Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Engineering therapeutics for the improved antitumor efficacy of chimeric antigen receptor T cell therapy
(USC Thesis Other)
Engineering therapeutics for the improved antitumor efficacy of chimeric antigen receptor T cell therapy
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
1
Engineering Therapeutics for the Improved
Antitumor Efficacy of Chimeric Antigen
Receptor T Cell Therapy
John Mac
Thesis
August 2019
A Dissertation Presented to the Faculty of the Graduate School
at the University of Southern California in Partial Fulfillment of the
Requirements of the Degree Doctor of Philosophy
(CHEMICAL ENGINEERING)
2
Committee Members
Dr. Pin Wang (Advisor)
Dr. Richard Roberts (Advisor)
Dr. Nicholas Graham
3
Table of Contents
Acknowledgements ............................................................................................ 4
Chapter 1: T Cell Immunotherapies Enhanced by Designer Biomaterials
1.1 Abstract and Introduction ......................................................................... 7
1.2 Discussion ................................................................................................ 10
1.3 Conclusion ............................................................................................... 39
1.4 Figures ..................................................................................................... 42
1.5 References ............................................................................................... 47
Chapter 2: A Tumor Stroma-Targeting Immunotoxin Engineered for Enhancing
the Efficacy of Immunotherapy in a B16 Melanoma Tumor Model
2.1 Abstract and Introduction .......................................................................... 56
2.2 Materials and Methods ............................................................................. 60
2.3 Results...................................................................................................... 65
2.4 Discussion and Conclusions .................................................................... 73
2.5 Figures ..................................................................................................... 78
2.6 References................................................................................................ 86
Chapter 3: mRNA Display .................................................................................... 93
Chapter 4: Identifying Peptide Inhibitors of STAT3 by mRNA Display
4.1 Introduction .............................................................................................. 96
4.2 Materials and Methods ............................................................................. 99
4.3 Results and Discussion ............................................................................ 109
4.4 Conclusions and Future Directions .......................................................... 119
4.5 Figures ..................................................................................................... 121
4.6 References ............................................................................................... 143
Chapter 5: Masking Peptides for the Tumor-Specific Activation of Chimeric
Antigen Receptors
5.1 Abstract and Introduction ......................................................................... 147
5.2 Materials and Methods ............................................................................. 150
5.3 Results and Discussion ............................................................................ 154
5.4 Conclusion and Future Directions ............................................................ 157
5.5 Figures ..................................................................................................... 159
5.6 References ............................................................................................... 164
4
Acknowledgements:
This work was made possible by both of my advisors: Professor Pin Wang and
Professor Richard Roberts. I could not have made it this far without their insightful
advice, endless patience, and willingness to explain things to me using simple, small
words. Their passion for science was infectious and I am very grateful for their roles in
facilitating my growth as a scientist. I consider myself very fortunate to have had two
great mentors and role models in my time here at USC. I am also very thankful for the
many memories that were formed, and lessons that were learned while working under
their guidance.
Furthermore, I would like to thank Professor Terry Takahashi, Dr. Farzad Jalali-Yazdi,
and Dr. Lan Huong Lai (of the Roberts lab) and Dr. Bi Liang Hu, Dr. Natnaree Siriwon,
Dr. Xiaoyang Zhang and Dr. Paul Bryson (of the Wang lab) for taking the time to teach
me the experimental and analytical skills necessary to make progress in my projects. In
my experience, experiments never work the first time you do them. My subsequent
experiments had a significantly higher, non-zero success rate largely because I have
had the privilege of receiving their advice and assistance. I also want to thank Dr. Yu
Jeong Kim and Professor Lili Yang, as well as Alireza Delfarah and Professor Nicholas
Graham for our fruitful collaborations. I would also like to thank Zhiyin Qin, Perry
Kumagai, and Anna Reich for their help in performing these experiments. I am
especially thankful for Guanmeng Wang and his help in figuring out and preparing for
our remaining experiments. Additionally, I would like to thank Daniel Linderman for both
his help in performing experiments, as well as for his role in enabling my speculation on
stocks. Because of you, my desire to gamble in Las Vegas has been replaced by the
5
more socially acceptable vice of gambling on the stock market. I would also like to thank
Zachary Dunn for helping me through all of my large scale experiments in my final year.
These experiments would not have been possible without your efforts or your friendship.
I'd like to say thank you to all of my colleagues from the Wang lab: Dr. Yarong Liu, Dr.
Liang Xiao, Dr. Chupei Zhang, Dr. Xiaolu Han, Dr. Jennifer Rohrs, Dr. Jinxu Fang, Dr.
Si Li, Dr. Elizabeth Siegler, Xianhui Chen, Gunce Cinay, Yun Qu, Melanie MacMullan,
and Fiona Guo for their continued help and support.
I also would like to thank all of my colleagues from the Roberts lab: Dr. Aaron Nichols,
Dr. Shannon Howell, Dr. Mehmet Cetin, Golnaz Kamalinia, and Christopher Hughes for
their friendship and continued support. I am particularly grateful the help of Dr. Kaori
Noridomi and William Evenson 4 for their role in demonstrating the importance of the
buddy system in both the laboratory and in lunch respectively.
Reflecting back on my time here at USC, I can only conclude that it has been a great
privilege to have been a part of both the Roberts Lab and the Wang Lab, and I am very
thankful for the relationships that exist because of it.
I want to thank my friend and mentor, Dr. Chawita Netirojjanakul, for the opportunity to
be her intern and for teaching me the skills necessary to work in industry. I consider my
summer working under your guidance to have been vital to my development as a
scientist. I am also very grateful for your willingness to play tennis with me despite my
awful stamina.
I would like to thank my family for their endless support in my endeavors. To my
brother Thomas: thank you for being by my side for my entire life and for all the times
you gave what little free time you had to help me move. Also, thank you for your
6
foresight and your explanation on how future hospital bills will outweigh any savings
from eating fast food for every meal. To both my sister Susanna and my brother-in-law
Mark: thank you both for being my role models and mentors in both science and life,
and thank you for having kids that are awesome. To my nephew Peter: thank you for
your positivity, as it seems to make the world a happier place. To my other nephew
Kevin: thank you for wishing me a happy birthday every time I see you. Your strategy
will pay off when I see you on my actual birthday. To my parents, Helen and Thac Mac,
thank you for raising me to be a responsible human being. I appreciate all the effort that
you put into ensuring that your children are in a position to succeed. I also appreciate
your unwavering love and support, as well as all the food you prepare for me in order to
prevent me from eating fast food for every meal.
Finally, I want to thank Marie Truong for her support and relentless positivity. Knowing
that you will be happy with everything makes my life much more enjoyable as a result. I
don't know what I did to deserve you, but I'm glad I did it. Your happiness is contagious
and I look forward to the years to come. Also, thank you for agreeing to meet my friends
from the Roberts lab multiple times and proving to them, once and for all, that you are
real.
7
Chapter 1: T Cell Immunotherapy Enhanced by Designer
Biomaterials
1.1: Abstract and Introduction
Abstract
Cancer immunotherapy has recently burst onto the center stage of cancer
treatment and research. T lymphocyte adoptive cellular transfer (ACT), a form of cancer
immunotherapy, has spawned unprecedented complete remissions for terminal patients
with certain leukemias and lymphomas. Unfortunately, the successes have been
overshadowed by the disappointing clinical results of ACT administered to treat solid
tumors, in addition to the toxicities associated with the treatment, a lack of efficacy in a
significant proportion of the patient population, and cancer relapse following the
treatment. Biomaterials hold the promise of addressing these shortcomings. ACT
consists of two main stages – T lymphocyte ex vivo expansion followed by reinfusion
into the patient – and biomaterials can improve the efficacy of ACT at both stages. In
this review, we highlight recent advances in the use of biomaterials for T lymphocyte
adoptive cellular cancer immunotherapy and discuss the challenges at each stage.
8
Introduction
Malignant neoplasms (cancer) remain the second leading cause of death in the
United States
1
, with 1 in 2 men and 1 in 3 women developing cancer in his or her
lifetime
2
. While the three longstanding pillars of cancer therapy – surgery, radiation, and
chemotherapy – are the basis for addressing this threat of cancer, a more recent
emerging fourth pillar is becoming the focus of oncology research
3
. That fourth pillar is
cancer immunotherapy, in which a patient’s own immune system is manipulated to fight
cancer. Thus far, cancer immunotherapy has made its most prominent strides in the form
of immune checkpoint blockades, cancer vaccines, and adoptive cellular transfer
4, 5
.
Given the recent success of T cell therapies, the scope of this review will be focused on
using biomaterials for enhanced T cell treatments.
Adoptive cellular transfer is a treatment in which a cell population is expanded ex
vivo and re-infused into the body. ACT for cancer immunotherapy typically utilizes
naturally occurring or genetically modified T cells
6
, although natural killer (NK) cells have
also been studied
7, 8
. Naturally occurring T cells that accumulate in tumors, termed
tumor infiltrating lymphocytes (TILs), have been proven to be an effective treatment for
metastatic melanoma and have shown promise for treating other cancers as well
9, 10
.
After recognizing the potential for the adoptive transfer of T cells, researchers began
genetically engineering T cells to target specific cancer markers and produce a more
durable treatment. Genetically modified T cells used for cancer immunotherapy fall into
two categories: T cells constructed to express transgenic T cell receptors (TCRs) or
chimeric antigen receptors (CARs)
11
. TCRs bind to peptide-major histocompatibility
9
complexes (MHC), which identify pathogen-derived antigens, and CARs are equipped
with a single chain variable fragment (scFv), which is an antibody fragment that can
identify and target extracellular biomarkers on a cell in an MHC-independent manner
11
.
The year 2017 marked the first FDA approvals of CAR T adoptive cellular therapies, one
for treating acute lymphocytic leukemia and a second for large diffuse Non-Hodgkin
lymphoma. This followed clinical trials in which CD19 CAR T cells resulted in complete
response rates over 80% for patients who had refractory B cell malignancies after
previously receiving two other forms of treatment
12-16
. CD19 is a biomarker present
solely on B cells
17
, which makes it the quintessential target for CAR T cells because
other cells in the body are not targeted and B cell aplasia is readily treated by
immunoglobulin injections
18
.
Despite the outburst of ACT cancer immunotherapy research and clinical trials,
the expansion of ACT beyond CAR T cell therapy for liquid cancer treatments remains in
question
19-22
. A few of the key challenges that currently prevent cancer immunotherapy
ACT from treating solid tumors are the immunosuppressive tumor microenvironment that
accompany solid tumors, the production of ample, able T cells during ex vivo expansion,
and toxicities associated with ACT. In this review we will discuss how biomaterials can
alleviate these problems by producing larger amounts of cytotoxic T cells and providing T
cells with additional tools for homing to and fighting cancer and cancer-associated cells.
In accordance with the schematic for ACT shown in Fig. 1, we will begin by discussing
biomaterials’ impact on T cell expansion methods followed by biomaterials’ potential role
in improving T cell in vivo efficacy.
10
1.2: Discussion
T cell expansion
Post leukapheresis, T cell expansion is performed to produce a sufficient number
of cells for successful treatment of a malignancy. To date, successful ACT for
immunotherapy is limited to autologous T cells to prevent graft versus host disease (in
which the body rejects the re-infused T cells that are not derived from the host)
23
. T cell
activation requires two signals: T cell receptor (TCR) activation and costimulation
24, 25
. A
third signal from a pro-survival cytokine such as Interleukin-2 (IL-2) is needed to maintain
to the expansion and differentiation of T cells
26, 27
. Antigen presenting cells (APCs),
especially dendritic cells (DC), provide these signals to T cells in the body in a specific
spatiotemporal manner
28, 29
.
Producing large quantities of T cells is not the only concern for ex vivo expansion
in cancer immunotherapy. Research has shown that the T cell type and differentiation
state upon re-infusion into the body strongly influences the efficacy of the treatment
30-34
.
CD8+ T cells, the cytotoxic T cells used to kill cancer cells in ACT, are usually composed
of multiple subsets undergoing various degrees of differentiation, including naïve,
effector, effector memory, central memory, and stem cell memory T cells, and T cells in
less differentiated states produce more durable ACT responses
6, 35-37
. Recently, Fraietta
et. al.’s characterization of T cell populations during CAR T cell infusion in patients with
chronic lymphocytic leukemia (CLL) revealed that an early memory cytotoxic T cell
population resulted in robust therapeutic response
33
. In a simple yet powerful study,
Ghassemi et. al. proved that minimally ex vivo expanded CD19 CAR T cells remain less
differentiated and exhibit improved effector function in vitro, and the in vivo efficacy of
11
CD19 CAR T cells in a murine xenograft ALL model varied inversely with expansion time
34
. Notably, despite most T cell engineering protocols requiring 9-14 days of ex vivo
expansion, CAR T cells administered to the xenograft model after 3 days of expansion
produced robust tumor control at a 6-fold lower dosage, in contrast to 9 day cultured
CAR T cells which failed to control leukemia at reduced doses. Although longer culture
periods increase cell count, “younger,” or less differentiated, T cells possess enhanced
anti-tumor capabilities. As discussed below, past expansion methods focus primarily on
the production of CD8+ cytotoxic T cells over CD4+ helper T cells, and studies centered
on expanding less differentiated CD8+ T cells will be needed to facilitate the
administration of these more potent T cells.
Natural APCs and Cellular aAPCs
Natural APCs, in the form of monocyte originated dendritic cells (moDCs), have
been harvested from patients and used to expand T cells ex vivo in a handful clinical
trials
38, 39
. Although designed by the body to activate and expand T cells, there are
several limitations to using autologous DCs for ex vivo expansion that prevent their
widespread usage. The maintenance of DCs for T cell ACT on a clinical scale requires
substantial manufacturing and labor costs
36, 40
, the functionality of DCs can be
compromised in diseased patients
41
, and DCs can cause T cell unresponsiveness
42
. In
an effort to mitigate these issues, researchers have experimented with various types of
cellular artificial antigen presenting cells (aAPCs). Cellular aAPCs, such as insect cells,
mouse fibroblasts, and human leukemic cell lines, have been genetically modified to
12
present antigen, adhesion, and costimulatory signals to T cells
43, 44
. While these
methods can achieve successful expansion of T cells, and clinical trials relying on human
leukemic cell line K562 for ex vivo T cell propagation saw significant clinical benefits of
the ACT treatment
45, 46
, cellular aAPCs run the risk of tumorigenicity and infection as a
result of the genetic modifications and, similar to natural APCs, require additional
culturing steps. But above all else, the aforementioned platforms do not allow for the
specific control of T cell signal frequency, orientation, and persistence. Engineered
biomaterials allow researchers to have full control over the spatio-temporal presentation
of signals to maximize the production of more desirable T cells
47, 48
. It is acellular aAPCs
that can create an off-the-shelf, simplified T expansion means which eliminate
supplementary culture and separation steps - streamlining T cell production.
Acellular aAPCs
Acellular (synthetic) aAPCs have become the mainstay for T cell expansion in the
clinic and laboratory. The key signaling molecules for T cell activation and proliferation
can be engineered into biomaterials of different shapes and sizes to maximize the
efficacy of T cell production, and thus far have been able to simplify, expedite, and
reduce the cost of T cell expansion. TCR agonists targeting CD3, such as anti-CD3
antibodies, can activate T cells in place of the TCR activation that naturally occurs upon
the recognition of a peptide-MHC complex
49, 50
. The second and third T cell stimulation
signals are most commonly delivered by anti-CD28 antibodies (costimulation) and
soluble IL-2 (prosurvival cytokine)
51-53
. The addition of free anti-CD3 and anti-CD28
13
antibodies for T cell culture fails to provide the simultaneous receptor stimulation
necessary for T cell expansion, which has led scientists to bind these signals to various
surfaces
54, 55
. The advent of Dynabeads in 1976, magnetizable and superparamagnetic
uniform polystyrene spherical beads of five or more microns in diameter, provided such a
surface. The signal antibodies are covalently bonded to the spherical beads, and given
their magnetic properties, the beads can be easily resuspended or removed from cell
culture
56
. The use of magnetic microspheres with conjugated anti-CD3 and anti-CD28
antibodies for ex vivo T cell expansion, with IL-2 culture supplementation, has been
utilized in clinical trials
15, 57, 58
as well as the first FDA approved CAR T cell therapy,
Novartis’ Kymriah (tisagenlecleucel)
59
. While anti-CD3/anti-CD28 Dynabeads are the
standard for translational T cell expansion, there are several limitations to this expansion
platform. Bead anti-CD3/anti-CD28 stimulation results in T cells achieving an effector
state prior to infusion resulting in diminished anti-tumor efficacy
60
, and favors the
proliferation of CD4+ helper T cells rather than tumor-killing cytotoxic CD8+ T cells
61
.
Furthermore, there is a growing body of work supporting the importance of aAPC shape
and surface mobility in optimizing T cell production for ACT
62
. The next section will
discuss alterations made to traditional, circular microbeads as well as contemporary T
cell expansion platforms like carbon nanotubes and supported lipid bilayers. While each
biomaterial platform tries to maximize T cell production in a different way, the underlying
principle remains the same – combine the three necessary T cells expansion signals. As
the complexity of the acellular aAPCs increases, it is important for researchers to be
conscientious of the final goal that is clinical application. Dynabeads are relatively cheap
and simple to produce, in addition to being highly reproducible, whereas several of the
14
more recent acellular APCs can be subject to batch to batch inconsistencies, increased
materials costs, and prolonged manufacturing procedures. With each new formulation
created it is vital that researchers provide scale up and cost analysis, and prove that the
platform warrants clinical investigation.
Advances in Acellular aAPCs
To more closely resemble natural APCs, researchers have strayed away from
Dynabeads towards synthetic aAPCs made of different sizes, shapes, and materials.
Although previous experimentation revealed that using microbeads less than 4
micrometers in diameter diminished T cell expansion
63
, Perica et. al. found that
magnetic 50-100 nm spherical nanoparticles and 30 nm nanocrystals were effective
aAPC platforms
64
. In a subsequent study, Perica et. al. proved that in the presence of
magnetic fields, nanoparticles are more advantageous for activating T cells than
microparticles
65
. This study highlighted the role of nanoscale TCR clustering in T cell
expansion – the nanoparticles allow TCR clusters to form on the T cells much like the
clusters that form when natural DCs present antigens and signals to T cells
66, 67
,
whereas microparticles are less conducive to TCR aggregation. Although the
immunological synapse formed between APCs and T cells is on the micrometer scale
68
,
the more recent discovery that TCRs on a given T cell conglomerate to form
nanoclusters for T cell activation
69, 70
can be exploited by nanoparticle aAPCs (naAPCs).
Hickey et. al. have since confirmed that in a static magnetic field 50 nm aAPCs produce
comparable CD8+ T cell expansion to microparticles without magnetic presence (and
15
showed that a magnetic field had no effect on microparticle T cell expansion), but using a
reductionist approach and further analysis concluded that 50 nm aAPCs were less
effective than larger aAPCs at stimulating T cells under subsaturated conditions
71
.
These results indicate that while nanospheres are useful for the study of TCR clustering,
they will not displace microbeads as the standard platform for T cell expansion.
It has also been shown that, while keeping particle volume and antigen content
held constant, ellipsoidal poly (lactic-co-glycolic) acid (PLGA) microbeads with high
aspect-ratio resulted in significantly enhanced in vitro and in vivo CD8+ T cell activity
compared with spherical PLGA beads
72
. Meyer et. al. found that PLGA ellipsoidal
naAPC significantly outperformed spherical naAPC for all degrees of stretch, or
elongation of a single radius (from 1.5 to 3.5-fold). For example, spherical particles
induced a 3-fold expansion of T-cells compared to a 15-fold expansion induced by
twofold stretch naAPCs
73
. Biodegradable materials, such as PLGA, hold promise as
aAPCs not only due to their structural flexibility, but also given their ability for the
paracrine delivery of soluble cues to T cells. The use of polymer aAPCs (paAPCs) for the
extended release of IL-2 in addition to activation and costimulation of T cells resulted in
enhanced CD8+ T cell proliferation by avoiding the T cell exhaustion that can be
instigated by the high concentrations of IL-2 required for polystyrene bead culture
methods
74
. paAPCs can result in a high concentration of IL-2 at the T cell activation site
despite a low overall culture concentration of IL-2. A distinct disadvantage of
biodegradable aAPCs is the loss of surface-bound molecules over time, but conversely
this can be a benefit to prevent the over stimulation of T cells – further investigation will
be needed to determine the optimal degradation rate of these platforms.
16
Carbon nanotubes and elongated “nanoworms” have emerged as two other
potent T cell expansion structures. Carbon nanotubes and flexible, water-soluble
polymer nanoworms provide increased surface area for T cell binding and the clustering
of surface receptors. In 2008, Fadel et. al. proved that carbon nanotubes decorated in T
cell signaling moieties resulted in more robust T cell stimulation than other high surface
area materials
75
, and have since matured the platform into a carbon nanotube–polymer
composite (CNP) that combines the carbon nanotube surface with magnetite-loaded
PLGA nanoparticles that allow for the paracrine delivery of IL-2
76
and the magnetic
extraction of the composite. Traditional Dynabead expansion required a 1000-fold
greater concentration of IL-2 than the CNP to produce a comparable level of cytotoxic T
cells, and the CNP produced a significantly higher percentage of less differentiated T
cells as determined by the presence of the CD27+ marker
59, 77, 78
. The nanoworm
platform consists of a poly(isocyano peptide) backbone with functionalized side chains
presenting signal molecules, and due to the semi-stiff backbone, was proven to allow
TCR clustering to occur readily and cause all the effector molecules on the nanoworm to
bind to receptors on the same cell
79
. The elongated structures mark a shift from
ellipsoidal and spherical synthetic aAPCs, and indicate an enhanced exploitation of TCR
clustering. The nanoworm and nanotube formulations use their pronounced aspect ratio
and unique nanoscale topography to promote TCR clustering, which was shown to
improve T cell expansion rates. Whereas nanospheres require saturated conditions and
the administration of magnetic fields to present T signals in a clustered presentation, the
elongated nanostructures can duplicate the T cell activation milieu inherently due to their
shape. Further investigation into the use of these nanostructures is warranted, and
17
determining the importance of backbone flexibility in enhancing T cell expansion should
be a focus.
The cell membrane, where receptor binding occurs, is composed a lipid bilayer.
Thus, to more closely resemble nature, aAPCs composed of lipid bilayers, in the form of
liposomes and 2D/3D supported lipid bilayers (SLBs), have been studied. The
concentration of MHC molecules on lipid rafts, microdomains on the plasma membrane
with high concentrations of cholesterol and glycosphingolipids
80
, facilitate antigen
presentation
81
. Raft microdomains that contained epitope/MHC complexes were isolated
from DCs and reconstituted on liposomes, named RAFTsomes, then successfully
expanded CD4+ cells and elicited a strong immune response in an OVA-expressing
E.G7 tumor model
82
. In another study, biotinylated monoclonal antibodies preclustered
in microdomains held on a liposome scaffold by neutravidin rafts improved polyclonal T
cells and MART-1-specific CD8+ T cells expansion compared to traditional microbeads
83
. Liposomes provide fluidity that is conducive for enhanced T cell expansion but lack
stability when compared to solid particles
46
and struggle to recreate the immunological
synapse, which has led to a focus on SLBs. SLBs are composed of lipid bilayers bonded
to a solid flat surface (2D) or a solid scaffold core (3D). 2D SLBs, typically utilizing glass
as the solid surface, are powerful tools for studying T cell activation and the
immunological synapse
84, 85
, but the level, continuous surface fails to recreate the
natural T cell – APC interaction resulting in subpar T cell expansion rates. This year
Cheung et. al. mimicked APCs by creating a mesoporous silica micro-rod supported fluid
lipid bilayer that presented anti-CD3 for TCR stimulation and anti-CD28 for costimulation
as membrane bound signals and allowed for the sustained release of soluble signals,
18
namely IL-2
86
. The novel 3D SLB platform, termed APC-ms for APC-mimetic scaffolds,
can be tailored to present the signals in spatiotemporal manner closely resembling
natural APCs. The APC-ms promoted up to a tenfold greater polyclonal expansion of
mouse and human T cells compared to commercially available Dynabeads, maintained a
substantial CD8+ biased skewing, and the CAR T cells expanded by APC-ms showed
high in vivo functionality in a disseminated xenograft model of Burkitt’s lymphoma.
Furthermore, whereas Dynabeads need to be removed, the evaluation of silicon content
in cell culture pellets confirmed the APC-ms degrades prior to cell re-infusion.
aAPCs are at the interface of T cell immunology and biomaterials, requiring a
deep understanding of both fields for the creation of improved T cell expansion
platforms. From the advent of Dynabead aAPCs, to naAPCs, elliposoidal aAPCS,
paAPCs, and most recently APC-ms (Fig. 2), new discoveries in T cell biology can be
incorporated into biomaterials and result in enhanced T cell expansion. Engineered
biomaterials provide investigators full autonomy over the presentation of signals to T
cells, and thus represent a promising method for achieving the T cell expansion
necessary for successful ACT immunotherapy. As research continues, a focus on
producing CD8+ T cells of a less differentiated state, rather than a sole focus on
producing CD8+ T cells over CD4+ T cells, will likely result in more effective treatments.
Research on producing younger T cells for ACT has focalized on preventing T cells
differentiation, using molecular inhibitors or cytokines such as IL-7 and IL-15
87
, but these
approaches require naïve, non-specific T cells. Emulating Yamanaka and Gurdon’s
Nobel Prize for inducing pluripotent cells from mature cells, induced stem cell memory T
cells have been generated
88
. The reversion of effector T cells provides a new
19
mechanism for achieving high expansion rates of antigen specific, stem cell memory T
cells. The next generation of aAPCs may look to combine the trend towards APC
mimicry using biomaterials with the ability to induce stem cell memory T cells. While
aAPCs have been investigated for three decades, recent studies involving the recreation
organs in vitro are providing additional pathways for expanding T cells, as briefly
reviewed next.
Organoids Supporting T Cell Expansion
Organoids, or 3D multicellular in vitro organ models, are biomaterials that present
a unique avenue for the improved expansion of immune cells, including B cells
89
and T
cells. Their three-dimensional nature has facilitated a variety of studies, including the
development of neuronal networks among brain cells, the interactions between different
types of cells
90
, such as the interaction between intraepithelial lymphocytes and
intestinal epithelial cells
91
, and the gene and pathway alterations associated with the
progression of pancreatic cancer
92
. However, in this review, we will focus on the use of
organoids to facilitate the generation and expansion of human T cells.
Organoids have also been used to assist in the modeling of the tumor
microenvironment and the study of the tumor T cell repertoire in patients. Neal et.al.
developed a method to extract patient-derived organoids (PDOs) in a manner that
preserved tumor-infiltrating immune cells. Using these PDOs, they were able to expand
and activate tumor antigen-specific TILs and demonstrate tumor cytotoxicity
93
. This
promising method could potentially facilitate the generation of patient-specific antitumor
20
T cells. Dijkstra et.al. presented an alternative, organoid-based method for the
generation of patient-derived TILs. In this report, tumor-reactive T cells were generated
from peripheral blood lymphocytes after coculture with autologous tumor organoids.
Remarkably, after two weeks of coculture, expanded T cells were shown to react solely
with the tumor organoid used in the coculture (mismatch repair-deficient colorectal
cancer), and had no reactivity towards organoids derived from another form of colorectal
cancer (mismatch repair-proficient colorectal cancer)
94
.
Recently, Montel-Hagen et.al. described a continuous 3D organoid system that
induced the differentiation of pluripotent stem cells (PSCs) into terminally differentiated
CD3+CD8αβ+ and CD3+CD4+ T cells bearing a diverse TCR repertoire (Fig. 3a)
95
. As
shown in Fig. 3b, T cells that were generated from PSCs in the presence of an artificial
thymic organoid (ATO) were shown to have a conventional T cell phenotype (CD8αβ+),
and were shown to be mature, naive T cells (CD45RA+CD45RO-). These newly
generated T cells were able to secrete TNFα, IFN-γ and IL-2 upon stimulation, as well as
proliferate in the presence of anti-CD3, anti-CD28 and IL-2 - demonstrating a functional
similarity to traditionally expanded T cells (Fig. 3c, 3d). The authors also demonstrated
that PSCs transduced with engineered TCR transgenes (specific for NY-ESO-1 ligand)
were also able to be expanded with the use of the athymic organoid, while maintaining
their functionality as well as antigen specificity (Fig. 3e).
Although further studies must be conducted to assess the feasibility of widespread
usage of organoids to generate patient-derived TILs, tumor organoids have the potential
to revolutionize the field of TIL therapy by providing the means to dramatically improve
TIL expansion ex vivo. The work of Montel-Hagen et.al. lays the foundation for the use of
21
organoids in the large scale generation of engineered, antigen-specific T cells from
PSCs – providing an alternative to traditional production methods for T cell therapies.
Whether its using organoids to expand TILs using tumor samples or PSCs, there is a
clear role for the use of organoids in the expansion of T cells for ACT. In concordance
with the use of biomaterials in T cell expansion, biomaterials have been studied for
enhancing T cell therapy in vivo. As discussed next, in the form of nanoparticles, small
molecular compounds, and scaffolds, biomaterials have the potential to improve the
treatment efficacy of T cell therapy through several avenues.
Enhancing T Cell Therapy Efficacy
One method in which biomaterials can improve T cell therapy is by allowing for
the targeted delivery of immune-modulating pharmaceuticals. Due to the often toxic
nature of cancer treatments, it is necessary to ensure the targeted delivery of therapeutic
compounds to tumor cells and avoid systemic toxicity. Furthermore, the therapeutic must
be protected from degradation, clearance, and off-target delivery prior to reaching its
intended destination. The study of nanoparticles has resulted in the development of
many safe and effective drug delivery vehicles. Their design
96
and successes
97
are
discussed in other reviews. Previous research has exploited the enhanced permeability
and retention (EPR) effect to deliver untargeted, drug-loaded nanoparticles to tumor cells
98, 99
. The EPR effect is a mechanism by which nanoparticles are able to accumulate in
tumor tissue due to the tumor's abundant vasculature and exclusion of immune cells.
Numerous groups have employed the preferential tumor targeting of the EPR effect to
22
deliver nanoparticles, including Wei et.al., who used nanomicelles to delivery doxorubicin
to breast cancer cells
100
, and Liu et.al., who developed crosslinked, multilamellar
liposomal vesicles (cMLVs) for the co-delivery of doxorubicin and paclitaxel
101
. However,
successful delivery of a single therapeutic is often insufficient due to the generation of
drug-resistant cancer cells
102
and antigen-loss variants
103, 104
. Additional steps must be
taken to ensure a lasting antitumor effect by recruiting and facilitating the help of immune
cells, and the recent advances in nanoparticle research provides a promising route for
this to occur.
Although immune cells have a vital role in killing tumor cells, additional barriers
due to the immunosuppressive tumor microenvironment (TME) can prevent immune cells
from functioning at full capacity. The TME facilitates many different avenues for the
inhibition of T cell activation including maintaining a hypoxic environment
99, 105
, and
producing excessive lactate, which can hinder export of intracellular lactate by T cells
106,
107
. Tumor cells can also employ other modes of immune suppression by recruiting
suppressive immune cells that secrete anti-inflammatory cytokines such as IL-10 and
TGF-β, and further deplete the microenvironment of metabolites necessary for T cell
function
108
. The immunosuppressive TME presents a unique obstacle that must be
considered when developing cancer immunotherapies. This section discusses the
research done to deliver small molecules, cytokines, and other immune cell function-
supporting molecules to T cells in order to circumvent the immunosuppressive TME and
promote antitumor effects. This section also reviews biomaterials that aid in the direct,
local delivery of T cells and biomaterials that work in conjunction with CAR T cells.
23
Targeted Nanoparticle Delivery
Cytokines play a vital role in T cell function. For instance, IL-2 is responsible for T
cell proliferation
20, 109
and IL-15 regulates T cell activation, proliferation, and survival
110,
111
. Although previous attempts at systemically administering IL-2 to cancer patients
resulted in the expansion of lymphoid cells, it also demonstrated rapid clearance of the
cytokine and significant side effects
112
. A later study involving the high dose
administration of IL-2 to metastatic melanoma patients achieved an objective response
rate of 16%, but still reported side effects in over half of the patients
113
. IL-2 treatment
has also been shown to increase the frequency of immunosuppressive regulatory T cells
in patients with melanoma or renal cell carcinoma
114
. Similarly, Berger et.al. showed that
IL-15 therapy, though able to expand memory CD8+ and CD4+ T cells, was still subject
to a high clearance rate and resulted in transient toxicity in nonhuman primates
115
.
These results indicate that cytokine therapies must be administered in a targeted manner
in order to limit side effects and have a higher antitumor efficacy.
The high clearance rate and complications with untargeted delivery make
cytokines ideal candidates for nanoparticle-based delivery. A study of antigen-
encapsulating nanoparticles presenting IL-15:IL-15Rα demonstrated a marked
improvement in the life spans of mice bearing OVA-expressing B16 tumors compared to
mice receiving antigen-encapsulating nanoparticles and free IL-15:IL-15Rα
116
. In a
similar manner, intratumoral injections of PEGylated liposomes with anti-CD137 and IL-
2Fc anchored to their surfaces were shown to slow the growth of B16-F10 tumors in vivo
while mitigating the toxicity associated with systemic administration of anti-CD137 and
24
IL-2
117
. In a related study, Zheng et.al. reported that intravenous injections of liposomes
with surface IL-2Fc molecules after adoptive cell transfer resulted in a fourfold
improvement in the expansion of adoptively transferred T cells
118
.
Although transforming growth factor β (TGF-β) is a cytokine that regulates cell
proliferation in healthy cells, mutations in the TGF-β pathway in cancer cells result in
uncontrolled proliferation and immune evasion
119, 120
. Intratumoral injections of
nanolipogels (nLG) delivering IL-2 and a TGF- β receptor inhibitor (SB505124), were
shown to significantly inhibit tumor growth and prolong survival in a mouse metastatic
melanoma model. Furthermore, intratumoral injections of nLG SB505124, IL-2, or both
compounds increased the amount of tumor infiltrating CD8+ T cells and natural killer
(NK) cells compared to injections with an empty nLG
121
. However, there have also been
reports of systemic administration of TGF-β receptor inhibitors resulting in cardiac toxicity
122, 123
, suggesting that TGF-β receptor inhibitors are a promising form of cancer therapy
that demonstrate a need for targeted delivery to mitigate any potential off-tumor side
effects. Zheng et.al. used liposomes targeted to CD90, an internalizing receptor on T
cells, to deliver a TGF-β receptor inhibitor (SB525334) to adoptively transferred T cells.
The targeted liposomes were shown to inhibit tumor growth compared to nontargeted
liposomes. Interestingly, liposomes targeted to CD45, a noninternalizing T cell receptor,
did not inhibit tumor growth as well as those liposomes targeted to CD90 - suggesting
that targeting an internalizing receptor can result in the enhanced uptake of delivered
drugs
124
. Schmid and coworkers employed a similar strategy, using PLGA and
polyethylene glycol (PLGA/PEG) nanoparticles to deliver a TGF-β receptor inhibitor (SD-
208) to exhausted T cells expressing programmed cell death protein 1 (PD-1). Treatment
25
of C57BL/6 mice bearing MC38 tumors with PD-1 targeted nanoparticles containing SD-
208 was shown to significantly inhibit tumor growth and prolong overall survival
compared to treatment with untargeted SD-208-loaded nanoparticles
125
.
The use of nanoparticles to support T cell function is not limited to therapies
directed towards cytokine signaling. Targeting cells that are responsible for maintaining
the immunosuppressive tumor microenvironment has also been an active area of
research. For instance, tumor associated macrophages (TAMs) are known to inhibit T
cell function
126
as well as contribute to angiogenesis and tumor invasion
127-129
. To
combat TAMs, Huang et.al. used PEG-hisitidine-modified alginate (PHA), a pH-sensitive
material, to encapsulate CpG oligodeoxynucleotides (ODN), anti-IL-10 ODN, and anti-IL-
10 receptor ODN. TAM targeting was achieved by associating ODNs with galactosylated
cationic dextran, which is known to bind to Mg1 - a lectin that is highly expressed on
TAMs. This strategy proved to be successful by reducing the expression of IL-10
receptor threefold, changing the TAM phenotype, and achieving a twofold inhibition of
tumor growth in vivo
130
. Regulatory T cells (Treg) are vital to the prevention of
autoimmune disease, but are also another type of immune cell that inhibits T cell function
and facilitates tumor evasion
131, 132
. In a recent study, Ou and coworkers targeted Treg
cells with imatinib-loaded nanoparticles decorated with a peptide targeting motif. They
observed that delivering imatinib to Treg cells resulted in enhanced tumor inhibition and
the increased presence of intratumoral CD8+ T cells
133
.
An alternative strategy to elicit T cell-mediated antitumor responses centers on
the activation of T cells in a cytokine-free manner. T cell activation requires antigen
binding to the T cell receptor (TCR), and a simultaneous costimulatory signal, typically
26
from binding to the CD28 or 4-1BB (CD137) receptors on the T cell. However, activated
T cells can become exhausted upon its PD-1 receptor binding to programmed death-
ligand 1 (PD-L1)
134, 135
. Although PD-L1 expression on healthy cells prevents
autoimmune diseases, its expression on cancer cells results in the evasion of tumor
specific T cells. A recent study described the use of nanoparticles coated with two
different antibodies that are specific for 4-1BB and PD-L1 (immunoswitches), resulting in
T cell stimulation while simultaneously preventing PD-1-mediated T cell exhaustion.
Intratumoral injection of immunoswitches were also found to inhibit the growth of B16-
SIY tumors in C57BL/6 mice, both in the presence and absence of adoptively transferred
CD8+ T cells
136
. Another approach involved redirecting CTLs towards tumors cells with
antigen-specific T cell redirectors (ATR) - nanoparticles decorated with moieties binding
to antigen-specific T cells and tumor cells. In this study, Schütz et.al. used either peptide-
loaded MHC complexes or a clonotypic anti-T cell receptor (TCR) antibody as the T cell
binding moiety, and anti-human CD19 antibody as the tumor binding moiety and showed
that ATRs could effectively facilitate the killing of CD19+ Raji tumor cells by T cells
lacking anti-CD19 TCRs, both in vitro and in vivo
137
.
Targeted delivery of therapeutics to tumors has been shown to facilitate the
"reawakening" of T cells within the tumor microenvironment. Nanoparticles can help
increase the presence of pro-immune cytokines and molecules within the tumor by
protecting them from degradation and clearance during the delivery process. Currently,
nanoparticle research has shifted its sights from creating protective nanoparticles to
developing innovative methods to direct nanoparticles and their payloads to specific cells
within the tumor microenvironment in a targeted fashion.
27
Backpacking Nanoparticles
The concept of backpacking involves conjugating nanoparticles to immune cells in
order to promote the successful delivery of a therapeutic to the "backpacked" immune
cell in a pseudoautocrine manner. This strategy utilizes the abundance of free thiol
groups on the cell surface as a means to conjugate nanoparticles bearing surface
maleimide groups. It was first developed by the Irvine group to deliver adjuvants to cells
commonly used in cell therapy, including CD8+ T cells and hematopoietic stem cells
138
,
but this method has since been adapted to attaching differing payloads to immune cells.
For instance, NK cells with a chimeric antigen receptor (CAR) targeting CD19
have been used to deliver cMLVs containing paclitaxel (cMLV PTX), a chemotherapeutic
agent, to SKOV3 CD19 tumors in NSG mice. In this study, Siegler et.al. demonstrated
the importance of conjugating cMLV (PTX) to the CAR NK cells by showing that mice
receiving coadministered cMLV (PTX) and CAR NK cells had significantly larger tumor
volumes than those receiving cMLV (PTX) directly conjugated to CAR NK cells.
Furthermore, they observed that the targeted nature of PTX delivery allowed for a
decreased dose of PTX and by extension, minimal side effects associated with PTX
therapy
139
. Similarly, Siriwon and coworkers conjugated cMLV containing SCH (cMLV
SCH), an A2a adenosine receptor antagonist, to CD19-targeting CAR T cells. Again, the
antitumor efficacy of cMLV SCH conjugated to CAR T cells was shown to be
substantially higher than coadministration of cMLV SCH and CAR T cells in NSG mice
with SKOV3 CD19 tumors. Additionally, conjugation of cMLV SCH to CAR T cells was
shown to significantly reduce intratumoral CAR T cell hypofunction and nearly double the
28
amount of tumor-infiltrating CAR T cells compared to the unconjugated combination
therapy
140
.
Similar to paclitaxel, SN-38 is a chemotherapeutic agent that is known to have
poor pharmacokinetic properties. The conjugation of SN-38 - loaded nanocapsules to T
cells was shown to promote SN-38 mediated killing of lymphoma cells in vitro and in
vivo. This delivery strategy exploited the innate ability of T cells to traffic to lymphoid
organs to target disseminated lymphoma tumors. Compared to systemically
administered SN-38, cell-mediated delivery of SN-38-loaded nanocapsules was reported
to increase the concentration of SN-38 in lymph nodes by 90-fold, and could successfully
prolong the survival of C57BL/6 mice with disseminated lymphoma tumors
141
. Jones et.
al. showed that conjugating drug-loaded lipid nanoparticles to cytotoxic T lymphocytes
(CTL) allowed the contents to be delivered to cells targeted by the CTLs in vivo. Drug
release was triggered by the release of perforins from CTL activation, which was shown
to result in the killing of targeted cells
142
.
Researchers have also used nanoparticles to backpack molecules that can
promote T cell function. For example, Stephan et. al. showed that the conjugation of
nanoparticles loaded with an inhibitor of Shp1 and Shp2 to T cells prior to adoptive
transfer resulted in a fourfold enhancement of T cell expansion at the tumor site
143
. In
another study, Tang et.al. employed this backpacking strategy to conjugate and deliver
protein nanogels containing IL-15 super-agonist (IL-15 Sa) complex to T cells. These
nanogels relied upon the increase in T cell surface reduction potential following antigen
binding to release their cargo. This novel release mechanism allowed for the transport of
protein therapeutics to the tumor microenvironment and for antigen recognition to trigger
29
cargo release, resulting in a 16-fold higher expansion of T cells in tumors compared to
treatment with soluble IL-15Sa, and the ability to deliver eightfold higher doses of
cytokines, while avoiding toxicity associated with systemic administration. Additionally,
compared to CAR T cell therapy alone, IL-15 Sa nanogels were able to substantially
improve the efficacy of EGFR-targeting CAR T cells and improve the survival of NSG
mice with adoptively transferred human T cells and U-87 MG tumors
144
.
Backpacking has the potential to provide even more control over the release of
the drug payload by the inclusion of stimuli-responsive nanoparticles, such as light-
responsive or temperature-responsive nanoparticles. However, additional research must
be conducted in order to evaluate their efficacy. Nonetheless, backpacking is a unique
workaround that bypasses the need to provide a targeting moiety to nanoparticles, while
still maintaining their ability to protect the drug payload. This promising delivery method
has the potential to enhance the efficacy of adoptive cellular therapies by providing the
stimuli needed to promote immune cell activation and proliferation, while counteracting
immune suppressive signals in the tumor microenvironment.
Biomaterials delivering molecular immunotherapies
Biocompatible scaffolds have also been investigated as a drug delivery vehicle.
Current T cell therapies require the large bolus systemic administration of cytokine IL-2,
which can cause cytokine release syndrome (CRS) and high toxicity in patients
145, 146
.
As a result, many groups have developed biocompatible scaffolds for the sustained
release of molecules such as IL-2 to enhance the effects of immunotherapy.
30
Injectable hydrogels have been used to deliver cytokines to tumor sites. A study
used alginate matrices to deliver IL-15Sa to B16-OVA tumors and showed that
peritumoral injections of IL-15Sa-carrying gels achieved about 40-fold higher
concentrations of IL-15Sa in the tumor site than systemically administered IL-15Sa and
suppressed tumor growth
147
. In an orthotopic bladder cancer mouse model, intravesical
treatment with chitosan/IL-12 (IL-12 coformulated with chitosan) cured 88% to 100% of
mice, compared to 38% to 60% of mice treated with IL-12 alone. Mice treated with
chitosan/IL-12 were also shown to have durable protection upon intravesical tumor
rechallenge
148
. Hydrogels have also been used to prevent tumor recurrence at tumor
resection sites. In a recent report, Park et.al. used hydrogels to deliver either R848, a
Toll-like receptor 7/8 agonist, or STING-RR, a stimulator of interferon genes (STING)
agonist, to tumor resection sites and demonstrated the importance of hydrogel delivery.
A majority of the mice receiving hydrogel-delivered R848 or STING-RR (60% - 70%)
survived beyond 90 days, while mice receiving local delivery of either agonist had a
median survival of under 60 days
149
.
Biomaterials have also been used to deliver cytokines for the purpose of attracting
host immune cells. The sustained release of granulocyte-macrophage colony-stimulating
factor (GM-CSF) from gelatin cryogels implanted into C57BL/6J mice was shown to
attract about 20 times more immune cells than a blank cryogel
150
. Another study
reported using a polyglyconate/gelatin scaffold to deliver CCL17, a chemoattractant for
CCR4+ CD8+ T cells, to pancreatic cancer cells in vivo. The CCL17-eluting scaffold
inhibited tumor growth and prevented the metastasis of pancreatic cancer cells to the
31
liver, while the non-eluting CCL17 scaffold was shown to be nonfunctional in both of
these regards
151
.
Researchers are constantly striving to develop novel biomaterials with expanded
drug delivery capabilities. Wen et. al. describe the use of EAK16-II peptides to generate
an in situ-forming system capable of enhancing the local retention of antibodies and
delaying antibody clearance. Structure assembly begins when the EAK16-II and EAKH6
peptides self-assemble into beta sheet bilayers. Upon beta sheet formation, the histidine
tags on the EAKH6 peptides are exposed and recognized by an anti-histidine tag
antibody, which is also recognized by Protein A/G. The avidity of Protein A/G allows for
the binding and retention of any nearby antibody that the structure encounters. The self-
assembled system was shown to improve the in vivo retention of intratumorally injected,
dye-labeled antibodies by at least three days - demonstrating the potential benefits of
this system when used in conjunction with checkpoint inhibitor therapies like anti-PD-1
and anti-CTLA-4
152
. A recent study described the use of βTail tags as a potential
material for the delivery of therapeutic proteins at precise ratios. Fusion proteins
containing βtails were shown to coassemble into fibrils when incubated with beta sheet
fibrillizing peptides. Using fluorescent proteins, Hudalla et. al. showed that this
preparation method allows for the precise control of the ratio of βTail-tagged therapeutic
proteins delivered by each fibril, which could potentially mitigate the side effects
associated with high dosages of molecular immunotherapies
153
.
As discussed below, biomaterials can also be used to deliver engineered cells to
support cancer immunotherapy. Aliperta et.al. used star-shaped PEG-heparin cryogels to
deliver engineered mesenchymal stromal cells (MSCs) that secrete an anti-CD33/anti-
32
CD3 bispecific antibody for the treatment for acute myeloid leukemia (AML). This
strategy was shown to effectively activate T cell-mediated antitumor responses by
activating T cells and bringing them into contact with CD33+ AML blasts
154
.
Biomaterial scaffolds delivering T cells
Biomaterials in the form of scaffolds can be used to deliver not only T cell
enhancing drugs but also adoptive T cells themselves. Scaffold can provide a means for
increasing the efficacy and longevity of T cell therapies while diminishing toxicity
associated with the large intravenous dosage of T cells as well as off-tumor reactivity.
Similar to synthetic aAPCS, biomaterial scaffolds for enhanced T cell therapy allow the
controlled delivery of soluble cues to a target area.
A broad range of materials can be used to create ACT enhancing scaffolds, and
each formulation has its own biocompatibility, biodegradability, structural integrity,
porosity, and rigidity – factors that play essential roles in the activation and structural
support of T cells
155
. Thus far the investigated T cell depots for cancer immunotherapy
are hydrogels composed of chitosan, polymerized alginate, or hyaluronic acid. Hydrogels
are three-dimensional, cross-linked structures of hydrophilic polymers swollen in water or
biological fluids
156
, and have received considerable attention over the past 50 years for
application in a wide range of fields
157
. The past two decades have seen a rise in the
use of hydrogels for immunomodulation, with a focus on enhancing cancer vaccines
158
,
and are only more recently under investigation for the improvement of T cell therapies. In
2014, Tsao et. al. created a biodegradable, thermal reversible hydrogel made of
33
poly(ethylene glycol)-g-chitosan (PCgel), which was proven to be a suitable reservoir for
the harbor and release of T lymphocytes for brain tumor immunotherapy
159
. PCgel is a
liquid at low temperatures and forms a gel at body temperature, allowing for the steady
release of viable T lymphocytes to a brain tumor without surgical intervention. Anti-EGFR
(epidermal growth factor receptor) CAR T cells released from the PCgel retained
increased antiglioblastoma activity compared those delivered in Matrigel, and PCgel is
clinically preferable over Matrigel and other animal-sourced gels because of its
biocompatibility, biodegradability, low immunogenicity, and low cost. Despite the
promising preliminary data, PCgel for T cell immunotherapy has yet to be verified in a
preclinical model. Two years following the work done by Tsao et. al., a similar chitosan-
based thermogel was investigated for the in situ encapsulation and gradual release of
cytotoxic T lymphocytes
160
(Fig 4a). The chitosan, sodium hydrogen carbonate, and
phosphate buffer thermogel has highly desirable mechanical and biocompatibility
properties for the local administration of T lymphocytes through a catheter or needle. The
optimized gel solution solidifies rapidly upon injection into mice, and in vitro studies
confirmed the sustained migration of cancer fighting T cells from the thermogel.
Stephan et. al. created a macroporous scaffold from polymerized alginate that
binds T cells via a synthetic collagen-mimetic peptide and contains porous silica
microparticles in the scaffold void spaces
161
(Fig 4b). The microparticles, similar to the
liposomes used as aAPCs, provide for the encapsulation and sustained release of
soluble biomolecules, and can provide membrane-bound signals mimicking nature when
coated with lipid bilayers. In their study, interleukin 15 superagonist was used as the
soluble factor and the membrane-bound ligands were anti-CD3, anti-CD28 and anti-
34
CD137 antibodies coupled to the microsphere bilayer. Integrating the microparticles into
the modified alginate scaffold resulted in a 22-fold jump in T-cell proliferation and a 8.3-
fold increase in T cell migration into the surrounding collagen gel, and additional testing
proved the scaffold-released T cells have improved antitumor cytotoxicity and reduced
apoptotic susceptibility. The efficacy of the T cell infused scaffolds were tested in a 4T1
mouse breast tumor resection model and an ovarian carcinoma unresectable tumor
model. The staggering results revealed that in the 4T1 model, when compared to T cells
injected intravenously or directly into the resection bed as well T cells expanded ex vivo
with IL-15 superagonist and anti-CD3, anti-CD28 and anti-CD137 antibodies before local
injection, none of the mice receiving the T cell bearing scaffolds experienced relapse
whereas relapse and subsequent death occurred in all other mice. Scaffold-delivered T
cells proliferated to numbers 167-fold higher than locally injected, prestimulated T cells in
the tumor resection bed, and despite the high expansion rates maintained a
nonexhausted phenotype. Furthermore, the scaffold-delivered T cells migrated to tumor-
draining lymph nodes (TDLNs) in significantly higher levels and took on a central
memory phenotype in the TDLNs. In the unresectable tumor model, the scaffold-
delivered T cells significantly increased survival time, with the disseminated cancer
eradicated in 6 of the 10 mice and the other 4 mice averaging 27 days longer survival.
Further studies proved the CAR T cells migrate from the scaffold and eliminate tumors
more efficiently than systemic delivery of the same cells in pancreatic and melanoma
models, and the codelivery of additional agonists from the scaffold improves immune
response
162
.
35
Recently, a hyaluronic acid-based low viscosity hydrogel (LVHydrogel) was
developed as a novel carrier for convection enhanced delivery (CED) of CAR T cells
163
.
Although the blood-brain-barrier limits the accessibility of pharmacological agents to the
central nervous system from the bloodstream
164
, CED uses a needle/catheter coupled
with a positive pressure pump to provide the steady flow of infusates directly into the
brain parenchyma
165
and results in significantly (up to 8 times) greater distribution of an
agent than delivery from a single direct intracranial injection
166
. Numerous studies and
clinical trials have been performed using CED and a variety of therapeutics to treat
central nervous diseases
167
, but when tasked with dispatching cells, CED has a low
efficacy, largely due to the sedimentation of the cells in
traditional CED carrier (phosphate buffer saline)
163
. LVHydrogel was demonstrated to be
a viable option for the CED of CAR T cells for glioblastoma immunotherapy, as it did not
cause acute toxicity in preclinical mouse models, prevented cellular sedimentation during
CED, and enhanced post-infusion migration and cytotoxicity profiles of the CAR T cells.
These studies have shown that scaffolds possess the unique ability to store,
deliver, and stimulate cancer killing T cells. The scaffolds are made with FDA-approved
materials, exhibiting minimal immunogenicity and high biodegradability. The injectable
thermosensitive hydrogels are attractive due to their ease of administration but research
into the preclinical efficacy of the T lymphocytes in tumor models is necessary to move
forward. Furthermore, the addition of T cell stimulators to the gels should be investigated
to provide a platform that not only delivers T cells but ensures their proliferation. Stephan
et. al.’s surgically implanted scaffold has shown outstanding preclinical success, and can
now be further modified to obtain the optimal scaffold formulation. It is possible that
36
including the paracrine delivery of anti-PD-1 and anti-CTLA-4 antibodies can improve the
therapy, and a deeper understanding of T cell migration can dictate the placement of the
adhesion molecules. The scaffolds will need to be tested on a larger scale for toxicity
and potency as humans typically require a 1000-fold increase in the number of T cells
administered – although the scaffolds demonstrate no toxicity in mice, the increase in
size for human application may invoke serious side effects.
CAR T Cell Therapy
T cell-mediated killing of tumor cells requires several external stimuli that are often
hindered by cancer cells and the tumor microenvironment: recognition of a tumor
antigen, activation of costimulatory receptors, and cytokine signals promoting the
expansion and continued antitumor effects of the T cell. CAR T cells express an
engineered receptor that is activated upon antigen binding - allowing for T cell activation
in the absence of a second signal from costimulatory receptors.
Although CAR T therapy has proven to be remarkably successful at treating
hematological malignancies
13, 168, 169
, there are several logistical considerations that
must be addressed in order to successfully treat patients. CAR T cells must be
generated from a patient's own immune cells in order to avoid graft-versus-host disease.
Currently, CAR T cell production requires harvesting T cells from patients, shipping them
to a facility to transduce the patient's T cells with the CAR gene, expanding the newly-
generated CAR T cells, and shipping the newly-generated, patient-specific CAR T cells
back to the hospital for reinfusion into the patient. The need to generate personalized
37
CAR T cells significantly increases the cost of therapy, which is subsequently passed on
to patients. Recently, Smith et.al. developed a method to generate leukemia-targeting T
cells in situ by using synthetic DNA nanocarriers with a CD3ε-targeting antibody
fragment to deliver CAR DNA to circulating T cells - potentially consolidating the steps of
CAR T cell generation into a single injection
170
. The in situ generated CAR T cells were
found to prolong the survival of C56BL/6 mice with systemically injected Eμ-ALL01
leukaemia cells as much as an infusion of CAR T cells that were generated ex vivo.
CAR T cell therapy comes with several significant limitations that could potentially
be addressed by biomaterials, including on-target, off-tumor effects, patient toxicity, and
decreased functionality due to the tumor microenvironment. In recent years, the profound
success of CAR T cells has inspired further research into improving the safety and
efficacy of CAR T cells. Previous research has shown that proteases such as cathepsin
B
171
, urokinase-type plasminogen activator
172
, legumain
173
, and different matrix
metalloproteinases are prevalent in the tumor microenvironment
174
- providing a
potential avenue for the activation of CAR T cells. To minimize on-target, off-tumor
effects, Desnoyers et.al. created a protease-activated "probody" by modifying
Cetuximab, an EGFR-binding antibody, to include a peptide linker susceptible to tumor-
associated proteases and a peptide "mask" to prevent off-tumor binding to EGFR. The
probody was shown to be relatively inactive in nonhuman primates, but could effectively
target EGFR+ tumor cells in a mouse xenograft model
175
. Han. et. al. adapted this
discovery and created a masked CAR T cell targeting EGFR, which was shown to be
inactive among tumor-associated protease negative cells and active among cells
expressing tumor-associated proteases. These masked CAR T cells also demonstrated
38
a level of in vivo antitumor efficacy similar to unmasked CAR T cells
176
. Other studies
have focused on modifying CAR T cells to secrete various agents that can improve their
function within the tumor microenvironment, including cytokines, such as IL-2
177
, IL-7
178
,
IL-15
179
, or IL-18
180
, and checkpoint inhibitors such as single chain antibodies targeting
PD-1
181
.
Researchers have also reported using liposomes to make the tumor microenvironment
more hospitable to CAR T cells. Zhang et.al. used liposomes with the tumor-targeting
iRGD peptide to encapsulate PI-3065, a P110δ PI3K kinase inhibitor, and DW8-5, an
immunostimulant-invariant natural killer T (iNKT) cell agonist, and demonstrate the
importance of preconditioning the tumor microenvironment prior to CAR T cell therapy.
Treating mice bearing 4T1 tumors with liposomes encapsulating PI-3065 and 7DW8-5
was shown to decrease the presence of monocytic myeloid derived suppressor cells and
TAMs threefold and sevenfold respectively, while increasing the amount of tumor-
infiltrating CD8+ T cells and iNKT cells fivefold and 20-fold respectively. Additionally, the
study reported that treatment with drug-loaded liposomes prior to anti-ROR1 CAR T cell
infusion effectively doubled the mediansurvival of BALB/cJ mice bearing 4T1-ROR1
tumors when compared to mice receiving drug-loaded liposomes after anti-ROR1 CAR T
cell infusion
182
.
Although the remarkable success of CAR T cell therapy is promising, there are
still several challenges that must be addressed. Their high cost, potential for on target,
off tumor effects, and susceptibility to inhibition by the tumor microenvironment are all
obstacles that may be overcome by additional research. As we continue to make new
39
discoveries and develop novel biomaterials, the field of CAR T cell therapy continues to
advance towards unlocking its true potential.
1.3: Conclusion
Immuno-oncology, in particular T cell therapies, has undergone tremendous
advancements in the past decade. Although significant process has been made,
remissions are only obtained in certain patients and are often short-lived.
Tisagenlecleucel, originally FDA approved to treat ALL, underwent further clinical trials
and has received approval for relapsed or refractory Diffuse Large B-Cell Lymphoma
treatment (DLBCL)
183
. This followed a tisagenlecleucel DLBCL study generating an
objective response rate (ORR) above 54%. Despite achieving an ORR of more than
double other treatments
184, 185
, the median overall survival (OS) was 11.1 months and
the OS probability at 18 months was 43%. T cell immunotherapy is altering the
oncological landscape, but we are far from our goal of curing cancer. Researchers
around the world are working to engineer an improved T cell – one that can successfully
home to cancer cells, proliferate inside the patient’s body, overcome the
immunosuppressive tumor microenvironment, and minimize side effects. T cells have
progressed through 1st, 2nd, 3rd, and now 4th generations, and a few of the
advancements can be seen through the introductory of suicide genes
186
, tandem
targeting moieties
187
, masking peptides
176
, secretory abilities
188
, and altered
receptor/gene expression
189
. While each of these developments holds promise in
40
enhancing T cell therapy, it is vital to explore solutions from a different angle. A shift from
a focus on T cells themselves has resulted in a wave of research on engineered
biomaterials for improved T cell therapy. Biomaterials were first used for the ex vivo
expansion of T cells, providing an off-the-shelf, simplified method for rapidly producing
large quantities of tumor fighting lymphocytes. While spherical microbeads coated with
anti-CD3 and anti-CD28 antibodies are the gold standard for clinical T cell expansion,
recent works have proven that manipulating synthetic aAPCs to mimic nature have
resulted in a higher quantity of higher quality T cells. As knowledge of T cell biology,
activation, and differentiation expands, the design of improved aAPCs will follow shortly
after. The discovery of TCR clustering has paved the path for the optimal placement of
signal antigens, and research into T cell development will allow less differentiated T cells
to be produced and maintained. The carbon nanotubes, nanoworms, and APC-ms
platforms have been proven to enhance T cell production, and warrant further
investigation, but it is also important to explore the reversion of differentiated T cells to
earlier states as an approach for achieving maximum cell expansion of younger, antigen-
specific T cells, and to see if this approach can be used in conjugation with these novel
platforms. In parallel to the immunotherapy buzz, research on biomaterials and
nanotechnology in medicine has experienced explosive growth
190, 191
. Despite their
proven effectiveness, the efficacy and accessibility of T cell immunotherapies still has the
opportunity for improvement. There is a constant need for improved in vitro T cell
expansion - especially given the recent interest in the generation of allogeneic, off-the-
shelf CAR T cells. Once infused into patients, the T cells must be able to efficiently traffic
to their intended destination, and proliferate and thrive in a hostile environment.
41
Researchers have recognized the potential of using biomaterials to address these
challenges and contribute to the improvement of immunotherapy in its entirety.
Nanoparticles and scaffolds have been proven to enhance T cell homing to cancer cells,
cytotoxic capacities, and T cell persistence in preclinical models – the next step is to
diminish the disparity between biomaterial publications and clinical impact
191
. Although
the biomaterials discussed in this review need to be tested in additional preclinical
models and scaled up to confirm their viability as therapeutics, they show great promise
in improving the efficacy of immunotherapy.
42
1.4: Figures
Figure 1. The Process of Adoptive T Cell Immunotherapy. T cells are harvested either from tumor (tumor-
infiltrating lymphocytes, TILs) or peripheral blood (peripheral blood lymphocytes, PBLs). TILs can be
expanded non-specifically since they are preferentially tumor-specific prior to culture. In contrast, tumor
specificity must be induced in PBLs, either through antigen-specific expansion or genetic engineering.
After several weeks of expansion in culture, tumor-specific T cells can be reinfused into the cancer patient.
Fig. 1 is reprinted with permission from Adoptive T cell immunotherapy for cancer. Perica K, Varela JC,
Oelke M, Schneck J. Rambam Maimonides medical journal. 2015;6:e0004.
43
Figure 2. Summary of antigen presentation platforms. APCs= antigen presenting cells; aAPCs = artificial
APCs; moDCs = monocyte originated dendritic cells; HLCs = human leukemic cells; SLBs = supported
lipid bilayers
44
Figure 3. Artificial thymic organoids for the differentiation of naive or engineered T cells from PSCs. a)
Schematic of workflow for the differentiation of polyclonal or engineered T cells using artificial thymic
organoids. b) Flow cytometry plots showing the differences between T cells from the thymus and T cells
differentiated from PSCs after six weeks of culture in ATOs. c) Flow cytometry plots demonstrating the
newly generated T cells’ ability to produce interferon γ, TNFα, and IL-2 upon stimulation. d) Flow cytometry
plots demonstrating the newly generated T cells’ ability to expand upon stimulation with anti-CD3 and anti-
CD28 in the presence of IL-2. e) Flow cytometry plots showing antigen-specific activation of the newly
generated T cells. a)-e) are reprinted with permission from Organoid-Induced Differentiation of
Conventional T Cells from Human Pluripotent Stem Cells, Montel-Hagen, A., Seet, C.S., Li, S., Chick, B.,
Zhu, Y., Chang, P., Tsai, S., Sun, V., Lopez, S., Chen, H-C., He, C., Chin, C.J., Casero, D., Crooks, G.M.,
Cell stem cell. 2019;24:376-89 e8.
45
Figure 4. Scaffolds delivering T cells for enhanced ACT therapy. a) Monette et. al. combined chitosan,
gelling agent (sodium hydrogen carbonate in phosphate buffer), and cell culture medium to create an
injectable chitosan-based thermogel (CTGel) that maintains the extended release of cytotoxic T
lymphocytes. At room temperature the components mix in solution, and upon injection into the thermogel
solidifies, creating a readily administered T cell reservoir. In vitro studies verified the gradual discharge of
desirable T cells from the CTGel. b) T cells are seeded onto Stephen et. al.’s polymerized alginate
macroporous scaffold, which are surgically implanted into a tumor resection bed or near an inoperable
tumor site. A synthetic collagen-mimetic peptide (CMP) is used to bind T cells to the scaffold, and porous
silica microparticles presenting membrane-bound ligands anti-CD3, anti-CD28 and anti-CD137 and loaded
with soluble interleukin-15 superagonist (IL-15Rα) are added into the scaffold void space. Degradation of
the scaffold results in the sustained release and enhanced proliferation of robust antitumor T cells.
46
Table 1. Methods for Enhancing T Cell Therapy Efficacy
Method Biomaterial Reference
Targeted nanoparticle
delivery
Antigen-encapsulating nanoparticles presenting IL-15:IL-15Rα [116]
PEGylated liposomes with anti-CD137 and/or IL-2Fc anchored
to their surfaces
[117, 118]
Nanolipogels delivering IL-2 and a TGF- β receptor inhibitor [121]
Liposomes targeted to CD90, an internalizing receptor on T
cells, to deliver a TGF-β receptor inhibitor
[124]
PLGA/PEG nanoparticles to deliver a TGF-β receptor inhibitor
(SD-208) to exhausted T cells expressing PD-1
[125]
PEG-histidine-modified alginate to encapsulate CpG
oligodeoxynucleotides (ODN), anti-IL-10 ODN, and anti-IL-10
receptor ODN targeting TAMs
[130]
Imatinib-loaded nanoparticles decorated with a peptide
targeting Treg cells
[133]
Backpacking
nanoparticles
cMLVs containing paclitaxel (chemotherapeutic) or SCH (an
A2a adenosine receptor antagonist) conjugated to NK cells
[139, 140]
SN-38 - loaded nanocapsules conjugated to T cells [141]
Nanoparticles loaded with an inhibitor of Shp1 and Shp2
conjugated to T cells
[143]
Protein nanogels containing IL-15 super-agonist (IL-15 Sa)
complex conjugated to T cells
[144]
Biomaterials delivering
molecular
immunotherapies
Alginate matrices to deliver IL-15Sa [147]
Chitosan coformulated with IL-12 for IL-12 delivery [148]
Hydrogels to deliver either R848, a Toll-like receptor 7/8
agonist, or STING-RR
[149]
Gelatin cryogel delivering granulocyte-macrophage colony-
stimulating factor
[150]
Polyglyconate/gelatin scaffold to deliver CCL17 [151]
PEG-heparin cryogels to deliver engineered mesenchymal
stromal cells (MSCs) that secrete an anti-CD33-anti-CD3
bispecific antibody
[154]
Biomaterial scaffolds
delivering T cells
Thermoreversible PEG-g-chitosan hydrogel [159]
Chitosan, sodium hydrogen carbonate, and phosphate buffer
thermogel
[160]
Polymerized alginate macroporous scaffold with synthetic
collagen-mimetic peptide and porous silica microparticles
[161]
Hyaluronic acid-based low viscosity hydrogel [163]
CAR T Cell Therapy
Synthetic DNA nanocarriers with a CD3ε-targeting antibody
fragment to deliver CAR DNA to circulating T cells
[170]
Protease activated probody to mask CAR T cells and prevent
on-target off-tumor effects
[176]
CAR T cells engineered to secrete cytokines [177-180]
Liposomes to make the tumor microenvironment more
hospitable to CAR T cells
[182]
47
1.5: References
1. Heron, M. Deaths: Leading Causes for 2016. National Vital Statistics Reports 67, 1-76
(2018).
2. Siegel, R. et al. Cancer treatment and survivorship statistics, 2012. CA Cancer J Clin 62,
220-41 (2012).
3. Kim, H.S., Chapiro, J. & Geschwind, J.H. Interventional Oncology: The Fourth Pillar of
Oncology. Cancer J 22, 363-364 (2016).
4. Mellman, I., Coukos, G. & Dranoff, G. Cancer immunotherapy comes of age. Nature
480, 480-9 (2011).
5. Pardoll, D.M. The blockade of immune checkpoints in cancer immunotherapy. Nat Rev
Cancer 12, 252-64 (2012).
6. Restifo, N.P., Dudley, M.E. & Rosenberg, S.A. Adoptive immunotherapy for cancer:
harnessing the T cell response. Nat Rev Immunol 12, 269-81 (2012).
7. Knorr, D.A., Bachanova, V., Verneris, M.R. & Miller, J.S. Clinical utility of natural killer
cells in cancer therapy and transplantation. Semin Immunol 26, 161-72 (2014).
8. Miller, J.S. et al. Successful adoptive transfer and in vivo expansion of human
haploidentical NK cells in patients with cancer. Blood 105, 3051-7 (2005).
9. Rosenberg, S.A. et al. Use of tumor-infiltrating lymphocytes and interleukin-2 in the
immunotherapy of patients with metastatic melanoma. N Engl J Med 319, 1676-80
(1988).
10. Geukes Foppen, M.H., Donia, M., Svane, I.M. & Haanen, J.B. Tumor-infiltrating
lymphocytes for the treatment of metastatic cancer. Mol Oncol 9, 1918-35 (2015).
11. Carl H. June, O’Connor, R.S., Kawalekar, O.U., Ghassemi, S. & Milone, M.C. CAR T cell
immunotherapy for human cancer. Science 359, 1361-1365 (2018).
48
12. Brentjens, R.J. et al. CD19-Targeted T Cells Rapidly Induce Molecular Remissions in
Adults with Chemotherapy-Refractory Acute Lymphoblastic Leukemia. Sci Transl Med 5
(2013).
13. Grupp, S.A. et al. Chimeric antigen receptor-modified T cells for acute lymphoid
leukemia. N Engl J Med 368, 1509-1518 (2013).
14. Davila, M.L. et al. Efficacy and Toxicity Management of 19-28z CAR T Cell Therapy in B
Cell Acute Lymphoblastic Leukemia. Sci Transl Med 6 (2014).
15. Lee, D.W. et al. T cells expressing CD19 chimeric antigen receptors for acute
lymphoblastic leukaemia in children and young adults: a phase 1 dose-escalation trial.
The Lancet 385, 517-528 (2015).
16. Maude, S.L. et al. Efficacy and Safety of CTL019 in the First US Phase II Multicenter
Trial in Pediatric Relapsed/Refractory Acute Lymphoblastic Leukemia: Results of an
Interim Analysis. Blood 128, 2801-2801 (2016).
17. Uckun, F.M. et al. Detailed studies on expression and function of CD19 surface
determinant by using B43 monoclonal antibody and the clinical potential of anti-CD19
immunotoxins. Blood 71, 13-29 (1988).
18. Tasian, S.K. & Gardner, R.A. CD19-redirected chimeric antigen receptor-modified T
cells: a promising immunotherapy for children and adults with B-cell acute lymphoblastic
leukemia (ALL). Ther Adv Hematol 6, 228-241 (2015).
19. Gilham, D.E., Debets, R., Pule, M., Hawkins, R.E. & Abken, H. CAR-T cells and solid
tumors: tuning T cells to challenge an inveterate foe. Trends Mol Med 18, 377-84 (2012).
20. Kakarla, S. & Gottschalk, S. CAR T Cells for Solid Tumors: Armed and Ready to Go?
Cancer J 20, 151-5 (2014).
21. Newick, K., O'Brien, S., Moon, E. & Albelda, S.M. CAR T Cell Therapy for Solid Tumors.
Annu Rev Med 68, 139-152 (2017).
49
22. Fesnak, A.D., June, C.H. & Levine, B.L. Engineered T cells: the promise and challenges
of cancer immunotherapy. Nat Rev Cancer 16, 566-81 (2016).
23. Perica, K., Varela, J.C., Oelke, M. & Schneck, J. Adoptive T cell immunotherapy for
cancer. Rambam Maimonides Med J 6, e0004 (2015).
24. Smith-Garvin, J.E., Koretzky, G.A. & Jordan, M.S. T cell activation. Annu Rev Immunol
27, 591-619 (2009).
25. Chen, L. & Flies, D.B. Molecular mechanisms of T cell co-stimulation and co-inhibition.
Nat Rev Immunol 13, 227-42 (2013).
26. Schwartz, R.H. Costimulation of T lymphocytes: The role of CD28, CTLA-4, and B7/BB1
in interleukin-2 production and immunotherapy. Cell 71, 1065-1068 (1992).
27. Boyman, O. & Sprent, J. The role of interleukin-2 during homeostasis and activation of
the immune system. Nat Rev Immunol 12, 180-90 (2012).
28. Hamilos, D.L. Antigen presenting cells. Immunol Res 8, 980117 (1989).
29. Guermonprez, P., Valladeau, J., Zitvogel, L., Thery, C. & Amigorena, S. Antigen
presentation and T cell stimulation by dendritic cells. Annu Rev Immunol 20, 621-67
(2002).
30. Berger, C. et al. Adoptive transfer of effector CD8+ T cells derived from central memory
cells establishes persistent T cell memory in primates. J Clin Invest 118, 294-305 (2008).
31. Hinrichs, C.S. et al. Adoptively transferred effector cells derived from naive rather than
central memory CD8+ T cells mediate superior antitumor immunity. Proc Natl Acad Sci U
S A 106, 17469-74 (2009).
32. Oliveira, G. et al. Tracking genetically engineered lymphocytes long-term reveals the
dynamics of T cell immunological memory. Sci Transl Med 7, 1-14 (2015).
33. Fraietta, J.A. et al. Determinants of response and resistance to CD19 chimeric antigen
receptor (CAR) T cell therapy of chronic lymphocytic leukemia. Nat Med 24, 563-571
(2018).
50
34. Ghassemi, S. et al. Reducing Ex Vivo Culture Improves the Antileukemic Activity of
Chimeric Antigen Receptor (CAR) T Cells. Cancer Immunol Res 6, 1100-1109 (2018).
35. Hinrichs, C.S. et al. Human effector CD8+ T cells derived from naive rather than memory
subsets possess superior traits for adoptive immunotherapy. Blood 117, 808-814 (2011).
36. June, C.H. Principles of adoptive T cell cancer therapy. J Clin Invest 117, 1204-12
(2007).
37. Golubovskaya, V. & Wu, L. Different Subsets of T Cells, Memory, Effector Functions,
and CAR-T Immunotherapy. Cancers 8 (2016).
38. Yee, C. et al. Adoptive T cell therapy using antigen-specific CD8+ T cell clones for the
treatment of patients with metastatic melanoma: in vivo persistence, migration, and
antitumor effect of transferred T cells. Proc Natl Acad Sci U S A 99, 16168-73 (2002).
39. Mackensen, A. et al. Phase I study of adoptive T-cell therapy using antigen-specific
CD8+ T cells for the treatment of patients with metastatic melanoma. J Clin Oncol 24,
5060-9 (2006).
40. Levine, B.L. Performance-enhancing drugs: design and production of redirected chimeric
antigen receptor (CAR) T cells. Cancer Gene Ther 22, 79-84 (2015).
41. Ratta, M. et al. Dendritic cells are functionally defective in multiple myeloma: the role of
interleukin-6. Blood 100, 230-237 (2002).
42. Hawiger, D. et al. Dendritic Cells Induce Peripheral T Cell Unresponsiveness under
Steady State Conditions in Vivo. The Journal of Experimental Medicine 194, 769-780
(2001).
43. Maus, M.V. et al. Ex vivo expansion of polyclonal and antigenspecific cytotoxic T
lymphocytes by artificial APCs expressing ligands for the T-cell receptor, CD28 and 4-
1BB. Nat Biotechnol 20, 143-148 (2002).
44. Kim, J.V., Latouche, J.B., Riviere, I. & Sadelain, M. The ABCs of artificial antigen
presentation. Nat Biotechnol 22, 403-10 (2004).
51
45. Butler, M.O. & Hirano, N. Human cell-based artificial antigen-presenting cells for cancer
immunotherapy. Immunol Rev 257, 191-209 (2014).
46. Hasan, A., Selvakumar, A. & O’Reilly, R. Artificial Antigen Presenting Cells: An Off the
Shelf Approach for Generation of Desirable T-Cell Populations for Broad Application of
Adoptive Immunotherapy. Adv Genet Eng 4 (2015).
47. Turtle, C.J. & Riddell, S.R. Artificial Antigen-Presenting Cells for Use in Adoptive
Immunotherapy. Cancer J 16, 374-381 (2010).
48. Neal, L.R. et al. The Basics of Artificial Antigen Presenting Cells in T Cell-Based Cancer
Immunotherapies. J Immunol Res Ther 2, 68-79 (2017).
49. Van Wauwe, J.P., De Mey, J.R. & Goossens, J.G. OKT3: a monoclonal anti-human T
lymphocyte antibody with potent mitogenic properties. J Immunol 124, 2708-2713
(1980).
50. Tsoukas, C.D. et al. Activation of resting T lymphocytes by anti-CD3 (T3) antibodies in
the absence of monocytes. J Immunol 135, 1719-1723 (1985).
51. Riddell, S.R. & Greenberg, P.D. The use of anti-CD3 and anti-CD28 monoclonal
antibodies to clone and expand human antigen-specific T cells. Journal of Immunological
Methods 128, 189-201 (1990).
52. Acuto, O. & Michel, F. CD28-mediated co-stimulation: a quantitative support for TCR
signalling. Nat Rev Immunol 3, 939-51 (2003).
53. Hollyman, D. et al. Manufacturing Validation of Biologically Functional T Cells Targeted
to CD19 Antigen for Autologous Adoptive Cell Therapy. J Immunother 32, 169-180
(2009).
54. Ledbetter, J.A., Gentry, L.E., June, C.H., Rabinovitch, P.S. & Purchio, A.F. Stimulation of
T Cells through the CD3/T-Cell Receptor Complex: Role of Cytoplasmic Calcium,
Protein Kinase C Translocation, and Phosphorylation of pp60c-src in the Activation
Pathway. Mol Cell Biol 7, 650-656 (1987).
52
55. Trickett, A. & Kwan, Y.L. T cell stimulation and expansion using anti-CD3/CD28 beads.
Journal of Immunological Methods 275, 251-255 (2003).
56. Neurauter, A.A. et al. in Cell Separation: Fundamentals, Analytical and Preparative
Methods (eds. Kumar, A., Galaev, I.Y. & Mattiasson, B.) 41-73 (Springer Berlin
Heidelberg, Berlin, Heidelberg, 2007).
57. Thompson, J.A., Figlin, R.A., Sifri-Steele, C., Berenson, R.J. & Frohlich, M.W. A Phase I
Trial of CD3/CD28-activated T Cells (Xcellerated T Cells) and Interleukin-2 in Patients
with Metastatic Renal Cell Carcinoma1. Clin Cancer Res 9, 3562-3570 (2003).
58. Wang, X. & Riviere, I. Clinical manufacturing of CAR T cells: foundation of a promising
therapy. Mol Ther Oncolytics 3, 16015 (2016).
59. (ed. FDA) (2017).
60. Gattinoni, L. et al. Acquisition of full effector function in vitro paradoxically impairs the in
vivo antitumor efficacy of adoptively transferred CD8+ T cells. J Clin Invest 115, 1616-26
(2005).
61. Deeths, M.J. & Mescher, M.F. B7-1-dependent co-stimulation results in qualitatively and
quantitatively different responses by CD4+ and CDS+ T cells. Eur J Immunol 27, 598-
608 (1997).
62. Eggermont, L.J., Paulis, L.E., Tel, J. & Figdor, C.G. Towards efficient cancer
immunotherapy: advances in developing artificial antigen-presenting cells. Trends
Biotechnol 32, 456-65 (2014).
63. Mescher, M.F. Surface contact requirements for activation of cytotoxic T lymphocytes. J
Immunology 149, 2402-2405 (1992).
64. Perica, K. et al. Nanoscale artificial antigen presenting cells for T cell immunotherapy.
Nanomedicine 10, 119-29 (2014).
53
65. Perica, K. et al. Magnetic Field-Induced T Cell Receptor Clustering by Nanoparticles
Enhances T Cell Activation and Stimulates Antitumor Activity. ACS Nano 8, 2252-2260
(2014).
66. Boyle, S. et al. Quantum dot fluorescence characterizes the nanoscale organization of T
cell receptors for antigen. Biophys J 101, L57-9 (2011).
67. Pageon, S.V. et al. Functional role of T-cell receptor nanoclusters in signal initiation and
antigen discrimination. Proc Natl Acad Sci U S A 113, E5454-E5463 (2016).
68. Purtic, B., Pitcher, L.A., van Oers, N.S. & Wulfing, C. T cell receptor (TCR) clustering in
the immunological synapse integrates TCR and costimulatory signaling in selected T
cells. Proc Natl Acad Sci U S A 102, 2904-9 (2005).
69. Molnar, E. et al. Cholesterol and sphingomyelin drive ligand-independent T-cell antigen
receptor nanoclustering. J Biol Chem 287, 42664-74 (2012).
70. Matic, J., Deeg, J., Scheffold, A., Goldstein, I. & Spatz, J.P. Fine tuning and efficient T
cell activation with stimulatory aCD3 nanoarrays. Nano Lett 13, 5090-7 (2013).
71. Hickey, J.W., Vicente, F.P., Howard, G.P., Mao, H.Q. & Schneck, J.P. Biologically
Inspired Design of Nanoparticle Artificial Antigen-Presenting Cells for
Immunomodulation. Nano Lett 17, 7045-7054 (2017).
72. Sunshine, J.C., Perica, K., Schneck, J.P. & Green, J.J. Particle shape dependence of
CD8+ T cell activation by artificial antigen presenting cells. Biomaterials 35, 269-277
(2014).
73. Meyer, R.A. et al. Biodegradable nanoellipsoidal artificial antigen presenting cells for
antigen specific T-cell activation. Small 11, 1519-25 (2015).
74. Steenblock, E.R., Fadel, T., Labowsky, M., Pober, J.S. & Fahmy, T.M. An artificial
antigen-presenting cell with paracrine delivery of IL-2 impacts the magnitude and
direction of the T cell response. J Biol Chem 286, 34883-92 (2011).
54
75. Fadel, T.R. et al. Enhanced Cellular Activation with Single Walled Carbon Nanotube
Bundles Presenting Antibody Stimuli. Nano Lett 8, 2070-2076 (2008).
76. Fadel, T.R. et al. A carbon nanotube-polymer composite for T-cell therapy. Nat
Nanotechnol 9, 639-47 (2014).
77. Cheung, A.S. & Mooney, D.J. Engineered Materials for Cancer Immunotherapy. Nano
Today 10, 511-531 (2015).
78. Ahlers, J.D. & Belyakov, I.M. Memories that last forever: strategies for optimizing
vaccine T-cell memory. Blood 115, 1678-89 (2010).
79. Mandal, S. et al. Therapeutic nanoworms: towards novel synthetic dendritic cells for
immunotherapy. Chemical Science 4, 4168 (2013).
80. Pike, L.J. Lipid rafts: bringing order to chaos. J Lipid Res 44, 655-67 (2003).
81. Anderson, H.A., Hiltbold, E.M. & Roche, P.A. Concentration of MHC class II molecules in
lipid rafts facilitates antigen presentation. Nat Immunol 1, 156-162 (2000).
82. Ding, Q. et al. RAFTsomes containing epitope-MHC-II complexes mediated CD4+ T cell
activation and antigen-specific immune responses. Pharm Res 30, 60-9 (2013).
83. Zappasodi, R. et al. The effect of artificial antigen-presenting cells with preclustered anti-
CD28/-CD3/-LFA-1 monoclonal antibodies on the induction of ex vivo expansion of
functional human antitumor T cells. Haematologica 93, 1523-34 (2008).
84. Irvine, D.J. & Doh, J. Synthetic surfaces as artificial antigen presenting cells in the study
of T cell receptor triggering and immunological synapse formation. Semin Immunol 19,
245-54 (2007).
85. Su, X., Ditlev, J.A., Rosen, M.K. & Vale, R.D. in The Immune Synapse: Methods and
Protocols (eds. Baldari, C.T. & Dustin, M.L.) 65-76 (Springer New York, New York, NY,
2017).
55
86. Cheung, A.S., Zhang, D.K.Y., Koshy, S.T. & Mooney, D.J. Scaffolds that mimic antigen-
presenting cells enable ex vivo expansion of primary T cells. Nat Biotechnol 36, 160-169
(2018).
87. Cieri, N. et al. IL-7 and IL-15 instruct the generation of human memory stem T cells from
naive precursors. Blood 121, 573-84 (2013).
88. Kondo, T. et al. Generation and application of human induced-stem cell memory T cells
for adoptive immunotherapy. Cancer Sci 109, 2130-2140 (2018).
89. Purwada, A. et al. Ex vivo engineered immune organoids for controlled germinal center
reactions. Biomaterials 63, 24-34 (2015).
90. Quadrato, G. et al. Cell diversity and network dynamics in photosensitive human brain
organoids. Nature 545, 48-53 (2017).
91. Nozaki, K. et al. Co-culture with intestinal epithelial organoids allows efficient expansion
and motility analysis of intraepithelial lymphocytes. J Gastroenterol 51, 206-13 (2016).
92. Boj, S.F. et al. Organoid models of human and mouse ductal pancreatic cancer. Cell
160, 324-38 (2015).
93. Neal, J.T. et al. Organoid Modeling of the Tumor Immune Microenvironment. Cell 175,
1972-1988 e16 (2018).
94. Dijkstra, K.K. et al. Generation of Tumor-Reactive T Cells by Co-culture of Peripheral
Blood Lymphocytes and Tumor Organoids. Cell 174, 1586-1598 e12 (2018).
95. Montel-Hagen, A. et al. Organoid-Induced Differentiation of Conventional T Cells from
Human Pluripotent Stem Cells. Cell Stem Cell 24, 376-389 e8 (2019).
96. Blanco, E., Shen, H. & Ferrari, M. Principles of nanoparticle design for overcoming
biological barriers to drug delivery. Nat Biotechnol 33, 941-51 (2015).
97. Shi, J., Kantoff, P.W., Wooster, R. & Farokhzad, O.C. Cancer nanomedicine: progress,
challenges and opportunities. Nat Rev Cancer 17, 20-37 (2017).
56
98. Maeda, H. Toward a full understanding of the EPR effect in primary and metastatic
tumors as well as issues related to its heterogeneity. Adv Drug Deliv Rev 91, 3-6 (2015).
99. Rosenblum, D., Joshi, N., Tao, W., Karp, J.M. & Peer, D. Progress and challenges
towards targeted delivery of cancer therapeutics. Nat Commun 9, 1410 (2018).
100. Wei, T. et al. Anticancer drug nanomicelles formed by self-assembling amphiphilic
dendrimer to combat cancer drug resistance. Proc Natl Acad Sci U S A 112, 2978-83
(2015).
101. Liu, Y., Fang, J., Joo, K.I., Wong, M.K. & Wang, P. Codelivery of chemotherapeutics via
crosslinked multilamellar liposomal vesicles to overcome multidrug resistance in tumor.
PLoS One 9, e110611 (2014).
102. Housman, G. et al. Drug resistance in cancer: an overview. Cancers 6, 1769-92 (2014).
103. Spiotto, M.T., Rowley, D.A. & Schreiber, H. Bystander elimination of antigen loss
variants in established tumors. Nat Med 10, 294-8 (2004).
104. Vyas, M., Muller, R. & Pogge von Strandmann, E. Antigen Loss Variants: Catching Hold
of Escaping Foes. Front Immunol 8, 175 (2017).
105. Tannock, I.F. & Rotin, D. Acid pH in Tumors and Its Potential for Therapeutic
Exploitation. Cancer Res 49, 4373-4384 (1989).
106. Hirschhaeuser, F., Sattler, U.G. & Mueller-Klieser, W. Lactate: a metabolic key player in
cancer. Cancer Res 71, 6921-5 (2011).
107. Fischer, K. et al. Inhibitory effect of tumor cell–derived lactic acid on human T cells.
Blood 109, 3812-3819 (2007).
108. Anderson, K.G., Stromnes, I.M. & Greenberg, P.D. Obstacles Posed by the Tumor
Microenvironment to T cell Activity: A Case for Synergistic Therapies. Cancer Cell 31,
311-325 (2017).
109. Spolski, R., Li, P. & Leonard, W.J. Biology and regulation of IL-2: from molecular
mechanisms to human therapy. Nat Rev Immunol 18, 648-659 (2018).
57
110. Zhang, X., Sun, S., Hwang, I., Tough, D.F. & Sprent, J. Potent and Selective Stimulation
of Memory-Phenotype CD81 T Cells In Vivo by IL-15. Immunity 8, 591-599 (1998).
111. Waldmann, T.A. Cytokines in Cancer Immunotherapy. Cold Spring Harb Perspect Biol
10 (2018).
112. Lotze, M.T. et al. In vivo adminsitration of purified human interleukin 2. J Immunol 135,
2865-2875 (1985).
113. Atkins, M.B. et al. High-Dose Recombinant Interleukin 2 Therapy for Patients With
Metastatic Melanoma: Analysis of 270 Patients Treated Between 1985 and 1993. J Clin
Oncol 17, 2105-2116 (1999).
114. Ahmadzadeh, M. & Rosenberg, S.A. IL-2 administration increases CD4+ CD25(hi)
Foxp3+ regulatory T cells in cancer patients. Blood 107, 2409-14 (2006).
115. Berger, C. et al. Safety and immunologic effects of IL-15 administration in nonhuman
primates. Blood 114, 2417-26 (2009).
116. Hong, E. et al. Configuration-dependent Presentation of Multivalent IL-15:IL-15Ralpha
Enhances the Antigen-specific T Cell Response and Anti-tumor Immunity. J Biol Chem
291, 8931-50 (2016).
117. Kwong, B., Gai, S.A., Elkhader, J., Wittrup, K.D. & Irvine, D.J. Localized immunotherapy
via liposome-anchored Anti-CD137 + IL-2 prevents lethal toxicity and elicits local and
systemic antitumor immunity. Cancer Res 73, 1547-58 (2013).
118. Zheng, Y. et al. In vivo targeting of adoptively transferred T-cells with antibody- and
cytokine-conjugated liposomes. J Control Release 172, 426-35 (2013).
119. Blobe, G.C., Schiemann, W.P. & Lodish, H.F. Role of transforming growth factor β in
human disease. N Engl J Med 342, 1350-1358 (2000).
120. Colak, S. & Ten Dijke, P. Targeting TGF-beta Signaling in Cancer. Trends Cancer 3, 56-
71 (2017).
58
121. Park, J. et al. Combination delivery of TGF-beta inhibitor and IL-2 by nanoscale
liposomal polymeric gels enhances tumour immunotherapy. Nat Mater 11, 895-905
(2012).
122. Yingling, J.M., Blanchard, K.L. & Sawyer, J.S. Development of TGF-beta signalling
inhibitors for cancer therapy. Nat Rev Drug Discov 3, 1011-22 (2004).
123. Gueorguieva, I. et al. Defining a therapeutic window for the novel TGF-β inhibitor
LY2157299 monohydrate based on a pharmacokinetic/pharmacodynamic model. British
Journal of Clinical Pharmacology 77, 796-807 (2014).
124. Zheng, Y., Tang, L., Mabardi, L., Kumari, S. & Irvine, D.J. Enhancing Adoptive Cell
Therapy of Cancer through Targeted Delivery of Small-Molecule Immunomodulators to
Internalizing or Noninternalizing Receptors. ACS Nano 11, 3089-3100 (2017).
125. Schmid, D. et al. T cell-targeting nanoparticles focus delivery of immunotherapy to
improve antitumor immunity. Nat Commun 8, 1747 (2017).
126. Dannenmann, S.R. et al. Tumor-associated macrophages subvert T-cell function and
correlate with reduced survival in clear cell renal cell carcinoma. Oncoimmunology 2,
e23562 (2013).
127. Condeelis, J. & Pollard, J.W. Macrophages: obligate partners for tumor cell migration,
invasion, and metastasis. Cell 124, 263-6 (2006).
128. Noy, R. & Pollard, J.W. Tumor-associated macrophages: from mechanisms to therapy.
Immunity 41, 49-61 (2014).
129. Mantovani, A., Marchesi, F., Malesci, A., Laghi, L. & Allavena, P. Tumour-associated
macrophages as treatment targets in oncology. Nat Rev Clin Oncol 14, 399-416 (2017).
130. Huang, Z. et al. Targeted delivery of oligonucleotides into tumor-associated
macrophages for cancer immunotherapy. J Control Release 158, 286-92 (2012).
131. Vignali, D.A., Collison, L.W. & Workman, C.J. How regulatory T cells work. Nat Rev
Immunol 8, 523-32 (2008).
59
132. Tanaka, A. & Sakaguchi, S. Regulatory T cells in cancer immunotherapy. Cell Res 27,
109-118 (2017).
133. Ou, W. et al. Regulatory T cell-targeted hybrid nanoparticles combined with immuno-
checkpoint blockage for cancer immunotherapy. J Control Release 281, 84-96 (2018).
134. Sharpe, A.H. & Pauken, K.E. The diverse functions of the PD1 inhibitory pathway. Nat
Rev Immunol 18, 153-167 (2018).
135. Riley, J.L. PD-1 signaling in primary T cells. Immunol Rev 229, 114-125 (2009).
136. Kosmides, A.K., Sidhom, J.W., Fraser, A., Bessell, C.A. & Schneck, J.P. Dual Targeting
Nanoparticle Stimulates the Immune System To Inhibit Tumor Growth. ACS Nano 11,
5417-5429 (2017).
137. Schütz, C. et al. Antigen-specific T cell Redirectors: a nanoparticle based approach for
redirecting T cells. Oncotarget 7, 68503-68512 (2016).
138. Stephan, M.T., Moon, J.J., Um, S.H., Bershteyn, A. & Irvine, D.J. Therapeutic cell
engineering with surface-conjugated synthetic nanoparticles. Nat Med 16, 1035-41
(2010).
139. Siegler, E.L. et al. Combination Cancer Therapy Using Chimeric Antigen Receptor-
Engineered Natural Killer Cells as Drug Carriers. Mol Ther 25, 2607-2619 (2017).
140. Siriwon, N. et al. CAR-T Cells Surface-Engineered with Drug-Encapsulated
Nanoparticles Can Ameliorate Intratumoral T-cell Hypofunction. Cancer Immunol Res 6,
812-824 (2018).
141. Huang, B. et al. Active targeting of chemotherapy to disseminated tumors using
nanoparticle-carrying T cells. Sci Transl Med 7, 1-9 (2015).
142. Jones, R.B. et al. Antigen recognition-triggered drug delivery mediated by nanocapsule-
functionalized cytotoxic T-cells. Biomaterials 117, 44-53 (2017).
60
143. Stephan, M.T., Stephan, S.B., Bak, P., Chen, J. & Irvine, D.J. Synapse-directed delivery
of immunomodulators using T-cell-conjugated nanoparticles. Biomaterials 33, 5776-87
(2012).
144. Tang, L. et al. Enhancing T cell therapy through TCR-signaling-responsive nanoparticle
drug delivery. Nat Biotechnol 36, 707-716 (2018).
145. Yang, J.C. et al. Randomized study of high-dose and low-dose interleukin-2 in patients
with metastatic renal cancer. J Clin Oncol 21, 3127-32 (2003).
146. Hershkovitz, L., Schachter, J., Treves, A.J. & Besser, M.J. Focus on adoptive T cell
transfer trials in melanoma. Clin Dev Immunol 2010, 260267 (2010).
147. Hori, Y., Stern, P.J., Hynes, R.O. & Irvine, D.J. Engulfing tumors with synthetic
extracellular matrices for cancer immunotherapy. Biomaterials 30, 6757-67 (2009).
148. Zaharoff, D.A. et al. Intravesical immunotherapy of superficial bladder cancer with
chitosan/interleukin-12. Cancer Res 69, 6192-9 (2009).
149. Park, C.G. et al. Extended release of perioperative immunotherapy prevents tumor
recurrence and eliminates metastases. Sci Transl Med 10, 1-14 (2018).
150. Koshy, S.T., Ferrante, T.C., Lewin, S.A. & Mooney, D.J. Injectable, porous, and cell-
responsive gelatin cryogels. Biomaterials 35, 2477-87 (2014).
151. Zhan, Q. et al. Drug-eluting scaffold inhibited in vivo pancreatic tumorigenesis by
engaging murine CCR4(+)CD8(+) T cells. Colloids Surf B Biointerfaces 158, 469-473
(2017).
152. Wen, Y. et al. Retaining antibodies in tumors with a self-assembling injectable system.
Mol Pharm 10, 1035-44 (2013).
153. Hudalla, G.A. et al. Gradated assembly of multiple proteins into supramolecular
nanomaterials. Nat Mater 13, 829-36 (2014).
154. Aliperta, R. et al. Cryogel-supported stem cell factory for customized sustained release
of bispecific antibodies for cancer immunotherapy. Sci Rep 7, 42855 (2017).
61
155. Weiden, J., Tel, J. & Figdor, C.G. Synthetic immune niches for cancer immunotherapy.
Nat Rev Immunol 18, 212-219 (2018).
156. Peppas, N.A. & Mikos, A.G. Preparation methods and structure of hydrogels (CRC
Press Inc, Boca Raton, FL, USA, 1986).
157. Ahmed, E.M. Hydrogel: Preparation, characterization, and applications: A review. J Adv
Res 6, 105-21 (2015).
158. Singh, A. & Peppas, N.A. Hydrogels and scaffolds for immunomodulation. Adv Mater 26,
6530-41 (2014).
159. Tsao, C.T. et al. Thermoreversible poly(ethylene glycol)-g-chitosan hydrogel as a
therapeutic T lymphocyte depot for localized glioblastoma immunotherapy.
Biomacromolecules 15, 2656-62 (2014).
160. Monette, A., Ceccaldi, C., Assaad, E., Lerouge, S. & Lapointe, R. Chitosan thermogels
for local expansion and delivery of tumor-specific T lymphocytes towards enhanced
cancer immunotherapies. Biomaterials 75, 237-249 (2016).
161. Stephan, S.B. et al. Biopolymer implants enhance the efficacy of adoptive T-cell therapy.
Nat Biotechnol 33, 97-101 (2015).
162. Smith, T.T. et al. Biopolymers codelivering engineered T cells and STING agonists can
eliminate heterogeneous tumors. J Clin Invest 127, 2176-2191 (2017).
163. Atik, A.F. et al. Hyaluronic acid based low viscosity hydrogel as a novel carrier for
Convection Enhanced Delivery of CAR T cells. J Clin Neurosci 56, 163-168 (2018).
164. Pardridge, W.M. Blood-brain barrier delivery. Drug Discov Today 12, 54-61 (2007).
165. Bobo, R.H. et al. Convection-enhanced delivery of macromolecules in the brain. Proc
Natl Acad Sci U S A 91, 2076-2080 (1994).
166. Chaichana, K.L., Pinheiro, L. & Brem, H. Delivery of local therapeutics to the brain:
working toward advancing treatment for malignant gliomas. Ther. Deliv. 6, 353-369
(2015).
62
167. Debinski, W. & Tatter, S.B. Convection enhanced delivery for the treatment of brain
tumors. Expert Rev. Neurother. 9, 1519-1527 (2009).
168. Maude, S.L. et al. Chimeric antigen receptor T cells for sustained remissions in
leukemia. N Engl J Med 371, 1507-17 (2014).
169. Kalos, M. et al. T Cells with Chimeric Antigen Receptors Have Potent Antitumor Effects
and Can Establish Memory in Patients with Advanced Leukemia. Sci Transl Med 3, 1-13
(2011).
170. Smith, T.T. et al. In situ programming of leukaemia-specific T cells using synthetic DNA
nanocarriers. Nat Nanotechnol 12, 813-820 (2017).
171. Gocheva, V. et al. Distinct roles for cysteine cathepsin genes in multistage
tumorigenesis. Genes & development 20, 543-556 (2006).
172. Duffy, M.J., McGowan, P.M., Harbeck, N., Thomssen, C. & Schmitt, M. UPA and PAI-1
as biomarkers in breast cancer: validated for clinical use in level-of-evidence-1 studies.
Breast Cancer Research 16, 1-10 (2014).
173. Zhen, Y. et al. Clinicopathologic significance of legumain overexpression in cancer: a
systematic review and meta-analysis. Sci Rep 5, 16599 (2015).
174. Mason, S.D. & Joyce, J.A. Proteolytic networks in cancer. Trends Cell Biol 21, 228-37
(2011).
175. Desnoyers, L.R. et al. Tumor-Specific Activation of an EGFR-Targeting Probody
Enhances Therapeutic Index. Sci Transl Med 5, 1-10 (2013).
176. Han, X. et al. Masked Chimeric Antigen Receptor for Tumor-Specific Activation. Mol
Ther 25, 274-284 (2017).
177. Markley, J.C. & Sadelain, M. IL-7 and IL-21 are superior to IL-2 and IL-15 in promoting
human T cell-mediated rejection of systemic lymphoma in immunodeficient mice. Blood
115, 3508-19 (2010).
63
178. Adachi, K. et al. IL-7 and CCL19 expression in CAR-T cells improves immune cell
infiltration and CAR-T cell survival in the tumor. Nat Biotechnol 36, 346-351 (2018).
179. Hoyos, V. et al. Engineering CD19-specific T lymphocytes with interleukin-15 and a
suicide gene to enhance their anti-lymphoma/leukemia effects and safety. Leukemia 24,
1160-70 (2010).
180. Avanzi, M.P. et al. Engineered Tumor-Targeted T Cells Mediate Enhanced Anti-Tumor
Efficacy Both Directly and through Activation of the Endogenous Immune System. Cell
Rep 23, 2130-2141 (2018).
181. Li, S. et al. Enhanced Cancer Immunotherapy by Chimeric Antigen Receptor-Modified T
Cells Engineered to Secrete Checkpoint Inhibitors. Clin Cancer Res 23, 6982-6992
(2017).
182. Zhang, F. et al. Nanoparticles That Reshape the Tumor Milieu Create a Therapeutic
Window for Effective T-cell Therapy in Solid Malignancies. Cancer Res 78, 3718-3730
(2018).
183. Schuster, S.J. et al. Tisagenlecleucel in Adult Relapsed or Refractory Diffuse Large B-
Cell Lymphoma. N Engl J Med 380, 45-56 (2019).
184. Crump, M. et al. Outcomes in refractory diffuse large B-cell lymphoma: results from the
international SCHOLAR-1 study. Blood 130, 1800-1808 (2017).
185. Hitz, F. et al. Outcome of patients with primary refractory diffuse large B cell lymphoma
after R-CHOP treatment. Ann Hematol 94, 1839-43 (2015).
186. Gargett, T. & Brown, M.P. The inducible caspase-9 suicide gene system as a "safety
switch" to limit on-target, off-tumor toxicities of chimeric antigen receptor T cells. Front
Pharmacol 5, 235 (2014).
187. Schneider, D. et al. A tandem CD19/CD20 CAR lentiviral vector drives on-target and off-
target antigen modulation in leukemia cell lines. J Immunother Cancer 5, 42 (2017).
64
188. Chmielewski, M., Hombach, A.A. & Abken, H. Of CARs and TRUCKs: chimeric antigen
receptor (CAR) T cells engineered with an inducible cytokine to modulate the tumor
stroma. Immunol Rev 257, 83-90 (2014).
189. Rupp, L.J. et al. CRISPR/Cas9-mediated PD-1 disruption enhances anti-tumor efficacy
of human chimeric antigen receptor T cells. Sci Rep 7, 737 (2017).
190. Emerich, D.F. & Thanos, C.G. Nanotechnology and medicine. Expert Opin Biol Ther 3,
655-63 (2003).
191. Gammon, J.M., Dold, N.M. & Jewell, C.M. Improving the clinical impact of biomaterials in
cancer immunotherapy. Oncotarget 7, 15421-15443 (2016).
65
Chapter 2: A Tumor Stroma-Targeting Immunotoxin
Engineered for Enhancing the Efficacy of Immunotherapy in
a B16 Melanoma Tumor Model
2.1 Abstract and Introduction
Abstract
The use of immunotherapies to treat cancer has increased rapidly due to its proven
efficacy in clinical trials. However, their efficacy when treating solid tumors are
undermined by the immunosuppressive tumor microenvironment's (TME) ability to
hinder the activity of endogenous T cells and prevent them from fully working in
conjunction with the therapeutic. Cancer associated fibroblasts (CAFs) are stromal cells
that are known to have a key role in maintaining the immunosuppressive nature of the
TME. As a result, we used an engineered immunotoxin (α-FAP-PE38hu) to target CAFs
in vivo and demonstrate its synergistic effects when used with PD-1 antibody blockade
or chimeric antigen receptor (CAR) T cell therapy - two widely adopted forms of
immunotherapy. We observed a marked inhibition of tumor growth in the mice receiving
α-FAP-PE38hu with either PD-1 or CAR T cell therapy, when compared to the groups
receiving any single treatment. Furthermore, we report an increase in the presence of
tumor infiltrating T cells as well as a decrease in the presence of immune suppressive
cells among mice receiving α-FAP-PE38hu combined with either PD-1 Blockade or CAR
T cell therapy the combined therapy. These results highlight the ability of α-FAP-
66
PE38hu, an engineered, CAF-targeting immunotoxin, to enhance the overall efficacy of
immunotherapies.
Introduction
Checkpoint inhibitors have revolutionized the field of cancer immunotherapy by allowing
treatments to circumvent the effects of the immunosuppressive tumor
microenvironment. PD-1 is a major checkpoint receptor that has been studied
extensively. It has been shown that PD-1 expressed on T cells downregulates signaling
induced by the interaction between antigens and T cell receptors, reduces secretion of
pro-inflammatory cytokines when bound to either of its two ligands (PD-L1 and PD-L2)
on antigen presenting cells(APCs) and tumor cells, and therefore hinders T cell effector
function.
1
As a result, blockade of the interaction between PD-1 and its ligands will
restore T cell effector functions.
2
Antibodies that block PD-1 binding have been
evaluated in clinical trials and have resulted in tumor regression in 30%-50% of patients
with various diseases, ranging from melanoma to lung cancer.
3-5
Nonetheless, only a
minority of patients experience a significant response with PD-1 therapy. Often, the
blockade of a single inhibitory pathway is not sufficient enough to overcome the
immunosuppressive TMEs that impair T cell functions and promote angiogenesis and
metastasis.
2, 6-9
Chimeric antigen receptor (CAR) T cells are another form of immunotherapy that has
transformed the landscape of cancer therapy. Their remarkable clinical success as a
treatment for acute lymphoblastic leukemia,
10, 11
has inspired further research into the
development of novel CARs for the targeting of different cancers. However, researchers
67
have found using CAR T cells to treat solid tumors to be considerably more difficult than
treating liquid tumors. This disparity can be attributed to the physical barriers presented
by the TME and the immunosuppressive cells present within the tumor, such as
regulatory T cells (Tregs), myeloid derived suppressor cells (MDSCs), and tumor
associated macrophages,
12
which work together to sabotage CAR T cell function.
13, 14
Cancer associated fibroblasts (CAFs) are the major non-hematopoietic cell types that
play crucial roles in the immunosuppressive functions of TMEs.
15
CAFs promote the
polarization of M2 macrophages, which are known to promote angiogenesis and induce
the proliferation of blood endothelial cells (BECs).
16, 17
It has been reported that BECs
upregulate PD-L1, which suppresses T cell functions through the inhibitory PD-1
pathway.
18, 19
Furthermore, the reduced expression of adhesion molecules on tumor-
associated BECs hinders lymphocyte extravasation.
20, 21
Additionally, CAFs enhance
the recruitment immune suppressive cells, such as MDSCs, through the secretion of
chemokines.
22, 23
Fibroblast activation protein (FAP) is the major phenotypic marker of
CAFs and studies have shown that depletion of FAP
+
cells result in reduced tumor
growth and metastasis.
24, 25
Therefore, therapeutics targeting FAP
+
stromal cells may be
a viable strategy to overcome the immunosuppressive TME and enhance antitumor
efficacy (Fig. 1A).
We previously developed a novel approach to target FAP
+
stromal cells with a truncated
version of Pseudomonas exotoxin A (PE38) named αFAP-PE38 and showed that the
immunotoxin bound to FAP with high affinity, specifically killed FAP
+
cells in vitro, and
68
inhibited tumor growth in vivo.
26
However, PE38, which is a toxin derived from bacteria,
has been shown to be highly immunogenic - especially in patients with solid tumors and
a fully functional immune system.
27-29
The first clinical trial with PE immunotoxins led to
100% of patients developing antibodies against the toxin. Subsequent clinical trials of
PE-based immunotoxins all resulted in an immunogenicity rate of at least 50%.
29, 30
Thus, there is an urgent need to reduce the immunogenicity of these bacterial toxins. To
address these needs, Liu et al designed a PE-based recombinant immunotoxin with low
immunogenicity through the identification and replacement of human B-cell epitopes.
31
In this study, we adopted the point mutations developed by Liu et al and attempted to
further reduce PE38's immunogenicity by removing domain II, which was shown to have
limited effects on its toxicity
32, 33
- thereby creating α-FAP-PE38hu: a less
immunogenic, recombinant immunotoxin. We examined the specificity and cytotoxicity
of α-FAP-PE38hu in vitro and measured its effects on tumor growth in a B16-F10
melanoma mouse tumor model. Afterwards, we examined the effects of α-FAP-PE38hu
combined with a PD-1 blockade treatment, as well as α-FAP-PE38hu with CAR T cell
therapy, on the enhancement of tumor growth suppression and tested the each
combination therapy's ability to alter the immunosuppressive TME.
69
2.2: Materials and Methods
Mice, cell line construction and cell culture
Female C57BL/6 mice were purchased from Charles River Laboratories (Wilmington,
MA) and housed in the animal facility in accordance with IACUC regulations. All animal
experiments and protocols were performed according to the guidelines set by the NIH
and the University of Southern California on the Care and Use of Animals. B16-F10 and
293T cells were purchased from ATCC (Manassas, VA) and cultured in high-glucose
Dulbecco’s modified Eagle medium (Hyclone, Logan, UT) with L-glutamine (Hyclone
Laboratories, Omaha, NE) supplemented with 10% fetal bovine serum (Sigma-Aldrich,
St. Louis, MO). The B16-CD19, 293T–hFAP and 293T–mFAP cell lines were generated
by stable transduction of cells with lentivirus pseudotyped with vesicular stomatitis virus
glycoprotein, as described previously.
34
T Cell Expansion and Transduction
The spleens of female C56BL/6 mice were harvested and mashed through a 70µm cell
strainer to obtain a single cell suspension. Red blood cells were then lysed and the
splenocytes were cultured in R10 (RPMI supplemented with 10% FBS) and 15 ng/mL
mouse IL-2, in 96 well plates previously coated with 8 µg/mL anti-CD3 and 2 µg/mL
anti-CD28 for 2 days. For transduction, 12 well plates were coated with 15 µg
RetroNectin per well (Takara) prior to centrifugation with lentiviral supernatant for 2
hours at 2,000 x g at 32 °C. The mouse T cells were then resuspended to a final
concentration of 1 x 10
6
/mL in R10 supplemented with 15 ng/mL mouse IL-2 and added
70
to the retrovirus-coated plate before centrifugation for 30 minutes at 1,050 x g at 32 °C.
After transduction, the cells were incubated overnight at 37 °C and 5% CO
2
. The
transduction procedure was then repeated once more.
Plasmid construction and protein purification
The sequences encoding both the original and the mutated truncated PE38 described
by Liu et al
31
were fused to the sequence of a species-crossreactive FAP-specific scFv
(MO36)
35
and were cloned into the pET-28a(+) vector (Life Technologies, Grand Island,
NY) separately. The plasmids were transformed into Escherichia coli BL21(DE3)
(Invitrogen) and were grown in luria broth media containing 100 µg/ml of kanamycin at
37°C. When OD
600
reached 0.6, isopropyl-β-D-1-thiogalactopyranoside (Sigma-Aldrich)
was added to 1 mM for 4 hours. The cells were then harvested and the recombinant
fusion protein was isolated from inclusion bodies by washing with 2M urea buffer and
dissolving in 8M urea. After renaturation by dialysis in gradient urea buffer, the
recombinant fusion protein was purified by Ni
2+
IDA column for His-tag purification
(Qiagen, Valencia, CA).
Dye labeling of α-FAP-PE38hu
Purified α-FAP-PE38hu protein was incubated with 50 nmol of Alexa488-TFP ester
(Invitrogen) for 2 hr in 0.1 M sodium bicarbonate buffer (pH = 9.3). The unbound dye
molecules were removed via buffer exchange into PBS (pH = 7.4) using a Zeba
desalting spin column (Thermo Fisher Scientific).
71
In vitro cytotoxicity of α-FAP-PE38hu
Standard XTT assays were performed to measure the dose-dependent cytotoxicity of α-
FAP-PE38hu in cultured cells using a commercial kit (Roche Scientific). Cells were
seeded on 96-well plates one day before the treatment, treated with α-FAP-PE38hu on
day 2 and XTT assay was conducted on day 4. PBS was used as a control for 0% cell
death. The OD values were normalized between the 100% cell death (0% line) and PBS
controls (100% alive) and fit to a standard 4-parameter sigmoidal curve with a variable
slope using the GraphPad Prism (version 5.03; GraphPad Software) program to obtain
the concentration of immunotoxin at which there was 50% cell death (IC
50
).
Tumor challenge and treatment
Female C57BL/6 mice (n = 6 per group) were subcutaneously inoculated with 2 ×10
5
B16-F10 cells on the right flank. Tumor growth was evaluated every other day by
measuring tumor diameter with calipers. Tumor volume was defined as (smallest
diameter) × (longest diameter) × (height). αFAP-PE38 at the dose of 0.5 mg/kg, α-FAP-
PE38hu at the dose of 0.5 mg/kg or 1.5 mg/kg and PD-1 antibody at the dose of 10
mg/kg were injected to mice via i.v. injection at day 12 post injection respectively. For
the CAR T cell experimental groups, mice were subcutaneously inoculated with 5 x 10
5
B16-CD19 cells on the right flank on Day 0 and given a 500 cGy dose of x-ray radiation
4 days later. On days 5, 7, and 9, the mice received 0.5 mg/kg α-FAP-PE38hu or a PBS
via tail vein injections, and on day 6, the mice received 3 x 10
6
CAR T or nontransduced
cells. Tumor size was monitored every other day. Mice were euthanized when they
72
displayed significant weight loss or weakness, ulceration of tumors, or tumor sizes in
excess of 1,000 mm
3
.
Pharmacokinetics
Two groups of three female C57BL/6 mice were injected with 10 µg of original αFAP-
PE38 and α-FAP-PE38hu in 0.2 mL of PBS with 0.2% HSA respectively. Blood samples
were taken from 3 separate mice within each group at time intervals of 2, 5, 10, 20, 30,
and 60 minutes from the time of injection, and each mouse was bled twice. Groups of 3
mice were bled at time intervals of 2 and 60 minutes, 5 and 30 minutes, or 10 and 20
minutes. Serum was harvested from the blood samples and analyzed by His Tag ELISA
Detection Kit (GenScript) following the manufacturer's protocol.
Flow cytometry analysis
Tumor tissue from treated mice was harvested, minced to single suspension cells and
filtered through 70 μm nylon strainers (BD Falcon, Franklin Lakes, NJ). The filtered cells
were washed twice with cold PBS and then incubated for 10 minutes at 4 °C with rat
anti-mouse CD16/CD32 mAbs (BD Biosciences) to block nonspecific binding. Cells
were then stained with monoclonal antibodies conjugated with fluorescent dyes. All
staining antibodies and isotype controls were purchased from eBioscience or
BioLegend, including anti-CD45 (30-F11), anti-CD3 (145-2C11), anti-CD4 (RM4-5), anti-
CD8 (53–6.7), anti-F4/80 (BM8), anti-ly-6G (1A8), anti-ly-6C(1A8), anti-PD-1 (RMP1-
30), anti-CD25 (PC61), anti-FoxP3 (FJK-16S), anti-CD11b (M1/70). Tregs were
identified by CD3
+
CD45
+
CD4
+
CD25
+
Foxp3
+
markers; CD4 and CD8 T cells were
73
identified by CD3
+
CD45
+
CD4
+
and CD3
+
CD45
+
CD8
+
markers, respectively. MDSCs
were identified by CD45
+
CD11b
+
F4/80
+
ly-6C
+
ly-6G
+
markers. Data were acquired on a
MACSquant cytometer (Miltenyi Biotec, San Diego, CA), and the analysis was
performed using FlowJo software (Tree Star, Ashland, OR).
Immunofluorescence imaging and Immunohistochemical analysis
For immunofluorescent staining, the frozen tumor sample slides were fixed with 4%
formaldehyde, permeabilized with 0.1% Triton X-100, stained with TUNEL antibody, and
followed by counterstaining with DAPI. All fluorescence images were acquired on a
Yokogawa spinning-disk confocal scanner system (Solamere Technology Group) using
a Nikon eclipse Ti-E microscope (Nikon) equipped with an x60/1.49 Apo TIRF oil
objective and a Cascade II: 512 EMCCD camera (Photometrics, Tucson). Then livers
and kidneys of mice were frozen and cut into sections and stained with hematoxylin and
eosin (H&E) for pathology analysis.
RNA isolation and transcripts analysis by RT-qPCR
Total tissue RNA was extracted from the flank tumor tissue using an RNeasy Mini Kit
(Qiagen, Valencia, CA), according to the manufacturer’s protocol. The cDNAs were
synthesized from equal amounts of total RNAs using the High-Capacity RNA-to-cDNA
Kit (Applied Biosystems, Grand Island, NY). Real-time qPCR with the appropriate
primers was used to measure the expression of Perforin, IL-12p35, IL-12p40, ICOS,
FAP, IFN-γ, MMP2, MMP9 and TNF-α genes. An ABI 7300 Real-Time PCR System
(Applied Biosystems) was used for real-time qPCR to measure the incorporation of
SYBR Green (Applied Biosystems). The ΔΔCt method was used to calculate changes in
74
gene expression level, and the raw values were normalized to the levels of GAPDH as a
reference gene.
Statistical analysis
Statistical analysis was performed by GraphPad (Prism) software to determine p values
by Student’s t-test where two groups were compared. When more than two groups were
compared, one-way ANOVA with the Tukey posttest was used to determine significant
differences between individual groups. Tumor growth curves were compared with two-
way ANOVA. Kaplan-Meier analysis was used to evaluate the survival of mice. A p
value below 0.05 was considered statistically significant, and data were presented as
means ± SEM.
2.3: Results
Construction and in vitro cytotoxicity of αFAP-PE38 and α-FAP-PE38hu
We made two changes to the previously described αFAP-PE38
26
in order to construct
α-FAP-PE38hu. First, the linker between the scFV and toxin domains was changed from
the human CD8 hinge to a furin-cleavable linker. Second, domain II of αFAP-PE38 was
removed and the seven point mutations described by Liu et al
31
were incorporated into
the truncated αFAP-PE38 in an effort to mitigate the immunogenicity of our immunotoxin
therapeutic, α-FAP-PE38hu (Fig. 1B).
75
The binding specificity of α-FAP-PE38hu to human and murine FAP-expressing 293T
cells was determined by using flow cytometry to measure the fluorescence of dye-
labeled immunotoxin bound to FAP. The K
D
of the interaction between the αFAP-PE38
or α-FAP-PE38hu and the FAP-expressing 293T cells was determined by Lineweaver-
Burk analysis.
36, 37
The K
D
of α-FAP-PE38hu was found to be 2.4 ± 0.6 nM (Fig. 1C) for
murine FAP (mFAP) and and 7.7 ± 0.4 nM (data not shown) for human FAP (hFAP). We
observed that α-FAP-PE38hu and αFAP-PE38 had similar binding affinities towards
mFAP and hFAP (data not shown), indicating that the modifications that were made to
αFAP-PE38 to create α-FAP-PE38hu have a minimal effect on its binding affinity.
Next, the in vitro cytotoxic effects of α-FAP-PE38hu on FAP-expressing cells was
evaluated by XTT assay. α-FAP-PE38hu was shown to have an IC
50
of 255 ng/mL and
3.18 µg/mL against 293T cells expressing mFAP and hFAP respectively, and showed
little to no toxicity towards 293T cells that did not express any form of FAP (Fig. 1D).
These results indicate that the modifications included in α-FAP-PE38hu did not
negatively impact its toxicity towards FAP-expressing cells and did not make it more
prone to off-target toxicity.
Both α-FAP-PE38hu and αFAP-PE38 restrain tumor growth in vivo
To determine the antitumor effect of the α-FAP-PE38hu compared to the original αFAP-
PE38 in vivo, we first subcutaneously inoculated C57BL/6 mice with 2x10
5
B16-F10
cells at the right flank. The original αFAP-PE38 (0.5 mg/kg) or α-FAP-PE38hu (0.5
76
mg/kg and 1.5 mg/kg) were then intravenously injected every other day from 12 days
after tumor inoculation for a total of four injections (Fig. 2A). Tumor volume and body
weight were measured every other day throughout the injections. PBS was injected as a
negative control. Mice receiving immunotoxin injections had inhibited tumor growth
compared to the control group. There was no significant difference between the tumor
sizes of the mice receiving injections of the original αFAP-PE38 and α-FAP-PE38hu
(Fig. 2B). Also, the mice receiving the higher dose of α-FAP-PE38hu did not show any
additional restraint of tumor growth (Fig. 2B). Based on these results, we used the 0.5
mg/kg dose of α-FAP-PE38hu for subsequent experiments. Furthermore, no significant
weight loss was observed in any group throughout the experiment (Fig. 2C). H&E
stainings of liver and kidney tissue indicated that no off-target toxicity was present and
further confirmed the safety of the selected dosages of α-FAP-PE38hu (Fig. 2D).
The pharmacokinetics of α-FAP-PE38hu were determined by first injecting C57BL/6
mice with a single dose of 10 µg of either α-FAP-PE38hu or αFAP-PE38 via the tail
vein. Blood samples were drawn at different time intervals between 2 mins and 60 mins
after injection and the concentration of immunotoxin was measured by His-Tag ELISA.
Data were fit into a single exponential decay function (Fig. 2E). We measured the half
lives of α-FAP-PE38hu and αFAP-PE38 to be 9.1 min and 15.3 min respectively,
suggesting that the alterations to the original αFAP-PE38 may have impacted its in vivo
half life.
77
Combinatorial therapy of PD-1 blockade and α-FAP-PE38hu shows improved
antitumor activity
Although PD-1 blockade therapy has been shown to have promising clinical results,
blockade of a single inhibitory pathway is not sufficient to overcome the attenuation of T
cell effector function by the immunosuppressive TME.
2, 6-9
To confirm our hypothesis
that α-FAP-PE38hu can eliminate FAP
+
stromal cells in TMEs and boost the efficacy of
PD-1 blockade by helping recover T cell effector functions, we first injected C57BL/6
mice with 2x10
5
B16-F10 cells at the right flank. α-FAP-PE38hu (0.5 mg/kg) or PD-1
antibody (10 mg/kg) or both were then intravenously injected every other day from 12
days after tumor inoculation for a total of four injections (Fig. 3A). PBS was injected in
the control group. Combinatorial injection of PD-1 antibody and α-FAP-PE38hu showed
a significant 4-fold tumor reduction (p< 0.001) compared to the control group (Fig. 3B).
Combinatorial therapy also displayed enhanced tumor suppression compared with
single injection groups (p< 0.01) (Fig. 3B). Additionally, no significant change of body
weight was observed (Fig. 3C). TUNEL staining of tumor tissues from day 22 also
indicated that both PD-1 blockade and α-FAP-PE38hu led to cell death and that
combining the therapies resulted in an even higher amount of apoptotic tumor cells (Fig.
3D).
78
Combinatorial therapy of PD-1 blockade and α-FAP-PE38hu alters
immunosuppressive TMEs to promote immune stimulatory TMEs.
To further verify that α-FAP-PE38hu can alter immunosuppressive TMEs to help PD-1
antibody to restore T cell effector function, we analyzed the TILs of treated mice at day
22. The α-FAP-PE38hu group showed a noticeable increase in the percentage of CD4
+
T
cells and the PD-1 antibody group displayed a pronounced rise in the percentage of
CD8
+
T cells (Fig. 4A and Fig. 4E), which is consistent with the results from Curran et
al.
38
Meanwhile, there was an evident additional increase in the percentage of CD8
+
and
CD4
+
T cells in the combinatorial treatment group (p< 0.001) (Fig. 4A and Fig. 4E). The
ratio of T cells to Tregs has been related to tumor development and repression in both
of mice and humans.
39, 40
The α-FAP-PE38hu group demonstrated an enhancement of
both CD8/Treg and CD4/Treg ratios, and the PD-1 antibody group also experienced a
noticeable increase in CD8/Treg ratio (Fig. 4B, Fig. 4F). The combinatorial treatment
group exhibited a further improvement in CD8/Treg and CD4/Treg ratios compared to
either of the single groups (p< 0.001) (Fig. 4B, Fig. 4F). Besides Tregs, MDSC ratios
were also investigated as a major composition of immunosuppressive TMEs. α-FAP-
PE38hu led to enhanced CD8/MDSC and CD4/MDSC ratios and the combinatorial
group displayed an additional increase of both ratios compared to the control group (p<
0.001) (Fig. 4C, Fig. 4G). Finally, the expression of PD-1, which functions as an
inhibitory pathway receptor on TILs was measured. The PD-1 antibody displayed an
evident upregulation of PD-1 in CD8
+
TILs and PD-1 in CD4
+
TILs and the combined
group showed similar results (Fig. 4D, Fig. 4H).
79
Next, we measured the gene expression of T cell activation markers and associated
cytokines to further investigate changes of TMEs by using real time-qPCR. Inducible T-
cell costimulator (ICOS) expression level was remarkably increased in the combined
treatment group compared with both single treatment groups (p<0.001) (Fig. 5A). IL-
12p70 is a heterodimer that consists of IL-12p35 and IL-12p40 subunit and promotes
Th1 polarization and cell-mediated immunity.
41, 42
The combination therapy elevated the
expression levels of both IL-12p35 and IL-12p40 (Fig. 5B, Fig. 5C). Perforin, a pore
forming protein in the granules of cytotoxic T lymphocytes, and TNF-α, a cytokine that
can cause acute and hypoxic death of cancer cells,
43
were all upregulated in both single
treatment groups. Combinatorial groups displayed further elevation of expression levels
(Fig. 5D, Fig. 5E).
Combinatorial therapy of α-FAP-PE38hu and CAR T cells demonstrates an
enhanced antitumor effect
After validating the ability of α-FAP-PE38hu to enhance the antitumor efficacy of PD-1
blockade in vivo, we wanted to confirm that the synergistic effects were not limited to
checkpoint inhibitor therapy. To address this, we chose to use α-FAP-PE38hu in
combination with another commonly used form of immunotherapy: CAR T cells. We first
inoculated 5x10
5
B16-CD19 cells into the right flanks of female C57BL/6 mice. On Day
4, tumor-bearing mice were irradiated with 500 cGy of X-ray radiation in order to enable
the adoptive transfer of CD19-targeting CAR T cells. Afterwards, the mice were
80
randomly divided into four different therapy groups (n = 8): one group only receiving
nontransduced T cells only (NT), one group receiving nontransduced T cells and α-FAP-
PE38hu, one group receiving CAR T cells only, and one group receiving both CAR T
cells and α-FAP-PE38hu. We then administered 20 ug of α-FAP-PE38hu or PBS on
days 5, 7, and 9, via tail vein injections. We also injected 10x10
6
CD19-targeting CAR T
cells (30% CAR positive) or nontransduced mouse T cells on day 6 via the tail vein (Fig.
6A). Although animals in all four groups showed tumor growth progression, the groups
receiving either α-FAP-PE38hu or CAR T cell therapy alone demonstrated a decreased
tumor growth rate compared to the NT group (p < 0.0001 and p = 0.0150 respectively,
two-way ANOVA). The combination therapy group the highest amount of tumor growth
inhibition (p < 0.0001 for NT, p = 0.0032 for α-FAP-PE38hu, p < 0.0001 for CAR T cells,
two-way ANOVA), indicating that there is a synergistic, antitumor effect between α-FAP-
PE38hu and CAR T cell therapy (Fig. 6B). We also observed no significant difference
between the body weights of each group during the course of treatment and tumor
growth monitoring - indicating that the combined therapy was well tolerated (Fig. 6C).
The decreased tumor growth rate of the combination therapy group also translated to a
statistically significant improvement in overall survival rate compared to each other
group (p < 0.0001 for NT, p = 0.0032 for α-FAP-PE38hu, and p = 0.0009 for CAR T, log-
rank test), further providing evidence of the synergy between the two therapies (Fig.
6D).
α-FAP-PE38hu supports CAR T cell function by modulating the tumor
microenvironment
81
We performed an ex vivo analysis of tumor tissues on Day 12 in order to elucidate
whether α-FAP-PE38hu was able to alter the tumor microenvironment by targeting
CAFs. We observed that the mice receiving injections of α-FAP-PE38hu had a
significantly higher percentage of CD3
+
CD45
+
TILs compared to the NT group (Fig. 7A,
p = 0.0121 for α-FAP-PE38hu only, and p = 0.0105 for α-FAP-PE38hu and CAR T cells,
one-way ANOVA). While analysis of the tumor tissues appeared to show that combining
α-FAP-PE38hu with CAR T cells raised the percentage of TILs when compared to the
group receiving CAR T cells alone, the difference was found not to be statistically
significant (Fig. 7A, p = 0.0575). Although we observed no difference in the populations
of Tregs between each group (data not shown), we found that the combination of α-
FAP-PE38hu and CAR T cells had the lowest amount of MDSCs compared to the NT (p
= 0.0035, one-way ANOVA) and CAR T cell (p = 0.0096) groups (Fig. 7B). However,
despite the appearance of a decrease in the MDSC population, there was no
statistically significant difference between the combination therapy group and the α-
FAP-PE38hu group (p = 0.1283).
Finally, we sought to explore α-FAP-PE38hu's mechanism of action by performing RT-
qPCR on the tumor tissues harvested on Day 12. We observed no significant difference
in the expression levels of ICOS and perforin between any two groups (Figs. 8A, 8B).
However, the group receiving the combined therapy had significantly higher levels of
IFN-γ, IL-12, and TNF-α than both the NT and CAR T cell therapy groups (Figs. 8C-F).
We also noticed a decrease in the expression of FAP for every therapy group when
82
compared to NT, with the combination therapy group exhibiting the largest decrease in
FAP expression (Fig. 8G). MMP2 and MMP9, which are enzymes that contribute to
tumor invasiveness and migration by degrading the extracellular matrix,
44-46
were found
to be significantly lower in the CAR T cell and combination therapy groups (Figs. 8H,
8I). Taken as a whole, these results indicate that combining α-FAP-PE38hu with CAR T
cell therapy results in an additive, synergistic improvement in the secretion of the pro-
immune cytokines: IL-12, TNF-α, and IFN-γ, as well as a decreased expression of
genes that are typically expressed within the tumor stroma (FAP, MMP2, MMP9),
supporting our hypothesis that α-FAP-PE38hu is able to support the function of CAR T
cell therapy, resulting in a combined therapy that is more effective than either singular
therapy.
2.4: Discussion and Conclusions
Our previous experiments with αFAP-PE38 exhibited its high affinity for FAP in vitro,
specific killing of FAP
+
cells, and robust antitumor activity in a 4T1 mouse breast cancer
model, which makes it a promising translational approach for treating human cancers.
26
However, the formation of neutralizing antibodies due to the immunogenicity of
immunotoxins is a major challenge that must be addressed in order to achieve
successful clinical applications. Liu et al used phage display and made point mutations
to identify and remove human B-cell epitopes in the PE38 immunotoxin and greatly
reduced its immunogenicity.
31
Previous research has also shown that domain II of PE38
was responsible for translocation into the cytosol.
33
However, we were able to remove
83
domain II and further mitigate any potential immunogenicity due to the presence of the
αFAP targeting moiety in our immunotoxin, and its ability to both target and facilitate
endocytosis into FAP
+
cells. In this study, we adopted these point mutations and deleted
domain II of PE38 to ultimately develop α-FAP-PE38hu: an improved, less
immunogenic, FAP-targeting immunotoxin.
Effective endocytosis of immunotoxins through antibody binding is crucial for the
efficacy of the therapy. Although we made no changes to the targeting modality (the α-
FAP scFv) of our immunotoxin, we measured the binding affinity of α-FAP-PE38hu and
compared it with that of the original αFAP-PE38 to confirm that the elimination of the
human B cell epitopes on the PE38 toxin had no effect on the binding affinity. Our
results indicated similar binding affinities between the original αFAP-PE38 and α-FAP-
PE38hu (data not shown), confirming that point mutations of PE38 did not significantly
impact the binding affinity of α-FAP-PE38hu. Despite having no effect on its binding
affinity, we observed that in regards to 293T cells expressing mFAP, the cytotoxicity of
α-FAP-PE38hu (IC
50
= 255 ng/mL, Fig. 1D) was greatly compromised in vitro - about 60
times lower than the previously reported IC
50
of 4 ng/mL for the original αFAP-PE38
26
.
However, in our in vivo study, there was no significant difference in the extent of tumor
regression between the original αFAP-PE38 and α-FAP-PE38hu (Fig. 2B). We expect
this discrepancy to be a result of two factors. First, the furin cleavable site of α-FAP-
PE38hu is located in the linker region and exposed, rather than hidden inside domain II
of PE38, which was the case for the original αFAP-PE38. Second, furin is a type I
transmembrane protein that is expressed by 293T cells.
47, 48
Therefore, the highly
84
exposed furin cleavable site, which is very susceptible to cleavage by the furin secreted
into the culture medium, may cause the immunotoxin to lose its targeting specificity
before endocytosis takes place and drastically decrease its cytotoxicity.
We also reported that α-FAP-PE38hu rendered the same in vivo antitumor effect as the
original αFAP-PE38 (Fig. 2B), showed improved antitumor efficacy with PD-1 blockade
(Fig. 3B), and enhanced the antitumor efficacy of CAR T cells (Fig. 6B). Furthermore,
H&E staining (Fig. 2D) and body weight monitoring from all in vivo experiments (Figs.
2C, 3C, 6C) showed no signs of off-target toxicity, which is consistent with the results
from previous studies involving FAP-targeting therapeutics.
24, 25, 49
Taken together,
these results indicate that α-FAP-PE38hu is able to enhance the antitumor effects of
immunotherapy .
Our next goal was to interrogate the mechanisms that enabled α-FAP-PE38hu to work
synergistically with PD-1 blockade and CAR T cell therapy to inhibit tumor growth. While
it is undeniable that checkpoint inhibitor therapy, such as PD-1 blockade, and CAR T
cell therapy have demonstrated an efficacious inhibition of tumor growth from clinical
trials, immunosuppressive TMEs can undermine the therapeutic effect through various
inhibitory pathways and promote tumor growth.
50, 51
For instance, immunosuppressive
TMEs secrete paracrine factors to inhibit effector T cell recruitment and inhibit pro-
inflammatory cytokine secretion to further impair effector T cell function.
52
CAFs secrete
CCL2, CCL3, CCL4 and CCL5 to increase the recruitment of myeloid cells such as
MDSCs and repel T cells by secreting CXCL12.
22, 23
CAFs also reduce the expression
85
of E-selectin on effector T cells and lower the expression of intercellular adhesion
molecule 1 (ICAM-1) and vascular cell adhesion molecule 1 (VCAM-1) on BECs to
hinder effector T cell extravasation. CAFs have been shown to upregulate mucosal
vascular addressin cell adhesion molecule 1 (MADCAM1) and CD166, which bind α4β7
integrin and CD6 respectively on Tregs to promote their extravasation.
20, 52
In this study, α-FAP-PE38hu, when combined with PD-1 blockade, successfully
inhibited tumor growth (Fig. 3), reduced the recruitment of MDSCs and Tregs into the
TME (Fig. 4), increased the recruitment of effector T cells (Fig. 4), and increased the
expression of pro-inflammatory cytokines within the tumor (Fig. 5). PD-1 expression in
CD8
+
and CD4
+
T effector cells was upregulated in the combined group, indicating that
effector T cells were accumulated and proliferated in the intratumoral region.
38
Similarly,
a combined therapy of α-FAP-PE38hu with CAR T cells inhibited tumor growth and
increased overall survival (Fig. 6) and reduced the amount of MDSCs in the TME while
increasing the amount of TILs (Fig. 7). Furthermore, the combined therapy was shown
to unite the benefits of CAR T cell therapy (decreased FAP, MMP2, and MMP9
expression) with the improved secretion of pro-inflammatory cytokines facilitated by α-
FAP-PE38hu - resulting in the enhanced inhibition of tumor growth discussed earlier
(Fig. 8). We speculate that the lack of improvement in ICOS and perforin expression
may be attributed to the lingering effects of partial myeloablation by X-ray irradiation, as
irradiation occurred 8 days before ex vivo tumor analysis. These results show that
combining an FAP-targeting immunotoxin with PD-1 blockade or CAR T cell therapy,
86
can restrain tumor growth by destroying a barrier to effector T cell function and
promoting the activation and recruitment of TILs.
We successfully developed α-FAP-PE38hu, a modified, less immunogenic immunotoxin
that targets tumor stromal cells, by introducing point mutations to known B cell epitopes
and removing domain II of PE38. We demonstrated that α-FAP-PE38hu bound to FAP
expressing cells with high affinities, specifically killed FAP
+
cells in vitro, mitigated some
of the immune suppressive effects of the tumor microenvironment, and showed
improved antitumor effects when combined with PD-1 blockade or CAR T cell therapy.
Our pharmacokinetic data indicated that α-FAP-PE38hu has an in vivo half-life of 9.1
min while the original αFAP-PE38 has an in vivo half-life of 15.3 min (Fig. 2E). The
decreased half-life of α-FAP-PE38hu is due to its smaller molecular weight of ~53 KDa
compared to that of original αFAP-PE38 which is 75 KDa, and thus makes α-FAP-
PE38hu susceptible to glomerular filtration.
53
PEGylation, which can improve a
therapeutic's circulation time and antitumor activity, is a possible avenue for
improvement of the current immunotoxin.
54
However, regardless of its decreased half-
life, we have shown that α-FAP-PE38hu works synergistically with two widely-adopted
forms of immunotherapy to combat tumor growth. The results discussed in this report
demonstrate the therapeutic potential of tumor stroma-targeting immunotoxins when
used to increase the efficacy of immunotherapy.
87
2.5: Figures
Figure 1. Illustration and Characterization of α-FAP-PE38hu
(A) Schematic illustration of the combination of α-FAP-PE38hu and PD-1 blockade.
(B) Schematic drawings of original αFAP-PE38 and α-FAP-PE38hu.
(C) Measurement of binding affinity of α-FAP-PE38hu toward mFAP
+
cells in vitro. The
KD of the interaction between α-FAP-PE38hu and cell-surface mFAP, as determined by
Lineweaver-Burk analysis. All the assays were conducted in triplicate for each cell line.
(D) Measurement of cell cytotoxicity of α-FAP-PE38hu. The cell cytotoxicity of α-FAP-
PE38hu against 293T, 293T-mFAP and 293T-hFAP cells was performed by a standard
XTT assay with a 48-hr treatment procedure. Data are given as an IC
50
: the
concentration of immunotoxin that causes 50% cell death after a 48-hr incubation with
immunotoxin. All the assays were conducted in triplicate for each cell line. Data are
shown as mean ± SEM.
88
Figure 2. Antitumor Efficacy of the Original αFAP-PE38 and α-FAP-PE38hu in B16-F10
Tumor-bearing mice
(A) Schematic representation of the in vivo α-FAP-PE38hu treatment protocol. Female
C57BL6 mice were inoculated s.c. with 2×10
5
B16-F10 cells in the right flank and
randomized into four groups (n = 6 per group). Each group was then treated with its
respective immunotoxin: original αFAP-PE38 (0.5 mg/kg) or α-FAP-PE38hu (0.5
mg/kg)/ (1.5 mg/kg) 12 days after tumor implantation through i.v. injection for total of
four times at the indicated days.
Tumor volume (B) and body weight (C) were monitored every other day posttreatment.
Tumor and body weight values are reported as mean ± SEM;
*
p < 0.05,
**
p < 0.01,
***
p<
0.001.
(D). H&E staining was performed on the kidney and liver of untreated mice and treated
mice with the original αFAP-PE38 or α-FAP-PE38hu on day 22. The scale bar
represents 20 µm
(E) Pharmacokinetics of original αFAP-PE38 and α-FAP-PE38hu. C57BL/6 mice were
injected intravenously with 10 µg of either original or mutant αFAP-PE38 and bled at
several intervals between 2 and 60 minutes from the time of injection. The concentration
of the immunotoxin in the plasma at the various intervals was determined by His-Tag
ELISA and fit to a single exponential decay function. The corresponding half-life (t
1/2
) is
indicated. Immunotoxin concentrations are reported as mean ± SEM.
89
Figure 3. Combinatorial Antitumor Efficacy of α-FAP-PE38hu and PD-1 Blockade
(A) Schematic representation of the in vivo α-FAP-PE38hu and PD-1 blockade
treatment protocol. Female C57BL6 mice were inoculated s.c. with 2×10
5
B16-F10 cells
in the right flank and randomized into four groups (n = 6 per group). They were then
treated with α-FAP-PE38hu (0.5 mg/kg), αPD-1(10mg/kg) or combined injections 12
days after tumor implantation through i.v. injections for a total of four injections on the
indicated days.
Tumor volume (B) and body weight (C) were monitored every other day posttreatment.
Tumor and body weight values are reported as mean ± SEM; *p < 0.05, **p < 0.01,
***p< 0.001.
(D). Representative images of apoptosis in tumor sections. Cell apoptosis was detected
by TUNEL staining (nuclei stained with DAPI, blue; apoptotic cells stained with FITC,
green). The scale bar represents 50 µm.
90
Figure 4. Combinatorial Therapy of α-FAP-PE38hu and PD-1 Blockade Increases
Tumor-Infiltrating T Cells and Decreases the Ratio of Immune suppressive Cells to
Effector T cells
Tumor tissues of treated mice from day 22 were harvested and purified. Cells were
stained by various markers, and analyzed by flow cytometry for the composition of
various subsets of immune cells. (A, E) Percentages of CD8
+
and CD4
+
T cells within
CD45
+
TILs. (B, F) The ratios of CD8
+
and CD4
+
T cells to CD4
+
CD25
+
FoxP3
+
Treg
cells. (C, G). The ratios of CD8
+
and CD4
+
T cells to CD45
+
CD11b
+
F4/80
+
ly-6C
+
ly-6G
+
MDSCs. (D, H) Percentages of CD8
+
and CD4
+
T cells expressing PD-1 within CD45
+
TILs. The data shown were individually analyzed from mice that received the indicated
therapy, and one-way ANOVA was used to determine the statistical significance
between samples (n = 6, Data are shown as mean ± SEM; *p < 0.05, **p < 0.01, ***p<
0.001).
91
Figure 5. Combinatorial Therapy of α-FAP-PE38hu and PD-1 Blockade Altered
Immunosuppressive Tumor Microenvironment.
mRNA expression levels of (A) ICOS, (B) IL-12 p35, (C) IL-12 p40, (D) Perforin, and (E)
TNF-α from treated mice were measured at day 22. Three tumors from each group were
resected and homogenized. Total RNA was extracted, and the mRNA expression levels
were determined by real-time qPCR. Graph depicts relative transcript expression genes
after normalizing to GAPDH mRNA levels (Data are shown as mean ± SEM; *p < 0.05,
**p < 0.01, ***p< 0.001).
92
Figure 6. Combinatorial Antitumor Efficacy of α-FAP-PE38hu and CAR T Cells
(A) Schematic representation of the in vivo α-FAP-PE38hu and CAR T cell treatment
protocol. Female C57BL6 mice were inoculated s.c. with 5×10
5
B16-CD19 cells in the
right flank and randomized into four groups (n = 8 per group). They were then irradiated
with 500 cGy of X-ray irradiation on day 4 and treated with 20 µg of α-FAP-PE38hu and
10 x 10
6
CAR T cells intravenously via tail vein on the indicated days, beginning 5 days
after tumor implantation.
Tumor volume (B) and body weight (C) were monitored every other day posttreatment.
Tumor and body weight values are reported as mean ± SEM, and statistical significance
was determined by two-way ANOVA.
(D) Kaplan-Meier curve showing the amount of mice with a tumor volume below 1,000
mm
3
. Significance was determined by log-rank test. (ns = not significant, *p < 0.05, **p <
0.01, ***p< 0.001, ****p< 0.0001).
93
Figure 7. Combinatorial Therapy of α-FAP-PE38hu and CAR T Cells Increases Tumor
Infiltration of T Cells and Decreases MDSCs Presence within the Tumor
Tumor tissues of treated mice from day 12 were harvested and purified. Cells were
stained by various markers, and analyzed by flow cytometry for the composition of
various subsets of immune cells. (A) Percentage of MDSCs. (B) Percentage of CD3
+
,
CD45
+
TILs. The data shown were individually analyzed from mice that received the
indicated therapy, and one-way ANOVA was used to determine the statistical
significance between samples (n = 3, Data are shown as mean ± SEM; ns = not
significant, *p < 0.05, **p < 0.01, ***p< 0.001).
94
Figure 8. Combinatorial Therapy of α-FAP-PE38hu and CAR T Cells Counteract the
Immunosuppressive Tumor Microenvironment.
mRNA expression levels of (A) ICOS, (B) Perforin, (C) IFN-γ, (D) IL-12 p35, (E) IL-12
p40, (F) TNF-α, (G) FAP, (H) MMP2, and (I) MMP9 from treated mice were measured at
day 12. Three tumors from each group were resected and homogenized. Total RNA
was extracted, and the mRNA expression levels were determined by real-time qPCR.
Graph depicts relative transcript expression genes after normalizing to GAPDH mRNA
levels (Data are shown as mean ± SEM; statistical significance was determined by one-
way ANOVA; ns = not significant, *p < 0.05, **p < 0.01, ***p< 0.001, ****p< 0.0001).
95
2.6: References
1. Keir, ME, Butte, MJ, Freeman, GJ, and Sharpe, AH (2008). PD-1 and its ligands in
tolerance and immunity. Annual review of immunology 26: 677-704.
2. Sharma, P, and Allison, JP (2015). The future of immune checkpoint therapy. Science
348: 56-61.
3. Lipson, EJ, Sharfman, WH, Drake, CG, Wollner, I, Taube, JM, Anders, RA, et al. (2013).
Durable cancer regression off-treatment and effective reinduction therapy with an anti-
PD-1 antibody. Clinical cancer research : an official journal of the American Association
for Cancer Research 19: 462-468.
4. Brahmer, JR, Tykodi, SS, Chow, LQM, Hwu, W-J, Topalian, SL, Hwu, P, et al. (2012).
Safety and Activity of Anti–PD-L1 Antibody in Patients with Advanced Cancer. N Engl J
Med 366: 2455-2465.
5. Wolchok, JD, Neyns, B, Linette, G, Negrier, S, Lutzky, J, Thomas, L, et al. (2010).
Ipilimumab monotherapy in patients with pretreated advanced melanoma: a randomised,
double-blind, multicentre, phase 2, dose-ranging study. Lancet Oncol 11: 155-164.
6. Mellman, I, Coukos, G, and Dranoff, G (2011). Cancer immunotherapy comes of age.
Nature 480: 480-489.
7. Drake, CG (2012). Combination immunotherapy approaches. Annals of oncology :
official journal of the European Society for Medical Oncology 23 Suppl 8: viii41-46.
8. Topalian, SL, Hodi, FS, Brahmer, JR, Gettinger, SN, Smith, DC, McDermott, DF, et al.
(2012). Safety, Activity, and Immune Correlates of Anti–PD-1 Antibody in Cancer. N Engl
J Med 366: 2443-2454.
96
9. Larkin, J, Chiarion-Sileni, V, Gonzalez, R, Grob, JJ, Cowey, CL, Lao, CD, et al. (2015).
Combined Nivolumab and Ipilimumab or Monotherapy in Untreated Melanoma. N Engl J
Med 373: 23-34.
10. Grupp, SA, Kalos, M, Barrett, D, Aplenc, R, Porter, DL, Rheingold, SR, et al. (2013).
Chimeric antigen receptor-modified T cells for acute lymphoid leukemia. N Engl J Med
368: 1509-1518.
11. Maude, SL, Frey, N, Shaw, PA, Aplenc, R, Barrett, DM, Bunin, NJ, et al. (2014).
Chimeric antigen receptor T cells for sustained remissions in leukemia. N Engl J Med
371: 1507-1517.
12. D'Aloia, MM, Zizzari, IG, Sacchetti, B, Pierelli, L, and Alimandi, M (2018). CAR-T cells:
the long and winding road to solid tumors. Cell death & disease 9: 282.
13. Gilham, DE, Debets, R, Pule, M, Hawkins, RE, and Abken, H (2012). CAR-T cells and
solid tumors: tuning T cells to challenge an inveterate foe. Trends in molecular medicine
18: 377-384.
14. Newick, K, O'Brien, S, Moon, E, and Albelda, SM (2017). CAR T Cell Therapy for Solid
Tumors. Annual review of medicine 68: 139-152.
15. Kalluri, R, and Zeisberg, M (2006). Fibroblasts in cancer. Nature reviews Cancer 6: 392-
401.
16. Coussens, LM, Zitvogel, L, and Palucka, AK (2013). Neutralizing Tumor-Promoting
Chronic Inflammation: A Magic Bullet? Science 339: 286-291.
17. Leek, RD, Lewis, CE, Whitehouse, R, Greenall, M, Clarke, J, and Harris, AL (1996).
Association of Macrophage Infiltration with Angiogenesis and Prognosis in Invasive
Breast Carcinoma. Cancer Research 56: 4625-4629.
18. Zang, X, Sullivan, PS, Soslow, RA, Waitz, R, Reuter, VE, Wilton, A, et al. (2010). Tumor
associated endothelial expression of B7-H3 predicts survival in ovarian carcinomas.
97
Modern pathology : an official journal of the United States and Canadian Academy of
Pathology, Inc 23: 1104-1112.
19. Rodig, N, Ryan, T, Allen, JA, Pang, H, Grabie, N, Chernova, T, et al. (2003). Endothelial
expression of PD-L1 and PD-L2 down-regulates CD8+ T cell activation and cytolysis.
European journal of immunology 33: 3117-3126.
20. Griffioen, AW, Damen, CA, Blijham, GH, and Groenewegen, G (1996). Tumor
Angiogenesis Is Accompanied by a Decreased Inflammatory Response of Tumor-
Associated Endothelium. Blood 88: 667-673.
21. Griffioen, AW, Damen, CA, Martinotti, S, Blijham, GH, and Groenewegen, G (1996).
Endothelial Intercellular Adhesion Molecule-1 Expression Is Suppressed in Human
Malignancies: The Role of Angiogenic Factors. Cancer Research 56: 1111-1117.
22. Paunescu, V, Bojin, FM, Tatu, CA, Gavriliuc, OI, Rosca, A, Gruia, AT, et al. (2011).
Tumour-associated fibroblasts and mesenchymal stem cells: more similarities than
differences. Journal of cellular and molecular medicine 15: 635-646.
23. Kim, JH, S.H., O, Kim, EJ, Park, SJ, Hong, SP, Cheon, JH, et al. (2012). The role of
myofibroblasts in upregulation of S100A8 and S100A9 and the differentiation of myeloid
cells in the colorectal cancer microenvironment. Biochemical and biophysical research
communications 423: 60-66.
24. Loeffler, M, Kruger, JA, Niethammer, AG, and Reisfeld, RA (2006). Targeting tumor-
associated fibroblasts improves cancer chemotherapy by increasing intratumoral drug
uptake. The Journal of clinical investigation 116: 1955-1962.
25. Liao, D, Luo, Y, Markowitz, D, Xiang, R, and Reisfeld, RA (2009). Cancer associated
fibroblasts promote tumor growth and metastasis by modulating the tumor immune
microenvironment in a 4T1 murine breast cancer model. PLoS One 4: e7965.
98
26. Fang, J, Xiao, L, Joo, KI, Liu, Y, Zhang, C, Liu, S, et al. (2016). A potent immunotoxin
targeting fibroblast activation protein for treatment of breast cancer in mice. International
journal of cancer 138: 1013-1023.
27. Pai, LH, Bookman, MA, Ozols, RF, Young, RC, Smith II, JW, Longo, DL, et al. (1991).
Clinical Evaluation of Intraperitoneal Pseudomonas Exotoxin Immunoconjugate OVB3-
PE in Patients With Ovarian Cancer. Journal of clinical oncology : official journal of the
American Society of Clinical Oncology 9: 2095-2103.
28. Pai, LH, Wittes, R, Setser, A, Willingham, MC, and Pastan, I (1996). Treatment of
advanced solid tumors with immunotoxin LMB-1: An antibody linked to Pseudomonas
exotoxin. Nature medicine 2: 350-353.
29. Hassan, R, and Ho, M (2008). Mesothelin targeted cancer immunotherapy. European
journal of cancer 44: 46-53.
30. Mazor, R, Onda, M, and Pastan, I (2016). Immunogenicity of therapeutic recombinant
immunotoxins. Immunological Reviews 270: 152-164.
31. Liu, W, Onda, M, Lee, B, Kreitman, RJ, Hassan, R, Xiang, L, et al. (2012). Recombinant
immunotoxin engineered for low immunogenicity and antigenicity by identifying and
silencing human B-cell epitopes. Proceedings of the National Academy of Sciences of
the United States of America 109: 11782-11787.
32. Taupiac, M-P, Bebien, M, Alami, M, and Beaumelle, B (1999). A deletion within the
translocation domain of Pseudomonas exotoxin A enhances translocation effciency and
cytotoxicity concomitantly. Molecular Microbiology 31: 1385-1393.
33. Jinno, Y, Ogata, M, Chaudhary, V, Willingham, MC, Adhya, S, FitzGerald, D, et al.
(1989). Domain II Mutants of Pseudomonas Exotoxin Deficient in Translocation. The
Journal of biological chemistry 264: 15953-15959.
99
34. Yang, L, Yang, H, Rideout, K, Cho, T, Joo, KI, Ziegler, L, et al. (2008). Engineered
lentivector targeting of dendritic cells for in vivo immunization. Nature biotechnology 26:
326-334.
35. Brocks, B, Garin-Chesa, P, Behrle, E, Park, JE, Rettig, WJ, Pfizenmaier, K, et al. (2001).
Species-Crossreactive scFv Against the Tumor Stroma Marker “Fibroblast Activation
Protein” Selected by Phage Display From an Immunized FAP-/- Knock-Out Mouse.
Molecular Medicine 7: 461-469.
36. Xiao, L, Hung, KC, Takahashi, TT, Joo, KI, Lim, M, Roberts, RW, et al. (2013). Antibody-
mimetic ligand selected by mRNA display targets DC-SIGN for dendritic cell-directed
antigen delivery. ACS Chem Biol 8: 967-977.
37. Benedict, CA, MacKrell, AJ, and Anderson, WF (1997). Determination of the binding
affinity of an anti-CD34 single-chain antibody using a novel, flow cytometry based assay.
Journal of Immunological Methods 201: 223-231.
38. Curran, MA, Montalvo, W, Yagita, H, and Allison, JP (2010). PD-1 and CTLA-4
combination blockade expands infiltrating T cells and reduces regulatory T and myeloid
cells within B16 melanoma tumors. Proceedings of the National Academy of Sciences of
the United States of America 107: 4275-4280.
39. Mandl, SJ, Rountree, RB, Dalpozzo, K, Do, L, Lombardo, JR, Schoonmaker, PL, et al.
(2012). Immunotherapy with MVA-BN(R)-HER2 induces HER-2-specific Th1 immunity
and alters the intratumoral balance of effector and regulatory T cells. Cancer
immunology, immunotherapy : CII 61: 19-29.
40. Chen, J, Zhang, L, Wen, W, Hao, J, Zeng, P, Qian, X, et al. (2012). Induction of
HCA587-Specific Antitumor Immunity with HCA587 Protein Formulated with CpG and
ISCOM in Mice. PLOS One 7: 1-11.
100
41. Del Vecchio, M, Bajetta, E, Canova, S, Lotze, MT, Wesa, A, Parmiani, G, et al. (2007).
Interleukin-12: biological properties and clinical application. Clinical cancer research : an
official journal of the American Association for Cancer Research 13: 4677-4685.
42. Fewell, JG, Matar, MM, Rice, JS, Brunhoeber, E, Slobodkin, G, Pence, C, et al. (2009).
Treatment of disseminated ovarian cancer using nonviral interleukin-12 gene therapy
delivered intraperitoneally. The journal of gene medicine 11: 718-728.
43. Kraman, M, Bambrough, PJ, Arnold, JN, Roberts, EW, Magiera, L, Jones, JO, et al.
(2010). Suppression of Antitumor Immunity by Stromal Cells Expressing Fibroblast
Activation Protein–a. Science 330: 827-830.
44. Thaker, PH, Han, LY, Kamat, AA, Arevalo, JM, Takahashi, R, Lu, C, et al. (2006).
Chronic stress promotes tumor growth and angiogenesis in a mouse model of ovarian
carcinoma. Nature medicine 12: 939-944.
45. Ma, XJ, Dahiya, S, Richardson, E, Erlander, M, and Sgroi, DC (2009). Gene expression
profiling of the tumor microenvironment during breast cancer progression. Breast cancer
research : BCR 11: R7.
46. Kessenbrock, K, Plaks, V, and Werb, Z (2010). Matrix metalloproteinases: regulators of
the tumor microenvironment. Cell 141: 52-67.
47. Tay, FP, Huang, M, Wang, L, Yamada, Y, and Liu, DX (2012). Characterization of
cellular furin content as a potential factor determining the susceptibility of cultured
human and animal cells to coronavirus infectious bronchitis virus infection. Virology 433:
421-430.
48. Seidah, NG, Day, R, Marcinkiewicz, M, and Chretien, M (1998). Precursor Convertases:
An Evolutionary Ancient, Cell-Specific, Combinatorial Mechanism Yielding Diverse
Bioactive Peptides and Proteins. Annals New York Academy of Sciences 839: 9-24.
101
49. Wen, Y, Wang, CT, Ma, TT, Li, ZY, Zhou, LN, Mu, B, et al. (2010). Immunotherapy
targeting fibroblast activation protein inhibits tumor growth and increases survival in a
murine colon cancer model. Cancer science 101: 2325-2332.
50. Tlsty, TD, and Coussens, LM (2006). Tumor Stroma and Regulation of Cancer
Development. Annu Rev Pathol Mech Dis 1: 119-150.
51. Bhowmick, NA, Neilson, EG, and Moses, HL (2004). Stromal fibroblasts in cancer
initiation and progression. Nature 432: 332-337.
52. Nummer, D, Suri-Payer, E, Schmitz-Winnenthal, H, Bonertz, A, Galindo, L, Antolovich,
D, et al. (2007). Role of tumor endothelium in CD4+ CD25+ regulatory T cell infiltration
of human pancreatic carcinoma. Journal of the National Cancer Institute 99: 1188-1199.
53. Kontermann, RE (2011). Strategies for extended serum half-life of protein therapeutics.
Current opinion in biotechnology 22: 868-876.
54. Filpula, D, Yang, K, Basu, A, Hassan, R, Xiang, L, Zhang, Z, et al. (2007). Releasable
PEGylation of Mesothelin Targeted Immunotoxin SS1P Achieves Single Dosage
Complete Regression of a Human Carcinoma in Mice. Bioconjugate chemistry 18: 773-
784.
102
Chapter 3: mRNA Display
mRNA display, invented by Roberts and Szostak, is a powerful combinatorial method
that allows for the simultaneous screening of libraries containing over ten trillion (10
12
)
unique peptide ligands
1
. Other commonly used molecular display methods, such as
yeast display and phage display, are only capable of screening libraries with lower
diversities (10
9
for yeast display and 10
11
for phage display)
2, 3
due to the fact that each
cell will express a unique peptide ligand, thereby limiting the overall library complexity
by the size of the culture being used. Similar to other display methods, mRNA display
relies on the ability to identify the sequence of selected ligands. But unlike other
methods, mRNA display generates a covalent linkage between each peptide ligand and
their respective mRNA coding sequence - creating a molecular Rosetta Stone.
Briefly, as shown in figure 1, the mRNA display cycle first begins with the amplification
of a DNA library by polymerase chain reaction (PCR), followed by in vitro transcription
to convert the DNA library into an mRNA library. Afterwards, the mRNA library is ligated
to a puromycin molecule, due to its structural similarity to aminoacyl-tRNA
4
, and
translated in vitro. Upon translation of the final amino acid in each peptide, the
puromycin inserts itself into the ribosome, which then covalently attaches the puromycin
molecule to the nascent peptide - effectively connecting each peptide to the mRNA that
was used to translate them. The peptide-mRNA fusions are reverse transcribed and
then allowed to bind to a protein of interest (though previous work has also been done
to select for RNA-binding peptides)
5
, after which unbound peptides are washed away.
103
The remaining, bound peptides are then PCR amplified and allowed to reenter the
mRNA display cycle.
The mRNA display process can be modified to allow for the selection of peptides with
unique properties. For instance, Howell and coworkers selected for serum-stable,
protease-resistant peptide binders to Gαi1-GDP.
6
Morelli et. al. used mRNA display to
evolve an artificial, thermostable RNA ligase with significantly higher activity at 65 °C.
7
mRNA display can also be used to select for peptides containing unnatural amino acids
via amber suppression codons.
8
Millward and coworkers have shown that peptide
libraries can be chemically cyclized to form macrocycles - demonstrating that mRNA
display allows for the post translational modification of peptide libraries.
9, 10
In this document, we will describe the use of mRNA display to generate therapeutic,
high affinity peptide ligands. In Chapter 4, we focus on the development of peptide
inhibitors for signal transducer and activator of transcription 3 (STAT3) - a transcription
factor that is constitutively active in many human cancers.
11
In Chapter 5, we discuss
the identification of masking peptides that can be used to prevent on-target, off-tumor
side effects in chimeric antigen receptor (CAR) T cells.
104
Figure 1: An mRNA display selection cycle as described by Takahashi and Roberts
12
105
Chapter 4: Identifying Peptide Inhibitors of STAT3 by mRNA Display
4.1: Introduction
Signal transducer and activator of transcription 3 (STAT3) is a transcription factor
that is constitutively active in many human cancers, including breast cancer
13
,
pancreatic cancer
14
, skin cancer
15
, and prostate cancer
16
. Although STAT3 is also
active in healthy cells, it lacks any oncogenic functions because its activation is well-
regulated and not constitutive. The constitutive nature of STAT3 signaling in cancer is
induced by secreted IL-6 family cytokines due to autocrine or paracrine signaling
between cancer cells.
11
This leads to tumor growth via the expression of genes that
promote angiogenesis (VEGF)
17
, survival (survivin, Bcl-xl)
18, 19
, and immune
suppression
20
.
STAT3 consists of five domains that each have a distinct purpose. The N-terminal
domain assists in the formation of a STAT3 tetramer between two STAT3 dimers.
Although tetramer formation helps stabilize the interaction between STAT3 dimers and
DNA, it is not necessary for DNA binding.
21
The coiled-coil domain is responsible for the
phosphorylation and recruitment of STAT3 to the cytoplasmic region of cytokine
receptors.
22
The DNA binding domain is responsible for recognizing and binding to
consensus DNA sequences.
23
The STAT3 SH2 domain is responsible for recognition of
phosphorylated tyrosine residues and mediates the formation of a homodimer upon
106
phosphorylation of tyrosine 705 in the transactivation domain.
24
In this document, we
will focus on the inhibition of STAT3 dimer formation via the SH2 domain.
Mechanistically, the activation of STAT3 begins with the engagement of cytokine or
growth receptors with their corresponding ligands. This prompts kinases, such as the
JAK and SRC family kinases, to phosphorylate the cytoplasmic tail of the receptors -
providing a docking site for STAT3 via its Src homology 2 (SH2) domain.
11
Binding to
the phosphorylated receptor leads to the phosphorylation of tyrosine 705 in the C-
terminal domain of STAT3. Phosphorylated STAT3 proteins then form a homodimer via
reciprocal interactions between their SH2 domains and phosphorylated tyrosine 705
residues.
25
Dimerized STAT3 then translocates to the nucleus where it can bind to
consensus promoter sequences and activate transcription.
26
In principle, every step of malignant STAT3 activation would be a useful target for
inhibition, but in practice, DNA binding inhibitors and dimerization inhibitors have been
the most successful. For example, decoy oligonucleotides are small DNA fragments that
mimic the consensus sequences that the STAT3 DNA binding domain can recognize.
These decoys are able to selectively abrogate activated STAT3 in vitro by binding to
activated STATs in the cytoplasm and preventing them from translocating to the
nucleus.
27
Dimerization inhibitors target the SH2 domain and can be divided into two groups:
peptides/peptidomimetics and small molecules. Many small molecules inhibitors of
STAT3 dimerization have been developed, including Stattic (IC
50
= 5.1µM)
28
, BP-1-102
(IC
50
= 6.8µM)
29
, S3I-201.166 (IC
50
= 35µM)
30
, and LLL12 (IC
50
= 0.9 µM)
31
. Peptide
107
and peptidomimetic inhibitors can have better affinities than small molecules, but they
are not as effective in vitro due to their poor cell permeability. Current peptides that
target the SH2 domain are based on a phosphopeptide sequence (pYLPQTV) derived
from gp130 - a known binding partner of STAT3.
32, 33
This phosphopeptide has been
shown to bind to STAT3 with a K
d
of 150 nM and inhibit STAT3 DNA binding activity via
electrophoretic mobility shift assays (EMSA).
33
Based on this research, Turkson and
coworkers developed additional phosphopeptides aimed at inhibiting SH2 domain
function.
34
However, these phosphopeptides have a high IC
50
in vitro due to the
negative charge of the phosphotyrosine residue severely hindering the peptides' ability
to cross the cell membrane.
Although cytotoxic and capable of inhibiting DNA binding, current STAT3 inhibitors
either have poor cell permeability or have a low binding affinity. In this document, we will
describe using mRNA display to identify high affinity, unphosphorylated peptide
inhibitors of the STAT3 SH2 domain. Cell permeability will be improved by the lack of a
phosphotyrosine residue and can be enhanced through the conjugation of fatty acids to
the peptide.
35
108
4.2: Materials and Methods
Plasmid Preparation
DNA encoding the SH2 domain (residues 582-702) of human STAT3 was PCR
amplified from a pDONR221 plasmid containing the cDNA for human STAT3 (kindly
provided by Dr. Terry Takahashi). The SH2 domain was PCR amplified to include a
SpeI restriction site on the 5' end and a BirA tag (GLNDIFEAQKIEWHE) followed by a
NotI restriction site on the 3' end. This gene was then subcloned into a pET24a vector
containing DNA encoding maltose binding protein (MBP) with a C-terminal 6X histidine
tag. The resulting plasmid (pET24a MBP-SH2) is shown at the end of this section in
Figure S1.
Protein Expression and Purification
pET24a MBP-SH2 was transformed into BL21(DE3) cells containing an expression
vector (chloramphenicol resistant) for biotin ligase. Transformed E. coli were then
inoculated into a 5mL overnight culture of LB containing 100µg/mL of kanamycin and
25µg/mL chloramphenicol. The overnight culture was then used to inoculate 1L of LB
containing 50µM biotin, 100µg/mL kanamycin, and 25µg/mL chloramphenicol, which
was then grown at 37 °C to an OD
600
of 0.6. Expression of MBP-SH2 was induced with
0.5 mM IPTG at 30 °C for 4 hours. The culture was then harvested by centrifugation.
Bacterial pellets were lysed in BPER (Thermo Fisher) containing 1 mM PMSF. The
lysate was then centrifuged to remove the insoluble fraction. The soluble fraction,
containing MBP-SH2, was purified via nickel affinity chromatography. Briefly, the soluble
109
fraction was loaded onto a column containing Nickel-NTA agarose resin (Qiagen) and
washed with wash buffer (50mM NaH2PO4, 300mM NaCl, 10mM Imidazole) followed
by a high salt wash buffer (50mM NaH2PO4, 1M NaCl, 10mM Imidazole). Bound
protein was eluted with a gradient of 0-100% elution buffer (50mM NaH2PO4, 300mM
NaCl, 250mM Imidazole). Fractions that contained MBP-SH2 were pooled,
concentrated, and buffer exchanged into 100mM NaCl, 50mM HEPES, 1mM EDTA,
1mM DTT, 10% Glycerol. The samples were then aliquoted, snap frozen and stored at -
80°C.
Biotinylation Verification
Biotinylation of MBP-SH2 was determined by SDS-PAGE (Figure S2). Briefly, MBP-
SH2 was incubated with streptavidin agarose resin (Thermo Scientific) for 1 hour.
Afterwards, the supernatant (Lane 1) was collected and the beads were washed with
PBST. Then, samples were boiled with 10 nmol excess biotin in solution and that
supernatant was collected as well (Lane 2).
mRNA Display: PCR and Transcription
Doped DNA libraries bearing a constant region on both the 5' and 3' ends were
purchased from Integrated DNA Technologies. Libraries were then PCR amplified and
phenol extracted with an equal volume of phenol:chloroform:isoamyl alcohol (25:24:1).
The aqueous layer was then ethanol precipitated and resuspended in 200 µL of 10 mM
Tris pH 8.5.
110
Transcription of the DNA libraries was performed by mixing 100 uL of the PCR
product with 200 µL 5X transcription buffer (400 mM HEPES–KOH, pH 7.5, 10 mM
spermidine, 200 mM DTT, 125 mM MgCl
2
), 200 µL 5X NTPs (20mM each NTP) and T7
RNA polymerase. Water was added to a final volume of 1mL and the transcription was
allowed to continue overnight at 37 °C. The reaction was stopped by adding 100µL of
0.5M EDTA and vortexing. Afterwards, we phenol extracted and ethanol precipitated the
sample. Then, the sample was resuspended in water and desalted with a Centri-Sep
spin column (Princeton Separations). Finally, mRNA concentration was determined via
nanodrop.
mRNA Display: Ligation
Ligation of the mRNA library to puromycin was performed by first combining 6 nmol
mRNA with 7.2 nmol F30P (Keck Oligo Synthesis Resource) and 8.4 nmol DNA splint to
bridge the mRNA to the F30P. Water was added to bring the final volume (excluding the
T4 DNA ligase and buffer) to 300 µL. The mixture was then heated at 65 °C for 10
minutes and cooled on ice. Then, 30 µL of T4 DNA Ligase buffer (New England Biolabs)
was added, followed by T4 DNA Ligase. The reaction was allowed to occur for 1 hour at
room temperature, and subsequently, 30 minutes at 37 °C. Then, the ligated product
was purified by Urea PAGE and ethanol precipitated.
mRNA Display: Translation and Purification
Ligated mRNA was translated using rabbit reticulocyte lysate. For a 100 µL translation
reaction, we mixed 40 pmol of ligated mRNA, 8 µL of 12.5X Amino Acid Mix, 4 µL 2.5M
KOAc, 2 µL 25mM MgOAc and added rabbit reticulocyte lysate to 100 µL. For a 100 µL
111
radioactive translation reaction, we used 40 pmol of ligated mRNA, 8 µL of 12.5X Amino
Acid Mix (without Met), 4 µL 2.5M KOAc, 2 µL 25mM MgOAc, 8 µL
35
S-labeled
methionine, and added rabbit reticulocyte lysate to 100 µL. The reaction was allowed to
proceed for one hour at 30 °C.
Fusion of the nascent peptide to its encoding mRNA was induced by the addition of 8
µL 1M MgCl
2
and 28 µL 2.5M KCl, followed by incubation at room temperature for 15
minutes. Afterwards, peptide-mRNA fusions were purified by incubation with oligo-dT
immobilized on streptavidin agarose beads while rotating for one hour at 4 °C. The
beads were then washed with Isolation buffer (100 mM Tris–HCl, pH 8.0, 1 M NaCl,
0.2% (v/v) Triton X-100) and bound peptide-mRNA fusions were then eluted with 200 µL
of 65 °C water. The eluted samples were then desalted with a Centri-Sep spin column.
mRNA Display: Reverse Transcription
Eluted peptide-mRNA fusions were reverse transcribed to generate the corresponding
cDNA for each peptide. For each 200 µL purified mRNA-peptide fusions, 25 µL 10X
dNTPS (2mM each dNTP), 25 µL 10X first strand buffer (500 mM Tris–HCl pH 8.3,
750 mM KCl, 30 mM MgCl
2
) and superscript II (SSII) were added. Reverse transcription
continued for an hour at 4 °C. Afterwards, SSII was heat inactivated by incubation at 65
°C for 10 minutes.
112
mRNA Display: Binding Assays
Standard
Biotinylated MBP-SH2 was immobilized on Dynabeads® MyOne™ Streptavidin T1
(Thermo Fisher). Unbound MBP-SH2 was washed away with STAT3 buffer (50 mM
HEPES pH 7.5, 100 mM NaCl, 1 mM DTT, 1 mM EDTA, 10% Glycerol). Washed beads
were resuspended in 1 mL STAT3 Blocking buffer (STAT3 buffer, 10 µM Biotin, 10
mg/mL BSA, 0.5% Tween 20, 1 mg/mL Yeast RNA) supplemented with 100X MBP in
solution and rotated at 4 °C for one hour. Afterwards, peptide-mRNA fusions were
added to the beads (the entire reverse transcribed sample was added for cold fusions
and approximately 100,000 cpm for hot fusions) and rotated at 4 °C or at room
temperature for another hour. After incubation, the beads were washed with STAT3
selection buffer (STAT3 buffer containing 10 µM biotin, 1mg/mL BSA, 10 µg/mL Yeast
RNA, and 0.05% Tween 20). The radioactivity of each wash was counted with a
scintillation counter (Beckman). Then, the radioactivity of the beads themselves was
counted in order to determine what percentage of the library remained bound to the
immobilized MBP-SH2. The cold peptide-mRNA fusion library was washed
simultaneously. After washing, the beads were resuspended in water for use as a
template in PCR.
Off-Rate Selection
The protocol for the off-rate binding assay is identical to the protocol for the standard
binding assay until the final wash of the beads. After the final wash, the beads are
resuspended in 1 mL STAT3 selection buffer containing 100X excess MBP-SH2 in
113
solution to prevent the reassociation of any peptides that dissociate from the
immobilized MBP-SH2. At various time points, 100 µL of the bead slurry was taken and
the percentage of counts remaining on the beads was determined by the scintillation
counter. We were able to estimate K
off
for the library by fitting these values to the
following equation, where Bound represents the percentage of radioactive counts
remaining on the beads, Max is the maximum percentage of remaining counts on the
beads, k
off
is the off-rate of the library, and t is the time at which the percent of
remaining counts was measured.
Using this model, we predicted the percentage of our target population of peptides
(with a fivefold k
off
improvement over the current peptide library) that remained bound.
We then compared the predicted percent bound values for both the actual library and
the target population library in order to determine the time at which the ratio of the
bound target population to the bound original library was maximal (t
optimal
). The cold
library was then bound to MBP-SH2, washed, and resuspended in 1 mL STAT3
selection buffer containing 100X free MBP-SH2 in solution. The sample was then
rotated at room temperature for a predetermined amount of time (t
optimal
), after which the
beads were quickly washed, resuspended in water, and used as a template for PCR.
Chymotrypsin Digestion
Chymotrypsin digestion was carried out by taking the purified peptide-mRNA fusions
and treating them with a specified amount of chymotrypsin agarose beads (Sigma-
Aldrich) for a specified amount of time. We used a Spin-X column (Corning Costar) to
114
stop the reaction with centrifugation by physically separating the solid beads from the
fusions in solution. The fusions were then subjected to the normal binding assay
protocol detailed previously.
Sanger Sequencing
Library DNA was subcloned into a TOPO vector (Thermo Fisher) and transformed into
TOP10 competent cells (Thermo Fisher). Individual colonies were the selected and
cultured overnight in 5 mL LB at 37 °C. The cultures were then miniprepped according
to the protocol for Nucleospin Plasmid columns (Macherey-Nagel). Plasmid
concentrations were determined by nanodrop and submitted for Sanger sequencing
(Genewiz) according to the provided sample submission guidelines.
115
Illumina Sequencing
Libraries were PCR amplified to include unique barcodes and Illumina adapters
(Forward Illumina Adapter Primer Sequence:
AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT,
Reverse Illumina Adapter Primer Sequence:
CAAGCAGAAGACGGCATACGAGATCGGTCTCGGCATTCCTGCTGAACCGCTCTTCCGATC)
. Prepared samples were then submitted to the UPC Genome Core for Illumina HiSeq
2500 analysis.
Peptide Synthesis
Peptides were chemically synthesized using Fmoc solid phase peptide synthesis by
an Initiator+ microwave synthesizer (Biotage). A 200 µmol synthesis was performed
with 5 molar equivalents of each amino acid and HATU. HATU and all amino acids were
dissolved in N-methyl-2-pyrrolidone (NMP) to a final concentration of 0.5 M. NMP was
used as the solvent for all of the synthesis reactions, and a solution of 25% (v/v) 4-
methyl piperidine in NMP was used to deprotect the peptide before the addition of each
amino acid. The peptides were synthesized on rink amide MBHA resin (AnaSpec),
which resulted in a final product with a C-terminal amide.
Synthesized peptides were cleaved off of the resin with a cleavage solution consisting
of 94% (v/v) trifluoroacetic acid (TFA), 2.5% 1,2-ethanedithiol (EDT), 2.5% deionized
water (DI), and 1% triisopropylsilane (TIS). Cleaved peptides were then precipitated in
methyl tert-butyl ether (MTBE) and lyophilized. The peptide was then dissolved in
dimethyl sulfoxide (DMSO) and purified via HPLC with a C
18
reverse phase column
(Vydac) using a gradient of 10-90% acetonitrile/0.1% TFA in water. Fractions
116
corresponding to large peaks on the HPLC trace were analyzed by MALDI-TOF (ABI).
Fractions with the correct molecular weight were pooled, lyophilized, dissolved in
DMSO, and stored at -80 °C.
Peptides containing a biotin were first synthesized with a protected lysine residue at
the C-terminus (Fmoc-Lys-MTT). Following attachment to the bead, the MTT group was
removed with a deprotection solution of 2% TFA, which was then confirmed by a Kaiser
test. With the MTT group removed, biotin was coupled to the side of the lysine group.
After verification of biotin coupling by a Kaiser test, peptide synthesis continued as
described above.
STAT3-Dependent Luciferase Assay
MCF-7 cells were electroporated with the Neon electroporator system (Invitrogen) with
three plasmids: two with a pcDNA 3.1 expression vector backbone for the constititutive
expression of GFP and the selected peptide (Winner, I3S or Scrambled), and one that
had STAT3-dependent luciferase expression. The luciferase plasmid was a gift from
Professor Abdelilah Soussi-Ghounni. After electroporation, we cultured the MCF7 cells
at 37 °C and 5% CO
2
for two days. Next, half of the MCF7 cells were analyzed on a flow
cytometer to measure GFP and determine the overall electroporation efficiency. The
remaining cells were used to prepare lysates using the Firefly Luciferase Glow Assay kit
(Pierce). The lysates were then analyed on plate reader (Synergy H4 Microplate reader)
at the USC nanobiophysics core, under the supervision of Dr. Shuxing Li.
117
Figure S1: Vector map for pET24a MBP-SH2
118
Figure S2: A sample SDS-PAGE gel verifying the biotinylation of MBP-SH2. Lane 1:
supernatant remaining after coupling MBP-SH2 to streptavidin agarose beads. Lane 2:
supernatant after boiling streptavidin agarose beads coupled to MBP-SH2 in the
presence of free biotin
4.3: Results and Discussion
SH2 Library Design
We designed our library based on the phosphopeptide identified by Ren et al:
GpYLPQTV
33
. The library consisted of a randomized amino acid (X), followed by a core
that was doped at 51% to resemble the unphosphorylated peptide (GYLPQTV), another
randomized amino acid (X) and a flexible GS-linker tail (SSSGS). This resulted in a
119
library that contained one copy of the wild-type peptide in every 10,000 peptides
(MXGYLPQTVXSSSGS).
SH2 Library Selection
We translated 400 pmol of ligated library mRNA for the first round of selection. We
then selected for binding to 1,000 pmol of MBP-SH2 immobilized on streptavidin
agarose beads by incubating the peptide fusions and the beads at 4 °C. For subsequent
rounds of selection, we translated 40 pmol of ligated library mRNA and used 30 pmol of
MBP-SH2 immobilized on magnetic streptavidin beads as our target. As shown in
Figure 1, we first observed binding above background levels with our Pool 3 library
(labeled "Round 4"). As a result, we increased the binding temperature from 4 °C to
room temperature (21 °C) in order to select for peptides that could bind at a higher
temperature. Despite the added selective pressure, we noticed a substantial increase in
radiolabeled binding in the next round - increasing from 0.1% for Pool 3 to 0.3% for Pool
4 (Labeled "Round 5"). By Round 7, the end of the selection, our peptide library was
shown to bind at 0.8%.
SH2 Library Sequencing
We first used Sanger sequencing to determine whether our library had converged.
We observed that our library, SH2 Pool 7, was divided into two distinct families: one
containing a frame shift and another, more hydrophobic family without a frame shift. As
shown in Figure 2, none of the peptides resembled the parent sequence, and only a few
had the same residues as the parent sequence at any given position. However, each of
120
the frame shifted peptides had the same tail (YHSSRLLAA), which began in the middle
of the peptide - in the randomized coding region. Interestingly, the DNA sequences for
each of the frame shifted peptides all had a single nucleotide deletion at the same
position in the randomized coding region. In Figure 3, we show that there is a high
degree of homology between all of the frame shifted peptides, which is not seen in the
peptides without a frame shift.
We Illumina sequenced SH2 Pool 7 to obtain a more complete view of the peptides
that comprised our library (Figure 4). Illumina sequencing produced over 5 million
sequences, of which 80,000 were unique. As shown in Figure 4, the top 11 sequences
made up over 80% of our pool, indicating that it had reached convergence. As
expected, Illumina sequencing also confirmed that the peptides of SH2 Pool 7 were
divided two distinct families.
SH2 Library Clones: Confirmation of Binding
Four clones were selected for further characterization based on their copy number:
7-2 (MHFVIIVIRR SSSGS), 7-3 (MKIFLYHSSR LLAAA), 7-4 (MYGGMFWITL SSSGS)
and 7-17 (MTVYLYHSSR LLAAA). The selected clones were PCR amplified,
transcribed, ligated to puromycin, translated radioactively, and used in binding assays
against 30 pmol MBP-SH2 immobilized on streptavidin magnetic beads. Binding assays
were performed at room temperature for one hour with 3000 pmol MBP in solution. Our
results indicate that 7-4 bound the best (1.4% binding), followed by 7-3 (0.8%), 7-2
(0.7%), and finally, 7-17 (0.4%).
121
Doped SH2 Library Design
We doped a new library based on clone 7-3 and added three randomized amino
acids on both ends of the core sequence in order to improve the binding affinity of our
peptide inhibitors. We avoided using clone 7-4 despite its better performance in
radioactive binding assays for two reasons. First, our ultimate goal is to identify and
synthesize peptide inhibitors. As a result, we were deterred by the presence of "GG" in
the sequence because of the difficulties associated with its synthesis.
36
Second, clone
7-4 is composed almost entirely of hydrophobic residues, which will likely result in
solubility issues. The Doped SH2 Library core was doped at 40%, resulting in a library
containing 7 copies of the wild type peptide per 10 million peptides. The DNA library
encoded MXXXKIFLYHSSRXXX GSGSS.
Doped SH2 Library K
off
Selection
We began our selection by using 1,200 pmol of MBP-SH2 immobilized on
streptavidin agarose beads as our target. This binding reaction was allowed to occur
while rotating at 4 °C. We used lower selective pressures for the first round in order to
capture all of the functional peptides in our beginning library. We used less target in
subsequent rounds of selection (30 pmol of MBP-SH2 immobilized on streptavidin
magnetic beads) and once again increased the binding temperature to 21 °C (room
temperature).
As shown in Figure 6, we proceeded with selection for four rounds until we observed
binding above background levels for Pool 3 (2%). After one more round of selection, our
122
library (Pool 4) was able to bind at nearly 10%. In order to select for the clones with the
best dissociation constants (
), we subjected Pool 5 to off-rate selective
pressure, as described in the materials and methods section. We normalized the
amount of radiolabeled counts remaining to the amount that was initially bound at the
beginning of our experiment (t = 0). As shown in Figure 7, the K
off
for Pool 5 was about
2.4 x 10
-5
as determined by Excel. Using our own model to fit the data, we calculated
the K
off
for Pool 5 to be about 5.0 x 10
-5
. Based on this value, we could conclude that
our current library contained a population of peptides that had an off rate that was
fivefold better than that of Pool 5 (1.0 x 10
-5
). Consequently, as shown by Figure 8, we
could compare the off-rates of both Pool 5 and higher affinity subpopulation of Pool 5
and determine the incubation time that would result in the greatest ratio between the
two pools.
As a result, we used an incubation time of 21.1 hours to generate Pool 6. We
determined the off-rate of Pool 6 to be 3.5 x 10
-5
(Figure 9). This value was surprising
because it indicated that there was no significant improvement between the off-rates of
Pool 5 and Pool 6 despite specifically performing an off-rate selection. When the data
are plotted together, there appears to be no difference between the off-rates of Pools 5
and 6 (Figure 10). One possible explanation for this is that there is not much diversity
remaining in Pools 5 and 6.
123
Doped SH2 Library Sequencing
We feared that the poor off-rate improvement was indicative of a converged library
because the ultimate goal of this work is to generate protease resistant peptides that
bind to the STAT3 SH2 domain. Without sufficient diversity, we would not be able to
identify any peptides that were functional and protease resistant. To determine whether
this was the case, we sequenced Pools 5 and 6 using Sanger sequencing. In Figure 11,
we can see that both Pool 5 was nearly converged and the off-rate selection essentially
enriched Pool 6 for the dominant clone, MIIRLLFRGTVYSLHV. Also, from these
sequencing results, we can see that none of the clones resemble the parent sequence.
Chymo-SH2 Library Design
We created a new library based on the dominant clone from the off-rate selection
(MIIRLLFRGTVYSLHV) to generate more diversity for the identification of protease
resistant peptides. We doped the library at 55% (resulting in one copy of the parent
peptide sequence for every 130 million peptides) and changed the GS linker tail
sequence to SGGSS to prevent any cross-contamination.
Chymotrypsin Resistance Selection
For the first round of selection, the Chymo-SH2 Library was translated and bound to
1,000 pmol MBP-SH2 immobilized on neutravidin agarose beads at 4 °C. We used a
large quantity of target to capture all functional clones in the library. Subsequent rounds
124
were selected against 30 pmol MBP-SH2 immobilized on streptavidin magnetic beads
at room temperature. We proceeded with a standard selection protocol (no protease
digestion) until we observed net binding above background levels in Pool 2 (Figure 12).
Pools 3-6 were treated with increasing amounts of chymotrypsin agarose for increasing
amounts of time to provide protease pressure. From Figure 13, it is clear that Pool 3
was highly susceptible to chymotrypsin digestion - decreasing from roughly 1.0%
binding to below 0.2% after only 10 seconds of digestion. The end result of the applied
protease selective pressure can be seen in the binding for Pool 6, which is not affected
when subjected to chymotrypsin digestion for two minutes. Because 2 minutes of
chymotrypsin digestion had no impact on the binding capabilities of Pool 6, we decided
to determine how long we could digest Pool 6 without experiencing a complete lack of
function. In Figure 14, we can see that digestion for ten minutes was sufficient to
significantly decrease the binding of Pool 6.
Chymo SH2 Library Sequencing
We submitted three pools from this selection for Illumina high throughput sequencing:
Pool 6, Pool 7 (2 min) - the pool that came from digesting Pool 6 for two minutes, and
Pool 7 (10 min) - the pool that came from digesting Pool 6 for ten minutes. In Figure 15,
we can see that the library was beginning to converge despite the increasing protease
pressure. The top clone accounts for nearly a quarter of all sequences, and the top ten
clones account for over half of all sequences. From Figure 16, it is obvious that the
library is continuing to converge - the top clone from Pool 7 (2 min) now accounts for
over 40% of all sequences and the top 10 clones now constitute over 75% of all
125
sequences in the pool. In Figure 17, we can see that ten minutes of chymotrypsin
digestion had a profound effect on the constituents of Pool 7 (10 min).
We then calculated every peptides' "concentration" (copy number divided by the sum
of all sequences for that pool) and compared these values (ppm ratio) for individual
peptides within each pool. By doing so, we are able to determine whether any particular
sequence was enriched by selection. We examined the ppm ratio between Pool 7 (10
min) and Pool 7 (2 min), as well as Pool 7 (10 min) and Pool 6.
Chymo SH2 Library Clones: Confirmation of Binding
Clones were selected for characterization based on several critera: their copy number
in Pool 7 (10 min), the ppm ratio between Pool 7 (10 min) and Pool 7 (2 min), and the
ppm ratio between Pool 7(10 min) and Pool 6. Clones with high copy number in all three
pools were selected regardless of ppm ratio, but clones with a high ppm ratio alone
were not selected due to the high likelihood of the high ppm ratio being an artifact of
sequencing. Ultimately, we selected clones 2, 4, 6, 7, 8, 9, 10, 16, 19 and 61
(sequences shown in Table 1), which were named according to their copy number rank
in Pool 7 (10 min). As shown in Figure 18, the binding of clones 2 (MCILLLINGIVYTIHL:
1.8%), 7 (MIIILNYKGTIIYIHV: 2.6%), and 9 (MIIRIVIGDTVYTIHL: 2.4%) were highest.
These clones will be synthesized and further characterized as discussed in the following
section.
126
A Sequence Comparison of "Winner" Clone to the Top 300 Clones from Chymo
SH2 Library
We analyzed nearly 300 of the most abundant clones from the Chymo SH2 selection
and compared their amino acid composition at each position in the peptide in order to
identify which residues were most conserved. The analysis revealed that isoleucine at
position 3 (I3) and histidine at position 15 were represented in over 80% of these top
sequences (Figure 19). Next, using the sequence of Winner as a template, we made the
point mutants I3S, H15Q and H15R in an attempt to reduce binding and validate the key
role that I3 and H15 had for maintaining binding to the SH2 domain. The decision to
mutate I3 to S and H15 to Q or R was due to the desire to select amino acids that had
different characteristics, but also had similar codons to the conserved residue. We
chose these criteria because we wanted to confirm that the mRNA display selection
process was able to enrich for functional sequences, and not nonfunctional sequences
that had a high probability of resembling the functional sequence. Figure 20 shows that
each point mutation was indeed able to decrease binding, but the I3S mutation had the
greatest impact on binding. As a result, we chose to use the I3S mutant as a negative
control for subsequent experiments.
Validation of Winner - STAT3 Interaction In Vitro
The sequences for Winner, I3S, and scrambled Winner were subcloned into a pcDNA
3.1 constitutive expression vector. Then, MCF7 cells were electroporated with one of
these constitutive expression vectors, along with a vector for the constitutive expression
127
of GFP and a vector for the STAT3-dependent expression of luciferase. After two days
of culture, the MCF7 cells were analyzed for GFP by flow cytometry to determine the
electroporation efficiency. Then, the cells were lysed and assayed for luciferase activity.
We observed no significant difference between the luciferase activity of the cells that
constitutively expressed I3S, the scrambled mutant, or the empty pcDNA3.1 vector.
However, we report that the MCF7 cells that constitutively expressed Winner or
received IL6 stimulation both had significantly higher STAT3 transcriptional activity than
any of the negative control peptides (Figure 21, * p < 0.05, ** p < 0.01). This result
indicates that although Winner is able to bind to STAT3, its binding results in an
increase in STAT3 transcriptional activity, rather than a decrease.
128
4.4: Conclusions and Future Directions
We used mRNA display to select for peptide ligands to the STAT3 SH2 domain.
After applying off-rate selective pressure, we were able to identify Winner
(MIIRLLFRGTVYSLHV) as a high affinity peptide ligand for the STAT3 SH2 domain.
However, Winner must be further characterized before it can be used as a therapeutic.
Although we demonstrated that Winner is able to affect STAT3 activity in vitro, it was
shown to function as an enhancer of STAT3 transcriptional activity, rather than an
inhibitor. As a result, further experiments must be conducted in order to utilize
Winner's STAT3 binding capabilities while minimizing its enhancement of STAT3
transcriptional activity.
Our proposed solution is to create a PROTAC (PRoteolysis TArgeting Chimera) with
Winner as a targeting moiety and a small molecule capable of binding to the E3 ligase.
Potential E3-binding molecules include those described by Buckley et.al.
37
and Raina
et.al.
38
One particular advantage to the molecule described by Raina et.al. is its
commercial availability and its compatibility with solid phase peptide synthesis.
Another approach is to use a 27 residue segment from hypoxia-inducible-factor-1α
(HIF-1α) as an E3 ligase targeting moiety.
Future experiments include performing a golgi-targeting assay in order to confirm
that Winner is able to colocalize with STAT3 within cells. Colocalization will further
confirm that Winner binds to STAT3 in vitro. Afterwards, we will confirm whether
Winner is able to facilitate the PROTAC-induced degradation of STAT3 in vitro. This
will be accomplished in one of two ways. In both cases, we would be repeating the
129
STAT3-dependent luciferase experiment with MCF7 cells as described above. The
key difference is whether we use a synthesized PROTAC consisting of Winner and the
commercially available E3-binding small molecule described by Raina et.al., or
whether we use a use an expression vector to constitutively express a Winner-HIF-1α
PROTAC in the MCF7 cells. We intend to validate the function STAT3-targeted
PROTAC with the expression vector first because using a synthesized peptide could
present challenges with cell membrane permeability. If this is shown to be successful,
we will then synthesize a more therapeutically relevant PROTAC consisting of Winner
and an E3-binding small molecule for further testing in vitro. Moreover, if protease
resistance is required, we would be able to conduct the same experiments with some
of the lower affinity, but more chymotrypsin-resistant clones identified in the Chymo
SH2 selection.
130
4.5: Figures
Figure 1: Percentage of radiolabeled counts from the SH2 library that bound to
immobilized MBP-SH2. Round 1: 1000 pmol target, Rounds 2-7: 30 pmol target.
Rounds 1-4 were bound at
4 °C, Rounds 5-7 were bound at room temperature.
131
Figure 2: Sanger sequencing results for SH2 Pool 7. Residues shown in bold are
identical to the parent sequence.
132
Figure 3: Comparison of DNA sequences of frame shifted peptides from SH2 Pool 7.
Red nucleotides indicate a deviation from the doped sequence.
133
Figure 4: Illumina sequencing results for SH2 Pool 7. a) Distribution of SH2 Pool 7 by
clone copy number ranking. b) Amino acid sequences and copy number for the top 20
clones in SH2 Pool 7. Yellow highlighting denotes frame shifted sequences.
134
Figure 5: Binding assays for clones from SH2 Pool 7 that were identified by both Sanger
and Illumina sequencing. Background binding is shown is red. Binding to the SH2
domain is shown in blue. Binding was done by rotation at room temperature for one
hour with 30 pmol of MBP-SH2 immobilized on magnetic streptavidin beads.
135
Figure 6: Percentage of radiolabeled counts from the Doped SH2 library that bound to
immobilized MBP-SH2. Pool 0: 1000 pmol target, Pools 1-4: 30 pmol target. Pool 0 was
bound at 4 °C, Pools 1-4 were bound at room temperature
136
Figure 7: Binding kinetics of Doped SH2 Pool 5 at room temperature with 100 pmol
MBP-SH2 immobilized on magnetic streptavidin beads.
137
Figure 8: Using the experimentally determined off-rate of Doped SH2 Pool 5 to predict
the off-rate of a higher affinity subpopulation. The incubation time that resulted in the
largest ratio between the library and the subpopulation was calculated to be 21.1 hours.
138
Figure 9: Binding kinetics of Doped SH2 Pool 6 at room temperature with 100 pmol
MBP-SH2 immobilized on magnetic streptavidin beads.
139
Figure 10: Comparison of Doped SH2 Pool 5 and Pool 6 binding kinetics using
experimentally calculated off-rates.
140
Figure 11: Sanger sequencing results for Doped SH2 Pool 5 and Pool 6. Bolded
residues denote residues that are conserved with the parent clone.
141
Figure 12: Binding assays for the first three rounds of selection for the Chymo-SH2
library. Selection for Pool 0 was performed with 1,000 pmol MBP-SH2 immobilized on
Neutravidin Agarose beads at 4 °C. Subsequent selections were performed with 30
pmol MBP-SH2 immobilized on Streptavidin magnetic beads at room temperature.
142
Figure 13: Binding assays for Chymo-SH2 Pools 3-6. Digested pools were used for
subsequent rounds of selection. Later pools were digested with larger amounts of
chymotrypsin agarose beads for longer periods of incubation. Binding of undigested
pools are shown for reference.
143
Figure 14: Binding assays for Chymo-SH2 Pool 6 with different digestion conditions.
Chymo-SH2 Pool 6 was digested with 1 mg of chymotrypsin agarose for either 2
minutes or 10 minutes.
144
Figure 15: Illumina sequencing results for Chymo-SH2 Pool 6. a) Sequence distribution
of Chymo-SH2 Pool 6. b) Sequence and copy number of the 20 clones with the highest
copy number.
145
Figure 16: Illumina sequencing results for Chymo-SH2 Pool 7 (2 min), generated by
digesting Chymo-SH2 Pool 6 for two minutes with chymotrypsin. a) Sequence
distribution of Chymo-SH2 Pool 7 (2 min). b) Sequence and copy number of the 20
clones with the highest copy number.
146
Figure 17: Illumina sequencing results for Chymo-SH2 Pool 7 (10 min), generated by
digesting Chymo-SH2 Pool 6 for ten minutes with chymotrypsin. a) Sequence
distribution of Chymo-SH2 Pool 7 (10 min). b) Sequence and copy number of the 20
clones with the highest copy number.
147
Clone Sequence
2 MCILLLINGIVYTIHL
4 MIIWIPVRVIVFLPTV
6 MIIWLPITITIVTLPT
7 MIIILNYKGTIIYIHV
8 MILIITLPDSFNVILV
9 MIIRIVIGDTVYTIHL
10 MTIRLVLGDTIISIHV
16 MIIRILVSGTVISIHL
19 MIIKVLIRGVIYYIHV
61 MIIWLPITITIITLPT
Table 1: Clones from Chymo-SH2 Pool 7 (10 min) selected for further characterization
148
Figure 18: Binding assays for clones selected out of Pool 7 (10 min). Binding reactions
were performed with 30 pmol MBP-SH2 immobilized on streptavidin magnetic beads,
followed by incubation with the specified peptides while rotating at room temperature.
149
Figure 19: A sequence comparison of Winner and the top 290 clones identified in the
Chymo SH2 selection. The representation of each amino acid residue at each position
was tallied in order to identify residues that were critical to binding to the SH2 domain.
The residues matching Winner at each position are shown in bold.
150
Figure 20: Point mutants of Winner were made based on a sequence comparison of
Winner and the top 290 clones identified in the Chymo SH2 selection. The binding
assay results of the point mutants (I3S, H15Q and H15R) are compared to Winner
below.
0%
2%
4%
6%
8%
10%
12%
H15Q H15R I3S Winner
Percent Bound (cpm)
Background
SH2 MBP
151
Figure 21: STAT3-dependent luciferase assay results. MCF7 cells were electroporated
with plasmids responsible for the STAT3-dependent expression of luciferase and the
constitutive expression of GFP and the indicated peptide. Electroporated MCF7 cells
were lysed, and luciferase activity was measured, normalized to GFP expression, and
shown below. (Data shown as mean ± SEM. ns = not significant, * p < 0.05, ** p < 0.01).
0%
50%
100%
150%
200%
pcDNA Winner Scrambled I3S Stimulated
Normalized luciferase activity
(% of control)
152
4.6: References
1. Roberts, R.W., Szostak, J.W. RNA-peptide fusions for the in vitro selection of peptides
and proteins. Proc Natl Acad Sci U S A (1997).
2. Bazan, J., Calkosinski, I. & Gamian, A. Phage display--a powerful technique for
immunotherapy: 1. Introduction and potential of therapeutic applications. Hum Vaccin
Immunother 8, 1817-28 (2012).
3. Pepper, L.R., Cho, Y.K., Boder, E.T., E.V. Shusta. A decade of yeast surface display
technology: Where are we now? Comb Chem High Throughput Screen. (2008).
4. Azzam, M.E., Algranati, I. D. . Mechanism of Puromycin Action: Fate of Ribosomes after
Release of Nascent Protein Chains from Polysomes. Proc Natl Acad Sci U S A (1973).
5. Barrick, J.E., Takahashi, T.T., Ren, J., Xia, T. & Roberts, R.W. Large libraries reveal
diverse solutions to an RNA recognition problem. Proc Natl Acad Sci U S A 98, 12374-8
(2001).
6. Howell, S.M. et al. Serum stable natural peptides designed by mRNA display. Sci Rep 4,
6008 (2014).
7. Morelli, A., Haugner, J., Seelig, B. Thermostable Artificial Enzyme Isolated by In Vitro
Selection. PLOS One (2014).
8. Li, S., Millward, S., Roberts, R. In Vitro Selection of mRNA Display Libraries Containing
an Unnatural Amino Acid. JACS Communications (2002).
9. Millward, S.W., Takahashi, T.T., Roberts, R.W. A General Route for Post-Translational
Cyclization of mRNA Display Libraries. JACS Communications (2005).
10. Millward, S.W., Fiacco,S., Austin, R.J., Roberts, R.W. Design of Cyclic Peptides That
Bind Protein Surfaces with Antibody-Like Affinity. ACS Chemical Biology (2007).
153
11. Yu, H. & Jove, R. The STATs of cancer--new molecular targets come of age. Nat Rev
Cancer 4, 97-105 (2004).
12. Takahashi, T.T. & Roberts, R.W. In vitro selection of protein and peptide libraries using
mRNA display. Methods Mol Biol 535, 293-314 (2009).
13. Banerjee, K. & Resat, H. Constitutive activation of STAT3 in breast cancer cells: A
review. Int J Cancer 138, 2570-8 (2016).
14. Wei, D. et al. Stat3 activation regulates the expression of vascular endothelial growth
factor and human pancreatic cancer angiogenesis and metastasis. Oncogene 22, 319-
29 (2003).
15. Pedranzini, L., Leitch, A., Bromberg, J. Stat3 is required for the development of skin
cancer. J. Clin. Invest. (2004).
16. Yu, H., Lee, H., Herrmann, A., Buettner, R. & Jove, R. Revisiting STAT3 signalling in
cancer: new and unexpected biological functions. Nat Rev Cancer 14, 736-46 (2014).
17. Niu, G., Wright, K.L., Huang, M., Song, L., Haura, E., Turkson, J., Zhang, S., Wang, T.,
Sinibaldi, D., Coppola, D., Heller, R., Ellis, L.M., Karras, J., Bromberg, J., Pardoll, D.,
Jove, R., Yu, H. Constitutive Stat3 activity up-regulates VEGF expression and tumor
angiogenesis. Oncogene (2002).
18. Aoki, Y., Feldman, G.M., Tosato, G. Inhibition of STAT3 signaling induces apoptosis and
decreases survivin expression in primary effusion lymphoma. Blood (2003).
19. Konnikova, L., Kotecki, M., Kruger, M.M., Cochran, B.H. Knockdown of STAT3
expression by RNAi induces apoptosis in astrocytoma cells. BMC Cancer (2003).
20. Ostrand-Rosenberg, S. & Sinha, P. Myeloid-derived suppressor cells: linking
inflammation and cancer. J Immunol 182, 4499-506 (2009).
21. Zhang, X. & Darnell, J.E., Jr. Functional importance of Stat3 tetramerization in activation
of the alpha 2-macroglobulin gene. J Biol Chem 276, 33576-81 (2001).
154
22. Zhang, T., Kee, W.H., Seow, K.T., Fung, W., Cao, X. The coiled-coil domain of Stat3 is
essential for its SH2 domain-mediated receptor binding and subsequent activation
induced by epidermal growth factor and interleukin-6. Molecular and Cellular Biology
(2000).
23. Horvath, C.M., Wen, Z., Darnell Jr., J.E. A STAT protein domain that determines DNA
sequence recognition suggests a novel DNA-binding domain. Cold Spring Harbor
Laboratory Press (1995).
24. Levy, D.E. & Lee, C.-k. What does Stat3 do? J. Clin. Invest. 109, 1143-1148 (2002).
25. Siveen, K.S. et al. Targeting the STAT3 signaling pathway in cancer: role of synthetic
and natural inhibitors. Biochim Biophys Acta 1845, 136-54 (2014).
26. Darnell Jr., J.E. STATs and Gene Regulation. Science (1997).
27. Leong, P.L. et al. Targeted inhibition of Stat3 with a decoy oligonucleotide abrogates
head and neck cancer cell growth. Proc Natl Acad Sci U S A 100, 4138-43 (2003).
28. Schust, J., Sperl, B., Hollis, A., Mayer, T.U. & Berg, T. Stattic: a small-molecule inhibitor
of STAT3 activation and dimerization. Chem Biol 13, 1235-42 (2006).
29. Zhang, X. et al. Orally bioavailable small-molecule inhibitor of transcription factor Stat3
regresses human breast and lung cancer xenografts. Proc Natl Acad Sci U S A 109,
9623-8 (2012).
30. Zhang, X. et al. A novel small-molecule disrupts Stat3 SH2 domain-phosphotyrosine
interactions and Stat3-dependent tumor processes. Biochem Pharmacol 79, 1398-409
(2010).
31. Lin, L. et al. A Novel Small Molecule, LLL12, Inhibits STAT3 Phosphorylation and
Activities and Exhibits Potent Growth-Suppressive Activity in Human Cancer Cells.
Neoplasia 12, 39-IN5 (2010).
32. Haan, S. et al. Characterization and Binding Specificity of the Monomeric STAT3-SH2
Domain. J Biol Chem (1999).
155
33. Ren, Z., Cabell, L.A., Schaefer, T.S. & McMurray, J.S. Identification of a High-Affinity
Phosphopeptide Inhibitor of Stat3. Bioorganic & Medicinal Chemistry Letters 13, 633-636
(2003).
34. Turkson, J. et al. Phosphotyrosyl peptides block Stat3-mediated DNA binding activity,
gene regulation, and cell transformation. J Biol Chem 276, 45443-55 (2001).
35. Goodwin, D., Simerska, P. & Toth, I. Peptides as Therapeutics with enhanced
bioactivity. Current Medicinal Chemistry (2012).
36. Aldrich, S.
37. Buckley, D.L. et al. Small-molecule inhibitors of the interaction between the E3 ligase
VHL and HIF1alpha. Angew Chem Int Ed Engl 51, 11463-7 (2012).
38. Raina, K. et al. PROTAC-induced BET protein degradation as a therapy for castration-
resistant prostate cancer. Proc Natl Acad Sci U S A 113, 7124-7129 (2016).
156
Chapter 5: Masking Peptides for the Tumor-Specific
Activation of Chimeric Antigen Receptors
5.1: Abstract and Introduction
Abstract
Chimeric antigen receptor (CAR) T cells have experienced remarkable clinical success
when used to treat hematological malignancies. This success has inspired the design
and creation of CAR T cells targeting many different tumor-associated antigens (TAAs).
However, because TAAs are also expressed on healthy tissues, CAR T cells that target
TAAs could have potentially serious on-target, off-tumor side effects, including cytokine
storm. To address this issue and prevent CAR activation by healthy tissues,
researchers have developed CARs with a receptor-masking peptide that is attached to
the CAR via a linker that consists of a peptide spacer sequence and a protease-
cleavable peptide that is susceptible to cleavage by proteases that are frequently found
in the tumor microenvironment. We discuss the use of mRNA display to identify a
peptide (Peptide 5-1) that is capable of preventing the premature activation of an
epidermal growth factor receptor (EGFR) CAR. Despite its sequence similarity to the
parent masking sequence (Mask 1), Peptide 5-1 bound cetuximab without the need for
a spacer sequence, unlike Mask 1, whose binding to cetuximab relied heavily on the
presence of the spacer sequence. Peptide 5-1 was also shown to diminish its inhibition
of CAR binding to EGFR in the presence of increasing protease concentrations in vitro.
157
These results provide a proof-of-concept, and show that mRNA display can be used to
identify novel masking peptides that mitigate the on-target, off-tumor effects of CAR T
cells.
Introduction
Chimeric antigen receptor (CAR) T cell therapy has become an increasingly popular
field of research due to its well-documented clinical successes as a treatment for blood
cancers such as acute lymphoblastic leukemia and chronic lymphoblastic leukemia.
39-41
The clinical successes of CAR T cell therapy has spurred further research into the
development of CAR T cells targeting a variety of tumor associated antigens (TAAs),
42
including CARs targeting mesothelin,
43
human epidermal growth factor receptor 2
(HER2),
44
fibroblast activating protein (FAP),
45
and epidermal growth factor receptor
(EGFR).
46, 47
However, because TAAs are often also expressed on normal, healthy
cells, CARs that target TAAs can be prone to on-target, off-tumor toxicity.
48
Researchers have developed several methods to mitigate the off-tumor toxicity of CAR
T cells. One method is to modify the CAR T cells to express an inducible caspase-9
suicide switch. This safety mechanism is triggered by a small molecule chemical
inducer of dimerization, which causes caspase-9 to dimerize and induce apoptosis.
49-51
Another approach, which was engineered by Roybal et.al., requires the recognition of a
TAA by a synthetic Notch receptor in order to induce the expression of a CAR that is
specific to another TAA. This method increases patient safety and decreases the risk of
on-target, off-tumor toxicity by forcing the T cell to recognize two different TAAs before it
is activated.
52
158
The focus of this report is on the use of masking peptides to prevent the premature
binding of CARs to TAAs on healthy tissues. The concept of a masking peptide was first
developed by Desnoyers et.al. to mask cetuximab, an EGFR-binding antibody, in an
effort to decrease the severity of its side effects. The masking peptide, denoted as
"Mask 1" in this report, was attached to cetuximab via a flexible spacer peptide and
protease-degradable peptide. Mask 1 was shown to effectively block cetuximab binding
to EGFR, but treatment with tumor-associated proteases restored cetuximab's binding
to EGFR, indicating that cleavage of the protease-degradable linker resulted in the
dissociation of Mask 1.
53
Han. et.al. adapted Mask 1 to create a masked, EGFR-
targeting CAR T cell that was shown to be inactive against tumor cells that expressed
EGFR but did not secrete any proteases. However, CAR binding and function was
restored upon treatment with urokinase-type plasminogen activator (uPA), a protease
that is upregulated in the tumor microenvironment, which indicates that Mask 1 was
able to dissociate and stop its inhibition of receptor binding upon degradation of the
protease-cleavable linker.
54
mRNA display is a combinatorial method that is used to scan libraries exceeding 10
13
unique peptide sequences and select for functional, high affinity ligands that bind to a
target of interest.
12, 55
Additional selective pressures may also be introduced to identify
more robust peptides. For instance, Howell et.al. introduced a serum incubation step in
the selection which resulted in the identification of serum stable peptides that bound to
Gαi1.
6
Here, we discuss the use of mRNA display to identify peptide masking
sequences that were able to bind to cetuximab at 37 °C. The peptides that
159
demonstrated the highest level of binding to cetuximab at 37 °C was chosen for further
characterization in CAR T cells in vitro.
5.2: Materials and Methods
Library design and amplification
A doped DNA library was designed based on the cDNA sequence for Mask 1, with two
random amino acids flanking each side of Mask 1 (MXXCISPRGCPDGPYVMYXX), and
was synthesized by Integrated DNA Technologies (IDT). Approximately 4.8 x 10
11
unique sequences were then PCR amplified and used for mRNA display
mRNA Display
mRNA display was conducted as described by Takahashi et.al.
12
The DNA from each
round of selection was PCR amplified using Taq polymerase and transcribed into
mRNA with T7 RNA polymerase. The mRNA was then ligated to a pF30P linker (Keck
Biotechnology Resource Laboratory, New Haven, CT) with T4 DNA Ligase. The ligated
mRNA was purified by urea-PAGE and resuspended in deionized water. The samples
were then translated in vitro using rabbit reticulocyte lysate (prepared according to
Jackson and Hunt),
56
100 mM K(OAc), 500 µM Mg(OAc)
2
, and 200 nM ligated mRNA, in
a 1X translation mix (20 mM HEPES-KOH pH 7.6, 100 mM creatine phosphate, 2 mM
DTT, and 312.5 M of each amino acid). When preparing radiolabeled peptides, we
used
35
S-labeled methionine (Perkin Elmer;40 µCi for each 50 µL of translation) and a
translation mix that lacked methionine, but was otherwise identical to the previously
160
described translation mix. Translation reactions were incubated at 30 °C for an hour.
Fusion of the mRNA to its corresponding peptide was initiated by the addition of 4 µL of
1M MgCl
2
and 14 µl of 2.5M KCl per 50 µL of translation, followed by incubation at room
temperature for 15 minutes.
The mRNA-peptide fusions were purified after incubation with oligo dT cellulose beads
(4 mg per 50 µL translation) for 1 hour at 4 °C. The dT cellulose beads were washed 6
times with dT buffer (50 mM HEPES-KOH pH7.5, 1M NaCl, 1mM EDTA, 0.05% Tween
20), before elution in 200 µL of 65 °C H
2
O. The eluted fusions were then desalted with a
CENTRI-SEP spin column (Princeton Separations). Next, RT Buffer was added to the
desalted fusions were added to a final concentration of 1X (50 mM Tris-HCl pH 8.3, 75
mM KCl, 3 mM MgCl2, 2.4 mM 3’ primer, 200 mM each dNTP). Then, after the 3' primer
was added to a final concentration of 2 µM, the sample was heated at 65 °C for 5
minutes and cooled on ice for 10 minutes to allow the 3' primer to anneal. After cooling,
5 µL of Superscript II reverse transcriptase was added, and the reaction was allowed to
proceed for one hour at 42 °C. Afterwards, Superscript II was inactivated by incubation
at 65 °C for 5 minutes.
The reverse transcribed mRNA-peptide fusions were then used for selection. Cetuximab
was immobilized on either agarose or magnetic Protein G beads (Pierce) by incubation
at 4 °C for 1 hour (Cetuximab was obtained as a gift from Professor Michael Wong).
After binding, the Cetuximab-loaded beads were blocked in Blocking Buffer (10 mg/mL
BSA, 1 mg/mL Yeast RNA, 0.5% Tween 20) at either 4 °C or 37 °C for 1 hour. Then the
mRNA-peptide fusions were added and allowed to bind at either 4 °C or 37 °C for 1
hour. Unbound mRNA-peptide fusions were removed by successive washes of
161
Selection Buffer (1 mg/mL BSA, 10 µg/mL Yeast RNA, 0.05% Tween 20) at either 4 °C
or 37 °C, and the bound fusions were PCR amplified for the next round of mRNA
display.
Retroviral Vector Production
Retroviral vectors that were used to transduce human peripheral blood mononuclear
cells (PBMCs) were generated by calcium phosphate transfection of 293T cells. 293T
cells were cultured in 15 cm tissue culture plates and were transfected with 37.5 µg of
the retroviral backbone plasmid encoding the masked or unmasked EGFR CAR, 18.75
µg of pGALV, and 30 µg of a packaging plasmid encoding gag-pol. After 48 hours of
culture at 37 °C and 5% CO
2
, the viral supernatant was harvested and filtered through a
0.45 µm filter and used for transduction.
T cell transduction and expansion
Frozen human PBMCs were obtained from AllCells. PBMCs were transduced and
expanded as described previously.
54
Briefly, PBMCs were activated by culturing with 50
ng/mL OKT3, 50 ng/mL anti-CD28, and 10 ng/mL recombinant human IL-7 and IL-15
(Peprotech) for two days. Next, retroviral supernatant was loaded onto 12 well plates
that were precoated with 15 µg retronectin (Takara) per well. Loading was done by
centrifugation at 2,000 x g for two hours at 32 °C. Next, the viral supernatant was
removed and activated PBMCs were resuspended to a final concentration of 5 x 10
5
/mL
in TCM supplemented with 10 ng/mL recombinant human IL-7 and IL-15 and added to
162
the virus-coated plate. The plates were spun at 1,050 x g for 30 minutes at 32 °C and
then incubated overnight at 37 °C and 5% CO
2
.
Surface staining by flow cytometry
Proteolytic activation of the masked EGFR CARs was conducted as described
previously.
54
Next, cells were stained with 2 µg/mL of a recombinant human EGFR-Fc
fusion protein for 30 minutes at 4 °C. The cells were then washed twice and stained with
PE-AffiniPure F(ab')
2
fragment of goat anti-human IgG (Jackson ImmunoResearch) for
30 minutes at 4 °C in order to detect EGFR binding. Finally, cells were washed twice
and fluorescence was measured using a Miltenyi Biotec flow cytometer. Data was
analyzed with FlowJo software.
Cytotoxicity Assay
2 x 10
6
H292 or MD-MB-231 cells were incubated with Cell Tracker Orange CMTMR
Dye (ThermoFisher Scientific) for 30 minutes at 37 °C, washed once with PBS, and
resuspended in D10. Then, the dye-labeled cells were added to a 96 well plate (50,000
per well) and cocultured with CAR T cells at the indicated effector to target ratios at 37
°C and 5% CO
2
overnight. Afterwards, the coculture was stained with 7AAD, and
viability was determined by flow cytometry. Cytotoxicity was determined by taking the
percentage of 7AAD
+
, CMTMR
+
double-positive tumor cells and dividing it by the total
number of CMTMR
+
tumor cells.
163
5.3: Results and Discussion
Identification of Masking Peptides with mRNA Display
The goal of this project was to prevent the on-target, off-tumor effects of EGFR CAR
cells by identifying peptides that were capable of preventing the CAR from recognizing
EGFR until the CAR T cell was in the presence of tumor cells. Because the EGFR-
binding scFv of the CAR is derived from cetuximab, all mRNA display experiments were
conducted with cetuximab as the protein target. We began with a DNA library that
encoded two random amino acids flanking either side of a core that was biased towards
the amino acid sequence of Mask 1. We used a 61% bias towards Mask 1 on the
nucleotide level, meaning that for any given DNA sequence in the library, each of the
highlighted nucleotides shown in Figure 1A had a 61% chance of being incorporated at
that position, while every other nucleotide had a 13% chance. This resulted in a
beginning library that resembled Mask 1, and contained nearly 5.0 x 10
11
unique
peptides.
We then performed three rounds of mRNA display, with cetuximab immobilized on
protein A/G agarose beads (Round 1) and magnetic protein G beads (Rounds 2 and 3),
and binding occurring at 4 °C for one hour, followed by washing with selection buffer at
4 °C as well. We observed no binding for Pools 0 and 1, but noticed substantial binding
above background levels for Pools 2 and 3 (Figure 1B). At this point in the selection, the
binding and selection buffer temperatures was shifted from 4 °C to 37 °C to better select
for ligands that were able to bind to the CAR at physiological temperature (Figure 2A).
164
To increase the stringency of our temperature selective pressure, we also increased the
number of 37 °C washes from 6 to 12 in the fifth and final round of selection. As
expected, we observed a marked decrease in binding after increasing the binding
temperature, but because the binding level was still well above background levels
(Figure 2B), we believed that Pools 4 and 5 contained robust, functional ligands of
cetuximab that were able to maintain a high level of binding at 37 °C.
High throughput DNA sequencing was then used to determine the composition of Pools
3, 4, and 5, and revealed that increasing the selective pressure with a higher binding
and washing temperature was able to enrich for peptide ligands that could better
tolerate these more stringent binding conditions. We highlight and focus on the
discovery and characterization of two ligands from our selection: Peptide 5-1
(MRMCISPRGCPDGPYERYTL) and Peptide 5-3 (MNLCISPRGCPDGPFEMYTL),
which were the first and third most abundant clones in Pool 5 (the second most
abundant clone was tested and found to be nonfunctional, data not shown). Peptides 5-
1 and 5-3 were selected due to their abundance in the final pool and their favorable
enrichment, which is defined as the change in representation of a particular sequence
as the selection progressed. In this case, enrichment of greater than 1 indicates that a
particular clone is better able to maintain binding at 37 °C than other clones within the
pool, whereas enrichment of less than 1 suggests that binding is substantially impaired
at 37 °C. Peptide 5-3, which had the highest enrichment, was shown to have enriched
22.74 times when going from Pool 3 to Pool 5, and Peptide 5-1, which was the most
abundant clone, enriched 5.91 times (Figure 2C).
165
Validation of Peptide 5-1 and Peptide 5-3
First, we performed radioactive binding assays to evaluate the binding of Peptide 5-1
and Mask 1 (both without the N-terminal spacer sequence) to cetuximab at 4 °C. In this
scenario, Peptide 5-1 was shown to have substantially higher binding to cetuximab than
Mask 1 (Figure 3A).
Binding assays were then performed at 37 °C to compare the binding of Peptides 5-3
and 5-1 to Mask 1. We found that the N-terminal spacer sequence (QGQSGQ) was
crucial for Mask 1 binding to cetuximab. As a result, we tested the binding of Mask 1
and Peptides 5-3 and 5-1 with an N-terminal spacer. Peptide 5-1 and Mask 1
demonstrated a similar degree of binding to cetuximab. The binding of Peptide 5-3
without the spacer is also shown (Figure 3B). These results indicate that Peptide 5-3,
both with and without the spacer, has the best binding to cetuximab at 37 °C. The high
enrichment and binding of Peptide 5-3 make it an excellent candidate for further
characterization in vitro. (Further experiments to evaluate the efficacy of Peptide 5-3 are
currently in progress.)
Next, we generated CAR T cells masked by Mask 1 and Peptide 5-1, and tested their
CAR masking capabilities using the protocols described by Han et.al.
54
Figure 4 shows
that despite the higher level of binding that Peptide 5-1 exhibited at 4 °C, there was no
substantial difference in EGFR binding between Peptide 5-1 and Mask 1 after treatment
with uPA. Similarly, coculture of either of the masked CARs with EGFR expressing
166
tumor cell lines that are known to secrete tumor associated proteases showed that both
masked CARs exhibited similar cytotoxicity towards H292 and MDA-MB-231 cells at
every effector to target ratio tested (Figures 5A and 5B). These results suggest that
mRNA display is a viable method for the identification of CAR masking peptides.
Although Peptide 5-1 has a similar core sequence to Mask 1, it is able to achieve the
same level of binding to cetuximab as Mask 1 without having an N-terminal spacer
segment. Despite this difference, Peptide 5-1 also demonstrated an ability to prevent
EGFR binding in vitro that was comparable to that of Mask 1.
5.4: Conclusion and Future Directions
CAR T cells have been shown to be remarkably effective at treating hematological
malignancies. However, due to the lack of tumor exclusive antigens, the antigens that
current CAR T cells target are often also expressed on healthy tissues. As a result,
there is a considerable risk of on-target, off-tumor effects that must be addressed when
engineering CAR T cells. In this report, we discuss using mRNA display to identify
masking peptides in order to mitigate the risk of on-target, off-tumor effects from CAR T
cell therapy. We identified Peptides 5-1 and 5-3, which demonstrated a substantially
higher level of binding to cetuximab than its parent, Mask 1. We then used each peptide
to prevent the recognition of EGFR by an EGFR-targeting CAR and demonstrated that
all masking peptides (Mask 1, Peptide 5-1, and Peptide 5-3) were able to block EGFR
binding successfully. Furthermore, we showed that CAR binding was restored for all
masking sequences when the masked CAR T cells were treated with an appropriate
167
amount of uPA, a tumor-associated protease, and that the masked CAR T cells were
able to kill protease-secreting tumor cells that expressed EGFR. Taken together, these
results show that the ligands identified by mRNA display against cetuximab are able to
be used to prevent on-target, off-tumor binding of cetuximab-derived, EGFR-targeting
CAR T cells.
168
5.5: Figures:
Figure 1: (A) Library construction. The DNA library was doped at 61% on the nucleotide
level (At each highlighted position, there is a 61% chance of the indicated nucleotide
being present, and a 13% chance of any other nucleotide). X represents a random
amino acid. N: 25%A, 25% T, 25% C, 25% G. S: 50% C, 50% G . (B) Binding of pools
generated by mRNA display against cetuximab at 4 °C.
169
Figure 2: Selective pressures are applied to allow for the identification of Peptides 5-1
and 5-3. (A) Schematic of the temperature selective pressure used to identify cetuximab
ligands that could bind at 37 °C. (B) Binding of Pools 3, 4, and 5 at the indicated
temperature and washing conditions. (C) Chart showing enrichment of Peptides 5-1 and
5-3 after subjecting the pools to binding and washing at 37 °C.
170
Figure 3: (A) Comparison of Peptide 5-1 and Mask 1 binding to cetuximab at 4 °C. (B)
Comparison of 5-3, 5-3, and Mask 1 (each with a peptide spacer, QGQSGQ) binding to
cetuximab at 37 °C.
171
Figure 4: Mask 1 and Peptide 5-1 are both capable of blocking CAR T cell recognition of
EGFR. CAR T cells expressing the indicated masked or unmasked CAR were treated
with the indicated concentration of uPA, washed, and incubated with recombinant
human EGFR-Fc (rhEGFR-Fc). Binding was measured by flow cytometry. NT =
nontransduced T cells.
0
25
50
75
100
0 5 10 20 50 100
Percent rhEGFR-Fc binding
Concentration of uPA (nM)
172
Figure 5: Masked CAR T cells are able to kill protease-secreting tumor cells in vitro. The
indicated CAR T cell, or nontransduced T cell (NT), were cocultured with H292 (A) or
MD-MB-231 (B) tumor cells at the indicated effector to target ratio. Cell killing was
measured by flow cytometry.
173
5.6: References
1. Roberts, R.W., Szostak, J.W. RNA-peptide fusions for the in vitro selection of peptides
and proteins. Proc Natl Acad Sci U S A (1997).
2. Bazan, J., Calkosinski, I. & Gamian, A. Phage display--a powerful technique for
immunotherapy: 1. Introduction and potential of therapeutic applications. Hum Vaccin
Immunother 8, 1817-28 (2012).
3. Pepper, L.R., Cho, Y.K., Boder, E.T., E.V. Shusta. A decade of yeast surface display
technology: Where are we now? Comb Chem High Throughput Screen. (2008).
4. Azzam, M.E., Algranati, I. D. . Mechanism of Puromycin Action: Fate of Ribosomes after
Release of Nascent Protein Chains from Polysomes. Proc Natl Acad Sci U S A (1973).
5. Barrick, J.E., Takahashi, T.T., Ren, J., Xia, T. & Roberts, R.W. Large libraries reveal
diverse solutions to an RNA recognition problem. Proc Natl Acad Sci U S A 98, 12374-8
(2001).
6. Howell, S.M. et al. Serum stable natural peptides designed by mRNA display. Sci Rep 4,
6008 (2014).
7. Morelli, A., Haugner, J., Seelig, B. Thermostable Artificial Enzyme Isolated by In Vitro
Selection. PLOS One (2014).
8. Li, S., Millward, S., Roberts, R. In Vitro Selection of mRNA Display Libraries Containing
an Unnatural Amino Acid. JACS Communications (2002).
9. Millward, S.W., Takahashi, T.T., Roberts, R.W. A General Route for Post-Translational
Cyclization of mRNA Display Libraries. JACS Communications (2005).
10. Millward, S.W., Fiacco,S., Austin, R.J., Roberts, R.W. Design of Cyclic Peptides That
Bind Protein Surfaces with Antibody-Like Affinity. ACS Chemical Biology (2007).
174
11. Yu, H. & Jove, R. The STATs of cancer--new molecular targets come of age. Nat Rev
Cancer 4, 97-105 (2004).
12. Takahashi, T.T. & Roberts, R.W. In vitro selection of protein and peptide libraries using
mRNA display. Methods Mol Biol 535, 293-314 (2009).
13. Banerjee, K. & Resat, H. Constitutive activation of STAT3 in breast cancer cells: A
review. Int J Cancer 138, 2570-8 (2016).
14. Wei, D. et al. Stat3 activation regulates the expression of vascular endothelial growth
factor and human pancreatic cancer angiogenesis and metastasis. Oncogene 22, 319-
29 (2003).
15. Pedranzini, L., Leitch, A., Bromberg, J. Stat3 is required for the development of skin
cancer. J. Clin. Invest. (2004).
16. Yu, H., Lee, H., Herrmann, A., Buettner, R. & Jove, R. Revisiting STAT3 signalling in
cancer: new and unexpected biological functions. Nat Rev Cancer 14, 736-46 (2014).
17. Niu, G., Wright, K.L., Huang, M., Song, L., Haura, E., Turkson, J., Zhang, S., Wang, T.,
Sinibaldi, D., Coppola, D., Heller, R., Ellis, L.M., Karras, J., Bromberg, J., Pardoll, D.,
Jove, R., Yu, H. Constitutive Stat3 activity up-regulates VEGF expression and tumor
angiogenesis. Oncogene (2002).
18. Aoki, Y., Feldman, G.M., Tosato, G. Inhibition of STAT3 signaling induces apoptosis and
decreases survivin expression in primary effusion lymphoma. Blood (2003).
19. Konnikova, L., Kotecki, M., Kruger, M.M., Cochran, B.H. Knockdown of STAT3
expression by RNAi induces apoptosis in astrocytoma cells. BMC Cancer (2003).
20. Ostrand-Rosenberg, S. & Sinha, P. Myeloid-derived suppressor cells: linking
inflammation and cancer. J Immunol 182, 4499-506 (2009).
21. Zhang, X. & Darnell, J.E., Jr. Functional importance of Stat3 tetramerization in activation
of the alpha 2-macroglobulin gene. J Biol Chem 276, 33576-81 (2001).
175
22. Zhang, T., Kee, W.H., Seow, K.T., Fung, W., Cao, X. The coiled-coil domain of Stat3 is
essential for its SH2 domain-mediated receptor binding and subsequent activation
induced by epidermal growth factor and interleukin-6. Molecular and Cellular Biology
(2000).
23. Horvath, C.M., Wen, Z., Darnell Jr., J.E. A STAT protein domain that determines DNA
sequence recognition suggests a novel DNA-binding domain. Cold Spring Harbor
Laboratory Press (1995).
24. Levy, D.E. & Lee, C.-k. What does Stat3 do? J. Clin. Invest. 109, 1143-1148 (2002).
25. Siveen, K.S. et al. Targeting the STAT3 signaling pathway in cancer: role of synthetic
and natural inhibitors. Biochim Biophys Acta 1845, 136-54 (2014).
26. Darnell Jr., J.E. STATs and Gene Regulation. Science (1997).
27. Leong, P.L. et al. Targeted inhibition of Stat3 with a decoy oligonucleotide abrogates
head and neck cancer cell growth. Proc Natl Acad Sci U S A 100, 4138-43 (2003).
28. Schust, J., Sperl, B., Hollis, A., Mayer, T.U. & Berg, T. Stattic: a small-molecule inhibitor
of STAT3 activation and dimerization. Chem Biol 13, 1235-42 (2006).
29. Zhang, X. et al. Orally bioavailable small-molecule inhibitor of transcription factor Stat3
regresses human breast and lung cancer xenografts. Proc Natl Acad Sci U S A 109,
9623-8 (2012).
30. Zhang, X. et al. A novel small-molecule disrupts Stat3 SH2 domain-phosphotyrosine
interactions and Stat3-dependent tumor processes. Biochem Pharmacol 79, 1398-409
(2010).
31. Lin, L. et al. A Novel Small Molecule, LLL12, Inhibits STAT3 Phosphorylation and
Activities and Exhibits Potent Growth-Suppressive Activity in Human Cancer Cells.
Neoplasia 12, 39-IN5 (2010).
32. Haan, S. et al. Characterization and Binding Specificity of the Monomeric STAT3-SH2
Domain. J Biol Chem (1999).
176
33. Ren, Z., Cabell, L.A., Schaefer, T.S. & McMurray, J.S. Identification of a High-Affinity
Phosphopeptide Inhibitor of Stat3. Bioorganic & Medicinal Chemistry Letters 13, 633-636
(2003).
34. Turkson, J. et al. Phosphotyrosyl peptides block Stat3-mediated DNA binding activity,
gene regulation, and cell transformation. J Biol Chem 276, 45443-55 (2001).
35. Goodwin, D., Simerska, P. & Toth, I. Peptides as Therapeutics with enhanced
bioactivity. Current Medicinal Chemistry (2012).
36. Aldrich, S.
37. Buckley, D.L. et al. Small-molecule inhibitors of the interaction between the E3 ligase
VHL and HIF1alpha. Angew Chem Int Ed Engl 51, 11463-7 (2012).
38. Raina, K. et al. PROTAC-induced BET protein degradation as a therapy for castration-
resistant prostate cancer. Proc Natl Acad Sci U S A 113, 7124-7129 (2016).
39. Kalos, M. et al. T Cells with Chimeric Antigen Receptors Have Potent Antitumor Effects
and Can Establish Memory in Patients with Advanced Leukemia. Sci Transl Med 3, 1-13
(2011).
40. Grupp, S.A. et al. Chimeric antigen receptor-modified T cells for acute lymphoid
leukemia. N Engl J Med 368, 1509-1518 (2013).
41. Maude, S.L. et al. Chimeric antigen receptor T cells for sustained remissions in
leukemia. N Engl J Med 371, 1507-17 (2014).
42. Townsend, M.H., Shrestha, G., Robison, R.A. & O'Neill, K.L. The expansion of
targetable biomarkers for CAR T cell therapy. J Exp Clin Cancer Res 37, 163 (2018).
43. Beatty, G.L. et al. Mesothelin-specific chimeric antigen receptor mRNA-engineered T
cells induce anti-tumor activity in solid malignancies. Cancer Immunol Res 2, 112-20
(2014).
177
44. Ahmed, N. et al. Human Epidermal Growth Factor Receptor 2 (HER2) -Specific Chimeric
Antigen Receptor-Modified T Cells for the Immunotherapy of HER2-Positive Sarcoma. J
Clin Oncol 33, 1688-96 (2015).
45. Wang, L.C. et al. Targeting fibroblast activation protein in tumor stroma with chimeric
antigen receptor T cells can inhibit tumor growth and augment host immunity without
severe toxicity. Cancer Immunol Res 2, 154-66 (2014).
46. Caruso, H.G. et al. Tuning Sensitivity of CAR to EGFR Density Limits Recognition of
Normal Tissue While Maintaining Potent Antitumor Activity. Cancer Res 75, 3505-18
(2015).
47. Liu, X. et al. Affinity-Tuned ErbB2 or EGFR Chimeric Antigen Receptor T Cells Exhibit
an Increased Therapeutic Index against Tumors in Mice. Cancer Res 75, 3596-607
(2015).
48. Bonifant, C.L., Jackson, H.J., Brentjens, R.J. & Curran, K.J. Toxicity and management in
CAR T-cell therapy. Mol Ther Oncolytics 3, 16011 (2016).
49. Hoyos, V. et al. Engineering CD19-specific T lymphocytes with interleukin-15 and a
suicide gene to enhance their anti-lymphoma/leukemia effects and safety. Leukemia 24,
1160-70 (2010).
50. Budde, L.E. et al. Combining a CD20 chimeric antigen receptor and an inducible
caspase 9 suicide switch to improve the efficacy and safety of T cell adoptive
immunotherapy for lymphoma. PLoS One 8, e82742 (2013).
51. Gargett, T. & Brown, M.P. The inducible caspase-9 suicide gene system as a "safety
switch" to limit on-target, off-tumor toxicities of chimeric antigen receptor T cells. Front
Pharmacol 5, 235 (2014).
52. Roybal, K.T. et al. Precision Tumor Recognition by T Cells With Combinatorial Antigen-
Sensing Circuits. Cell 164, 770-9 (2016).
178
53. Desnoyers, L.R. et al. Tumor-Specific Activation of an EGFR-Targeting Probody
Enhances Therapeutic Index. Sci Transl Med (2013).
54. Han, X. et al. Masked Chimeric Antigen Receptor for Tumor-Specific Activation. Mol
Ther 25, 274-284 (2017).
55. Takahashi, T.T., Austin, R.J. & Roberts, R.W. mRNA display: ligand discovery,
interaction analysis and beyond. Trends in Biochemical Sciences 28, 159-165 (2003).
56. Jackson, R.J. & Hunt, T. Preparation and Use of Nuclease-Treated Rabbit Reticulocyte
Lysates for the Translation of Eukaryotic Messenger RNA. Methods in Enzymology 96,
50-74 (1983).
Abstract (if available)
Abstract
The use of immunotherapies to treat cancer has increased rapidly due to its proven efficacy in clinical trials. However, their efficacy when treating solid tumors are undermined by the immunosuppressive tumor microenvironment's (TME) ability to hinder the activity of endogenous T cells and prevent them from fully working in conjunction with the therapeutic. Cancer associated fibroblasts (CAFs) are stromal cells that are known to have a key role in maintaining the immunosuppressive nature of the TME. As a result, we used an engineered immunotoxin (α-FAP-PE38hu) to target CAFs in vivo and demonstrate its synergistic effects when used with PD-1 antibody blockade or chimeric antigen receptor (CAR) T cell therapy—two widely adopted forms of immunotherapy. We observed a marked inhibition of tumor growth in the mice receiving α-FAP-PE38hu with either PD-1 or CAR T cell therapy, when compared to the groups receiving any single treatment. Furthermore, we report an increase in the presence of tumor infiltrating T cells as well as a decrease in the presence of immune suppressive cells among mice receiving α-FAP-PE38hu combined with either PD-1 Blockade or CAR T cell therapy the combined therapy. These results highlight the ability of α-FAP-PE38hu, an engineered, CAF-targeting immunotoxin, to enhance the overall efficacy of immunotherapies.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Combination therapy for solid tumor
PDF
Engineering chimeric antigen receptor-directed immune cells for enhanced antitumor efficacy in solid tumors
PDF
Improving antitumor efficacy of chimeric antigen receptor-engineered immune cell therapy with synthetic biology and combination therapy approaches
PDF
Engineering chimeric antigen receptor (CAR) -modified T cells for enhanced cancer immunotherapy
PDF
Novel design and combinatory therapy to enhance chimeric antigen receptor engineered T cells (CAR-T) efficacy against solid tumor
PDF
Engineering immunotoxin and viral vectors for cancer therapy
PDF
Self-secretion of checkpoint blockade enhances antitumor immunity by murine chimeric antigen receptor-engineered T cells
PDF
Mechanistic model of chimeric antigen receptor T cell activation
PDF
Site-specific PEGylation of recombinant immunotoxin αFAP-PE38 for half-life extension
PDF
Chimeric Antigen Receptor targeting Prostate Specific Membrane Antigen (PSMA)
PDF
Phospho-proteomic analysis of immune cell activation
PDF
Enhancing chimeric antigen receptor-engineered immune cell therapy with synthetic biology and nanomedicine
PDF
Lym-1 epitope targeted chimeric antigen receptor (CAR) T cells for the immunotherapy of cancer
PDF
Generation and characterization of anti-CD138 chimeric antigen receptor T (CAR-T) cells for the treatment of hematologic malignancies
PDF
Optimization of nanomedicine based drug delivery systems for the treatment of solid tumors
PDF
Development of TCR-like antibody and novel chimeric antigen receptor for cancer immunotherapy
PDF
The effect of tumor-mediated immune suppression on prostate cancer immunotherapy
PDF
Enhancing the anti-cancer specificity of chimeric antigen receptor T cells through targeting HLA loss
PDF
Engineered control of the CAR-T-tumor synapse using customized DNA linkers
PDF
Engineered CAR-T cells for treatment of solid cancers
Asset Metadata
Creator
Mac, John Edward
(author)
Core Title
Engineering therapeutics for the improved antitumor efficacy of chimeric antigen receptor T cell therapy
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Chemical Engineering
Publication Date
01/11/2020
Defense Date
05/13/2019
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
cancer therapeutic,CAR T cell,immunotherapy,immunotoxin,mRNA display,OAI-PMH Harvest,Stat3
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Roberts, Richard (
committee chair
), Wang, Pin (
committee chair
), Graham, Nicholas (
committee member
)
Creator Email
jmac@usc.edu,johnmac1990@g.ucla.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-180856
Unique identifier
UC11660425
Identifier
etd-MacJohnEdw-7544.pdf (filename),usctheses-c89-180856 (legacy record id)
Legacy Identifier
etd-MacJohnEdw-7544.pdf
Dmrecord
180856
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Mac, John Edward
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
cancer therapeutic
CAR T cell
immunotherapy
immunotoxin
mRNA display
Stat3