
50
)
measured from irINS-Tf treated hepatoma cells. The increased binding affinity of irINS-Tf to IR
was contributed by the bivalent binding to both IR and TfR on liver cells, leading to sustained
and hepatoselective hypoglycemic effects in Type 1 Diabetes mice. Taken together, studies from
this dissertation project indicate that ProINS-Tf is a highly liver-targeted INS prodrug with a
combination of 3 specific actions in liver cells: (1) TfR mediated binding and uptake of the
prodrug on the cell surface, (2) liver-specific, TfR-mediated conversion of the prodrug into its
active form, and (3) the bifunctional binding of the active fusion protein to both Tf and INS
receptors in the liver to achieve prolonged retention and thus enhanced anti-diabetic activities.
1
CHAPTER 1. INTRODUCTION
1.1 Strategies for Liver-Targeted Drug Delivery
The liver as the largest internal organ in mammals plays critical roles in metabolism,
detoxification, lipids and cytokines secretion, and inflammatory responses. Diseases like
hepatitis, alcoholic or non-alcoholic fatty liver disease, liver cirrhosis, and hepatocellular
carcinoma could cause mild to severe liver damages and pose great challenges to normal liver
functions. To improve the specificity and efficacy of drugs for treating liver diseases, several
systems have been utilized to achieve improved liver-targeted drug delivery.
1.1.1 Passive Targeting
The blood is supplied to the liver via sinusoids, which are larger than normal capillaries
and lined with endothelial cells and Kupffer cells [1]. The liver sinusoidal endothelium forms
open pores (fenestrae) with diameters of 150-175 nm, which allow rapid exchange of
macromolecules, solutes and fluid between the blood and the surrounding liver tissue [2, 3]. The
large gaps of the fenestrae, therefore, are often utilized for molecules or particles with relatively
large molecular size to achieve liver selective delivery by passive targeting. Several INS analogs
with hepatoselectivity, e.g. Peglispro [4] and Thyroxyl-INS analog [5], are designed to have
increased size and therefore have restricted transport across normal capillaries but can readily
diffuse through the liver sinusoidal endothelial cell fenestrae.
1.1.2 Receptors
Liver cells express a variety of receptors and conjugation of drugs to molecules that bind
to liver-specific receptors can be used to develop delivery scaffolds targeting to different liver
2
cells. Hepatocytes are rich in receptors including the asialoglycoprotein receptor, low-density
lipoprotein (LDL) and high-density lipoprotein (HDL) receptors, INS receptor (IR), transferrin
receptor (TfR), as well as receptors for other proteins such as epidermal growth factor and
transforming growth factor-beta. Kuffer cells contain receptors for agalactoglycoprotein, a2-
macroglobulin, and fibronectin, while endothelial cells have receptors for agalactoglycoprotein,
sulfated polysaccharides and immune complexes, as well as scavenger receptors that bind to
negatively charged proteins [6-8]. Many molecules or particles have been modified to
incorporate ligands targeting the hepatic receptors [9-12]. However, not every hepatic receptor
mediated delivery system is sufficient to achieve targeting effect to liver cells, especially when
drugs are administrated through systemic administration, because certain receptors are also
expressed in other tissues [6].
1.1.3 Transporters
Another strategy for liver specific drug deposition is to target the transporters that are
predominantly expressed at the surface of liver cells. Hepatic transporters facilitate the uptake
and excretion of xenobiotics and play important roles in the liver detoxification process [13].
Several families of transporters have been well characterized, including the organic anion-
transporting polypeptide (OATPs), the organic cation transporters (OCTs) and the organic anion
transporters (OATs) [14]. OATP transporters are involved in the uptake of bile acids, steroid
hormones, thyroid hormone, as well as drugs such as enalapril and atorvastatin, and their
substrates are often anionic and amphiphilic compounds [15]. The charged molecules can be
effluxed from hepatocytes into either the sinusoid or the bile, and therefore more efficient
transport by transporters on the sinusoidal membrane is required for a drug to maintain enough
3
residency time in the hepatocytes for achieving a pharmacological significance [16]. To design
the molecules towards OATP substrates for hepatic delivery, polar acidic moieties are commonly
incorporated because they contribute to the recognition by the transporters and reduce the
passive diffusion of the drug into off-target tissues [14].
1.1.4 Enzymes
The liver expresses a large number of enzymes to carry out its function in metabolizing
various compounds. The enzymes the liver can be employed to cleave pharmacologically
inactive prodrugs and thus generate hepatoselective effects. Such liver-specific prodrug
activation requires certain conditions: first, the catalyzing enzymes should express
predominantly in the liver; second, the enzymes should catalyze efficient activation of the
prodrug; and the cleavage of the prodrug moiety is independent of the drug [16]. Oseltamivir as a
prodrug example is activated in the liver by carboxylesterase 1A into a much more active
carboxylate form to inhibit the activity of viral neuraminidase and prevent the release of new
viral particles [17]. While the liver is rich in enzymes, many enzymes are not liver specific or
less liver specific. These enzymes can also be utilized for liver targeting if additional features of
the prodrug, such as molecular weight or recognition by hepatic receptors, could facilitate the
uptake of the prodrug by the liver [16].
1.2 Current Therapies for INS Resistance/Type 2 Diabetes
Insulin resistance and Type 2 Diabetes
Insulin (INS) is an essential endogenous ligand that functions in maintaining glucose
homeostasis and promote efficient glucose utilization. The metabolic activities of INS are
4
mediated through its binding to INS receptor (IR), which is a receptor tyrosine kinase whose
activation initiates a series of downstream activities involved in promoting glucose uptake,
inhibiting gluconeogenesis, as well as other functions to regulate glucose homeostasis [18]. INS
resistance commonly defined as decreased sensitivity or responsiveness in INS mediated
metabolic actions to a given level of INS. The primary compensatory response to INS resistance
by the body is increasing INS secretion, which leads to hyperinsulinemia. If the defect in INS
action could be overcome efficiently by hyperinsulinemia, glucose tolerance stays normal;
otherwise, the impaired glucose tolerance results in type 2 diabetes (T2D) [19]. T2D accounts for
more than 90% of the 415 million diabetes patients worldwide and leads to microvascular and
macrovascular complications, causing intensive distress to both the patients and the health-care
providers [20, 21].
1.2.1 Lifestyle Changes
Most T2D patients have some degree of overweight or obesity and managements in
dietary intake and physical exercise are therefore recommended as the primary intervention for
the treatment of diabetes [22, 23]. Dietary approaches are considered to attenuate the
abnormalities related to INS resistance syndrome from two aspects including enhancing INS
sensitivity and minimizing the adverse effects of the compensatory hyperinsulinemia [19].
Studies have shown that INS sensitivity was improved with weight loss in overweight/obese
individuals, in whom decreased plasma INS response to oral glucose and decreased triglyceride
concentration were measured [24]. Although there is not enough evidence to suggest an ideal
macronutrient content for T2D patient, research results have indicated that changing the relative
amounts or kinds of dietary carbohydrates, proteins and fat could help to modify the
5
abnormalities associated with INS resistance syndromes [19, 23]. In addition to facilitating
weight loss, physical exercise also helps to improve glycemic control and other metabolic
parameters in patients with T2D [23].When changing lifestyle alone is not sufficient to achieve
or maintain glycemic control, other pharmacological treatments will be added.
1.2.2 Metformin
Metformin is considered the first-line therapy for T2D unless specifically contraindicated
[20, 22, 23]. It can be given to T2D patients as a monotherapy or in combination with other
antidiabetic agents, e.g. sulfonylureas or thiazolidinedione [22, 25]. Metformin, an antidiabetic
biguanide derivative, improves insulin sensitivity by reducing hepatic glucose production and
enhancing peripheral glucose uptake, as well as increasing intestinal use of glucose and
decreasing fatty acid oxidation [26]. Although the primary target of metformin is the
mitochondria, the antidiabetic activities of metformin are mainly mediated through its secondary
effect on activating AMP-activated protein kinase, which is critical in protecting cellular
functions under energy-restricted conditions [27]. Metformin in general is well tolerated, but
contraindicated in patients with renal impairment or other factors that predispose to lactic
acidosis [23, 25].
1.2.3 Sulfonylureas
Sulfonylureas binds to specific receptors on pancreatic b-cells to stimulate INS secretion,
and is one of the most commonly used second-line agents for addition to metformin if glycemic
goals are not reached [20, 23, 25]. It is widely utilized because of its low cost and glucose-
lowering efficacy. Although sulfonylureas can be employed as monotherapy or dual therapy, the
6
efficacy of being a monotherapy is not as durable as metformin or thiazolidinedione [20]. This
type of drug is also associated with risks of hypoglycemia and weight gain as seen in other INS
therapy, due to the increased INS concentration resulted from the treatment [23].
1.2.4 Thiazolidinediones
The thiazolidinedione (TZD) is a class of agents that act as peroxisome proliferator
activated receptor g agonists to activate the receptor, whose activation induces responses that
affect INS action and lipid metabolism [28]. TZD improves INS sensitivity by acting on target
organs to increase glucose utilization and decrease glucose production [23, 29]. The use of TZD
has been controversial due to the associated side effects including weight gain and an increased
risk of congestive heart failure. Troglitazone (first-in-class) was withdrawn because of severe
hepatocellular injury [30] and rosiglitazone was suspended in Europe based on its adverse
cardiovascular outcomes [23, 31].
1.2.5 Insulin
INS remains the most effective treatment for patients with all types of diabetes. As
exogenous INS preparation varies from endogenous INS secretion (basal and prandial), INS
analogs with different pharmacokinetic (PK) profiles were developed, ranging from long-acting
INS analogs (Glargine, Detemir, Degludec), to intermediate-acting analog (Human NPH) and
rapid-acting analogs (Lispro, Aspart, Glulisine) [22]. As the disease stage of T2D progresses,
although some patients may be able to achieve glycemic control with only oral medications,
most patients eventually need to use INS therapy to manage the consequences of the diseases
[32]. While higher INS doses imply greater risk of hypoglycemia and weight gain, the
7
intensification of INS therapy in T2D patients needs to be well managed in order to minimize
those side effects. Long-acting INS analogs are usually initiated in patients to increase the basal
INS level because of the convenient treatment regimen and the lower risk of hypoglycemia [25,
32]. When basal INS has been up-titrated but no longer effective in lowering HbA1c, additional
agents including GLP-1 receptor agonists and mealtime INS or rapid-acting INS analogs, need to
be combined to cover the postprandial glucose excursions [22]. As the INS therapy gets
intensified with the progression of T2D, greater cost and complexity of the treatment regimen
will be resulted, which may require more patient education on dose adjustment and injection
techniques to guarantee the compliance.
1.3 Proinsulin-Transferrin Fusion Protein
A proinsulin-transferrin (ProINS-Tf) fusion protein was previously developed in our lab
and characterized to be a novel long-acting INS analog in type 1 diabetes (T1D) animal model
[33, 34]. This fusion protein is composed of two parts: a proinsulin (ProINS) moiety, which is
the inactive precursor of human INS, and a transferrin (Tf) moiety, which is a protein widely
used in improving drug delivery and prolonging drug half-life. ProINS-Tf displaces low potency
on initiating IR regulated activities and undergoes a Tf receptor (TfR) mediated slow recycling
process that converts the inactive ProINS to an active INS form in hepatocytes (Figure 1.1),
generating an activated immune-reactive INS-Tf (irINS-Tf) [33]. When being administered into
streptozotocin (STZ)-induced T1D mice, ProINS-Tf demonstrated a delayed but sustained blood
glucose (BG) lowering effect [34]. The hypoglycemic activity of ProINS-Tf is exerted after its
conversion to irINS-Tf and the conversion process occurred primarily in liver cells [35].
Therefore, ProINS-Tf behaves like a liver-specific prodrug, resulting in IR activation exclusively
8
in the liver but not in peripheral tissues [34]. Lack of peripheral activity from ProINS-Tf
treatment minimized the risk of causing severe hypoglycemia as seen in many other INS
therapies, making ProINS-Tf a potentially better basal INS analog for diabetes patients.
Although evidences have suggested the preferential activity of ProINS-Tf in the liver, the in vivo
liver-targeting effect remains unconfirmed and the exact function mechanism leading to the
prolonged and liver specific antidiabetic activity of ProINS-Tf has yet to be determined.
Figure 1.1. ProINS-Tf conversion to irINS-Tf in H4IIE hepatoma cells [33].
ProINS-Tf initially binds to TfR on cell membrane and is then endocytosed along with TfR to
the early endosome. Instead of being sorted to the degradation pathway, ProINS-Tf undergoes
the TfR-mediated slow recycling pathway, during which the compartment containing the fusion
protein encounters vesicles with secretory proteases that can convert ProINS-Tf to irINS-Tf.
Upon activation, irINS-Tf is recycled back to cell surface and released into the culture media.
9
CHAPTER 2. BINDING MECHANISM CHARACTERIZATION OF ACTIVATED
PROINSULIN-TRANSFERRIN FUSION PROTEIN
2.1 Background
ProINS, as a precursor of INS, has much lower binding affinity to INS receptor (IR) and
the resulted biologic potency was only 1% or less relative to INS [36]. Previous data has shown
that ProINS-Tf initially was inactive and required a lag time to be activated before it exhibited
activity on stimulating Akt phosphorylation in H4IIE cells [33]. The long-acting hypoglycemic
effect of ProINS-Tf observed in vivo therefore could be attributed to the active version of the
fusion protein, i.e. irINS-Tf. The activated fusion protein can be considered as a bifunctional
fusion protein because it contains two functional moieties INS and Tf. By combining two protein
domains together, bifunctional fusion proteins could often exhibit the properties of each of the
functional moiety and therefore achieve a better therapeutic profile [37-39]. Pharmacokinetics
(PK) and pharmacodynamics (PD) studies of bifunctional Tf fusion proteins suggested that the
addition of the Tf domain allowed retention of physiological functions of the native protein while
increasing the stability and plasma half-life, directing the protein to target specific tissues, and
assisting the delivery of the drug to otherwise inaccessible sites [40]. irINS-Tf as a bifunctional
protein with two active binding groups should be able to target to both IR and TfR. When bound
by two receptors simultaneously on the cell surface, this fusion protein may function as a
bivalent ligand, of which the binding to one receptor potentially facilitates the binding to another.
Therefore, as illustrated in Figure 2.1, we hypothesized that irINS-Tf might have an enhanced
binding affinity to IR as compared to native INS and thus result in higher level of IR activation
and subsequent cellular response, which then contributed to better BG lowering effect in vivo.
Here we planned to study the effect of ProINS-Tf fusion protein in binding to and activating IR
10
in hepatoma cells to characterize its binding mechanism, which might help us to better
understand the in vivo hypoglycemic effect seen for ProINS-Tf.
INS resistance leads to the eventual development of Type 2 diabetes (T2D) and it is often
referred as a state in which a higher than normal level of INS is required to achieve the normal
response [18, 41, 42]. The decreased cellular or tissue sensitivity to INS due to INS resistance
could be compensated by increasing the INS secretion from b-cells, however, hyperinsulinemia
would cause additional concerns such as abnormal regulation of free-fatty acid metabolism and
eventually lead to deterioration of INS resistance state [41, 43]. INS resistance could be resulted
from alterations at different cellular levels: prior to interacting with IR, at the level of the
receptor, or during the subsequent cellular activities involved in IR regulated pathway [42]. If the
alteration causing INS resistance involved interaction of INS to its receptor, a decreased
biological response would be resulted at a given concentration of INS, and such compromised
response might be attenuated if the receptor affinity to the ligand could be improved. Therefore,
the binding effect of ProINS-Tf in INS resistant cells was also going to be investigated in this
chapter. If the active irINS-Tf could achieve enhanced binding affinity to IR as compared to
native insulin, a better hypoglycemic efficacy in the treatment could be expected not only in
T1D, but also in T2D, of which INS resistance is a major concern.
11
Figure 2.1.Bivalent binding mechanism proposed for activated ProINS-Tf.
irINS-Tf possessing two active binding moieties is expected to be able to bind to IR and TfR
simultaneously. The binding to TfR may potentially facilitate the binding of the fusion protein to
IR, thus lead the fusion protein to achieve enhanced activation of IR-mediated signaling
pathways.
12
2.2 Materials and methods
2.2.1 Cell Cultures
Human embryonic kidney cell line HEK293, rat hepatoma cell line H4IIE and human
hepatocellular carcinoma cell line HepG2 were purchased from ATCC (Manassas, VA). All
three cell lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented
with 10% fetal bovine serum (FBS) and per ml 0.1 unit of penicillin/0.1 mg streptomycin
(Invitrogen). All mammalian cells were maintained in a 5% CO
2
humidified incubator
at 37°C.
Confluent cells were subcultured using 0.05% trypsin-EDTA on a regular basis.
2.2.2 Protein Preparation
ProINS-Tf fusion protein was produced and purified as previously described [33]. In
brief, DNA plasmids containing the ProINS-Tf fusion gene with His-tag at c-terminus were
transfected into HEK293 cells (ATCC, Manassas, VA). Fusion protein was collected from
conditioned medium, concentrated using tangential flow filtration systems (Millipore, Billerica,
MA), and further purified by applying through Ni-NTA chromatography. Tf and ProINS
moieties were both confirmed in the purified fusion protein by recognition of anti-Tf and anti-
ProINS antibodies in Western Blot assay.
Cell-activated ProINS-Tf fusion protein was prepared as previously described [33].
Briefly, ProINS-Tf (10 nM) was activated by incubation with confluent H4IIE cells for 24 h. The
conditioned medium containing partially converted ProINS-Tf was collected and used as cell-
activated ProINS-Tf in subsequent experiments, denoted as irINS-Tf.
Recombinant human INS and apo-Tf were purchased from Sigma (St. Louis, MO). INS
was dissolved in 0.1 M HCl and further diluted in PBS to desired concentrations. Recombinant
13
human ProINS was purchased from R&D systems (Minneapolis, MN). Apo-Tf was saturated
with iron by incubating with ferric ammonium citrate followed by dialysis in PBS. Receptor
grade radioactive INS labeled with I
125
on tyrosineA
14
was purchased from Perkin Elmer
(Waltham, MA).
2.2.3 Akt Phosphorylation Assays
Confluent HepG2 cells were starved in serum-free medium for 16-18 h before the
experiment, and subsequently starved cells were treated with different proteins for the indicated
period of time. After treatment, cells were washed and lysed with cell extraction buffer (Cell
Signaling Technology) supplemented with protease and phosphatases inhibitor cocktail (Thermo
Scientific) on ice. After BCA quantification, equal amount of total cellular proteins from each
treatment group was subjected to Western Blot analysis using anti-phospho-Akt antibody
(Ser473, Cell Signaling) and anti-GAPDH antibody (D16H11, Cell Signaling). Band densities
were quantified by Image Lab software (Bio-Rad, Hercules, CA).
In pulse-chase Akt phosphorylation assay, starved HepG2 cells were first pulsed in the
dosing medium containing different proteins for 30 min at 4°C. Cells were then washed 3 times
with ice-cold PBS and then incubated in protein-free DMEM medium at 37°C for indicated
period of time.
2.2.4 Insulin Resistance Induction by Palmitate
Bovine serum albumin (BSA)-palmitate complex preparation was adapted from previous
studies [44, 45]. In brief, 8% BSA and 8 mM sodium palmitate were dissolved separately in 150
mM NaCl solution. The two solutions were mixed at a 1:1 ratio and stirred for 1 h at 37°C to
14
produce a stock complex solution of 4% BSA/4 mM palmitate. To induce INS resistance in
HepG2 cells, cells were treated in BSA/palmitate solution diluted in serum-free DMEM for 16-
18 h and then applied to Akt phosphorylation assays.
2.2.5 Insulin Receptor Binding Assays
IR binding assays were performed in HepG2 cells, which were seeded in 24-well plates
and reached confluence before each experiment. Prior to dosing, cells were incubated in DMEM
with 0.1% BSA at 37°C for 30 mins to reach the equilibration and depletion of residual serum
components. Cells were then washed three times with PBS and dosed with ice-cold dosing
solutions containing
125
I-Tyr(A14)-INS and various concentrations of proteins diluted in DMEM
with 0.1% BSA for 2 h at 4°C. After incubation, cells were washed three times with ice-cold
PBS and then dissolved in 1N NaOH. Radioactivity associated with dissolved cell monolayers
was counted by the Packard Cobra II gamma counter and normalized to the amount of total
cellular proteins. Binding curves and corresponding IC
50
values were generated using GraphPad
Prism software by fitting the data with the non-linear regression model of one-site competitive
binding.
To prepare activated fusion protein, ProINS-Tf (10 nM) was incubated with confluent
H4IIE cells for 24 h. The conditioned medium containing partially activated ProINS-Tf was
serially diluted to subsequent lower concentrations, or concentrated by approximated 10-fold to
get a higher concentration of the active form. The concentration of the activated form, irINS-Tf,
in the conditioned medium was estimated by using INS specific RIA. INS or activated ProINS-
Tf were also applied to the IR binding assay on HepG2 cell monolayers with the addition of 100-
fold free Tf at each protein concentration.
15
2.2.6 Statistical Analysis
All data are represented as mean ± standard deviation (n ≥ 3) for all experiments. The
student’s t test was used to compare data sets, where P<0.05 was considered as statistically
significant different.
2.3 Results
2.3.1 Akt Phosphorylation Induced by Active and Inactive ProINS-Tf
INS binding to insulin receptor (IR) leads to activation of PI3K signaling pathway, in
which Akt will be phosphorylated as one of the downstream targets. In order to measure the
activity of ProINS-Tf in eliciting active INS signaling, Akt phosphorylation assay was performed
in HepG2 cells, which unlike H4IIE cells, have been shown lack of the activity to activate
ProINS-Tf [35]. ProINS-Tf (10 nM) was either directly applied to phospho-Akt experiment or
pre-treated with H4IIE cells for activation [19] and then the conditioned medium containing
active irINS-Tf (approximately 0.8 nM from INS RIA) was applied to HepG2 cells. As shown in
Figure 2.2, after 10 min stimulation, both INS and irINS-Tf induced significant levels of Akt
phosphorylation in the cells. Low Akt phosphorylation was observed in ProINS treated cells due
to the known low potency of ProINS to bind to and activate IR. ProINS-Tf demonstrated an even
less activity than ProINS on inducing Akt phosphorylation. The activity of ProINS-Tf in terms of
initiating IR signaling increased significantly after its conversion to irINS-Tf in H4IIE cells
(Figure 2.2).
16
2.3.2 Time Course Akt Phosphorylation Assay
To compare both onset and prolonged effect of INS and active irINS-Tf on inducing IR
activation, a time course Akt phosphorylation experiment was conducted. irINS-Tf demonstrated
much more increased onset effect in 10 min and 1 h treatments on causing Akt phosphorylation
compared to both 1 nM and 10 nM native INS treatment, suggesting its enhanced ability on
activating IR regulated pathway (Figure 2.3). In the more prolonged treatment, 4 h or 8 h, the
active fusion protein caused higher level of p-Akt than INS groups (Figure 2.3), indicating its
binding effect to IR was also sustained. The partially activated ProINS-Tf should contain no
more than 10nM of the active form irINS-Tf (assuming 100% conversion, the total concentration
would be 10nM), and in fact the concentration of irINS-Tf usually estimated as 0.6-0.8 nM by
INS specific RIA. Therefore, the activity of irINS-Tf observed here, which was higher than
10nM INS and still could be underestimated, confirmed its enhanced ability on binding to and
activating IR.
2.3.3 Pulse-Chase Akt Phosphorylation Assay
When cells were placed into fresh medium after the pulse phase, the cell associated
irINS-Tf caused significantly more Akt phosphorylation at 10 min and 1 h as suggested by
Figure 2.4, while INS demonstrated very low activity. After 4 h of chasing, the p-Akt level
stimulated by irINS-Tf decreased to a similar level as of INS (Figure 2.4). The higher activity of
irINS-Tf could be attributed to either more irINS-Tf than INS was bound by cells during the
pulse phase, or the cell associated irINS-Tf induced stronger IR activation or retention during the
chase phase.
17
2.3.4 Akt Phosphorylation Assay with Competition of Excess Tf
The effect of Tf in irINS-Tf on the IR binding affinity was investigated by including a
high concentration of Tf in a short-term Akt phosphorylation assay. As shown in Figure 2.5, with
addition of 100-fold excess Tf (1 µM), Akt phosphorylation level induced by 10 nM INS was not
significantly altered. On the other hand, for irINS-Tf the resulted Akt phosphorylation was
significantly decreased when excess Tf was present, suggesting a compromised activation of IR.
2.3.5 INS Receptor Binding of INS and ProINS-Tf
Relative IR binding affinities were measured for both INS and ProINS-Tf on HepG2 cell
monolayers. Compared to INS, which had a high affinity to IR, ProINS-Tf demonstrated a very
weak binding ability to IR (Figure 2.6 and 2.7). The estimated IC
50
values for INS and ProINS-
Tf were 7.29 nM and 1856 nM respectively, indicating the low affinity of the fusion protein on
binding to IR. The corresponding binding curve generated for ProINS-Tf barely dropped even at
the highest concentration (100nM) (Figure 2.7). After H4IIE cells-mediated conversion, the
partially activated ProINS-Tf showed significantly increased binding affinity to IR, with its IC
50
decreased to 18.54 nM (Figure 2.7), suggesting the conversion of the ProINS moiety to much
more active INS form that had higher binding affinity to IR. As estimated by INS specific RIA,
only 6.7% of the H4IIE cell-activated ProINS-Tf was converted irINS-Tf.
2.3.6 INS Receptor Binding of INS and Activated ProINS-Tf with Competition of Excess Tf
IR binding assay was carried out for INS and activated ProINS-Tf with the presence of
100-fold Tf at each protein concentration. The binding profile of INS was not affected with
addition of excess Tf, with IC
50
changed from 7.29 nM to 9.87 nM. The IR binding affinity of
18
H4IIE cell-activated ProINS-Tf was significantly decreased when 100-fold of Tf was present.
The IC
50
of activated ProINS-Tf increased from 18.54 nM to 321.5 nM, corresponding to an
approximate 17-fold decrease in IR binding affinity.
2.3.7 Effect of Palmitate on Inducing INS Resistance in HepG2 Cells
Sodium palmitate and BSA were formulated together to generate a 4 mM/4% palmitate-
BSA complex. Diluted concentrations of the palmitate complex ranging from 0.1mM to 1mM
were used to induce INS resistance in HepG2 cells. As shown in Figure 2.9A, the treatment of
palmitate caused decreased Akt phosphorylation in response to INS stimulation, without
affecting the expression level of IR. The induction of INS desensitization by palmitate was dose-
dependent, with higher concentration of palmitate causing lower p-Akt response. In cells treated
by 0.25mM palmitate, although p-Akt levels were lowered compared to non-palmitate treated
cells, INS was still able to induce increased amount of Akt phosphorylation with a higher
concentration (Figure 2.9B), indicating cells were resistant but still in response to INS
stimulation. In subsequent experiments, 0.25 mM palmitate was chosen to induce INS resistance
in HepG2 cells.
2.3.8 Akt Phosphorylation Assay in INS Resistant Cells
Akt phosphorylation stimulated by INS and irINS-Tf was evaluated in palmitate treated
HepG2 cells to compare their abilities on activating IR in an INS resistance model. INS
resistance was confirmed in palmitate treated cells by decreased level of p-Akt induced by INS
and irINS-Tf (Figure 2.10A). Also, more reduction of Akt phosphorylation in INS treatment was
noted in palmitate treated cells, as compared to the reduction in irINS-Tf treatment. Similar to
19
the time course Akt phosphorylation assay in regular HepG2 cells, irINS-Tf caused significantly
higher levels of p-Akt than INS both onset (10min) and prolonged (up to 8h) (Figure 2.10B).
10nM INS induced more Akt phosphorylation than 1nM INS (Figure 2.10B), suggesting this INS
desensitized cell model still showed dose-dependent response to INS. irINS-Tf demonstrated its
enhanced binding ability to IR in INS resistant cells by exhibiting a much stronger effect on
activating Akt with the same or even less concentration of INS.
Figure 2.2. Akt phosphorylation in HepG2 cells.
Cells were treated for 10 min with 10 nM of INS, ProINS, ProINS-Tf or irINS-Tf converted
from 10 nM of ProINS-Tf. Phospho-Akt band density was normalized with corresponding
GAPDH band density. Data were presented as the average values with error bars indicating the
standard deviations (N=3). ** indicated p<0.01.
20
Figure 2.3. Time-course Akt phosphorylation in HepG2 cells.
Starved HepG2 cells were treated with 1 nM of INS, 10 nM of INS, or irINS-Tf converted from
10 nM of ProINS-Tf for indicated period of time. Phospho-Akt band density was normalized
with corresponding GAPDH band density. Data were presented as the average values with error
bars indicating the standard deviations (N=3).
21
Figure 2.4. Pulse-chase Akt phosphorylation assay in HepG2 cells.
Starved HepG2 cells were treated with ice-cold dosing solution containing 10 nM of INS or
irINS-Tf converted from 10 nM of ProINS-Tf for 30 min at 4°C, and then chased in DMEM only
medium for indicated period of time at 37°C. Phospho-Akt band density was normalized with
corresponding GAPDH band density. Data were presented as the average values with error bars
indicating the standard deviations (N=3).
22
Figure 2.5. Akt phosphorylation assay with Tf competition in HepG2 cells.
Starved HepG2 cells were treated with 10 nM of INS or irINS-Tf converted from 10 nM of
ProINS-Tf for 10 min, with or without addition of 100-fold Tf (1µM). Phospho-Akt band density
was normalized with corresponding GAPDH band density. Data were presented as the average
values with error bars indicating the standard deviations (N=3). ** indicated p<0.01.
23
Figure 2.6. IR binding profile of INS on HepG2 cells.
Following binding at 4°C for 2 h, the cell-associated radioactivity was counted and normalized
by total cell protein. Data were presented as an average of percentage of binding relative to the
binding of the lowest INS concentration (N=3). IC
50
value of INS was estimated to be 7.29 nM
by using GraphPad Prism.
-12 -11 -10 -9 -8 -7
0
20
40
60
80
100
120
Log Concentration(M)
%Binding
INS
24
Figure 2.7. IR binding profiles of ProINS-Tf on HepG2 cells before and after its activation.
ProINS-Tf was either directly applied to the binding assay, or pre-incubated with H4IIE cells for
activation and then applied to the binding assay. Following binding at 4°C for 2 h, the cell-
associated radioactivity was counted and normalized by total cell protein. Data were presented as
an average of percentage of binding relative to the binding of the lowest ProINS-Tf
concentration (N=3). IC
50
values estimated for ProINS-Tf before and after activation were 1856
nM and 18.54 nM, respectively.
-12 -11 -10 -9 -8 -7
0
20
40
60
80
100
120
Log Concentration(M)
%Binding
ProINS-Tf
Partially activated
ProINS-Tf
25
Figure 2.8. IR binding profiles of INS and activated ProINS-Tf on HepG2 cells with or
without Tf competition.
At each concentration, 100-fold of Tf was added to compete for TfR binding. ProINS-Tf was
pre-incubated with H4IIE cells for activation and then applied to the binding assay. Following
binding at 4°C for 2 h, the cell-associated radioactivity was counted and normalized by total cell
protein. Data were presented as an average of percentage of binding relative to the binding of the
lowest protein concentration (N=3). IC
50
values estimated for INS and activated ProINS-Tf with
100-fold Tf competition were 9.87 nM and 321.5 nM, respectively.
-12 -11 -10 -9 -8 -7
0
20
40
60
80
100
120
Log Concentration (M)
%Binding
INS
INS + 100X Tf
-12 -11 -10 -9 -8 -7
0
20
40
60
80
100
120
Log Concentration (M)
%Binding
Activated ProINS-Tf
Activated ProINS-Tf+100XTf
26
Figure 2.9. Palmitate induced INS resistance in HepG2 cells.
A. Palmitate was able to cause INS resistance in HepG2 cells in a dose-dependent manner.
HepG2 cells were incubated with indicated concentration of palmitate complex for 16 h, and
then stimulated with 1 nM of INS for 10 min. B. INS was able to cause dose-dependent
response in palmitate treated HepG2 cells. HepG2 cells treated with 0.25 mM of palmitate
was stimulated by different concentrations of INS for 10 min.
B
A
27
Figure 2.10. A. Palmitate induced INS resistance in HepG2 cells.
HepG2 cells were incubated with 0.25 mM of palmitate complex for 16 h, and the development
of INS resistance was confirmed by decreased level of Akt phosphorylation when stimulating
with the same treatment of INS or irINS-Tf, i.e., 10 nM for 10 min. Levels of p-Akt induced by
INS or irINS-Tf decreased by approximated 70% and 21% respectively in palmitate treated cells.
B. Time-course Akt phosphorylation assay in INS resistant HepG2 cells.
HepG2 cells with INS resistance were treated with 1 nM of INS, 10 nM of INS, or irINS-Tf
converted from 10 nM of ProINS-Tf for indicated period of time. Phospho-Akt band density was
normalized with corresponding GAPDH band density. Data were presented as the average values
with error bars indicating the standard deviations (N=3).
A
B
28
2.4 Discussion
The hypoglycemic effects of ProINS-Tf as an inactive prodrug could be mainly attributed
to two key factors. Activation of the prodrug, and the interaction of the active form with IR.
Nearly all physiological activities of INS or INS-related analogs were exerted through binding to
IR. Therefore, how well a ligand can bind to and activate IR would be positively related to its
activity on triggering the blood glucose lowering events that were controlled within the IR
signaling pathway.
The mechanism study of ProINS-Tf on binding to IR was carried out on HepG2 cells.
HepG2 cells were reported to express extremely low levels of hepatic metabolism enzymes and
altered TfR recycling mechanism [46, 47], and showed in our previous study to be ineffective at
converting ProINS-Tf to its active form [48]. Using this cell line for studying the active and
inactive forms of the fusion protein could help to eliminate the background created by addition
activation of the fusion protein during the course of an actual experiment. HepG2 cells were also
shown to express relatively equal amount of IR and TfR [49], making it a good model to
investigate the potential bivalent binding mechanism of irINS-Tf.
When comparing to INS for the capability on inducing Akt phosphorylation, active
irINS-Tf demonstrated much higher activity under both onset (at 10 min) and prolonged (up to 8
h) conditions (Figure 2.2 and 2.3), suggesting its enhanced ability on activating IR. ProINS-Tf by
itself had little activity on Akt phosphorylation and became much more active after the H4IIE
cells-mediated conversion (Figure 2.2). This observation was consistent with our previous
finding and further proved this fusion protein to be an inactive prodrug. The increased p-Akt
levels induced by irINS-Tf could be due to two possible reasons: one was that the fusion protein
29
was simply more stable in the dosing medium and therefore could continuously act on the cells
to stimulate Akt activation; the other reason could be with the same amount of protein, the active
fusion protein could achieve enhanced and sustained binding with IR and thus cause increased
and prolonged activation effect. To see which one explained the action of irINS-Tf, a pulse-chase
Akt phosphorylation was performed, in which cells were pulsed with dosing proteins under 4°C,
and then chased in protein-free medium. In this case, all the Akt phosphorylation should be
induced by the proteins remained associated at the cell membrane after the pulse phase, therefore
could be a more reliable indication of the relative effect on activating IR. Compared to INS,
whose activity diminished very fast in the pulse-chase experiment, irINS-Tf showed a sustained
effect on causing Akt phosphorylation (Figure 2.4), suggesting its prolonged association with IR
in addition to the improved stability.
To see if the activity on causing Akt phosphorylation was actually correlated with the
binding affinity of the protein to IR, a series of IR competitive binding experiments were
performed. ProINS-Tf initially had very low affinity to IR and barely displaced any binding of
the radioactive
125
I-INS tracer even at its highest concentration (Figure 2.7). When the fusion
protein was pre-incubated with H4IIE cells for activation, its affinity to IR significantly
increased (Figure 2.7). This observation correlated very well with the Akt phosphorylation result
shown in Figure 2.2, in which the activity of ProINS-Tf increased significantly after conversion.
Since the inactive ProINS-Tf had very minimal binding effect to IR, the binding activity of the
partially activated fusion protein mixture should be mostly contributed by the active irINS-Tf.
After INS-specific RIA quantification, there was 0.67 nM irINS-Tf generated after the
conversion from 10 nM ProINS-Tf, approximately an 6.7% activation. The IC
50
of the irINS-Tf
30
could therefore be roughly estimated as 1.24 nM, which was even lower than the estimated IC
50
of native INS, indicating the active fusion protein possessed higher affinity to IR.
The enhanced affinity of irINS-Tf to IR could possibly be explained by the fact that the
active fusion protein was able to bind to both IR and TfR simultaneously as a bivalent ligand. As
shown in Figure 2.5 and Figure 2.8, with the addition of 100-fold Tf to partially block the TfRs
on the cell surface, the activity of irINS-Tf on both inducing Akt phosphorylation and binding to
IR was significantly reduced, suggesting loss of TfR binding compromised the binding of the
irINS-Tf to IR. The IC
50
of irINS-Tf in presence of excess Tf could be considered as the intrinsic
binding affinity of INS domain in irINS-Tf, which was the affinity of the fusion protein to IR
without the assistance of TfR binding. In other words, in this bivalent fusion protein, the irINS
moiety might have a decreased affinity to IR binding as compared to native INS. However, the
Tf moiety facilitated the membrane binding of the fusion protein through its binding to TfR,
resulting in the overall increased binding affinity of the fusion protein to IR.
As the enhanced binding affinity of the active fusion protein to IR being noticed, we started to
hypothesize that if such improved effect could be utilized in treating INS resistance in type 2
diabetes. INS resistance sometimes could be compensated by using increased concentration of
INS to achieve the original level of response. With the increased activity seen for irINS-Tf, we
hoped this fusion protein could achieve improved efficacy than the same concentration of INS in
INS resistant cells. In the preliminary study using a simplified INS resistance cell model,
although Akt phosphorylation caused by either INS or irINS-Tf was lower in INS-resistant cells
than in regular cells, irINS-Tf remained more potent than INS in the desensitized cells (Figure
2.10). The ratios of p-Akt levels induced by INS or irINS-Tf between INS resistant cells and
normal cells were estimated to be 0.3 and 0.79, corresponding to 70% decreased activity for INS
and 21% decreased activity for irINS-Tf. In addition, the roughly estimated difference in Akt
phosphorylation between irINS-Tf and 10 nM INS was 3-fold in regular cells and 5-fold in INS-
resistant cells (Figure 2.3 and 2.10), suggesting the greater advantage of irINS-Tf in INS-
31
resistant cells and therefore the potential application of irINS-Tf in restocking IR response to
attenuate INS resistance syndromes.
2.5 Summary
In this chapter, the binding mechanism of ProINS-Tf was evaluated for both the original
form and the active form. Both Akt phosphorylation assay and IR competitive binding assay
suggested ProINS-Tf had very low potency on activating IR, and it became a much more active
form after H4IIE cell-meidated conversion. As compared to INS, the active irINS-Tf possessed
higher affinity to IR through its bivalent binding to both IR and TfR. With assistance of TfR
binding, irINS-Tf was able to achieve improved binding affinity and prolonged activation of IR.
The active fusion protein also showed enhanced potency in INS-resistant cells and could be
further explored as a potential INS analog for the treatment of type 2 diabetes.
32
CHAPTER 3. BIODISTRIBUTION, ACTIVATION, AND RETENTION OF
PROINSULIN-TRANSFERRIN FUSION PROTEIN IN THE LIVER
3.1 Background
Endogenous INS is secreted by the pancreas directly into the liver through the hepatic
portal vein. After the liver extraction, the remaining approximately 20-50% of INS is then
dispersed into the systemic circulation. Therefore, liver is exposed to 2 to 4-fold higher
concentration of INS than peripheral tissues that are supplied by systemic circulation [50-52].
Conventional INS therapeutics are mostly administered to diabetes patients via the subcutaneous
route. Due to the lack of preferential delivery to the liver for establishing the liver/periphery INS
gradient, subcutaneously injected INS can render liver underinsulinized and peripheral tissues
hyperinsulinized [53, 54]. Overexposure of INS to periphery can cause many metabolic
abnormalities including severe hypoglycemia, dyslipidemia, excessive glycemic fluctuations,
weight gain and long-term micro- and macrovascular complications [55-58].
Several currently employed methods to restore the hepatic/peripheral INS gradient
include implantation of intraperitoneal insulin pumps [59, 60], pancreatic transplantation [61],
development of oral INS analogs [62] and INS analogs with increased molecular size to utilize
the larger vascular pore size of the hepatic endothelial cell layer [63, 64]. These methods are
associated with either practical difficulties and high cost or concerns of long-term use due to
hepatotoxicity or severe hypoglycemia caused by large amounts of active drug in the systemic
circulation.
Another potential strategy is to develop an INS analog that can target directly to, and
work preferentially in, the liver. As there are many different receptors expressed on liver cells,
33
high affinity ligand-receptor interactions can be utilized to develop hepatic delivery systems [6].
Transferrin (Tf), a high affinity ligand to transferrin receptor (TfR), has been actively explored as
a drug carrier for small molecules, peptides, proteins and nanoparticles to achieve improved
selectivity in TfR rich tissues [65, 66]. ProINS-Tf initially exhibits a very low anti-diabetic
activity. However, through TfR-mediated endocytosis and recycling process in hepatoma cells,
ProINS-Tf can be converted to an active form irINS-Tf [33]. In this chapter, we investigated the
biodistribution, retention and conversion of ProINS-Tf to demonstrate that the liver preferential
activity of ProINS-Tf is a combination of site-selective biodistribution, activation and retention
in the liver; the 3 factors that have long been considered as the criteria for the evaluation of the
targeted effect of a prodrug [67].
3.2 Materials and Methods
3.2.1 Animals
Male CF-1 mice (6-7 weeks old, 30-35 g) were purchased from Charles River
Laboratories (Wilmington, MA). Mice were housed at 12h light/ 12h dark cycles with room
temperature at 22±3°C and relative humidity at 50±20%. All mice had access to regular rodent
diet (Labdiet, St. Louis, MO) and water ad libitum. All animal studies were conducted in
accordance with NIH guidance and approved by the University of Southern California
Institutional Animal Care and Use Committee.
3.2.2 Protein Iodination
To prepare
125
I-labeled ProINS-Tf, Tf and INS, the proteins were iodinated with Na
125
I
using the chloramine-T method as previously described [68]. Solution containing 750-1000 µg of
34
Tf, INS or ProINS-Tf was mixed with 50 µl (about 0.25 mCi) of I
125
-Na (Perkin Elmer) and 50
µl of Chloramine-T solution (8 mg/ml) in a 1.5ml tube. The reaction was allowed to proceed for
10 min on ice with a gentle vortex every 3-4 mins in between. After 10 min reaction, 50 µl of
sodium metabisulfite solution (4.8 mg/ml) was added to stop the reaction and the reaction
mixture was incubated for another 5 min on ice. 100 µl of potassium iodine solution (10 mg/ml)
was then added and the mixture was kept on ice for 5 min. All of Chloramine-T, sodium
metabisulfite and potassium iodine solutions were freshly prepared before each iodination
process. After the conjugation reaction, the whole reaction mixture was loaded onto a 10 ml
PBS-equilibrated Sephadex G-50 size-exclusion column (GE Healthcare) for separation of I
125
-
labeled protein of interest. Ten of 1 ml fractions were collected from the column using PBS as
the mobile phase and counted by a gamma counter (Packard, Downers Grove, IL). Fractions
containing the highest counts were combined and stored at -20°C.
3.2.3 ProINS-Tf Conversion in Precision-Cut Tissue Slices
Tissues, including liver, intestine, lung, kidney, muscle and brain, were collected from
wild type male CF-1 mice and precisely sliced at 250 µm thickness using the Compresstome VF-
200-0Z (Precisionary Instruments Inc., Greenville, NC). Freshly prepared tissue slices were
incubated in William’s E medium (Life Technologies, Carlsbad, CA) with 25 mM D-glucose and
0.1% bovine serum albumin (BSA) under oxygenated (95% O
2
/5% CO
2
) condition. For the
conversion assay, medium containing 10 nM ProINS-Tf with or without 10 µM Tf was incubated
with different tissue slices on an orbital shaker shaking at 50 rpm under 37°C. At indicated time
points, concentrations of the conversion product irINS-Tf in the conditioned medium was
determined using human INS specific radioimmunoassay (RIA) kit (Millipore, Billerica, MA).
35
At the end of the incubation, tissue slices were homogenized and the total protein concentration
was quantified by bicinchoninic acid assay (BCA).
3.2.4 ProINS-Tf Retention in Liver and Muscle Precision-Cut Tissue Slices
Precision-cut liver and muscle slices were first incubated in 10 nM of I
125
-ProINS-Tf
diluted in William’s E medium containing 0.1% BSA. The incubation was placed on an orbital
shaker shaking at 50 rpm under 37°C. After 4 h incubation, the tissue slices were washed 3 times
in ice-cold PBS then continued to incubate in fresh medium. At 0, 60 and 120 min, tissue slices
were washed with ice-cold PBS and tissue associated radioactivity was counted by the gamma
counter and normalized to total tissue protein. The percentage of association was calculated as
the amount of tissue associated radioactivity relative to time 0.
3.2.5 Biodistribution Study
Biodistribution study was designed and conducted according to previous reports [69, 70].
Briefly,
125
I-Tf,
125
I-INS or
125
I-ProINS-Tf was injected into CF-1 mice intravenously (i.v.) at a
dose of 5 nmole/kg. At indicated time points, the mice were euthanized and blood, heart, lungs,
spleen, liver, intestine, kidneys, muscle and brain were collected and weighted. Tissue-associated
radioactivity was counted by using the gamma counter for the estimation of the amount of
protein in each tissue sample. For biodistribution analysis, protein concentrations in tissues
(nmole/g) were normalized to protein concentration in the blood (nmole/ml) to analyze the
relative targeting effect of the protein [69]. To measure levels of intact protein, tissues were
homogenized in lysis buffer and precipitated with 20% trichloroacetic acid (TCA) for 10 min at
36
4°C. The precipitate pellets which contained TCA-insoluble intact proteins, were counted after
centrifugation.
3.2.6 Statistical Analysis
All data are represented as mean ± standard deviation (n ≥ 3) for all experiments. The
student’s t test was used to compare data sets, where P<0.05 was considered as statistically
significant different.
3.3 Results
3.3.1 Ex Vivo Conversion of ProINS-Tf in Precision-Cut Tissue Slices
Precision-cut tissue slices were prepared and used for ex vivo activation of ProINS-Tf to
irINS-Tf. As shown in Figure 3.1A, only 2 out of 6 organs, i.e. liver and intestine, exhibited the
capacity of converting ProINS-Tf to irINS-Tf. In the liver slices incubation, the conversion of
ProINS-Tf to irINS-Tf was completely inhibited in the presence of 1000-fold excess Tf (Figure
3.1B), indicating the activation in liver was blocked when no TfR was accessible. This result was
consistent with the previous finding from the liver slices of C57BL/6J mice [21]. In intestine
slices incubation, however, no inhibition of the conversion was observed in the presence of
excess Tf, suggesting the conversion in intestine was not a TfR-mediated process. When the non-
TfR mediated conversion was subtracted from the total conversion, the result indicated that the
TfR-mediated conversion of ProINS-Tf occurred only in the liver slices (Figure 3.1C).
37
3.3.2 ProINS-Tf Retention in Liver and Muscle Precision-Cut Tissue Slices
125
I-ProINS-Tf demonstrated longer retention time in liver slices than in muscle slices.
The amount of muscle slices associated
125
I-ProINS-Tf dropped significantly to below 20% of
the initial amount within the first 60 min of incubation, and further decreased to no more than
10% after 2 h (Figure 3.2). In liver slices, the associated
125
I-ProINS-Tf maintained at a level that
was similar the initial amount throughout the course of 120 min incubation, without any
significant decrease being detected (Figure 3.2).
3.3.3 Biodistribution of
125
I-Tf,
125
I-INS and
125
I-ProINS-Tf in CF-1 Mice
CF-1 mice were injected with radioactive proteins to study the biodistribution. Blood and
major organs were collected at 1h, 4h and 8h post injection. The total radioactive equivalent
concentrations (nmole eq./g) in tissues of interest were calculated and normalized to radioactive
equivalent concentrations in blood (nmole eq./ml) to compare the targeting effects of Tf, INS and
ProINS-Tf in mice following a single i.v. dose (Figure 3.3).
As shown in Figure 3.3A and 3.3B, both Tf and INS were distributed evenly in most
organs (Table 3.1 for statistical analysis). The constant level of tissue concentration to blood
concentration ratio from 1 h to 8 h in all organs, except muscle, in
125
I-INS-injected mice
indicated the lack of tissue-specific INS accumulation. On the other hand, the significant
decrease of muscle-specific radioactivity after 1 h (Figure 3.3B) was consistent with the fact that
INS would initially bind to IR in muscle, followed by a rapid clearance [71].
125
I-Tf showed low
accumulation in all organs, including the liver (Figure 3.3A). The intravenously injected
125
I-Tf
was expected to be diluted by the high concentration of endogenous Tf in the blood, which
would compete with
125
I-Tf for TfR binding in many organs expressing high levels of TfR. On
38
the other hand, the tissue/blood (T/B) ratio of
125
I-ProINS-Tf in the liver increased significantly
over time, suggesting its targeting effect to the liver (Figure 3.3C). The hepatic targeting effect of
125
I-ProINS-Tf seen here was not simply due to the high expression of TfR or IR in the liver
because neither Tf nor INS exhibited increased T/B ratios over time.
3.3.4 Retention of
125
I-ProINS-Tf in Various Organs in CF-1 Mice
Biodistribution study of
125
I-ProINS-Tf was extended to 48h to study its retention in
different tissues. Radioactive equivalent concentrations of the fusion protein in organs and blood
were calculated at 1, 4, 8, 16, 24 and 48h after a single dose i.v. injection. As shown in Figure
3.4, highest concentrations in all tissues were observed at 1h post injection and the
concentrations continued to decease throughout the course of the experiment. To quantify the
relative rate of elimination of
125
I-ProINS-Tf from different tissues, the concentration curves
were analyzed and an estimated retention half-life was generated for each organ (Table 3.2 and
3.3). Retention half-lives in all tissues except liver were within 1.5-fold of that in the blood
(Table 3.2).
125
I-ProINS-Tf in the liver, on the other hand, possessed a much longer retention
half-life than that in the blood (~2.5-fold). This result was consistent with the observation in the
T/B biodistribution analysis as shown in Figure 3.3C that
125
I-ProINS-Tf specifically targeted to
and accumulated in the liver.
3.3.5 Intact
125
I-ProINS-Tf Levels in Liver and Other Organs
Since the tissue concentration of
125
I-ProINS-Tf was determined based on the
radioactivity, the detected signals could come either from the iodine associated with intact
protein or from small peptides or even free iodine as the products of protein degradation. To
39
determine the tissue associated
125
I-ProINS-Tf was in intact or degraded form, TCA precipitation
was performed with homogenized tissue lysates of livers collected at different time points. TCA-
insoluble radioactivity was normalized to total radioactivity in the lysate to generate the
percentage of intact
125
I-ProINS-Tf presented in the tissue (Figure 3.5A). The amount of intact
protein in the liver was approximately 75% at 1 h and stayed relatively stable from 8 h to 48 h at
90%, suggesting the integrity of the fusion protein accumulated in liver (Figure 3.5A).
TCA precipitation was also performed on other tissues collected at 48 h to compare the
intact protein contents in different organs (Figure 3.5B). For intestine, kidney, heart and spleen,
although 50 - 60% of the radioactivity were associated with intact
125
I-ProINS-Tf at 48 h, the
levels were significantly lower than that in the liver, i.e., 90%. This result indicated that
considerable amount of radioactivity in these organs was associated with free iodine or
degradation product(s). The difference of intact
125
I-ProINS-Tf distribution in the liver and in
other organs was even more striking when the amount of the fusion protein per gram tissue was
compared to each other (Figure 3.5C).
40
Figure 3.1. Conversion of ProINS-Tf to irINS-Tf in precision-cut tissue slices.
10nM ProINS-Tf was incubated with tissue slices with or without 10µM Tf for TfR blockage.
Concentration of irINS-Tf at indicated time point was determined from aliquots of the
conditioned medium by INS specific RIA and normalized to total protein concentration. A. Total
conversion in tissue slices. B. Non-TfR mediated conversion. C. TfR mediated conversion (total
– non-TfR mediated). Data were presented as the average values with error bars indicating the
standard deviations (N=3)
A
C
B
41
Figure 3.2. I
125
-ProINS-Tf retention in liver and muscle slices.
Liver and muscle slices were incubated with I
125
-ProINS-Tf for 4h then washed and placed in
fresh medium. Tissue associated amount of I
125
-ProINS-Tf was measure at different time points
and normalized to total tissue protein. Data were presented as a mean of percentage of the
associated I
125
-ProINS-Tf amount relative to the amount at time 0 (N=3).
30 90 0 60 120
0
20
40
60
80
100
120
Time(min)
% Tissue Associated I
125
-ProINS-Tf
Liver
Muscle
42
Figure 3.3. Biodistribution of (A)
125
I-Tf, (B)
125
I-INS and (C)
125
I-ProINS-Tf in CF-1 mice
after a single dose i.v. injection.
Tissues of interest were collected at 1, 4 and 8h post protein injection(5nmole/kg). The ratios of
radioactive equivalent protein concentrations in tissues (nmole/g) to protein concentrations in
blood (nmole/ml) were calculated. Data were presented as the mean ratios with error bars
indicating the standard deviations (N=4).
A B
C
43
A
B
44
Table 3.1. Dunnett’s multiple comparison test for biodistribution of (A)
125
I-Tf, (B)
125
I-INS
and (C)
125
I-ProINS-Tf.
For each protein at each time point, the data set of the overall tissue concentration (all
measurements at this time point, including all organs and all repeats, except for brain) was
compared with the data set of each organ using Dunnett’s multiple comparison test. *p<0.01,
***p<0.0001
C
45
Figure 3.4. Retention of
125
I-ProINS-Tf in various organs in CF-1 mice after a single i.v.
injection (5nmole/kg).
Tissue concentration of
125
I-ProINS-Tf in blood, heart, lungs, spleen, liver, intestine, kidney,
muscle, and brain were determined 1, 4, 8, 16, 24, and 48h post injection. Data were presented as
the average values with error bars indicating the standard deviations (N=4).
46
Table 3.2. Estimated retention half-life of
125
I-ProINS-Tf in different tissues.
Retention curves were analyzed by linear regression and a corresponding retention half-life was
generated for each organ.
Table 3.3. Parameters summary of retention half-life analysis. To calculate the retention half-
life shown in Table 3.2, linear regression analysis was performed to each organ using the tissue
concentrations at 1, 4, 8, 16 and 24 h. 48 h concentrations were not included in the analysis
because the concentrations did not decrease further from 24 h to 48 h in many organs.
Organ Retention Half-life (h)
Blood 4.61
Heart 7.27
Lungs 6.93
Spleen 6.58
Liver 11.72
Intestine 5.84
Kidneys 6.62
Muscle 6.06
Brain 7.77
47
Figure 3.5. A. Percentage of intact
125
I-ProINS-Tf in liver over time.
Livers samples collected at indicated time points were homogenized and subject to TCA
precipitation. %Intact
125
I-ProINS-Tf was calculated as the TCA-insoluble count over total count
of the tissue lysate. B. Percentage of intact
125
I-ProINS-Tf in liver, intestine, kidney, heart
and spleen at 48h post injection. C. Intact
125
I-ProINS-Tf concentration in liver, intestine,
kidney, heart and spleen at 48h post injection. Data were presented as the average values
with error bars indicating the standard deviations (N=4).
A B
C
48
3.4 Discussion
It has been demonstrated previously that precision-cut liver slices were capable of
activating ProINS-Tf ex vivo [35]. Here we wanted to know if liver was the only organ where the
activation of ProINS-Tf was possible. Among all the tested organs, the conversion of ProINS-Tf
to irINS-Tf was only observed in the liver and intestine slices (Figure 3.1). The conversion of
ProINS-Tf to irINS-Tf could happen under either a specific TfR-mediated activation [33] or non-
specific enzymatic digestion by various proteases [72, 73]. Although concentration of irINS-Tf
increased after incubation with both liver and intestine slices, only the conversion in liver slices
was a TfR-mediated process (Figure 3.1C). The high conversion rate in the intestine incubation
was most likely due to extracellular proteolysis by secreted proteases like trypsin. It has been
demonstrated previously that trypsin could efficiently convert ProINS to INS-like molecules [72,
73]. Therefore, large amount of ProINS-Tf could be converted in intestine and the concentration
of the active product did not increase further as the incubation time extended from 4 to 8 h. The
in vivo study also showed that, compared to liver in which nearly 90% of the fusion protein was
in intact protein form, most of the fusion protein had been degraded in the intestine (Figure
3.5B). The ex vivo and in vivo studies together confirmed that TfR-mediated intracellular
ProINS-Tf conversion occurred specifically in liver, which resulted in an intact and active form
of the fusion protein; while non-specific digestion of ProINS-Tf could happen in intestine, in
which the fusion protein might be degraded extracellularly and some of debris could be
recognized by the INS antibody used in the RIA.
Our previous study have demonstrated that in ProINS-Tf treated mice, prolonged IR
phosphorylation was only seen in liver but not in skeletal muscle, indicating the liver preferential
activity [34]. Here the biodistribution of ProINS-Tf was further explored in mice to confirm its
49
hepatic targeting effect. From the biodistribution study of Tf, INS and ProINS-Tf, liver targeting
effect was only seen in the ProINS-Tf treatment (Figure 3.3). Compared to INS, ProINS-Tf
demonstrated improved accumulation in liver and decreased accumulation in muscle, making it a
good INS analog to overcome the diminished liver/periphery INS gradient seen in conventional
INS therapeutics. However, only TfR-mediated binding and endocytosis is not sufficient to
achieve the liver-targeting effect because TfRs are expressed widely on the surface of all types of
cells [74, 75]. Additionally, Tf alone did not demonstrate significant liver targeting effect, as
shown in Figure 3.3A, that there was no selective uptake of
125
I-Tf by the liver. In fact, the
uptake of injected
125
I-Tf in all organs were relatively low, possibly due to the rapid recycling of
Tf-TfR complex in the cells and the competition by the high concentration of endogenous Tf in
the blood circulation (~ 5µM) [76].
Upon activation, it was demonstrated here that the retention of ProINS-Tf in the liver was
longer than in any other organs (Figure 3.4, Table 3.2). The concentration of
125
I-ProINS-Tf in
liver started to exceed the concentration in the blood 8 h post injection, indicating the onset of
the hepatic accumulation of the fusion protein (Figure 3.4). Previously we have observed that
there was a lag time of 4 to 6 h for ProINS-Tf to be activated in vivo before it reached a maximal
hypoglycemic effect in type 1 diabetes mice [34]. Therefore, the long liver retention we observed
here was likely contributed by the activated form of the fusion protein, irINS-Tf, due to its
enhanced bivalent-binding effect. The short retention of ProINS-Tf in ex vivo muscle slices
(Figure 3.2), therefore, could be possibly explained by two factors: the lack of ProINS-Tf
conversion to irINS-Tf in muscle and thus the absence of bivalent-binding effect of the protein to
achieve enhanced tissue association.
50
3.5 Summary
In this chapter, the ex vivo tissue slice conversion assay suggested that the TfR-mediated
conversion of ProINS-Tf to active irINS-Tf specifically occurred in liver but not in other organs.
Non-specific digestion of ProINS-Tf might happen in intestine due to the presence of excess
proteases, resulting in degraded form of proteins. Correspondingly, the in vivo biodistribution
analysis confirmed the liver-targeting effect of ProINS-Tf, while even distribution among most
organs were seen for both INS and Tf. Prolonged retention of ProINS-Tf with liver tissue both ex
vivo and in vivo further supported its enhanced association, which was thought to be contributed
by the liver-specific activation and bivalent-binding mechanism of the active fusion protein.
51
CHAPTER 4. EFFICACY STUDY OF PROINSULIN-TRANSFERRIN FUSION
PROTEIN IN NON-OBESE DIABETES MICE
4.1 Background
ProINS-Tf has been characterized previously in STZ-induced T1D mice to have
sustained and liver specific hypoglycemic effect. STZ-induced T1D mouse model although has
been widely used in the field of diabetes study, certain limitations are still associated with this
system and might have caused the results less relevant to T1D in humans. STZ is toxic to
pancreatic b-cells and resulted in loss of b-cells activity, leading to insulin deficiency and
development of T1D [77]. In addition to b-cells, STZ also could induce unselective toxic effects
on all macromolecules as well as in other organs including kidney and liver [77, 78]. The STZ-
induced T1D model is only suitable for short-term studies because the animals begin to exhibit
STZ-related toxicity and complications of severe diabetes mellitus starting at approximately 10
days post STZ induction [79]. In order to study the long-term effect of ProINS-Tf, as well as re-
evaluate the efficacy in T1D treatment, we decided to switch to use a well-established T1D
model, Non-Obese Diabetes (NOD) mice, to continue the in vivo study of ProINS-Tf. In the
genetic NOD mice, development of T1D is spontaneous and the identified mechanism shares
extended level of pathogenic and genetic similarities to the human condition, making NOD mice
a valuable tool for the study of T1D [80, 81].
The INS analog related side effects would also be evaluated in this chapter in normal CF-
1 mice. Development of several INS analogs with prominent hypoglycemic were suspended due
to increased incident of cancer risk [82]. The potential of inducing increased mitogenic effects
through activation of INS-like growth factor-I receptor (IGF-1R) had been noticed in many INS
analogs, raising more attention to the safety concerns of novel INS analogs. IGF-1R shared
52
similar structure to IR and therefore could bind to INS, but with much lower affinity.
Distinguished from IR signaling pathway, through which metabolic activities in regulating BG
and lipid homeostasis are induced, the activation of IGF-1R mainly enables mitogenic related
events in promoting cell proliferation, cell migration and cell survival [83]. INS analogs with
modified molecular structures for achieving improved therapeutic profiles might also have
altered the affinity to IGF-1R, causing increased mitogenic potential. As summarized in one
review study [84], long-acting INS analogs Glargine and Determir all showed decreased affinity
to IR and increased affinity to IGF-1R comparing to INS or other short-acting analogs,
suggesting long-acting analogs might be more reactive on stimulating the mitogenic pathway.
The effect of ProINS-Tf on activating IGF-1R therefore was going to be investigated to reveal its
mitogenic potential.
In addition to the mitogenic effect, the risk of myocardial infarction would also be
analyzed for ProINS-Tf. High ProINS concentration was suggested to be involved in
atherosclerosis and cardiovascular disease, and increased ratio of ProINS-to-INS had been
noticed in the progression of aging and T2D, both were associated with cardiovascular
complications [85, 86]. ProINS and INS had shown to be factors that could upregulate
plasminogen activator inhibitor-1 (PAI-1), which is a main regulator in fibrinolysis and
predisposes cardiovascular risks [87, 88]. ProINS-Tf fusion protein will be undergoing a
sustained conversion process once administered into the animal, leaving both ProINS and INS
moieties around. It was not known if coexistence of INS and ProINS resulted from ProINS-Tf
would cause higher level of PAI-1 secretion. Therefore, the effect of ProINS-Tf on PAI-1
secretion would be examined and compared to that of INS and ProINS.
53
4.2 Materials and Methods
4.2.1 Animals
Male CF-1 mice (6-7 weeks old, 30-35 g) were purchased from Charles River
Laboratories (Wilmington, MA) and female NOD/ShiLtJ mice (8 weeks old) were purchased
from Jackson Laboratory (Bar Harbor, ME). Mice were housed at 12h light/ 12h dark cycles with
room temperature at 22±3°C and relative humidity at 50±20%. All mice had access to regular
rodent diet (Labdiet, St. Louis, MO) and water ad libitum. Blood glucose (BG) levels of NOD
mice were measured through tail vein blood sampling weekly by using OneTouch glucose meter
(LifeScan, Milpitas, CA) with detection range between 20 and 600 mg/dL to monitor the
development of diabetes. NOD mice were classified into different stages of diabetes based on
measurement of the non-fasting BG concentrations. Mice with BG levels between 200 to 400
mg/dL were classified as in mild diabetic condition, 400 to 500 mg/dL as in moderate diabetes
and with two consecutive non-fasting BG measurements greater than 500 mg/dL as in severe
diabetes. All animal studies were conducted in accordance with NIH guidance and approved by
the University of Southern California Institutional Animal Care and Use Committee.
4.2.2 Hypoglycemic Efficacy Study in NOD Mice
NOD mice developed with severe diabetes (BG>500 mg/dL) were used for hypoglycemic
efficacy studies. For experiments under fasting condition, mice were pre-fasted for 3 h and then
given a single subcutaneous (s.c.) injection of proteins or PBS (vehicle). After injection, mice
were remained fasted and given free access to water until the end of experiment. For experiments
under free-feeding or alternating fasting/feeding conditions, mice were given free access to food
54
and water during feeding period and food was withdrawn during indicated fasting period. Small
blood samples (~10µl) were taken from the tail vein at indicated time points for BG
measurement. BG concentration higher than the detection limit the of glucose meter (600 mg/dL)
will be recorded as 600 mg/dL for subsequent statistical analysis.
4.2.3 PAI-1 Secretion in CF-1 Mice
CF-1 mice were s.c. injected with 22.5 nmole/kg of INS, ProINS-Tf or PBS once a day
for 5 days. At the end of the 5-day treatment, mice were euthanized and the blood was collected.
For each sample, serum was prepared from whole blood and tested by a mouse PAI-1 ELISA kit
(Abcam) to determine the serum concentration of PAI-1.
4.2.4 IGF-1 and IGF-1R Phosphorylation Measurements in CF-1 Mice
Normal CF-1 mice were intravenously (i.v.) injected with 5 nmole/kg of INS, ProINS-Tf
or PBS. Mice were euthanized 1 h or 8 h post injection and mice serum and liver samples were
harvested. Serum concentration of IGF-1 was quantified by IGF-1 ELISA (R&D System). Liver
samples were homogenized in lysis buffer supplemented with protease inhibitor cocktail. The
extracted proteins were applied to immunoprecipitation using anti-IGF-1R antibodies (Cell
Signaling) and protein A/G agarose beads (Santa Cruz). Precipitated proteins were subjected to
both anti-phospho-Tyrosine (Millipore) and anti-IGF-1R Western Blot analysis.
4.2.5 IGF-1R Phosphorylation in HepG2 Cells
Trypsin digested ProINS-Tf was prepared as a mimic of the activated ProINS-Tf [34].
Briefly, ProINS-Tf was mixed with trypsin in the ratio of 10:1 (w/w) and incubated at 37°C for
55
10 min. The reaction was stopped by adding Bowman-Birk soybean protease inhibitor to inhibit
excess trypsin activity. HepG2 cells were starved from serum for 16 h before experiment.
Starved HepG2 cells were treated with 100nM of INS, ProINS-Tf or trypsin digested ProINS-Tf
for 1 h or 8 h. After protein incubation, cells were lysed and the total protein concentration was
quantified by BCA. Equal amounts of total protein from each treatment were applied to phospho-
IGF-1R ELISA (Cell Signaling) to determine the relative levels of IGF-1R phosphorylation.
4.2.6 Long-Term Protein Treatment in NOD Mice
NOD mice with fasting BG concentrations greater than 200 mg/dL were given s.c.
injections of INS (22.5 nmole/kg), ProINS-Tf (22.5 nmole/kg) or PBS (200µl) once a day for 28
days. Mice were kept on a 12 h fasting/ 12 h feeding cycle daily after injection. BG and body
weight of the mice were monitored every 4 days by the end of the fasting period. Blood and
livers were collected from all of the treated mice and a group of untreated normal NOD mice
(BG<100 mg/dL) by the end of the study. Hemoglobin A1c (HbA1c) levels were measured by
using a DCA Vantage Analyzer (Siemens). Lipid panels were taken for all the mice using the
CardioChek Plus analyzer (PTS Diagnostics), which measured concentrations of total
cholesterol, HDL cholesterol and triglyceride.
4.2.7 Detection of Anti-Drug Antibodies Developed in NOD Mice
PVDF membranes were cut into 1cm
2
pieces and each piece was dot-blotted with 100 ng
of human Tf or ProINS and blocked in 5% milk at room temperature for 1 h. Serum of long-term
ProINS-Tf or PBS treated NOD mice was diluted in 5% milk and used to incubate with protein
dotted membranes. HRP-conjugated anti-mouse IgG secondary antibody (Thermo Fisher) was
56
used as secondary antibody to detect the existence of corresponding primary antibody in mice
serum.
4.3 Results
4.3.1 Diabetes Development in Female NOD Mice
The process of diabetes development in one batch of 40 female NOD mice was tracked to
estimate the incidence rate and onset of diabetes development in those mice. Mice with well-
developed hyperglycemia (BG>400 mg/dL) were characterized with diabetes for our study. All
mice were in normal glycemic condition (70-100 mg/dL) when received (Table 4.1). During
week 15, 2 out the 40 mice developed diabetes (Table 4.1). The majority of diabetes
development occurred between week 18 and week 22, indicating the idea period of time for
conducting experiments with sufficient number of diabetic mice. Another 7 mice developed
diabetes during the last 7 weeks before we finished monitoring this batch of mice. In total there
were 30 out of 40 mice developed diabetes within 30 weeks, corresponding to an incidence rate
of 75%.
4.3.2 Hypoglycemic Efficacy of ProINS-Tf in Fasted NOD Mice
Fasted diabetic NOD mice were injected with 22.5 nmole/kg of ProINS-Tf, INS or PBS
to evaluate their hypoglycemic efficacy. As shown in Figure 4.1, for mice treated with PBS, BG
levels exhibited slight decrease during the fasting experiment. However, such decrease was not
significant and the lowest average BG concentration within the group was still above 300 mg/dL.
Mice injected with INS showed a rapid decrease in BG levels within the first two hours after
injection, with the lowest BG achieved at 2 h post injection. Starting at 2 h, the BG lowering
57
effect of INS gradually diminished and the BG of INS treated mice returned to similar levels as
the vehicle group at 6 h. ProINS-Tf treated mice also showed BG decrease after injection,
although not as rapid as in the INS group. The maximum hypoglycemic effect of ProINS-Tf was
shown starting at 4 h, when BG of INS treated mice started to return to higher levels, and
continued till the end of the experiment. Starting at 6 h post injection, the BG of ProINS-Tf
group was significantly lower than both INS and PBS groups (Figure 4.1). Compared to INS,
although the onset of maximum hypoglycemic effect was delayed for ProINS-Tf, its effect was
more sustained, with the BG controlled under 100 mg/dL for 6 hours or longer (from 4 h to end
of experiment), while INS only controlled BG under 100 mg/dL for about 2 hours.
4.3.3 Hypoglycemic Efficacy of ProINS-Tf in NOD Mice under a Single Fasting/Feeding
Cycle
Hypoglycemic efficacy of ProINS-Tf at a medium dose (45 nmole/kg) was evaluated in
diabetic NOD mice under alternating fasting/feeding condition. As shown in Figure 4.2A, BG
levels of mice treated with ProINS-Tf gradually decreased under fasting condition, reached a
normoglycemic level (< 100 mg/dL) between 4 to 8 h. After switching to free-feeding at 8 h post
injection, the effect of ProINS-Tf soon diminished and BG levels returned to similar levels of the
control group (~ 550 mg/dL). When mice were first subjected to free-feeding condition after
injection at time 0, no BG lowering effect was observed in ProINS-Tf treated mice (Figure
4.2B). However, after the mice were switched to fasting condition at 8 h post injection, the BG-
lowering effect of ProINS-Tf started to show at 4 h and a normal BG concentration was achieved
between 6 to 8 h after the initiation of the fast, i.e., 14 to 16 h after the initial injection.
58
4.3.4 Hypoglycemic Efficacy of ProINS-Tf in NOD Mice under Multiple Fasting/Feeding
Cycles
To investigate the prolonged hypoglycemic efficacy of ProINS-Tf, NOD mice were given
an injection of 135 nmole/kg of the fusion protein or INS Glargine, an established long-lasting
INS analog on the market. After injection, mice were maintained under multiple 8 h fasting/8 h
feeding cycles for total of 72 hours. BG was monitored every 8 h by the end of each fasting or
feeding period. As shown in Figure 4.3, ProINS-Tf exhibited significant hypoglycemic effect for
up to 48 hours, which was much longer than the effect of Glargine (less than 16 hours). Similar
as in previous single cycle fasting/feeding experiment, the hypoglycemic response to ProINS-Tf
treatment was only significant during fasting periods. After each feeding period, the average BG
returned to high levels. Although BG lowering effect from a single dose of 135 nmole/kg of
ProINS-Tf could last for more than 40 h, severe hypoglycemia (BG<20mg/dl) was observed in
one mouse by the end of second fasting period (24 h), indicating a risk of hyperglycemia in NOD
mice from a high dose ProINS-Tf treatment.
4.3.5 24-Hour Meal Challenge Test in NOD Mice
NOD mice were administered with a single dose of 45 nmole/kg ProINS-Tf to explore
the potential of using ProINS-Tf as a once daily injection for achieving glycemic control within
24 h. Since the hypoglycemic effect of ProINS-Tf was minimal under feeding condition as
previously observed, the feeding period in this experiment was shortened to 2 h to better control
the food intake. Mice were kept at 8 h fasting/2 h feeding cycles for totally 24 h. As shown in
Figure 4.4, compared to the control, a single injection of ProINS-Tf at a dose of 45 nmole/kg was
59
able to control the BG level under fasting condition up to 24 h. The hypoglycemic effects soon
diminished each time when the mice were switched to feeding condition.
4.3.6 PAI-1 Secretion in CF-1 Mice
Serum concentrations of PAI-1 were measured in normal CF-1 mice after 5-day treatment
of INS, ProINS-Tf, ProINS or PBS. The average level of PAI-1 in ProINS-Tf treated mice
appeared to be slightly higher than that in other groups, but the difference was not statistically
significant due to the large variation within each group (Figure 4.5). INS, ProINS-Tf and ProINS
groups were all at similar levels of the control group, suggesting no significantly increased level
of PAI-1 was caused by any of the three protein treatments under our experimental conditions.
4.3.7 IGF-1 and IGF-1 Receptor Phosphorylation in CF-1 Mice
For both the serum IGF-1concentration and the liver IGF-1R phosphorylation, no
significant difference was detected among INS, ProINS-Tf and PBS treatments (Figure 4.6 and
Figure 4.7). As compared to the levels at 1 h, slight increase of IGF-1R phosphorylation was
seen in all three groups at 8 h (Figure 4.7). IGF-1 concentrations remained relatively constant at
both time points for all treatment groups (Figure 4.6).
4.3.8 IGF-1 Receptor Phosphorylation in HepG2 Cells
Absorbance at 450 nM was quantified as an indirect measure of the relative amount of
IGF-1R phosphorylation in HepG2 cells as instructed by the ELISA kit. As shown in Figure 4.8,
INS treatment resulted in increased IGF-1R phosphorylation after both 1 h and 8 h incubation as
compared to the control (Figure 4.8). ProINS-Tf by itself had no significant effect at both time
60
points on increasing IGF-1R phosphorylation. Trypsin digested ProINS-Tf, on the other hand,
resulted in higher amount of IGF-1R phosphorylation than both the control and INS treated
groups (Figure 4.8). The levels of IGF-1R phosphorylation for all four groups remained similar
between the two incubation periods, with no significant increase being noticed from 1 h to 8 h.
4.3.9 Long-Term ProINS-Tf Treatment in NOD Mice
Initially the average BG levels for ProINS-Tf, INS and PBS groups are 351±142,
251±135 and 358±173 mg/dL, respectively. As compared to INS or PBS treated groups, a
significant decrease of fasting BG level 12 h after injection was observed in ProINS-Tf treated
NOD mice up to 12 days (Figure 4.9). However, the extent of BG lowering effect of ProINS-Tf
was gradually reduced starting Day 8. No difference was observed between ProINS-Tf group and
the control group after 20 days of treatment. Daily injection of INS did not show significant
hypoglycemic effects after 12-h fasting as the BG level continued to increase throughout the
course of the experiment (Figure 4.9). By the end of the 28-day treatment, mice from all three
groups developed severe hyperglycemia with average BG near 600 mg/dL.
As shown in Table 4.2, body weight, blood level of lipids and HbA1c were measured for
all mice by the end of the 28-day treatment. Comparing to PBS treatment, ProINS-Tf and INS
both showed beneficial effects in maintaining body weight, lowering lipids and HbA1c levels.
4.3.10 Anti-Drug Antibody Detection in NOD Mice
Serum from the NOD mice that had been subjected to the 28-day treatment of ProINS-Tf
or PBS was applied to immune-dotting assays for detection of antibodies developed against
human Tf and human ProINS. As suggested by Figure 4.10A, anti-hTf antibody was detected in
61
all 3 mice treated with ProINS-Tf even at the highest titer (1:20000). No anti-hTf antibody was
detected in mice treated with PBS. No detectable anti-ProINS antibody was found in the serum
of mice treated with either ProINS-Tf or PBS (Figure 4.10B).
Table 4.1. Diabetes development process in female NOD mice.
40 female NOD mice were received in non-diabetic condition at week 8. BG levels were tracked
for each mouse on a weekly basis and mice with BG>400 mg/dL were characterized as in
diabetes.
62
Figure 4.1. Hypoglycemic efficacy of ProINS-Tf in fasted NOD mice.
NOD mice were pre-fasted for 3 h and s.c. injected with 22.5 nmole/kg of ProINS-Tf, INS or
200 µl of PBS at time 0. Mice were kept fasted and BG concentrations were measured at
indicated time post injection. Readings above 600mg/dL were presented as 600mg/dL due to
limitation of the detection range. Data were presented as the average values with error bars
indicating the standard deviations (N=3).
0 2 4 6 8 10
0
100
200
300
400
500
600
700
Time (h)
BG (mg/dL)
ProINS-Tf (22.5nmol/kg)
INS (22.5nmol/kg)
PBS
63
Figure 4.2. Hypoglycemic efficacy of ProINS-Tf under one fasting/feeding cycle.
A. Diabetic NOD mice were s.c. injected with 44.5 nmole/kg ProINS-Tf or PBS and subjected to
one 8 h fasting/8 h feeding cycle. BG concentrations were measured at indicated time points. B.
Diabetic NOD mice were administered with the same injections as in A and subjected to one 8 h
feeding/8 h fasting cycle. White zone indicated fasting phase and grey zone indicated free-
feeding phase. Readings above 600mg/dL were presented as 600mg/dL due to limitation of the
detection range. Data were presented as the average values with error bars indicating the
standard deviations (N=3).
0 4 8 12 16
0
100
200
300
400
500
600
700
Time (h)
BG (mg/dL)
ProINS-Tf (45nmole/kg)
PBS
A
0 4 8 12 16
0
100
200
300
400
500
600
700
Time (h)
BG (mg/dL)
ProINS-Tf(45nmol/kg)
PBS
B
64
Figure 4.3. Hypoglycemic efficacy of ProINS-Tf in NOD mice under multiple
fasting/feeding cycles.
NOD mice were s.c. injected with 135 nmole/kg of ProINS-Tf or INS Glargine at time 0. Mice
were kept under 8 h fasting/8 h feeding cycles for a total of 72 h. White zone indicated fasting
phase and grey zone indicated feeding phase. BG concentrations were measured at indicated time
points. Readings above 600mg/dL or below 20 mg/dL were presented as 600mg/dL or 20 mg/dL
due to limitation of the detection range. Data were presented as the average values with error
bars indicating the standard deviations (N=3).
0 8 16 24 32 40 48 56 64 72
0
100
200
300
400
500
600
700
Time (h)
BG (mg/dL)
ProINS-Tf (135nmol/kg)
INS-Glargine (135nmole/kg)
65
Figure 4.4. 24-H meal challenge test in NOD mice.
NOD mice were s.c. injected with 45 nmole/kg of ProINS-Tf or 200 µl of PBS at time 0. Mice
were kept under 8 h fasting/2 h feeding cycles for a total of 24 h. White zone indicated fasting
phase and grey zone indicated feeding phase. BG concentrations were measured at indicated time
points. Readings above 600mg/dL were presented as 600mg/dL due to limitation of the detection
range. Data were presented as the average values with error bars indicating the standard
deviations (N=3).
0 4 8 12 16 20 24
0
100
200
300
400
500
600
700
Time (h)
BG (mg/dL)
ProINS-Tf (45nmole/kg)
PBS
66
Figure 4.5. PAI-1 secretion in CF-1 Mice.
Normal CF-1 mice were s.c. injected daily with 22.5 nmole/kg of INS, ProINS-Tf, ProINS or
200 µl of PBS for 5 days. Mice serum was collected by the end of 5-day treatment and applied to
PAI-1 ELSIA to determine PAI-1 concentrations. Data were presented as the average values with
error bars indicating the standard deviations (N=3).
INS
ProINS-Tf
ProINS
PBS
0
10
20
30
40
PAI-1 concentation
(ng/ml)
PAI-1
67
Figure 4.6. Serum concentrations of IGF-1 in CF-1 mice.
CF-1 mice were i.v. injected with 5 nmole/kg of INS, ProINS-Tf or 50 µl of PBS and mice
serum was collected 1 h and 8 h post injection. Concentrations of IGF-1 was determined using
mouse IGF-1 ELISA. Data were presented as the average values with error bars indicating the
standard deviations (N=3).
1h 8h
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
IGF-1 concentration
(ng/ml)
INS
ProINS-Tf
PBS
IGF-1
68
Figure 4.7. IGF-1R phosphorylation in liver tissue of CF-1 mice.
CF-1 mice were i.v. injected with 5 nmole/kg of INS, ProINS-Tf or 50 µl of PBS and liver
samples were collected 1 h and 8 h post injection. After immunoprecipitation, relative liver IGF-
1R phosphorylation levels were estimated by normalizing p-Tyrosine band density to the
corresponding IGF-1R band density. Data were presented as the average values with error bars
indicating the standard deviations (N=3).
0.0
0.2
0.4
0.6
0.8
1.0
p-IGF-1R / IGF-1R
INS
ProINS-Tf
PBS
1h 8h
69
Figure 4.8. IGF-1R phosphorylation in HepG2 cells.
Serum starved HepG2 cells were treated with 100 nM of INS, ProINS-Tf or trypsin digested
ProINS-Tf for 1 h or 8 h. Equal amounts of total protein from each treatment were applied to p-
IGF-1R ELISA. Relative IGF-1R phosphorylation levels was quantified by measuring
absorbance at 450 nM. Data were presented as the average values with error bars indicating the
standard deviations (N=3).
70
Figure 4.9. 28-Day Fasting BG in NOD mice with different treatment.
NOD mice with symptom of hyperglycemia (BG> 200 mg/dL) were initiated with daily s.c.
injection of 22.5 nmole/kg of ProINS-Tf, INS or 200 µl of PBS for a total of 28 days. Mice were
subjected to 12 h fasting/ 12 h feeding after each injection. BG level was measured by the end of
12-h fasting period every 4th day. Readings above 600mg/dL were presented as 600mg/dL due
to limitation of the detection range. Data were presented as the average values with error bars
indicating the standard deviations (N=3).
0 4 8 12 16 20 24 28
0
100
200
300
400
500
600
700
Day
BG (mg/dL)
ProINS-Tf
INS
PBS
71
Table 4.2. Long-term treatment effect in NOD mice.
All measurements were taken from the mice by the end of 28-day treatment. Concentrations of
cholesterol, triglyceride and HbA1c were measured from whole blood. Data were presented as
the average values with error bars indicating the standard deviations (N=3). *Measurement
range: 100-400 mg/dL.
72
Figure 4.10. Anti-Tf and Anti-ProINS antibody detection in NOD mice serum.
Serum from NOD mice treated with ProINS-Tf or PBS for 28 days were serial diluted (1:20 to
1:20000) and used to incubate with membranes that were dotted with 100ng of (A) human Tf or
(B) ProINS. Secondary anti-mouse antibody was used to detect existence of corresponding
antibody
A
B
73
4.4 Discussion
The effect of ProINS-Tf was evaluated in NOD mice to compare and confirm its
hypoglycemic effects seen in STZ-induced T1D mouse model. The incidence rate of T1D in
female NOD varied greatly in the literature, ranging from 60% to 90% [89, 90]. In order to better
plan the experiments, it was important for us to establish the diabetes incidence rate of the NOD
mice under the current setting of our animal facility. As shown in Table 4.1, over a period of 29
weeks after birth, about 75% of the NOD mice developed T1D and the incidence rate was within
the suggested range from the literature. The majority of mice developed hyperglycemia around
20 weeks after birth (Table 4.1). Subsequent experiments were planned around this period in
order to obtain enough TID mice for different treatment groups.
ProINS-Tf exhibited similar effects on lowering BG levels in fasting T1D NOD mice as
previously seen in STZ-induced T1D mice. Different from INS, which induced the BG level to
rapidly decrease right after the injection but re-bounced 4 h after (Figure 4.1), ProINS-Tf
resulted in a slow decrease of BG to a level below 100 mg/dL and the BG maintained at low
level until 10 h or longer after the injection. The delayed onset of ProINS-Tf corresponded to the
time it took to be converted to the active form and sustained effect was contributed by both the
increased stability and enhanced IR affinity of the active irINS-Tf. Comparing with STZ-induced
T1D mice, NOD mice seemed to be more sensitive to ProINS-Tf treatment. In NOD mice treated
with a high dose of ProINS-Tf (135 nmole/kg), severe hypoglycemia was observed in one
mouse. Such overdose effect has not been observed in STZ-induced T1D mice treated with the
same or even higher concentration of the fusion protein [34, 35]. Another observation worth
noticing was that there were larger variations in T1D NOD mice within the same treatment group
74
than the variations seen in STZ-induced T1D mice. This could possibly be explained by the fact
that each NOD mouse with hyperglycemia might have different severity of the diabetic
syndromes at the time of an experiment and therefore might have responded to the same ProINS-
Tf treatment differently.
The liver selective hypoglycemic effect of ProINS-Tf was also identified in NOD mice
through the fasting/feeding and feeding/fasting experiments. The BG lowering effect of ProINS-
Tf was only significant under fasting condition and became very minimal under feeding state
(Figure 4.2). Since hepatic glucose production is the major BG source during fasting period, but
not feeding period [91, 92], results in Figure 4.2 further supported the previous finding that
ProINS-Tf was a liver selective INS analog for the control of BG level in T1D mice [34].
Although the hypoglycemic effect of ProINS-Tf always diminished and resulted in BG bouncing
back to the same level as in the control group during feeding periods, a single dose of the fusion
protein was still able control the fasting BG in NOD mice over an extended period of time.
ProINS-Tf at a medium dose (45 nmole/kg) was able to lower the BG in NOD mice under
fasting condition for up to 24 h, while mice were subjected to two 2 h-feeding periods in between
(Figure 4.4). A high dose of ProINS-Tf (135 nmole/kg) exhibited capacity of controlling the BG
for a much longer period of time (>40 h) under fasting condition (Figure 4.3). On the other hand,
INS Glargine, which is the currently available long-lasting INS analog on the market, only
managed the BG for less than 16 h during the same experiment. Efficacy of ProINS-Tf shown in
the meal-challenge test confirmed its long-lasting effect and could be applied to adjust the dosing
frequency when optimizing ProINS-Tf toward a more practical long-acting INS analog.
The increase of the average PAI-1 level in mice serum after 5 days ProINS-Tf treatment was
insignificant due to the large within group variation (Figure 4.5). In order to reduce the variation
75
of serum PAI-1 between individual mouse, the study could be repeated in the future with a large
number of mice in each treatment group to improve the significance of PAI-1 measurements. In
HepG2 cells, the IGF-1R phosphorylation resulted from trypsin digested ProINS-Tf treatment,
which was a mimic of the activated ProINS-Tf, was higher than that from INS, while ProINS-Tf
remained less active than INS (Figure 4.8). The higher IGF-1R phosphorylation level seen from
trypsin digested ProINS-Tf treatment might imply the potentially increased affinity of activated
ProINS-Tf to IGF-1R than INS, however, no increased IGF-1R phosphorylation was observed in
the liver of mice treated with ProINS-Tf. In the mice, liver cells initially were exposed to
ProINS-Tf instead of activated ProINS-Tf. Therefore, one possible explanation of the
discrepancy between the in vivo and in vitro observations could be that only a small portion of
ProINS-Tf would be activated over an extended period of time in vivo, leaving significant
amount of the fusion protein was in the inactive form and caused minimal effect on increasing
IGF-1R phosphorylation. It was also possible that, unlike in HepG2 cells, there were many other
factors in the liver, e.g., growth hormone [83], that could regulate the IGF-1R phosphorylation in
liver, making the effect of ProINS-Tf on IGF-1R phosphorylation less significant.
Long-term use of ProINS-Tf was evaluated in NOD mice, with a daily injection of 22.5
nmole/kg protein for a total of 28 days. During the first 12 days of the study, ProINS-Tf
managed to lower BG to a much lower level than PBS, making it look promising on slowing
down the progression of diabetes symptoms in NOD mice (Figure 4.9). However, the significant
BG lowering effects of ProINS-Tf gradually diminished after day 12, and eventually resulted in
no difference from the control group. The cause of the loss of effect was likely due to the
development of anti-human Tf antibody in the ProINS-Tf treated mice (Figure 4.10). Since
ProINS-Tf was constructed with human ProINS and human Tf sequences, the immunogenicity of
76
this fusion protein in mice was somewhat expected. However, only anti-hTf but not anti-ProINS
antibodies were detected from the ProINS-Tf treated mice, suggesting the immune response was
only due to the immunogenicity of human Tf protein but not any new antigenic sites created in
the construction of the fusion protein. Therefore, the formation of anti-hTf antibody in mice
could not be applied to predict the immunogenicity of the fusion protein in human. Other
humanized models could be used to further investigate the antigenicity of the fusion protein in
human immune system. Even with anti-drug antibody formation to compromise the effect of
ProINS-Tf, the fusion protein was still able to show considerable level of beneficial effects in
maintaining body weight, lowering lipids and HbA1c concentrations in T1D NOD mice (Table
4.2). Therefore, more promising results could be expected for the long-term effect of ProINS-Tf
if a more proper animal model could be used to eliminate the chance of anti-drug antibody
formation, e.g. NOD scid gamma (NSG) mice that have extreme immunodeficiency [93, 94].
4.5 Summary
Studies in this chapter demonstrated that ProINS-Tf was effective for the control of BG
in T1D NOD mouse model. The delayed but sustained hypoglycemic effect of ProINS-Tf in
NOD mice was similar to its effect in STZ-induced T1D mice. Possibly due to the different
pathogenic basis of the two mouse models, ProINS-Tf appeared to be more effective in NOD
T1D mice than in STZ-induced T1D mice, with the same concentration causing more severe
hypoglycemic effect in NOD mice. Comparing to INS and ProINS, no increased level of related
side effect was found for ProINS-Tf. Long-term use of ProINS-Tf in NOD mice showed
beneficial effects in managing body weight and decreasing levels of lipids and HbA1c as
compared to non-treated mice. However, ProINS-Tf exhibited gradually diminished efficacy
77
during the long-term study due to development of anti-hTf antibody in the treated mice. Future
studies could be done using a different system to get a more proper evaluation of the long-term
use of ProINS-Tf.
78
CHAPTER 5. SUMMARY
5.1 Conclusions
In this dissertation, we studied the mechanism of ProINS-Tf from multiple aspects to
demonstrate that this fusion protein is a liver-selective INS prodrug with prolonged
hypoglycemic activity in T1D animal model. The liver-targeting effect of ProINS-Tf is resulted
from a combination of multiple factors, including the bulky size of the fusion protein, the TfR-
mediated primary binding and subsequent activation in liver, and the bivalent binding
mechanism of the activated protein to achieve prolonged liver association.
As suggested by several studies, the size of INS or INS analogs could play a role on
influencing the availability of the drug to peripheral tissues and liver [53]. The tight endothelium
barrier at muscle and fat restricted the entrance of large size molecules, whereas the sinusoids in
the liver are lined by fenestrated endothelium, leaving large pores for high molecular weight
proteins to pass through [2]. Therefore, many of currently developed INS analogs with
preferential effects on inhibiting hepatic glucose output are mostly designed to have large
molecular weight or ability to bind to other circulating protein to gain increased size [63, 64, 95].
In the case of ProINS-Tf, the molecular size of ProINS has been largely increased by the fusion
of Tf domain, which is an 80-kDa serum glycoprotein. The liver/periphery gradient effect
applying to other large size INS analogs would also apply to ProINS-Tf. However, this strategy
is mostly dependent on passive targeting and a significant amount of INS macromolecular
analogs will remain in the peripheral compartments at higher doses, which will decrease the
hepatoselectivity [48]. The improved targeting effect to liver not only is associated with the large
size of ProINS-Tf, but also related to TfR mediated binding and uptake. Compared to other INS
analogs, after reaching the live cells, ProINS-Tf could utilize TfR binding as an additional step to
79
enhance its association with the liver tissue since TfR’s are highly expressed on Kupffer cells
and hepatocytes [74]. In addition to increasing the molecular size and providing a primary
binding action, the protein domain of Tf in ProINS-Tf also served as an activation platform for
the conversion of the fusion protein to the active form via the TfR-mediated recycling pathway
[33]. ProINS-Tf originally elicited a very low potency on activating IR as indicated by the Akt
phosphorylation assay (Figure 2.2). Upon the pre-incubation with H4IIE hepatoma cells, the
potency of the fusion protein was significantly increased, suggesting it had been activated into a
much more active form, irINS-Tf.
As proposed and identified in chapter 2, with the ability to bind to both IR and TfR as a
bifunctional fusion protein, the active irINS-Tf possesses enhanced affinity and prolonged
association with IR. The bivalent and high affinity binding of irINS-Tf to IR was consistent with
our previous report that, after washing off unbound protein, ProINS-Tf showed a much longer
Akt dephosphorylation process than INS [35]. Here we also noticed the enhanced effect of
irINS-Tf on inducing Akt phosphorylation when compared to INS (Figure 2.3). Taken together,
these results indicate the potential of the bifunctional irINS-Tf to achieve prolonged association
with the targeted tissue, and thus explain the enhanced accumulation and sustained activity of
ProINS-Tf in the liver.
The improved liver distribution of ProINS-Tf correlated well with its in vivo efficacy on
lowering BG levels in diabetic mice. As demonstrated in the NOD mice experiment, the
hypoglycemic activity of ProINS-Tf was only significant when the mice were being fasted
(Figure 4.2). During the fasting period, the BG level is largely influenced by the hepatic glucose
production, while under feeding condition, glucose removal is mainly dependent on disposal into
peripheral tissues and the effect of reduced hepatic glucose production is not as significant as
80
under fasting condition [91, 92]. The more potent efficacy of ProINS-Tf observed in fasted mice
indicated its preferential anti-diabetic effect on reducing hepatic glucose output.
It has been long recognized that in order for a prodrug to achieve site-specific delivery,
properties of both the active drug and the targeting site are critical. The selectivity can be further
improved when transport, selective enzyme cleavage and retention all work in concert [67]. As
illustrated in Figure 5.1, the liver-specific effect of ProINS-Tf in type 1 diabetes mice was a
combination of TfR-mediated endocytosis in liver, followed by the conversion in TfR recycling
compartments to an active form with a high retention in the liver. With a combination of
selective delivery, activation and retention, ProINS-Tf appears to be an ideal liver-targeted INS
prodrug, which can achieve anti-diabetic effect without the concern of severe hypoglycemia and
other possible side effects due to the INS activity in peripheral tissues. In addition, our finding
also provides a model for the design of fusion protein as prodrug for the selective release of an
active therapeutic protein inside target cells.
81
Figure 5.1. Mechanism of ProINS-Tf as a long-acting and liver-targeted INS prodrug.
ProINS-Tf initially binds to TfR at the surface of liver cells and gets endocytosed. During the
TfR-mediated recycling process, ProINS-Tf is converted to active irINS-Tf and released back to
the cell surface. With bivalent binding to both IR and TfR, irINS-Tf is able to achieve enhanced
binding to IR, resulting in sustained effect on lowering BG in T1D animals.
82
5.2 Future Perspectives
5.2.1 Identification of irINS-Tf Structure and In Vitro Activation of ProINS-Tf
So far we have already confirmed the INS-like activity of irINS-Tf, however, the exact
structure of the irINS moiety remained unclear. It was not sure whether the irINS group in the
activated fusion protein resembles the intact structure of native INS or is a derivative of ProINS
or INS. Before using any analytical method to uncover the sequence of the protein, the primary
challenge is to isolate the active irINS-Tf from the inactive ProINS-Tf. Currently the most
efficient method to obtain irINS-Tf is by incubating the protein with H4IIE cells. After
incubation, a mixture of irINS-Tf and ProINS-Tf will be left in the conditioned medium and the
purification of irINS-Tf will be difficult due to the similar molecular size of the two proteins and
the low conversion efficiency (less than 10%). To obtain enough amount of pure irINS-Tf for
analyzation, the cell conversion system needs to be further optimized in order to get more
amount of the active form.
If the cell-based conversion failed to give enough irINS-Tf for structure analyzation,
another potential strategy is to activate ProINS-Tf in vitro by using enzyme digestion. It has been
done in previous reports that ProINS could be converted to INS-like molecules by using trypsin
and carboxypepdidase B [73, 96]. The enzyme digestion process in vitro is expected to be much
more efficient than the cell-mediated conversion and therefore should be able to generate large
quantity of the activated fusion protein. In early study, trypsin digested ProINS-Tf had been
preliminarily characterized to be a more active form than original ProINS-Tf [34]. However,
there might be structural variations between the cell-activated and enzyme-activated forms. If the
condition of in vitro enzyme digestion could be optimized to generate a form that is as active as
the cell-converted ProINS-Tf, the enzyme digested ProINS-Tf could then be applied for structure
83
identification to provide more insight about the activated fusion protein. In addition, the BG
lowering efficacy of directly using the activated ProINS-Tf can be evaluated if enough quantity
of the active form could be obtained from enzyme digestion. Without requirement for in vivo
activation, the pharmacodynamics profile of irINS-Tf might be different from that of ProINS-Tf.
5.2.2 Exploration of Therapeutic Potentials of ProINS-Tf in T2D Animal Model
Given the superior bivalent binding effect of irINS-Tf, the therapeutic efficacy of the
fusion protein could be further explored for the treatment of T2D. In chapter 2, the preliminary
study in a simplified INS-resistant cell model has shown promising result of irINS-Tf on
overcoming INS resistance to restore IR activation. Therefore, there is a great potential of
ProINS-Tf to be used as an INS analog to restore the attenuated INS response in INS resistance
symptoms and therefore to prevent the progression of the disease conditions.
To add to the Akt phosphorylation assay, glucose production inhibition activity of
ProINS-Tf could be evaluated in INS resistant liver cells to measure the downstream
physiological effect after turning on IR activation. ProINS-Tf was inactive on promoting glucose
uptake in adipocytes due to lack of protein activation mechanism in the periphery [33]. Since
irINS-Tf is active on INS related function, it could be directly applied to test its efficacy on
promoting glucose uptake in both regular and INS resistant cultured adipocytes.
Evaluation of both ProINS-Tf and irINS-Tf could be carried out in a T2D animal model
characterized with INS resistance, e.g. db/db mice, the most widely used T2D mouse model.
db/db mice have been identified to be able to develop obesity, hyperlipidemia, hyperinsulinemia
and INS resistance, which are the typical symptoms of T2D [78]. In case too many complications
are developed in db/db mice with severe INS resistance, causing unpredictable variations when
84
evaluating the efficacy of ProINS-Tf, another T2D model, KK mice, could serve as a backup
mouse model. KK mice exhibit only mild INS resistance and obesity [78], and therefore could be
useful for preliminary in vivo characterization of ProINS-Tf.
If ProINS-Tf or irINS-Tf could demonstrate good therapeutic potential in T2D animal
model to attenuate INS resistance symptoms, the application of this fusion protein could then be
further expended from only treating T1D to treating both T1D and T2D, as well as many other
INS resistance associated diseases, e.g., hypertension, non-alcoholic fatty liver disease (NAFLD)
and non-alcoholic steatohepatitis (NASH). With successful in vitro enzyme activation, irINS-Tf
as a derivative of ProINS-Tf could also be developed towards a new protein drug entity for
overcoming INS resistance in both liver and peripheral tissues.
85
REFERENCES
1. Bissell, D.M., Cell-matrix interaction and hepatic fibrosis. Prog Liver Dis, 1990. 9: p. 143-
55.
2. Braet, F. and E. Wisse, Structural and functional aspects of liver sinusoidal endothelial
cell fenestrae: a review. Comp Hepatol, 2002. 1(1): p. 1.
3. Irie, S. and M. Tavassoli, Transendothelial transport of macromolecules: the concept of
tissue-blood barriers. Cell Biol Rev, 1991. 25(4): p. 317-33, 340-1.
4. Madsbad, S., LY2605541--a preferential hepato-specific insulin analogue. Diabetes,
2014. 63(2): p. 390-2.
5. Shojaee-Moradie, F., et al., Novel hepatoselective insulin analogues: studies with
covalently linked thyroxyl-insulin complexes. Diabet Med, 1998. 15(11): p. 928-36.
6. Kang, J.H., R. Toita, and M. Murata, Liver cell-targeted delivery of therapeutic molecules.
Crit Rev Biotechnol, 2016. 36(1): p. 132-43.
7. Mishra, N., et al., Efficient hepatic delivery of drugs: novel strategies and their
significance. Biomed Res Int, 2013. 2013: p. 382184.
8. Beljaars, L., D.K. Meijer, and K. Poelstra, Targeting hepatic stellate cells for cell-specific
treatment of liver fibrosis. Front Biosci, 2002. 7: p. e214-22.
9. Craparo, E.F., et al., Galactosylated micelles for a ribavirin prodrug targeting to
hepatocytes. Biomacromolecules, 2013. 14(6): p. 1838-49.
10. Craparo, E.F., et al., Galactosylated polymeric carriers for liver targeting of sorafenib. Int
J Pharm, 2014. 466(1-2): p. 172-80.
86
11. Ma, P., et al., Lactose mediated liver-targeting effect observed by ex vivo imaging
technology. Biomaterials, 2010. 31(9): p. 2646-54.
12. Yu, B., et al., Lipid nanoparticles for hepatic delivery of small interfering RNA.
Biomaterials, 2012. 33(25): p. 5924-34.
13. Kullak-Ublick, G.A., Regulation of organic anion and drug transporters of the sinusoidal
membrane. J Hepatol, 1999. 31(3): p. 563-73.
14. Tu, M., et al., Medicinal chemistry design principles for liver targeting through OATP
transporters. Curr Top Med Chem, 2013. 13(7): p. 857-66.
15. Niemi, M., Role of OATP transporters in the disposition of drugs. Pharmacogenomics,
2007. 8(7): p. 787-802.
16. Erion, M.D., Prodrugs for Liver-targeted Drug Delivery. In: Stella V.J., Borchardt R.T.,
Hageman M.J., Oliyai R., Maag H., Tilley J.W. (eds) in Prodrugs. Biotechnology:
Pharmaceutical Aspects, vol V. . 2007, Springer, New York, NY.
17. Lew, W., X. Chen, and C.U. Kim, Discovery and development of GS 4104 (oseltamivir): an
orally active influenza neuraminidase inhibitor. Curr Med Chem, 2000. 7(6): p. 663-72.
18. Wilcox, G., Insulin and insulin resistance. Clin Biochem Rev, 2005. 26(2): p. 19-39.
19. Reaven, G.M., The insulin resistance syndrome: definition and dietary approaches to
treatment. Annu Rev Nutr, 2005. 25: p. 391-406.
20. Chatterjee, S., K. Khunti, and M.J. Davies, Type 2 diabetes. Lancet, 2017. 389(10085): p.
2239-2251.
21. Vijan, S., In the clinic. Type 2 diabetes. Ann Intern Med, 2010. 152(5): p. ITC31-15; quiz
ITC316.
87
22. American Diabetes, A., (7) Approaches to glycemic treatment. Diabetes Care, 2015. 38
Suppl: p. S41-8.
23. Marin-Penalver, J.J., et al., Update on the treatment of type 2 diabetes mellitus. World J
Diabetes, 2016. 7(17): p. 354-95.
24. Olefsky, J., G.M. Reaven, and J.W. Farquhar, Effects of weight reduction on obesity.
Studies of lipid and carbohydrate metabolism in normal and hyperlipoproteinemic
subjects. J Clin Invest, 1974. 53(1): p. 64-76.
25. Lew, K.N. and A. Wick, Pharmacotherapy of Type 2 Diabetes Mellitus: Navigating Current
and New Therapies. Medsurg Nurs, 2015. 24(6): p. 413-9, 438.
26. Bailey, C.J. and R.C. Turner, Metformin. N Engl J Med, 1996. 334(9): p. 574-9.
27. Viollet, B., et al., Cellular and molecular mechanisms of metformin: an overview. Clin Sci
(Lond), 2012. 122(6): p. 253-70.
28. Lebovitz, H.E. and M.A. Banerji, Insulin resistance and its treatment by
thiazolidinediones. Recent Prog Horm Res, 2001. 56: p. 265-94.
29. Nanjan, M.J., et al., Thiazolidinediones as antidiabetic agents: A critical review. Bioorg
Chem, 2018. 77: p. 548-567.
30. Chojkier, M., Troglitazone and liver injury: in search of answers. Hepatology, 2005. 41(2):
p. 237-46.
31. Mahaffey, K.W., et al., Results of a reevaluation of cardiovascular outcomes in the
RECORD trial. Am Heart J, 2013. 166(2): p. 240-249 e1.
32. Meece, J., Basal Insulin Intensification in Patients with Type 2 Diabetes: A Review.
Diabetes Ther, 2018.
88
33. Wang, Y., et al., Receptor-mediated activation of a proinsulin-transferrin fusion protein
in hepatoma cells. J Control Release, 2011. 155(3): p. 386-92.
34. Wang, Y., et al., Proinsulin-transferrin fusion protein as a novel long-acting insulin analog
for the inhibition of hepatic glucose production. Diabetes, 2014. 63(5): p. 1779-88.
35. Shao, J., J.L. Zaro, and W.C. Shen, Proinsulin-Transferrin Fusion Protein Exhibits a
Prolonged and Selective Effect on the Control of Hepatic Glucose Production in an
Experimental Model of Type 1 Diabetes. Mol Pharm, 2016. 13(8): p. 2641-6.
36. Peavy, D.E., et al., In vitro activity of biosynthetic human proinsulin. Receptor binding
and biologic potency of proinsulin and insulin in isolated rat adipocytes. Diabetes, 1984.
33(11): p. 1062-7.
37. Osborn, B.L., et al., Pharmacokinetic and pharmacodynamic studies of a human serum
albumin-interferon-alpha fusion protein in cynomolgus monkeys. J Pharmacol Exp Ther,
2002. 303(2): p. 540-8.
38. Duttaroy, A., et al., Development of a long-acting insulin analog using albumin fusion
technology. Diabetes, 2005. 54(1): p. 251-8.
39. Muller, N., et al., Superior serum half life of albumin tagged TNF ligands. Biochem
Biophys Res Commun, 2010. 396(4): p. 793-9.
40. Chen, X., et al., Effects of receptor binding on plasma half-life of bifunctional transferrin
fusion proteins. Mol Pharm, 2011. 8(2): p. 457-65.
41. Reaven, G.M., Banting lecture 1988. Role of insulin resistance in human disease.
Diabetes, 1988. 37(12): p. 1595-607.
89
42. Kahn, C.R., Insulin resistance, insulin insensitivity, and insulin unresponsiveness: a
necessary distinction. Metabolism, 1978. 27(12 Suppl 2): p. 1893-902.
43. Shulman, G.I., Cellular mechanisms of insulin resistance. J Clin Invest, 2000. 106(2): p.
171-6.
44. Lee, J.Y., H.K. Cho, and Y.H. Kwon, Palmitate induces insulin resistance without
significant intracellular triglyceride accumulation in HepG2 cells. Metabolism, 2010.
59(7): p. 927-34.
45. Gao, D., et al., The effects of palmitate on hepatic insulin resistance are mediated by
NADPH Oxidase 3-derived reactive oxygen species through JNK and p38MAPK pathways.
J Biol Chem, 2010. 285(39): p. 29965-73.
46. Tanaka, T., Y. Fujishima, and Y. Kaneo, Receptor mediated endocytosis and cytotoxicity
of transferrin-mitomycin C conjugate in the HepG2 cell and primary cultured rat
hepatocyte. Biol Pharm Bull, 2001. 24(3): p. 268-73.
47. Wilkening, S., F. Stahl, and A. Bader, Comparison of primary human hepatocytes and
hepatoma cell line Hepg2 with regard to their biotransformation properties. Drug Metab
Dispos, 2003. 31(8): p. 1035-42.
48. Shao, J., J.L. Zaro, and W.C. Shen, Tissue barriers and novel approaches to achieve
hepatoselectivity of subcutaneously-injected insulin therapeutics. Tissue Barriers, 2016.
4(2): p. e1156804.
49. Ciechanover, A., A.L. Schwartz, and H.F. Lodish, The asialoglycoprotein receptor
internalizes and recycles independently of the transferrin and insulin receptors. Cell,
1983. 32(1): p. 267-75.
90
50. Horwitz, D.L., et al., Proinsulin, insulin, and C-peptide concentrations in human portal
and peripheral blood. J Clin Invest, 1975. 55(6): p. 1278-83.
51. Chap, Z., et al., First-pass hepatic extraction and metabolic effects of insulin and insulin
analogues. Am J Physiol, 1987. 252(2 Pt 1): p. E209-17.
52. Meier, J.J., J.D. Veldhuis, and P.C. Butler, Pulsatile insulin secretion dictates systemic
insulin delivery by regulating hepatic insulin extraction in humans. Diabetes, 2005. 54(6):
p. 1649-56.
53. Herring, R., R.H. Jones, and D.L. Russell-Jones, Hepatoselectivity and the evolution of
insulin. Diabetes Obes Metab, 2014. 16(1): p. 1-8.
54. Kruszynska, Y.T., P.D. Home, and K.G. Alberti, Comparison of portal and peripheral
insulin delivery on carbohydrate metabolism in streptozotocin-diabetic rats.
Diabetologia, 1985. 28(3): p. 167-71.
55. Reaven, G.M., Insulin resistance and compensatory hyperinsulinemia: role in
hypertension, dyslipidemia, and coronary heart disease. Am Heart J, 1991. 121(4 Pt 2): p.
1283-8.
56. Hall, J.E., et al., Insulin resistance, hyperinsulinemia, and hypertension: causes,
consequences, or merely correlations? Proc Soc Exp Biol Med, 1995. 208(4): p. 317-29.
57. Sonksen, P.H., D. Russell-Jones, and R.H. Jones, Growth hormone and diabetes mellitus.
A review of sixty-three years of medical research and a glimpse into the future? Horm
Res, 1993. 40(1-3): p. 68-79.
58. Abdul-Wahed, A., et al., A link between hepatic glucose production and peripheral
energy metabolism via hepatokines. Mol Metab, 2014. 3(5): p. 531-43.
91
59. Ruotolo, G., et al., Effects of intraperitoneal versus subcutaneous insulin administration
on lipoprotein metabolism in type I diabetes. Metabolism, 1990. 39(6): p. 598-604.
60. Saudek, C.D., et al., Implantable insulin pump vs multiple-dose insulin for non-insulin-
dependent diabetes mellitus: a randomized clinical trial. Department of Veterans Affairs
Implantable Insulin Pump Study Group. JAMA, 1996. 276(16): p. 1322-7.
61. Stratta, R.J., et al., Pancreas transplantation. Int J Pancreatol, 1995. 17(1): p. 1-13.
62. Iyer, H., A. Khedkar, and M. Verma, Oral insulin - a review of current status. Diabetes
Obes Metab, 2010. 12(3): p. 179-85.
63. Shojaee-Moradie, F., et al., Novel hepatoselective insulin analog: studies with a
covalently linked thyroxyl-insulin complex in humans. Diabetes Care, 2000. 23(8): p.
1124-9.
64. Edgerton, D.S., et al., Changes in glucose and fat metabolism in response to the
administration of a hepato-preferential insulin analog. Diabetes, 2014. 63(11): p. 3946-
54.
65. Widera, A., F. Norouziyan, and W.C. Shen, Mechanisms of TfR-mediated transcytosis and
sorting in epithelial cells and applications toward drug delivery. Adv Drug Deliv Rev,
2003. 55(11): p. 1439-66.
66. Li, H. and Z.M. Qian, Transferrin/transferrin receptor-mediated drug delivery. Med Res
Rev, 2002. 22(3): p. 225-50.
67. Stella, V.J. and K.J. Himmelstein, Prodrugs and site-specific drug delivery. J Med Chem,
1980. 23(12): p. 1275-82.
92
68. Sonoda, S. and M. Schlamowitz, Studies of 125I trace labeling of immunoglobulin G by
chloramine-T. Immunochemistry, 1970. 7(11): p. 885-98.
69. Vugmeyster, Y., et al., Biodistribution of [125I]-labeled therapeutic proteins: application
in protein drug development beyond oncology. J Pharm Sci, 2010. 99(2): p. 1028-45.
70. Chen, J., et al., Comparison of succinimidyl [(125)I]iodobenzoate with iodogen iodination
methods to study pharmacokinetics and ADME of biotherapeutics. Pharm Res, 2014.
31(10): p. 2810-21.
71. Dimitriadis, G., et al., Insulin effects in muscle and adipose tissue. Diabetes Res Clin
Pract, 2011. 93 Suppl 1: p. S52-9.
72. Melani, F., A.H. Rubenstein, and D.F. Steiner, Human serum proinsulin. J Clin Invest,
1970. 49(3): p. 497-507.
73. Kemmler, W., J.D. Peterson, and D.F. Steiner, Studies on the conversion of proinsulin to
insulin. I. Conversion in vitro with trypsin and carboxypeptidase B. J Biol Chem, 1971.
246(22): p. 6786-91.
74. Gatter, K.C., et al., Transferrin receptors in human tissues: their distribution and possible
clinical relevance. J Clin Pathol, 1983. 36(5): p. 539-45.
75. Ponka, P. and C.N. Lok, The transferrin receptor: role in health and disease. Int J Biochem
Cell Biol, 1999. 31(10): p. 1111-37.
76. Qian, Z.M., et al., Targeted drug delivery via the transferrin receptor-mediated
endocytosis pathway. Pharmacol Rev, 2002. 54(4): p. 561-87.
77. Kolb, H., Mouse models of insulin dependent diabetes: low-dose streptozocin-induced
diabetes and nonobese diabetic (NOD) mice. Diabetes Metab Rev, 1987. 3(3): p. 751-78.
93
78. Kitada, M., Y. Ogura, and D. Koya, Rodent models of diabetic nephropathy: their utility
and limitations. Int J Nephrol Renovasc Dis, 2016. 9: p. 279-290.
79. Deeds, M.C., et al., Single dose streptozotocin-induced diabetes: considerations for study
design in islet transplantation models. Lab Anim, 2011. 45(3): p. 131-40.
80. Driver, J.P., D.V. Serreze, and Y.G. Chen, Mouse models for the study of autoimmune
type 1 diabetes: a NOD to similarities and differences to human disease. Semin
Immunopathol, 2011. 33(1): p. 67-87.
81. Thayer, T.C., S.B. Wilson, and C.E. Mathews, Use of nonobese diabetic mice to
understand human type 1 diabetes. Endocrinol Metab Clin North Am, 2010. 39(3): p.
541-61.
82. Baricevic, I., et al., A framework for the in vitro evaluation of cancer-relevant molecular
characteristics and mitogenic potency of insulin analogues. Carcinogenesis, 2015. 36(9):
p. 1040-50.
83. Westley, R.L. and F.E. May, A twenty-first century cancer epidemic caused by obesity: the
involvement of insulin, diabetes, and insulin-like growth factors. Int J Endocrinol, 2013.
2013: p. 632461.
84. Vigneri, R., S. Squatrito, and L. Sciacca, Insulin and its analogs: actions via insulin and IGF
receptors. Acta Diabetol, 2010. 47(4): p. 271-8.
85. Bryhni, B., E. Arnesen, and T.G. Jenssen, Associations of age with serum insulin,
proinsulin and the proinsulin-to-insulin ratio: a cross-sectional study. BMC Endocr Disord,
2010. 10: p. 21.
94
86. Wiberg, B., et al., Insulin sensitivity measured by the euglycaemic insulin clamp and
proinsulin levels as predictors of stroke in elderly men. Diabetologia, 2009. 52(1): p. 90-6.
87. Ginsberg, H.N., Insulin resistance and cardiovascular disease. J Clin Invest, 2000. 106(4):
p. 453-8.
88. Schneider, D.J. and B.E. Sobel, PAI-1 and diabetes: a journey from the bench to the
bedside. Diabetes Care, 2012. 35(10): p. 1961-7.
89. Delovitch, T.L. and B. Singh, The nonobese diabetic mouse as a model of autoimmune
diabetes: immune dysregulation gets the NOD. Immunity, 1997. 7(6): p. 727-38.
90. Leiter, E.H., M. Prochazka, and D.L. Coleman, The non-obese diabetic (NOD) mouse. Am J
Pathol, 1987. 128(2): p. 380-3.
91. Aronoff S.L., B.K., Sheriener B., Want L., Glucose Metabolism and Regulation: Beyond
Insulin and Glucagon. Diabetes Spectrum, 2004. 17(3): p. 183-191.
92. Samuel, V.T. and G.I. Shulman, Mechanisms for insulin resistance: common threads and
missing links. Cell, 2012. 148(5): p. 852-71.
93. Shultz, L.D., et al., Human lymphoid and myeloid cell development in NOD/LtSz-scid IL2R
gamma null mice engrafted with mobilized human hemopoietic stem cells. J Immunol,
2005. 174(10): p. 6477-89.
94. Coughlan, A.M., et al., Myeloid Engraftment in Humanized Mice: Impact of Granulocyte-
Colony Stimulating Factor Treatment and Transgenic Mouse Strain. Stem Cells Dev,
2016. 25(7): p. 530-41.
95
95. Hordern, S.V., et al., Comparison of the effects on glucose and lipid metabolism of
equipotent doses of insulin detemir and NPH insulin with a 16-h euglycaemic clamp.
Diabetologia, 2005. 48(3): p. 420-6.
96. Chance, R.E., R.M. Ellis, and W.W. Bromer, Porcine proinsulin: characterization and
amino acid sequence. Science, 1968. 161(3837): p. 165-7.
Abstract (if available)
Abstract
A recombinant proinsulin-transferrin fusion protein (ProINS-Tf) has been previously reported to be a novel long-lasting insulin analog, acting specifically on the inhibition of hepatic glucose output. In this dissertation, ProINS-Tf was investigated to elucidate the mechanism responsible for its sustained and liver targeted anti-diabetic effects. Studies were carried out to determine its binding properties, in vivo biodistribution, as well as the in vivo hypoglycemic effects in Type 1 Diabetes mice. The biodistribution study revealed that ProINS-Tf exhibited liver specific accumulation after a single intravenous injection, whereas transferrin (Tf) or insulin (INS) showed relatively even distribution among different organs. The conversion of inactive ProINS-Tf into an active immune-reactive INS-Tf form (irINS-Tf) via a Tf receptor (TfR) mediated process only occurred in the liver, but not in other organs. In addition, ProINS Tf demonstrated a prolonged retention in the liver after an intravenous injection, suggesting the enhanced association of the bifunctional active form, irINS-Tf, within liver cells. ProINS-Tf after being activated possessed enhanced binding affinity to INS receptor (IR), as suggested by the high Akt phosphorylation level and low half maximal inhibitory concentration (IC50) measured from irINS-Tf treated hepatoma cells. The increased binding affinity of irINS-Tf to IR was contributed by the bivalent binding to both IR and TfR on liver cells, leading to sustained and hepatoselective hypoglycemic effects in Type 1 Diabetes mice. Taken together, studies from this dissertation project indicate that ProINS-Tf is a highly liver-targeted INS prodrug with a combination of 3 specific actions in liver cells: (1) TfR mediated binding and uptake of the prodrug on the cell surface, (2) liver-specific, TfR-mediated conversion of the prodrug into its active form, and (3) the bifunctional binding of the active fusion protein to both Tf and INS receptors in the liver to achieve prolonged retention and thus enhanced anti-diabetic activities.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Proinsulin-transferrin recombinant fusion protein: mechanism of activation and potential application in diabetes treatment
PDF
Characterization of rice-derived proinsulin-transferrin fusion protein and its oral bioactivity as a basal insulin analogue
PDF
Reconstruction of a proinsulin-transferrin fusion protein
PDF
Proinsulin-transferrin fusion protein as a liver targeting insulin therapeutic in Type 1 diabetes treatment
PDF
Characterization of the transferrin oligomer and its potential application in drug delivery
PDF
Preparation and characterization of Tf-G-CSF fusion protein
PDF
Human growth hormone-transferrin recombinant fusion protein for oral delivery
PDF
Insulin sensitivity in cognition, Alzheimer's disease and brain aging
PDF
The intracellular trafficking of transferrin in Caco-2 cells and its implication to transferrin-mediated oral drug delivery
PDF
Developing recombinant single chain Fc-dimer fusion proteins for improved protein drug delivery
PDF
Cell penetrating peptide-based drug delivery system for targeting mildly acidic pH
PDF
Integrin-mediated targeting of protein polymer nanoparticles carrying a cytostatic macrolide
PDF
Characterization of actin based motility in mammalian cells through LIM and SH3 domain protein 1 (LASP1) and elastin like polypeptide (ELP) fusion protein
PDF
Characterization of the role of LIM and SH3 protein (LASP-1) in endocytosis and actin cytoskeleton remodeling
PDF
Factors that may impact drug disposition and metabolism
PDF
Harnessing environmental and culture conditions to alter fungal ‘omics’
PDF
Site-specific PEGylation of recombinant immunotoxin αFAP-PE38 for half-life extension
PDF
Development of dihydromyricetin (DHM) as a novel therapy for alcoholic liver disease (ALD) and alcohol use disorder (AUD)
PDF
Stable cell lines expressing elastin-like polypeptide fusions with epidermal growth factor receptor modulate gene expression in a heat dependent manner
PDF
Role of purinergic P2X7 receptors in inflammatory responses in the brain and liver: a study using a mouse model of chronic ethanol and high-fat diet exposure
Asset Metadata
Creator
Liu, Yuqian
(author)
Core Title
Mechanism study of proinsulin-transferrin fusion protein as a long-acting and liver-targeting insulin prodrug
School
School of Pharmacy
Degree
Doctor of Philosophy
Degree Program
Pharmaceutical Sciences
Publication Date
07/12/2020
Defense Date
05/04/2018
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
Diabetes,fusion protein,insulin analog,insulin prodrug,liver-targeting,OAI-PMH Harvest,transferrin
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Shen, Wei-Chiang (
committee chair
), Okamoto, Curtis (
committee member
), Stiles, Bangyan (
committee member
)
Creator Email
liuyuqian527@gmail.com,yuqianl@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-15575
Unique identifier
UC11671728
Identifier
etd-LiuYuqian-6393.pdf (filename),usctheses-c89-15575 (legacy record id)
Legacy Identifier
etd-LiuYuqian-6393.pdf
Dmrecord
15575
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Liu, Yuqian
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
Tags
fusion protein
insulin analog
insulin prodrug
liver-targeting
transferrin