Formulation of antigens in immunomodulating microparticles for cancer vaccination

Today cancer remains one of the main causes of death worldwide and it is predicted that the
incidence rate of men and women diagnosed with cancer in their lifespan will increase up to 40
percent. Fortunately, extensive research has dramatically reduced the mortality rate due to
improvement of the existing therapies together with the emergence of new therapeutic
strategies in the immuno-oncology field. As stated in Chapter 1, the immune system is designed
to recognize and eliminate infections through a smart, combined attack of immune cells of the
innate and adaptive immune system. Interestingly, it has been elucidated that the immune
system plays a dual role in cancer development through protection of the host against
tumorigenesis (immunosurveillance) and, in contrast, promotion of tumor growth (tumor
immune-escape).
This complex relationship is discussed in detail in Chapter 2 and unravels the immunesuppressive strategies used by cancer cells in order to evade immune recognition and
eradication. In this regard, it is clear that subtle differences in immune cell populations can
drastically change the action of the immune system from immunosurveillance to immuneescape. This points out the potential of cancer immune-therapy to shift the balance of a pro-
tumoral environment towards an unfavorable setting for cancer cells via manipulation of the
immune-suppressive cells and the tumor microenvironment.
This work aims to contribute to the development of cell-lysate based anti-cancer vaccines. The
potential advantage over conventional or other immunotherapeutic strategies are the less
severe side effects and the induction of immunological memory to provide prolonged
protection against metastasis or relapse. Although cancer vaccination is a promising
immunotherapeutic strategy in the battle against cancer and despite extensive research in this
field, it has not lived up to the expectations. There is still an unmet need for formulation
strategies that allow easy, mild and efficient incorporation of cancer antigens into
immunogenic vaccines.
In Chapter 2, the requirements for vaccines to efficiently target and activate dendritic cells in
vivo are described and the hypothesis is raised that cancer vaccines need to resemble to
pathogens in order to evoke a strong anti-tumor response. This involves formulation of cancer
antigens into particulate carriers, to enhance uptake efficiency by dendritic cells, engineered
with pathogen- and/or damage-associated molecular patterns (DAMPs and/or PAMPs), to
evoke DC maturation via stimulation of pathogen recognition receptors (PRRs). PRR-stimulation
is essential for the priming of cytotoxic CD8+ T-cells that hold the capacity to recognize and
eliminate tumor cells. In addition, another important hurdle in cancer vaccine design is the lack
of efficacy and the low applicability of many of the current approaches. This can be attributed
to the use of single or multiple tumor-associated antigens that are limited in use as antigens of
many types of cancer are still unidentified and are more prone to treatment failure due to
mutation or lack of expression. This can be circumvented with personalized cancer vaccination
that primes the patientU+2019s immune system not only against tumor-associated antigens but also
against the patientU+2019s individual tumor-specific antigens potentially leading to more potent
immune responses tailored to the patientU+2019s unique tumor mutanome. Note that personalized
cancer vaccines contain cancer antigens obtained from patient-derived tissue and requires
sufficient amount of tumor and thus is only applicable for solid tumors that can be surgically
resected. On one hand, neo-antigen vaccines involve individual tumor-specific antigens
identified via genomic analysis which is costly, labor intensive and complex. Patient-derived
cancer cells or cell lysate containing vaccines, on the other, could have an important advantage
in terms of cost and labor burden.
In this regard, in this thesis four different strategies are developed that allow for the
formulation of potent personalized cell-derived cancer vaccines as illustrated in Figure 1.
Chapter 3 and Chapter 4 deal with the formulation of soluble cancer cell lysates whereas Chapter
5 and 6 focus on the formulation of intact cancer cells into microparticles which additionally
comprise cell membrane components, and when translated to whole tumor tissue, also offer
the possibility to co-encapsulate stromal proteins.
Figure 1. Overview of the formulation strategies that are developed in this thesis, divided into two cell
lysate-based vaccines (green) and two intact cell-based vaccines (blue).
In general, extensive focus is devoted in this work to obtain simple, efficient and potent vaccine
formulation strategies of cell-derived antigens into pathogen-like particles regarding size and
immunogenicity. The latter is either obtained by introduction of TLR-agonists (PAMPs) or
immunogenic pre-treatment of the cells considering heat shock protein (DAMPs) expression.
Chapter 3, deals with the design of polymer-protein conjugates formed by disulfide exchange.
This concept was grounded on three attractive properties: [1] disulfides can be readily formed
with antigens via reaction with free thiols on cysteine residues; [2] disulfides are stable under
extracellular conditions but are reduced to free thiols in the cytoplasm of cells; and [3] disulfide
exchange. For this purpose, a co-polymer of HPMA and APMA (poly(HPMA-co-APMA)) was
synthesized that bears pending pyridyldisulfide moieties (further denoted as poly(HPMA-PDS))
followed by assessment of the conjugation efficiency of the polymer to a model antigen
ovalbumin (OVA) whether or not substituted with additional thiols. It was found that
poly(HPMA-PDS) is well suited for efficient reversible conjugation of OVA, providing that the
protein is modified with protected thiols. In vitro analysis revealed that the polymer-protein
conjugates show increased cellular uptake, relative to unconjugated protein. This is attributed
to disulfide exchange between remaining pyridyldisulfide moieties and exofacial thiols present
on the cell surface. Furthermore, the formulation demonstrated to increase antigen
presentation by bone-marrow derived DCs (bmDCs) to CD8+ T-cells in vitro.
Following characterization of this formulation with OVA, attempts to conjugate cancer cell
lysate unfortunately failed. It was not possible to introduce protected thiols to cancer cell lysate
due to aggregate formation. Nevertheless, this formulation strategy has potential as a
formulation platform for the design of vaccines containing tumor-associated antigens or neoantigens. Therefore, it would be interesting to conjugate TAAs or neo-antigens to poly(HPMAPDS) together with further optimization of the formulation regarding the amount of introduced
protected thiols. In detail, lowering the percentage of introduced protected thiols on cancer
antigens in conjunction with increasing the pending pyridyldisulfide moietes on the polymer is
of interest to obtain similar conjugation efficiencies with a minimum risk of epitope loss.
Furthermore, assessment of the effect of polymer conjugation on lymphatic antigen
transportation as well as conjugation of molecular adjuvants such as TLR-agonists to increase
immunogenicity should be performed in future research.
As this work aims to develop, however, cell-derived cancer vaccines, Chapter 4 elaborates on
an alternative strategy to encapsulate cancer cell lysate as such, without the need for
functionalization of the proteins and thus avoiding epitope loss and aggregation issues. For this
purpose, porous calcium carbonate (CaCO3) microparticles obtained by a one-step
precipitation reaction, in the presence of cancer cell lysate, were explored. This approach was
chosen based on multiple attractive properties such as its widespread use in protein
encapsulation, its high loading capacity for macromolecules, its ease of production and its low
cost. In addition, the synthesis can be performed under very mild conditions in aqueous
medium without the need of any organic solvents, reactive chemistry or high energy input.
Indeed, this approach resulted in the efficient incorporation of cancer cell lysate into nonaggregated spherically shaped CaCO3 microparticles that strongly enhanced uptake efficiency
leading to an improvement of cross-presentation by dendritic cells in vitro opposed to nonparticulate cell lysate. To increase the potency of CaCO3 microparticles as vaccine carriers,
immunogenicity was introduced via adsorption of a small molecule toll like receptor 7/8-agonist
CL264 conjugated to poly(HPMA-APMA) (further denoted as CL264-poly(HPMA-APMA)) to the
microparticle surface. This is essential in order to enable efficient priming of a robust antitumor immune response via the induction of DC maturation through TLR-activation. TLR7/8-
triggering in particular is attractive in the context of tumor vaccination as this leads to elevated
levels of type I IFN and IL-12, which are key cytokines to promote TH1- and cytotoxic T-cell
responses required for potent anti-tumor immune responses. The TLR7/8-ligand was
conjugated to the polymer backbone as it has been shown previously that lipid-, polymer- and
nanoparticle-conjugation of small molecule ligands strongly reduces systemic inflammation
and yields potent lymph node localized responses that enhance the adaptive immune response
against co-delivered antigens. In vitro activation of bmDCs and RAW blue macrophages
revealed that the polymer-conjugation of the TLR7/8-agonist did not reduce the activity of the
ligand. Moreover, the agonist was more potent when adsorbed onto the microparticle surface
which can be explained by the more efficient uptake which leads to enhanced interaction of
the TLR-ligand with the receptor upon cell uptake.
Regarding these results, formulation of cancer cell lysate into immunogenic CaCO3
microparticles shows promise as a mild and efficient strategy to encapsulate cancer lysates.
Further assessment of the potency of the vaccine can include introduction of one or more
additional TLR-agonists alongside follow-up experiments to unravel the induced cytokine
spectrum.
The encapsulation of cancer cell lysate, however, does not include cell membrane proteins and
stromal proteins, when translated to whole tumor tissue. Thererfore, this work also devoted
focus to vaccin formulations involving intact cancer cells. In Chapter 5, live cancer cells were
used as templates for layer-by-layer (LbL) assembly of complementary interacting components
followed by hypo-osmotic treatment to obtain bio-hybrid capsules loaded with cancer cell
lysate within the hollow void of the obtained capsules. The LbL assembly technique, to design
a synthetic semi-permeable membrane onto non-planar substrates, is an appealing strategy as
it allows easy, all-aqueous and mild encapsulation of a wide variety of species. Initial
experiments were performed using the oppositely charged polyelectrolytes, poly-L-arginine
(PLARG) and dextran sulfate (DEXS) based on previous work that showed multilayer capsules
composed of these polyelectrolytes are biocompatible, degradable in vitro and in vivo and
induce broad cellular and humoral immune responses against the encapsulated antigen. This
led, however, to instantaneous aggregation, cell lysis and cell death upon incubation of the
living cells with poly-L-arginine. Less harsh complementary interacting components were
therefore chosen to coat the cancer cells. This strategy was based on the use of
poly(vinylpyrrolidone) (PVP) and tannic acid (TA) that form hydrogen-bonded complexes. As
the aim of our work is to encapsulate whole cancer cells, it is important to preserve cell integrity
as much as possible while affecting cell viability as little as possible to retain a maximum amount
of cellular proteins within the LbL coating. In this regard, particular focus was devoted to
elucidate the optimal coating components and conditions. It was found that deposition of two
PVP/TA bilayers followed by hypo-osmotic lysis yielded cell-templated bio-hybrid capsules
containing a high amount of encapsulated proteins. Furthermore, it was confirmed that, upon
hypo-osmotic lysis, the cancer cells were dead which is of interest as, upon administration,
regrowth of new tumors due to residual living cells often occurs in case of whole cell-based
lysates. Further, immunogenic properties were engineered into the capsules, in a proof of
concept study, by pre-treatment of the cancer cells with heat shock to induce expression of
DAMPs, important endogenous immune-activators.
To assess the potential of the bio-hybrid cell-templated capsules, preliminary in vitro uptake
experiments were performed which revealed only 5% of the capsules to be taken up. We
hypothesized that this could be attributed to the large size (above 10 µm) of the cell-templated
capsules for efficient uptake by dendritic cells. Several attempts were subsequently made to
decrease the cell-templated capsules in size through exposure to salt or high temperature. This
resulted, however, in degradation or aggregation of the particles.
In this regard, an alternative strategy was developed in Chapter 6 to formulate intact cancer
cells into vaccine particles by a single-step method which is substantially less labor-intensive
and time-consuming and thus avoids unnecessary cell loss opposed to the layer-by-layer
coating of cancer cells. A simple, yet efficient single-step method was characterized that
encapsulates whole cancer cells in polyelectrolyte microparticles by spray drying. Porous and
non-aggregated polyelectrolyte-enrobed microparticles loaded with dead cancer cells were
obtained by admixing mannitol and live cancer cells with the oppositely charged
polyelectrolytes, DEXS and PLARG in aqueous medium prior to spray drying. Similar to the celltemplated bio-hybrid capsules (Chapter 5), the polyelectrolyte-enrobed cancer cells were dead
which could avoid tumor regrowth upon administration. The polyelectrolyte-enrobed cancer
cells, upon redispersion in PBS buffer, were stable as the microparticles did not release cell
proteins in the supernatant. In vitro evaluation revealed that the microparticles were
internalized to a much larger extent by dendritic cells and significantly enhanced antigen crosspresentation, relative to cell lysate. In analogy to the cancer cell lysate-containing CaCO3
microparticles described in Chapter 4, immune-stimulating cues were introduced by co-spray
drying of the vaccine components with CL264-poly(HPMA-APMA) yielding immunogenic
microparticles that strongly promoted TLR-activation.
These results show the potential of the polyelectrolyte-enrobed cancer cells as immunogenic
antigen carriers. Introduction of multiple TLR-ligands and subsequent assessment of the
optimal conditions for T-cell priming, as proposed for the lysate-containing CaCO3
microparticles, is thereby also of interest to further increase the potency of this formulation.
Furthermore, immunogenic treatment or induction of immunogenic cell death of the cancer
cells, prior to spray drying, to induce DAMP-production could be an additional approach to
increase the potency of DC activation and CTL-mediated anti-tumor responses.
In conclusion, regarding the cell lysate-based vaccines, the polymeric CaCO3 microparticles
appeared to be more promising opposed to the polymer-protein ligated nano-conjugates in
terms of applicability. CaCO3 microparticles allowed for efficient encapsulation of cell lysate,
without the need of functionalization, into immunogenic particles that efficiently activate
dendritic cells in vitro. The cell-based vaccines, on the other, revealed the polyelectrolyteenrobed cancer cells to be superior compared to the bio-hybrid tumor cell-templated capsules
as this method allowed for one-step formulation of cancer cells into polyelectrolyte particles
that show promising results in vitro in terms of uptake efficiency, MHC-I cross presentation
induction and immunogenicity.
For future experiments, it would be of interest to assess the potency of these two promising
vaccines in vivo, investigating the induction of anti-tumor immunity. Initial in vivo experiments
could be performed using an immunogenic murine cancer cell line that stably expresses OVA
as this enables thorough screening with readily available assays. In addition, it would be
particularly interesting to compare the formulation that appears to be the most promising in
vivo with a neo-antigen containing vaccine.
Overall, this thesis explored four different strategies to efficiently encapsulate cancer cell lysate
or cancer cells into immunogenic personalized vaccine particles. This thesis shows the
beneficial effect of antigen formulation into pathogen-like particles engineered with immunestimulating cues, in this case TLR-agonists, with the respect to antigen uptake and activation of
dendritic cells.