5. Artificial antigen

advertisement
The right touch:
T cell activation via artificial antigenpresenting cells.
Joep van der Weijden
UU – Drug Innovation
May 2013
I. Abbreviations
aAPC
ACT
APC
CAR
CCL
CD
CTL
DC
DN
DP
EPR effect
GMP
HLA
ICAM-1
IFN-γ
Ig
IL
IS
Kd
LFA-1
MHC (I/II)
nLGs
Ag
PAMPs
PBMC
PEG
PGA
PLA
PLGA
pMHC
PS
RBC
SLB
SMAC
TCR
TGF-β
Th
TIL
artificial antigen-presenting cell
adoptive cell transfer
antigen-presenting cell
chimeric antigen receptor
(C-C) motif ligand
cluster of differentiation
cytotoxic T cell
dendritic cell
double negative
double positive
enhanced permeability and retention effect
good manufacturing practice
human leukocyte antigen
intercellular adhesion molecule 1
interferon gamma
immunoglobulin
interleukin
immune synapse
dissociation constant
lymphocyte function-associated antigen 1
major histocompatibility complex class I/II
nanosized liposomal polymeric gels
antigen
pathogen associated molecular patterns
peripheral blood mononuclear cells
polyethylene glycol
polyglutamic acid
polylactic acid
poly(lactic-co-glycolic acid)
peptide-major histocompatibility complex
polystyrene
red blood cells
supported lipid bilayer
supramolecular activation cluster
T cell receptor
transforming growth factor beta
helper T cell
tumor infiltrating lymphocyte
2
Table Of Contents
I. Abbreviations ....................................................................................................................................... 2
1. Introduction ......................................................................................................................................... 5
2. The immune system[7,8] ....................................................................................................................... 6
2.1 Innate immunity ............................................................................................................................ 6
2.2 Adaptive immunity ........................................................................................................................ 7
2.3 Principles of T cell activation ......................................................................................................... 9
2.4 The immunological synapse ........................................................................................................ 11
3. Cancer immunotherapy..................................................................................................................... 15
3.1 Principles of cancer immunotherapy .......................................................................................... 15
3.2 Cancer immunotherapy via dendritic cells .................................................................................. 16
3.3 Cancer immunotherapy via adoptive T cell transfer ................................................................... 17
4. Biomimetic immunomodulation ....................................................................................................... 19
4.1 Biomimetic material design ......................................................................................................... 19
4.2 Physical and mechanical material properties.............................................................................. 20
4.2.1 Size........................................................................................................................................ 20
4.2.2 Shape .................................................................................................................................... 24
4.2.3 Rigidity .................................................................................................................................. 25
4.2.4 Zeta potential ....................................................................................................................... 25
4.3 Ligand presentation..................................................................................................................... 26
4.3.1 Presentation of surface bound ligands................................................................................. 26
4.3.2 Presentation of soluble ligands ............................................................................................ 27
4.3.3 Surface bound ligands that prolong circulation time ........................................................... 29
5. Artificial antigen-presenting cells ...................................................................................................... 30
5.1 Basic components of an aAPC ..................................................................................................... 30
5.2 Scaffolds used for artificial antigen presentation ....................................................................... 31
5.2.1 Liposomes ............................................................................................................................. 32
5.2.2 Supported lipid bilayers........................................................................................................ 32
5.2.3 Polystyrene beads ................................................................................................................ 33
5.2.4 Biodegradable particles ........................................................................................................ 34
5.2.5 Other scaffolds ..................................................................................................................... 34
6. Conclusions ........................................................................................................................................ 36
II. Summary............................................................................................................................................ 38
3
III. Samenvatting.................................................................................................................................... 39
IV. Bibliography ..................................................................................................................................... 41
4
1. Introduction
The immune system constantly protects our body from a myriad of attacks from potentially lethal
pathogens. Key to the success of our immune system lies in its ability to develop immunity to
pathogens it has never encountered before. Professional antigen-presenting cells (APCs) known as
dendritic cells (DCs) are vital in directing this so called adaptive immune response. Next to their role
of protecting us from external pathogens, DCs are required for the generation of anti-tumor
immunity[1,2]. If we could mimic DCs synthetically, we could potentially harness the power of the
immune system for cancer immunotherapy.
When a virus or bacteria manages to break through the skin, it is likely to encounter a dendritic cell.
Via a process known as phagocytosis, the dendritic cell engulfs the pathogen and breaks down its
structure. Next, it will present parts of the pathogens as antigens via a membrane protein known as
the Major Histocompatibility Complex (MHC) on its cell surface. After the DC migrate to the lymphoid
organs, this MHC-antigen complex is then used to activate T cells, a second set of specialized cells of
the immune system capable of rapid and efficient killing of the pathogen or pathogen-infected cells.
Similarly, the immune system scans the body for abnormalities in cellular secretion products, which is
a potential sign of cancerous growth. When abnormal growth is detected, the cells responsible are
removed. The generation of such anti-tumor immunity is mediated by DCs[1,2] and their function as
APC has already successfully been exploited in the field of cancer immunotherapy. Unfortunately, the
high costs of culturing each patients individual cells and processing them into a vaccine under good
manufacturing practice (GMP) is a potential roadblock for clinical use[3,4].
The search for modular, off-the-shelf DC mimics has led researchers to the production of artificial
antigen-presenting cells (aAPCs). A wide range of acellular scaffolds have been functionalized with
the basic requirements for T cell activation[5] and consequently used to activate T cells. Choosing the
right combination of scaffold and set of T cell activating ligands is vital to eliciting the right T cell
response. Control over T cell activation makes aAPCs not only an attractive alternative to DCs in
cancer immunotherapy, but also an important tool to investigate the factors that influence native T
cell activation.
The importance of scaffold morphology in intercellular signal transduction is increasingly recognized.
Factors such as avidity or domain formation of presented signals on the one hand, and overall
scaffold shape, rigidity and size on the other can influence the outcome of cellular signaling events.
With this in mind, there have been increased attempts to use techniques from (bio)chemical
engineering to more closely mimic natural signal transduction[6].
This review explores the role scaffold morphology in cellular interaction and signaling. Its focus will
be on the biomimicry of APCs in the context of cancer immunotherapy using artificial scaffolds. To
this end, the basic biology behind DC-based cancer immunotherapy is explained, followed by an
overview of aAPCs described in literature.
5
2. The immune system[7,8]
The immune system is the name given to the cells and molecules that protect our body from damage
caused by infectious agents or other unwanted intrusion. An immune response is generated when
the immune system recognizes an antigen as “non-self”. This sets the various effector cells and
molecules of the immune system into action to clear the body of the antigen source.
A wide range of pathogens are always recognized by the immune system, because they possess
common antigens. These antigens trigger an innate immune response. Other pathogens, such as
influenza, are more elusive and may not be recognized immediately. They require a more specific
response (the production of antibodies, for example) in order to be eliminated. Such a response is
called an adaptive immune response, because it requires the body to adapt to infection from that
particular pathogen. An adaptive immune response is a form of immunological memory, a
phenomenon gratefully exploited in vaccination.
2.1 Innate immunity
White blood cells, or leukocytes, form the cellular part of both the innate and adaptive immune
system, shown in figure 2. Leukocytes originate in the bone marrow and following maturation, they
differentiate into more specialized cell types as they move out to the bloodstream, the lymphatic
system and peripheral tissues. Essential cell types of the innate immune system are macrophages,
granulocytes, mast cells and DCs. All of these are formed from the common myeloid precursor, a
lineage of leukocytes. A common feature of these cells lies in their ability to engulf, or phagocytose,
and destroy particles such as bacteria or even complete cells.
The major population of leukocytes is called granulocytes, owing to their grainy staining when
studied under a microscope. The peculiar staining comes from granules in their cytoplasm, which
consist of a mixture of pore forming proteins capable of destroying microbes via lysis of the
membrane. Granulocytes can be divided into three subtypes; eusinophils, basophils and neutrophils,
of which the latter is the most abundant leukocyte in the blood.
Macrophages, a mature form of monocytes shown in figure
1, are long lasting, phagocytic cells present throughout the
body. Their primary function is to engulf and enzymatically
degrade pathogens they may encounter and since they are
distributed in most tissues, they are often the first line of
defense of the innate immune system. In addition to the
phagocytosis of microbes, macrophages act as the bodies’
janitor, clearing cells and cellular debris. By secreting
signaling proteins, they can recruit other immune cells and
induce inflammation, which is vital to a successful immune
Figure 1. Left, a schematic representation of a response.
macrophage. Right, a micrograph of a
macrophage surrounded by leukocytes[169].
Dendritic cells (DCs) get their name from the finger-like
protrusions, or dendrites formed by the cell membrane.
Similar to macrophages and granulocytes, DCs engulf particles by phagocytosis. In addition, DCs
6
Figure 2. Dendritic cells form a key link between the innate immune system and the adaptive immune
system. They are phagocytic and are specialized to ingest a wide range of pathogens and to display their
antigens at the dendritic cell surface in a form that can be recognized by T cells[169].
continuously “drink” extracellular fluid in a process called macropinocytosis. In contrast to
macrophages and granulocytes, DCs are specialized to process engulfed material intracellularly and
present antigenic fragments through specialized receptors on their surface. This form of antigen
presentation activates a type of cell from the adaptive immune system known T lymphocytes and
gives DCs their role as a ‘bridge’ between the innate and the adaptive immune system. Although
some other cell types such as macrophages are also capable of antigen presentation, DCs are the
most potent type of APC[9]. The exact mechanism by which DCs present antigens to activate T
lymphocytes is discussed in detail in section 2.3.
2.2 Adaptive immunity
An important lineage from which many different cell types associated with the adaptive immune
system arise is the common lymphoid progenitor. Collectively known as lymphocytes, these cells are
mainly found in the blood and the lymphatic system. Most lymphocytes are antigen-specific cells,
meaning that they can respond to one antigen only. The only exception comes from the natural killer
cell (NK cell), which is specialized in killing tumor or virus infected cells.
The remainder of the lymphocyte subtypes are antigen-specific. Because of the vast number of
pathogens that our body might encounter during our lifetime, a huge repertoire of different antigenspecific cells exists in our body. This is possible through the modular way in which an antigen
receptor is produced. By continuous recombination of the genes that code for the antigen receptor,
our body can produce up to a billion different lymphocytes. Only when a lymphocyte is presented
with the antigen specific to its antigen receptor, it goes from a naïve to an activated state. In the
activated state, lymphocytes are called effector cells and perform various immunological functions in
the body.
There are two types of lymphocytes, the B cells and the T cells. B cells are most popularly known as
the source for antibodies. After their antigen receptor recognizes an antigen, B cells transform into
plasma cells or effector B cells. These cells produce and secrete large amounts of immunoglobulin
(Ig) proteins, which is a soluble form of the antigen receptor on their cell membrane. This protein,
7
also known as an antibody, helps the immune system to recognize a specific pathogen, thereby
promoting its removal from the body.
Although T cells also recognize antigens through an antigen receptor (the T cell receptor, or TCR),
their effector functions are different from B cells. T cells can differentiate into several different
effector subtypes, which can be
categorized
as
cytotoxic
(specialized in killing), helper
(stimulates other T cells and B
cells)
or
regulatory
(controls/suppresses other T
cells).
T cells can be found in the
blood and the lymph. They
differentiate from the common
lymphoid precursor in the bone
marrow, a primary lymphoid
organ. From there, they migrate
to the thymus where they
mature. They then go on to
secondary lymphoid organs
such as the lymph nodes and
the spleen where fully matured
naïve T cells wait to be
stimulated
by
antigenpresenting
cells.
The
distribution of the different lymphoid organs through the body is shown in figure 3.
When lymphocytes enter the thymus, they lack a functional T cell receptor nor do they express
cluster of differentiation 4 (CD4) or CD8, both TCR co-receptors. At this stage, they are called double
negative (DN) cells. As the DN cells proliferate, they start to express low levels of CD4, CD8 and the
TCR, which consists of an α and β chain as
Figure 3. The distribution of lymphoid tissues in the body. T cells arise from stem
cells in bone marrow and differentiate in the thymus (yellow). They migrate from shown in figure 4.
the thymus and are carried in the bloodstream to the peripheral lymphoid organs
(blue), which include lymph nodes and the spleen. The peripheral lymphoid
organs are the sites of lymphocyte activation by antigen, and lymphocytes
recirculate between the blood and these organs until they encounter their
specific antigen[169].
The incredible diversity of
TCRs comes from the
somatic recombination of
the genes that code for the α
and β chains respectively. This process, known as V(D)J recombination, is the
origin of TCR diversity. After translation of the genes that code for the TCR, the
full α:β heterodimer is formed at the surface of the cells, which are now called
double positive (DP) cells because they express both the CD4 and CD8 coreceptor.
Cells that engage with TCRs, such as DCs, do this through antigens Figure 4. Schematic structure of a T cell
bound to MHC molecules on their surface. There are two classes antigen receptor. The TCR is composed
of two chains, an α chain (yellow) and a
β chain (green), each of which has a
variable and a constant part. The
variable parts of the two chains create a
variable region, which forms the
antigen-binding site[169].
8
of MHC molecules; those that bind endogenous antigens (MHC class I) and those that bind
extracellular antigens (MHC class II).
To ensure that the newly formed TCRs function properly, they
need to be able to bind to antigen/MHC complexes but avoid
strong binding to self-antigens since that may lead to autoimmunity.
The desired binding properties of TCRs is determined by
positive and negative selection in the thymus. Positive selection
ensures that the newly formed TCR has at least a weak affinity
for an MHC complex. Double positive T lymphocytes that fail to
bind an MHC complex through their TCR during positive
selection die within 3-4 days. DP cells that survive positive
selection will eventually become single positive (CD4 or CD8) T
lymphocytes. During negative selection, apoptosis is induced in
cells that bind too strongly to self-antigen/MHC complexes.
Less than 5% of all DP cells survive this process, but such strict
selection ensures that all single positive T lymphocytes that
leave the thymus can recognize MHC and are self-tolerant.
These matured cells, now called single positive (CD4 or CD8)
naïve T lymphocytes, move to the periphery and end up in
secondary lymphoid organs where they are maintained in low
numbers until they meet “their” antigen and rapidly expand in
numbers. This process is briefly illustrated in figure 5.
Figure 5. Clonal selection. A single lymphocyte precursor gives rise to a large
number of lymphocytes, each bearing a distinct antigen receptor. Lymphocytes
with receptors that bind ubiquitous self antigens are eliminated before they
become fully mature, ensuring tolerance to such self antigens. This is known as
negative selection. When a foreign antigen interacts with the receptor on a
mature naive lymphocyte, that cell is activated and starts to clonally expand.
Antigen specificity is thus maintained as the clonally expanded cells proliferate
and differentiate into effector cells. Once antigen has been eliminated by these
effector cells, the immune response ceases, although some lymphocytes are
retained to mediate immunological memory[169].
2.3 Principles of T cell
activation
Mature, naïve T cells that
circulate the blood and the lymphatic system need to be activated by antigen-presenting cells before
they are fully functional. Once activated, T cells become effector T cells and perform various
immunological tasks.
Most nucleated cells display MHC molecules on their cell surface, but only DCs are capable of
activating T cells. DCs continuously sample their surroundings for pathogens and depending on the
way a pathogen enters the dendritic cell, the antigen fragments derived from the pathogens are
loaded onto different types of MHC molecules. Extracellular pathogens that enter the cell through
either receptor-mediated phagocytosis or macropinocytosis are loaded onto MHC class II receptors
9
and pathogens that are located directly in the cytosol (such as most viruses) are loaded onto MHC
class I molecules after proteolytic degradation. However, some exogenous proteins internalized by
DCs end up on MHC class I complexes, a phenomenon referred to as cross presentation. The
distinction between peptide-MHC I (pMHC I) and pMHC II complexes on the APC surface is important
because it determines the type of T cell that gets activated.
After DCs capture an antigen, they move from the peripheral tissues like the skin to draining lymph
nodes via lymphatic vesicles and mature[10]. In the lymph nodes, antigen presentation to naïve T cells
takes place[11].
Activation of naïve T cells by APCs requires three signal, as illustrated in figure 6. The first signal in T
cell activation is the recognition of a pMHC complex by a TCR on the T cell surface. The bond
between a single pMHC:TCR is rather weak, but it can be stabilized by either CD4 or CD8, both co
receptors of the TCR on the T cell surface. The pMHC II:TCR complex is stabilized by CD4 and the
pMHC I:TCR complex is stabilized by CD8. Since single positive, naïve T cells express only one of the
two co receptors, this gives rise to two different populations of
effector T cells (CD4+ or CD8+) after activation.
A second signal is provided by co stimulatory molecules on the
APC, such as CD80 (B7.1) or CD86 (B7.2). These surface
molecules interact with CD28 on naïve T cells. This second
signal is needed for further activation of the T cell. Stimulation
solely through the TCR, without co stimulatory signals, can lead
to a state of anergy in the T cell. In this state, T cells are
selectively non-responsive when presented with an antigen
later on[12]. Adhesion molecules, such as the integrin LFA-1 on
the T cell which interacts with ICAM-1 on the APC surface can
be seen as another co stimulatory signal, since they serve to
prolong and stabilize APC:T cell contact. A third signal comes
from soluble signaling molecules, collectively known as
cytokines. This signal influences to which effector cell a naïve T
cell will differentiate.
Figure 6. Three kinds of signals are
involved in activation of naive T cells by
antigen-presenting cells. Binding of the
pMHC complex by the TCR and, in this
example, a CD4 co-receptor, transmits an
activation signal. Effective activation of
naive T cells requires a second signal, the
co-stimulatory signal, to be delivered by
the same antigen- presenting cell (APC).
For CD4 T cells in particular, different
pathways of differentiation produce
subsets of effector T cells that carry out
different effector responses, depending on
the nature of a third signal (arrow 3)
delivered by the antigen-presenting cell.
Cytokines are commonly, but not
exclusively, involved in directing this
differentiation[169].
The nature of the T cell response is strongly related to the
nature of the presentation of the different activation signals.
Activation of CD8+ T cells occurs through TCR recognition of a
pMHC I complex (signal 1) on the APC and co stimulation
through CD28 on the T cell surface. The effector cells of the
CD8+ subtype are called cytotoxic T cells (CTLs) and these cells
are capable of killing cells in an antigen-selective fashion. For
example, when a CTL encounters a virus-infected cell
presenting a viral peptide derived antigen on its MHC I
complex, it will kill that cell.
Activation of CD4+ T cells can lead to a variety of effector cells,
of which the subtype is determined by the different cytokines
(signal 3). When a naïve CD4+ T cell receives a signal 1 (through
10
pMHC II) and signal 2 from an APC in combination with the cytokine IL-12, it is likely to differentiate
towards the T helper 1 (TH1) subtype. As the name implies, this effector type is important in aiding
the (cellular) immune response by secreting another cytokine called IL-2, a potent T cell growth
factor. Cells of the TH1 subtype are also efficient stimulators of macrophage activity, thereby
increasing macrophage microbial activity. When IL-4 is present as a third signal, CD4+ T cells are
prone to differentiate towards the TH2 subtype. T helper 2 cells are involved in humoral immune
response by the activation of B cells. Another group of CD4+ T cells is called regulatory T cells, or Treg
and these cells are important in preventing over-reactive CTLs from causing excess damage to
tissues. One of the ways that regulatory T cells are induced is by antigen-loaded DCs which have
failed to mature[13,14].
2.4 The immunological synapse
Here, we zoom in on the T cell:DC interaction to discuss the spatial and temporal elements of T cell
activation and the minimal requirements for signal transduction.
The site of signaling from a DC to a naïve T cell is known as the immunological synapse (IS), or
supramolecular activation cluster (SMAC). This is a highly organized structure on the cell surface
several micrometers across, with different stimulatory ligands clustered in spatially organized
domains[15].
Figure 7. Structure of the immunological synapse. (A) The three layers on the T cell side of the immunological synapse are
pictured here. The receptor layer contains the T cell receptor (TCR)–CD3 complex, CD4 or CD8, CD28 and leukocyte
function-associated antigen 1 (LFA1). The signaling layer includes downstream phosphorylation proteins and the
cytoskeletal layer contains mainly filamentous actin (F-actin). (B) The figure represents the supramolecular activation
clusters (SMACs) of an immunological synapse, as seen looking down on a T cell. The TCR-rich central SMAC (cSMAC) core is
shown in red, the CD28-rich cSMAC periphery is shown in green and the LFA1-rich peripheral SMAC (pSMAC) is shown in
blue. The entire contact area between the T cell and antigen-presenting cell is outlined, and the yellow area represents Factin, which forms dynamic protrusions that move around the periphery of the synapse (the distal SMAC (dSMAC)) in a
11
radial wave. Each SMAC contains hundreds to thousands of receptors, and the dSMAC is the site at which TCR microclusters
are first detected[16].
The interaction of a naïve T cell with a dendritic cell can be divided into distinct phases, based on the
strength and duration of contact[17]. In the lymph nodes, naïve T cells constantly scan their
surroundings in search for activating signals on DCs. In this scanning phase which can last from 30
minutes up to 8 hours, DCs are contacted by up to 5000 T cells in 1 h[18]. Initiation of contact between
a DC and a naïve T cell occurs through the recognition of activating pMHC complexes in a sea of selfpMHC on the DC surface by the TCR. As few as 8-10 activating pMHC complexes can be enough to
trigger a response (formation of a synapse) in T cells[19,20].
Following initial contact, activating TCR microclusters, enriched in costimulatory molecules such as
CD28, move inward in an F-actin dependent manner to form the central supramolecular activation
cluster (cSMAC). The central cluster of TCR-pMHC is surrounded by a ring of adhesion molecules
known as the peripheral SMAC or pSMAC, which consists of adhesive contacts between LFA-1 on the
T cell surface and ICAM-1 on the DC[21]. Together, the cSMAC and the pSMAC form a typical bull’s eye
pattern, as shown in figure 7b. The interaction between a DC and a naïve T cell can last up to 8 hours
and this is seen as a requirement of proper T cell activation. Finally, T cells return to their state of
rapid migration and the activated T cell proliferates and exits the lymphoid system in search for its
antigen in the periphery.
From solution based affinity measurements we know that
TCR:pMHC affinity is rather low, which is surprising given
the highly specific interactions that occur between T cells
and APCs. There are various explanations for this
discrepancy, most of which have to do with the fact that
solution-based affinity is not comparable to the in vivo
situation. It is important to acknowledge the physical
restraints of both TCR and pMHC, as illustrated in figure 8.
These membrane bound proteins have limited degrees of
freedom, which means that upon binding, the loss of
entropy is less than what is to be expected from 3D,
solution-based measurements[22,23]. In addition, the
formation of TCR microclusters makes for a synergistic
enhancement of single TCR:pMHC affinity[12,24]. The tight
junction between the T cell and the DC at the IS, caused by
the LFA-1 enriched pSMAC network also serves to enhance T
cell sensitivity to antigen.
Figure 8. Receptor-ligand kinetics in solution
versus membranes. In theory, dissociation
constants (Kd) for bound ligands in solution are
calculated based on dissociation occurring in 3D
space, with six degrees of freedom. However,
when two opposed cell membranes are
interacting, both receptor and ligand are
confined to 2-D translation and 1-D rotation,
which may stabilize and prolong these receptorligand interactions. Also, rebinding by
neighboring receptors (such as clustered TCR)
can further trap ligands and generate longer
binding than would be predicted in solution[22].
The geometry of the immunological synapse is intriguing,
since it adds contextual information to the traditional ligandreceptor view of signal transduction[25]. Many researchers
have sought to image the IS, either by using ligands
embedded in a supported lipid bilayer (SLB) to mimic the DC
or by in vivo imaging. An antibody against one of the TCR
coreceptors, CD3, shown in figure 7a, can be used to mimic
12
the pMHC complex since it will engage with TCRs and activate naïve T cells regardless of their antigen
specificity.
Figure 9. Immunological synapse spatial mutations. (A) Physical barriers to protein transport on a fluid supported lipid
bilayer. Thin chrome lines create barriers to the diffusion of bilayer-tethered proteins (such as pMHC) and cellular proteins
(such as TCRs) interacting with them (left). The spatial organization of the immunological synapse (TCRs; green) and ICAM1;
red) without (1) and with (2–4) barriers of different geometries (right). (B) Subcellular size protein patterns functionalized
on a surface. A TCR-activating antibody (anti-CD3; green) and an adhesion molecule (ICAM1; purple) are patterned on a flat
surface (left). Anti-CD3, shown in schematics and cell overlays (in which anti-CD3 is blue) can be seen in a wild-type central
zone pattern or in two variant patterns: multifocal and a peripheral ring (right). (C) The subcellular pattern of a TCRactivating antibody (anti-CD3ε; green) and a co-stimulatory antibody (anti-CD28; blue) on an adhesion molecule (ICAM1;
purple)-rich surface (left). Different patterns are tested for their effect on T cell activation: TCR and CD28 follow the pattern
of anti-CD3 and anti-CD28 antibodies, respectively, which can be either co-localized or segregated[26].
13
Vale et al. were able to show TCR clustering and IS formation in Jurkat cells using anti-CD3ε and
ICAM-1 tethered to a planar lipid bilayer, similar to the system shown in figure 9a[27]. Wind et al.
performed a well-based single-cell study in combination with a cellular artificial APC to activate
Jurkat cells and clearly show initial TCR clustering prior to colocalization to a single cSMAC[28]. By
introducing constraints to the lateral mobility of the lipid-tethered ligands, or by patterning the
ligands to a surface, it is possible to study the effect of spatial (re)organization on the T cell response.
Some of these surface manipulations are shown in figure 9. From these studies, it was shown that
TCR microclusters that are physically hindered to move towards the cSMAC continue to signal
intracellularly and that spatial orientation of ligands strongly influences T cell response[16,26,29,30].
These studies are of importance when trying to mimic antigen-presenting cells synthetically, since
they tell us which factors (i.e. ligand mobility, ligand density) are most influential for T cell activation.
Although the studies presented here show the importance of spatial orientation of T cell activating
ligands, we must to keep in mind that these systems are often incomplete mimics of what is
happening in the native cellular environment. When a T cell is presented with pre-patterned ligands
on an artificial surface, it will try to accommodate to the presented pattern, thus influencing the
outcome of the experiment. Even though this is not the case when ligands are presented on a SLB
surface without constraints, such SLB experiments might not translate completely to native T cell
activation. Other factors, such as the choice of ligands that are used in these studies or temporal
context in which they are presented to a T cell, might also affect the size and shape of the IS and thus
the signal received by the T cell[22]. Direct, in vivo imaging of the IS, although not yet technically
feasible, will provide a better picture of the importance of spatial organization in the IS[31]. Such
information could be very useful for the construction of biomimetic aAPCs.
14
3. Cancer immunotherapy
As stated by Palucka and Banchereau in 2012, “cancer immunotherapy attempts to harness the
power and specificity of the immune system to treat tumors”[32]. The idea of using the immune
system to combat diseases is not new. In fact, a major goal of vaccination is to induce T cell mediated
long-lived immunity. The discovery that our immune system can recognize and eliminate cancerous
cells in a DC-mediated manner[1,2], has inspired researchers to study the therapeutic effects of DCs
and T cells in cancer. One example of a recent breakthrough in cancer immunotherapy is the full
recovery of a patient with leukemia after treatment with chimeric antigen receptor (CAR) modified T
cells[33].
3.1 Principles of cancer immunotherapy
Cancer cells often have altered protein expression levels compared to normal cells, which can result
in the display of antigens that are associated with tumors. In order to establish antitumor immunity,
the immune system has to complete several steps, illustrated in figure 10. First, a tumor associated
antigen must be taken up by dendritic cells in an inflammatory context to ensure full maturation of
the DCs. Next, these antigen loaded DCs must mature and find their way to the lymph nodes where
they can interact with naïve T cells.
The exact composition of T cell
subtypes required to establish
antitumor immunity is not known,
but it is certain that it must include
CD8+ cytotoxic T cells (CTLs) [34]. In
the presence of an inflammatory
signal, such as IL-12, these CD8+ T
cells develop into a memory subtype
which
persists
long
after
[35]
activation . Although most of the
work on T cell mediated cancer
immunotherapy has focused on the
generation of CTLs, the CD4+ T cell
subtype has an important role in
sustaining an active immune
response
by
secreting
the
[36]
stimulatory cytokine IL-2 . After
the establishment of an activated
repertoire of tumor-antigen specific
T cells, the CTLs have to reach the
tumor site where they can execute
their cytotoxic function. Several
factors
hamper
the
proper
functioning of CTLs in the tumor
microenvironment, such as limited
Figure 10. Generation and regulation of antitumor immunity. Antitumor
immune responses must begin with the capture of tumor-associated
antigens by dendritic cells, either delivered exogenously or captured from
dead or dying tumor cells. The dendritic cells process the captured antigen
for presentation or cross-presentation on MHC class II and class I
molecules, respectively, and migrate to draining lymph nodes. If capture
and presentation occurred in the presence of an immunogenic maturation
stimulus, dendritic cells will elicit anticancer effector T cell responses in the
lymph node; if no such stimulus was received, dendritic cells will instead
induce tolerance leading to T-cell deletion, anergy or the production of Treg
cells. In the lymph node, antigen presentation to T cells will elicit a
response depending on the type of dendritic cell maturation stimulus
received and on the interaction of T-cell co-stimulatory molecules with
their surface receptors on dendritic cells. Antigen-educated T cells will exit
the lymph node and enter the tumor bed, where a host of
immunosuppressive defense mechanisms can be produced by tumors that
oppose effector T cell function[34].
15
expression of tumor-antigen presenting MHC I molecules by the cancer cells and the presence of
immunosuppressive cytokines.
To help the immune system to generate an effective immune response against cancer, two
fundamentally different promising approaches have evolved. One approach is aimed at the delivery
of (exogenous) tumor antigen to dendritic cells to enhance antigen presentation to T cells, while the
other focuses on ex vivo activation and expansion of patient-derived naïve T cells, followed by
reinfusion of these cells. The latter process is also known as adoptive cells transfer, or ACT. Both
approaches are discussed in more detail below, as they form the therapeutic context for the
development of aAPCs.
3.2 Cancer immunotherapy via dendritic cells
Dendritic cells are capable of capturing tumor antigens and cross present them to naïve T cells in the
lymph nodes to induce an anti-tumor CTL response[1,2], a process which can be exploited in cancer
immunotherapy as illustrated in figure 11.
Within dendritic cell-based cancer
immunotherapy there are two
approaches to increase antigen
presentation by DCs in vivo. One
approach is to grow DCs from
patient-derived monocytes ex vivo,
load them with the antigen of
interest and then inject them back
into the patient. The tumor
antigens can be introduced by
culturing the DCs in the presence
of the antigen; when added in
excess, the tumor antigen will
Figure 11. Antigen capture and presentation by antigen-presenting cells. simply exchange with the antigens
Tumor-associated antigens are exogenously delivered to provide a source of already present on the MHC
antigen. Antigens can then be taken up by antigen-presenting cells,
expressing co stimulatory molecules such as CD80, to be processed and molecules on the DC surface. As an
presented to T cells. Phenotypes of CD4+ T cells can either enhance (eg, Th1) alternative to coculturing the DCs
or inhibit (eg, Treg) the immune response[46].
with a single antigen, tumor cell
lysates can be added to the culture
medium to allow DCs to process tumor antigens through natural pathways. Although this ex vivo
approach has been widely explored[32,37], the labor required to generate antigen-loaded DCs for each
individual patient and the limited lifetime of DCs in culture (a few days), limits its broad applicability
to a clinically relevant setting[3].
Antigens can also be delivered directly to DCs in vivo using targeted delivery of antigen to peripheral
DCs. It was shown by Steinman and colleagues that efficient delivery of antigen to DCs and
consequent CTL response is possible by coupling the antigen to a DEC-205 antibody, a DC specific
surface protein[38]. Other means of targeted antigen delivery include, but are not limited to; antigen-
16
loaded, biodegradable polymeric nanoparticles[39,40], antigen-coupled surface modified polymeric
particles[41] or magnetic particles[42].
Often in DC targeting, a danger signal or adjuvant is provided along with the antigen in the form of an
immunogenic carrier or directly through the coadministration of the cytokine IFN-γ. Collectively, such
immunogenic danger signals are known as pattern-associated molecular patterns (PAMPs). Without
such a danger signal, targeted DCs may become tolerogenic, inducing regulatory T cells that may
hamper an effective immune response[14,43].
Overall, DC based cancer immunotherapy is far from being commonly used as a treatment modality
for cancer. Due to the complex nature of T cell activation by DCs, control over DC maturation and
consequent T cell activation at the level of DCs is poor.
3.3 Cancer immunotherapy via adoptive T cell transfer
The adoptive transfer of ex vivo stimulated, tumor-antigen specific T cells, a therapeutic approach
illustrated in figure 12, bypasses the difficult task of dendritic cell manipulation in vivo, where a
suppressive tumor micro-environment can interfere with DC development. For example, some
tumors are infiltrated by a population of cytotoxic T cells, but due to the immunosuppressive
microenvironment generated by the tumor, these tumor infiltrating lymphocytes (TILs) are not
capable of eliminating the cancerous cells. Systemic administration of IL-2 is known to stimulate T cell
growth and survival, but the high dose required to expand TILs in vivo often leads to significant
toxicities. Combining these findings, researchers have successfully expanded populations of TILs ex
vivo from a tumor biopt by adding IL-2 to the culture medium[44]. The adoptive T cell transfer (ACT) of
such ex vivo expanded TILs has shown to be efficient in the treatment of patients with melanoma[45].
Figure 12. Isolation and ex vivo stimulation of antigen-specific tumor-reactive T cells for adoptive cell transfer. Autologous
lymphocytes are stimulated ex vivo with either antigen-loaded dendritic cells or artificial antigen-presenting cells. Enriched
tumor-specific T cells are then re-infused for treatment[46].
17
Genetic manipulation of ex vivo cultured T lymphocytes to provide them with antigen-specificity
provides another interesting therapeutic opportunity for ACT[47]. The previously mentioned success
with CAR-modified T cells[33] is an example of an ACT-based therapeutic intervention. For the genetic
manipulation of T cells, peripheral blood mononuclear cells (PBMCs) are harvested from a patient
and non-specifically expanded ex vivo by repeated stimulation with anti-CD3 and anti-CD28
antibodies. When the cells are grown to the desired number, they are transfected with the gene of
interest using a viral delivery system after which they are reinfused.
The potential beneficial outcome of ACT is greatly improved if the ex vivo cultured T cells are infused
only after lymphodepletion, in combination with a dose of IL-2. Lymphodepletion reduces the
presence of negative regulators for inflammation, prolonging survival and enhancing the effector
functions of the adoptively transferred cells[48].
Antigen-presenting cells can be used to generate antigen-specific T cells from a polyclonal population
of PBMCs (figure 12). Patient-derived, antigen-pulsed DCs have been used in this way to generate
effector T cells for the treatment of melanoma[49,50]. However, DCs are not ideally suited for ex vivo
culture due to their short life-time and batch-to-batch variations. Moreover, autologous DCs may be
dysfunctional due to the diseased state of the donor. To overcome this problem, T cells have been
clonally expanded using aAPCs[5,51], an approach which is also illustrated in figure 12. The use of
aAPCs to clonally expand T cells ex vivo allows for a high degree of control over the signals 1,2 and 3
that are presented to the T cells, which makes this technique not only well suited for the production
of antigen-specific T cells for ACT under GMP, but also for more fundamental research on T cell
activation. Chapter 5 covers various artificial antigen-presenting systems in more detail.
Adoptive cell transfer shows great promise in the prevention and treatment of cancer, with
numerous clinical trials performed and many more on the way[52]. However, ACT is far from becoming
a routine clinical practice as its implementation up to now is laborious and costly. In this regard, the
use of ACT in cancer immunotherapy would greatly benefit from the development of a well defined,
cost-effective “off-the-shelf” aAPC platform for the ex vivo expansion and activation of T cells.
18
4. Biomimetic immunomodulation
Advances in material engineering provide an opportunity for tailor made biomimetic signaling
platforms. In this chapter, we look at recent advances in material chemistry and how material
properties such as size and shape affect interaction with biological systems, specifically in the context
of T cell activation. In addition, we discuss the various approaches to ligand presentation on artificial
scaffolds.
4.1 Biomimetic material design
Biomaterials can provide a temporal and spatial context for the delivery of information, something
which is especially important in intercellular communication. The biomimetic properties of materials
have traditionally been limited by available chemistries, but advances in biomaterial synthesis now
allow us to effectively incorporate lessons from nature into material design[53,54]. Differences in
physical properties of natural objects, such as size and shape, are obvious when observed at the
macroscopic scale, but such properties are less well appreciated at the microscopic scale. Figure 13
illustrates the extreme diversity in shape encountered at the microscopic level in nature, ranging
from viruses to pollen and the immune synapse. In bacteria alone, numerous examples can be found
in which shape and size contribute to biological function[55].
Figure 13. Examples of natural biological objects that have diverse physical properties. (A) Human herpesvirus; scale bar,
100 nm (image: Frank Fenner). (B) Ebola virus; scale bar: 500 nm (image: Frederick A. Murphy). (C) Enterobacteria phage λ;
scale bar: 50 nm (image: University of Wisconsin-Madison). (D) Human erythrocytes; scale bar approximately 10 μm. (E)
Escherichia coli; image size approximately 7 μm x 6 μm. (Panels D and E © Dennis Kunkel Microscopy). (F) Surface texture in
alveolar macrophages; scale bar: 5 μm (© 2008 STM). (G) Pollen; image size approximately 50 μm x 45 μm (Dartmouth
Electron Microscope Facility). (H) Intestinal villi; approximate magnification x5,950 (© Dennis Kunkel Microscopy). (I) The
immunological synapse; T cell forming a synapse with a supported membrane containing GPI-linked pMHC and ICAM; ©
1999 AAAS). (J) A schematic of cellular compartmentalization showing several organelles surrounding the nucleus. The
images clearly establish that nature uses physical parameters such as size, shape, texture and compartmentalization in
designing life[53].
Smart biomimetic material design has taken lessons from nature and applied it to for example, drug
delivery, where biodegradable polymers such as poly(lactic-co-glycolic acid) (PLGA) enable the slow
release of bioactive molecules, and particle coating with polyethylene glycol (PEG) prolongs
circulation time and minimizes non-specific interactions. Targeted drug-delivery with PEGylated
nanoparticles is common practice in research and has led to clinically approved drug formulations
already two decades ago[56]. In the context of immunotherapy, it is vital that the biomaterial, which
19
usually serves as a carrier, is immunologically non-reactive, meaning it is not recognized by the
immune system.
As we increase our understanding of biological systems on the one hand, while adding complexity to
synthetic systems on the other, we are slowly able to bridge the gap between the synthetic and the
biological world, as illustrated in figure 14). This hold great promise for immunology, where many
translational challenges like synthetic implants, tissue engineering and tools for (cancer)
immunotherapy can benefit from advances in (bio)engineering.
Figure 14. Bioengineered, bio-inspired and biomimetic systems. The gap between synthetic and biological
systems has traditionally been very large. However, recent advances in the synthesis of novel materials and
understanding of biological systems have paved the way towards bridging this gap. Combining perspectives
from the synthetic and biological fields will provide a new paradigm for the design of immunological tools. PEG,
polyethylene glycol[54].
4.2 Physical and mechanical material properties
Smart material engineering allows us to explore open questions in basic immunological function,
specifically how dendritic cells and T cells interact. Recently, the properties of materials used for
immunomodulation has been extensively reviewed[6,57–59]. Here, we discuss the findings from these
reviews and relate those findings to T cell stimulation and in addition, consider which signals are
appropriate for T cell activation and what would be the effect of avidity and paracrine delivery of
such signals. Ultimately, this information should provide a picture to which material properties are
especially important in T cell activation by artificial antigen-presenting cells.
4.2.1 Size
Size matters. At least when it comes to the way in which micro- and nanoparticles interact with their
surroundings. The profound effect of particle size on its interaction with its surroundings has been
observed at the cellular level in terms of endocytosis and signal transduction, but also in body
biodistribution and in vivo motility.
All cells are capable of engulfing extracellular material through processes collectively known as
endocytosis, shown in figure 15. Endocytosis provides a means for cells to absorb nutrients and
probe their surroundings. There are some size restrictions to the different endocytotic pathways;
while most cells are only capable of ingesting sub-micron sized particles by pinocytosis, only
phagocytotic cells readily take up particles that are over one micron in size[60]. However, endocytosis
of particles larger than 1 µm has been shown for non-phagocytotic cells. Gratton et al. showed
uptake of solid microparticles up to 3 µm in size after incubation with cervical cancer cells (HeLa)[61].
Particle size is well known to influence cellular uptake, which is illustrated by the rate of phagocytosis
20
of polystyrene microparticles by macrophages[62]. This was shown to be size-dependant and
maximum phagocytosis was observed at a particle diameter of 2 to 3 µm, which interestingly
coincides with the typical diameter of bacteria. For nanoparticles in the range of 30-100 nm, with ζpotentials ranging from -23 mV to +9 mV, size is the major determinant for non-specific uptake by
macrophages, where 100 nm particles are taken up fastest[63]. The rate of particle uptake is of
importance in intercellular signaling, since slower uptake usually more time for the particle to reach
its target.
Figure 15. Modes of cellular internalization of nanoparticles and respective size limitations. (A) Internalization of large
particles is facilitated by phagocytosis. (B) Nonspecific internalization of smaller particles (>1 μm) can occur through
macropinocytosis. (C) Smaller nanoparticles can be internalized through several pathways, including caveolarmediated endocytosis, (D) clathrin-mediated endocytosis and (E) clathrin-independent and caveolin-independent
endocytosis, with each being subject to slightly different size constraints. Nanoparticles are represented by blue
circles (> 1 μm), blue stars (about 120 nm), red stars (about 90 nm) and yellow rods (about 60 nm) [56].
At the cellular level, particle size may also influence signal transduction when the particle is coated
with signaling molecules, something that is evident from T cell stimulation with artificial antigenpresenting cells. Microparticles are better at activating T cells than nanoparticles coated with the
same stimulatory ligands[64,65]. This may be explained by the requirement of a large, continuous
surface for the formation of a mature immune synapse, a structure several microns in diameter.
21
The in vivo biodistribution and clearance of particles is largely dictated by the particle size and the
route of administration[66]. In general, small particles (<100 nm) show increased in vivo mobility
compared to larger particles, which is due to the ability of smaller particles to pass through natural
barriers in the body. On the other hand, particles over 5 µm in diameter are likely to cause embolism
after intravenous administration[67], since their circulation is limited by the diameter of the smallest
capillaries (2-3 µm). The passive targeting of nanoparticles is often exploited in cancer therapy,
where it is known that particles up to 100 nm often preferentially leave the vasculature at sites of
inflammation, due to the enhanced permeability and retention (EPR) effect. In healthy tissues,
particles between 100-200 nm are optimally suited for prolonged circulation, since they are large
enough to avoid uptake in the liver, but small enough to avoid filtration in the spleen[56]. However,
non-rigid, red blood cell like particles show prolonged circulation even at sizes up to 5 µm, probably
due to their ability to deform when passing through biological barriers[68].
The lymphatic system is an important size-dependant targeted organ for immunomodulation, the full
dynamics of which is illustrated in figure 16. After intradermal administration, nanoparticles are
known to reach the draining lymph nodes in a size dependant manner through the interstitial fluid,
with an upper size limit of around 100 nm dependant on tissue and injection pressure. In DC targeted
immunotherapy, passive migration of sub-100 nm nanoparticles to draining lymph nodes is often
used to deliver antigen to dendritic cells[39,69–71]. Although particles up to 180 nm can also reach the
draining lymph nodes after subcutaneous administration, this can take up to several days[72].
Recently, Hubbell and coworkers showed that OVA-antigen coupled to a solid core, 30 nm
nanoparticle via a reduction-sensitive linker elicited a CD8+ T cell response whereas OVA-antigen
encapsulated in 125 nm polymeric vesicles (polymersomes) preferentially elicited a CD4+ T cell
response after subcutaneous administration[71].
22
Figure 16. Tissue and cell targeting for antigen delivery in nanoparticulate subunit vaccines. (A) Nanomaterials injected into most
tissues flow into draining lymphatic capillaries or are taken up by tissue-resident APCs. Small particles (red, <100 nm) more easily
penetrate the extracellular matrix and thus distribute more broadly than larger particles (orange), which are retained more in the
tissue. Interstitial APCs can be exclusively targeted by particles that are >500 nm or decorated with targeting antibodies (blue).
Upon entering lymphatic vessels, cells and antigens travel to the draining lymph node. (B) Inhaled or consumed antigens or
nanomaterials mostly target the nasal- or gastrointestinal- associated lymphoid tissue (NALT or GALT, respectively), which include
Peyer’s patches. Because of the tight epithelial barrier, most antigens and particulates are taken up by immune cells, where they
then traffic to the lymph node. (C) Antigens, particulates, and immune cells from the afferent lymphatics encircle the lymphnode
in the subcapsular sinus and enter different areas according mainly to size. Subcapsular macrophages take up large particles (>7
nm) or opsonized antigens (orange), whereas smaller antigens flow into conduits (red). (D) Polymeric nanoparticles (<50 nm, red)
are rapidly cleared after intradermal injection and drain into the lymph node, where they are taken up by APCs surrounding the
lymphatic endothelium (blue) and T cell stroma (green). Scale bar: 200 mm. (E) Multilamellar lipid vesicles (180nm,blue) were
drained into the lymph node, after which germinal centers (green) formed nearby, generating B cells (red) specific to the antigen
borne by the vesicles. Scale bar 200mm. (F) APCs collect polymer nanoparticles (<50nm, red) in the airways (lung parenchyma,
blue) and migrate to the pulmonary lymph node. Scale bar: 10 mm[58].
23
4.2.2 Shape
Although it has long been known that viral and bacterial shape influences in vivo behavior, the
systematic study of shape-effects on the behavior of micro- and nanoparticles has only recently
gained attention with the advent of techniques that allow for the synthesis of non-spherical
particles[73,74].
The shape of micro- and nanoparticles can have a profound effect on uptake by phagocytotic[75–78]
and non-phagocytotic cells[61,79]. As a rule of thumb, particles with a higher aspect ratio show slower
uptake, independent on particle size[78] or cell type[79]. As a high aspect ratio particle is more likely to
approach a cell with one of its low curvature sides, extensive actin remodeling, an energy dependent
process, is required to internalize the particle. This can help to explain the observations made by
Sharma et al., who found that ellipsoid polystyrene microparticles attached more readily to
macrophage surfaces than their spherical counterparts, but are taken up less efficiently[75]. The sizeindependent, aspect ratio-dependent uptake of micro- and nanoparticles by macrophages as found
by Champion et al. is illustrated in figure 17.
Particle shape also significantly influences the circulation time in vivo, an effect which can be
observed from the disk shaped red blood cell (RBC), which can stay in circulation up to 120 days.
Although the prolonged circulation time observed with RBCs is not only due to its shape, prolonged
circulation of synthetic microparticles that mimic the dimensions close to those of RBCs has been
observed[80]. Geng et al. found that elongated micelles (filomicelles) show circulation times of up to a
week, which was significantly longer than their spherical counterparts[81]. A tempting explanation for
the prolonged circulation time of disks and filomicelles is their ability to align with the blood stream,
which would minimize contact with the cell wall and phagocytotic cells present in the blood[81].
Figure 17. Scanning electron micrographs (A-C) of
particles during phagocytosis and (D) a phagocytosis
phase diagram. Top: Micrographs (A–C) of cells and
particles were colored brown and purple,
respectively. (A) The cell body can be seen at the end
of an opsonized elliptical disk, and the membrane has
progressed down the length of the particle (Scale bar:
10 µm). (B) A cell has attached to the flat side of an
opsonized elliptical disk and has spread on the
particle. (Scale bar: 5 µm). (C) An opsonized spherical
particle has attached to the top of a cell, and the
membrane has progressed over approximately half
the particle. (Scale bar: 5 µm). (D) Phagocytosis phase
diagram with Ω and dimensionless particle volume V*
(particle volume divided by 7.5 µm radius spherical
cell volume) as governing parameters (n = 5 for each
point). There are three regions. Cells attaching to
particles at areas of high Ω, >45°, spread but do not
initiate internalization (region C). Cells attaching to
particles at areas of low Ω, <45°, initiate
internalization (regions A and B). If V* is ≤1,
internalization is completed (region A). If V* > 1,
internalization is not completed because of the size of
the particle (region B).The arrows above the plot
indicate the point of attachment for each shape that
corresponds to the value of Ω on the x axis. Each case
was classified as phagocytosis or no phagocytosis if
>95% of observations were consistent. Each data
point represents a different shape, size, or aspect
ratio particle. Adapted from[78].
24
Particle shape may also influence signal transduction, since some signaling events, like DC:T cell
interactions, exhibit clear shape dependency. In a comparative study, spherical, rod- and disk-shaped
polystyrene micro- and nanoparticles were coated with the breast cancer cell recognizing antibody
trastuzumab and assessed for specific and nonspecific uptake. Rods exhibited higher specific uptake
and lower nonspecific uptake in all cells compared with spheres, implying that there is an increased
binding to rod-shaped particles compared to spherical particles[82]. This could be of interest to microor nanoparticles mediated cellular signaling or delivery of cytotoxic drugs to cancerous cells.
4.2.3 Rigidity
One mechanical property that is often overlooked in biomimetic material design is the rigidity of the
material, although this property may affect cellular interaction in a number of ways. In an attempt to
mimic red blood cells, Doshi et al deposited layers of protein on a preformed, RBC shaped PLGA
template. After cross linking of the protein layer, the PLGA template was dissolved, leaving a hollow,
flexible synthetic RBC (sRBC) with an elastic modulus close to that of mouse RBCs. By virtue of their
flexibility, these sRBCs are able to pass through microchannels with a diameter below that of their
resting diameter[83]. In a study with soft and rigid polyacrylamide microparticles, macrophages
showed a strong preference to engulf the rigid particles[84]. Using the Particle Replication in Nonwetting Templates (PRiNT) technology, Merkel et al. were able to produce sRBC with a finely tuned
elastic modulus ranging from 8-64 kPa, depending on the degree of cross-linking. After in vivo
administration, the particles showed differences in circulation time dependent on their elastic
modulus, with an eight-fold decrease in circulation time correlating to a 30-fold increase in
circulation time[68].
With regard to ligand recognition, substrate adhesion by ligand-coated polyethylene oxide-bpolybutadiene polymersomes was found to be lower for flaccid polymersomes compared to their
more rigid counterparts[85], and in the context of T cell activation, stiff polyacrylamide gels
incorporating anti-CD3 and anti-CD28 were more effective at T cell activation than soft gels,
measured by IL-2 secretion[86].
4.2.4 Zeta potential
Most mammalian cells exhibit a negative overall charge on their surface due to the presence of sialic
acids in the carbohydrate coating of the cellular membrane. As a consequence, the ζ-potential, a
measure for the potential difference between a (particle) surface and its surroundings, can be used
to predict the behavior of a particle when it encounters a cell. Positively charged particles show
stronger non-specific interaction with cells[61,63,87] and can penetrate the cell membrane, a property
exploited in drug delivery.
These findings have an impact on biomimetic particle design; for prolonged circulation and minimal
non-specific interaction, a particle must preferably have a ζ-potential that is neutral or slightly
negative. As reactive functional groups such as thiols, amines and carboxylic acids strongly influence
surface charge, care must be taken to control the display of these functional groups on the particle
surface. Specifically for particles functionalization, it could be useful to select non-charged functional
groups ligand attachment[87]. Soluble proteins found in serum will affect the net charge of by forming
a protein corona around the particle, so for any biological application the ζ-potential must be
determined only after incubation with serum[63].
25
4.3 Ligand presentation
In the classical approach to ligand-receptor interactions, both the ligand and receptor are considered
to be free in solution, whereas in cell to cell communication usually at least either the ligand or the
receptor is bound to a cellular membrane. Intercellular signaling events involve multiple ligandreceptor interactions, examples of which include immunological and neuronal synapses[88]. Here, we
will discuss both surface bound and soluble ligand presentation, specifically for T cell activation and
prolonged circulation.
4.3.1 Presentation of surface bound ligands
In intercellular communication processes, such as immune recognition or cellular adhesion, multiple
surface bound ligands are often presented to enhance signal strength. The combined strength of
multiple binding interactions is known as avidity, not to be mistaken with the affinity associated with
a single ligand-receptor interaction. An excellent example of the power of multivalent binding is
found in the significant (100-fold) increase in affinity of TCR:pMHC binding in situ vs TCR:pMHC
binding in solution[89]. In fact, the interaction between a single TCR and peptide-MHC complex is very
weak, with a dissociation constant (Kd) in the order of 1-100 µM resulting in the requirement of
multiple pMHC complexes to stimulate T cells[24]. The power of multivalent binding has led to the
development of scaffolds that can display multiple ligands[90]. Such scaffolds can be anything from
proteins to dendrimeric polymers or solid particles.
Figure 18. Three ligands (recognition (red), costimulatory (green), and adhesion (blue)) that could be presented on the
surface of artificial antigen presenting cells (aAPCs) in various ways. (A) Isotropic surface presentation of randomly
distributed ligands. All three ligands are presented uniformly over the particle surface. (B) Anisotropic presentation of ligands
in a patch pattern on the surface of a particle. Recognition and costimulatory ligands are randomly distributed in the patch,
with a surrounding field of adhesion ligands. (C) Dynamic anisotropic presentation of ligands on a fluid supported lipid bilayer
(SLB; yellow). Before interactions with a cell (e.g., T cell), lower initial surface densities of randomly distributed ligands may
be placed on the SLB surface. Anisotropic reorganization of ligands occurs in response to interactions with a cell. The resulting
bull’s eye pattern would be characteristic of a natural supramolecular activation clusters (SMAC) formed at immune synapse
between a T cell and APC[6].
Ligands can be bound to a scaffold in a random fixed (isotropic),site selective fixed (anisotropic) or
dynamic fashion, illustrated in figure 18. The ability of ligands to move in response to a binding event
can be crucial for signal transduction, and this dynamic remodeling of surface bound ligands is
increasingly recognized as an important aspect of cellular communication[91]. Although working with
solid micro- and nanoparticles is more appealing than liposomes from an engineering point of view
due to the poor stability of liposomes[92], solid particles only allow for the attachment of fixed
surface bound ligands. An elegant solution for this problem is to coat solid particles with a lipid
26
bilayer, effectively mimicking the fluid nature of natural
cell membranes. Ashley et al. coated silica nanoparticles
with a lipid membrane that was targeted to
hepatocellular carcinoma[93]. A combination of targeting
and fusogenic peptides bound to lipids in the bilayer
ensured rapid uptake and release of the drug-loaded
silica core inside the targeted cells. The influence of the
fluidity of the spherical supported lipid bilayer (SLB) on
binding to Hep3B cells was investigated using either a
Figure 19. Recruitment of Alexa Fluor 647-labelled
fluid DOPC or an non-fluid DPPC lipid for the construction SP94 peptides (white) to the surface of a Hep3B cell
of the SLB. At similar targeting peptide concentrations when peptides are displayed on a
(~6 per particle), a 100-fold increase in binding affinity nitrobenzoxadiazole-labelled SLB (green) composed
of DOPC (open circles) or DPPC (closed circles).
was observed when the fluid DOPC was used to construct These data were collected at 4°C. Hep3B cells were
the SLB. This dramatic increase in binding was attributed labeled with CellTracker Red CMTPX (red) and
Hoechst 33342 (blue). Inset scale bars=5 µm[93].
to the ability of ligands in the fluid SLB to reorganize, as
shown in figure 19. Similarly, lipid coated silica microparticles that were designed to capture viral
particles through multiple ligand-receptor interactions showed a reduction in captured viral particles
at a lower temperature due to reduced lateral mobility of the ligands in the SLB[94]. The ability of
ligands to dynamically reorganize in response to a binding event has been studied for leukocyte
mimics, or leuko-polymersomes (bilayer vesicles constructed of amphiphilic block copolymers)[95].
Surface bound ligands can be attached to particles via different synthetic strategies and an optimal
ligation strategy most be chosen for each type of ligand-particle pair. A discussion of these strategies
is outside the scope of this thesis, but the reader is referred to a recent review on the topic[96].
4.3.2 Presentation of soluble ligands
Cells can communicate with their environment via
paracrine signaling, a process that describes the release
of soluble factors, such as cytokines and chemokines, by
cells into their surrounding in order to elicit a response
from neighboring cells. Cells use paracrine signaling in
processes such as tissue regeneration, neuronal signaling
and immunity. Paracrine signaling allows for a local, high
concentration of signal, without the need for direct
contact, as illustrated in figure 20[97]. Moreover, paracrine
release allows cells to create a signal gradient, which
other cells can home to or move away from. The
biomimetic release of soluble protein-based ligands in a
paracrine fashion was first shown in 1976[98] and has
evolved significantly with the design of novel
biodegradable polymers such as poly(lactic-co-glycolic
acid (PGLA)[99,100] and other systems, recently reviewed by
Rothstein and Little[101]. Here, we discuss how the
presentation of soluble ligands and in particular paracrine
signaling can steer biomimetic immunomodulation.
Figure 20. Cytokine release from a paracrine
signaling antigen-presenting cell (paAPC). IL-2
release from paAPCs in isolation and in the vicinity
of T cells was modeled using a diffusion equation
to illustrate the level of IL-2 accumulation that T
cells experience during initial interaction with
paAPC. Left , a T cell paAPC interaction at 3 µm
apart and right, 20 nm apart[97].
27
Collectively, signaling molecules released by cells are called cytokines. Some cytokines, specifically
those released by APCs, strongly influence T cell maturation and differentiation. During T cell
development, dendritic cells secrete cytokines to promote T cell growth and differentiation towards
a specific lineage[102]. It should come as no surprise then, that these cytokines have been tested for
therapeutic applications, either by systemic administration or as adjuvant in vaccines[103]. The toxicity
of high-dose IL-2 therapy, which was previously discussed, has promoted research into the
biomimetic release of IL-2 and other soluble factors for therapeutic applications to enable high local
concentrations of the cytokine while avoiding toxic side effects[104–109].
In one recent example, Fahmy and coworkers described the combined paracrine release of a TGF-β
inhibitor and IL-2 from nanosized liposomal polymeric gels (nLGs)[108]. The TGF-β inhibitor serves to
counteract the immunosuppressive function of TGF-β in the tumor microenvironment, allowing IL-2
to optimally stimulate cytotoxic natural killer and CD8+ T cells. Evaluation of these nLGs in a
melanoma mouse model revealed a delayed tumor growth, increased survival of tumor-bearing
mice, and increased the activity of natural killer cells and of intratumoral-activated CD8+ T cell
infiltration[108]. Paracrine release of IL-2 has also been explored for the non-specific stimulation of
naïve T cells. Naïve T cells were cultured in the presence of a non-specific activation signal provided
by aCD3/aCD28 coated beads and either soluble or paracrine delivered IL-2. The slow, sustained
release of IL-2 from microparticles significantly increased CD8+ T cell proliferation in comparison with
exogenously added IL-2[104]. This demonstrates that a local cytokine microenvironment created by
paracrine release can provide additional information for T cell development.
In addition to cytokines, some cells also secrete chemokines to recruit nearby cells to for instance a
site of inflammation. Local release of chemokines can be used to manipulate cell trafficking in vivo
and has been explored for therapeutic applications, including recruitment of TILs[110], DCs[111] and
regulatory T cells[112]. Recruitment of cells could be interesting for vaccination strategies, in which
one could image a local depot of antigen being injected together with a chemokines as attractant for
Figure 21. Migration paths of T cells chemotaxing toward CCL21 chemokine-releasing microspheres (CRMs). Resting or
activated human T cells were embedded in collagen gels and imaged by videomicroscopy for 1.5 h. In parallel, control
samples of T cells and empty CRMs in collagen mixed with 10 μg/mL “free” CCL21 were imaged. Shown are single-cell paths
for cells whose starting positions (ro) were greater than or less than 100 μm from the nearest bead, color-coded by cell
motility[114].
28
dendritic cells and naïve T cells to form a synthetic lymph node[113]. Leukocyte migration in alginate
gels was studied using microbeads that secreted CCL19 or CCL21, showing either leukocyte
colocalization with or “hopping” from bead to bead, dependant on the chemokine concentration and
release rate[114]. Figure 21 shows the migration of activated and resting T cells towards CCL21
releasing microspheres.
4.3.3 Surface bound ligands that prolong circulation time
After administration, micro- and nanoparticles are often cleared from the bloodstream by
phagocytes before they reach their desired target in the body. This has motivated researchers to find
ways of prolonging particle circulation time by preventing phagocytosis. A significant breakthrough
came in 1994 when it was discovered that coating PLGA particles with the synthetic polymer
polyethylene glycol, or PEG, could significantly reduce the number particles captured by the liver
after injection[115]. Ever since, PEG coatings have been used to enhance circulation time and in vivo
stability of particles, proteins and therapeutic drugs[116]. The effect of prolonged circulation due to
PEGylation can be explained by the observation that the PEG corona reduces the adsorption of
opsonins, a natural marker protein for macrophages. The PEG polymer length influences circulation
time to a similar extent as particle size, with a long (10 kDa) PEG chain and small particle size (20nm)
resulting in a half life greater than 48 hours[117].
Synthetic particles are still no match to red blood cells (RBCs) when it comes to circulation time. Once
RBCs enter the blood stream, they may circulate up to 120 days before they are engulfed by
macrophages. The unique features of RBCs have been translated to nanoparticles via a top-down
approach, in which the membranes of RBCs were used to coat PLGA nanoparticles[118]. These RBCcoated nanoparticles showed a marked increase in circulation time compared to bare nanoparticles
and even outlasted PEGylated nanoparticles. A key protein in preventing engulfment of RBCs by
phagocytes has been identified as CD47, which acts as a “marker of self”[119]. Discher and coworkers
showed that CD47 could prevent phagocytosis of opsonized microbeads, but only at a low surface
density[120]. Recently, a minimal CD47 derived “self” peptide was designed and tested for inhibition of
phagocytosis, which showed that this peptide was equally effective as the whole length protein[121].
For biomimetic intercellular signaling, the identification of surface ligands that prevent or slow down
clearance of particles from the bloodstream is of interest because biomimetic intercellular signaling
can only occur with particles that are not engulfed.
29
5. Artificial antigen-presenting cells
Immunotherapy through DC vaccination or adoptive cell transfer, holds a great therapeutic potential
for diseases that can be controlled using the body’s own immune system, such as cancer or autoimmune diseases. Cell-based approaches to elicit the desired cellular and humoral immune
responses are often hampered to by the complex and unpredictable nature of the immune system.
As touched upon previously, a promising alternative to cellular-based immunotherapy has been the
development of acellular, artificial antigen-presenting cells (aAPCs), of which the exact composition
can be tightly controlled. Artificial APCs have found their way in vaccination and ACT. In this chapter,
we will discuss the different acellular aAPC scaffolds reported in the literature thus far to see what
has driven scaffold design for aAPCs.
5.1 Basic components of an aAPC
Antigen presentation by natural APCs can lead to variety of T cell responses, depending on which
signals are transmitted. Therefore, control over the signals incorporated into artificial antigenpresenting cells improves control over the therapeutic outcome[122]. The information that is
transmitted by an APC to activate, expand and differentiate a naïve T cell is classically divided into
three signals, as illustrated in figure 22 and discussed in chapter 2.
Figure 22. Schematic of signal classes presented by an antigen–presenting cell (APC). Three signals are essential for
optimal T cell stimulation: 1. Recognition signals that ligate the T cell antigen receptor, pMHC complexes or
antibodies cross-linking the T cell receptor (TCR), 2. Costimulatory molecules of the CD80/86 or TNF family and
adhesive molecules that strengthen interactions between cells, 3. Cytokines secreted by APC or other immune cells
that bind to receptors on the T cell surface[5].
The first signal, recognition, takes place when a T cell receptor (TCR) on a T cell recognizes a peptideMHC (pMHC) complex on an APC surface. For artificial antigen presentation, either an pMHC class I
(for expanding CD4+ T cells) or pMHC II (for expanding CD8+ T cells) can be used as recognition signal.
Often, an MHC I/II non-specific antibody, anti-CD3, is used as an alternative recognition signal on
surface of aAPCs. The use of an antibody as opposed to the more biosimilar pMHC I/II should come
30
as no surprise; antibodies are more easily produced in large quantities and only one aAPC is required
for the activation and expansion of a diverse repertoire of T cells.
Costimulation through the interaction of CD80/86 receptors on the APC and CD28 on T cells is known
to enhance the strength of the antigen-specific T cell response. Many aAPCs therefore present either
CD80/86 or anti-CD28 on their surface as a second signal, although stimulation with aCD28 may only
lead to T cell proliferation but not differentiation[123]. In addition to costimulatory ligands, adhesive
interactions though ICAM-1 on the APC surface with LFA-1 on T cells may serve to enhance affinity
and prolong APC:T cell interaction. As such, anti-LFA-1 has been used in artificial APC systems[124].
Lastly, an important factor in the rapid expansion and differentiation of T cells comes from cytokines,
which are either released by the APC or by neighboring activated T cells. Cytokines are extensively
used for the ex vivo culture of T cells (in adoptive transfer for example) or for direct in vivo
administration as a form of immunotherapy. Cytokine release has only recently been mimicked in
local delivery strategies, mainly in biodegradable PLGA particles[104] or anchored to liposomes
through an Fc-fragment[125].
It is perhaps the modular, systematic description of T cell stimulation that has attracted (bio)chemical
engineers to the field of artificial antigen presentation[5,6,59]. The type of T cell response can be
precisely tuned, depending on the signals provided by the aAPC. Controlled display of information to
T cells does not only help to increase our fundamental understanding of the nature of T cell
activation, but can eventually also lead to well defined immunotherapies.
5.2 Scaffolds used for artificial antigen presentation
Ever since they
were introduced by
Matthew Mescher
in 1978, a wide
variety of scaffolds
have been reported
for the construction
of aAPCs[126]. Several
reviews on the use
of aAPCs in T cell
activation can be
found[5,127,128],
of
which the review by
Steenblock et al. is
the
most Figure 23. Schematic representations of four types of acellular aAPC. Particles can be coupled with
comprehensive and recognition, costimulatory or adhesion molecules by different binding schemes. Examples of
configurations are shown here[5].
lists a table of the
different scaffold that are used for the construction of aAPCs. An updated overview of aAPC scaffolds
can be found in table 1 at the end of this chapter and the most common scaffold classes are shown in
figure 23.
31
5.2.1 Liposomes
Liposomes are spherical vesicles which can be prepared from amphiphilic lipid molecules such as
phosphatidylcholine or cholesterol, both present in natural membranes. Their biocompatibility and
ease of preparation has led many researchers to study the use of liposomes as delivery vehicles for
therapeutic and diagnostic purposes early on[129,130]. Liposomes self-assemble under aqueous
conditions and because of the low molecular weight of most lipids, their membrane remains fluid
which allows for mechanical resizing to create vesicles from 60nm to 1 µm.
Mescher and coworkers studied the use of liposomes as potential artificial antigen-presenting cells as
early as 1978[126]. These crude aAPCs, consisting of purified HLA-A and HLA-B antigens reconstituted
in phospholipid liposomes, were one of the first systems used to study T cell activation on a
molecular level and were used to show that T cell activation is dependent on the antigen density on
the liposome surface.
More than two decades later, pMHC II decorated liposomal
nanoparticles were found to be able to bind MHC II restricted T
cells and induce interfacial pMHC:TCR clustering, demonstrating
the potential of signal clustering on the aAPC[124]. Signal
clustering was further explored with the use of neutravidin
"rafts" comprised of pMHC complexes, anti-CD28 and anti-LFA-1
on liposomes (see figure 24)[131,132]. These preclustered,
liposomal aAPCs were compared with commercially available
Dynabeads® (discussed in section 5.2.3) for their ability to
expand T cells. The results from this experiment showed that
these aAPCs were similarly effective as the commercial
microbeads, but preferentially expanded CD8+ over CD4+ T
cells[132].
Recently, the pioneering work by Mescher was revisited when
Ding et al. isolated lipid rafts containing pMHC complexes from
DCs and reconstituted them in 200-300 nm-sized liposomes[133].
These so-called RAFTsomes effectively expand T cells in vitro
and elicit both cellular and humoral immune responses in mice.
Figure 24. Schematic of an artificial APC
depicting a section of the lipid bilayer.
GM-1 ganglioside incorporated into the
bilayer binds one molecule of choleratoxin β (CTB) biotin. In turn, one molecule
of neutravidin anchors the biotinylated
anti-CD28, the pMHC complex and biotinanti-LFA-1 to the aAPC through the CTBbiotin molecule[170].
Liposomes share interesting properties with natural APCs, such
as membrane fluidity and biocompatibility. Unfortunately, the
trade-off for the fluid nature of liposomal membranes is their
poor stability, which greatly hampers their use as an aAPC[134].
To overcome this problem, other, more stable scaffolds have
been developed.
5.2.2 Supported lipid bilayers
Supported lipid bilayers (SLB) on a planar surface are often used to study the molecular details of the
immune synapse, as discussed in chapter 2. However, supported lipid bilayers can also be
constructed on spherical surfaces, in which lipids are coated around a solid bead[135,136]. Such
spherical SLBs allow for a lateral movement of surface bound ligands, while their stability is increased
compared to liposomes[93]. Cell-sized, spherical SLBs were prepared by coating solid microparticles
32
with a lipid layer that anchored pMHC I. These SLB-based aAPCs were not able to expand CTLs from
naïve T cells, but they were capable of restimulating previously primed CTLs[137]. Micrometer sized,
artificial APCs prepared by coating silica microparticles with melanoma cell-lysates showed no
significant responses in stage IV melanoma patients, possibly due to lack of exogenous activating
cytokines[138].
5.2.3 Polystyrene beads
Uniform, solid polystyrene beads can easily be prepared in different sizes from the polymerization of
styrene via controlled emulsion polymerization. Although such polystyrene (PS) beads lack the fluid
outer membrane and aqueous inner core of liposomes, they are easily handled and are readily
surface modified through non-covalent interactions with the particle surface. These beads have been
used as aAPCs to study the effects of, amongst other things, particle size and ligand density on T cell
stimulation[64,139–147].
For example, it was shown that in the range of 1-5 µm, 4-5 µm PS particles coated with pMHC I were
optimally suited for CTL activation. Below 4 µm, activation of CTLs decreased rapidly, an effect which
could not be compensated by the addition of more particles[64]. These findings suggest that a large,
continuous surface contact between an aAPC and a T cell is required for optimal signal transduction.
Additionally, it was found that stimulation by surface bound CD86 alone did not result in activation of
T cells, but that this required the presence of additional anti-TCR antibodies, and that the
effectiveness of CD86 costimulation was dependant on its surface density[139]. Although signal 1 and 2
on an aAPC surface is enough to activate T cells, exogenous addition of cytokines is required to
ensure proliferation and effector differentiation after initial activation[35,148].
In a move towards the clinical use of aAPCs, PS beads coated with (HLA)-A2/pIL-13Rα2345-354
tetramers, anti-CD28 and CD83 were used to expand antigen specific CTLs from peripheral blood
derived mononuclear cells (PMBCs) obtained from healthy donors. The ex vivo expanded CTLs
showed lytic activity HLA-A2+ glioma cells that expressed IL-13Rα2, whereas no lytic activity was
observed for HLA-A2- that did not express IL-13Rα2[143]. More recently, PS-based aAPCs were used to
generate a melanoma-specific CTLs response in vivo. Artificial APCs consisting of H-2K(b)/TRP2
tetramers, anti-CD28 and anti-4-1BB coupled to 5 µm PS beads, were injected intravenously and
subcutaneously into naïve or antigen-primed mice which resulted in a vigorous CTL response[147].
Interestingly, no harmful side-effects such as embolism was reported in these studies.
A special class of PS-bead based aAPCs incorporates an iron oxide core within a PS coating and an
outer layer of reactive tosyl-, or epoxide groups. These beads, commercially sold under the
Dynabead® brand name, allow for the attachment of a variety of ligands, and the iron oxide core
enables easy magnetic removal of the aAPCs from the culture medium, making this system optimally
suited for ACT based therapies. These beads are often used to non-specifically expand T cells from a
population of PBMCs, although they can be used for a wide range of applications[149].
Magnetic aAPCs have been equipped with aCD3[123,150–153], MHC tetramers[154] or HLA-Ig dimers[155–161]
in combination with surface bound or soluble costimulatory molecules in the context of T cell
stimulation. Genetically produced HLA-Ig constructs confer mechanical stability to the HLA:TCR
interaction, which increases T cell binding affinity[155,156]. HLA-Ig coated magnetic aAPCs, in
combination with exogenously delivered IL-2, effectively expanded MART-1 specific CTLs from a
population of CD8+ PBMCs, generating up to 109 cells within two months[155].
33
Although both the magnetic and the regular PS beads are versatile and easy to handle platforms for T
cell activation, their material properties limit the extent to which biomimetic T cell stimulation can be
achieved. Both paracrine release of cytokines and dynamic redistribution of ligands is not possible
with this system. Moreover, the method of attachment to these particles can interfere with the
proper orientation of the displayed ligands, thereby hampering signal transduction to T cells[162].
5.2.4 Biodegradable particles
When transient stimulation by aAPCs is required, for example in active vaccination strategies or
other in vivo applications, biodegradable polymeric scaffolds may be superior over nonbiodegradable scaffolds. In addition, biodegradable scaffolds are capable of encapsulating cytokines,
which can be released in a paracrine manner, allowing for a local, sustained cytokine-rich
environment.
A team led by Terek Fahmy has focused on the production of such biodegradable aAPCs, based on
PLGA polymers[65,104,163]. This copolymer can be made into spherical particles with well defined sizes
and the rate of degradation can be finely tuned by adjusting the PLA/PGA ratio. To functionalize
these polymers, palmitate-conjugated streptavidin is mixed in with the polymer during the
fabrication process, which ends up on the outside of the particle after production. This streptavidin
moiety can then be modified to display any biotinylated biomolecule of interest.
In a comparative study, either micro- or nano-sized PLGA particles were modified to display pMHC I
dimers and anti-CD28 and consequently used to stimulate CD8+ T cells[65]. Micron-sized aAPCs
outperformed both nano-sized aAPCs and soluble pMHC I dimers, as measured by cytokine secretion.
In the same study, it was shown that costimulation through CD28 significantly enhances T cell
proliferation and cytokine secretion and that the overall T cells count is increased when aAPCs
display pMHC I dimers instead of aCD3 as a recognition signal. Finally, enhanced expansion of T cells
was observed when IL-2 was encapsulated in the polymer matrix. These biodegradable aAPC were
shown to outperform autologous APCs for the expansion of antigen-specific T cells in vitro[164]. A
possible limitation to these biodegradable systems is their intrinsic instability, which could limit the
time that ligands stay at the aAPC surface.
5.2.5 Other scaffolds
Next to the scaffolds mentioned previously, there numerous additional materials that could serve as
a scaffold for the multimeric display of T cell activating ligands. Some interesting materials that have
been tested as scaffolds for artificial antigen presentation include carbon nanotubes[165],
polyisocyanides[166] and mannose nanofibers[167]. The development of novel scaffolds is of interest to
the field of artificial antigen presentation, since each material gives us more insight into the chemical
and mechanical requirements for optimal T cell activation.
34
Table 1. Artificial antigen-presenting platforms reported in the literature (adapted from[5])
Platform
Liposomes
Supported lipid
bilayers
Polystyrene
beads
Size
(common)
60-1000
nm (100
nm)
5 µm
Signal 1
Signal 2
HLA-A/HLA-B, pMHC Anti-CD28, antiII, HLA-DR4, antiLFA-1
CD3
pMHC I
-
IL-2/ IL-15
1-10 µm (5
µm)
Tumor cell
membrane, pMHC I,
anti-CD3, HLA-A, H2Kb-Ig dimers, MHC
II dimers, HLA-A2
tetramers
Anti-CD3, HLA
tetramers, HLA-A2Ig dimers
Tumor cell
membrane, antiCD28, anti-CD2,
4-1BB ligand,
anti-4-1BB,
CD83
Anti-CD28, anti4-1BB, anti-FasIgM, CD80-Ig,
CD86-Ig
Anti-CD28
Magnetic
polystyrene
beads
4.5 µm
PLGA beads
130 nm-8
µm (7 µm)
Carbon
nanotubes
500 nm
(bundles)
Polyisocyanides
2 x 120 nm
Anti-CD3
-
Self-asembled
fibers
40nm x
~3µm
-
-
Anti-CD3, MHC-KB
dimers, HLA-A2-Ig
dimers
pMHC I
Signal 3
Ligand
attachment
Hydrophobic,
Avidin-Biotin
In vitro stimulation
Murine splenocytes,
Human PBMCs
In vivo
Ref.
use
[124,126,131,132,168]
Active
(humans)
Con A
Hydrophobic
Murine splenocytes
-
[137]
IL-2, IL-4,
IL-7, IL-12
Adsorption,
avidin-biotin,
covalent non
selective
Murine splenocytes,
Human PBMCs
[64,141–146]
IL-2, IL-7,
TCGF
Covalent, nonselective
Murine splenocytes,
Human PBMCs
Active
(mice),
adoptive
(mice
and
humans)
Adoptive
(mice
and
humans)
Active
(mice)
IL-2
Avidin-biotin
(paracrine)
Murine splenocytes,
Human PBMCs
[123,150–160]
[65,104,163,164]
Avidin-biotin
Ag specific murine
CD8+ lymphocytes
-
[165]
-
Avidin-biotin
Human PBMCs
-
[166]
Con A
Electrostatic
Jurkat cells
-
[167]
Anti-CD28
(soluble)
Nb. Signals 1 and 2 are surface bound unless state otherwise. Signal 3 is exogenously added unless stated otherwise.
6. Conclusions
Work with dendritic cell-based vaccines has proven that clinically relevant results can be obtained by
harnessing the power of the immune system to fight cancer. However, such APC-based therapies are
often laborious and costly, limiting their widespread application. This, in addition to the
unpredictable outcome of DC manipulation, has led to the development of modular, tunable systems
for T cell activation known as artificial antigen-presenting cells.
As the name suggests, artificial antigen-presenting cells should mimic antigen-presentation by
professional antigen-presenting cells. The activation of naïve T cells by DCs (a professional APC) has
been studied extensively, from the molecular to the systemic level. Naïve T cells are primarily
activated in the lymph nodes, where they scan their surroundings at high speeds until they
encounter their cognate pMHC complex on the surface of a DC. Only a few (<10) pMHC complexes
are required for the T cell to bind, but once the T cell is bound, it can stay bound for hours. Initial
contact leads to the formation of an immunological synapse (IS) between the DC and the T cell. The
macromolecular structure of the IS influences T cell activation, through the formation of multiple,
spatially separated ligand-receptor interactions. Although the precise role of the overall IS structure
on T cell activation is not known, it seems to be essential that TCRs are preclustered, preferably
together with costimulatory receptors such as CD28, on the T cell surface. In addition, stimulatory
cytokines, such as IL-2, need to be present to assure proliferation of the activated T cells.
The stimulatory capabilities of an APC have successfully been mimicked by artificial antigenpresenting platforms, varying from liposomal to solid polymeric constructs. Especially in vitro, such
systems have proven their value as modular platforms for the stimulation of T cells. However, some
essential characteristics of natural T cell stimulation, such as dynamic remodeling of surface bound
ligands or the paracrine delivery of cytokines, remain hard to mimic. Another point of concern
related to the possible therapeutic use of aAPCs is related to the large size (5-10 µm) that seems to
be required for optimal T cell stimulation. Although ex vivo cultured T cells put no restriction on aAPC
size, the intravenous delivery of particles larger than a few micrometers may lead to embolism. For
this reason, it may be of interest to develop nano-sized aAPCs for the in vivo stimulation of T cells[163].
The choice of the right material for the construction of an aAPC is definitely not trivial, as basic
scaffold material properties partially determine how signals are presented by the aAPC. An ideal
aAPC is structurally stable and allows for a well defined ligand decoration, while preferably being
biocompatible/biodegradable as well. Since the scaffolds reported thus far have primarily been
selected based on availability of the scaffold, development of novel scaffolds could open up the
possibility to systematically study unexplored biomimetic factors such as shape or surface fluidity.
For the development of new artificial antigen-presenting cells, the ultimate application of such aAPCs
should be kept in mind. When used as a model system for the stimulation of T cells in vitro, the
choice of scaffold is almost unrestricted. However, when such aAPCs are expected to be used in
active vaccination strategies, the scaffold must be biocompatible and small enough to prevent
embolism.
36
Artificial antigen-presenting cells can provide a powerful tool for unraveling the molecular details of T
cell activation and show potential use for cancer immunotherapy. The artificial antigen-presenting
platforms presented in the literature up to date show that it is possible to stimulate T cells in a
biomimetic way using simplified constructs. Although not all requirements for T cell activation, and
particularly differentiation, are known, these constructs provide a solid basis from which new
artificial antigen-presenting cells can be constructed.
37
II. Summary
The immune system is perfectly adapted to protect against disease by constantly monitoring the
body for anything which is not the body’s own. Any unfamiliar agent, either a virus or a misfolded
protein, may lead to an immune response. Key players in mediating the immune response are
dendritic cells (DCs), a professional type of antigen-presenting cell (APC). DCs capture, process and
present compounds that are derived from undesired agents (antigens) via MHC/antigen complexes
on their cell surface to activate T cells, which in turn track down the antigen source and clear it from
the body.
Through their function as antigen-presenting cells, DCs are vital mediators in the development of
cancer immunity, and it is this finding that has led researchers to seek ways of manipulating the
immune system in the context of cancer immunotherapy. Broadly, two approaches to cancer
immunotherapy exist; direct DC manipulation in order to promote DC-mediated antigen presentation
or adoptive transfer of ex vivo cultured, DC-activated tumoricidal T cells. Both approaches rely on
antigen-presentation by natural DCs, which is still problematic due to the lack of knowledge on DC:T
cell communication and the difficulty of culturing DCs. For a better understanding of T cell activation,
artificial antigen-presenting cells (aAPCs) have been developed. These simplified DC mimics usually
incorporate the basic elements required for T cell activation on a material scaffold and have proven
useful for the production of large amounts of activated T cells.
APCs deliver three kind of signals to T cells; activation via the recognition of an MHC complex by a T
cell receptor (TCR), survival though additional interactions with receptors on the T cell surface, and
differentiation via soluble cytokines released by the APC. During T cell activation, a large interface is
formed between the APC and the T cell, known as an immunological synapse (IS). From studies
where the APC is mimicked by a supported lipid bilayer (SLB), we know that only a few MHC:TCR
interactions are enough for the formation of an IS and that the structural orientation of the
activation and survival signals strongly influence the outcome of overall T cell activation.
Although there is little variation in the ligands that are incorporated on the surface of aAPCs, various
completely different material scaffolds have been tested over the years, with different success rates.
Liposomes were classically used, due to their ease of production, but were abandoned because of
stability issues. A commonly used scaffold nowadays is the solid polystyrene bead, which has proven
optimal for T cell activation at sizes comparable to cells. For potential in vivo use, biodegradable
polymeric aAPCs have been developed, which in addition to activation and survival signals, also
release cytokines during their degradation.
The physical properties of the particles that are used for the production of aAPCs are of importance
to biomimetic antigen presentation. It is well known that particle size affects their interaction with
cells, but less is known about particle shape or the lateral mobility of ligands on a particle surface.
Advances in material design allow for the production of ever more complex materials on a nano- and
micrometer scale, the scale at which cells operate. Extrapolation of the results from cellular
interaction with such novel materials to T cell activation, suggests that such novel materials could be
used as scaffolds for the production of aAPCs.
38
III. Samenvatting
Ons immuun system heeft zich ontwikkeld tot een geavanceerd afweermechanisme tegen infecties
van buitenaf (virussen, bacteriën), maar ook verkeerd gevouwen lichaamseigen eiwitten of
abnormaal groeiende cellen. Gespecialiseerde cellen van het immuunsysteem, dendritische cellen
(DCs), houden zich bezig met het vangen van kwaadaardige deeltjes in het lichaam. De kwaadaardige
deeltjes worden verwerkt tot kleine stukjes, ook wel antigenen genoemd, die vervolgens middels
receptoren op het oppervlak van DCs aan T cellen worden gepresenteerd. De functie van een T cel is
als een speurhond; nadat een DC een antigen aan een T cel heeft gepresenteerd, gaat de T cel op
zoek naar kwaadaardige deeltjes waar het antigen vandaan kwam en ruimt deze op.
Het lichaam past deze vorm van cellulaire immuniteit ook toe bij het opruimen van kankercellen,
alhoewel deze kankercellen zich niet altijd even makkelijk laten vinden. Dit heeft onderzoekers
gemotiveerd om het immuunsysteem een handje te helpen in het bestrijden van kanker. Er zijn twee
verschillende manieren om dit te doen; ervoor zorgen dat de DCs in het lichaam extra antigen tot
hun beschikking hebben om te presenteren aan T cellen, of T cellen buiten het lichaam opkweken
onder optimale omstandigheden en ze terugplaatsen. Voor beide methodes worden DCs gebruikt,
maar door de complexiteit van de interacties tussen DCs en T cellen levert dit niet altijd functionele T
cellen op. Om de complexiteit die DCs met zich meebrengen te omzeilen, kan er ook gebruik gemaakt
worden van kunstmatige DCs. Kunstmatige DCs bestaan vaak uit kleine (micro- of nano)deeltjes, die
de minimale hoeveel informatie bevatten om T cellen te kunnen activeren. Zulke kunstmatige DCs
zijn met succes gebruikt voor het opkweken van kanker-opruimende T cellen.
Dendritische cellen dragen drie verschillende soorten signalen over aan T cellen; activeren, overleven
en differentiëren. De signalen voor activeren en differentiëren worden direct overgedragen via het
cel oppervlak en het signaal voor differentiëren wordt indirect uitgescheiden door DCs. Via
verschillende experimenten weten we dat DCs en T cellen tijdens signaaloverdracht een synaps
vormen, en dat binnen deze synaps de signalen voor activeren en overleven zich ordenen in een
uniek patroon. Uit deze experimenten blijkt dus dat het belangrijk is dat de signalen op het oppervlak
van een (kunstmatige) DC vrij kunnen bewegen of in de juiste structuur moeten zitten.
Kunstmatige DCs bestaan in verschillende soorten. Wat ze gemeen hebben is dat ze een over
activerend en overlevingssignaal beschikken en ze daarmee, met behulp van een toegevoegd
differentiatiesignaal, T cellen kunnen activeren. De eerste kunstmatige DCs werden gemaakt van
liposomen, waarop oppervlak-gebonden signalen vrijelijk kunnen bewegen. Helaas bleken deze
deeltjes niet stabiel genoeg om gebruikt te worden als kunstmatige DCs. Een stabieler systeem voor
de productie van kunstmatige DCs is gebaseerd op rubberen bolletjes, waarbij bolletjes ter grootte
van cellen het beste waren in het activeren van T cellen. Een recenter systeem is gebaseerd op
afbreekbare polymeren, die niet alleen over activerend en overlevingssignaal beschikken, maar ook
een differentiatiesignaal kan afgeven tijdens het afbraakproces.
Het materiaal waaruit een kunstmatige DC bestaat, heeft een grote invloed op de manier waarop de
verschillende signalen gepresenteerd worden. Nieuwe ontwikkelingen in de productie van micro- en
nanodeeltjes maken het mogelijk om niet alleen naar de effecten van grootte, maar ook bijvoorbeeld
vorm en stijfheid op T cel activatie te kijken. Vanuit andere studies met cellen is al het een en ander
39
bekend over de invloed van deze “nieuwe” eigenschappen van micro- en nanodeeltjes, maar binnen
het gebied van T cel activatie met kunstmatige DCs valt nog een hoop te ontdekken.
40
IV. Bibliography
[1]
M. S. Diamond, M. Kinder, H. Matsushita, M. Mashayekhi, G. P. Dunn, J. M. Archambault, H.
Lee, C. D. Arthur, J. M. White, U. Kalinke, et al., The Journal of experimental medicine 2011,
208, 1989–2003.
[2]
M. B. Fuertes, A. K. Kacha, J. Kline, S.-R. Woo, D. M. Kranz, K. M. Murphy, T. F. Gajewski, The
Journal of experimental medicine 2011, 208, 2005–16.
[3]
C. G. Figdor, I. J. M. de Vries, W. J. Lesterhuis, C. J. M. Melief, Nature medicine 2004, 10, 475–
80.
[4]
D. H. Schuurhuis, P. Verdijk, G. Schreibelt, E. H. J. G. Aarntzen, N. Scharenborg, A. de Boer, M.
W. M. M. van de Rakt, M. Kerkhoff, M.-J. P. Gerritsen, F. Eijckeler, et al., Cancer research
2009, 69, 2927–34.
[5]
E. R. Steenblock, S. H. Wrzesinski, R. a Flavell, T. M. Fahmy, Expert opinion on biological
therapy 2009, 9, 451–64.
[6]
S. C. Balmert, S. R. Little, Advanced materials (Deerfield Beach, Fla.) 2012, 24, 3757–78.
[7]
K. Murphy, Janeway’s Immunobiology, 8th Edition, Garland Science, 2012.
[8]
T. T. Doan, R. Melvold, C. Waltenbaugh, Concise Medical Immunology, Lippincott Williams &
Wilkins, 2005.
[9]
P. Guermonprez, J. Valladeau, L. Zitvogel, C. Théry, S. Amigorena, Annual review of
immunology 2002, 20, 621–67.
[10]
K. Shortman, Y.-J. Liu, Nature reviews. Immunology 2002, 2, 151–61.
[11]
G. J. Randolph, V. Angeli, M. a Swartz, Nature reviews. Immunology 2005, 5, 617–28.
[12]
J. E. Smith-Garvin, G. a Koretzky, M. S. Jordan, Annual review of immunology 2009, 27, 591–
619.
[13]
R. M. Steinman, D. Hawiger, M. C. Nussenzweig, Annual review of immunology 2003, 21, 685–
711.
[14]
A. W. Thomson, H. R. Turnquist, A. F. Zahorchak, G. Raimondi, Transplantation 2009, 87, S86–
90.
[15]
A. Grakoui, Science 1999, 285, 221–227.
[16]
M. L. Dustin, D. Depoil, Nature reviews. Immunology 2011, 11, 672–84.
[17]
M. J. Miller, O. Safrina, I. Parker, M. D. Cahalan, The Journal of experimental medicine 2004,
200, 847–56.
41
[18]
M. J. Miller, A. S. Hejazi, S. H. Wei, M. D. Cahalan, I. Parker, Proceedings of the National
Academy of Sciences of the United States of America 2004, 101, 998–1003.
[19]
D. J. Irvine, M. a Purbhoo, M. Krogsgaard, M. M. Davis, Nature 2002, 419, 845–9.
[20]
M. a Purbhoo, D. J. Irvine, J. B. Huppa, M. M. Davis, Nature immunology 2004, 5, 524–30.
[21]
A. Chisholm, A. Zahler, J. Sisson, V. Ambros, Nature 1998, 340, 764–766.
[22]
D. R. Fooksman, S. Vardhana, G. Vasiliver-Shamis, J. Liese, D. a Blair, J. Waite, C. Sacristán, G.
D. Victora, A. Zanin-Zhorov, M. L. Dustin, Annual review of immunology 2010, 28, 79–105.
[23]
L. J. Edwards, V. I. Zarnitsyna, J. D. Hood, B. D. Evavold, C. Zhu, Frontiers in immunology 2012,
3, 86.
[24]
J. Xie, J. B. Huppa, E. W. Newell, J. Huang, P. J. R. Ebert, Q.-J. Li, M. M. Davis, Nature
immunology 2012, 13, 674–80.
[25]
P. Friedl, A. T. den Boer, M. Gunzer, Nature reviews. Immunology 2005, 5, 532–45.
[26]
B. N. Manz, J. T. Groves, Nature reviews. Molecular cell biology 2010, 11, 342–52.
[27]
Y. Kaizuka, A. D. Douglass, R. Varma, M. L. Dustin, R. D. Vale, Proceedings of the National
Academy of Sciences of the United States of America 2007, 104, 20296–301.
[28]
M. J. P. Biggs, M. C. Milone, L. C. Santos, a Gondarenko, S. J. Wind, Journal of the Royal
Society, Interface / the Royal Society 2011, 8, 1462–71.
[29]
J. Doh, D. J. Irvine, Proceedings of the National Academy of Sciences of the United States of
America 2006, 103, 5700–5.
[30]
M. L. Dustin, J. T. Groves, Annual review of biophysics 2012, 41, 543–56.
[31]
J. J. Illingworth, P. Anton van der Merwe, Immunology 2012, 135, 198–206.
[32]
K. Palucka, J. Banchereau, Nature reviews. Cancer 2012, 12, 265–77.
[33]
D. L. Porter, B. L. Levine, M. Kalos, A. Bagg, C. H. June, The New England journal of medicine
2011, 365, 725–33.
[34]
I. Mellman, G. Coukos, G. Dranoff, Nature 2011, 480, 480–9.
[35]
M. F. Mescher, J. M. Curtsinger, P. Agarwal, K. a Casey, M. Gerner, C. D. Hammerbeck, F.
Popescu, Z. Xiao, Immunological reviews 2006, 211, 81–92.
[36]
P. a Antony, C. a Piccirillo, A. Akpinarli, S. E. Finkelstein, P. J. Speiss, D. R. Surman, D. C. Palmer,
C.-C. Chan, C. a Klebanoff, W. W. Overwijk, et al., Journal of immunology (Baltimore, Md. :
1950) 2005, 174, 2591–601.
[37]
E. H. J. G. Aarntzen, I. J. M. De Vries, W. J. Lesterhuis, D. Schuurhuis, J. F. M. Jacobs, K. Bol, G.
Schreibelt, R. Mus, J. H. W. De Wilt, J. B. a G. Haanen, et al., Cancer research 2013, 73, 19–29.
42
[38]
L. Bonifaz, D. Bonnyay, K. Mahnke, M. Rivera, M. C. Nussenzweig, R. M. Steinman, Journal of
Experimental Medicine 2002, 196, 1627–1638.
[39]
Z. Zhang, S. Tongchusak, Y. Mizukami, Y. J. Kang, T. Ioji, M. Touma, B. Reinhold, D. B. Keskin, E.
L. Reinherz, T. Sasada, Biomaterials 2011, 32, 3666–78.
[40]
T. Yoshikawa, N. Okada, A. Oda, K. Matsuo, K. Matsuo, H. Kayamuro, Y. Ishii, T. Yoshinaga, T.
Akagi, M. Akashi, et al., Vaccine 2008, 26, 1303–13.
[41]
S. T. Reddy, A. J. van der Vlies, E. Simeoni, V. Angeli, G. J. Randolph, C. P. O’Neil, L. K. Lee, M. a
Swartz, J. a Hubbell, Nature biotechnology 2007, 25, 1159–64.
[42]
N.-H. Cho, T.-C. Cheong, J. H. Min, J. H. Wu, S. J. Lee, D. Kim, J.-S. Yang, S. Kim, Y. K. Kim, S.-Y.
Seong, Nature nanotechnology 2011, 6, 675–82.
[43]
S. L. Demento, A. L. Siefert, A. Bandyopadhyay, F. a Sharp, T. M. Fahmy, Trends in
biotechnology 2011, 29, 294–306.
[44]
N. P. Restifo, M. E. Dudley, S. a Rosenberg, Nature reviews. Immunology 2012, 12, 269–81.
[45]
S. a Rosenberg, J. C. Yang, R. M. Sherry, U. S. Kammula, M. S. Hughes, G. Q. Phan, D. E. Citrin,
N. P. Restifo, P. F. Robbins, J. R. Wunderlich, et al., Clinical cancer research : an official journal
of the American Association for Cancer Research 2011, 17, 4550–7.
[46]
M. L. Disis, H. Bernhard, E. M. Jaffee, Lancet 2009, 373, 673–83.
[47]
C. H. June, B. R. Blazar, J. L. Riley, Nature reviews. Immunology 2009, 9, 704–16.
[48]
L.-X. Wang, S. Shu, G. E. Plautz, Cancer research 2005, 65, 9547–54.
[49]
A. Mackensen, N. Meidenbauer, S. Vogl, M. Laumer, J. Berger, R. Andreesen, Journal of clinical
oncology : official journal of the American Society of Clinical Oncology 2006, 24, 5060–9.
[50]
C. Yee, J. a Thompson, D. Byrd, S. R. Riddell, P. Roche, E. Celis, P. D. Greenberg, Proceedings of
the National Academy of Sciences of the United States of America 2002, 99, 16168–73.
[51]
C. M. Paulos, M. M. Suhoski, G. Plesa, T. Jiang, S. Basu, T. N. Golovina, S. Jiang, N. a Aqui, D. J.
Powell, B. L. Levine, et al., Immunologic research 2008, 42, 182–96.
[52]
L. Galluzzi, E. Vacchelli, A. Eggermont, W. H. Fridman, J. Galon, C. Sautès-Fridman, E. Tartour,
L. Zitvogel, G. Kroemer, Oncoimmunology 2012, 1, 306–315.
[53]
S. Mitragotri, J. Lahann, Nature materials 2009, 8, 15–23.
[54]
J.-W. Yoo, D. J. Irvine, D. E. Discher, S. Mitragotri, Nature reviews. Drug discovery 2011, 10,
521–35.
[55]
K. D. Young, Microbiology and molecular biology reviews : MMBR 2006, 70, 660–703.
[56]
R. a Petros, J. M. DeSimone, Nature reviews. Drug discovery 2010, 9, 615–27.
43
[57]
J. a Hubbell, S. N. Thomas, M. a Swartz, Nature 2009, 462, 449–60.
[58]
M. a Swartz, S. Hirosue, J. a Hubbell, Science translational medicine 2012, 4, 148rv9.
[59]
J. J. Moon, B. Huang, D. J. Irvine, Advanced materials (Deerfield Beach, Fla.) 2012, 24, 3724–
46.
[60]
S. D. Conner, S. L. Schmid, Nature 2003, 422, 37–44.
[61]
S. E. a Gratton, P. a Ropp, P. D. Pohlhaus, J. C. Luft, V. J. Madden, M. E. Napier, J. M.
DeSimone, Proceedings of the National Academy of Sciences of the United States of America
2008, 105, 11613–8.
[62]
J. a Champion, A. Walker, S. Mitragotri, Pharmaceutical research 2008, 25, 1815–21.
[63]
S. S. Yu, C. M. Lau, S. N. Thomas, W. G. Jerome, D. J. Maron, J. H. Dickerson, J. a Hubbell, T. D.
Giorgio, International journal of nanomedicine 2012, 7, 799–813.
[64]
M. F. Mescher, Journal of immunology (Baltimore, Md. : 1950) 1992, 149, 2402–5.
[65]
E. R. Steenblock, T. M. Fahmy, Molecular therapy : the journal of the American Society of Gene
Therapy 2008, 16, 765–72.
[66]
N. Bertrand, J.-C. Leroux, Journal of controlled release : official journal of the Controlled
Release Society 2012, 161, 152–63.
[67]
D. Kohane, Biotechnology and bioengineering 2007, 96, 203–209.
[68]
T. J. Merkel, S. W. Jones, K. P. Herlihy, F. R. Kersey, A. R. Shields, M. Napier, J. C. Luft, H. Wu,
W. C. Zamboni, A. Z. Wang, et al., Proceedings of the National Academy of Sciences of the
United States of America 2011, 108, 586–91.
[69]
K. Y. Dane, C. Nembrini, A. a Tomei, J. K. Eby, C. P. O’Neil, D. Velluto, M. a Swartz, L. Inverardi,
J. a Hubbell, Journal of controlled release : official journal of the Controlled Release Society
2011, 156, 154–60.
[70]
A. Stano, C. Nembrini, M. a Swartz, J. a Hubbell, E. Simeoni, Vaccine 2012, 30, 7541–6.
[71]
A. Stano, E. a Scott, K. Y. Dane, M. a Swartz, J. a Hubbell, Biomaterials 2013, 34, 4339–46.
[72]
J. J. Moon, H. Suh, A. V Li, C. F. Ockenhouse, A. Yadava, D. J. Irvine, Proceedings of the
National Academy of Sciences of the United States of America 2012, 109, 1080–5.
[73]
J. a Champion, Y. K. Katare, S. Mitragotri, Journal of controlled release : official journal of the
Controlled Release Society 2007, 121, 3–9.
[74]
N. Daum, C. Tscheka, A. Neumeyer, M. Schneider, Wiley interdisciplinary reviews.
Nanomedicine and nanobiotechnology 2012, 4, 52–65.
[75]
G. Sharma, D. T. Valenta, Y. Altman, S. Harvey, H. Xie, S. Mitragotri, J. W. Smith, Journal of
controlled release : official journal of the Controlled Release Society 2010, 147, 408–12.
44
[76]
J. a Champion, S. Mitragotri, Pharmaceutical research 2009, 26, 244–9.
[77]
J.-W. Yoo, S. Mitragotri, Proceedings of the National Academy of Sciences 2010, 107, 11205–
11210.
[78]
J. a Champion, S. Mitragotri, Proceedings of the National Academy of Sciences of the United
States of America 2006, 103, 4930–4.
[79]
L. Florez, C. Herrmann, J. M. Cramer, C. P. Hauser, K. Koynov, K. Landfester, D. Crespy, V.
Mailänder, Small (Weinheim an der Bergstrasse, Germany) 2012, 8, 2222–30.
[80]
S. Muro, C. Garnacho, J. a Champion, J. Leferovich, C. Gajewski, E. H. Schuchman, S.
Mitragotri, V. R. Muzykantov, Molecular therapy : the journal of the American Society of Gene
Therapy 2008, 16, 1450–8.
[81]
Y. Geng, P. Dalhaimer, S. Cai, R. Tsai, M. Tewari, T. Minko, D. E. Discher, Nature
nanotechnology 2007, 2, 249–55.
[82]
S. Barua, J.-W. Yoo, P. Kolhar, A. Wakankar, Y. R. Gokarn, S. Mitragotri, Proceedings of the
National Academy of Sciences of the United States of America 2013, 110, 3270–5.
[83]
N. Doshi, A. S. Zahr, S. Bhaskar, J. Lahann, S. Mitragotri, Proceedings of the National Academy
of Sciences of the United States of America 2009, 106, 21495–9.
[84]
K. a Beningo, Y. Wang, Journal of cell science 2002, 115, 849–56.
[85]
G. Robbins, D. Lee, J. Katz, P. Frail, Soft Matter 2010, 769–779.
[86]
E. Judokusumo, E. Tabdanov, S. Kumari, M. L. Dustin, L. C. Kam, Biophysical journal 2012, 102,
L5–7.
[87]
H.-S. Han, J. D. Martin, J. Lee, D. K. Harris, D. Fukumura, R. K. Jain, M. Bawendi, Angewandte
Chemie (International ed. in English) 2013, 52, 1414–9.
[88]
M. L. Dustin, The Journal of clinical investigation 2012, 122, 1149–55.
[89]
J. B. Huppa, M. Axmann, M. a Mörtelmaier, B. F. Lillemeier, E. W. Newell, M. Brameshuber, L.
O. Klein, G. J. Schütz, M. M. Davis, Nature 2010, 463, 963–7.
[90]
L. L. Kiessling, J. E. Gestwicki, L. E. Strong, Angewandte Chemie (International ed. in English)
2006, 45, 2348–68.
[91]
A. Dolganiuc, World journal of gastroenterology : WJG 2011, 17, 2520–35.
[92]
B. M. Discher, Science 1999, 284, 1143–1146.
[93]
C. E. Ashley, E. C. Carnes, G. K. Phillips, D. Padilla, P. N. Durfee, P. A. Brown, T. N. Hanna, J. Liu,
B. Phillips, M. B. Carter, et al., Nature materials 2011, 10, 389–97.
[94]
M. Porotto, F. Yi, A. Moscona, D. a LaVan, PloS one 2011, 6, e16874.
45
[95]
D. a. Hammer, G. P. Robbins, J. B. Haun, J. J. Lin, W. Qi, L. a. Smith, P. Peter Ghoroghchian, M.
J. Therien, F. S. Bates, Faraday Discussions 2008, 139, 129.
[96]
W. R. Algar, D. E. Prasuhn, M. H. Stewart, T. L. Jennings, J. B. Blanco-Canosa, P. E. Dawson, I. L.
Medintz, Bioconjugate chemistry 2011, 22, 825–58.
[97]
M. Labowsky, T. M. Fahmy, Chemical engineering science 2012, 74, 114–123.
[98]
R. Langer, J. Folkman, Nature 1976, 263, 797–800.
[99]
G. Crotts, T. G. Park, Journal of microencapsulation 1998, 15, 699–713.
[100] T. G. Park, H. Yong Lee, Y. Sung Nam, Journal of controlled release : official journal of the
Controlled Release Society 1998, 55, 181–91.
[101] S. N. Rothstein, S. R. Little, Journal of Materials Chemistry 2011, 21, 29.
[102] L. Zhou, M. M. W. Chong, D. R. Littman, Immunity 2009, 30, 646–55.
[103] M. G. Tovey, C. Lallemand, Vaccine Adjuvants, Humana Press, Totowa, NJ, 2010.
[104] E. R. Steenblock, T. Fadel, M. Labowsky, J. S. Pober, T. M. Fahmy, The Journal of biological
chemistry 2011, 286, 34883–92.
[105] J. Park, W. Gao, R. Whiston, T. B. Strom, S. Metcalfe, T. M. Fahmy, Molecular pharmaceutics
2011, 8, 143–52.
[106] H. Nakase, K. Okazaki, Y. Tabata, M. Ozeki, N. Watanabe, M. Ohana, S. Uose, K. Uchida, T.
Nishi, M. Mastuura, et al., The Journal of pharmacology and experimental therapeutics 2002,
301, 59–65.
[107] J. W. Koten, M. J. a. Van Luyn, J. a. Cadée, L. Brouwer, W. E. Hennink, C. Bijleveld, W. Den
Otter, Cytokine 2003, 24, 57–66.
[108] J. Park, S. H. Wrzesinski, E. Stern, M. Look, J. Criscione, R. Ragheb, S. M. Jay, S. L. Demento, A.
Agawu, P. Licona Limon, et al., Nature materials 2012, 11, 895–905.
[109] S. Jhunjhunwala, S. C. Balmert, G. Raimondi, E. Dons, E. E. Nichols, A. W. Thomson, S. R. Little,
Journal of controlled release : official journal of the Controlled Release Society 2012, 159, 78–
84.
[110] U. K. Kar, M. K. Srivastava, A. Andersson, F. Baratelli, M. Huang, V. a Kickhoefer, S. M.
Dubinett, L. H. Rome, S. Sharma, PloS one 2011, 6, e18758.
[111] X. Zhao, S. Jain, H. Benjamin Larman, S. Gonzalez, D. J. Irvine, Biomaterials 2005, 26, 5048–63.
[112] S. Jhunjhunwala, G. Raimondi, A. J. Glowacki, S. J. Hall, D. Maskarinec, S. H. Thorne, A. W.
Thomson, S. R. Little, Advanced materials (Deerfield Beach, Fla.) 2012, 24, 4735–8.
[113] D. J. Irvine, A. N. Stachowiak, Y. Hori, Seminars in immunology 2008, 20, 137–46.
46
[114] Y. Wang, D. J. Irvine, Integrative biology : quantitative biosciences from nano to macro 2013,
5, 481–94.
[115] R. Gref, Y. Minamitake, M. Peracchia, V. Trubetskoy, V. Torchilin, R. Langer, Science 1994, 263,
1600–1603.
[116] K. Knop, R. Hoogenboom, D. Fischer, U. S. Schubert, Angewandte Chemie (International ed. in
English) 2010, 49, 6288–308.
[117] S. D. Perrault, C. Walkey, T. Jennings, H. C. Fischer, W. C. W. Chan, Nano letters 2009, 9, 1909–
15.
[118] C.-M. J. Hu, L. Zhang, S. Aryal, C. Cheung, R. H. Fang, L. Zhang, Proceedings of the National
Academy of Sciences of the United States of America 2011, 108, 10980–5.
[119] P. -a. Oldenborg, A. Zheleznyak, Y.-F. Fang, C. F. Lagenaur, H. D. Gresham, F. P. Lindberg,
Science 2000, 288, 2051–2054.
[120] R. K. Tsai, P. L. Rodriguez, D. E. Discher, Blood cells, molecules & diseases 2010, 45, 67–74.
[121] P. L. Rodriguez, T. Harada, D. a. Christian, D. a. Pantano, R. K. Tsai, D. E. Discher, Science 2013,
339, 971–975.
[122] J. V Kim, J.-B. Latouche, I. Rivière, M. Sadelain, Nature biotechnology 2004, 22, 403–10.
[123] C. P. Broeren, G. S. Gray, B. M. Carreno, C. H. June, Journal of immunology (Baltimore, Md. :
1950) 2000, 165, 6908–14.
[124] B. Prakken, M. Wauben, D. Genini, R. Samodal, J. Barnett, a Mendivil, L. Leoni, S. Albani,
Nature medicine 2000, 6, 1406–10.
[125] B. Kwong, S. A. Gai, J. Elkhader, K. D. Wittrup, D. J. Irvine, Cancer research 2013, 1–12.
[126] V. H. Engelhard, J. L. Strominger, M. Mescher, S. Burakoff, Proceedings of the National
Academy of Sciences of the United States of America 1978, 75, 5688–91.
[127] C. J. Turtle, S. R. Riddell, The Cancer Journal 2010, 16, 374–81.
[128] M. Mescher, J. Curtsinger, in Vaccinology: Principles and Practice, 2012, pp. 286–299.
[129] G. Gregoriadis, Trends in biotechnology 1995, 13, 527–37.
[130] V. P. Torchilin, Nature reviews. Drug discovery 2005, 4, 145–60.
[131] B. J. Prakken, R. Samodal, T. D. Le, F. Giannoni, G. P. Yung, J. Scavulli, D. Amox, S. Roord, I. de
Kleer, D. Bonnin, et al., Proceedings of the National Academy of Sciences of the United States
of America 2004, 101, 4228–33.
[132] R. Zappasodi, M. Di Nicola, C. Carlo-Stella, R. Mortarini, A. Molla, C. Vegetti, S. Albani, A.
Anichini, A. M. Gianni, Haematologica 2008, 93, 1523–34.
47
[133] Q. Ding, J. Chen, X. Wei, W. Sun, J. Mai, Y. Yang, Y. Xu, Pharmaceutical research 2013, 30, 60–
9.
[134] D. J. Irvine, Nature materials 2011, 10, 342–3.
[135] T. Jin, P. Pennefather, P. I. Lee, FEBS letters 1996, 397, 70–4.
[136] K. Shen, J. Tsai, P. Shi, L. C. Kam, Journal of the American Chemical Society 2009, 131, 13204–
5.
[137] S. a Goldstein, M. F. Mescher, Journal of immunology (Baltimore, Md. : 1950) 1986, 137,
3383–92.
[138] A. Z. Dudek, M. F. Mescher, I. Okazaki, V. T. Math, X. Luo, J. M. Curtsinger, J. S. Miller,
American journal of clinical oncology 2008, 31, 173–81.
[139] M. J. Deeths, M. F. Mescher, European journal of immunology 1997, 27, 598–608.
[140] M. J. Deeths, M. F. Mescher, European journal of immunology 1999, 29, 45–53.
[141] B. L. Musgrave, C. L. Watson, S. M. M. Haeryfar, C. a Barnes, D. W. Hoskin, Cellular
immunology 2004, 229, 1–12.
[142] K. Schilbach, G. Kerst, S. Walter, M. Eyrich, D. Wernet, R. Handgretinger, W. Xie, H.-G.
Rammensee, I. Müller, H.-J. Bühring, et al., Blood 2005, 106, 144–9.
[143] X. Jiang, X. Lu, R. Liu, F. Zhang, H. Zhao, Clinical cancer research : an official journal of the
American Association for Cancer Research 2007, 13, 7329–34.
[144] S. Caserta, P. Alessi, J. Guarnerio, V. Basso, A. Mondino, Cancer research 2008, 68, 3010–8.
[145] X. Lu, X. Jiang, R. Liu, H. Zhao, Z. Liang, Cancer letters 2008, 271, 129–39.
[146] D. Rudolf, T. Silberzahn, S. Walter, D. Maurer, J. Engelhard, D. Wernet, H.-J. Bühring, G. Jung,
B. S. Kwon, H.-G. Rammensee, et al., Cancer immunology, immunotherapy : CII 2008, 57, 175–
83.
[147] C. Shen, K. Cheng, S. Miao, W. Wang, Y. He, F. Meng, J. Zhang, Immunology letters 2013, 150,
1–11.
[148] J. M. Curtsinger, M. F. Mescher, Current opinion in immunology 2010, 22, 333–40.
[149] M. Oelke, J. P. Schneck, Immunologic research 2010, 47, 248–56.
[150] B. L. Levine, W. B. Bernstein, M. Connors, N. Craighead, T. Lindsten, C. B. Thompson, C. H.
June, Journal of immunology (Baltimore, Md. : 1950) 1997, 159, 5921–30.
[151] L. G. Lum, a V LeFever, J. S. Treisman, N. K. Garlie, J. P. Hanson, Journal of immunotherapy
(Hagerstown, Md. : 1997) 2001, 24, 408–19.
48
[152] A. K. Thomas, M. V. Maus, W. S. Shalaby, C. H. June, J. L. Riley, Clinical Immunology 2002, 105,
259–272.
[153] F. Ito, A. Carr, H. Svensson, J. Yu, A. E. Chang, Q. Li, Journal of immunotherapy (Hagerstown,
Md. : 1997) 2003, 26, 222–33.
[154] M. V Maus, J. L. Riley, W. W. Kwok, G. T. Nepom, C. H. June, Clinical Immunology 2003, 106,
16–22.
[155] M. Oelke, M. V Maus, D. Didiano, C. H. June, A. Mackensen, J. P. Schneck, Nature medicine
2003, 9, 619–24.
[156] M. Oelke, J. P. Schneck, Clinical immunology (Orlando, Fla.) 2004, 110, 243–51.
[157] H. Zhang, K. M. Snyder, M. M. Suhoski, M. V Maus, V. Kapoor, C. H. June, C. L. Mackall, Journal
of immunology (Baltimore, Md. : 1950) 2007, 179, 4910–8.
[158] C. Schütz, M. Fleck, A. Mackensen, A. Zoso, D. Halbritter, J. P. Schneck, M. Oelke, Blood 2008,
111, 3546–52.
[159] S. Ugel, A. Zoso, C. De Santo, Y. Li, I. Marigo, P. Zanovello, E. Scarselli, B. Cipriani, M. Oelke, J.
P. Schneck, et al., Cancer research 2009, 69, 9376–84.
[160] M. Durai, C. Krueger, Z. Ye, L. Cheng, A. Mackensen, M. Oelke, J. P. Schneck, Cancer
immunology, immunotherapy : CII 2009, 58, 209–20.
[161] Z. M. Ndhlovu, M. Oelke, J. P. Schneck, D. E. Griffin, Proceedings of the National Academy of
Sciences of the United States of America 2010, 107, 3669–74.
[162] S. a Fuertes Marraco, P. Baumgaertner, A. Legat, N. Rufer, D. E. Speiser, Journal of
immunological methods 2012, 385, 90–5.
[163] S. Demento, E. R. Steenblock, T. M. Fahmy, Conference proceedings : ... Annual International
Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in
Medicine and Biology Society. Conference 2009, 2009, 1161–6.
[164] H. Han, J.-R. Peng, P.-C. Chen, L. Gong, S.-S. Qiao, W.-Z. Wang, Z.-Q. Cui, X. Yu, Y.-H. Wei, X.-S.
Leng, Biochemical and biophysical research communications 2011, 411, 530–5.
[165] T. R. Fadel, N. Li, S. Shah, M. Look, L. D. Pfefferle, G. L. Haller, S. Justesen, C. J. Wilson, T. M.
Fahmy, Small (Weinheim an der Bergstrasse, Germany) 2012, 1–7.
[166] S. Mandal, Z. H. Eksteen-Akeroyd, M. J. Jacobs, R. Hammink, M. Koepf, A. J. A. Lambeck, J. C.
M. van Hest, C. J. Wilson, K. Blank, C. G. Figdor, et al., Unpublished n.d.
[167] T. Kim, H. Lee, Y. Kim, J.-M. Nam, M. Lee, Chemical communications (Cambridge, England)
2013, 49, 3949–3951.
[168] F. Giannoni, J. Barnett, K. Bi, The Journal of Immunology 2005, 174, 3204–3211.
[169] K. Murphy, Janeway’s Immunobiology, 8th Edition, Garland Science, 2012.
49
[170] E. Koffeman, E. Keogh, M. Klein, B. Prakken, S. Albani, Methods in molecular medicine 2007,
136, 69–86.
50
Download