PD-L1-GM-CSF fusion protein-loaded DC vaccination activates PDL1-specific humoral and cellular immune responses

PDF

Download

Open document

Flip pages

Summarize

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
PD-L1-GM-CSF FUSION PROTEIN-LOADED
DC V ACCINATION ACTIV ATES PDL1-SPECIFIC
HUMORAL AND CELLULAR
IMMUNE RESPONSES

Tiffany Chunmei Jehng

University of Southern California
Molecular Microbiology and Immunology

A thesis submitted for the degree of Master of Science (M.S.)

August 2017

2
DEDICATION

I dedicate this dissertation to my parents, Jihmirn Jehng and Connie Hsu, who have
always been there for me during difficult times throughout my life. Thank you for your
strong and constant support throughout my educational journey. I would not be who I am
today without the love and efforts they have put forth for me all the time. Because of my
parents, I could easily make up my mind to conquer those obstacles in front me and
become a more persistent and strong-minded person.

3
ACKNOWLEDGEMENTS

First and foremost, I would like to thank my advisor, Dr. Si-Yi Chen. He has guided me
into a brand new world of cancer immunotherapy that I have enjoyed and learned a lot.
He has constantly brought many novel ideas and the big picture as well as trained me to
gain independent-thinking ability and good reading habits. I greatly value the time and
efforts he has devoted over the past two years and really thank him for the opportunities
he had given me.
I would like to thank my dissertation committee members, Dr. Joseph Landolph and Dr.
Hayoun Lee, for their help, guidance and support. Their knowledge and expertise are
valuable and greatly enrich my work and research study. I truly appreciate their time
throughout my master’s courses and the defense process.
I would like to express gratitude towards all the members of the Chen Laboratory: Drs.
Xue-Fen Huang, Hui Liu, Guan Wang, Shruti Lal, Xi Kang and Ms. Lindsey Jones. They
all helped me with the experimental designs,techniques and identifying solutions to the
problems encountered, to improve the experiments for my thesis project.
I would like to thank Dr. Alan L. Epstein for his kind gift of the GL-261 cancer cell line
and all the members of Dr. W. Martin Kast’s Lab for the technical assistance and the
sharing of some experimental materials and equipment.
I would also like to thank my dear friends who have supported on this journey. Especially
my lovely roommates, Luping Chen, Xinyan Liang and He Zhao, who are the master
students in the Department of Biochemistry and Molecular Medicine or Department of
Molecular Microbiology and Immunology. I thank all of them for the constant support,
love and care I received. I will remember this precious camaraderie, which was essential
for my growth as a person and as a scientist as well in the mental aspects.
Lastly, I would like to thank my parents, my elder brother, Jonathan, and my family for
their unconditional love and support. The sense of gratitude I feel toward them is beyond
description. Thank you, and I love you all.

4
TABLE OF CONTENTS

LIST OF FIGURES …………………………………………………………………..... 5

CHAPTER 1: INTRODUCTION

1.1 Immune checkpoints (PD-1/PD-L1)…………………………………………………. 6
1.2 Clinical efficacy of anti-PDL1 antibody immunotherapy…………………………… 8
1.3 The role of dendritic cells in activating T cell responses…………………………… 11
1.4 The role of dendritic cells in activating B cell responses…………………………... 13
1.5 Tumor-associated antigens: up-regulation of PD-L1 on tumor cells….……………. 15
1.6 Breaking in immune tolerance by linking tumor-associated antigens with an immune
stimulus (GM-CSF)………………………………………………………………… 20
1.7 DC-based tumor vaccines…………………………………………………………... 23

CHAPTER 2: PDL1-GMCSF FUSION PROTEIN-LOADED DC V ACCINATION
ACTIVATES PDL1-SPECIFIC HUMORAL AND CELLULAR
IMMUNE RESPONSES
2.1 Abstract…………………..…………………..…………………..…………………. 26
2.2 Introduction…………………..…………………..…………………..……………... 28
2.3 Materials and Methods………………………………..…………………………….. 32
2.4 Results
2.4.1 Production of a recombinant PD-L1-GM-CSF fusion protein……….…….… 40
2.4.2 Maturation of the PD-L1-GM-CSF fusion protein-loaded DCs…..………….. 43
2.4.3 PD-L1-GM-CSF-loaded DCs enhance cytokine production…………………. 45
2.4.4 Inhibition of tumor growth by PD-L1-GM-CSF-loaded DC vaccines….……. 47
2.4.5 Induction of humoral and cellular immune responses by PD-L1-GM-CSF-
loaded DC vaccines…………………………………………………………… 49
2.5 Discussion……………………………………………………………………...…… 51

CHAPTER 3: SUMMARY AND FUTURE DIRECTIONS
3.1 Summary……………………………………………………………………………. 57
3.2 Future Directions…………………………………………………………………… 58

REFERENCES……………………………………………………………………….. 60

5
LIST OF FIGURES

1.1.1 The regulation of PD-1/PD-L1 immune checkpoint in an immune response….... 7
1.4.1 Function of dendritic cells in the immune response to viruses….……………... 14
2.4.1 Production of a pET-21a
+
/PD-L1-GM-CSF protein vaccine………………….. 42

2.4.2 Morphology of murine BMDCs at different days of culture and costimulatory
molecule (CD40, CD80 and CD86) expression of murine mature-BMDCs.. 44

2.4.3 PD-L1-GM-CSF fusion protein-loaded DC vaccine generates increased
IFN-γ and IL-2 secretion from activated CD4
+
and CD8
+
T cells………….. 46

2.4.4 Vaccination with PD-L1-GM-CSF fusion protein-loaded DCs can efficiently
inhibit tumor growth in mice………………………...……………………... 48

2.4.5 Fusion protein-loaded DC Vaccination increases anti-PD-L1 antibody
responses and generates significantly enhanced cytotoxic T cell responses.. 50

† Abbreviations:
APCs, Antigen-presenting cells; CTL, Cytotoxic T lymphocyte; DC, Dendritic cell;
DMEM, Dullbecco’s modified eagle’s medium; FACS, Fluorescence-activated cell
sorting; GM-CSF, Granulocyte-macrophage colony-stimulating factor; IFN, Interferon;
IL, Interleukin; LPS, Lipopolysaccharide; mAb, Monoclonal antibody; MHC, Major
histocompatibility complex; PD-L1, Programmed death ligand 1; TAAs, Tumor-
associated antigens; Th1, T helper type 1; TNF, Tumor necrosis factor

6
CHAPTER 1: INTRODUCTION

1.1 Immune Checkpoints (PD-1/PD-L1)
Our immune systems have the abilities to differentiate between normal cells and
“abnormal or foreign” cells in our bodies. This will lead to the attack of the foreign cells
by the immune system without damaging the normal cells. In order to accomplish this,
the immune system uses “checkpoints” to maintain the balance between immune
activation and suppression. Immune checkpoints refer to the inhibitory pathways in the
immune system that are critical for regulating the duration and amplitude of physiological
immune responses in human body tissues. The immune checkpoints maintain self-
tolerance and minimize collateral body tissue damage (Drew M. Pardoll. 2012).
However, it is now known that tumors co-operate with particular immune-checkpoint
pathways to gain resistance and then to avoid being attacked by the immune system,
especially against T cells that are specific for some tumor antigens. In conclusion, the
“blockade of immune checkpoints” is obviously the most promising approach to activate
therapeutic anti-tumor immunity. Therefore, it was found that activated antibodies that
directly target those immunogenic regulators (checkpoints), could possibly hold a huge
promise in cancer treatments. In our studies, we focus on a potentially novel targeted
therapy, which is the interaction between the programmed death receptor-1 (PD-1) and its
ligand, PD-L1. PD-1 and PD-L1 are recognized as potent targets to activate and enhance
tumor antigen-specific cytotoxic T-cell function.

7
PD-L1 is a major inhibitory immune pathway exploited by cancer. Under normal
circumstances, PD-L1 plays a crucial role in maintaining immune homeostasis. PD-L1
binds to specific receptors on the surface of T-cells. As binding to their receptors,
cytotoxic T-cell activity is down-regulated, thus, protecting normal cells from collateral
tissue damage. However, in many cancers, PD-L1 is over-expressed on the surface of
tumors and tumor-infiltrating immune cells, such as dendritic cells and macrophages. It
binds to PD-1 on cytotoxic T-cells, suppressing the anti-tumor immune response.
Therefore, PD-L1 is known as a potential biomarker in cancer immunotherapy.

Figure 1.1.1 The regulation of PD-1/PD-L1 immune checkpoint in an immune response. The
cytotoxic T-lymphocyte-associated antigen 4 (CTLA-4)-mediated immune checkpoint is induced
in T-cells as their initial response to an antigen. In contrary to the CTLA-4, instead of regulating
at the initial stage of T-cell activation, PD-1 plays its main role in modulating inflammatory
responses in body tissues by effector T-cells through the recognition of antigen in peripheral
tissues. Activated T-cells are continuously over-expressing PD-1 in the tissues. Inflammatory
cytokine signals produced by infiltrating T-cells in the tissues induce the expression of PD-1
ligands (PD-L1 and PD-L2). These ligands down-regulate the effect of T-cells and further avoid
collateral tissue damage in response to antigen in the tissue. Among those secreted inflammatory
cytokines, the best representative signal for the induction of PD-L1 is interferon-γ (IFN-γ)
primarily produced by T helper 1 (TH1) cells. Figure is taken and adapted from (Pardoll. 2012)

8
1.2 Clinical efficacy of anti-PDL1 antibody immunotherapy
The development of human cancer is a multistep process containing several genetic
and epigenetic alterations that drive tumor progression (Topalian et al. 2012). These
alterations tell cancer cells apart from their counterparts, which are the normal cells,
allowing tumors to be recognized as “foreign” by the human immune system through
their generated neo-antigens. Nevertheless, tumors are rarely eliminated spontaneously,
reflecting their abilities to uphold an immunosuppressive microenvironment in the
immune system. PD-L1 (programmed death-ligand 1), which is also called B7-H1 or
CD274, is expressed in many cancers and tumor-infiltrating immune cells. It is the
primary PD-1 ligand up-regulated in numerous solid tumors, where it can suppress
cytokine secretion and cytotoxic activity of PD-1-positive CD4+ and CD8+ T-cells. PD-
L1 leads a significant role in blocking the “immune defense” by binding its receptor, PD-
1 and B7.1 (CD80), both of which are negative immunogenic regulators of T-cell
activation (Herbst et al. 2014). Tumors are found to develop multiple resistance
mechanisms to evade the endogenous immune responses in preclinical models and
patients. For example, binding of PD-L1 to its receptors inhibits T-cell proliferation,
migration and release of cytotoxic mediators such as perforin and granzymes, which
further suppress the killing of tumor cells (Santarpia et al. 2015). These characteristics
make PD-L1 a promising target for cancer immunotherapy. Thus, the development of
immunotherapeutic approaches for cancer by blocking the interaction between PD-1 and
PD-L1 could probably enhance T-cell responses and anti-tumor immunity.
To evaluate the efficacy of anti-PD-L1 antibody blockade of PD-1/ PD-L1 immune
checkpoint toward enhancing anti-tumor activity, several preclinical and clinical studies

9
have been conducted. A study from Brahmer et al in 2012 was designed for administering
intravenous anti-PD-L1 antibody at escalating dosages to patients with selected advanced
cancers. A high affinity, human PD-L1-specific monoclonal antibody, Bristol-Myers
Squibb’s BMS-936559 (MDX-1105) displayed the therapeutic efficacy in a phase I
clinical trial (ClinicalTrials.gov number, NCT00729664). BMS-936559 inhibited the
binding of PD-L1 to both PD-1 and CD80. This antibody-mediated blockade of PD-L1
elicited durable tumor regression and prolonged stabilization of disease in patients with
selected advanced cancers (Brahmer et al. 2012).
There was another study designed to evaluate the safety, clinical activity and
biomarker status of PD-L1 inhibition through an application of the engineered humanized
antibody MPDL3280A (Genentech/Roche). It is an anti-PD-L1 IgG1 monoclonal
antibody that is able to block PD-L1 interactions with both PD-1 and B7-1 receptors. The
study of metastatic urothelial bladder cancer (UBC) indicated that MPDL3280A has
noticeable clinical activity in controlling tumor growth. The objective response rates
(ORRs) among patients were 43% for those with PD-L1-postitive tumors and 11% for
those with PD-L1-negative tumors (Herbst et al. 2014). In an extended phase I trial
across several cancer types (non-small cell lung cancer, melanoma and other types of
tumors), patients with tumors expressing high levels of PD-L1, especially when PD-L1
was expressed by tumor-infiltrating lymphocytes (TILs), were observed for positive
responses to MPDL3280A. The data in this study imply that MPDL3280A is most
efficient in patients where pre-existing immunity has been suppressed by PD-L1, and is
re-invigorated through antibody-mediated blockade therapy. This anti-PDL1-antibody
drug (MPDL3280A) makes a breakthrough in cancer immunotherapy. Furthermore, in

10
June 2014, it gained a designation status by the FDA for the treatment of metastatic
urothelial bladder cancer (Chen and Han. 2015).
In recent years, multiple anti-PD-L1 monoclonal antibodies are under investigation
and are in clinical trials. To date, antibody treatment has generated outstanding clinical
benefits by inducing regression of advanced and metastatic solid tumors and increasing
survival rates as well. Furthermore, therapeutic antibodies which block the PD-1/PD-L1
axis induce durable clinical responses, tolerable toxicity, minimum side effects, and
applicable to a broad range of cancer types, particularly against a growing list of solid
tumors. In conclusion, anti-PD-L1 monoclonal antibody-therapy has been shown to be
applicable and efficacious in advanced human cancers compared to other conventional
therapies.

11
1.3 The role of dendritic cells in activating T cell responses
The mechanisms employed by professional antigen-presenting cells (APCs) to sense
microorganisms and then initiate immune responses have been well discovered, greatly
on account of the significant role of dendritic cells (DCs), which is a kind of APC, in the
initiation and regulation of immune responses (Kabler et al. 2006). Dendritic cells are
well known to function as a “bridge” between the innate and the adaptive immune
systems. The APC involved in activating T cells is usually a dendritic cell. Instead of
recognizing and responding to a free or soluble antigen, T cells can only recognize and
respond to an antigen that has been processed and further presented by cells via MHC
molecules. Only certain “professional” APCs like “dendritic cells” can exploit the class I
as well as the class II antigen presenting pathways. The activation of T cells induced by
dendritic cells requires three essential signals as follows, (1) the TCR/MHC recognition
(2) the interaction between costimulatory molecules (CD28/ B7 ligands) (3) cytokine
production.
Helper T cells (CD4+ T cells) can recognize exogenous antigen presented on MHC
class II molecule while cytotoxic T cells (CD8+ T cells) can recognize endogenous
antigen presented on MHC class I molecule. The peptide/MHC complex is recognized by
a specific T cell receptor (TCR), leading to an interaction of TCR/MHC complex which
is referred to as “signal 1”. DCs establish interaction with T cells by forming an
immunological synapse, where the TCR/MHC interaction and costimulatory molecules
congregate in a center surrounded by adhesion molecules (Dustin and Cooper. 2000).
Thus, “signal 2” is the costimulatory interaction between CD28 and CD80/CD86 (B7
ligands). There are several accessory receptors on T cells that modulate responses toward

12
signal 1 and 2, the activation and proliferation of T cells, and the cytokine production
which is referred to as “signal 3” (McMaster et al. 2015).
An effective HIV vaccination approach based on suppression of antigen presentation
attenuators in dendritic cells has been confirmed. A previous study demonstrated that
through the inhibition of suppressor of cytokine signaling 1 (SOCS1), the ability of DCs
was restored to elicit anti-HIV-1 immunity. The investigators found that SOCS1-silenced
DCs could elicit enhanced antibody responses, increased HIV-1 envelope-specific CD8
+

cytotoxic T lymphocytes and CD4
+
helper T lymphocytes in mice. Thus, in this study,
SOCS1 functions as an antigen presentation attenuator to harness both HIV-1-specific
humoral and cellular immune responses. Hence, this study has provided evidence that
functionally restored DCs are able to induce specific T-cell and antibody responses (Song
et al. 2006). There have been studies focusing on the effective DC vaccination in HIV
infection. Both mucosal and systemic immune responses are required for the prevention
of HIV transmission and chronic infection. In another study, A20, a negative regulator
that controls the maturation, cytokine production, and immunostimulatory potency of
DCs was silenced to activate DCs. The hyper-activated DCs displayed to elicit both
robust mucosal and systemic HIV-specific humoral and cellular immunity (Hong et al.
2011).

13
1.4 The role of dendritic cells in activating B cell responses
Dendritic cells (DCs) play a critical role not only in activating T cell responses, but
also in B-cell functions. Although B and T lymphocytes are well-known mediators of
immune systems, their functions are tightly regulated by dendritic cells. DCs can capture,
retain, and transport unprocessed intact antigen (Ag). Additionally, DCs can transfer
preserved Ag to naïve B cells to initiate antigen-specific antibody responses (Wykes and
Macpherson. 2000).
B-lymphocytes are both antigen-receiving and antigen-presenting cells that present
antigens transferred from DCs to helper T cells (T
H
cells) through the interaction and
binding of TCR-MHC-II/peptide. B cells bind antigens with their receptors (BCRs), thus,
as an activated T cell senses the peptide presented by the B cell on its surface, the ligand
on the T cell binds to the receptor on the B cell that the specific antigen is bound. For
instance, the CD40L or the CD28 on the helper T cell will individually bind to the CD40
or the B7.1/B7.2 receptor on the B cell, resulting in the activation of resting B cell. The T
cell also produces cytokines such as IL-2, which directly affect and stimulate B cells
through the binding of IL-2 to the ILR on a B cell’s surface.
On account of these signals, the B cell will be highly activated and undergo division,
antibody isotype switching, and further differentiation into plasma B cells or Ag-specific
memory B cells. Eventually, the end-result will be a B cell that is able to massively
produce specific antibodies against a particular antigenic-target. In conclusions, DCs
function to recruit and activate CD4
+
T cells via those three signals mentioned above,
then the activated CD4
+
T cells interact with B cells which results in the proliferation and

14
differentiation of antigen-specific B cells. Thus, DCs are able to provide B cells with
isotype-switch signals independent of T cells while that the assistance of T cells is
essential for antibody production.
A previous study demonstrated that the protective efficacy of broadly neutralizing
antibodies targeted to the HIV-1 envelope glycoprotein in vivo, which is beneficial to the
prevention of HIV-1 transmission. HIV-1 infection, is dependent on effective viral entry
regulated by the interaction between its envelope glycoprotein and specific cell surface
receptors such as CD4. Therefore, the investigators have attempted to block HIV-1
envelope glycoprotein or the HIV-1 primary cell surface receptor, CD4, through the
application of neutralizing antibodies generated by either passive or active immunization
(Pegu et al. 2014).

Figure 1.4.1 Function of dendritic cells in the immune response to viruses. After the uptake
of viral antigen, myeloid and plasmacytoid DCs (mDCs & pDCs) migrate to lymph node to prime
naïve CD4
+
and CD8
+
T-cells. Additionally, activated DCs produce plenty of cytokines, which
affect T-cell survival and differentiation. CD4
+
T-cells differentiate into T helper 1 (T
H
1) and T
helper 2 (T
H
2), dependent on the cytokine signal. T
H
1 cell-mediated IFN-γ release induces the
production of immunoglobulin G2a antibodies by B cells. As for T
H
2, its cell-mediated cytokine
secretion stimulates the production of immunoglobulin G1 antibodies by B cells. Virus-specific
antibodies produced by B cells can neutralize and prevent viral reinfection. Figure taken and
adapted from (Lambotin et al. 2010).

15
1.5 Tumor-associated antigens: Up-regulation of PD-L1 on tumor cells
Human cancer immunotherapy based on vaccination has been a hot issue for its
potential capability to successfully threat cancer patients for decades. Early tumor
vaccines consisted of whole tumor cells, fragments of tumor cells, or protein lysate from
tumors (Lewis et al. 2003). However, limited outcomes along with these methods drove
researchers to the development of a novel aspect of cancer vaccines based on specific
tumor-associated antigens (TAAs). Recently, investigators have put emphases on the
research in identification, classification and characterization of TAAs on human cancers,
additionally, the development of new strategies to properly utilize and deliver the defined
antigens as therapeutic vaccines applied to experimental animals as well as to patients.
Originally, PD-L1 is absent in most normal body tissues; however, its expression can
be stimulated by IFN-γ cytokine in any nucleated cells. IFN-γ is primarily secreted by
inflammatory cells of hematopoietic origin, which especially refer to T cells, thus, it is
reasonable to suggest that cancer-induced inflammation will lead to the up-regulation of
PD-L1. Besides tumor cells, tumor antigens can be presented by tumor stromal cells and
hematopoietically derived infiltrating cells, containing dendritic cells, macrophages,
neutrophils and lymphocytes as well. “Adaptive resistance” derived from PD-1/PD-L1-
mediated evasion of tumor immune response may be initiated by the recognition of tumor
antigens by tumor-infiltrating lymphocytes (TILs) via their receptors (TCRs). Following
the specific recognition of tumor antigen, TILs (CD8
+
T cells) secrete IFN-γ and will
elicit PD-L1 expression on tumor cells, infiltrating hematopoietic cells or stromal cells in
tumor site.

16
This assumption is confirmed by immunohistochemistry-based detections that PD-L1
expression on cell surface is observed only in cells that are adjoining to T cells. The
release of IFN-γ can boost TIL effector functions through enhancing TIL differentiation
and inducing antigen processing and presentation, then PD-L1 presented on the cell
surface binds its receptors such as PD-1 and B7-1 on effector T cells (Teffs), hence
resulting in T cell dysfunction. In normal tissues, up-regulation of PD-L1 functions to
prevent extensive inflammation as well as to avoid collateral tissue damage; nevertheless,
over-expression of PD-L1 in the tumor microenvironment serves as a negative feedback
mechanism to inhibit tumor immunity (Chen and Han. 2015).
The assumption has further been supported by previous studies, which found that there
is a strong connection between the presence of TILs and up-regulation of PD-L1 in
human melanocytic lesions. It showed an over-expression of PD-L1 on tumor cells, and
those clusters of PD-L1 ligands directly bound surrounding TILs. Then, the expression
level of IFN-γ was determined at the interface of TILs and PD-L1
+
tumors while it
remained undetected in PD-L1
-
tumor cells. Additionally, in a mouse tumor model, IFN-γ
was neutralized by the application of a monoclonal antibody and the expression of PD-L1
on tumors was shown to be highly reduced, implying that IFN-γ is a critical mediator and
inducer of PD-L1 expression event. In 2013, a study from Spranger et al. indicated that
the increased expression of PD-L1 was dependent on CD8
+
T cells (infiltrated immune
cells that secrete IFN-γ) and interferon-γ (Spranger et al. 2013). Accordingly, the
adaptive resistance mechanism perfectly explains how cancer evades immune
surveillance, regardless of the endogenous anti-tumor immunity.

17
In spite of the fact that IFN-γ is regarded as a major inducer of the up-regulation of
PD-L1 on tumor cells, there is still an intrinsic IFN-γ-independent PD-L1 expression
mechanism possessed by tumor cells. As the evidence shows, some portions of human
cancers still over-express PD-L1 without TILs in the tumor microenvironment. There are
numerous studies that have manifested the mutations of EGFR (Akbay et al. 2013), the
loss of the tumor suppressor gene, phosphatase and tensin homolog (PTEN) (Parsa et al.
2007), activation of AKT and signal transducer and activator of transcription 3 (STAT3)
signaling pathways, and the constitutive expression of anaplastic lymphoma kinase
(ALK) via its signaling pathway are able to directly cause “up-regulation of PD-L1” on
cancer cells. Together, PD-L1 expression can be induced by the intrinsic signals and
extrinsic signals of tumor cells (Chen and Han. 2015).
In 2013, an article from Akbay et al. demonstrated that mutant EGFR induced the
elevated expression of PD-L1 in mouse lung tumors. PD-1 antibody immune checkpoint
blockade enhanced T-cell function, which further ameliorated the survival rate of mice
with oncogenic mutant EGFR-driven adenocarcinomas. Thus, this study indicates that
PD-1/PD-L1 immune-checkpoint pathway modulated immune evasion of lung tumors
driven by the mutant EGFR signal. In addition, KRAS signaling pathway could increase
PD-L1 expression as well. Nevertheless, KRAS-driven mouse lung tumors were unable
to be suppressed and controlled via PD-L1 blockade, showing the fact that this immune
checkpoint blockade alone is not sufficient for efficacious cancer immunotherapy in those
KRAS-driven immunosuppressive pathways of cancers (Akbay et al. 2013).
Furthermore, recent studies showed that the over-expression of PD-L1 could be
induced by the application of radiotherapy (Deng et al. 2014). Irradiation enhanced PD-

18
L1 expression on tumors and myeloid derived suppressor cells (MDSCs), as a result,
leading to an inhibition of radiation-induced immunity. Combination therapy with PD-L1
blockade and irradiation that applies both immunotherapy and radiotherapy
simultaneously with the participation of CD8
+
T cells and secreted interferon-γ gives rise
to optimal antitumor immune responses and therapeutic efficacy (Xu et al. 2014).
All together, the up-regulation of PD-L1 on tumor cells or tumor-infiltrating immune
cells brings out an optimal target site for immunotherapy with the application of PD-
1/PD-L1 immune checkpoint blockade by anti-PD-1 or anti-PD-L1 antibodies. Moreover,
the expression of PD-L1 antigen on tumors is elicited by intrinsic signals (gene
mutations, the activation of oncogenes or the loss of tumor suppressor genes) and
extrinsic signals (the induction of TILs, mainly CD8
+
T cells, and the release of IFN-γ).
PD-L1 has been proved to be immunogenic, which can provoke PD-L1-specific
immune responses. PD-L1-specific CTLs may boost immunity by the killing of
immunosuppressive tumor cells as well as regulatory cells (Munir et al. 2013). The
critical role of the PD-1 pathway is not at the initial T-cell activation stage. Instead, it is
to regulate effector T-cell responses by down-regulating T-cell activity to avoid collateral
tissue damage. The surface expression of PD-L1 on cancer cells suppresses cytotoxic T-
lymphocytes because of enhanced levels of PD-1 on the surface of these T cells.
A previous study described that the respective counteractive PD-L1-specific cytotoxic
T-lymphocytes (CTLs) have been established by the immune system. PD-L1-reactive
CTLs were able to be isolated mostly from peripheral blood of cancer patients and from
blood of healthy donors to a lesser extent. However, these PD-L1-specific CTLs not only
recognized and killed cancer cells, but also PD-L1-expressing APCs in a PD-L1-

19
dependent manner. Thus, the regulating role of PD-L1-specific CTLs in the immune
system may vary. Results in that study indicated that the presence of PD-L1-specific
CTLs during the activation stage of an immune response suppresses immunity on account
of the removal of PD-L1-expressing APCs. On the contrary, PD-L1-specific CTLs
effectively boost the immunity at the stage of "effector phase" of an immune response via
the removal of PD-L1-expressing regulatory cells that suppress PD-1-expressing effector
T cells. Hence, the overall biological function of PD-L1-specific CTLs may depend on
the microenvironment and the stage of an immune response. It is noticeable that PD-L1-
specific cytotoxic T-lymphocytes are beneficial to be applied for anti-leukemia
immunotherapy in order to combat immune evasion exploited in diverse cancers
mediated by the PD-1 signaling pathway (Ahmad et al. 2014).

20
1.6 Breaking in immune tolerance by linking tumor-associated antigens with an
immune stimulus (GM-CSF)
Granulocyte-macrophage colony-stimulating factor (GM-CSF) is a growth factor first
identified as an inducer of proliferation and differentiation of granulocytes (neutrophils,
eosinophils, and basophils) and macrophages derived from hematopoietic progenitor cells
(Wicks and Roberts. 2016). GM-CSF also has profound influences on the functions of
several circulating leukocytes. It is secreted by a wide range of cell types containing T
lymphocytes, macrophages, fibroblasts and endothelial cells when those cells are
activated via immune stimuli. Thus, it plays a part in the immune/inflammatory cascade,
including the biological process for fighting infection.
Studies have shown that GM-CSF functions across numerous tissues and biological
processing pathways. It is involved in both innate and adaptive immune responses. This
monomeric glycoprotein functions as a cytokine that activates macrophages to suppress
fungal survival, involving in the signaling pathway of STAT3. GM-CSF elicits deprival
of intracellular free zinc and the increase of the release of reactive oxygen species that
result in fungal zinc starvation and toxicity. As a result, GM-CSF is regarded as a critical
hematopoietic growth factor and modulator in immune system (Shi et al. 2006).
A critical determinant of host immune response is the mixture of cytokines produced
in the tumor microenvironment. In 2002, an article from Dr. Dranoff indicated that the
ability for many secreted and surface molecules to boost tumor immunity through
transferring genes into cancer cells have been compared. By using several mouse models,
GM-CSF was confirmed to be the most potent immunostimulator. It was found that after
DC vaccination with engineered GM-CSF protein product to tumor cells, tumor antigen

21
presentation by DCs and macrophages has increased. The results also showed enhanced
functions of CD4
+
and CD8
+
T cells, increased CD1d-restricted NKT cells and antibody-
mediated protective immune responses upon the vaccination. The efficacy of this
vaccination strategy applied to patients with advanced melanoma cancer was proved by
the consistent induction of coordinated humoral and cellular antitumor immune responses
following by a considerable necrosis of distant metastases (Dranoff. 2002).
The secretion of GM-CSF in tumor microenvironment caused the recruitment of
substantial professional APCs, which revealed that one vital function of this cytokine was
the augmentation of tumor antigen presentation. DCs are already well known for their
abilities to efficiently express antigen fused with MHC molecules on their surfaces,
express great amounts of co-stimulatory molecules, and release cytokines following
maturation and migration toward lymph nodes to further activate T lymphocytes and
antibody responses. Thus, GM-CSF-expressing B16 melanoma cells induced higher
levels of protective immunity and enhanced antitumor immune responses. This
convincing evidence implies that appropriate pharmacological delivery of GM-CSF may
bring therapeutic efficacy for vaccination in cancer immunotherapy. In this study, an
initial clinical evaluation of GM-CSF-based tumor vaccines has illustrated the consistent
induction of immune responses without significant toxicity (Dranoff. 2002).
This cytokine, GM-CSF, has been and is still widely used as an immunologic adjuvant
in clinical trials of vaccination with tumors, peptides, recombinant proteins and/or
dendritic cells in a wide range of human neoplasms. GM-CSF was applied either as the
product of gene-transfected tumor cells (GM-CSF-expressing tumors) or as recombinant
protein along with the vaccine given. The dosage and vaccine-specific immune response

22
in patients with different solid tumors were studied, and early phase clinical testing
indicated promising evidence of safety and activity (Gupta and Emens. 2010). Some of
the result of these studies demonstrated that this cytokine appeared to assist in generating
immune responses. Together, the preclinical and initial clinical studies all could be
regarded as convincing evidence to support continued clinical research and development
of these targeted vaccines. By integrating them with novel and effective cancer therapies
could lead to the promotion of vaccine efficacy through boosting co-stimulatory signaling
pathways of T-cell activation and conquering immune suppression.
Recent vaccination studies toward cancer immunotherapy have shown that many
diverse cytokines have been regarded as effective immune adjuvants and utilized in
facilitation of antigen recognition and T cell proliferation in both animal models and
human bodies. The application of GM-CSF as an immune adjuvant for its ability to
promote DC maturation and function has been studied.
For patients with advanced disease, GM-CSF-based tumor vaccines should be
synergistic with PD-L1 antibody blockade therapy, producing the prospective
recombinant protein-loaded DC tumor vaccines. GM-CSF-based tumor vaccines have
brought out new ideas and prospects for their induction of antitumor activity by applying
immunotherapy. However, clinical trials to date lacking potent and supportive data, have
only dropped a hint of prospective clinical activity. Accordingly, it is assumed that
integrating these GM-CSF-secreting tumor vaccines with novel immune modulators to
overcome immune tolerance can finally highlight the utility of these vaccines in cancer
therapy (Gupta and Emens. 2010).

23
1.7 DC-based tumor vaccines

Cancer has continuously been a major cause of mortality, even though there have been
many advances in radiation therapy and the application of those selected and proved
chemotherapeutic drugs. In recent years, cancer immunotherapy has been a hot issue
through applying monoclonal antibodies and cytokines to combat cancers. Up to now,
several licensed anti-cancer monoclonal antibodies have already been used. In the area of
cancer immunotherapy, expanded preclinical research in mice and many clinical trials
have been conducted. However, a standardized human cancer immunotherapy based on
vaccination, has not been well developed (Proudfoot et al. 2007).
Dendritic cells (DCs) are critical antigen-presenting cells (APCs) that are unique for
their abilities to stimulate and activate naive CD4
+
(helper) and CD8
+
(cytotoxic) T cells,
and induce effective immune responses when loaded with antigen. Antigen presentation
allows for the induction of antigen-specific adaptive immunity, and the enhanced immune
responses are able to fight against both intracellular and extracellular pathogens. Also,
this induced antigen-presenting feature of adaptive immunity leads to the defense against
specific antigen-expressing tumor types. Thus, some cancer immunotherapies apply
artificial DCs to prime the adaptive immune system for targeting advanced cancers.
Dendritic cells are often called “nature’s adjuvants” due to their ability to engulf the
target antigen and further present it on their surfaces; thus, the provision of an antigen
along with an adjuvant, the dendritic cell, to elicit therapeutic antigen-specific T cells. As
a result, dendritic cells have become the natural agents for antigen delivery. Furthermore,
the special properties of dendritic cells that keep them at the core of the immune system
is their ability to control between immunity and immune tolerance. The development of

24
DC vaccination is due to the characteristics of these dendritic cells in coordinating innate
and adaptive immunity. The goal of the strategies of DC vaccination by means of the
provision of DCs with tumor-specific antigens is to induce tumor-specific effector T cells
that can decrease the tumor volume/mass specifically and that can further induce
immunological memory to control tumor relapse. Therefore, Dendritic cells are essential
targets in order to generate therapeutic immune responses and through the identification
of cancer-specific antigens to develop antigen-specific immunotherapy against cancer
(Palucka and Banchereau. 2012).
One of the essential immune processes is antigen presentation. Owing to the property
that T cells only recognize antigens displayed on the cell surfaces, an antigen-presenting
complex is needed for the detection of infectious cells. After infection with viruses,
bacteria or microorganisms, the cells will present endogenous peptide fragments derived
from the pathogen through major histocompatibility complex, abbreviating in MHC
molecules on the cell surface. There are two types of MHC, according to the source of the
antigens: MHC class I molecules (MHC-I) combine with peptides from the cytosol, while
those peptides generated in vesicles are bound to MHC class II molecules (MHC-II).
Thus, DCs can deliver exogenous antigens to the MHC class I processing pathway for the
activation of CTLs (Melief. 2008). T cells can recognize among ten to hundreds of
thousands of peptides, owing to the ability of each MHC molecule that can bind a
different range of peptides. Activated T cells drive DCs towards their ultimate maturation,
inducing further expansion, proliferation and differentiation of T lymphocytes into
effector T cells. The microbial and inflammatory inducing “danger signals” and T cell-
derived activation signals elicit DC maturation and the release of interleukin 12 (IL-12).

25
Vaccination with tumor antigen-loaded DCs has been demonstrated to induce
antitumoral CTL responses in vivo and cause therapeutic tumor regression in cancer
patients. Pancreatic carcinoma cells can be suppressed in vivo after the application of
tumor antigen-loaded DC vaccines (Bauer et al. 2007). Nevertheless, in order to more
efficiently overcome the tumor immunosuppressive environment, combination therapy
with radiation or chemotherapy appears to promote the therapeutic efficacy of DC-based
tumor vaccines to successfully combat cancers.
Chemotherapies, radiotherapies and passive immunotherapies can bring immediate
clinical benefits when it comes to treating the existing disease. However, they cannot
bring out long-term protection from metastatic disease recurrence (Proudfoot et al. 2007).
On the other hand, the function of those therapeutic vaccines mostly depends on antigen-
specific CD8
+
T cells that will differentiate into CTLs to eventually reject cancer cells.
An effective cellular immunotherapy may have two potent favorable aspects. First, the
vaccines should prime naïve T cells and mediate existing memory T cells to enlarge an
immune response as well as to improve long-term efficaciousness. Second, through
establishing disease-specific memory immune responses, the vaccines can reduce and
avoid metastatic disease recurrence. For instance, DC-based vaccination should result in
the production of long-lived memory CD8+ T cells that are capable to prohibit cancer
relapse (Palucka and Banchereau. 2013).

26
CHAPTER 2: PDL1-GMCSF FUSION PROTEIN-LOADED DC
V ACCINATION ACTIV ATES PDL1-SPECIFIC
HUMORAL AND CELLULAR IMMUNE RESPONSES

ABSTRACT
Tumor antigen-specific cytotoxic T lymphocytes (CTLs) could be activated and
enhanced by the novel approach of dendritic cell (DC)-based tumor vaccines, inducing
effective anti-tumor activity. Programmed death-ligand 1 (PD-L1) was found to be highly
and commonly expressed on the surface of tumor cells and tumor-infiltrating immune
cells as well, such as antigen-presenting cells (APCs), the dendritic cells in several solid
malignancies including pancreatic carcinoma, suggesting an ideal tumor-specific antigen
for cancer immunotherapy and the development of immunotherapeutic vaccine. However,
a major hindrance for the therapy of pancreatic carcinoma with immune-associated
strategies is the development of immune tolerance toward tumor-associated antigens,
which leads to the failure of those natural tumor antigens to induce efficiently effective
immune response. In this study, a synthetic vaccine composed of the extracellular domain
of PD-L1 with a link to human GM-CSF (functions as a cytokine) is generated and
introduced to increase the immunogenicity of the tumor-specific antigen (human PD-L1).
We also have generated the mouse pancreatic tumor originated from murine Panc02 cells
that have been transduced with a membrane-bound form of retroviral PD-L1 to ensure the
Panc02 tumor cells express PD-L1 tumor antigen at a high level. After the vaccination of
PD-L1-GM-CSF fusion protein-loaded DCs, those immunized mice released an increased
amount of Interferon-gamma (IFN-γ) and Interleukin-2 (IL-2) and elicited stronger anti-
PD-L1 antibody responses and protective T cell-dependent CTL immunity as well.
Furthermore, this DC-based PD-L1-GM-CSF fusion protein vaccine showed significant

27
immunotherapeutic anti-tumor activity and prolonged survival of tumor-bearing mice.
These findings suggest that there is a potential efficient and promising strategy for the
immunotherapy of pancreatic carcinoma via the application of this novel PD-L1-GM-
CSF fusion protein-loaded DC vaccination.

28
INTRODUCTION
Pancreatic cancer (PC) is currently found to be the fourth major cause of cancer-
related deaths in both men and women in the United States (Dragovich. 2016), with a
five-year survival rate of only 2% (< 5%) (Siegel, Miller et al. 2015). Data indicate that
pancreatic cancer escape from the immune system via the secretion of chemicals from
tumor cells that actively steer immune cells away from the target tumor (Kotteas et al.,
2016). Pancreatic cancer is one of the most lethal malignancies found to be resistant to
conventional therapies. Even though some chemotherapy drugs have been applied to
cancer, they only provide modest survival benefit. Also, targeted therapy in combination
with chemotherapy has not shown significant improvement in outcome of treatments
(Bauer et al. 2007; Louvet et al. 2005). For pancreatic cancers, to overcome the
immunosuppressive tumor environment and enhance immunity, an optimal approach is
required, including mechanisms that are able to target specific tumor-associated antigens
for PC therapy. Pancreatic tumor cells can be recognized by tumor-specific T
lymphocytes (Ito et al. 2001; Schnurr et al. 2001). Tumor-infiltrating immune cells, T
helper cells and cytotoxic T lymphocytes (CTLs) differentiated from CD8
+
T cells
represent beneficial prognostic factors for patients with pancreatic cancers (Fukunaga et
al. 2004). Also, pancreatic carcinoma cells can be suppressed in vivo after the application
of tumor antigen-loaded DC vaccines (Bauer et al. 2007). Together, these factors provide
a potent rationale for the research and development of PC immunotherapy.
Tumor-specific T lymphocytes can be elicited and activated by Dendritic cell (DC)-
based vaccines (Palucka and Banchereau. 2013). DCs are major professional APCs of the
immune system that are unique for their abilities to induce and recruit naïve CD4
+
and

29
CD8
+
T cells. This allows for the elicitation of antigen-specific adaptive immune
responses and the defense against specific antigen-expressing tumors, resulting in the
enhanced antitumor CTL response and tumor regression (Melief. 2008). An antigen-
specific adaptive immunity derived from the stimulation of DCs can further elicit
immunological memory to control tumor relapse (Palucka and Banchereau. 2012).
DCs are often loaded with synthetic peptides from defined tumor-associated antigens
(TAAs), tumor lysate, or chemotherapy to generate immune responses (Amin and
Lockhart. 2015). Early tumor vaccines consisted of whole tumor cells, fragments of
tumor cells, or protein lysate from tumors (Lewis et al. 2003). However, limited
outcomes along with these methods drove researchers to the development of a novel
aspect of cancer vaccines based on specific tumor-associated antigens (TAAs). Defined
synthetic peptides in accordance with the antigenic determinants recognized by T
lymphocytes in the immune system have been utilized to activate MHC class I pathways
to further increase antitumor CTL responses (Kavanagh et al. 2007; Melief. 2008).
Recent vaccination studies toward cancer immunotherapy have shown that many
diverse cytokines have been regarded as effective immune adjuvants and utilized in
facilitation of antigen recognition and T cell expansion in both animal models and
humans. The application of GM-CSF as an “immune adjuvant” for its ability to promote
DC maturation and function that primes tumor specific CD4
+
and CD8
+
T cells to
enhance antitumor activity has been studied and confirmed (Gupta and Emens.
2010). Furthermore, functional CD8
+
T cells induced by DC-based vaccines can
differentiate into CTLs and provide long-lived memory CD8
+
T cells that can generate
new effector T cells, releasing cytotoxic molecules such as perforin and granzymes to

30
further prevent the recurrence of metastatic cancer diseases (Palucka and Banchereau.
2013). Treatment with DC-based vaccines is seen to be a logical and feasible approach,
due to the minimal toxicity. To date, vaccine therapy has not provided the anticipative
outcomes (Kotteas et al. 2016). Nevertheless, there was evidence showing the induction
of dendritic cell-mediated antitumor cytotoxic response against allogeneic pancreatic
carcinoma cells in vivo (Stift et al. 2003). On account of the rare case that has been
studied and what has been discussed above, combinational treatments including DC-
based vaccines are required to get involved in optimal pre-clinical as well as clinical trials
for PC immunotherapy.
Programmed death-ligand 1 (PD-L1) is up-regulated in multiple solid tumors and
tumor-infiltrating immune cells, which can suppress cytokine secretion and T cell-
mediated immune response upon binding its receptors, PD-1 or CTLA-4 that are negative
immunogenic regulators of T-cell activation (Herbst and Soria et al. 2014). Binding of
PD-L1 to PD-1 results in malfunction of effector T cells and evasion of tumor immunity.
PD-L1 has been proved to be immunogenic antigen, which can provoke PD-L1-specific
immune responses. PD-L1-specific CTLs may boost immunity by the killing of
immunosuppressive tumor cells as well as regulatory cells (Munir et al. 2013). The
clinical studies of immunotherapeutic PD-L1 antibody drugs (BMS-936559 (MDX-
1105), Brahmer and Tykodi et al. 2012; MPDL3280A (Genentech/Roche), Herbst and
Soria et al. 2014) for selected advanced cancers have proved that via PD-1/PD-L1
immune checkpoint blockade can ultimately enhance T-cell responses, anti-tumor
activity, and control tumor growth. However, for PC treatments, the neutralization of
immune checkpoint blockade antibodies (CTLA-4 or PD-1/PD-L1) did not reach

31
therapeutic efficacy (Brahmer et al. 2012). New clinical trails with antibody-mediated
blockade therapy involving nivolumab and pembrolizumab are continuous in
combinations with other vaccines or chemotherapies (Kotteas et al. 2016). Numerous
studies regarding PC therapy gained similar outcomes implying that up-regulation of PD-
L1 in human pancreatic carcinoma cells is associated with poor prognosis (Wang et al.
2010; Geng et al. 2008; Nomi et al. 2007). PD-L1 is rarely expressed on normal tissues
while highly expressed on tumor cells, tumor-infiltrating hematopoietic cells, or stromal
cells in tumor site (Chen and Han. 2015; Dong et al. 2002). This demonstrates that the
selective PD-L1 expression might be related to clinical results and could be a promising
target for cancer immunotherapy. In recent studies, a fusion protein with PD-L2 and Fc
was shown to improve the antitumor immune effect of GM-CSF-releasing whole cell
vaccine (Kojima et al. 2014). Whereas, whether a potential PD-L1-targeting DC-based
vaccine, which is a combination of checkpoint inhibitors with vaccination, can generate
antitumor CTL response in tumor-bearing mouse models or human patients and provide
anticipated therapeutic efficacy, still remains elusive.
In this study, a new synthetic fusion protein DC-based vaccine comprises human PD-
L1 extracellular domain and human GM-CSF that employs the pET-21a
+
-His tag system
was generated. This PD-L1-GM-CSF fusion protein-loaded DC vaccine was capable to
elicit an effective PD-L1-specific antitumor CTL response and suppress the growth of
stably expressing PD-L1-Panc02 tumors in mice. These results provide a novel and
potent strategy for pancreatic cancer immunotherapy.

32
MATERIALS AND METHODS

Mice and cell lines
Female C57BL/6 mice (7-8 week old) were purchased from Jackson Laboratories
(New York, NY). Mice were housed in animal facilities of the USC institution, and all
animal experiments were performed in accordance with an animal protocol approved by
Institutional Animal Care and Use Committee (IACUC) of the University of Southern
California in compliance with the Guide for the Care and Use of Laboratory Animals.
There are two types of murine tumor cell lines used in our studies, described as follows:
The murine pancreatic adenocarcinoma Panc02 cell line was originally established by
Corbett et al., and was a kind gift from Dr. Q Yao (Baylor College of Medicine, Houston,
Texas). Murine colon adenocarcinoma MC-38 cell line was kindly provided by Dr. John
C. Morris (National Cancer Institute, Bethesda, Maryland). Both of these two cell lines
were maintained in DMEM medium (4.5 g/L glucose) supplemented with 10% fetal
bovine serum (FBS), 1% L-glutamine, 1% sodium pyruvate, and 1% streptomycin-
penicillin-neomycin solution (Biological Industries) at 37°C in 5% CO
2
in cell culture
incubator. Additionally, the cell lines, Panc02 and MC-38, stably expressing human PD-
L1 were maintained in the same DMEM medium with all the supplements described as
above while with the individual addition of 3.5 µg/ml and 2 µg/ml puromycin
(Invitrogen).

33
Protein production and purification
The recombinant vector pET-21a-PD-L1-GM-CSF-his was constructed by introducing
a synthetic PD-L1 (nucleotide: 420 bps; 120 aa)-spacer-hGM-CSF (nucleotide: 432 bps;
144 aa) sequence into the Nde I/Xho I cut pET-21a
+
vector (EMD Millipore, Darmstadt,
Germany). The recombinant plasmid was transformed into BLR (DE3) competent E. coli
(EMD Millipore). The protein expression of transformed cells was induced by 1mM
IPTG as the OD reached between 0.6 and 0.8, and the cells were cultured/shaken at 24°C,
225 rpm, overnight. Those recombinant proteins were highly expressed in the inclusion
bodies rather than in the supernatant. The cell pellets were washed by wash buffer (100
mM Tris-Cl, pH 7.0, 5 mM EDTA, 2 M urea and 2% Triton X-100) for the removal of
unnecessary proteins. 20 mM Tris-HCl pH 7.5 was used to resuspend and stabilize the
pellets with the addition of 1% PMSF protease inhibitor. Cells were lysed by sonication
on ice at 400 Watts (sonication for 10 s, intermission for 10 s, repeated for 5 times). The
expressed recombinant proteins were purified in denaturing conditions through the
application of Ni-NTA Superflow Columns (Qiagen) for gravity-flow purification of His-
tagged proteins according to the manufacturer’s instructions (Denaturing
Lysis/Wash/Elution Buffer: 8M urea, 100mM NaH
2
PO
4
, 10mM Tris-Cl, pH 8.0/pH 6.3/
pH 4.5). The purified recombinant proteins were then further dialyzed by a 10K MWCO
(3 mL) Slide-A-Lyzer Gamma Irradiated Dialysis Cassette in PBS following the
manufacturer’s instructions (Thermo Fisher Scientific, Waltham, Massachusetts). The
purified (before dialysis) and the already dialyzed recombinant proteins were
analyzed/confirmed separately through running the SDS-PAGE with the following

34
analysis of Coomassie Blue Staining and Western Blotting. The purified recombinant
protein stocks were stored at -20°C for further application.
Retroviral-transduced tumor cell lines
To produce retroviral vectors, 293-T packaging cells were cultured in 100-mm culture
dishes with DMEM medium (4.5 g/L glucose) supplemented with 10% fetal bovine
serum (FBS), 1% L-glutamine, 1% sodium pyruvate, and 1% A/A at 37°C in 5% CO
2
and
transfected with 1 ml Transfectagro
TM
Reduced serum medium (Corning) containing 10
µg PD-L1-retroviral vector plasmid, pCMV-Gag-Pol (10 µg) and pCMV-VSV-G (3.5
µg) and with the addition of polyethylenimine (PEI, 1 µg/µl) (the volume of PEI used
was based on a 3:1 ratio of PEI (µg): total DNA (µg)). After overnight incubation, the
medium was replaced with DMEM containing 5% FBS. 24 and 48 hours later, the culture
medium containing recombinant retroviruses was collected and filtered through a 0.45
µm filter (EMD Millipore, Darmstadt, Germany). The non-concentrated harvested
retroviruses were transduced into the exponentially growing tumor cell lines (Panc02,
MC-38, PY-230 and GL-261) containing 8 ng/ml Polybrene. I selected the positive PD-
L1 expressing tumor cells with the addition of optimal final concentration of puromycin
(1 mg/ml) as follows: Panc02 (4 µg/ml), MC-38 (2 µg/ml), PY-230 (2.5 µg/ml) and GL-
261 (0.8 µg/ml). I cultured those stably-expressing PD-L1 tumor cells and then harvested
them for the further Flow Cytometry analysis.

35
Preparation of bone marrow-derived dendritic cells
C57BL/6 mouse bone marrow (BM)-derived dendritic cells (DCs) were prepared as
described previously (Shen et al. 2004). Briefly, mouse bone marrow cells were flushed
from the bones of mouse limbs, passed through a nylon mesh, and depleted of red blood
cells with ammonium chloride. After extensive washing with RPMI-1640 medium, cells
were cultured in RPMI-1640 medium supplemented with 10% FBS, 1% PSG, 1% L-
glutamine, 1% NEAA, 1% HEPES, 0.1% 2-Mercaptoethanol, recombinant mouse GM-
CSF/ml (20 ng/ml; PeproTech) and recombinant mouse IL-4 (20 ng/ml; PeproTech).
Non-adherent granulocytes were eliminated after 2 days (48 hours) of culture. To
generate DCs, every other day, the supernatant was removed and then replaced with fresh
medium containing mGM-CSF (20 ng/ml) and mIL-4 (20 ng/ml). All cultures were
incubated at 37°C in 5% humidified CO
2
. Bacterial lipopolysaccharides (LPS; Sigma)
was added (100 ng/ml) at the time point of the last 16 hours for BM-DC maturation. After
7 days of cell culture, above approximate 80% of the cells expressed characteristic DC-
specific markers such as CD11c, CD40, CD80 and CD86 as determined through FACS.

DC immunization and tumor challenge studies
For the preparation of tumor-specific antigen-pulsed DCs, Bone marrow-derived DCs
were incubated overnight with the addition of protein vaccine (100 µg/ml) or the control
(without the addition of protein) in complete RPMI-1640 culture medium supplemented
with the above cytokines (mGM-CSF and mIL-4). Dendritic cells were then double-
pulsed with a lower amount of protein vaccine (50 µg/ml) or the control for 2 hours
before DC vaccination. After being washed by PBS for 2-3 times, these tumor-specific

36
antigen-pulsed DCs were ready for immunization. C57BL/6 mice were randomly split
into three different groups (5 mice per group, three groups) as the followings: (1) PBS
control (2) hGM-CSF (3) PD-L1-GM-CSF. Mice were immunized via footpad injection
(0.5-1 x 10
6
cells per mouse) for two or three times with an interval of 7-10 days. To
further test the efficacy of DC vaccines, after 7-10 days of the last round of DC
immunization, the exponentially growing stably expressing human PD-L1 Panc02 or
MC-38 tumor cell lines (5 x 10
4
cells) were subcutaneously injected into the flanks of
different groups of the immunized mice. Tumor sizes were measured every 3-4 days by a
caliper, with tumor volumes calculated as follows: (longest diameter) x (shortest
diameter)
2
(You et al. 2000).

Analysis of antibody responses – ELISA
Peripheral blood was collected from the murine tail veins, and anti-PD-L1 antibodies
in the sera of immunized mice were determined by enzyme-linked immunosorbent assay
(ELISA) (BioLegend, San Diego, California) following the manufacturer’s instructions.
In brief, microtiter plates coated with recombinant PD-L1 proteins (100 ng per well) were
incubated with serially diluted sera in a blocking buffer (KPL Inc.) at room temperature
for 2 hours with constant shake. Bound antibodies were detected after incubation with
avidin-horseradish peroxidase (HRP)-conjugated antibodies against mouse IgG (Sigma-
Aldrich) that have been diluted in the blocking buffer. Read absorbance at 450 nm, and
non-immunized murine sera served as negative controls.

37
Intracellular staining (ICS) and Flow Cytometry analysis
Splenocytes were isolated and harvested from immunized C57BL/6 mice three days
after the last round of DC immunization. Prepared splenocytes (1 x 10
6
cells per well)
were re-stimulated with protein vaccine (5 µg/ml) in a U-bottom 96-well plate for 6 hours
with the addition of recombinant mouse IL-2 (20 ng/ml; PeproTech). An inhibitor of
intracellular protein transport, Brefeldin A (5 µg/ml; eBioscience, San Diego, California),
was added to block the protein transportation toward the Golgi complex and accumulate
proteins in the ER, leading to enhanced detection of intracellular cytokines. After the re-
stimulation process, the activated cells were firstly incubated with anti-mouse CD11c
(dendritic-cell surface marker), CD4 and CD8 (T-cell surface markers) antibodies for cell
surface staining. Subsequently, intracellular staining for IFN-γ, IL-2, perforin and
Granzyme B were conducted after the fixation and permeabilization of those activated
splenocyte cells. Fixable viability dye was used to sort out live cells. Flow Cytometry
was performed using a MMI FACS CANTO II (BD Biosciences, Los Angeles, CA, U.S.)
and data were analyzed with FlowJo software (Tree Star Inc., Ashland, OR, U.S.). Those
antibodies used in this study were acquired from BD Biosciences (San Diego, CA, U.S.)
and listed as follows: anti-mouse CD11c (APC), anti-mouse CD40 (FITC), anti-mouse
CD80 (APC), anti-mouse CD86 (FITC), anti-mouse CD4 (FITC), anti-mouse CD8a
(APC), anti-mouse IFN-γ (PE-cy7), anti-mouse IL-2 (PE), anti-mouse CD107a (PE-cy7),
anti-mouse perforin (PE), anti-mouse Granzyme B (eFluor 660), anti-mouse PD-L1
(CD274) (APC). Fixable viability dye, ghost dye
TM
violet 510 was purchased (TONBO
biosciences, San Diego, CA, U.S.).

38
Enzyme-linked immunospot (Elispot) assay
The ELISPOT assays were conducted as described in our previous studies (Huang et
al. 2003). In brief, MultiScreen
TM
96-well assay plates (Millipore, Darmstadt, Germany)
were coated with anti-mouse IFN-γ monoclonal antibody (100 ng per well) overnight at
4°C in ELISPOT coating buffer. Plates were blocked with complete RPMI-1640 (with
10% FBS, 1% PSG, 1% L-glutamine) at room temperature for one hour. Then,
splenocytes (1 x 10
5
or 2 x 10
5
cells per well (duplicate)) from immunized mice were co-
cultured with protein (50 ng per well) in the coated plate at room temperature for 24
hours. After washing, biotinylated anti-mouse IFN-γ detection antibody was added to
each well, and plates were incubated at 37°C for 2 hours. After washing, Avidin-HRP
reagent was added to each well and then incubated at room temperature for one hour.
Lastly, freshly-prepared AEC Substrate Solution (Thermo Fisher Scientific, Waltham,
Massachusetts) was added to each well and then monitored the development of spots.
After spots develop, the plate was rinsed completely with ddH
2
0 and air-dried the plate.
The results were evaluated and analyzed through a dissecting microscope and automated
ELISPOT plate reader system (Zellnet Consulting Inc., Mineola, New York).
Western blotting analysis
To verify the target proteins, protein fractions were collected individually as the
following: (1) Cell Lysis (CL) (2) Flow Through (FT) (3) Wash Fraction (WF) (4) Eluted
fraction (E) (5) Dialyzed fraction (D) to run SDS-PAGE using precast NuPAGE Bis-Tris
4-12% Gel (Novex, Invitrogen, Waltham, Massachusetts) and then transferred the protein
through iBlot 2 NC mini Stacks (Novex)with an iBlot® 2 Dry Blotting System. The
membranes were blocked in 1:1 Odyssey Blocking Buffer with TBST or 5% nonfat dry

39
milk with TBS at room temperature for one hour and then probed with primary antibodies
(1) anti-6X His tag® mouse mAb (Abcam) or (2) anti-PD-L1 rabbit polyclonal Ab (Santa
Cruz Biotechnology), both at a dilution of 1:1000 in the cold room, overnight. Second
antibodies were then applied, including (1) goat anti-mouse (2) goat anti-rabbit, both at a
dilution of 1:10,000 and incubated at room temperature for one hour. Membranes were
analyzed with Odyssey Imaging System (LI-COR Biosciences). Protein concentrations
were determined using the BCA Protein Assay Kit (Thermo Fisher Scientific).

Statistical analysis
All in vitro and in vivo results described in this study represent of the mean ± SEM of
at least three-independent experiments. Statistical significance was determined using the
t-test and One Way ANOV A test with SigmaStat software. Differences were considered
significant at p<0.05. Regression plots were constructed using SigmaPlot software and
PRISM 7 software (GraphPad Software, Inc., La Jolla, California).

40
RESULTS
2.4.1 Production of a recombinant PD-L1-GM-CSF fusion protein
The over-expression of PD-L1 molecule in cancer (tumor cells or tumor-infiltrating
immune cells such as dendritic cells and macrophages) has been associated with the
resistance of several human solid malignancies including breast, ovarian, oral, head and
neck, brain (Yao et al. 2009 and Jacobs et al.), lung, liver, colorectal (Dong et al. and Hua
et al.), pancreatic (Nomi et al. 2007, Huang et al. 2009 and Wang et al. 2010) and skin
cancers, etc. toward anti-cancer therapies (Afreen and Dermime. 2013); however, PD-L1
is rarely expressed on normal human tissues. These findings lead to a powerful
assumption that the PD-L1 molecule is an ideal antigen target for the research on cancer
immunotherapy. Allogeneic granulocyte-macrophage colony-stimulating factor (GM-
CSF)-secreting tumor vaccines were found to be able to treat established tumors in the
mouse (Jaffee et al. 2001). GM-CSF is an attractive adjuvant for a DNA vaccine due to
its ability to recruit antigen-presenting cells (macrophages, dendritic cells, lymphocytes
and monocytes) to the site of antigen synthesis, thus, this cytokine has been used as an
immune stimulus to elicit an immune response (Yoon and Aleyas et al. 2006).
To evaluate the anti-tumor efficacy of tumor antigen-loaded DCs in priming CTLs
directed against murine pancreatic cancer cells, a recombinant peptide segment
containing the extracellular domain of human PD-L1 (amino acid 19-159, total 140 a.a.)
which the T cell epitopes were included in, fusing with SGSG linker and human full-
length GM-CSF (144 a.a.), was synthesized (Figure 1A). To obtain an assembly pET-
21a
+
-PD-L1-GM-CSF-His fusion protein vaccine, the synthetic peptide was then cloned

41
in-frame into pET-21a
+
-His expression plasmid (Figure 1B). The recombinant assembly
plasmid (pET-21a
+
-PD-L1-GM-CSF-His) was transformed into BLR (DE3) E.coli and
further induced by 1 mM IPTG. The expression of the fusion protein was then confirmed
by SDS-PAGE, where it was highly expressed in pellet containing IPTG induction with
an expected molecular weight of 33.79 kDa (Figure 1C, indicated as the arrow points at).
Next, the His-tagged fusion protein was purified through Ni-NTA superflow columns
under denaturing conditions, and the cell lysis, flow through, wash fraction and elutes
were further analyzed by SDS-PAGE with coomassie blue staining (Figure 1D) and
western blotting analysis (Figure 1F). In order to get an isotonic and non-toxic biological
environments for cells, in other words, to reach the physiological consistency, the urea-
rich elutes (purified protein) were further dialyzed and then confirmed by SDS-PAGE
with coomassie blue staining (figure 1E) and western blotting analysis (Figure 1G, (a)
anti-His Ab and (b) anti-PDL1 Ab). The other three kinds of fusion proteins containing
the PD-L1-GM-CSF segment were also purified and dialyzed, and were analyzed by
SDS-PAGE (Figure 1H and 1I) for future applications. Applying the purification
procedures mentioned previously, approximate 15 mg high purity recombinant pET-21a
+
-
PD-L1-GM-CSF fusion protein vaccine was obtained from 1 L bacterial culture for
further functional characterization.

42

Figure 2.4.1 Production of a pET-21a
+
-PD-L1-GM-CSF protein vaccine.
(A) Amino acid sequences of the gene encoding variable extracellular regions of human PD-L1
(140 a.a.) and human full-length GM-CSF (144 a.a.) protein, separated by a SGSG spacer.
(B) Schematic map of the pET-21a
+
-PDL1-GM-CSF-His plasmid DNA vaccines. The fusion
segment containing extracellular domain of human PD-L1, linker and human GM-CSF was
designed with the Nde I and Xho I restriction sites of pET-21a
+
downstream of promoter.
(C) The fusion protein vaccines of pET-21a
+
-PD-L1-GM-CSF-His expressed in pellet and
supernatant were induced or non-induced by IPTG (1 mM) shown at the expected size through
SDS-PAGE analysis (Bio-Safe™ Coomassie Stain, Bio-Rad). (D) Coomassie blue staining and
(F) Western blotting analysis of the protein expression in different collected fractions during the
purification process were shown. (E) Coomassie blue staining of the elute and dialyzed protein.
(G) Western blotting analysis of the dialyzed protein with the incubation of anti-his-tag antibody
shown in the left and anti-PDL1 antibody shown in the right. (H) Western blotting analysis of the
four different purified/ (I) dialyzed fusion proteins while all containing the PDL1-GMCSF
segment with the incubation of anti-his-tag antibody was shown. M, Marker; S, Supernatant; P,
Pellet; ∆, Non-induced by IPTG; ▲, IPTG induction; CL, Cell lysis; FT, Flow through; WF,
Wash fraction; E, Elutes; D, Dialyzed protein; V, PDL1-GMCSF target protein vaccine (33.79
kDa); 1, IDH1(R/H)-PDL1-GMCSF (44.66 kDa); 2, EVIII-PDL1-GMCSF (50.73 kDa); 3,
MSLN-PDL1-GMCSF (58.24 kDa)

43
2.4.2 Maturation of the PD-L1-GM-CSF fusion protein-loaded DCs

C57BL/6 mouse bone marrow-derived dendritic cells (BMDCs) were prepared as
described previously (Shen et al. 2004). On day 0, BM progenitor cells seeded into
culture dishes showed a spherical morphology (Figure 2A). Cells were small, had defined
cell membranes, and were in great health. On day 3, the initiation of colony formation
happened at various sites with some aggregations of BMDCs and some cells were
converted into adherent macrophages (Figure 2B). Then, on day 6, the immature BMDCs
were obtained with the culture in the medium containing recombinant mouse GM-CSF
(rmGM-CSF). Clusters of colonies of BMDCs were formed, and amounts of floating and
semi-adherent BMDCs were seen (Figure 2C). Subsequently, the cells were treated with
maturation stimulus LPS (100 ng/ml) for 24 hours. Increased cell population and
obviously huge clusters of colonies were seen. Simultaneously, the dendrites of cells were
extended and more adherent macrophages were observed (Figure 2D and 2E (with the
addition of fusion protein as well)) (Madaan et al. 2014).
In addition, to detect the maturation of BMDCs, on day 6 of culture, the DCs were
individually loaded with PD-L1-GM-CSF protein, PBS and LPS. After overnight
stimulation, the expression levels of T-cell co-stimulatory molecules-CD40, CD80 and
CD86, were determined by surface staining and further flow cytometry analysis. An
approximate 5.5-fold increased expression of CD40 was observed after stimulation by
LPS compared to the PBS control (Figure 2F). However, if DCs were loaded with only
PD-L1-GM-CSF protein without stimulation of the additional LPS, the expression levels
of CD40 would remain low. Thus, LPS was known as an essential inflammatory stimulus
to induce DC maturation. In the meanwhile, the prominently increased release of T-cell

44
costimulatory molecules, CD80 and CD86, was observed after LPS stimulation (Figure
2G). The results show that the antigen uptake and DC maturation events have occurred
during this culture and stimulation process.

Figure 2.4.2 Morphology of murine BMDCs at different days of culture and costimulatory
molecule (CD40, CD80 and CD86) expression of murine mature-BMDCs. Murine bone
marrow cells were cultured in RPMI-1640 10% FBS containing 20 ng/ml rmGM-CSF and 20
ng/ml IL-4 for 6 days. Fresh medium was supplemented every two days (at day-2 and day-4). (A)
Murine bone marrow progenitor cells at day-0. (B) BMDCs at day-3. (C) Immature un-treated-
BMDCs at day-6. (D) BMDCs were treated with maturation stimulus, LPS (100 ng/ml) for 24
hours. (E) BMDCs were loaded with PD-L1-GM-CSF fusion protein at day-6 and then after 4
hours, stimulated by LPS (100 ng/ml) for another 24 hours. The red arrows indicate the sites of
colony formation at day-3, day-6 and day-7 (magnification-100X). (F) & (G) The expression
level of costimulatory molecules, CD40, CD80 and CD86 was determined by surface staining and
further flow cytometry analysis. The CD11c (murine DC-specific marker) positive cells were
gated in all groups.

45
2.4.3 PD-L1-GM-CSF-loaded DCs enhance cytokine production

To evaluate the efficacy of PD-L1-GM-CSF fusion protein vaccine to induce effective
cellular immune responses, naïve C57BL/6 mice were individually immunized twice at
one-week intervals with 100 µg/ml of PD-L1-GM-CSF fusion protein vaccine, 100 µg/ml
of GM-CSF protein and PBS-DC only. Seven days after the second immunization, the
splenocytes from the immunized mice were isolated and digested into single-cell
suspensions, and then were further stimulated by 5 µg/ml protein vaccine in an U-bottom
96- well plate for another 6 hours. Then, the frequency of IFN-γ-producing T cells was
determined by Elispot. According to expectation, the activated murine T lymphocytes
through immunization with PD-L1-GM-CSF fusion protein-loaded DCs released a
significantly higher level of IFN-γ (type II interferon) compared to cells either
immunized with hGM-CSF protein-loaded DCs or PBS-DC alone. An approximate 5-fold
and 3-fold upregulation of IFN-γ secretion was observed as 1×10
5
and 2×10
5
murine
splenocytes were coated in a 96-well plate respectively (Figure 3A).
The indicators such as CD4 and CD8 molecules have been selected to assess the
efficacy of cancer immunotherapy. The changes of these two indexes should be
statistically significant in comparison of the effectively treated group and the control
group. Thus, as anticipated, the prominent increase of CD4
+
and CD8
+
populations in the
murine splenocytes of the PD-L1-GM-CSF fusion protein-loaded DC vaccine group was
observed (Figure 3B, indicated in red marks). IL-2, a type of cytokine signaling molecule
in the immune system, functions primarily via its direct effects on T cells. The frequency
of IL-2-producing CD4
+
T cells was determined and analyzed by intracellular staining
and then flow cytometry. The result showed that the expression level of IL-2-producing

46
CD4
+
T cells was enhanced in a 15-fold increased drift (Figure 3C, indicated in red mark)
of the fusion protein-immunized group compared to the control group. These data
perfectly indicate that this novel PD-L1-GM-CSF fusion protein-loaded DC vaccine
elicits enhanced IFN-γ and IL-2 secretion from activated T lymphocytes; in other words,
would generate better potential to trigger CTL responses.

Figure 2.4.3 PD-L1-GM-CSF fusion protein-loaded DC vaccine generates increased IFN-γ
and IL-2 secretion from activated CD4
+
and CD8
+
T cells. Naïve C57BL/6 mice were
individually immunized twice at one-week intervals with 100 µg/ml of PD-L1-GM-CSF fusion
protein vaccine, 100 µg/ml of GM-CSF protein and PBS-DC only. 7 days after the second
immunization, murine splenocytes were isolated and stimulated with additional protein vaccines
(5 µg/ml) for another 6 hours. (A) The frequency of IFN-γ-producing T cells was examined by
Elispot. Data represent as mean ± SEM. ** p < 0.01 (B) Analysis of CD4 and CD8 population in
murine splenocytes by intracellular staining and further flow cytometry. (C) The frequency of IL-
2-producing CD4
+
T cells was determined by intracellular staining and further flow cytometry.

47
2.4.4 Inhibition of tumor growth by PD-L1-GM-CSF-loaded DC vaccines

After that, mouse cancer models have been used to test the efficacy of this PD-L1-
GM-CSF fusion protein vaccine for the treatment of pancreatic carcinoma. Thus, a stably
expressing PD-L1 Panc02 pancreatic tumor cell line was produced through retroviral PEI
transfection into 293 T cells performed as standard protocols. Then I added the selectable
marker, puromycin, and successfully selected the positively expressing PD-L1 Panc02
tumor cells after the transduction was performed by standard methods. The increased
expression levels of PD-L1 on this established stably expressing PD-L1 Panc02 tumor
cell line was analyzed by flow cytometry analysis (Figure 4A).
Next, stably expressing PD-L1 Panc02 tumor cells (2.5 x 10
5
cells per mouse) were
subcutaneously injected into mice after one week of the second DC vaccination, and the
tumor growth was observed and monitored at 5-day intervals. We surprisingly found that
the group vaccinated with the PD-L1-GM-CSF fusion protein more efficiently delayed
tumor growth compared to both the GM-CSF vaccine and the PBS control group (Figure
4B). Interestingly, the immunized mice from GM-CSF protein vaccine group also showed
efficacy for the inhibition of tumor growth. Remarkably, the PD-L1-GM-CSF fusion
protein vaccine significantly enhanced the survival of tumor-bearing mice compared to
other two groups, and 40% of the target protein vaccine treated mice survived over 80
days (Figure 4C). The similar trend of tumor growth was observed for the same grouping
of mice inoculated with MC38 tumor cells (2.5 x 10
5
cells per mouse) (Figure 4D).
Therefore, these data suggest that this novel fusion protein vaccine (PD-L1-GM-CSF
pulsed DCs) could be a potently therapeutic vaccine by comparison with the conventional

48
non PD-L1 DC targeting protein vaccines.

Figure 2.4.4 Vaccination with PD-L1-GM-CSF fusion protein-loaded DCs can efficiently
inhibit tumor growth in mice. C57BL/6 mice were individually immunized twice at one-week
intervals with 100 µg/ml of PD-L1-GM-CSF fusion protein vaccine, 100 µg/ml of GM-CSF
protein and PBS-DC only. A week after the second vaccination, the immunized mice were all
inoculated with 2.5 ×10
5
stably expressing PD-L1 Panc02 or 2.5 ×10
5
MC38 tumor cells by
subcutaneous injection. (A) The expression level of PD-L1 on the stably expressing PD-L1
Panc02 cells was determined by flow cytometry. (B) The tumor size of mice inoculated with PD-
L1-Panc02 was measured every 5 day and growth curve was calculated as described previously.
(C) The survival curve of mice inoculated with PD-L1-Panc02 was indicated (n = 5). (D) The
tumor size of mice inoculated with MC38 was monitored and growth curve was calculated as
well. All data represent as mean ± SEM. ** p < 0.01

49
2.4.5 Induction of humoral and cellular immune responses by
PD-L1-GM-CSF-loaded DC vaccines
Furthermore, to study the ability of the PD-L1-GM-CSF fusion protein-loaded DCs
for the induction of specific anti-PD-L1 antibodies, the sera from the tails of immunized
mice were collected and ELISA assays were applied to measure antibody levels. Levels
of anti-PD-L1 antibodies (IgG) in sera of PD-L1-GM-CSF-loaded-DC immunized mice
significantly increased (p < 0.05) compared to other two immunized groups (hGM-CSF-
loaded-DC and PBS-DC/control) and wild type (non-immunized with DCs) as well
(Figure 5A).
The inter-DC antigen transfer functions to amplify antigen presentation across a
broader network and interaction of lymphoid-resident DCs for effective T-cell activation
(Allan et al. 2006). Once getting into the murine bodies, DCs migrate to the lymphoid
tissues and lead to peptide-reactive CTL responses (Proudfoot et al. 2007) with the
release of granzymes (cytotoxic T lymphocyte-associated serine protease) and perforins
(glycoproteins), resulting in lysis of the targeted tumor cells (Anguille et al. 2013;
Trapani and Smyth. 2002). Although CD8
+
T cells are the main source of Granzyme B
and perforin during immunological responses, activated CD4
+
T cells express Granzyme
B and a low amount of perforin as well (Lin et al. 2014; Osińska et al. 2014).
Accordingly, murine splenocyte T cells were obtained 3 days after the second DC
immunization (naïve C57BL/6 mice were immunized with fusion protein-loaded DCs
twice) as reported previously. Then, not significantly while notably, the increased
expression level of Granzyme B was observed in PD-L1-GM-CSF fusion protein-loaded
DC-immunized murine splenocytes through intracellular staining and flow cytometry

50
analysis (Figure 5B). These data imply that vaccination with these fusion protein-loaded
DCs, activated-humoral immune responses, and protective CTL responses occur.

Figure 2.4.5 Fusion protein-loaded DC Vaccination increases anti PD-L1 antibody responses
and generates significantly enhanced cytotoxic T cell responses. Naïve C57BL/6 mice were
individually immunized twice at one-week intervals with 100 µg/ml of PD-L1-GM-CSF fusion
protein vaccine, 100 µg/ml of GM-CSF protein and PBS-DC only. (A) 3 days after the second
immunization, murine sera were collected and analyzed by ELISA. (B) 7 days after the second
immunization, murine splenocytes were isolated from each group and cultured with RPMI-1640,
following stimulation by IL-2, protein vaccines (100 µg/ml) and GolgiPlug (protein transport
inhibitor) for 6 hours. The splenocyte T cells were directly incubated with surface marker
antibodies, anti-CD4
+
and anti-CD8
+
, and the intracellular staining with the application of anti-
mouse granzyme-B antibody and anti-mouse perforin antibody (result not shown). The results
were analyzed by flow cytometry analysis.

51
DISCUSSION

The cancer survival rate overall has doubled over the past 40 years; while for some
cancer types, there has been seldom improvement in the clinical cancer treatment (Smith.
2016). This is especially true for pancreatic cancer, where the survival rate has hardly
enhanced for decades. Pancreatic cancer is a lethal disease and remains one of the most
resistant cancer types toward traditional cancer therapies such as chemotherapy and
radiotherapy (Gunturu et al. 2013). Therefore, the emergence of cancer immunotherapy
might shed light on the treatment for advanced pancreatic cancer on account of applying
a totally different mechanism. It works by triggering the immune system through
recruiting and activating T cells that recognize tumor-specific antigens to enhance the
antitumor activity.
Immunotherapy in pancreatic carcinoma (PC) is an evolving, still unexplored prospect.
During particular stages of PC pathogenesis and evolution, tumor-associated antigens
may be generated by the mutation of oncogenes and can be overexpressed in PC cells at
solid stages of tumor pathogenesis, leading to hindrance of the therapeutic efficacy of
vaccine immunotherapy (Kotteas et al. 2016). The evolution of PC pathogenesis requires
an inflammatory response. KRAS mutations are major events in the early stages of PC
pathogenesis, which induce transforming growth factor-β and interleukine-10 (IL-10)
expression to increase the inflammatory response (Evans and Costello. 2012). Cytokines
involved in PC mediate tumor microenvironment and stimulate escape of tumor
immunity. Cancer cells can express immunosuppressive cytokines as well (Dennis et al.
2013). In PC, IL-10 is found to be highly released and boosts the activation of T-

52
regulatory cells (Tregs) (Bellone et al. 2006). Also, metastasis of PC cells has been found
to be dependent on the signaling pathway via IL-17B receptor (Wu et al. 2015). Together,
KRAS-targeting vaccines could prove only useful in delaying or preventing PC
progression but could not be a great benefit to invasive disease (Khvalevsky et al. 2013).
The obstacles for DNA-based vaccine therapy are the generation of functionally
mature DC and effective DC loaded with tumor antigens (Subbotin. 2014). Unfortunately,
tumor cells persistently mutate by their nature, unpredictably modifying mutated tumor
antigens (Subbotin. 2014). According to the illustration in the introduction, PD-L1 is
known to be mainly expressed on tumor cells or tumor-infiltrating immune cells in PC
while is rarely expressed on normal tissues. These make PD-L1 an optimal and ideal
target antigen for PC immunotherapy. In our current study, we have constructed and then
successfully expressed a synthetic recombinant protein vaccine that comprises the
integral part of human PD-L1 extracellular domain fused with human GM-CSF to
improve the immunogenicity of tumor antigen-specific vaccine. Tumor antigens linked to
GM-CSF were confirmed to activate ex vivo autologous cells and peripheral-blood
mononuclear cells (PBMCs) such as DCs, inducing the activation of tumor infiltrating
lymphocytes and further leading to effective antitumor activity in humans (Kantoff et al.
2010; Schellhammer et al. 2013). Thus, the PD-L1 along with GM-CSF in this
constructed protein vaccine was considered to bring an extra therapeutic efficacy via the
induction of antitumor CTL response. Mouse Panc02 cells transduced with human
retroviral PD-L1 were utilized to establish a stably expressing PD-L1-Panc02 tumor cell
line for subsequent generation of murine pancreatic tumors. This strategy was
surprisingly found to be able to elicit enhanced antibody response as well as effective

53
antitumor CTL response to significantly suppress Panc02 tumor volume and prolong the
survival rate of Panc02 tumor-bearing mice. Our study provides an approach based on
protein/peptide loaded-DC-derived vaccine, which brings out an effective strategy for PC
immunotherapy. This design may be beneficial to the therapies that apply target of
diverse neo-antigens (Delamarre et al. 2015).
In the complex invasive PC microenvironment, an immunosuppressive state is
generated through decreased cytotoxic CD8
+
and helper CD4
+
T cells, while increased
regulatory T cells (Tregs), and tumor-associated macrophages (Evans and Costello. 2012),
cause the insensitivity toward reduced amount of CTLs. To increase the amount of tumor
antigen presented by DCs, an alternative method containing the culture and
differentiation of mature DCs from monocytes in vitro, along with specific antigen and
immune stimulus/adjuvants, being inoculated into patients as therapeutic vaccines, was
used. Cancer cells exploit the immune checkpoint pathway to escape from tumor
immunity. The PD-L1-PD-1 axis plays a critical role in down-regulating antitumor CTL
responses and tumor growth control. Through the application of immune checkpoint
blockade, immune system is resumed and is triggered to send out the alarming signals,
then launching a system-wide attack on cancer cells (Hammerich et al. 2015). Consistent
with our hypothesis, statistically significant anti-PD-L1 antibody response was detected
in murine sera upon the immunization with this novel PD-L1-GM-CSF DC-based
vaccine in the study. This indicated that the enhanced PD-L1 antibodies may somehow
neutralize the activity of PD-L1/PD-1 immune checkpoint. Therefore, antibody-mediated
immune response against tumor cells may be one of the vital mechanisms toward
combating advanced cancers. Antibody therapy has been confirmed to have great potency

54
for cancer treatment. Two critical mechanisms utilized by antibody drugs to kill targeted
tumor cells are antibody-dependent cell-mediated cytotoxicity (ADCC) and complement-
dependent cytotoxicity (CDC) (Kellner et al. 2014). Thus, we need to determine whether
the production of PD-L1-specific antibodies in mice immunized with PD-L1-GM-CSF-
DCs is correlated with the levels of ADCC and CDC activities.
In addition, to investigate the production of antigen-specific T cell responses induced
by this novel PD-L1-GM-CSF DC-based vaccine in murine models, antitumor CTL
responses were measured in vitro. Higher levels of CD4
+
/CD8
+
T cells, IFN-γ and IL-2
secretion were found in the effector cells differentiated from murine splenocytes
immunized with PD-L1-GM-CSF DC-based vaccine. This implys that helper T-cell
responses and cytotoxic T-cell responses were primed. The production of IFN-γ and IL-2
promotes CTL development and also delays hypersensitivity reactions, which will be
beneficial to anticancer immune response.
Our murine pancreatic tumor models were generated by the inoculation of the PD-L1
retroviral transduced Panc02 tumor cell line. Even though in our study, only 20% of the
PD-L1 transduced Panc02 tumor cells expressed PD-L1, this novel immunotherapeutic
strategy for the immunization with PD-L1-GM-CSF DC-based vaccine could
significantly suppress Panc02 tumor growth and prolong the survival of Panc02 tumor-
bearing mice as well. The reason for studying the murine Panc02 tumor model is on
account of its poor immunogenicity. This is similar to one of the characteristics of human
pancreatic carcinoma (Turnquist et al. 2004). Nevertheless, Panc02 tumor cells can be
targeted and killed by CTLs despite that fact that low levels of MHC class I expression,
and immunotherapeutic approaches can be effective in this tumor model (Schmidt et al.

55
2006). Based on the knowledge of the PD-L1 up-regulation event in several tumor cells,
we were interested in applying this novel vaccine to another type of tumor. Thus, we
generated another murine colon adenocarcinoma tumor model through the inoculation of
the MC-38 tumor cell line. The result shows that this PD-L1-GM-CSF DC-based vaccine
could also significantly suppress non-artificially expressing PD-L1-MC-38 tumor growth.
GM-CSF, an immune stimulus, has already been known to be capable of inducing
immunity. Thus, we found that the GM-CSF DC vaccine could inhibit tumor growth as
well. However, the GM-CSF alone DC vaccine could not effectively extend survival of
tumor-bearing mice. Together, these results, though not that elegant, occurred in some
assays. This indicated that this newly constructed PD-L1-GM-CSF fusion protein-loaded
DC vaccine can successfully elicit strong anti-PD-L1 T cell-dependent humoral and
cellular immune responses, control tumor growth, and improve survival for therapeutic
cancer immunotherapy.
The neutralization of checkpoint inhibitors via receptor/ligand target of CTLA-4 or
PD-1/PDL-1 has been proved to be ineffective (Anjali et al. 2015; Brahmer et al. 2012;
Royal et al. 2010). Until now, GVAX containing the transduced GM-CSF gene is found
to be the most promising pancreatic cancer vaccine; whereas, vaccine therapy has not
reached the desirable outcomes (Kotteas et al. 2016). Additionally, CARs in solid tumors
have not produced significant results to date (Abate Daga et al. 2014), according to their
limited efficacy and unacceptable toxicity. Increasing evidence indicates that well-
established therapeutic strategies, including chemotherapy, radiation and debulking
surgery, may be successfully combined with immunotherapies (Broomfield et al. 2005;
Correale et al. 2005). Thus, based on the results obtained from our study, combinations of

56
this PD-L1-GM-CSF DC-based vaccine with more optimal trial designs are possible to
bring promising approaches for pancreatic cancer treatment by immunotherapy.

57
CHAPTER 3: SUMMARY AND FUTURE DIRECTIONS

3.1 Summary

We demonstrated that the PD-L1-GM-CSF DC-based vaccine can stimulate robust T
cell-dependent immune responses. This new vaccine, consisting of the fusion protein of
human PD-L1 protein fragment as an antigen and human GM-CSF protein as an immune
stimulus, loaded with DCs to elicit stronger anti-PD-L1 antibody responses and
protective T cell-dependent PD-L1-specific CTL immunity. Furthermore, this DC-based
PD-L1-GM-CSF fusion protein vaccine shows significant immunotherapeutic antitumor
activity and prolongs survival of PD-L1-expressing tumor-bearing mice. Therefore, the
feasibility of this novel immunotherapeutic strategy has been confirmed. This study
provides a newly feasible strategy to overcome immunosuppressive characteristics of
tumor cells overexpressing PD-L1. It also supports the ideas of combinational
immunotherapies, which the DC vaccination is included, may bring potential benefits in
cancer treatment.

58
3.2 Future Directions

A highly (80-90%) stably expressing PD-L1-Panc02 tumor cell line should be
generated, which can be regarded as a subpopulation compared to those without the up-
regulation of PD-L1. We believe that increasing the PD-L1 expression level on Panc02
tumor cells might lead to a greater significant difference of the immune effect toward
tumor regression among the PD-L1-GM-CSF vaccine group, the GM-CSF vaccine group
and the control group. Among them, we expect that the therapeutic efficacy of PD-L1-
GMCSF vaccine may surpass the results compared to the GM-CSF alone vaccine via its
ability to suppress the overexpressing PD-L1-Panc02 tumor growth.
One of the most difficulties is to figure out the optimal timing for the immunoassays.
To set up a correct timeline for the collection of murine sera, for the frequency of DC
immunization, for extraction of murine splenocytes after the sufficient immunization, and
for tumor injection that will lead to the most significantly therapeutic immunoassay
results is absolutely a big challenge. The magnitude of stimulated immune responses in
immunized mice is unknown and undetectable, unless through the increase of sample size,
which is the amount of experimental mice due to the sacrifice of several mice each time
when an immunoassay is applied. Thus, repetition of this study is essential to achieve
powerful data and convincing outcomes.
Clinical trials must confirm meaningful clinical benefit with the least toxicity. Thus,
whether the antitumor effect that results from PD-L1-GM-CSF DC vaccination can
damage normally adjacent liver and kidney tissue cells should be further studied.
Therefore, liver and kidney tissues from mice immunized with PD-L1-GM-CSF-DCs and

59
PBS-DCs can be isolated. Then, we will apply Hematolxylin & Eosin (H&E) staining on
sections and analyze this to detect whether there is a positive labeling in the cytoplasm of
livers and kidneys. We hypothesize this PD-L1-GM-CSF DC-based vaccine that induces
effective antitumor CTL responses will not attack adjacent normal tissues.
The combinations of vaccines with checkpoint inhibitors and more optimal trial
designs must be made to enhance the therapeutic efficacy on PC treatment. We then
suggest that blockade of PD-L1 can improve the therapeutic effect of PD-L1-GM-CSF
DC-based vaccine. Thus, our next step may test whether blocking PD-L1/PD-1 signaling
pathway may significantly enhance the antitumor activity induced by this vaccine by the
application of anti-PD-L1 antibody treatment. The combination of PD-L1-GM-CSF DC-
based vaccine with anti-PD-L1 antibody treatment is highly anticipated to dramatically
inhibit the tumor growth and promote survival of tumor-bearing mice as well.
Vaccination with tumor antigen‐loaded DCs has been shown to elicit anti-tumoral
CTL responses in vivo and to induce tumor regression. Thus, many appropriate neo-
antigens can be applied to the DC vaccination. To test whether these antigens loaded with
DCs can induce antigen-specific humoral and cellular immune responses, and inhibit the
specific antigen-expressing tumor growth and increase the survival of specific antigen-
expressing tumor-bearing mice.

60
REFERENCES

Aaron Goodman, Sandip P. Patel and Razelle Kurzrock (2017). PD-1–PD-L1 immune-checkpoint blockade in
B-cell lymphomas. Nat Rev Clin Oncol 14(4): 203-220.

Ahmad, S. M., S. K. Larsen, I. M. Svane and M. H. Andersen (2014). Harnessing PD-L1-specific cytotoxic
T cells for anti-leukemia immunotherapy to defeat mechanisms of immune escape mediated by the PD-1
pathway. Leukemia 28(1): 236-238.

Akbay EA, Koyama S, Carretero J et al. (2013). Activation of the PD-1 pathway contributes to immune
escape in EGFR-driven lung tumors. Cancer Discov. 3: 1355–1363.

Alka Madaan, Ritu Verma, Anu T Singh, Swatantra Kumar Jain, Manu Jaggi (2014). A stepwise procedure
for isolation of murine bone marrow and generation of dendritic cells. Journal of Biological Methods 1(1)

Alsheikhly AR, Zweiri J, Walmesley AJ, Watson AJ, Christmas SE (2004). Both soluble and membrane-
bound forms of Flt3 ligand enhance tumor immunity following "suicide" gene therapy in a murine colon
carcinoma model. Cancer Immunol Immunother 53: 946-954.

Amin M, Lockhart AC (2015). The potential role of immunotherapy to treat colorectal cancer. Expert Opin
Investig Drugs 24: 329-344.

Anguille S, Lion E, Van den Bergh J, Van Acker HH, Willemen Y, Smits EL et al (2013). Interleukin-15
dendritic cells as vaccine candidates for cancer immunotherapy. Hum Vaccin Immunother 9: 1956-1961.

Banchereau J, Steinman RM (1998). Dendritic cells and the control of immunity. Nature 392: 245-252.

Barth RJ, Jr., Fisher DA, Wallace PK, Channon JY, Noelle RJ, Gui J et al (2010). A randomized trial of ex
vivo CD40L activation of a dendritic cell vaccine in colorectal cancer patients: tumor-specific immune
responses are associated with improved survival. Clin Cancer Res 16: 5548-5556.

Bauer C, Bauernfeind F, Sterzik A, Orban M, Schnurr M, Lehr HA et al (2007). Dendritic cell-based
vaccination combined with gemcitabine increases survival in a murine pancreatic carcinoma model. Gut
56(9): 1275-1282.

Brahmer, J. R., S. S. Tykodi, L. Q. Chow, W. J. Hwu, S. L. Topalian, P. Hwu, C. G. Drake, L. H. Camacho,
J. Kauh, K. Odunsi, H. C. Pitot, O. Hamid, S. Bhatia, R. Martins, K. Eaton, S. Chen, T. M. Salay, S.
Alaparthy, J. F. Grosso, A. J. Korman, S. M. Parker, S. Agrawal, S. M. Goldberg, D. M. Pardoll, A. Gupta
and J. M. Wigginton (2012). Safety and activity of anti-PD-L1 antibody in patients with advanced cancer. N
Engl J Med. 366(26): 2455–2465.

61
Broomfield S, Currie A, van der Most RG, Brown M, van Bruggen I, Robinson BW et al (2005). Partial,
but not complete, tumor-debulking surgery promotes protective antitumor memory when combined with
chemotherapy and adjuvant immunotherapy. Cancer Res 65: 7580-7584.

Caminschi I, Vremec D, Ahmet F, Lahoud MH, Villadangos JA, Murphy KM et al (2012). Antibody
responses initiated by Clec9A-bearing dendritic cells in normal and Batf3(-/-) mice. Mol Immunol 50: 9-17.

Chen, L. and D. B. Flies (2013). Molecular mechanisms of T cell co-stimulation and co-inhibition. Nat Rev
Immunol 13(4): 227-242.

Chen, L. and X. Han (2015). Anti-PD-1/PD-L1 therapy of human cancer: past, present, and future. The
Journal of Clinical Investigation 125(9): 3384-3391

Chiodoni C, Paglia P, Stoppacciaro A, Rodolfo M, Parenza M, Colombo MP (1999). Dendritic cells
infiltrating tumors cotransduced with granulocyte/macrophage colony-stimulating factor (GM-CSF) and
CD40 ligand genes take up and present endogenous tumor-associated antigens, and prime naive mice for a
cytotoxic T lymphocyte response. J Exp Med 190: 125-133.

Corbett TH, Roberts BJ, Leopold WR, Peckham JC, Wilkoff LJ, et al. (1984). Induction and
chemotherapeutic response of two transplantable ductal adenocarcinomas of the pancreas in C57BL/6
mice. Cancer Res 44: 717–726.

Cornelis J.M Melief (2008). Cancer immunotherapy by dendritic cells. Immunity 29: 372-383.

Correale P, Cusi MG, Tsang KY, Del Vecchio MT, Marsili S, Placa ML et al (2005). Chemo-
immunotherapy of metastatic colorectal carcinoma with gemcitabine plus FOLFOX 4 followed by
subcutaneous granulocyte macrophage colony-stimulating factor and interleukin-2 induces strong
immunologic and antitumor activity in metastatic colon cancer patients. J Clin Oncol 23: 8950-8958.

Delamarre L, Mellman I, Yadav M (2015). Cancer immunotherapy. Neo approaches to cancer vaccines.
Science 348: 760-761.

Deng L, Liang H, Burnette B et al. (2014). Irradiation and anti-PD-L1 treatment synergistically promote
antitumor immunity in mice. J Clin Invest. 124: 687–695.

Dong H, Strome SE, Salomao DR, Tamura H, Hirano F, Flies DB et al (2002). Tumor-associated B7-H1
promotes T-cell apoptosis: a potential mechanism of immune evasion. Nat Med 8: 793-800.

Dranoff, G. (2002). GM-CSF-based cancer vaccines. Immunol Rev 188: 147-154.

Dustin, M.L., and Cooper, J.A. (2000). The immunological synapse and the actin cytoskeleton: molecular
hardware for T cell signaling. Nat. Immunol. 1: 23–29.

62
Esche C, Subbotin VM, Maliszewski C, Lotze MT, Shurin MR (1998). FLT3 ligand administration inhibits
tumor growth in murine melanoma and lymphoma. Cancer Res 58: 380-383.

Evel-Kabler, K., X. T. Song, M. Aldrich, X. F. Huang and S. Y. Chen (2006). SOCS1 restricts dendritic
cells’ ability to break self tolerance and induce antitumor immunity by regulating IL-12 production and
signaling. J Clin Invest 116(1): 90-100.

Fukunaga A, Miyamoto M, Cho Y, Murakami S, Kawarada Y, Oshikiri T et al (2004). CD8+ tumor-
infiltrating lymphocytes together with CD4+ tumor-infiltrating lymphocytes and dendritic cells improve the
prognosis of patients with pancreatic adenocarcinoma. Pancreas 28: e26-31.

Geng L, Huang D, Liu J, Qian Y, Deng J, Li D et al (2008). B7-H1 up-regulated expression in human
pancreatic carcinoma tissue associates with tumor progression. J Cancer Res Clin Oncol 134: 1021-1027.

Granucci, F., E. Ferrero, M. Foti, D. Aggujaro, K. Vettoretto and P. Ricciardi-Castagnoli (1999). Early
events in dendritic cell maturation induced by LPS. Microbes Infect 1(13): 1079-1084.

Gunturu, K. S., G. R. Rossi and M. W. Saif (2013). Immunotherapy updates in pancreatic cancer: are we
there yet? Ther Adv Med Oncol 5(1): 81–89.

Gupta, R. and L. A. Emens (2010). GM-CSF-secreting vaccines for solid tumors: moving forward. Discov
Med. 10(50): 52-60

Hammerich L, Binder A, Brody JD (2015). In situ vaccination: Cancer immunotherapy both personalized
and off-the-shelf. Mol Oncol 9: 1966-1981.
Hartkopf, A. D., F. A. Taran, M. Wallwiener, C. B. Walter, B. Kramer, E. M. Grischke and S. Y . Brucker
(2016). PD-1 and PD-L1 Immune Checkpoint Blockade to Treat Breast Cancer. Breast Care (Basel) 11(6):
385-390.

Herbst, R. S., J. C. Soria, M. Kowanetz, G. D. Fine, O. Hamid, M. S. Gordon, J. A. Sosman, D. F.
McDermott, J. D. Powderly, S. N. Gettinger, H. E. Kohrt, L. Horn, D. P. Lawrence, S. Rost, M. Leabman,
Y. Xiao, A. Mokatrin, H. Koeppen, P . S. Hegde, I. Mellman, D. S. Chen and F . S. Hodi (2014). Predictive
correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients. Nature 515(7528): 563–
567.

Hong, B., X. T. Song, L. Rollins, L. Berry, X. F. Huang and S. Y. Chen (2011). Mucosal and systemic anti-
HIV immunity controlled by A20 in mouse dendritic cells. J Clin Invest 121(2): 739-751.

Huang XF, Ren W, Rollins L, Pittman P, Shah M, Shen L et al (2003). A broadly applicable, personalized
heat shock protein-mediated oncolytic tumor vaccine. Cancer Res 63: 7321-7329.

Ito M, Shichijo S, Tsuda N, Ochi M, Harashima N, Saito N et al (2001). Molecular basis of T cell-mediated
recognition of pancreatic cancer cells. Cancer Res 61: 2038-2046.

63
Jacobs JF, Idema AJ, Bol KF, Nierkens S, Grauer OM, Wesseling P et al (2009). Regulatory T cells and the
PD-L1/PD-1 pathway mediate immune suppression in malignant human brain tumors. Neuro Oncol 11:
394-402.

Jaffee EM, Hruban RH, Biedrzycki B, Laheru D, Schepers K, Sauter PR et al (2001). Novel allogeneic
granulocyte-macrophage colony-stimulating factor-secreting tumor vaccine for pancreatic cancer: a phase I
trial of safety and immune activation. J Clin Oncol 19: 145-156.

Kantoff PW, Higano CS, Shore ND, Berger ER, Small EJ, Penson DF et al (2010). Sipuleucel-T
immunotherapy for castration-resistant prostate cancer. N Engl J Med 363: 411-422.

Katsuya Y, Fujita Y, Horinouchi H, Ohe Y, Watanabe S, Tsuta K (2015). Immunohistochemical status of
PD-L1 in thymoma and thymic carcinoma. Lung Cancer 88: 154-159.

Kavanagh B, Ko A, Venook A, Margolin K, Zeh H, Lotze M et al (2007). Vaccination of metastatic
colorectal cancer patients with matured dendritic cells loaded with multiple major histocompatibility
complex class I peptides. J Immunother 30: 762-772.

Kellner, C., S. Derer, T. Valerius and M. Peipp (2014). Boosting ADCC and CDC activity by Fc
engineering and evaluation of antibody effector functions. Methods 65(1): 105-113.

Kim HJ, Cantor H (2014). The path to reactivation of antitumor immunity and checkpoint immunotherapy.
Cancer Immunol Res 2: 926-936.
Kojima M, Murata S, Mekata E, Takebayashi K, Jaffee EM, Tani T (2014). Fusion protein of mutant B7-
DC and Fc enhances the antitumor immune effect of GM-CSF-secreting whole-cell vaccine. J Immunother
37: 147-154.

Kotteas E, Saif MW, Syrigos K (2016). Immunotherapy for pancreatic cancer. J Cancer Res Clin Oncol
142: 1795-1805.

Lambotin, M., S. Raghuraman, F. Stoll-Keller, T. F. Baumert and H. Barth (2010). A look behind closed
doors: interaction of persistent viruses with dendritic cells. Nat Rev Microbiol 8(5): 350-60

Louvet C, Labianca R, Hammel P, Lledo G, Zampino MG, Andre T et al (2005). Gemcitabine in
combination with oxaliplatin compared with gemcitabine alone in locally advanced or metastatic pancreatic
cancer: results of a GERCOR and GISCAD phase III trial. J Clin Oncol 23: 3509-3516.

Melief CJ (2008). Cancer immunotherapy by dendritic cells. Immunity 29: 372-383.

Munir, S., G. H. Andersen, O. Met, M. Donia, T. M. Frosig, S. K. Larsen, T. W. Klausen, I. M. Svane and
M. H. Andersen (2013). HLA-restricted CTL that are specific for the immune checkpoint ligand PD-L1
occur with high frequency in cancer patients. Cancer Res 73(6): 1764-1776.

64
Nomi T, Sho M, Akahori T, Hamada K, Kubo A, Kanehiro H et al (2007). Clinical significance and
therapeutic potential of the programmed death-1 ligand/programmed death-1 pathway in human pancreatic
cancer. Clin Cancer Res 13: 2151-2157.

Palucka, K. and J. Banchereau (2012). Cancer immunotherapy via dendritic cells. Nat Rev
Cancer 12(4): 265-277.

Palucka, K. and J. Banchereau (2013). Dendritic-cell-based therapeutic cancer vaccines. Immunity 39(1):
38-48.

Pardoll, D. M. (2012). The blockade of immune checkpoints in immunotherapy. Nature Reviews
Cancer 12: 252-264.

Parsa, A. T., J. S. Waldron, A. Panner, C. A. Crane, I. F. Parney, J. J. Barry, K. E. Cachola, J. C. Murray, T.
Tihan, M. C. Jensen, P. S. Mischel, D. Stokoe and R. O. Pieper (2007). Loss of tumor suppressor PTEN
function increases B7-H1 expression and immunoresistance in glioma. Nat Med 13(1): 84–88.

Pegu, A., Z. Y. Yang, J. C. Boyington, L. Wu, S. Y. Ko, S. D. Schmidt, K. McKee, W. P. Kong, W. Shi, X.
Chen, J. P. Todd, N. L. Letvin, J. Huang, M. C. Nason, J. A. Hoxie, P. D. Kwong, M. Connors, S. S. Rao, J.
R. Mascola and G. J. Nabel (2014). Neutralizing antibodies to HIV-1 envelope protect more effectively in
vivo than those to the CD4 receptor. Sci Transl Med 6(243): 243ra288.

Powles, T., J. P. Eder, G. D. Fine, F. S. Braiteh, Y. Loriot, C. Cruz, J. Bellmunt, H. A. Burris, D. P. Petrylak,
S. L. Teng, X. Shen, Z. Boyd, P. S. Hegde, D. S. Chen and N. J. Vogelzang (2014). MPDL3280A (anti-PD-
L1) treatment leads to clinical activity in metastatic bladder cancer. Nature 515(7528): 558–562.

Proudfoot O, Pouniotis D, Sheng KC, Loveland BE, Pietersz GA (2007). Dendritic cell vaccination. Expert
Rev Vaccines 6: 617-633.

Santarpia M, González-Cao M, Viteri S, Karachaliou N, Altavilla G, Rosell R (2015). Programmed cell
death protein-1/programmed cell death ligand-1 pathway inhibition and predictive biomarkers:
understanding transforming growth factor-beta role. Transl Lung Cancer Res. 4(6): 728–742.

Schellhammer PF, Chodak G, Whitmore JB, Sims R, Frohlich MW, Kantoff PW (2013). Lower baseline
prostate-specific antigen is associated with a greater overall survival benefit from sipuleucel-T in the
Immunotherapy for Prostate Adenocarcinoma Treatment (IMPACT) trial. Urology 81: 1297-302.

Schmidt J, Ryschich E, Sievers E, Schmidt-Wolf IG, Buchler MW, Marten A (2006). Telomerase-specific
T-cells kill pancreatic tumor cells in vitro and in vivo. Cancer 106: 759-64.

Schnurr M, Galambos P, Scholz C, Then F, Dauer M, Endres S et al (2001). Tumor cell lysate-pulsed
human dendritic cells induce a T-cell response against pancreatic carcinoma cells: an in vitro model for the
assessment of tumor vaccines. Cancer Res 61: 6445-50.

65
Schultz ES, Schuler-Thurner B, Stroobant V, Jenne L, Berger TG, Thielemanns K et al (2004). Functional
analysis of tumor-specific Th cell responses detected in melanoma patients after dendritic cell-based
immunotherapy. J Immunol 172: 1304-10.

Sharma P, Allison JP (2015). The future of immune checkpoint therapy. Science 348: 56-61.

Shen L, Evel-Kabler K, Strube R, Chen SY (2004). Silencing of SOCS1 enhances antigen presentation by
dendritic cells and antigen-specific anti-tumor immunity. Nat Biotechnol 22: 1546-53.

Shi, Y., C. H. Liu, A. I. Roberts, J. Das, G. Xu, G. Ren, Y. Zhang, L. Zhang, Z. R. Yuan, H. S. Tan, G. Das
and S. Devadas (2006). Granulocyte-macrophage colony-stimulating factor (GM-CSF) and T-cell
responses: what we do and don't know. Cell Res 16(2): 126-133.

Siegel RL, Miller KD, Jemal A (2015). Cancer statistics, 2015. CA Cancer J Clin 65: 5-29.

Song, X. T., K. Evel-Kabler, L. Rollins, M. Aldrich, F. Gao, X. F. Huang and S. Y. Chen (2006). An
alternative and effective HIV vaccination approach based on inhibition of antigen presentation attenuators
in dendritic cells. PLoS Med 3(1): e11.

Spranger, S., R. M. Spaapen, Y. Zha, J. Williams, Y. Meng, T. T. Ha and T. F. Gajewski (2013). Up-
regulation of PD-L1, IDO, and T(regs) in the melanoma tumor microenvironment is driven by CD8(+) T
cells. Sci Transl Med. 5(200): 200ra116.
Steinman, R. M. and M. Pope (2002). Exploiting dendritic cells to improve vaccine efficacy. J Clin Invest.
109(12): 1519-1526.

Stift A, Friedl J, Dubsky P, Bachleitner-Hofmann T, Benkoe T, Brostjan C et al (2003). In vivo induction of
dendritic cell-mediated cytotoxicity against allogeneic pancreatic carcinoma cells. Int J Oncol 22: 651-6.

Subbotin VM (2014). Dendritic cell-based cancer immunotherapy: the stagnant approach and a theoretical
solution. Drug Discov Today 19: 834-7.

Taube, J. M., R. A. Anders, G. D. Young, H. Xu, R. Sharma, T. L. McMiller, S. Chen, A. P. Klein, D. M.
Pardoll, S. L. Topalian and L. Chen. (2012). Colocalization of inflammatory response with B7-h1
expression in human melanocytic lesions supports an adaptive resistance mechanism of immune escape. Sci
Transl Med. 4(127): 127ra137.

Topalian SL, Drake CG, Pardoll DM (2015). Immune checkpoint blockade: a common denominator
approach to cancer therapy. Cancer Cell 27: 450-61.

Trapani JA, Smyth MJ (2002). Functional significance of the perforin/granzyme cell death pathway. Nat
Rev Immunol 2: 735-47.

66

Turnquist HR, Kohlgraf KG, McIlhaney MM, Mosley RL, Hollingsworth MA, Solheim JC (2004). Tapasin
decreases immune responsiveness to a model tumor antigen. J Clin Immunol 24: 462-70.

Vigneron, N. (2015). Human Tumor Antigens and Cancer Immunotherapy. Biomed Res Int 2015: 948501.

Wang L, Ma Q, Chen X, Guo K, Li J, Zhang M (2010). Clinical significance of B7-H1 and B7-1
expressions in pancreatic carcinoma. World J Surg 34: 1059-65.

Wang L, Rollins L, Gu Q, Chen SY, Huang XF (2009). A Mage3/Heat Shock Protein70 DNA vaccine
induces both innate and adaptive immune responses for the antitumor activity. Vaccine 28: 561-70.

Watts, C. (1997). Capture and processing of exogenous antigens for presentation on MHC molecules. Annu.
Rev. Immunol. 15: 821.

Wicks, I. P. and A. W. Roberts (2016). Targeting GM-CSF in inflammatory diseases. Nature Reviews
Rheumatology 12: 37-48

Wykes, M. and G. MacPherson (2000). Dendritic cell-B-cell interaction: dendritic cells provide B cells
with CD40-independent proliferation signals and CD40-dependent survival signals. Immunology 100: 1-3.

You Z, Hester J, Rollins L, Spagnoli GC, van der Bruggen P , Chen SY (2001). A retrogen strategy for
presentation of an intracellular tumor antigen as an exogenous antigen by dendritic cells induces potent
antitumor T helper and CTL responses. Cancer Res 61: 197-205.

Abstract (if available)

Abstract Tumor antigen-specific cytotoxic T lymphocytes (CTLs) could be activated and enhanced by the novel approach of dendritic cell (DC)-based tumor vaccines, inducing effective anti-tumor activity. Programmed death-ligand 1 (PD-L1) was found to be highly and commonly expressed on the surface of tumor cells and tumor-infiltrating immune cells as well, such as antigen-presenting cells (APCs), the dendritic cells in several solid malignancies including pancreatic carcinoma, suggesting an ideal tumor-specific antigen for cancer immunotherapy and the development of immunotherapeutic vaccine. However, a major hindrance for the therapy of pancreatic carcinoma with immune-associated strategies is the development of immune tolerance toward tumor-associated antigens, which leads to the failure of those natural tumor antigens to induce efficiently effective immune response. In this study, a synthetic vaccine composed of the extracellular domain of PD-L1 with a link to human GM-CSF (functions as a cytokine) is generated and introduced to increase the immunogenicity of the tumor-specific antigen (human PD-L1). We also have generated the mouse pancreatic tumor originated from murine Panc02 cells that have been transduced with a membrane-bound form of retroviral PD-L1 to ensure the Panc02 tumor cells express PD-L1 tumor antigen at a high level. After the vaccination of PD-L1-GM-CSF fusion protein-loaded DCs, those immunized mice released an increased amount of Interferon-gamma (IFN-γ) and Interleukin-2 (IL-2) and elicited stronger anti-PD-L1 antibody responses and protective T cell-dependent CTL immunity as well. Furthermore, this DC-based PD-L1-GM-CSF fusion protein vaccine showed significant immunotherapeutic anti-tumor activity and prolonged survival of tumor-bearing mice. These findings suggest that there is a potential efficient and promising strategy for the immunotherapy of pancreatic carcinoma via the application of this novel PD-L1-GM-CSF fusion protein-loaded DC vaccination.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Dendritic cell-specific vaccine utilizing antibody-mimetic ligand and lentivector system

PDF

Engineering viral vectors for T-cell immunotherapy and HIV-1 vaccine

Asset Metadata

Creator Jehng, Tiffany Chunmei (author)

Core Title PD-L1-GM-CSF fusion protein-loaded DC vaccination activates PDL1-specific humoral and cellular immune responses

School Keck School of Medicine

Degree Master of Science

Degree Program Molecular Microbiology and Immunology

Publication Date 07/19/2017

Defense Date 04/19/2017

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

Tag APCs, antigen-presenting cells,CTLs, cytotoxic T lymphocytes,DC vaccination,DC, dendritic cell,FACs, fluorescence-activated cell sorting,fusion protein,GM-CSF, granulocyte-macrophage colony-stimulating factor,interferon,interleukin,lipopolysaccharide,major histocompatibility complex,monoclonal antibody,OAI-PMH Harvest,PD-L1, programmed death ligand 1

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Chen, Si-Yi (committee chair), Landolph, Joseph (committee member), Lee, Ha-Youn (committee member)

Creator Email jehng@usc.edu,tiffyjehng@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-403909

Unique identifier UC11264411

Identifier etd-JehngTiffa-5553.pdf (filename),usctheses-c40-403909 (legacy record id)

Legacy Identifier etd-JehngTiffa-5553.pdf

Dmrecord 403909

Document Type Thesis

Rights Jehng, Tiffany Chunmei

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